Hesperia 69, 2000

E$PE H THE OF JOURNAL OF THE CL A S S I C A L AMERICAN AT STUDIES VOLUME SCHOOL ATHENS 69: NUMBER JANUARY-MARC...

Author: The American School of Classical Studies at Athens; various authors

23 downloads 2196 Views 63MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

E$PE

H THE OF

JOURNAL

OF THE

CL A S S I C A L

AMERICAN AT

STUDIES

VOLUME

SCHOOL ATHENS 69:

NUMBER

JANUARY-MARCH

K

2000

A91

r1

*,~~~~~~~~~~~~~~~~~~lsia /2'-~~~~~~T >

Stuie atAten A~~~~~~~~~~20

A

Schooo

I

The American School of Classical Studies at Athens is a research and teaching institution dedicated to advanced study of the archaeology,art, history, philosophy, language, and literature of Greece and the Greek world. Established in 1881 by a consortium of nine American universities, the School now serves graduate students and scholars from more than 150 affiliated colleges and universities, acting as a base for research and study in Greece. The main buildings of the School and its library are located in Athens, with administrative and publications offices in Princeton, New Jersey. As part of its mission, the School directs ongoing excavations in the Athenian Agora and at Corinth and sponsors all other American-led excavations and surveys on Greek soil. It is the official link between American archaeologists and classicists and the Archaeological Service of the Greek Ministry of Culture and, as such, is dedicated to the wise management of cultural resources and to the dissemination of knowledge of the classical world. Inquiries about membership in the School or participation in the Summer Sessions should be sent to the American School of Classical Studies at Athens, 6-8 Charlton Street, Princeton, NewJersey 08540-5232. lJesperiais published quarterlyby the American School of Classical

Studies at Athens. Founded in 1932 and devoted primarily to the timely publication of reports on School-sponsored and School-directed projects, Hesperia welcomes submissions from all scholars working in the fields of Greek archaeology,art, epigraphy,history, and literature,from earliest prehistoric times onward. Hesperia is a refereedjournal.

VOLUME

NUMBER

69:

JANUARY-MARCH

I

2000

THE

JOURNAL

OF CLASSICAL

PUBLICATIONS STAFF

KerriCox Hesperia TraceyCullen

EDITOR,

ASSOCIATE

AsSISTANT

SuzanneAbrams PRODUCTION

L.

1

LAWALL

Graffiti, Wine Selling, and the Reuse of Amphoras in the Athenian Agora, ca. 430 to 400 B.C.

3

DAVID

R. JORDAN

A Personal Letter Found in the Athenian Agora

91

LEE

ANN

RICCARDI

ASSOCIATE

Carol Ford MARKETING

AT ATHENS

MANAGER

SarahGeorge Figueira PRODUCTION

SCHOOL

EDITOR

Michael Fitzgerald EDITORIAL

STUDIES

AMERICAN

Note to Potential Contributors

MARK

EDITOR-IN-CHIEF

OF THE

MANAGER

PatriciaTanner

PUBLICATIONS COMMITTEE Carol C. Mattusch (Chairman) George Mason University Darice Birge Loyola Universityof Chicago Kevin Clinton Cornell University Jack L. Davis Universityof Cincinnati JeniferNeils Case Western ReserveUniversity KathleenW. Slane Universityof Missouri-Columbia Stephen V. Tracy (ex officio)

The Ohio State University

Uncanonical Imperial Portraits in the Eastern Roman Provinces:The Case of the Kanellopoulos Emperor

105

Submissions: Manuscripts and communications should be addressed to Dr. TraceyCullen, Editor,Hesperia, American School of ClassicalStudies at Athens, 6-8 CharltonStreet,Princeton,NewJersey08540;tel. 609-683-0800; fax 609-924-0578; e-mail [email protected] photocopies of illustrationsmust be submittedin triplicate;originalartworkand photographs should not be sent unless prior arrangementsare made with the Editor. A short abstractsummarizingthe majorconclusionsof the articleshould also be included.Articles aresubmittedto a double-blindreviewprocessand authors arerequestedto preparetheir manuscriptsaccordingly,without their name or affiliationappearing.The style for manuscriptpreparation,notes, bibliography,and other informationon submissionscan be found in the Guidelines for Authors published in Hesperia 62, 1993, pp. i-xvi; on the School's website (www.ascsa.org);or by writing to ASCSA Publicationsat the above address. The American School of ClassicalStudies at Athens will not knowinglyprint in Hesperia or any of its other publicationsthe announcementor initial scholarlypresentationof anyobjectacquiredafterDecember30,1973, by anymeans other than through an officially sanctioned excavationor survey,unless the objectwas partof a previouslyexistingcollection or was legallyexportedfrom the countryof origin. Subscriptions:The annualsubscriptionprice,payablein advancein dollars,is $55.00 for individuals,$65.00 for institutions, and $33.00 for studentswith valid I.D. Canada and other countries add $9.50 postage. We accept VISA and MasterCard.Published quarterly.Reprintsof Hesperia 1-41, Index Volume 1 (Hesperia1-10 and Supplements 1-6), and Supplements 1-12 and Supplement 18 should be orderedfrom Swets and Zeitlinger,b.v., P.O. Box 810, 2160 SZ Lisse, Netherlands. Single issues (currentand back numbers when available)of Hesperia 42 and following areavailablefor $15.00 eachplus postagefromthe AmericanSchoolof ClassicalStudiesatAthens, 6-8 Charlton Street,Princeton,NewJersey 08540-5232 U.S.A. OrderIndex Volume2 and Supplements13 and following from UniversityMuseum Publications,Universityof PennsylvaniaMuseum of Archaeologyand Anthropology,33rd and SpruceStreets,Philadelphia,Pennsylvania19104 U.S.A.

Copyright ?C2000 The American School of ClassicalStudies at Athens Producedat EdwardsBrothers, Inc., Ann Arbor,Michigan. Design by Ellen McKie. Periodicalspostage paid at Princeton,New Jersey,and at additionalmailing offices. Postmaster:Send addresschanges to Hesperia, American School of

ClassicalStudies at Athens, 6-8 Charlton Street,Princeton, New Jersey08540-5232 U.S.A. ISSN 0018-098X ISBN 87661-500-0

HESPERIA

69,

2000

NOTETOP OT

CONTRI

ENTIAL

BUTORD

We are pleased to report that the new format of Hesperia has met with general enthusiasm. With the publication of the first volume complete, and a new yearunderway,it seems an auspicioustime to clarifythe journal's current mission and invite submission of new manuscripts. The scope of Hesperiahas been the subject of debate for many years, with the result that some confusion still exists as to what sorts of articles are considered appropriatefor the journal.When Hesperiawas founded in 1932, its purpose was closely circumscribed:to make accessible to colleagues and the public the results of research undertaken by members of the American School of Classical Studies at Athens. In 1983, however, the Managing Committee of the School passed a resolution (regulation IX.4) that opened the journal to a wider range of submissions. While still primarily a forum for School members, Hesperiaalso invites scholars not affiliated with the School to submit articles if their researchis of "particular relevance to the School's work." The work of the School is far-reaching, encompassing the history, archaeology, art, epigraphy,and literature of the Greek world, extending well outside the borders of modern Greece and from antiquity up to the present day.New fellowships at the School supportadvancedresearchacross a wide range of specializations, including anthropology, philosophy, political science, and religious studies, in addition to archaeology,art history, classics, and history.The Gennadius Library,established by the School in 1926 and currentlybeing renovated, provides an unparalleledresource for the study of Byzantine, Balkan, and Ottoman culture, as well as modern Greek language,literature,and history. Scientific studies now too find their place at the School with the opening in 1993 of the Wiener Laboratory for Archaeological Science. The range of articles currentlysubmitted to Hesperiais relativelylimited when compared with the scope of research and activities undertaken by members of the American School. The traditional strengths of Hesperia have been the presentation of field reports on School-sponsored excavations (and, more recently, surveys), Attic epigraphy and topography, and studies of Greek monuments. In particular, the journal has long given prominence to the results of work in the Athenian Agora and Corinth.

2

NOTE

TO

POTENTIAL

CONTRIBUTORS

These studies will continue to be of major importance in the journal. The researchsupportedby the School has expandedsubstantiallyover the course of the past century,however, as have the central themes of classical studies in America. It is our hope that Hesperia, as the School's official journal, will reflect the full breadth of disciplinary approaches taken today in the study of the Greek world. The focus of the journal remains centered on Greek material culture. The geographical limits are those of the entire Greek world, with no chronological restrictions imposed. Articles on all aspects of the field are sought-primary research,interdisciplinarystudies,theoreticaldiscussions, and syntheses of topics and problems in Greek art, archaeology,and epigraphy.Articles on the history and practice of archaeology and ethnography in Greece are also welcome. We may publish an occasional issue devoted to a single theme such as funerary ritual, the beginnings of literacy,or early travelersin Greece. Scholars might consider contributing an article to Hesperiaon Byzantine or Ottoman Greece, ceramicpetrography, Hellenistic Egypt, provenancestudies, Black Sea colonies, or Roman Crete, to name only a few areas of interest. Publication of a wide range of articles can only increase the vitality (and readership)of the journal. No page limit exists for contributions, although very long pieces may better appear as monographs. In order to ensure that articles published are of high quality,all submissions are refereed in a double-blind process by two outside reviewers and a member of the Publications Committee. At present, an author whose manuscript has been accepted for publication can expect to see his or her article in print in less than a year and a half from the time of submission. Fifty free offprints are provided, with an option to purchase more. Contributors are encouraged to consult Hesperia's stylistic guidelines on the School's website (www.ascsa.org),although an article need not be preparedin the journal's format to be considered for publication. We warmly invite submissions from all scholars whose researchintersects with the School's work. -

The Editors and the Committee on Publications

HESPERIA

69,

2000

Pages3-90

GRAFFITI, WINE ANLD SELLING, REUSE IN

AGO

OF

THE

THE

AMPHORAS ATHENIAN

RAP CA.

400

430

TO

B.C.

ABSTRACT Graffition transportamphorasof the 5th centuryB.C. recordvolume, weight, price, and abbreviationsthat mayreferto the jar'scontents. The graffitioften appearin the southeast corner of the Agora excavations.While some price marks may have been applied outside Athens, many graffiti resulted from retail practices in the Agora, such as refilling jars from local suppliers, decanting fromjars in the shop, and selling differentproductssuch as honey or meat. Most of the graffiti date between 430 and 400 B.C. Political and economic conditions at this time encouraged the use of graffiti on an unprecedented scale. 1. Reviews of Agora XXI are fairly representativeof the generalacceptance of the interpretationsof these graffiti: Oliver (1977) has some criticismsof the interpretationsof other classes of graffitibut writes that "Langis at her best ... when she edits the commercial and tax notations,which areworthwhile"(p. 210). Of the reviewsI have seen, onlyJohnston (1978a) takes issue with the interpretationsof the amphoragraffiti.Others,who do suggest differentreadingsfor other graffiti, do not comment on the numerical notations (e.g., Jordan1978; Oikonomides 1986 and 1988). My researchon amphorasfound in the Athenian Agora, including the graffitidiscussedhere, has been facilitated by permissionsfromT. Leslie ShearJr.andJohn McK. Camp II. The graffitidrawingsareby the author, preparedfor publicationwith the assistance of Craig Mauzy (all graffitiare illustratedat scale 1:1).The photographsareby Craig Mauzy and are

Of the many classes of evidence from antiquity,artifactswith writing should "speak for themselves" the most clearly. As the massive volume of epigraphic and philological scholarship makes very clear, however, speaking and being understood are two very different phenomena. The same gap between the evidence and our understanding holds true for graffiti on pottery. The graffiti on 5th-century transport amphoras from the Athenian Agora, the topic of this article,have been thought by many to speak clearly in comparison with more debated scratchings.1 Reexamination of this material, however, highlights difficulties with the previously accepted inreproducedhere by permissionof the American School of ClassicalStudies, Agora Excavations.The plans were preparedby RichardAnderson.This articledevelopedgraduallyout of my dissertation(Lawall 1995), a subsequent paperdeliveredat the 1997 Annual Meeting of the Archaeological Institute of America (abstractin AJA 102 [1998], pp. 401-402), and a much shorterversion initially submittedto Hesperiain December 1998. Improvements over the courseof the article's life are due to feedbackfrom various

people, but especiallyfrom the anonymousHesperiareviewersand from discussionswith CarolynKoehler, Molly Richardson,Lynn Snyder,and Malcolm Wallace.Thanks are also due to the funding agencies and institutionsthat supportedperiods of researchin Athens: the University of Michigan, the Universityof Manitoba, the Social Sciences and Humanities ResearchCouncil of Canada,the Solow Art and Architecture Foundation,and the M. Aylwin Cotton Foundation.

MARK

4

L. LAWALL

terpretations,and in doing so reveals many new aspects of daily life in the

Figure1. The Agoraca. 400 B.C. with

Athenian Agora.

findspots of 5th-century commercial

Any difficulties remaining in the interpretation of graffiti from the Agora are not for lack of scholarlyinterest. Ever since LucyTalcott's publication of "Attic Black-glazed Stamped Ware and Other Pottery from a Fifth Century Well" in 1935, graffiti on pottery at the Agora have received frequent attention.2 Mabel Lang stands out in this regard, having published two majorstudies,"NumericalNotation on Greek Vases"in 1956 and Graffiti and Dipinti in 1976, and a volume in the Agora PictureBook

graffition amphoras

series.3

Lang's publications offered readings and interpretations of hundreds of markings, including many numerical marks that are likely to be related to buying and selling in the Agora. These marks often appearon transport amphoras, and Lang suggested that some such graffiti resulted from recording the measured capacity of the jars for their reuse.4Lang read other, far less frequent numerical marks as price labels; rarerstill for the 5th century are notations of weight. Virginia Grace and Malcolm Wallace both connected the appearanceof the many volumetric markings to uncertainties arising from a change in the standardChian amphora capacity,from 7 to 8 Chian choes, to become commensuratewith 7 Attic choes. Grace saw this change as being in accordancewith the Athenian StandardsDecree.5 This mid- to late-5th-century decree dictated the use of Athenian standard measures (including standards for coinage) among all members of the Delian League.6 These earlier studies thus associated the graffiti not only with specific actions related to the sale of the amphorasbut also with a specific historical circumstance. This picture, however, was developed with minimal attention to the contexts of the graffiti:which amphora types carrywhich sorts of graffiti? What are the dates of the markedjars?Where are the graffiti found in the Agora? Inquiry along these lines revealsvery clear patterns. First, not only Chian jars but many amphora types carrygraffiti. Second, although such graffiti do appearsporadicallythroughout the 5th century B.C., most numerical markings on amphoras appearin the last third of the century.Finally, while examples have been found across the Agora, most amphora graffiti were excavated in the southeast corner (Figs. 1-2). These patterns are difficult to reconcile with either the postulated use of graffiti in checking the capacity of jars for personal reuse or the proposed implications of the Standards Decree. If graffiti were applied when confirming amphora capacities, then marked amphoras should be distributed across the Agora, wherever amphoras were found in large numbers. This is not the case. Moreover, the connection to the Standards Decree 2. Talcott1935,pp.495-496,515516. 3. Lang 1956; AgoraXXI; Lang 1974. Other comments on Agora graffitiincludeTalcott 1936, pp. 344 and 352; Boulter 1953, pp. 99-101; and Rotroff and Oakley 1992, especiallypp. 27-28, 35-37, figs. 2122. 4. Lang 1956, pp. 23-24.

5. GraceandSawvvatianouPetropoulakou1970, pp.359-360; Grace 1979b, pp. 121-122. Wallace (1984, pp. 12-13; 1986, p. 88) accepts the change in Chian capacityand its relationshipto the graffiti,but expressesuncertaintyas to the precise role of the StandardsDecree. 6. For discussionsof the date of this decree,see, among many others,

Lewis1987,Mattingly1993and 1999; for the intentions and impacts of the decree see Finley 1973, pp. 168169; Mattingly 1981; Schonhammer 1993, p. 190; and Figueira 1998, whose discussionof the goals of the Athenians emphasizes,even more so than that of Lewis (1987), the range of dates of copies of the decree.

GRAFFITI,

WINE

AND

SELLING,

REUSE

OF

AMPHORAS

R

ICAVATI ATEIENS SHOWING DISTRIBUTION OF

TTURYB .C. AMPHORAS WITH C

I-?...

...j

.

...... . ,,..!...........................

.~ , ???????...........

ction T

9/Ar

\

X

./-^

019:4 ..*- .-..."...--."; '".

Section NN/110-111fKE-KH Fill/. f./

AREOPAGOS

-** -:::...::' ............. 1

m .

i-t r-n X-

I

X \

;

.

5

PP - PP'

(PART OF) SECTION --

SURVIVING CLASSICAL WALLS RESTORED CLASSICAL WALLS

E

SURVIVING WALLS WALLS m RESTORED CONJECTURAL WALLS CONJECTURAL WALLS

1

OFRECEDING BULIBRARY LIBRARY O PANTAINO

Figure2. Restoredplanof the kapeleionareaaroundwell R 13:4,with otherlate-5th- and early-4th-centurywells

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

7

and Chian amphoras raised the expectation of a limited period of uncertainty and, therefore,of intense markingjust after the initial promulgation of the decree (perhaps ca. 449 B.C., perhaps ca. 425 B.C.). Instead, roughly thirty-five years are indicated by the range of dates for the amphoras carrying the marks.7Furthermore,a connection to the decree might prompt the expectation that the markswould appearprimarily on those amphora types whose capacities changed in accordancewith the decree. It is still a matter of debate as to which, if any,types were modified,8 but the fact that volumetric graffiti are not limited to specific types casts doubt on this element of the interpretation as well. Finally, while some commentators argue that the decree was meant to facilitate collection of taxes, tribute, and other payments, whether in money or other products,9the graffiti indicate instead some sort of difficulty arising in later-5th-century sales in Athens. If such contexts of the graffiti requireus to question, or even abandon, earlierexplanations, new interpretations are needed. The following reconsideration of the 5th-century amphora graffiti from the Agora excavations begins with a detailed surveyof the markingsthemselves, reviewing Lang's readings and taking into account other examples.10In this review I emphasize the wide range of topics addressed by the graffiti and consider problems of interpretation not discussed by Lang. After types of markings are described and the findspots, dates, and amphoras carrying the graffiti are presented, first in the catalogue and then synthesized in the text, it will be possible to consider what activities and transactions resulted in the application of the graffiti. The third and final step to be taken here is to explain why graffiti appear so frequently on amphoras datable to the later decades of the 5th century B.C. in Athens. 7. If Figueira(1998) is correctto

of emphasizemultiplepromulgations the decree,thenperhapsanongoing senseof uncertainty concerningstandardsshouldbe expected.Grace,however,saw the graffitias connected to a single decreeca. 449 B.C. (Grace and 1970, Savvatianou-Petropoulakou pp. 359-360; Grace 1979b, pp. 121122). Wallace (1986, p. 88, note 8) places the decree"somewherebetween 449 and 414 B.C."and notes the possibility that the change in capacitywas not relatedto the decree.It should be noted that Grace'sdate for the new amphoratype with the higher bulge on the neck and the proposedlarger capacitywas based entirelyon the date of ca. 449 for the decree.The jars themselvesdo not appearin Agora contexts or elsewherebefore deposits closed after440. In terms of the archaeologicalevidence,ca. 440 and not ca. 450 is the apparentdate for the introductionof the high-bulge Chian IX, pp. 23jars (type C/3 in Kerameikos 24; and see Lawall 1995, pp. 99-102).

Without acceptingthe date of 449 for the decree,following Grace,there is no evidencefor the high-bulge Chians startingca. 449. 8. Published argumentsconcerning these changes arebased on a very limited numberof examples,far too few to make a case one way or the other (e.g., Grace and Savvatianou-Petropoulakou [1970, p. 360, note 4] cite only two capacitiesof Chianjars of 7 Chian choes dating before 450 B.C.). On the difficultiesof measuringamphorassee WallaceMatheson and Wallace 1982, pp. 302-320; Wallace 1984, p. 13; and Koehlerand Wallace 1987, p. 57. In connectionwith Wallace's(1984, p. 13) comment that "nobody,so far as I know,has doubted that [Chios] changed her wine exportcapacityfrom 7 to 8 Chian choes,"see Brashinskiy 1984, pp. 98-99, suggestingno significant change in the Chian standard amphoracapacity. 9. See Martin 1985, pp. 196-207. Martin opposes Finley'sview of minting as a symbol of a state'sautonomy,

arguinginstead for "economic"motives: the decree,in his view, facilitatedcollection of tributeand port taxes.Finley, while explicitlyacknowledgingthe decree'sfacilitationof tax and tribute collection, called that a "politicalelement"(1973, p. 169). Finley'smain concernwas to deny any Athenian policy to benefit Athenian merchantsat others'expense.Martin implies that merchantswould benefit from the unification of standards(p. 199, note 7, and p. 204), but never addresses Finley'scontention that all Aegean merchants-not only Athenianswould find such a situationeasier. 10. Johnston (1978a, p. 218) criticizedthe incompletenessof Agora XXI, noting that much of the Agora graffitiknown to him had not been included.I presenthere all numerical amphoragraffitiof the 5th century known to me. While there are more graffitipublishedhere than had been publishedearlier,there are still bound to be furtherpieces that I have not found.

8

MARK

L. LAWALL

THE GRAFFITI The graffiti in question were applied to jars before they were broken and not to discrete sherds.l1Some markedjars are still nearly complete; markings preserved only on fragments never fit neatly on the sherd, in contrast to the case with ostraka.With one exception, 67, the marks were incised after the jars were fired, so the graffiti were applied at some time during the jar'speriod of use. Most of these graffiti involve a series of symbols, sometimes letters, often repeating. Since vertical strokes are common and since some letters are best read as TTfor "5"and A for "10,"many graffiti appear to be numerical notations. These notations include tallies of vertical and/or horizontal strokes often with acrophonic numerals, tallies using letters to abbreviate units of measure,and price marks involving monetary symbols or abbreviations.If we assume that these graffiti refer to the vessels on which they are found and often to their contents, the marksshould relate in some way to the production, filling, distribution, and even refilling of the amphoras. Hence, these marks are safely considered economic graffiti-at least if one includes in a definition of "economic"the production and distribution of goods and services.12 There is, too, a second broad class of graffiti that is of less certain economic relevance: short abbreviations commonly found on amphoras. The letters could simply abbreviatenames: perhaps an owner of the amphora, perhaps a merchant, and so on. On the other hand, the repeated occurrence of some of these marks across many different sites and amphora types suggests that the letters refer to some activity common to amphora use. There are, of course, other kinds of graffiti on amphoras. These include names, other abbreviations, or simply letters of uncertain significance. Since these marks are not repeated over many amphoras, their significance for the production and distribution of the jars may have been limited to naming an owner during the "lifetime"of the jar. Such marks are included here only when they appearin the same context as examples of economic graffiti. They are not considered in detail nor were examples from other parts of the Agora studied.l3 11. Price labels or accounting ostrakadiffer distinctlyfrom the graffiti discussedhere.The graffiti on such ostrakaoften follow one or more edges of the sherd.Even if the sherd has brokenacrossthe original graffiti,it is often clearthat the graffiti on these ostrakaoriginallyfollowed the lines of a sherd.They were not appliedto a complete vessel. Amphora fragmentsused for accounting ostrakaare not coveredin this article. 12. Some have arguedthat this

definition is too narrowand that it excludesthe element of risk and choices in allocationat the core of economic study (e.g., Burling 1962). Productionand distribution,however, aretwo very visible behaviorsin the archaeologicalrecordand may have involvedthe sorts of allocation issues attendedby more traditional economic historiansand anthropologists. On problemsof defining economic studies,see Lowry 1987, pp. 8-9. 13. I collected the graffitipresented

in this articlefrom a handwrittenand largelyup-to-date list of all inventoried graffitiin the Agora, and by searching for uninventoriedgraffitifrom closed deposits and stratifiedfills throughoutthe Agora as part of the researchfor both my dissertation and, more recently,a typological study of Classicaland Hellenistic amphorasin the Agora. I thank John Camp for alertingme to graffiti finds in the building fill of the late-5thcenturyMint.

GRAFFITI,

EARLIER

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

9

READINGS

Lang assigned most of the numerical graffiti involving repeated symbols to two categories, capacity and price. She interpreted the symbols as follows:

H-

14.This sizeforanAtticchousis usedanddiscussedin Graceand 1970, Savvatianou-Petropoulakou p. 360. 15. Larsen(1938, pp. 394-395) associatesthe term xsp6Lc0ov in papyri with a unit of 8 choes, but he emphasizes the late date of common usage of this term. For furtherdiscussionof the size of Classicalamphoras,see, among many others,Pritchett 1956, pp. 195196; Grace 1949, p. 175; Brashinskiy 1984, pp. 92-127. For readingsthat would be improvedby applyingLang's method, see Grakov1968, pp. 104-105. 16. Lang 1956, pp. 13-16. For the Attic Stelai and prices of wine and amphorasfrom other sources,see Pritchett 1956, pp.195-196,199-203; and Amyx 1958, pp. 174-178. FurtheYon wine prices:Reger 1994, pp. 234-235; andJohnston 1996, p. 87. 17. AgoraXXI, pp. 12-13, C 8, C 9. 18. Lang 1956, p. 9, no. 29. 19. For 25 the link to Chios is suggestedin AgoraXXI, p. 76, He 2; for 13, the Chian attributionis made in Lang 1956, p. 9, no. 30. 20. See Lang 1956, p. 5, no. 10. The theory that amphorasin the 5th centurywere potted to fit certaincities' standardsmore preciselythan others remainsan open question awaiting more thoroughpublicationof amphora capacitiesand carefulconsiderationof the implicit assumptionsabout the natureof commercerequiredby the theory (cf. Lawall 1995, pp. 297-300 and Garlan 1985, p. 243). 21. Lang 1956, p. 2; cf. Johnston (1978a and 1996, passim),emphasizing such uncertainties.

K 0 X H rT TTHmonogram TTXmonogram A AX monogram 2

1 chous, kotyle, or drachma 1 drachma 1 kotyle or 1 obol 1 kotyle (1/12chous) 1 oxybaphon (1/4 kotyle)

1 chous chous, 1 hydria, or alphabetic "8" 5 choes or 5 drachmas 5 hydriae 5 choes 10 choes, 10 drachmas, or 10 staters 10 choes stater 1/2

Lang was thereby able to read graffiti as indicating that certain amphora types held between 7 and 8 Attic choes (defined as 3.2 liters/chous),14 figures matched by measurements of complete jars.15The price graffiti, according to Lang, involve either fairly clear monetary symbols (the drachma symbol or the 2 for stater) or a series of less explicit tally marks exceeding the likely capacity of the given jar. Lang noted that a price of 2 drachmas per chous of wine, as she read in some of the price graffiti, fit well with other epigraphic and literary evidence for 5th-century wine prices.16 Alongside these general practices, Lang highlighted certain idiosyncrasies. She interpreted the letter E as abbreviatingeither heUL- or hsis (4, 13, 25, 64), the aspirate omitted. In her discussion of 4, in which she read an E for hesit-, Lang noted that E also replaced the hs- in the name Hegestratos on two graffiti from the Agora.17In the case of 64, where the h is dropped from E abbreviating hEiXouvV,Lang writes: "The writer, being psilotic or an h-dropper, was probably not Athenian."18For 13 and 25, Lang attributed this psilosis more directly to Chios.19Lang identified another example of "foreign"graffiti in the possible indication of the nonAttic monetary unit, staters, in two examples (19 and 26). Finally, Lang suggested that certain tallies of measurement were based on a Chian, not an Athenian, unit of measure (22).20 Lang acknowledgedthat there was "anelement of uncertainty"in some readings,21and this is not surprisinggiven the informal natureof the markings. To what extent, however, are other readings of the graffiti possible? Are subjects other than volume and price indicated with any frequency by the graffiti?While many of the readings require detailed, individual discussion (see catalogue),some comments pertain more generallyto the types of numerical graffiti.These are discussed in the following order:volumetric notations, notations of weight (?), price marks, abbreviations,summations, and combined notations.

MARK

IO

VOLUMETRIC

L.

LAWALL

NOTATIONS

(I, 7?, 9, 10, II, I2, 13, 25, 26, 27, 28, 29?, 30?, 33, 34, 35, 36, 37?, 38, 39, 40, 41, 42, 45?, 46, 48, 49, 50, 51, 52, 53,

63, 64, 65, 67, 68, 70, 7I?, 72, 73, 93?, 94, 95, 96, 97)22 With forty-five entries in this catalogue, volumetric notations comprise the largest group among the numerical graffiti. The markings here range from simple strokes and the acrophonic numerals A and TTto more explicit notations using abbreviations for the units of measure. Following Lang's analysis, the acrophonic numerals and the vertical tallies may be read as whole choes; horizontal strokes may indicate kotylai.23As noted above, Lang found that graffiti read in this way matched amphora capacity measurements quite well. Even markings indicating just over 10 choes fall within the range of possible capacities for 5th-century amphoras.Amphoras of such a large size are rare,however, as are graffiti that necessarily indicate more than 8 choes. When simple tallies exceed indications of 10 units, or even 8, they are unlikely to be volumetric notations.24In some such cases, Lang suggested that the notations might be better read as price marks,and for a very few examplesshe proposed the identification as weight notations; both possibilities will be considered below. Far less ambiguous are the many graffiti read by Lang as including abbreviationsof chous (X), hemichous (H or E), and kotyle (K).25In many cases, this sequence of abbreviationsresults in quite reasonable and likely readings. These readings, like the simpler tallies described above, rarely give values beyond 8 to 10 choes. In one detail, however, Lang's interpretation of these graffiti should be corrected. She suggested that the use of E for hVLutor as an abbreviation for hesl indicated an h-dropping speaker as the writer. Such a writer, according to Lang, would not have been Attic but probably Chian (where the dialect did not use the aspirate).26A Chian, however, would presumably spell hStL- starting with an Ionian eta, utu-;and Athenians sometimes dropped the aspirate in words beginning with epsilon.27 The E/H marks in numerical graffiti present still further problems of interpretation.When the H or E follows whole-unit symbols (e.g., X for chous), only one H or E is usually present, and this single letter often precedes a series of even smaller unit symbols. In such cases, the E/H is read 22. Numbersfollowed by question marksin these lists indicate that the classificationof the graffitoas to reference (e.g., volume, weight) is uncertain. 23. Similarly,horizontallines can indicate obols on monetaryinscriptions. Tod (1911-1912, p. 101) notes that a verticalline is most common for obol in Attic inscriptions,but elsewhere he lists numerousexamplesof the horizontal line for obol (1911-1912, passim); see, too,Johnston 1979, fig. 12:c, 14F, 15(E). 24. Brashinskiy(1984, pp. 170-204)

lists hundredsof capacitymeasuresfor Archaicthrough Hellenistic amphora types. Of these only sixteen exceed 30 liters. 25. The readingsof H or Efor halfmetretesand Hfor hydriaare discussed furtherbelow.Although the Greek term for half a chous is tro ujll)(xoov, I use the term "hemichous"here as being a more straightforward,English halving of chous. 26. See summaryof Lang'sargument above,p. 9. 27. Collitz and Bechtel 1905, p. 703, and line no. 5653b, line 5 (j'?p-YLCaLv)

16 (jEpjv); p. 712, no. 5664, line 8 (aX?pi-Covyiuocu). I owe this obser-

vation to the comments of one of Hesperia'sanonymousreviewers,who very generouslypointed out this difficulty in Lang'sinterpretation.Lang, indeed, mentions other Agora graffitiwhere the aspirateis droppedfrom names (see note 17, above).The aspirateis left out in variousother stone, ceramic,and numismaticinscriptions;for examples see Guarducci1970, p. 694; Patitucci 1991, p. 249, no. 5; Blond6 1989, p. 518, no. 170; Buck 1955, pp. 156, 159; and Tod 1911-1912, p. 118.

GRAFFITI,

28. Apart from the amphoragraffiti discussedhere, Lang (1956, p. 5, no. 13) also publishesa cook-pot fragmentwith E followed by four verticalstrokes.The E could indicate the fifth unit of measureand the total number,five units.Johnston (1978a, p. 218) expressessome concernover such earlyusage of alphabeticnumerals. In these summations,however-if they arebeing correctlyinterpretedas summations-the numericaluse of the letter seems likely. 29. Blonde 1989, p. 518, no. 170; Johnston 1996, p. 82, note 5 refersto a furtherexamplefrom Olbia. 30. Tod 1911-1912, passim. 31. Lang 1956, p. 11.

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

II

as half of the preceding unit. In other cases the meaning of E/H is more ambiguous. In 25, EE follows a 1T,and the Es are read as acrophonic numerals abbreviating hsi; (with the h dropped). Along the same lines, the EEEE graffito, 29, might be read as four of some unit, but that unit is uncertain (perhaps choes, perhaps mnas; see below). E might also serve as an alphabetic numeral, 5, and H as 8. The alphabetic use of H is most securely read in summation notations (see below).28When E/H stands as the first symbol in a numericallist, the letter may indicate a half-metretes (6 choes) (13, though the interpretation is uncertain; 24; 70). This seems especially likely when the E/H and subsequent marks allow a reading of something less than 8 choes (but see discussion of solitary E/H marks below). E/H, therefore, might stand for various half-units, alphabetic numerals, or-very rarely,and only with E-as an acrophonic numeral for iFC;. Multiple instances of E/H rarelyappearon amphora graffiti;there are three among the graffiti considered here (29, 6, and 65). The first, 29, with EEEE (with the possibility of more letters on either side of the preserved fragment), is from the Mint construction fill. If, from a practical standpoint, it seems unlikely that multiple half-units (e.g., hemichoes) would appear on numerical notation, perhaps this graffito should be read as 4 (sc. choes). This reading, however, depends on the rare acrophonic use of E. If E is read as an alphabetic numeral, then we might have 20 units recorded here, a large volumetric notation but possibly a weight notation (see below). Finally, reading E as an acrophonic numeral for 100, as paralleled on numerical ostraka,29would create an unusually large number for reference to the amphora.The fragment does not seem to be an ostrakon since the writing does not follow any particular edge of the sherd. On balance, the readings either of E for etS, or of E for hesitXouv,despite the rarity of the former and the apparent impracticality of the latter, seem most likely. A second multiple E/H graffito in this collection, 6, may well be an ostrakon;the marks do fit neatly across the sherd. A reading of 300 for the HHH,however,would be difficult to reconcilewith the subsequent T, which is most often used to express a fraction or a talent,30the latter being most unlikely here. If this graffito was applied to, and therefore referred to, a complete vessel, the Hs might stand for YtXooc, followed by the T for onefourth (or one-third), and then a series of 4 kotylai. Alternatively, the T might summarize the four following strokes (see below, "Summation Marks").Such decreasing sizes of units occur throughout these numerical graffiti. The use of multiple half units, however, might seem odd when more than one whole unit was used (why not XHT..., for example?). A further alternative,to read the H for sf;, is unparalleled in this collection. The third and last multiple H graffito, 65, bears a much closer resemblance to the other volumetric notations in terms of syntax or arrangement of the letters. Instead of X being used in a monogram with TTto indicate 5 units, H is used, and a further H precedes two Ks. Lang rightly rejects the possibility that the TTHmonogram indicates 500, and she rejects the possibility here that H would stand for hemichoes, doubting especially "thatfive hemichoes would be used as a unit at all."31Whether the

MARK L. LAWALL

12

presence of the TTHmonogram requiresthere to have been a 5-hemichous unit in use seems to be an open question. Perhaps5 hemichoes were poured in or out before the writer felt the need to note down smaller amounts. Lang ends her discussion of this piece with the suggestion that the H stands for hydria.32This unit of measure is known in two contexts: Epiphanius's treatise on weights and measures, from which it may be equated to 7.29 liters; and the lexicographerHesychius, who defines the term as half of an Attic metretes (s.v. o8Spocia). In both cases, if 65 is read as a tally of hydriae,the total far exceeds the likely capacity of the amphora (more than 44 liters by Epiphanius'sdefinition; more than 115 (!) liters by Hesychius's definition).33As with the previous two multiple E/H graffiti, the difficulties here with interpretations other than hemichous encourage the acceptance of the use of multiple half-unit measures. These particularexamples notwithstanding, the volumetric notations tend to be the most straightforward of the numerical graffiti. The data from measured amphora capacities clearly define the most likely upper range for such markings, and many of the graffiti, especially those using the abbreviations X and K, indicate figures well within this range. Once tallies exceed the number 8, however, the possibility of a nonvolumetric notation must be considered. NOTATIONS

OF WEIGHT?

(3, 5, 7?, I4, I5, i6, I7, I8, 29?, 43, 45?, 74, 79?, 80o, 98) Graffiti and dipinti recording the empty and gross weights of amphoras from the Hellenistic period through Late Antiquity are published from the Agora and elsewhere.34For the 5th century B.C.,however, such readings of the graffiti or any explicit connection between clay vessels and mnas are very rare.35Lang published one example: a small table amphora whose graffito reads (AM ligature, MA ligature, M M) "'AM(popEcoq) M(voc) A(?cX0)M(voc) M(voc)"and (MA ligature MA ligature) "M(voa) A(exoc) M(vaoc)A(ixCx)."36 Lang proposed that the first marks refer to the weight of the empty jar (tare weight) of 12 mnas (approximately5.5 kg) and that the second group refers to the net weight of the jar'scontents, 20 mnas (just over 9 kg),37the equivalent of roughly 9 liters of wine, just under 3 choes. Given the postulatedweight of the emptyjar,and the weights of intact amphoras of various sizes (see below), this jar is likely to have held more than 3 choes, and this weight notation should record a partial 32. Lang'ssuggestion (1956, p. 11) that the graffitowas written by a slave misspellingthe initial sound of chous certainlyseems possiblebut would requirefurthersupportof parallel examplesof misspellings. 33. See Hultsch 1882, p. 574 for interpretationof Epiphanius. 34. Lang 1956, pp. 17-18;Agora XXI, pp. 64-72 and pp. 77-81, passim.

35. Tolstoi (1953, p. 97) published a lagynos graffitoof the 2nd or

1st centuryB.C. from Pantikapaion that reads: jvocSocpyuopou 8Exoc ('A)x-tolou Msvse&opoS with AY2

written on a line abovethe main graffito abovethe end of the name There areproblemsinter'AxTocuou. preting the AYE, but the present interestis the possibilitythat this clay vessel was weighed as 10 silvermnas. By the 2nd centuryB.C. in Attica this would be the equivalentof ca. 6.5 kg. This figuremight indicate the weight of the vessel and its contents (the

empty vessel alone would weigh much less). It is also possible that the lagynos actuallycontained 10 mnas of silver. 36. AgoraXXI, p. 76, He 3 (= P 23948). 37. The conversionfactorfor much of the 5th centuryB.C. is 1 emporic mna to 105 coin drachmasor 0.457 kg/mna; see AgoraX, pp. 4, 19, and 20. For the definition of an emporic mna as opposed to a monetarymna, see AgoraX, pp. 2-4.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

I3

filling of the jar. Another very similar example (3) appears on a transport amphora of the mid-5th century B.C.:AM ligature with MA written out, which may be read as indicating an empty amphora of 10 mnas. Despite the rarity of graffiti referring explicitly to mnas in the 5th century B.C., these examples raise the possibility that tallies indicating large numbers are measures of weight. For example, Lang read a 3rd-century B.C. amphora fragment with AAI11111 as 26 mnas, arguing that the number is too large to indicate the volume of the jar.38She used the same argument to read four other Late Classical or Hellenistic graffiti as weight measurements but suggested such a reading for only one of the larger 5th-century tallies: a Chian jar, 14, with a graffito AA as indicating a tare weight of 20 mnas.39Might other 5th-century graffiti pertain to weight as well? Weights of well-preserved jars may suggest a possible range of tare weights for late-5th-century amphoras in emporic mnas. Weights of intact amphoras are rarelyreported since sufficiently well preservedjars are not very common. Published weights (converted to 5th-century emporic mnas of 105 drachmas,0.457 kg/mna) range from 6.7 mnas to 37.2 mnas.40 Other weighed amphoraslisted by Lang inAgoraXXI are of much smaller, Late Roman types. David Peacock and Dyfri Williams list a series of weights for major earlier Roman amphora types (note that all of these later forms are much larger with thicker walls than their Classical counterparts).41I review the averageshere: Dr 1B, 25 kg (54.7 mnas); Dr 2-4, 15 kg (32.8 mnas); Haltern 70, 18 kg (39.4 mnas); Dr 20,28.42 kg (62.1 mnas); Africana Grande, 17.83 kg (39 mnas); and Tripolitanian, 15.86 kg (34.8 mnas). I weighed four intact 5th- and 4th-century B.C.Greek amphoras using a simple spring balance to determine a range of possible empty weights for the amphoras carrying the graffiti (none of which are intact themselves). An early-5th-century jar (P 23750), likely to be from northern Greece, weighs ca. 5 kg (11.0 mnas); a late-5th-century jar (P 30685), perhapsfromThasos or the surroundingarea,weighs ca. 6.5 kg (14.2 mnas); an unidentified jar (P 27420) of similar size as the late-5th-century Chian type weighs ca. 7.5 kg (16.4 mnas);42and a largerjar,possibly Mendean, of the mid-4th century (SS 14826), weighs ca. 10.5 kg (23.0 mnas). This small sampling of Classical amphora weights, suggesting a general range of 5 to 10 kg or 10 to 20 mnas, fits well with the range of weights of 38. AgoraXXI, p. 66, Hb 1 (= Lang 1956, p. 17, no. 73). Here Lang uses an emporicmna standardof 0.654 kg/mna (150 coin drachmas),but this standardwas legislatedin the late 2nd centuryB.C. as replacinga standardof 0.60168 kg/mna (138 coin drachmas), which itself must have replaced the earlier 0.457 kg/mna (105 coin drachmas); see Agora X, pp. 19-20 for the history of fluctuations in this standard. 39. Agora XXI, p. 66, Hb 2, Hb 3; Lang 1956, p. 18, nos. 75 and 76 are the other 4th and 3rd-century weight notations. Lang 1956, p. 17, no. 72 =

14 (note that Lang's conversion of these 20 mnas to kilograms here is based on a coin mna of 100 drachmas [0.436 kg/mna]; the conversion should be based on an emporic mna of at least 105 drachmas). 40. Johnston and Jones 1978, p. 104, Attic SOS amphora, 17 kg (37.2 mnas); Bertucchi 1992, p. 102, type 6 Augustan amphora, 11.6 kg (25.4 mnas);

p. 114, type 7,1st centuryA.c., 11.5 kg (25.2 mnas); p. 128, Graeco-Italic, early 2nd century B.C.,5.9 kg (12.6 mnas); p. 135, Dr. 7-11 form, 10.1 kg (22.1 mnas); Agora XXI, p. 66, Hb 5,

P 3467, Early Roman amphora, 8.150

kg with dipinto indicating 8.194 kg (17.8 mnas);p. 66, Hb 9, P 26602, RobinsonM 238 type amphora,4th centuryA.C.,3.065 kg (6.7 mnas); pp. 66-67, Hb 10, P 9881, Robinson M 232 type amphora,4th centuryA.C., 5.030 kg (11 mnas). 41. Peacockand Williams 1986, p. 52, table 1. 42. Lang (1956, p. 17, no. 72) cites a weight of 8.640 kg for a Chian jar,but the jar is restoredwith plaster,which may have added considerablyto its weight.

I4

MARK

L. LAWALL

amphoras of comparable size from other periods and regions. With wine or oil added, a large Classical amphora could reach or exceed 50 mnas (over 20 liters). A full jar could weigh over 70 mnas. The largernumericalgraffitimight, therefore,referto aspectsof weight. Figures indicating between 10 and 20 mnas, for example, could record the empty weight of the jar. An early graffito, 5, shows a A followed by two rows of four vertical strokes each. Lang's reading of 10 choes, 8 kotylai as the capacity of the jar is possible, but unusually large.43A slightly laterjar, If the similarity between these two 15, shows a similar graffito, A 1111111. a of indicates graffiti similarity meaning, both the earlier and the later record could tare weights (18 mnas for the earlier,17 mnas for the graffito later). On the other hand, if the empty weight was apparent simply from observing the balance,then the tallied figures might recordeither the number of weights added to determine the net weight of the jar's contents or the total number of balancing weights present once the liquid was added (gross weight ofjar and contents). Graffiti clearly indicating amounts over 30 units (74 and 81) may indicate either net or gross weight. Other markings less clearly indicate such large numbers. Some graffiti are composed of stacked, short, horizontal marks along the first or final of a series of vertical strokes (16, 18, and possibly 45).44 By comparison with graffiti in which Ks (abbreviatingkotylai) are stacked on a single vertical stroke (35, 36), the horizontal strokes could be small units or fractions in relation to the vertical strokes. Given that 10-mna units have been postulated for other weight graffiti, it is possible that the verticals here indicate units of 10 and the horizontal strokes indicate single mnas. By this process, 16 would be read as 50 mnas, and 18 indicates 28 mnas. A third such graffito, 45, preserves the ends of seven short horizontal segments followed by six verticals arranged in pairs. If the verticals do refer to units of 10, 67 mnas would be indicated. On the other hand, it seems odd to have the smaller units before the larger ones. Here, only the ends of the horizontals arepreserved,and it cannot be determined whether they were stacked on a single stroke as in the other graffiti just discussed. The use of horizontal lines to indicate 10 drachmas is attested in monetary inscriptions, so here the graffiti may use the horizontals to account for 70 mnas followed by 6 more mnas, for a total, perhaps gross, weight of 76 mnas.45 All of the possible weight notations considered so far involve quite ambiguous and simply rendered tallies. The only likely abbreviationsfor units of weight so far encountered are the two MA monograms treated above.The two multiple E/H graffiti, 6 and 29, could be read as indicating a hemistater, the equivalent of 1 mna, thus obviating their interpretation as multiple half-unit marks.In support of such a reading of multiple E/ H marks, a 4th-century amphora graffito with TTHHH[ ] AATT[was read by Lang as a tare notation, perhaps with 8+ mnas as empty weight and 25+ mnas as the gross or net weight (Lang does not provide a specific interpretation).46There is no clear evidence, however, that the term hemistater was used as a synonym in Athens for mna even though the mna was defined in terms of the stater.47

43. Lang 1956, p. 6, no. 16. 44. Johnston (1979, pp. 30-31) discussestallies "tied"or "bundled" together in this same way; a particularly close parallelis providedbyARV666, no. 13 (Boston 01.18, unpublished) with six short markscoming off a single longer one. There is no clearindication what the bundled units referto, whether numbersof vessels in a batch, price, or some other factor.See also Babinovet al. 1978, p. 134, no. 1755, pl.31. 45. For inscriptionswith horizontal lines for 10 drachmas,see Tod 19111912, pp. 104 (Epidauros),113 (Euboea);1936-1937, p. 241 (Epidauros).Tod (1926-1927, pp. 149-150 and 1936-1937, pp. 255-257) discusses the use of a horizontalbar for 1 mna at Cyrene.Johnston (1979, p. 31) also notes the use of the horizontalline for "10"on finewaregraffiti. 46. Lang 1956, p. 18, no. 76. 47. AgoraX, pp. 2-3.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

I5

The use of the term stater in the Athenian system of weights raises the possibility that E-graffiti (19 and 26), which Lang read as prices, might refer to weights of the jars or their contents. 26 has A 2 (10 staters) on one side of the neck and a clear abbreviationfor 10 choes, 2 kotylai (AX monogram KK)on the opposite side of the neck. Ten staters (20 mnas) is unlikely to be the weight of the recorded volume of contents since it is difficult to imagine what materialwould give such a low weight for more than 30 liters; even 30 liters of barleywould weigh 19.2 kg or ca. 42 mnas.48The 10 staters could, however, describe the weight of the empty jar to which the 10 choes, 2 kotylai were added or whose measured capacity was 10 choes, 2 kotylai. With 19, the seven staters recordedwould equal 14 mnas, a very likely weight for an empty jar of this type. Along this line of thought, however, the 14 staters I interpret as the reading of 25, following the volumetric note of 7 choes, seem unlikely to indicate the empty weight of the jar; 28 mnas seems high for a jar of this type. If the staters (?) on 25 are taken as an indication of price, then the other stater notations are perhaps also better considered as prices. The possibility that they refer to weight requiresmore evidence from weighed intactjars of preciselythe same types as those marked. None of these possible amphora weight graffiti juxtaposes a notation of the tare weight with either the net or gross weight. Lang wrote that without the tare weight, weight graffiti referring to contents alone would be meaningless.49This is true, however, only if the weighing did not occur in a face-to-face context. If the customer sees the empty jar balanced by the weighing device, then the customer knows that any further weight added-and perhaps recorded by marks on the jar-is the weight of the added goods. PRICE MARKS

48. Garnsey1992, p. 148. 49. AgoraXXI, p. 66, Hb 1.

50.This is the practicegenerally followedin finewarepricemarks, see althoughthereareexceptions; Johnston 1979, fig. 14:f.

(19,

25, 26, 30?,

44, 56, 69, 79?)

Among the many numerical graffiti discussed here, there are very few markings attributableto price. The most securely interpreted price marks are those that carryeither the common epigraphic symbol for drachma, F, or 2 abbreviatingstater.Lang interpreted other tally marks as prices, often in cases where the numbers exceeded the expected figures for volumetric notation (7-8 choes). Since, however, those tallies give no indication of being price marks, and given the presence of more explicit monetary symbols when prices are certainly intended, it seems more reasonable to read these largertallies as notations of weight (as above).Calculatingboth weight and volume might involve some process by which units are successively recorded and so are especially appropriatefor tallying. Price, on the other hand, would more likely have been conceived of as one figure and recorded using symbols for the largest units of the figure (e.g., A for 10 instead of ten vertical strokes) followed by smaller units.50 Perhaps surprisingly,given Athenian use of drachma coinage instead of staters,the 2 abbreviatingstater occurs nearly as often as the symbol for drachma.In the case introduced above, 26, we have an apparentjuxtaposition of price and volume-10 staters and 10 choes, 2 kotylai-on opposite sides of a Mendean neck. Lang noted that a price of 10 staters for just over

i6

MARK

L. LAWALL

10 choes would fit well with the generalization that "two drachmae per chous was the standard price for wine."51Similarly, for 19, a Chian jar I may be read as 7 staters.Later,in likely to have held ca. 7 Attic choes, T 11 the early 4th century, a jar of unidentified type carries the letters TTI on its neck in black paint (or charcoal?);this relativelysmalljar may have held 6 choes.52In discussing the first two pieces, Lang noted that staters in Chios and Mende were considered as didrachms,so the prices in the Athenian Agora would indicate a price of 2 drachmas per chous for both Mendean and Chian wine. Alan Johnston notes that if the graffiti were applied in the places of export, the writer may have been thinking in terms of local coin standards,thereby casting considerableuncertaintyfor us now as to the intended price in Attic drachmas.53Nevertheless, the 2 in all cases seems to indicate stater. Graffiti using the standard epigraphic symbol for drachma (F) are much simpler to read as price marks,but they arerare.54Drachmas were so indicated on only two 5th-century pieces published by Lang. The first, 44, has two clear drachma signs preceded by a difficult symbol read by Lang as two deltas sharing a common side,55giving a total of 22 drachmas.The arrangement of the deltas, however, is not paralleled elsewhere. Reading the symbol as a monogram TTA,for 50, has parallels in many fineware and pithos graffiti.56On this interpretation,the price would read 52 drachmas. Such a price might seem high, but the total number of known wine prices (if that is even the material being priced) is not so great that we can claim any certainty as to what is a normal price.57The second price in drachmas published by Lang is a dipinto, which partly covers a volumetric (?) graffito (56).58The price reads 16 drachmas, certainly at the lower end of prices associated with amphora contents. The only other securely read amphoraprice graffito known to me from the 5th-century Agora, 69, gives a price of 27 or 28 drachmas on a Chian jar. Depending on the volume of liquid in the jar,this graffitomight indicate a price of 3 or 4 drachmas per chous.59 51. Lang 1956, p. 13. Papadopoulos and Paspalas(1999, p. 177) remindus of Pritchett'scomment (1956, pp. 202-203, note 112) when writing about the Attic Stelai prices, that amphorascarriedmore than just wine or oil (perhapstheir intended primarycontents), so that the prices markedon amphorasonly indicate the price of their contents-whatever those may have been. While Lang is right to place emphasison the fame

of Chianwineandhencethe strong

likelihood that the jars contained wine, the many possibilitiesof other contents and of the pricesbeing inscribedat any of variouspoints in the life of the jar should be borne in mind (see below on interpretingthe graffiti).

52. P 30714, with a fully preserved rim, neck, handles, and part of the shoulder.The type is not published but is found extensivelyin late-5thand early-4th-centurycontexts in the Agora. See Lawall 1995, p. 167, note 212. 53. Johnston 1996, p. 82. 54. The rarityof these graffitiis particularlynotable in contrastto the many examplesof F-used on Attic finewaregraffiti. 55. Lang 1956, p. 15, no. 65, pl. 3.

56. Forexamplessee Lang1956,

p. 22, no. 99;Johnston 1979, p. 28; Solomonik 1984, pp. 121-122, pls. 2728, nos. 284-290; Jefremov1998, p. 74. 57. Prices in literaryand epigraphic

referencesand from graffitisuch as the ones discussedhere give a range from 4 obols per chous to 50 obols per chous;for studies of wine prices see note 16 above.Bagnall (1989, pp. 70-71) highlights the variability of commodity prices in papyriof the 4th centuryA.C. 58. AgoraXXI, p. 59, Ha 5. Johnston (1978a, p. 218) notes that this piece "allowsa numberof interpretations."For furtherdiscussion of the relationshipbetween the graffito markand the price dipinto and the possible interpretationsof the graffitoitself, see the catalogueentry below. 59. This fragmentis cited in Papadopoulos and Paspalas1999, p. 177, note 78.

GRAFFITI,

WINE

COMMERCIAL

SELLING,

AND

REUSE

OF

I7

AMPHORAS

ABBREVIATIONS

(8, 20, 2I, 54, 55, 66, 75, 76, 77, 78) The letters and other marks in the graffiti discussed above all appear in groups, and it was this circumstance that led to their identification as numerical graffiti pertinent to units of measurement.A much more common form of graffiti-for any sort of vessel-is an isolated letter or short abbreviation often thought to indicate an owner or maker of the pot. The ambiguity of these marks and the fact that they are found on so many different kinds of vessels make them hazardous to study solely in terms of graffiti on amphoras. Certain letters, however, often appear on amphoras in deposits with numericalgraffiti.60The letters E and H, alreadydiscussed abovewith other symbols, often appearas isolated letters. A second common graffito of this type is the letter M, sometimes contained in the abbreviationME.While it is possible that these abbreviations, E/H and M/ME, are simply owner's marks, their repeated presence on amphoras sets them apart from other isolated-letter graffiti and raises the possibility that they refer to aspects of the amphora or its contents other than ownership. E/H

GRAFFITI

As noted above, the letters E and H often appear as components of numerical graffiti. It is tempting to transfer some element of their meanings in that context to instances in which they stand alone. Given the frequent appearanceof volumetric graffiti on the amphoras studied here, a reading seems reasonof the E/H graffito as a half-metretes (,u[o.u [i?sTp.YcTOV) able. However, there is no single term in ancient Greek for half-metretes does not exist), so it is more likely that the H here stands for (y],st1,(?TpY]T defined by Hesychius as half of an Attic metretes.61The letter might hydria, indicate that the jar contains only a half-metretes,62which may be considerably less than its full capacity. If H, when found alone, stands for hydria, however, it is no longer so readily interchangeable with E. Indeed, the two cases where E stands alone, 54 and 66, differ from the H graffiti in terms of the type of amphora on which they are found. Both 54 and 66 are found on Solokha I ampho60. One furthersuchabbreviation (KA)is seen in 84 and, possibly,85 in Q15:2. The same abbreviationmay be restoredon amphorasat other sites and may,therefore,fall into the same class as the E/Hand M/MEmarkshere, i.e., not simply owner'smarksbut relatedto amphorasin a broadersense. Given the scarcityof these marks, however,in comparisonwith the abbreviationsdiscussedin this section, I have left the discussionof their possible meaningsto the catalogue entries.For examplesof KAgraffiti outside Athens, see Samothrace 11.2,

p. 100, no. 246, and Solomonik 1984, nos. 172 and 175. 61. Hesychius,s.v. ob8pocxic,and see p. 12, above. 62. The assumptionthat an Attic metretesheld 12 choes is supported by only one papyrusdocument,to my knowledge:P Ryl. 4, document 564rp, ctr, 17, line 5; but even here, 8o6?x6yXoot"is restored(though with good parallelswhere metretesis modified, but not with the specificlabel of "Attic").Otherwise,"Atticmetretes"is expressedonly in Hesychius'sdefinition of 1 hydria(see note 61). Modern

authorshave defined the term assuminga duodecimalsystem for the relationshipsbetween units of measureand using comparisonswith Roman units of measure;see, e.g., Hultsch 1882, p. 101, note 6, which begins:"Aneinem direktZeugnisse iiber die Einteilung des attischen Metretes fehlt es...."

Epiphanius

(Treatiseon Weightsand Measures,43) distinguishesbetween a "sacredchous" as a twelfth of a metretes and a "complete"or "greater"chous as a "ninthof a metretes;see Dean 1935,

p. 56.

18

MARK

L. LAWALL

ras, a type that is rarelyresinated. If one follows the communisopinio that resinated jars were for wine and some nonresinated jars more likely for oil,63then perhaps the E here specifies that these contain oil, EXxaov. A few Solokha I jars are known, however, with fairly small capacities, so the E could serve as an abbreviation for half-metretes (in the longer form hetLoLuUTvp-yroo, which would allow the E to abbreviateYi,lUo). Lang suggested half-metretes as the reading for the E followed by two vertical strokes on 13.64As discussed in the catalogue entry below, however, such a reading does not fit the possibilities of dialect associated with reading E for hSiL- nor the measured capacity. M/ME

GRAFFITI

Given the argument above for H abbreviating hydria/half-metretes, the most direct reading for M or ME would be as a single complete metretes. Such a reading certainly finds support in the large size of the Solokha I jars on which these letters are often found (8, possibly Solokha I form; 55; 77; and 78, Samian jar in Solokha I form); some jars of this form can hold slightly more than 1 Attic metretes, 12 choes.65Problematic for such a reading, however, is the repeated presence of M-graffiti on necks that belong to much smaller jars (20 and 21). On these two amphoras, both from the same deposit, a letter M is found on either side of the neck. All four letters are very carefully and visibly inscribed. It seems unlikely that the writer would feel the need to be so emphatic about the fact that the amphora contained 1 metretes. One possible reason that the Ms are so clearly inscribed is that it was very important for the jars' contents to be known to the buyer. Here the absence of resinous coatings on jars with the M/ME graffiti at the Agora seems relevant.As noted above, a lack of resin is sometimes taken to indicate oil as the primarycontents of an amphora.It is difficult to reconcile M with a term related to oil given cases where the abbreviationis lengthened to ME;66had the abbreviationbeen lengthened to MY, then perhaps perfumed oil (ji6pov) might have been the meaning. The presence of ME, however, suggests reading the M as signifying tiSXt,honey. An unresinated amphorawould seem to be quite an appropriatecontainer for honey. Honey weighs much more than an equivalent amount of oil.67If honey was sold from reused oil containers or reused, unlined wine jars, it would not be surprising to see the containers in which it was sold very clearly marked as such. Later amphora dipinti clearly identify am63. Koehler(1986, pp. 50-52) collects many of the ancient references relatingresin/pitchto wine amphoras, and (p. 52) lists wax, gum, and the dregs of olive oil productionas possible lining agents for oil amphoras. Dupont (1998, p. 182, note 257) questionsthe exclusionof other productsincluding oil from resinated jars. Resin was also transportedin amphoras,accordingto G. Bass

(pers.comm., November 1996). 64. Lang 1956, p. 9, no. 30. 65. Brashinskiy(1984, p. 124, with charton pp. 198-199) lists capacities for Solokha I amphorasof 39,700, 32,200, 17,500, and two at 14,000 cm3. In addition,an MEgraffito appearson a CorinthianB type shoulderfrom deposit B 13:5 (Lawall 1995, pp. 342, 384, CrB2), which is not included in the cataloguehere since it is not found

with any other numericalgraffiti;such a jar would have a capacityof much less than a complete metretes. 66. See, too, ME(as monogram)Al, Solomonik 1984, no. 341. 67. Lang, in AgoraXXI, p. 80, He 29, notes that ancient metrological writers consideredhoney to weigh one and a third times the equivalentvolume of wine or water.

WINE

GRAFFITI,

SELLING,

AND

REUSE

OF

AMPHORAS

I9

phora contents as honey.68Perhaps to be cited in support of this reading is a fragmentarybeehive reported among the finds from well R 13:1,69a well immediately adjacent to deposits containing examples of the M/ME graffiti (see catalogue and "Findspots,"below) and containing many amphora graffiti, although none with the M/ME markings. A related reading for the M/ME graffiti could be a label for honeyed wine (L?XtTT(tY otvoS, wine prepared with honey, or VsXChpoo; olvo;, honey-sweetened wine).70 Such wine is mentioned both in the medical authors and in later Roman amphora dipinti from the Agora.71While the lack of resin does not fit quite so well with this reading, here too the seller of the jar might be concerned to indicate very clearly that the jar contains honeyed wine, not normal wine. SUMMATION (2,

68. AgoraXXI, p. 73, and He 29, He 33, He 34, and He 36. 69. Liidorf 1998-1999, p. 76. has manyvariant 70. MeMiXpooo spellings,especiallyin papyri;see LSJ, s.v. tisXXpoo;. Agora XXI, p. 79, He 21 reads olvo)u ?XiLTLVOO, which should

mean wine made from honey,not honeyed wine. The writerof He 21 may have intended olvoouV?XLTTirou; there seem to be some extraneouspainted strokestowardthe end of the second word that could be readin this way. 71.AgoraXXI, p. 73; p. 79, He 21, He 30; and see Papadopoulosand 80, p. Paspalas1999, p. 175. For ancient and references,see LSJ,s.v. i.eXEMpooS 72. See Lang 1956, p. 5, nos. 10-14 andAgoraXXI, p. 59, Ha 5-Ha 7 for examplesof summationgraffiti. 73. AgoraXXI, p. 59, He 7.

4, 6?,

22,

MARKS

23, 24?, 46?, 56, 57, 80?,

91,

98?)

In a few cases, a simple tally of vertical strokes is accompanied by an acrophonic or alphabetic numeral, which can be read as the summary of the tally.72These summation marks occur in three different patterns: 1) simple vertical strokes are summarizedby a single alphabeticor acrophonic numeral; 2) the summarizing numeral follows a combination of singleunit strokes and largerunit symbols; and 3) the summation occurs midway through the tallying process and is followed by further notations. The first pattern is the most common here with perhaps as many as seven examples catalogued:2, 22, 23, 56, 57, and possibly 46 and 80. The tally marks are sometimes more sketchily cut than the summation marks, but more often all marks are cut in roughly the same fashion. The fact that the number of strokes equals the alphabetic or acrophonic numeral suggests a summation. The tallies might stop one unit short of the total if the summation mark was intended also to include the final small unit. The second pattern only occurs twice and both examples are open to question (24, 98). In 24 an H might be read as a half-metretes (6 choes) followed by four vertical strokes to give a total of 10 choes. Below this graffito is an incised delta that could be read as an acrophonic 10 recording the total indicated by the smallerunits. For 98, the largest numeral,a delta, may have been inscribed first, then the series of four very sketchy strokes cut as the writer worked his way up to the "5"indicated by TT.This piece is unusual among the summation marks as it seems to indicate a measurement of weight rather than volume. The third pattern is also fairly rare(4, 6, and 91). Here the summary is inscribed part way through the counting process and is followed by symbols for additional, usually smaller units. Lang read these graffiti as the result of pouring in a certain number of small units of liquid and then noting the total before adding more to top off the jar.73It may be significant that two of the three examples of this pattern, 4 and 6, are earlierthan the bulk of the graffiti and the third example, 91 (which comprises two graffiti), may have a slightly different interpretation.High on the shoulder near the base of the neck is the volumetric notation 111i.Lower on the shoulder is another marking of five vertical strokes. Following the third pattern of summation graffiti, the five verticals would be read as being

20

MARK

L.

LAWALL

summarized by the TT.After these 5 choes, another three were added, for a total of eight. Alternatively, the eight may have been noted first, and the five strokes noted later to account for removed contents.74 This last example raisesthe question of whether the summariesrecord liquid poured in or poured out (decanted, for example, into a customer's smallerjar). In examples of the third pattern, especially in 4 and perhaps 6, in which larger to smaller and more precise units of measure are used, it seems likely that Lang is right to see the jar on which we have the graffiti being filled and topped off. For most of the other examples, however, it is possible that the summation represents the known quantity of liquid in the jar and that the smaller tallies result from liquid being poured out or decanted into smaller containers (for more on such processes, see discussion of practices, below). COMBINED

NOTATIONS

(25, 26, 56)

Jars carrying more than one type of notation are quite rare among the graffiti treated here. In later periods, among finds from across the Agora excavationsand elsewhere, combined notations are much more common.75 It is possible that had more fragments preserved both sides of the amphora neck or broaderparts of the shoulder, more combinations of markings might have appeared.The three combined notations that are preserved all involve volume and price. In two of these cases, 25 and 26, the volume is likely to represent the complete filling of the jar and the price indicates the cost of the contents (probablywith the jar as well).76As noted above, the A2 on 26 might instead be an indication of the weight of the empty vessel. If so, the marks on 26 would indicate the specific capacity of the vessel of the marked weight. For the reasons presented earlier,however, an indication of price on 26 is more likely. 56 seems to reflect at least two stages of marking. The graffito on the jar is a summation mark of 5 choes, with the simple pattern of a series of strokes summarized by a numeral (in this case a three-letter abbreviation)for the total. This graffito, however, is then covered by a red stripe. Further to the side on the same neck fragment is a dipinto that may be read as 16 drachmas, also in red paint. It seems likely that the graffiti and the red dipinti represent two occasions of marking. First, 5 choes (unlikely to be the full capacity of the jar) were poured in or decanted. This graffito was then painted over, perhaps to cancel its message, and the price mark was painted. There is no need to see the price as related to the 5 choes; these may have long since left the jar by the time the price was marked. SUMMARY

Broad patterns in the nature of the graffiti may be summarized as follows. Volume does seem to be the point of information predominant among the graffiti, but the marksdo not always record the full capacity of the amphora. The differing implications of those markings that might account for the complete capacityof the jar and of those that recordsmalleramounts are considered below. Notations of volume appear occasionally with price marks and in the format of a summation of smaller measures by a single larger unit.

74. Lang (1956, p. 6, no. 17) suggested that the five strokesmight indicate kotylaiif there were only two strokesfollowing the n. This readingis not well supportedby other examples. Kotylaitend to be listed with the rest of a notation. It is unclearto me why an indication of only 7 choes in the main part of the graffitiwould recommend the simple tally to be read as kotylai. 75. AgoraXXI, pp. 75-81. 76. For 98, Lang (1956, p. 15, no. 64) suggests a separateprice for the jar and for the contents. It is conceivable that jarswere sold separately,given that much of the wine might have been decanted and sold in smallerunits (see below). 98, however,would be the only examplehere of a price for contents separatefrom the price of the jar.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

21I

Notations of weight are, at best, moderately explicit on a very few Classical amphoras. Even these, where M is read as abbreviatingmna, are open to question. The other markings interpreted here as notations of weight come as tallies that seem to surpass the likely volume of the amphora and fall within the range of the possible empty weight of the jar and the gross weight of the filled jar. Price markings are generally rare on these amphoras.The same holds true to some extent with fineware graffiti:in relative terms, general "trademarks"are far more common than price marks in particular.A second significant feature of the amphora price marks is that they often seem to involve reference to staters even though Athens did not use that particular unit of coinage in the 5th century.The rarityof price marks and this use of "foreign"terminology are useful factors in identifying likely contexts for the application of price marks (see below). Isolated letters of abbreviationsare fairly common on amphoras and other pottery, and in most cases it is not at all certain whether they have any particular commercial significance. On the other hand, two sets of abbreviations, E/H and M/ME, recur sufficiently often on amphorasboth within the major graffiti-bearing deposits in the Agora, considered here, and at other sites-that they seem to have some particularmeaning related to amphora use. E/H for half-metretes, or H alone for hydria, E for elaion (oil), and M/ME for metretes, meli (honey), or honey-flavored wine seem to be the most reasonable interpretations given the nature of the amphoras on which the marks appear. Summation markings, as noted above, tend to involve notations of volume, but in one anomalous case weight may be the topic of the graffito (98). By far the most common pattern for these is also the most simple, a series of single strokes for counting up to or down from an overall total. Indeed, the primaryissue of interpretationfor these marksis whether they attest to decanting from or to filling the marked amphora (see below). The scarcity of combined notations has also been noted. In the few cases found here, the focus seems to be on volume and price. While both price and volumetric notations often occur in isolation, it is noteworthy to find this juxtaposition of information. Such paired notations may help determine which stage in the distribution of the amphoras might be referred to by the price inscriptions. Above all, in considering the readings of individual markings and determining why these markswere applied, it is important to emphasize the difficulties of interpretation and the limits on our certainty.At best, support of parallel syntaxes in other graffiti provides reasonable security in approachingindividual problems, but many uncertainties and possibilities remain. On the other hand, the fact that so many graffiti, especially those that use abbreviationsfor units of volume, may be read as matching or approximating measurementsof preserved,contemporaryamphorasstrongly suggests that we are on the right track. Readings of weight notation are more problematic. They are suggested here as an alternative to the difficulty in reading tallies as price marks when more recognizable monetary symbols are commonly used on amphoras, pithoi, and finewares. On the whole, it seems useful to make suggestions as to readings and to explore where those readings lead in terms of interpreting the graffiti.

MARK

22

CATALOGUE AMPHORAS

OF GRAFFITI

L.

LAWALL

ON 5TH-CENTURY

The pieces illustrated here include both previously unpublished or unillustratedgraffitiand previouslypublished graffitiof debatedreading.Lang provides illustrations of most pieces not shown here.77 The deposits arelisted in approximatechronologicalorder.All findspots are indicated on Figure 1; a detail of the walls and Classical well deposits in the vicinity of grid unit R 13 is illustrated in Figure 2. Within each deposit, the entries are presented in the following order:volumetric notations; notations of weight (?);prices; abbreviationsE/H and M/ME; summations; and others. Within each entry, the following information is provided: catalogue number, Agora inventory number, general type of graffito present, amphora type, previous publication, preservation of the amphora fragment, description of the graffito, discussion of the reading of the graffito, and the likely date of the amphora type. The description of the cuttings as light, moderate, or heavy-referring to the force applied in creating the graffito-is admittedly subjective but is meant to give some impression of the variation among the markings. In discussing the readings of each marking, I have tried to consider differentpossibilities;in many cases I see no way of confirming one interpretation over the others, and the reader may have even further interpretations. Comments on the dates of the amphora fragments and the deposits containing the graffiti are based on my current research on Late Archaic through Hellenistic amphora typologies. Since many of the better-preserved pieces and larger deposits will be discussed in depth in a future publication, I do not consider chronological issues in detail here. EARLY

EXAMPLES

OF NUMERICAL

GRAFFITI

The following graffiti are from deposits closed before ca. 425 B.C.These pieces are listed in roughly chronological order, as dated by either the findspot or, when possible, the amphora carrying the marking. 1

(P 11068). Volume. Chian C/l.

Fig. 3

D 15:1; well filled during cleanup after the Persian sack in 480 B.C. (Shear 1993, pp. 434-435). Lang 1956, p. 3, no. 2. Preserves half of rim, neck, and one handle. Seven horizontal, parallel incisions across handle, each cut quite deeply. Lang interprets this graffito as indicating 7 Attic choes, assuming that early Chian jars held 8 Chian choes. Brashinskiy (1984, p. 171, nos. 20-23 and 30-33), however, cites a range of sizes from less than 7 to nearly 10 choes. For the form of this jar and all three of the major 5th-century Chian forms (C/1, C/2, and C/3), see Kerameikos IX, pp. 23-24; the C/1 form is datable between 520 and 480 B.C.

77. Lang 1956;AgoraXXI.

GRAFFITI,

WINE

SELLING,

AND

REUSE

2

1

3 7

8 Figure 3. Graffiti from deposits closed before ca. 425 B.C.

OF

AMPHORAS

23

MARK

24

2

(P 9340). Summation. Unclassified type.

L. LAWALL

Fig. 3

Section D grid 65/MB, notebook p. 279, context dated to late 5th or early 4th century,just south of South Stoa I. Lang 1956, p. 5, no. 14. Unclassified handle fragment. Preserves only lower part of handle. Graffito at the base of the handle: H11111. Very short, small, carefully cut symbols. Lang reads a final alphabetic numeral for "8,"created by linking the sixth and seventh verticals, after 7 individual choes were poured into the jar.The vertical tallies may record subtraction from the total of 8 choes as each chous is poured out. The careful cutting of the letters may suggest a single marking occasion for the filling or decanting. Neither the form nor the fabric is precisely datable, but the fragment is likely to date to the late 6th or early 5th century. 3

(P 5176). Weight. Northern Greek, possibly Thasian.

Fig. 3

H 6:5, well found under the Stoa of Zeus, cut through Persian destruction level (Talcott 1936, p. 333, note 2). Talcott 1936, pp. 344 and 352. Nearly complete jar with minor restorations. AM ligature on shoulder with MA written out in full. Placement of AM relative to MA not indicated on Figure 3. Lightly incised, large letters. Following Lang's reading of a similar ligature on a small table amphora (AgoraXXI, He 3), this graffito may be read as "amphoreus10 mnas."Talcott (1936, pp. 344, 352) discusses graffiti on two other amphoras from the same well. One (P 5174; AgoraXXI, p. 33, F 60 = PAA 122810) carries the graffito AMA; the other (P 5175; AgoraXXI, p. 34, F 61) is marked XAP. By comparison with P 5174, the AM monogram on 3 may be simply an owner's mark. Lang (AgoraXXI, p. 33) notes that names starting with AMA tend to be either "heroic or later than the 5th century."Cf. PAA 122818, possibly ca. 400 B.C.3 was chosen for inclusion in this catalogue on account of the MA mark incised higher on the shoulder, near the AM; MA seems unlikely to abbreviatea name. The fabric of this amphora type bears some resemblance to later stamped Thasian amphoras, and the form is widely paralleled across the north shore of the Aegean (Lawall 1997). Date ca. 460. 4

(P 26070). Summation. Northern Greek.

Section W; surface find in the areajust south of the Mint. AgoraXXI, p. 59, Ha 7, pl. 32. Preserves only the lower part of handle. IIIIIIIZEKKrunning toward the shoulder along the outer face of the handle. Deeply cut symbols, with tally strokes cut more deeply than the following letters.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

25

Lang reads as 7 choes, summed up by the Z (=7) followed by half a chous and 2 kotylai. Although it is difficult to establish whether this was a roughly 7-choes jar (few capacity measures are published from this period), there is no ready alternative to Lang's reading. The low-set angle of the handle against the shoulder finds best parallels in the first half of the 5th century,perhaps the second quarter; cf. Grace 1953, p. 106, nos. 158-159, pl. 39. 5

(P 21965). Weight? Unknown type.

N 7:3, well just northwest of the north end of the later Stoa of Attalos, published in full by Boulter (1953) with catalogue of amphoras by Grace (1953). Boulter 1953, p. 100, no. 140, fig. 4; Lang 1956, p. 6, no. 16. Preserves only a small, poorly diagnostic body fragment of a relatively thin-walled vessel. Graffito A followed by two rows of four vertical strokes, one row above the other. Lightly cut. Lang (1956, p. 6, no. 16) considers that this would be "anunusual price inscription"and suggests instead 10 choes, 8 kotylai. While such a large capacity is possible, the large number indicated could record the empty weight of the jar, 18 mnas (ca. 8.2 kg). Date before ca. 440 B.C. 6

(P 30085). Summation? Type uncertain, possibly northern Greek.

H 4:5, a pit filled just west of the Royal Stoa, published in full by Rotroff and Oakley (1992). Rotroff and Oakley 1992, p. 125, no. 356, fig. 22, pl. 60. Neck sherd only. Graffito in small, carefully cut letters: HHHT 1111.The vertical strokes are set slightly apart from the preceding letters. The T might sum up the four individual tallies: perhaps then we have 3 hemichoes and 4 kotylai. Alternatively, the T might indicate a third or quarterof a chous following the accounting of the 3 hemichoes. While it is possible that this sherd was inscribed after breaking, the T is more likely to indicate a fraction than either the numbers 30 or 40; and 300+ talents seems like a large sum to inscribe on a simple ostrakon. Although Rotroff and Oakley (1992, esp. pp. 53-57) suggest a closing date for the deposit around 425, the diagnostic amphora pieces and many of the other ceramics are no later than ca. 450. The graffito may also be much earlier than ca. 425. 7

(P 17124). Volume or weight? Chian C/3.

Fig. 3

A 20-21:1, deep cutting for drain.The latest datable piece in the fill bell is a krater,P 17000, ca. 400-397 B.C.See Young 1951, pp. 254-257. Full rim, neck, both handles, part of shoulder. Large red-painted A on neck; below this is a graffito: Al. The

MARK

26

L. LAWALL

dipinto does not cover the area of the graffito, so it is not certain which mark was applied first. The incised letters are fairly large in comparison with other pieces in this catalogue, and the incisions are of moderate depth. Eleven choes would be fairly large for a jar of this type; 11 mnas is a possibility though this would be fairly light for such a thick-walled and tall jar (no intact jars were availablefor more precise estimation of the range of weights for this particulartype). The jar form is late in the bulging neck series, either just before or very early among the pieces found in well R 13:4 (see below); see KerameikosIX, pl. 64:4 and 8. Date ca. 440-430 (much earlier than the bulk of the deposit, which seems to date late in the 5th century). 8

(P 25896). M-graffito. Possibly Solokha I.

Fig. 3

M 17:7, a pit just south of South Stoa I. Preserves only handle. ME cut heavily on outer surface of handle; the letters are very visible. On the basis of this handle and ME graffiti from other sites, it is likely that the M graffiti abbreviatea word beginning with ME rather than MY.Although it is possible that this is a noncommercial initial, the common appearanceof M or ME in Athenian graffiti suggests the reading of [jierpYj-Tncor seTp-YjTo(measured). Another

possibility, however, is that the M/ME stands for ,usXt(honey). It seems noteworthy that neither this piece nor any of the amphoras marked with M have a resinous coating on the interior surface. The profile and fabric of this handle are best identified as belongto ing the Solokha I type, mushroom-rimmed jar (cf. Lawall 1995, p. 346, NG 22). For this type, see Lawall 1995, pp. 218-233; Zeest 1960, pp. 91-92; and Mantsevich 1975. The amphora material and other pottery are datable early in the third quarterof the 5th century and clearly earlier than the latest finds in R 13:4. WELL

R 13:4

Well R 13:4 is just south of the later Stoa of Attalos, perhaps associated with very poorly preserved 5th-century walls. The well was filled ca. 425 and the datable contents range from 440 to 425 (for complete publication, see Talcott 1935). Rotroff and Oakley (1992, p. 56) propose a link between this deposit and the earthquake of 425 B.C. 9

(P 33411). Volume. Probably Chian straight-neck.

Fig. 4

Preserves shoulder only. Graffito TTIII1.The second leg of the 11 is shorter than the first; the vertical strokes vary in length; marks lightly cut. A tally of 9 choes might indicate the use of Chian units, and, if one follows the 9:8 ratio for Chian to Athenian units (Barron 1986, p. 98, note 49; Wallace Matheson and Wallace 1982, p. 300, note 21;

WINE

GRAFFITI,

I9

SELLING,

(

\

AND

REUSE

OF

11))//,I\

AMPHORAS

2/

9

11

10

/

/

\ \

14

rr7

15

Ori

18 -

FFr 18

20

/n

11

I111\11 21

23

i Figure 4. Graffiti from well R13:4

25

27

I

28

MARK

L. LAWALL

Wallace 1986, p. 88, note 7, citing a ratio of 50:56, which gives 8.03 Attic choes for 9 Chian), then this marking would indicate a "convenient"match to Athenian standards (less convenient if one uses the 8:7 ratio proposed by Lang 1956, p. 3; Grace and SavvatianouPetropoulakou 1970, p. 360; and Mattingly 1981). Nine Attic choes would be an unusually large capacity for this amphora type; indeed such a figure is unattested. For this reason, 9 and 10, if they are to be read as volumes, must have been inscribed before the jars reached Athens. The identification of this fragment as probably being from a straight-neck Chian jar is based on the wide shoulder with a sharp outer edge and somewhat finer fabric than is commonly seen in the latest of the bulging neck jars of Chios datable to the third quarter of the 5th century; see, e.g., Grace 1979a, figs. 44-45. The closing date of this well limits this fragment's date to ca. 430-425. 10 (P 33413). Volume. Probably Chian straight-neck.

Fig. 4

Shoulder only. Nine vertical strokes remain-the sherd breaks off without leaving enough blank space to guarantee that the tallies did not continue. Lightly cut marks;uneven lengths and spacing. Resinated interior. As with 9 above, 9 choes could be indicated. Same type as 9 above. Date ca. 430-425. 11 (SS 1845). Volume. Chian straight-neck.

Fig. 4

Talcott 1935, p. 496, no. 85, fig. 17 (jaronly), and pp. 514-515. Restored jar with both handles preserved, nearly complete rim, and complete toe; large areas of neck and body restored with plaster. Graffito of six vertical strokes. Lightly cut, low on neck between handles. The neck is broken immediately left of this graffito, raising the possibility that there were originally more strokes. A reading of 6 or more choes seems mostly likely. It is unfortunate that the beginning of the graffito is not preservedwith certainty. Were it possible to know that only 6 choes were marked as the capacity of this jar (if one reads these marks as necessarily checking the full capacity of the jar, as need not be the case), then the attendant stamp would not guarantee even a minimum capacity of the "standard"8 Chian choes. The kantharos stamp here reappearson other Chian jars of roughly the same period and somewhat later; the same stamp appears on 69, below, and SS 14080 from R 11:3 (closed ca. 420-410). Date ca. 430-425. 12 (SS 1839). Volume. Chian straight-neck. Lang 1956, p. 6, no. 15, pl. 6; Grace 1979a, fig. 45, far left. Mended and restored to complete amphora form, missing only parts of body (restored in plaster).

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

29

Graffito on neck: TTwith two vertical strokes followed by three horizontals. Very lightly cut marks. Lang reads 7 choes, 3 kotylai. Jar carries sphinx stamp at the base of the handle (see Grace 1979a, text with figs. 48-49, and Zeest 1960, p. 77, pl. 3). Date ca. 430-425. 13 (P 2368). Volume. Chian straight-neck. Talcott 1935, p. 516, fig. 28f. Heavily restoredjar, missing handles, large part of shoulder, fragments of body; toe in very poor condition from salt damage. Graffito Ell. Fairly light strokes but larger letters than many in this series. Talcott reads this mark as an alphabetic numeral E indicating 5 choes, followed by two more for 7 choes. Lang (1956, p. 9, no. 30) reads 1/2metretes and 2 choes, and this is equivalent to the measured capacity of the jar of 8 Chian choes. Lang proposed that the use of E for 1/2indicates Chian psilosis, but E would not be used by an Ionian Chian to abbreviatea word beginning ,uI-. An Athenian or other non-Chian-based writer would not record volume by Chian standards. Talcott's reading of 7 choes fits both the measured capacity (8 Chian choes = 7 Attic choes) and the requirements of dialect, but such a use of E instead of the very common TTmight have been confusing. The E could also stand for either 1 or 1/2chous, followed by 2 kotylai. Such a tally would not describe the full capacity of the jar, and the E for "sIS;" would be unusual (though not without parallel;see 25, below). Date ca. 430-425. 14 (SS 1841). Weight? Chian straight-neck.

Fig. 4

Lang 1956, p. 17, no. 72. Preserves three-quarters of rim, both handles, and complete toe, but much of the body is restored in plaster. Graffito on neck: AA. Light to moderately deep incisions. Lang reads as a tare weight of 20 mnas with supporting evidence of the weight of the jar.The large amount of plaster used in the restoration of the vessel renders the weight measurement of the jar itself of uncertain value. Nevertheless, weighing intact jars without plaster restorations indicates that 20 mnas is a possible weight for an empty vessel. Date ca. 430-425. 15 (P 33412). Weight? Chian C/3 or very early straight-neck

Fig. 4

Preserves only fragment of neck and shoulder. Graffito on lower part of neck: A I1111111. II Light to moderately deep incisions. The marks indicate 17, perhaps a price, just over 2 drachmas per chous; 10 choes and 7 kotylai would seem too large a capacity for a jar of this type. Without the standard drachma symbol a price

30

MARK

L.

LAWALL

interpretation is uncertain. Seventeen mnas (ca. 7.5 kg) might correspond to the empty weight of a large, heavily built Chian C/3 type amphora. Both the Chian C/3 amphoras and the earliest straight-neck jars have a coarser fabric, as is the case here, than is usually associated with the fully developed straight-neck type. Date ca. 440-430. 16 (P 2367). Weight? Very late variant of Chian C/3. Talcott 1935, p. 516, fig. 28c; Lang 1956, p. 14, no. 63. Preserves rim, neck, and one handle. Graffito on neck: five vertical strokes, on the fifth of which are ten short horizontal marks. Fairly coarse strokes, each one trailing off at the ends. Resinated interior. Lang reads as 14 drachmas.The use of such a long, complex symbol for 10 drachmas seems unnecessary,and it is unusual to place the larger unit of price after individual marks indicating, by Lang's reading, smaller units. An alternate reading would be 4 choes and 10 kotylai. As each kotyle was removed, another would be added to the short stroke tallies; decanted choes would be accounted for with the longer strokes. With this reading, only part of the complete potential volume of the jar was decanted in the recorded operations. The verticals could also be read as units of 10 mnas to reach a total of 50 mnas for the marking. Once four 10-mna weights were balancing the jar with its contents, single-unit weights were added until the jar was filled, and each of these was marked along the fifth vertical stroke. Fifty mnas of wine would be equivalent to ca. 22 liters, a common capacity measure for Chian jars of this period. Short strokes tied together by a single line are discussed by Johnston (1979, pp. 30-31) without clear indication of whether price is being recorded or simply numbers of vessels in the batch; Tod (1911-1912, pp. 108 and 116) presents epigraphic examples of verticals linked by a single horizontal as indicating multiple drachmas,but in these cases the practice of linking drachma signs seems to begin only in the 3rd century B.C. Date ca. 440-430/25. 17 (SS 1840). Weight? Chian straight-neck. Lang 1956, p. 4, no. 8, pl. 1. Mended and restored to nearly complete amphora form; missing toe and minor fragments of the body. Graffito on neck: three horizontal strokes over two circular marks, followed by seven vertical strokes. Coarsely and clearly cut graffito. Lang reads 7 choes, 3 kotylai, 2 "smallerunits."The horizontals preceding the verticals may indicate units of 10, with the dots below as single units, and the verticals-despite their greater

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

3I

AMPHORAS

length-as smaller units. This system is used on the early-4thcentury building accounts from Epidauros; see Tod 1911-1912, pp. 103-105, in which the markings refer to drachmas and obols. It seems odd, however, to list 7 obols instead of another drachma and 1 obol. A listing of 32 mnas and seven smaller units seems more likely. The amphora has a sphinx stamp at the base of the handle. This stamped Chian type is the same as 12 above, date ca. 430-425. 18 (SS 1842). Weight? Chian straight-neck.

Fig. 4

Restored jar missing one handle, parts of rim and neck, and fragments of the body. Graffito on the neck: two verticals followed by a third from which there project eight short horizontals. There is a trace of a horizontal line between the first two vertical tallies, but this may not have been deliberate. Light vertical strokes, slightly coarser and clearerhorizontals. It is difficult to apply Lang's method of reading a price mark on 16 to this mark (8 drachmas on the vertical plus two more marked with the larger strokes to give 10 drachmas). A volumetric notation might be read as follows (corresponding to a possible reading for the similar 16): the verticals indicate 2 choes and the shorter, horizontal tallies refer to 8 kotylai. Such a reading has two implications: 1) the graffito does not "check"the actual capacity in light of the stamp since there are too few marks;and 2) given that there was still far more than 1 chous remaining, the kotylai were not added to top up the capacity measurement.The vertical strokes might, therefore, account for choes of decanted wine with the bar of horizontals used to keep track of smaller amounts (kotylai). The weight-mark reading method, which worked well for 16, would give us 28 mnas for this marking. Such a weight falls between the likely empty and gross weights of amphoras, but the mark could indicate either the net weight of the added liquid partly filling the jar or the net weight of a grain like barley filling the jar (20 liters barley = 12.8 kg = 28 mnas). Amphora with sphinx stamp at top of neck, date ca. 430-425. 19 (P 2372). Price mark. Chian straight-neck.

Fig. 15

Lang 1956, p. 14, no. 62; Talcott 1935, p. 516, fig. 28e. Nearly complete jar missing only small bit of the rim and parts of the body. Graffito on shoulder near the base of the neck: T I12. Very light strokes. Lang reads a price of 7 staters.The use of a non-Attic unit of currency supports the view that many of the price marks seen in these graffiti were applied for use at the Peiraieus (or another Attic port), where merchants would be accustomed to moving between many different currencies,or before reaching it; see below, p. 75. Date ca. 430-425.

32

MARK

L. LAWALL

20 (P 11384). M-graffito. Possibly Chian straight-neck.

Fig. 4

Preserves part of rim, neck, one handle. Neck with graffiti: one M on either side of neck. The position of each M relative to the other is not indicated on Figure 4. Moderately coarse strokes. No resin. As in 21, the writer seems to have wished to make the M very clear and visible, from either side of the jar. Perhaps the contents were honey or honeyed wine, not some lighter liquid or standard wine. Possibly an overfired Chian straight-neck; however, the grayish surface color and the quite short neck are somewhat unusual for Chios. The form, whether Chian or another similar type, is unlikely to date much before 430-425. 21 (P 11385). M-graffito. Possibly Chian straight-neck.

Fig. 4

Preserves rim, neck, upper attachment of one handle. Same graffiti as 20 with same arrangementof one M on either side of the neck, moderately coarse strokes. The jar type is the same as 20, but the neck here is slightly taller; date ca. 430-425. 22 (P 2371). Summation. Chian C/3. Talcott 1935, pp. 515-516, fig. 28b; Lang 1956, p. 5, no. 10. Jar restored, missing a few bits from handles, neck, and shoulder. Graffito consisting of an enigmatic ligature (see Talcott 1935, followed by eight vertical strokes;a bit further from these, 28b) fig. set above them, is a smaller H. Moderately deep to deep markings with the initial ligature being especially carefully cut. Talcott suggests a false start writing an alphabetic 6 to explain the ligature, then a difficult time of counting up to eight before inscribing the H, giving a total of 14 (drachmas) as a price. Lang reads the tally as 8 Chian choes, summarized by the alphabetic numeral H, but does not comment on the preceding ligature. Date ca. 440-430/25. 23 (P 11386). Summation. Chian handle.

Fig. 4

Lang 1956, p. 5, no. 12. Preserves only the lower half of the handle to the point where it attaches to the shoulder. Graffito on the outer surface of the handle near base: H set over seven vertical strokes.The marks are exceptionally small and carefully cut, but the H is cut much more deeply than the vertical strokes. Lang reads as seven single-chous measures poured in and the summation 8 (H) inscribed after the eighth chous was poured in. Given the difference in weight between the verticals and the H, Lang's reading of relatively less formal tallies followed by a final, formal summation

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

33

seems very likely. A chous-by-chous emptying of the jar would result in the same pattern of markings. The form and fabric of the handle require a Chian identification; the relative fineness of the fabric suggests the later straight-neck type, ca. 430-425. 24 (P 11383). Summation? Probably Chian straight-neck. Lang 1956, p. 8, no. 26, pl. 1. Preserves only neck and a bit of the shoulder. Graffito on lower part of neck: HIII , with A below, closer to the shoulder.The incisions are very light and tend to become fainter from left to right, perhaps evidence that the tallies were all cut at one time. The A, however, is cut in an even more sketchy manner, so it may not have been cut at the same time as the other marks. Very slight traces of resin inside. Lang reads as 10.5 choes, 4 kotylai. Alternative readings include a half chous and 4 kotylai decanted against a total of 10 choes; or a half metretes (hydria) and 4 more choes for a total, recorded by the A, of 10 choes. The A might also simply be an initial, unrelated to the capacity,cost, or weight of the jar. Date ca. 430-425. 25 (P 2366). Volume and price. Chian C/3.

Figs. 4, 14

Lang 1956, p. 12, no. 58; AgoraXXI, p. 76, He 2, pl. 42; Talcott 1935, p. 516, fig. 28a;Johnston 1996, p. 82, fig. 1.1. Most of rim, both handles, much of body, and most of toe preserved; large parts of lower neck restored in plaster. I restore as "-nWVTS Graffito: nEEXAEKATETOPE77. etS e, X)oS xcxT?T?opesG[TTcc]p?e;]"(see discussion below). Small letters fairly carefully cut vertically down the neck and onto the shoulder. Deep strokes;letters tend to increase in size further along the graffito. Lang reads TTEEXas t?VT? etS ets Xo6e. She notes (1956, p. 12) that the jar measured the equivalent of 7 Attic choes. The use of epsilon as an acrophonic numeral for Et; is not securely paralleled among the graffiti studied here; however, the alternative of hELuseems unlikely with two epsilons (one-half and one-half) on such a formally inscribed graffito.Johnston questions Lang's reading of 7 choes by citing the use of E for 100 at Olbia. That graffito, however, seems to have been written on a sherd rather than an amphora (for similar use of multiple Es on an ostrakon, see Blonde 1989, p. 518, no. 170). The 8sxocTTeopsc, 14, when considered with the dropped aspirates implied by the epsilon abbreviationsearlier in this graffito, provides some limits to the possible dialect of the writer. Buck (1955, pp. 154-160) places -c&RopeSin his West Greek and Northwest Greek groups. Few members of these groups, however, drop the aspirate. Those that do include Delphi (at times, p. 156) and Elis (p. 159). The final, poorly preserved letter that I restore as a sigma introduces another significant area of debate concerning this graffito. Lang

MARK

34

L. LAWALL

illustrates the upper bar but interprets the graffito as 7 choes, 14 drachmas (see also Lang 1956, p. 13), with the drachmas implied but not abbreviatedor written. Johnston wonders whether drachmas or staters should be supplied as the unit of currency.The upper angle of the sigma supports the restoration of oTactypseS,either abbreviated or written out. The surface of the jar is worn below the preserved area of the graffito. Further along is an X, but this might not be a deliberate mark. Had oca-cYpS been written out, especially given the varying size of letters as preserved in this graffito, the word may have fit before the was inscribed. X. The X, however, could have been cut after ocYocxTpss the reading Given the poor preservation of the initial 2 of ocYrcTYpes, cannot be considered secure, but the mark interpreted here as the upper angle of a 2 encourages the reading of staters rather than the implied drachmas. Date ca. 440-430. 26 (P 11382). Volume and price. Mendean.

Fig. 17

Lang 1956, p. 10, no. 44; AgoraXXI, p. 76, He 1, pl. 42; Johnston 1996, p. 82, fig. 1.2. Preserves rim, most of neck, and both of the upper handle attachments. Graffiti on either side of the neck: A KKwith X in the A, and opposite, A2 (three-bar sigma); also a red-painted stroke behind one handle. The AX KKare incised with slightly heavier lines than the A2, whose strokes are quite light and sketchy. The AX KKgraffito resembles 27 from R 13:12, but that piece lacks the additional price mark. Lang (1956, p. 10, no. 44) reads as 10 choes, 2 kotylai with a price of 10 staters, making Mendean wine have the same 2 drachmas per chous price read elsewhere for Chian wine (see p. 16 above).Johnston (1996, p. 82) points out the ambiguities of this mark:"whose choes and whose staters?"Staters often appear on amphora price marks, and these foreign currencies may have been applied by wholesale shippers. Such marks seem too ambiguous for any role in facilitating retail commerce in the Agora. The stater notation here could indicate the tare weight of the precisely measuredjar (10 staters equal 20 emporic mnas, or just over 9 kg). The figure cannot be read as an indication of the weight of the jar's contents, since no contents likely in a volume of just over 10 choes (over 30 liters) would weigh only 20 mnas. The amphora neck is fairly short, but even so it is not any more closely datable than the general range of the contents of the well: ca. 440-425. WELL

R 13:12

R 13:12 is a well deposit east of R 13:4. The amphora forms are similar to those in R 13:4, but the finewares and plainwareswere not reexamined for the present study. Camp (1977, p. 220) suggests a date for the contents of the well in the third quarterof the 5th century.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

27 (P 30835). Volume. Mendean.

I\

Kft

27 Figure 5. Graffito from well R 13:12

35

Fig. 5

Complete neck, both handles, part of shoulder. Graffito on neck: A with X inside, followed by three kappas, two set above the third. Very light incisions, small carefully cut letters. Resinated. Very similar arrangement of letters as in 26, but the particularstyles seem a bit different (the range of allowable variation in such graffiti in terms of hands and letter forms is uncertain). A reading of 10 choes, 3 kotylai seems required.It seems possible that some Mendean (?) shipper marked his jars with specific capacity notes. It will be interesting to see how similar to 26 and 27 is the example from Kommos cited byJohnston (1996, p. 82, "agraffito similar as far as it is preserved"). Profile of the jar shows no significant difference from material in R 13:4, date ca. 425. CONSTRUCTION

FILLS

FOR THE

MINT

BUILDING

The pottery lots listed below are all from construction fills of the late-5thcentury Mint building (the topic of a forthcoming study by John Camp andJohn Kroll).The pottery in these fills continues into the last decade or so of the 5th century (Camp, pers. comm., June 1998), but the diagnostic amphora fragments are uniformly datable to the third quarterof the century. It seems possible that the rubbish used to fill under the Mint floor was the same that was also used to fill well R 13:4. 28

(P 3346). Volume. Chian C/3.

Fig. 6

Lot IIA 294A, section IIA, grid MF, MA/104, 105, construction fill under the floor of the Mint. Preserves only neck wall. Graffito 11111111. Lightly cut, with uneven length and spacing. choes is a Eight likely interpretation.The unevenness of the lines could be the result of either one hasty filling or measurement, or successive decantings. Date ca. 440-430. 29 (P 33415). Volume or weight? Possibly Solokha I.

Fig. 6

Lot PA 294A. Preserves neck fragment only; break at left of graffito. Graffito EEEE.Deeply but unevenly cut small letters. The two most likely interpretations of this graffito seem to be either 4 or more choes, reading the E for ?C5,or 4 or more hemichoes. If the E is read as an acrophonic numeral for sercthe units of measure could also be mnas, thereby making this a notation of net weight (probably only partially filling the vessel). Another reading for weight would involve reading the E as an alphabetic numeral, 5, and thereby reading 20 or more mnas. Poorly diagnostic, but other amphora material in this context is datable ca. 425 or slightly later.

MARK

36

L. LAWALL

IX X,H2 29 28

30 31

32b

32a

Figure6. GraffitifromMint constructionfill 30 (P 3420). Volume or price? Possibly northern Greek.

Fig. 6

Lot IIA 293, section VIA,grid M, MET/101, 102, construction fill under the floor of the Mint. Preserves lower part of neck and shoulder. Graffito X (or F-)II111. Cleanly cut lines breaking though the grayish surface of the sherd to the red core. Unfortunately the graffito is not sufficiently preserved to determine whether the initial sign refers to choes or drachmas. If choes, then we have 1 or more choes and 3 kotylai. If drachmas are indicated by the first sign, the subsequent verticals might indicate obols. Poorly diagnostic fragment; pre-425?

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

37

AMPHORAS

31 (P 33418). Unclassified graffito. Possibly Thasian.

Fig. 6

Lot IA 294A. Preserves neck and one handle; fairly low-slung handle. Graffiti on lower part of neck: MO I. Deeply cut large letters. First part of the word missing. Uncertain commercial significance. Poorly diagnostic fragment; pre-425? 32a-b

(P 33419, P 33417). Unclassified graffiti.

Fig. 6

Lot HIA294A. Lot includes two poorly diagnostic sherds, one (a) with EV graffito, another (b) with 0. Apart from their findspot, there is no reason to see these as commercial graffiti. WELL

R 13:1

Well R 13:1 is located north of R 13:4, and south of the later Stoa of Attalos. The well was closed at the end of the 5th century,with the range of dates for the bulk of the fill being ca. 430-400 B.C.(AgoraXII, p. 398). The amphoras in this deposit are very fragmentary; no complete vessels were ever assembled from these finds. The amphoras here show some development beyond the forms in R 13:4, and the best parallels occur in deposits closed ca. 410. None of the R 13:1 amphoras necessarily dates beyond ca. 410. 33 (P 9242). Volume. Possibly Chian straight-neck.

Fig. 7

Lang 1956, p. 4, no. 5. Preserves large portion of neck. Graffito on neck of four parallel strokes. Strokes of moderate depth. Resinated interior surface. Lang reads at least 4 choes. Form is unlikely to be much before ca. 425-410. 34 (P 9245). Volume. Probably Chian straight-neck.

Fig. 7

Lang 1956, p. 4, no. 6. Small bit of rim, bit of handle, and neck. Six parallel marks on the neck, closer to one handle. Lightly cut incisions. Lang reads as 5 or more choes. By my reading, 6 or more choes. The form need not be any later than the straight-neck Chian jars in R 13:4, ca. 425. 35 (P 9239). Volume. Mendean. Lang 1956, p. 7, no. 21, pl. 1. Preserves handle fragment only. Graffito on the outer face of the handle shaft, moving from top to bottom: three horizontal strokes, followed by a fourth on which

MARK

38

I

I

l

(

L. LAWALL

Il

34i

33

36 K

38 40

V 43

41

44

47

46 Figure 7. Graffiti from well R 13:1

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

39

are stacked three kappas,with a single larger kappa at the bottom of the preserved fragment. The fragment is broken at the top, leaving open the possibility of more strokes, and at the bottom. At this lower end there is enough uninscribed preserved surface that it is unlikely that there were further related markings. Very coarsely cut incisions. There are extra cuts over the middle of the stacked kappas as though for the correction of a mistake. Lang reads some amount greater than 3 choes and 4 kotylai. Despite the apparent mistake in incising the series of aligned kappas, the intention may have been as Lang reconstructs it, or perhaps (see 36) we are only meant to read the two unaltered kappas atop the vertical along with the last sign, to give 3 or more choes and 3 kotylai. The handle form could date anytime in the last third of the century, to ca. 410. 36 (P 9240). Volume. Mendean.

Fig. 7

Lang 1956, p. 7, no. 22. Preserves only the lower part of the neck and shoulder. Graffito on lower part of neck, just above shoulder: three kappas facing left one above the other, followed by three vertical strokes. Very coarsely cut marks. Resinated interior. Lang reads as 3 or more choes and 3 kotylai. It is not common to find retrogradeinscriptions among these late-5th-century amphora graffiti; it is possible that this graffito is meant to be read from above. This graffito and 35 are the only ones that I know with stacked kappas. The two fragments are of very similar fabric, the graffiti show a similar style of cutting, and the handle (35) clearly carries a mistake. Perhaps the handle belongs to the same jar as this neck, and the neck graffito offers a cleaner, correct version of the marking. The fragment is not sufficiently diagnostic to narrow the date beyond the general date for the contents of the deposit, ca. 410. 37 (P 9241, P 9253). Volume? Mendean. Lang 1956, p. 13, no. 59, pl. 3. Preserves lower neck and shoulder. Graffito on lower part of neck: N (or I I) IIHOOO. Coarsely incised letters. Lang reads four verticals before the H to give 4+ choes, /2chous, and 3 oxybapha (= 3/4 kotyle); however, she notes the problematic reading of the initial signs. The diagonal of the possible N could be an extraneous mark. Context date ca. 425-410. 38 (P 9244). Volume. Chian or Solokha II.

Fig. 7

Lang 1956, p. 11, no. 49. Preserves lower neck wall. Partiallypreserved TTwith an X inside, with two horizontal lines below. Wide, but not especially deep, strokes.The second leg of the TTis much shorter than the first. Resinated interior.

40

MARK

L.

LAWALL

Lang reads as "five choes or more"perhaps referring to the two horizontals below the TT.Interpretations of either 7 choes or 5 choes and 2 kotylai would take account of the lines below the TTX. Fragment is poorly diagnostic. The Solokha II type amphora is related in form very closely to the straight-neck Chian types. This is particularlytrue of the rim and upper parts of the handles. The differences lie in the lower parts of the handles, the shapes of the toe, and the somewhat darkerred-brown fabric of the Solokha II amphoras;for illustrations and discussion, see Doulgeri-Intzessiloglou and Garlan 1990 (identifying Peparethos as at least one production area for this type) and Mantsevich 1975. Context date ca. 425-410. 39 (P 9249). Volume. Possibly Solokha II. Lang 1956, p. 12, no. 54, pl. 3. Preserves lower neck and shoulder fragment. Graffito on lower part of the neck, close to the transition to the shoulder: TTXHIfollowed by a vertical stroke that could be one side of another H. Shorter second leg of TT.Moderately coarse but clean strokes (similar coarseness as in 46). Some resin preserved on interior. Lang reads "6 1/2choes and ?"The last vertical mark suggests a reading of 6.5 choes and 1 or more kotylai; however, the break in the sherd allows for the possibility of another H instead of a simple vertical stroke. The Solokha II attribution is based on the fact that the fabric here seems consistently redder than is common on Chian amphoras of the same time. The piece is not sufficiently preserved to give much indication of date. Context date ca. 425-410. 40 (P 9250). Volume. Possibly Chian.

Fig. 7

Lang 1956, p. 9, no. 33. Preserves lower neck. Graffito XXHII.Lightly cut incisions. Lang reads 2.5 or more choes and 1 or more kotylai on the basis of reading K after the H. In either case, we seem to have 2 or more choes, 1/2chous, and one or more fractions, probably kotylai. Context date ca. 425-410. 41 (P 9251). Volume. Mendean.

Fig. 7

Lang 1956, p. 10, no. 42. Preserves shoulder and small bit of the neck. Graffito over transition from neck to shoulder: TTXmonogram created by a horizontal line crossing the short second leg of the TT.Two incompletely preserved vertical strokes follow this monogram. Light incisions. Lang reads two 5-chous signs. If Lang intended the two freestanding verticals as the legs of the second IT,the X for this monogram

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

4I

is not preserved.This interpretation gives the second TTlegs of equal length, while the first TThas legs of different length. Such a difference in orthography seems unlikely, and most TTsamong the graffiti here have uneven leg lengths (cf. 42). If Lang intended the first freestanding vertical to be the right leg of the first TT,then only the left leg of the second TT1 is preserved.With so little indication of the second 5-chous sign, and given the common pattern among the graffiti studied here of single units following the 5-chous symbol, I suggest a reading of 7 choes: TTXII.A possible parallel for this arrangementof the TTXmonogram is offered by 53, unfortunately also incomplete. Lang's reading does create a volumetric notation for this jar that is in keeping with the two 10-chous notations on 26 and 27. Around 400 B.C., however, Mendean amphora capacities are known to have fallen closer to 7 choes; see Eiseman and Ridgway 1987, p. 52, table 7. Poorly diagnostic fragment, context date ca. 425-400. 42 (P 9252). Volume. Chian straight-neck. Lang 1956, p. 10, no. 43, pl. 2. Preserves small bit of neck wall. Graffito: a faintly preserved vertical stroke followed by X and four vertical strokes.The third vertical after the X has a short diagonal angling down to the right from the point where the stroke breaks off. This diagonal does not appear deliberate to me. Lang reads the graffito as a TTXmonogram (using the verticals on either side of the X, with equal leg lengths for the TT)followed by H1I.The H, if the diagonal stroke was deliberate, would have an unusually sloping crossbar.Without the H, but keeping the rest of Lang's reading, we have 8 choes, a common enough capacity for late-5th-century Chian jars.The initial 5-chous sign, however, is also problematic as it would require a TTwith equal leg lengths, which is rare among these graffiti. Reading a TTbefore the X would find a parallel in 96 and would give a reading of 6 choes, 4 kotylai. Poorly diagnostic fragment, context date ca. 425-410. 43 (SS 6918). Weight? Possibly Solokha II.

Fig. 7

Preserves small amount of rim and neck wall with small fourspoked wheel-stamp near rim. Graffito on neck: A followed by one vertical and two poorly preserved diagonal strokes. Light to moderately cut strokes. Resinated interior. Given that this vessel type is of roughly the same size as late-5thcentury Chian amphoras, a reading of 11 choes and 2 kotylai seems unlikely.Without explicit drachma signs, 11 drachmas, 2 obols, is also problematic. A tare weight of 11 mnas and two fractions,just over 5 kg, seems the most likely interpretation. Poorly diagnostic fragment, context date ca. 425-410.

42

MARK

L. LAWALL

44 (P 9248). Price mark. Probably Chian straight-neck.

Fig. 7

Lang 1956, p. 15, no. 65, pl. 3. Preserves neck wall only. Graffito TTA(monogram) followed by 1-. Fairly light but uneven incisions. Some resin preserved on interior. Lang reads as a price of 22 drachmaswith two deltas sharing a common side. The bottom line of this parallelogrammonogram of AA is not as clearly incised as the other lines, and I do not consider it a deliberate stroke. I read the numeral 50, which is quite commonly found among ceramic graffiti (e.g., Lang 1956, no. 68, pl. 3, and no. 99, pl. 5; Jefremov 1998, passim; Solomonik 1984, nos. 284-286). The price then should be read as 52 drachmas.This price seems high in comparison with other known late-5th-century wine prices,but the commodity referredto is not known, nor are enough wine prices known to say with any certainty what is a high price and what is not (as emphasized in Johnston 1996). Poorly diagnostic fragment, context date ca. 425-410. 45 (P 9247). Volume or weight? Mendean. Lang 1956, p. 5, no. 9, pi. 1. Preserves neck, rim, one upper bit of handle. Graffito on neck: the ends of seven horizontal marks followed six vertical marks arrangedin pairs of two. Fairly light incisions by but consistent depth throughout. Resinated interior surface. Lang reads 6 horizontals with the interpretation of 6 choes, 6 kotylai. The seventh horizontal is visible at the bottom of the stack in the photograph published by Lang. With the reading of a seventh short mark, 6 choes, 7 kotylai seems likely. As in other examples with more than 6 kotylai marked, the short marks could record small amounts poured out. Mendean jars from the later Porticello shipwreck measured between roughly 6 and 7 Attic choes (Eiseman and Ridgway 1987, pp. 51-52). While the volumetric reading works reasonably well, a weight reading would give either 67 or 76 mnas depending on which unit is assigned to each set of strokes;the latter seems more likely since the horizontals precede the verticals.This weight would be the gross weight of a jar weighing ca. 5.5 kg with 24.5 liters of wine. The short neck suggests that this piece may not be any later than the R 13:4 pieces, ca. 425; but the short neck does continue into the last quarterof the 5th century (see Brashinskiy 1976 and Lawall 1995, pp. 121-122). 46 (P 9243). Volume, possible summation. Probably Chian straightneck. Fig. 7 Lang 1956, p. 6, no. 18. Preserves lower neck wall and bit of shoulder. Graffito on neck: three vertical strokes on the first line, TTon the line below. Moderately deep and clean strokes. Second leg of T shorter than first. Resinated interior.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

43

may, however, represent a summaLang reads as 8 choes. The T11 tion of the vertical tallies above, as is more clearly the case in other examples where the larger number is on a separate line (22, 23, possibly 24). Poorly diagnostic fragment, context date ca. 425-410. 47 (P 9246). Unclassified graffito. Possibly Chian.

Fig. 7

Lang 1956, p. 8, no. 23. Preserves neck wall only. Graffito K lying horizontally.There may be a horizontal mark below the K, but its identification as a stroke is uncertain. Deep clean strokes. Resinated interior. Lang includes this piece ("forcompleteness' sake")with other kotylai graffiti, but the isolation of this letter makes any interpretation very difficult. This is not a letter that is frequently encountered alone on amphoras. Poorly diagnostic fragment, context date ca. 425-410. WELL

S i6:i

Well S 16:1 was discovered in R. R. Holloway's excavations in the Kolletis House garden and lies roughly 50 m southeast of R 13:4, on the east side of the Panathenaic Way. This fill is generally dated ca. 425-400 and is noted for the many amphora fragments found (Holloway 1966, pp. 83-84; AgoraXII, p. 398). Closer study of the amphorasin the fill, however,places them with R 13:1 as no later than ca. 410 B.C. 48 (P 27513). Volume. Chian straight-neck.

Fig. 16

AgoraXXI, p. 59, Ha 3, pl. 32. Preserves part of rim, one complete handle, neck, shoulder with tightly rounded outer edge. Graffito on neck: four vertical lines of moderate depth. Resinated interior surface. Lang (AgoraXXI, p. 59, Ha 3) reads four units "measuredas they were poured in," but she notes that Chian jars held more than 4 choes, so "this might not be a permanent record of the total capacity but a temporary note about a smaller quantity put in (or taken out)."The parenthetical suggestion would certainly fit the kinds of activities described below that have concentrated these graffiti in the area. The tall neck and sharply turned shoulder of this jar clearly place it after the jars in R 13:4 and probably near the end of the 5th century. 49 (P 27515). Volume. Chian straight-neck. AgoraXXI, p. 59, Ha 4, pl. 32. Preserves complete rim, one handle, and part of neck and shoulder. The markings here are quite irregularand may not all be deliberate strokes. Near the base of the neck are seven vertical strokes, lightly incised; they fit well with the graffiti discussed here and are surely deliber-

44

MARK

L. LAWALL

ate. Above this, startingjust below the rim, are the following marks: 1) a relativelylarge A lying on one side immediately under the rim; 2) two very irregularhorizontal marks below the A; 3) then, under the horizontals, two very short and accidental-looking verticals; and 4) under these verticals, two more, somewhat more regular,horizontals. The longer strokes here seem more deliberate and more likely to be part of an intended graffito than do the very short marks. Resinated interior surface. Lang reads a temporary accounting of 7 choes (the upper markings, leaving aside the A) followed by the more permanent, formal 7 tallies below. The A is not a common abbreviationin this group of graffiti and is best left aside as unrelated.The short strokes in the upper group, however, seem too informal and accidental to be connected to the tallying process. What remain, therefore, are the three horizontals and the seven, more regular,verticals. Perhaps the two sets of marks pertain to two or more activities in the use of the jar, e.g., a partial decanting (or filling) and a more complete decanting (or filling). The thick rim of this jar, both rounded outward and thickened inward, places this jar fairly early in the Chian straight-neck series, ca. 430-420. 50 (P 27516). Volume. Chian straight-neck.

Fig. 8

Preserves much of rim, upper parts of handle, neck, and small bit of shoulder. Graffiti low on the neck near the shoulder: five stacked horizontal lines. The upper two lines are less heavily and carefully cut than those below. Resinated interior. This variation in cutting may suggest multiple occasions for making the marks. Five choes may be indicated (though more may have been incised originally). If these five were all that were inscribed, then this only accounts for a portion of the total possible capacity of the jar.These strokes seem likely to record goods decanted over a period of time. The neck is fairly short but could still fall anywhere between the constraints of the starting date for the type and the likely closing date of the context. Date ca. 430-410. 51 (P 27519). Volume. Chian straight-neck.

Fig. 8

Preserves part of rim and neck. Graffito on neck: stack of four horizontal lines, with no trace of further lines above or below. Some end cleanly and others taper more gradually. The differences among the strokes may imply separate occasions for their cutting. Assuming that the jar originally held at least 7 choes, this partial accounting may attest to material taken out over a period of time. Context date ca. 425-410.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

54 53

51 50

55 57

58

61

59

62 Figure 8. Graffiti from well S 16:1

45

46

MARK

L.

LAWALL

52 (P 27525). Volume. Mendean. AgoraXXI, p. 59, Ha 6, pl. 32. Preserves small bit of rim, both handles, much of neck, large part of shoulder. The graffiti appear around the transition area between neck and shoulder. On one side of the jar is the marking TTX(monogram) XXXH; below these on the same shoulder is the abbreviationAE; then in the thumbprint at the base of one handle is an incised X (or K); finally, on the other side of the neck are 8 vertical tallies. The vertical tallies are very lightly and unevenly incised; the tally of abbreviatedchoes is heavily cut as is the letter in the thumbprint. The AE is cut very lightly and sketchily.The TTXmonogram is unusual for lacking the second leg of the TTaltogether. Heavily resinated interior. Lang reads an informal tally of the vertical strokes-8 choesformalized in the more complex notation and made more precise with the additional hemichous. It is equally possible that the deeply cut TTXXXXH may be the permanent total count against which decanted choes have been recorded as simple, irregularvertical strokes. There is no clear indication of whether the letter in the thumbprint is meant to go with either of the other marks.The AE seems likely to be an owner's graffito. The shortness of the neck and very rounded shoulders could place this jar very early among those seen in well R 13:4, ca. 440. 53 (P 27543). Volume. Possibly northern Greek.

Fig. 8

Preserves neck wall sherd only. Possible traces of resin inside. Graffito TTX(monogram). Probably created by crossing second of leg the TTwith a horizontal bar,but it is possible that the second leg of the TTis not at all preserved and the X is freestanding under the TT.Small and carefully cut graffito, very much like that of 65 in well 0 19:4. A volumetric notation of 5 (and probably more) choes seems most likely. Possible Mendean or northern Greek amphora, but the identification by fabric alone is uncertain. Context date ca. 425-410. 54 (P 27418). E-graffito. Solokha I.

Fig. 8

Holloway 1966, p. 84, pl. 28:h (jar only). Complete jar preserved. Graffito E on shoulder. Moderately deep strokes.The size of the letter and its prominent place on the shoulder make it quite visible. The E here may replace the H found in other graffiti (cf. 75, 76); however, if that H is read as indicating 1 hydria instead of the clumsier then E cannot abbreviatethe unit of measurement. [?T-pYTYco, YJLOOU I type tends to be quite large, so it is difficult the Solokha Furthermore, to see these single letters as referringto a half-metretes; both 54 and 66

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

47

were cut on Solokha I jars. Such jars are rarelyresinated and seem, therefore, to have been used at least in part as oil jars. E abbreviating ScXOLov (oil) is a possible interpretation of these marks. At this point the dating of the jar cannot be narrowedbeyond the context date, ca. 425-410, but it does seem fairly early in this range. 55 (P 27526). M-graffito. Solokha I.

Figs. 8, 21

Preserves complete rim, neck, one complete handle, part of other handle, part of shoulder. Graffito M at middle of neck on one side, irregularstroke on opposite side of neck of uncertain significance. M-graffito cutting is heavy with clean ends to each stroke. See discussions at 8 and 20 above. Context date ca. 425-410. 56 (P 27517). Summation graffito and price dipinto. Chian straightneck. AgoraXXI, p. 59, Ha 5, pl. 32. Preserves neck, rim, both handles, part of shoulder. Resinated interior surface. Graffito on lower part of neck: THEN11 II 1.These marks-both the letters and the vertical tallies-are heavily incised, with careful stops to each stroke. The marks seem to have been applied at one time. The second leg of the T11 is shorter than the first. The first three letters are covered a wide vertical stripe of red paint. Further around the partly by behind one neck, starting handle, is a price mark dipinto also in red: ATTI-.Below the main graffito is a sketchily incised A. Although the price mark itself does not cover the main graffito, the vertical red stripe may have been applied at the same time as the price mark dipinto, thereby making the price mark necessarily later than the graffito. Lang suggests that the dipinto attests to the original price in reference to the common 7-8 choes size of these amphoras (with the common 2 drachmas per chous), but this reading does not fully explain the graffito and its functional relationship to the dipinto. Various explanations are possible. The TTENmay summarize the five tallies from a decanting operation of 5 choes; then the jar was refilled; the vertical stripe cancels the graffiti; and the new contents' price was set by the price mark. More in line with Lang's belief that the graffito refers to fillings, the jar may have been filled with 5 choes and then had its price set at 16 drachmas.As in other cases examined here, the sketchy single letter, A, need not be part of the numerical notation and may refer to an owner of the jar or to its contents at some point in the use of the jar. The form of the jar could fall anywhere between the constraints of the starting date for the type and the likely closing date of the context. Date ca. 430-410.

MARK

48

57 (P 27544). Summation. Unattributed type.

L. LAWALL

Fig. 8

Preserves only the transition area from neck to shoulder. Resinated interior. Graffito on the shoulder at the base of the neck: IIIITT.Lightly incised, uneven strokes. Multiple occasions of cutting may be indicated here. The verticals could refer to 4 choes decanted, then summed up by the TTwhen the fifth was poured out; the TTcould also represent a total from which the four tallies are subtracted.Finally, a total filling of 9 choes could be described. Retrograde graffiti, such as this one, are not commonly encountered in the pieces published here. Not a diagnostic fragment; context date ca. 425-410. 58 (P 27518). Unclassified graffito. Chian straight-neck.

Fig. 8

Preserves rim and part of neck. Resinated interior. Graffito on neck: TT?Light uneven cutting; unusually tall and narrowletter. It is uncertain how deliberate this marking was or whether it is of a numerical nature. Context date ca. 425-410. 59 (P 27520). Unclassified graffito. Mendean.

Fig. 8

Preserves part of rim, neck, and one handle. Graffito on one side of the neck: TTTnearly touching one another. The lTT graffito is unevenly cut, with shallow to moderate depth, but the T is clearly cut as a letter separatefrom the TT. The meaning of this abbreviationis quite uncertain. Five talents would bring this abbreviationclosest to epigraphic parallels,but it is hard to imagine placing 5 talents in this amphora.Any other reading would require assuming more vagueness as to the units involved than is typical for these amphora graffiti (5+4?, 5 and 1/4?,5 and 1/3?). The height of the extant portion of this neck places it late in the 5th century,well after the closing of R 13:4. Date ca. 410. 60 (P 27527). Unclassified graffito. Solokha I. Complete rim, part of neck, one complete handle, part of other. Graffito on neck: V?Lightly cut. Uncertain if these are deliberate strokes. Context date ca. 425-410. 61 (P 27529). Unclassified graffito. Unknown type.

Fig. 8

Complete rim, neck, handles, part of shoulder. Graffito AK or NK monogram. Graffito sits behind handle; moderate depth of strokes, carefully cut. Of the markings listed here, only 66 has a similarly hidden position near or behind the handle. Uncertain if either of these are commercial graffiti.

WINE

GRAFFITI,

SELLING,

AND

REUSE

OF

49

AMPHORAS

Dark redwarejar with very heavy handles. The type develops through the 5th century,but it is not certain if the date of this piece in particularcan be narrowed further than the context date, ca. 425-410. 62 (P 27540, P 27541). Unclassified graffito.

Fig. 8

Unclassified body sherds. Graffito X A. There is room for a letter between the two preserved, and a large portion is missing at just that point. The scantiest trace of this possible letter is present. Light to moderate depth of cutting, and these letters are larger than most of the graffiti seen here. Resinated interior. Such size and carelessness of cutting is common among possible owner's marks.When such letters appear on fragments with numerical graffiti, the numerical notation is more carefully cut (e.g., 52, 93, and 98). Context date ca. 425-410. WELL

0

19:4

Well 0 19:4 is roughly 60 m south of South Stoa I and the Mint, on the northern slope of the Areopagus; ca. 140 m southwest and upslope of R 13:4. The fill is generally dated to ca. 425-400. None of the graffiti fragments is necessarily later than ca. 425, but there is only one that is sufficiently preservedto allow an independent suggestion of its date. Other amphoras in this deposit are datable to the end of the 5th century (see AgoraXII, p. 396; Lawall 1995, pp. 332-333). 63 (P 12635). Volume. Possibly northern Greek.

Fig. 9

Lang 1956, p. 9, no. 36. Preserves lower neck fragment. Partiallypreserved graffito XXXXX. Heavy incisions. Lang reads 4 or more choes. There is a trace of a fifth X at the end of the preserved graffito. The very micaceous fabric suggests the possibility of a northern Greek origin for the jar, but it is not an otherwise diagnostic fragment. Context date ca. 425-400. 64 (P 12657). Volume. Unidentified type. Lang 1956, p. 9, no. 29, pl. 2. Preserves rim, upper part of one handle, neck, and part of shoulder. Graffito on neck: poorly preserved tips of perhaps three horizontal lines (see below) followed by XEIII.Uneven incision of strokes, from light to moderate. Lang reads "atleast 1 chous,1/2 chous and 3 kotylai."If all of the three horizontals before the X are from deliberate cutting, then we have another E or at least three horizontal lines preceding the X; either mark would be difficult to explain. The middle stroke of these

MARK

50

L.

LAWALL

63

65

66

Figure9. Graffitifromwell 0 19:4

three horizontals, however, is the best preserved and may be the only deliberate mark. If so, another X is unlikely since the extant X is composed of diagonals rather than a vertical and a horizontal. One possibility is that the line is from the horizontal crossbarof a TTX monogram, 5-chous mark (see 41 and 53). Six and a half choes and 3 kotylai would match common late-5th-century amphora capacities. The use of E instead of H recalls 4 and, perhaps, 13. The white slip of the fragment, the short neck, and the very narrow,outwardly rounded rim all seem unusual for a Chian amphora; otherwise 64 might have been identified as such. The low height of the neck, however, makes possible a date of ca. 425. 65 (P 12962). Volume. Unidentified type.

Fig. 9

Lang 1956, p. 11, no. 46, pl. 2. Preserves only shoulder and bit of neck. Graffito TTH(as monogram) HKK.Lightly incised markings though very small and carefullycut. Shorter second leg of TT.Break in the sherd allows for the possibility that there were more Ks following those preserved. Lang reads 6 hydriae and 2 kotylai, questioning the idea of a 5-hemichous measure. Reading hydria here, however, is problematic if one accepts Hesychius's definition of the hydria (s.v. oipco(aCa) as one half-metretes. In this case, the graffito would represent 36 choes and 2 kotylai (even 12-chous amphoras are very rare in the 5th century B.C.). Six hemichoes would, by contrast, at least fit in the jar.

The fabric of the vessel is dark red-brown, very hard and compact, but any attribution of the place of manufactureof this jar would be uncertain. Context date ca. 425-400.

GRAFFITI,

WINE

SELLING,

AND

REUSE

66 (P 12658). E-graffito. Early Solokha I.

OF

5I

AMPHORAS

Fig. 9

Preserves rim, neck, one handle, and part of shoulder. Graffito E (retrograde)on lower part of neck wall nearly behind the handle. Moderate depth of strokes. For the interpretation of single E graffiti, see 54. Here, however, the somewhat hidden letter may have a different significance than that proposed for 54, where the letter is much more visibly inscribed. The form of the jar is best paralleled in well N 7:3, closed ca. 440 B.C.,so this piece may be earlier than others in this group. WELL

R 13:5

Well R 13:5 is immediately adjacent to and west of R 13:4, but the fill is somewhat later, with material datable as late as 390 (AgoraXII, p. 398). The amphora material in this fill most closely resembles that of 0 19:4 and the later material in Q15:2, so a date at or just beyond the end of the 5th century seems quite likely. 67 (P 11375). Volume. Unclassified type.

Fig. 10

Lang 1956, p. 12, no. 55. Preserves transition from neck to shoulder only. Prefiring marking AX. The break of the sherd allows for further letters following those preserved, and there does appear to be a bit of a vertical line preserved. Small, very carefully cut letters. Lang reads as either 11 or 12 choes (or more). Long-standing criticisms of the theory that stamps guaranteed the capacity of the jar are 1) that the guarantee was applied before the vessel was fired and, presumably,shrunk in the process, and 2) that an unfired jar, especially one with only partly dried clay,would be difficult to measure.This is the only apparentlyvolumetric marking we have that must have been applied before drying was complete (stamps may be volumetric marks in a manner of speaking, but this has yet to be proven). Perhaps the mark was applied to remind the potter or someone else of the intendedvolume. Unlike many other graffiti published here, this was clearly not inscribed in Athens. Fairly hard, compact fabric. Context date ca. 425-390. 68 (P 2067). Volume. Solokha I.

Fig. 10

Lang 1956, p. 11, no. 50, pl. 3. Preserves lower neck and bit of shoulder. Graffito on the shoulder at the base of the neck: 1TXHand a further symbol, possibly A, H, or K.The letters are fairly small, unevenly incised, and lightly to heavily cut. The last letter is covered with further scratchings, perhaps intended to efface it. Further to the side, fully on the shoulder is a second graffito, AEON, in very faint, small, neat letters. as 6.5 choes and 2 kotylai. Lang notes that the Lang reads VTXHKK, are not kappas clearly legible. The secondary cuttings make it difficult to know if there are even two letters intended after the H. It is possible that

MARK

52

67

68

L. LAWALL

69 Figure10. Graffitifromwell R 13:5

the last letters and scratches attest to a mistake, and that the 6.5 choes are all that were intended. The second graffito presumably represents the name of an owner or someone otherwise associated with the jar at some point in its use. The amphora type is identified on account of the sharpness of the neck-shoulder join, which is not seen to this degree on other late-5thcentury amphora types of similar fabric. Context date ca. 425-390. 69 (P 33414). Price mark. Chian straight-neck.

Fig. 10

Cited by Papadopoulos and Paspalas 1999, p. 177, note 78. Complete rim, part of neck, one complete handle and upper part of second, bit of shoulder;with circularkantharos stamp near rim. Graffito running from neck to partly behind handle: AATTr-Fwith one further vertical stroke preserved. Light strokes, as is the case in other price graffiti published here. Slight traces of resin inside. The graffito can be read as 27 drachmas and 1 obol or as 28 drachmas (the lower part of the last vertical, where the short horizontal to create the last F-may have been inscribed, is not preserved). A 7- or 8chous jar would make the price of wine contained in the jar, if that was its contents, between 3 and 4 drachmas per chous. The same stamp appears on 11 and SS 14080 from R 11:3 (closed ca. 420-410). The stamp might date early in the last quarterof the 5th century,but there is no certainty as to how long the image was used on Chian stamps. The form of this particularamphora seems somewhat more developed than those in R 13:4. Date early in the last quarterof the 5th century. WELL (Q15:2

Well Q15:2 is just west of the Panathenaic Way, north of the Mint, ca. 40 m southwest of R 13:4. The fill contained some debris possibly attributable to the Mint operations and two bronze official measures, perhaps from the metronomoi offices in South Stoa I (Thompson 1955, pp. 6970; Camp 1977, p. 218), but the fill also contained large quantities of animal bone, both from food and from the bone-working industry (Lynn Snyder, pers. comm., October 1999). The lower parts of this fill include late-5th-century material comparable to finds in R 13:1 and S 16:1; the higher parts of the fill more closely resemble finds in 0 19:4 and R 13:5 and may be datable into the first decade of the 4th century.The lower part of the fill also contained an unusual series of five plainware oinochoai with comic scenes (Crosby 1955).

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

53

70 (P 26388). Volume. Possibly Chian straight-neck.

Fig. 11

Preserves shoulder fragment only. Graffito: possibly HX or simply two verticals followed by a horizontal. Very little of the third stroke is preserved and perhaps only the upper part of the first and second. Large letters of light to moderate depth. The latter reading is better paralleled in this collection. It is unusual among the graffiti discussed here to see a smaller unit (hemichous) before a larger unit (chous). Alternatively the H might stand for one hydria (six choes), followed by a seventh chous; this would be a common size for a late-5th-century Chian amphora. Poorly diagnostic, context date ca. 425-390. 71 (P 26365). Volume? Possibly northern Greek.

Fig. 11

Preserves neck wall only. Base of neck wall graffito: I1111.Uneven spacing and depth of line, with lines trailing off unevenly. Such uneven and irregulartallies tend to be either volumetric, here indicating 4+ choes, or to refer to the weight of the jar.The former seems more likely here, with an incomplete tally accounting for only part of the total capacity of the jar. Context date ca. 425-390. 72 (P 26349). Volume. Mendean.

Figs. 11, 20

Preserves part of rim, neck, and one handle. Graffito at the base of the neck: the upper parts of two vertical strokes and K. Moderately deep strokes, with the K set slightly higher than the other marks. Resinated interior. On analogy with other Mendean jars (26, 27) there may have been another K below the one preserved. Perhaps 2 choes and 1 kotyle; however, the verticals could also be the tops of an H as a hemichous sign, after a tally of choes (not preserved), or it might stand for hydria (half-metretes). This tally (at least as far as it is preserved) only accounts for a portion of the total capacity of the jar; if no other markings were ever inscribed, an accounting of dispensed wine seems a more likely interpretation than a complete refilling of the vessel. Height of neck suggests a date of ca. 400 or later. 73 (P 26381). Volume. Chian straight-neck.

Fig. 11

Preserves neck sherd only. Graffito X followed by a second possible X, incompletely preserved. Light cuttings. Resinated interior. Uncertain which side of the sherd points toward the rim, so this could be either the start or the end of a multiple choes count. Context date ca. 425-390.

MARK

54

L. LAWALL

72 70

f

71

'5 Ii

LA4

74

73

75

77

78

76

i/

80 79

if81 81

82 Figure 11. Graffiti from well Q15:2

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

55

AMPHORAS

74 (P 24194). Weight? Unattributed type.

Fig. 11

Neck fragment only, and fairly worn. Graffito: one vertical stroke present along the break at the left side of the sherd, followed by AAA III1.Moderate to light incision, uneven spacing. The trace of a letter before the first A and the presence of three deltas raise the comparison with 81. The first vertical may be the righthand leg of a TTwhose upper bar is not carefullyjoined to this right leg. A lTA monogram would create a reading of 84 for this graffito. 84 mnas could record the gross weight of the amphora and some heavy contents. Context date ca. 425-390, but the wear on the sherd could indicate that it was used even earlier. 75 (P 23946). H-graffito. Northern Greek.

Fig. 11

Preserves complete rim, handles, neck, and much of shoulder. Graffito H fully on the shoulder. Light, but fairly neat, incision. A half-metretes, or hydria, as Lang (1956, p. 11, no. 46) interpreted another H-based graffito (65), seems unlikely in this case at least as a measure of the complete jar.The jar, however, could certainly hold 6 choes (a half-metretes or a hydria), partially filled. Alternatively, the mark could refer to 8 choes as in certain summation marks seen in these graffiti. Without any accompanying tallies, however, it is unclear if this letter has commercial significance. Height of neck suggests a date of ca. 400 or slightly later. 76 (P 26355). H-graffito. Unattributed type.

Fig. 11

Preserves neck wall only. Graffito on middle part of neck: H. Smaller letter than 75, but similarly neat. For interpretation see 75. Context date ca. 425-390. 77 (P 23990). M-graffito. Solokha I.

Fig. 11

Preserves lower part of neck and part of shoulder. Graffito M on shoulder just below the neck. Similar size of letter as on 20 and 21, but the style is slightly different; moderately deep strokes. Sharp turn at neck suggests Solokha I identification. Context date ca. 425-390. 78 (P 24003). M-graffito. Possibly Samian, Solokha I form.

Fig. 11

Shoulder fragment only. Graffito on shoulder: M. Heavily incised letter with extensions beyond the apices. Dark brown fabric with much fine mica, possibly a Samian jar of the Solokha I form (see discussion by Grace 1971, pp. 67 and 78, note 68). Context date ca. 425-390.

56

MARK

79 (P 23979). Weight or price? Mendean.

L. LAWALL

Fig. 11

Preserves fragment of neck and shoulder only. Graffito at the transition from neck to shoulder: two vertical strokes followed by an upward pointing arrow.A preceding diagonal line does not appearto be part of the graffito and may not even be a deliberate mark. Very heavy,thickly cut lines, with cleanly rendered endpoints. Traces of possible resin on interior. The upward pointing arrow,an "arrowdelta,"is interpreted as indicating 10 byJohnston (1979, pp. 29-30; and 1982, pp. 208209). The graffito here may indicate 12 units, which is a bit low for a mna weight notation for a late-5th-century Mendean amphora, and certainly high for a capacity measure.This could be a vague price notation. The upward pointing arrow mark occurs on finewares fairly often, but to my knowledge this is the only occurrence on amphora graffiti from the Agora. A series of arrow deltas all sharing the same line appearson a fineware graffito published by Yailenko (1980, p. 92), where he refers to earlier publications of such graffiti on Ionian cups (rejecting the earlier numerical interpretation, preferringinstead to see the stacked arrowsas a tree or three-barbed arrowhead). Context date ca. 425-390. 80 (P 26378). Summation or weight? Chian straight-neck.

Fig. 11

Preserves part of shoulder, larger part of neck, but no rim. The top edge of a graffito on shoulder near where the neck rises: four vertical strokes followed by A or A (only the tops of each symbol preserved);preceded by a much longer diagonal that does not appearto be part of the same graffito.To the extent that they are prreserved,the lines of the graffito are cut lightly and unevenly. Resinated interior. If the last letter is a A, then the graffito could read 4 choes with the alphabetic 4 (A) as a summary mark off to the side. Alternatively, the graffito may be a retrogradetally of 14 mnas, a possible empty weight for a jar of this type; however, there are few necessarily retrogradegraffiti among the graffiti discussed here. The piece is darkly fired but seems more like Chian than Solokha II. Context date ca. 425-390. 81 (P 23949). Numerical graffito. Possibly Solokha I.

Fig. 11

Preserves neck and shoulder only. Graffito on neck, well above shoulder:very uncertain monogram, perhaps combining P and P or HTand X, followed by AAA. The letters are fairly small and moderately to deeply cut. The horizontals of the deltas form a continuous line, which was perhaps cut first to carry the deltas. does not suggest 5 choes The reading of the monogram as TTIX The closest parallel for the here on account of the deltas following.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

57

AMPHORAS

monogram is a counting-table from Naxos (IG XII.5 99) on which a very similar monogram signifies 500 (Tod 1911-1912, p. 116). Either a combination of P and P or a TTXmonogram gives a very large number: 580 or 530. There is no unit of measure that can fit such a large number to this jar alone, so perhaps this graffiti should be seen as having been applied prior to importation, indicating something about the production or exportation stages. The sharp transition from shoulder to neck suggests a Solokha I form amphora, but the jar is not datable more narrowlythan to the late 5th or early 4th centuries. 82 (P 23942). Unclassified numerical notation. Possibly Solokha I, though very uncertain. Fig. 11 Preserves neck and shoulder only. Graffito on shoulder near neck: TPIH followed by an upright stroke at a slight angle, perhaps the vertical line of a K missing the lower part of the letter. Various interpretations are possible: tptaxocx;for 30 or a division of 30; Tpvlxoa7OS6 for (a duty of?) 1/30th;or TpYIqxotcalo; o for "30-days" as an adjective (e.g., wine that is 30 days old). The eta replacing the alpha could indicate an Ionian writer. The fabric of this fragment is very micaceous and very bright orange on the interior surface,yellower tan on the exterior.The fabric very closely resembles that of a later-4th-century water jar (P 30802, which also happens to carry a numerical graffito: A above HKKKK). There is a possibility then that this is not an amphora but a local (?), large water jar. Context date ca. 425-390. 83 (P 23991). Owner's graffiti? Possibly Chian straight-neck.

Fig. 12

Preserves shoulder only. Graffito on shoulder near neck: - followed by two shorter strokes angling up toward each other, and a bit of a third stroke angling up away from these first two. Perhaps to be restored as -AN. Clean incisions with moderate depth. Resinated interior. Eocvappears as a possible owner's mark on two other late5th-century pots: see AgoraXXI, p. 37, F 106, P 24774, a lekane rim of the late 5th century; and F 109, P 3736, a black-glaze bowl of the second half of the 5th century.The letter forms involved are all similar. Eoave6vis an adjective sometimes used to describe honey (AgoraXXI, p. 80, He 36, 4th century A.C.). While the M-graffiti here may label some amphoras as honey amphoras, none of those so marked had resinated interiors as this sherd does. For this reason, it seems unlikely that this graffito refers to "yellow honey." Context date ca. 425-390.

MARK

58

83

L. LAWALL

84

85 86

87

Figure 12. Graffiti from well Q_15:2

84 (P 23968). Unclassified graffito. Chian straight-neck.

Fig. 12

Preserves part of rim, much of neck, one upper handle segment. Graffito KA on middle of neck wall between the handles. Light to moderate, uneven incisions; fairly large sloppy letters. Resinated interior. Various possibilities exist for the meaning of this abbreviation. washed out or washed over, might indicate that this jar KAXuoCSrC, has been rinsed out and is ready to be refilled. One example of a KA graffito, from Samothrace (Samothrace11.2, p. 100, no. 246), is inscribed upside down near the toe of the jar-a likely place to note that a jar, now perhaps draining, has been washed out. On the other hand, the Delian temple inventories mention waterproofing jars, possibly for oil, using wax (IG XI.2 219A, line 40 "... OsaioC xo7c0 H1111Tv ,jlTiW av As^LOL GTs[yveoavc0v]Xr

xct xpCOt xXDAaocv-nF- . . .").

Finally, in style of cutting, this graffito tends to resemble other appar-

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

59

ently noncommercial marks that are adjuncts to commercial markings; so this may be an owner's mark. Other examples of this graffito are published by Solomonik (1984, nos. 172 and 175). As is expectable in a late-Sth-century deposit, the neck of this piece is taller than most in R 13:4. Date ca. 400. 85 (P 26387). Unclassified graffito. Uncertain type.

Fig. 12

Preserves only shoulder and part of neck. Graffito on shoulder near the rise of the neck; possibly reading KA, KF,or rr. Large, lightly incised letters. KA graffito is better preserved in 84. In terms of the possible readit may be significant that the previous ing x;u-cezos or xX6AwoavTo, had a resinated interior whereas this one does not. Frrmay refer example to 3+3 choes, but such a use of alphabetic numerals is otherwise unparalleled. KF could stand for 23, perhaps a record of the weight of the jar's contents-but only of a partial filling of the jar.The sloppiness of the letters is paralleled by 84. Context date ca. 425-390. 86 (P 26343). Unclassified graffito. Possibly Mendean.

Fig. 12

Preserves bit of neck and shoulder only. Graffito on the transition from neck to shoulder: TTI (possibly H) on first line, V on second line beneath the I. Uneven, fairly light incisions. Resinated interior. Five and a half choes (and more) is a possible reading of the first line, but the second line is of uncertain significance (perhaps not even a deliberate graffito?). Context date ca. 425-390. 87 (P 33422). Unclassified graffito. Unidentified type.

Fig. 12

Fragment preserves roughly half of rim and part of neck wall. Very similar fabric and rim form in another set of fragments from the same container; however, the rim here does not actuallyjoin the larger set of fragments. Graffito on neck just below the rim: M followed by three vertical lines. All very lightly incised. All of the cuttings after the M seem to be the upper parts of further letters or other symbols. This piece was kept out of the M/ME series since the graffito is more complex than those in that series. Without more of the symbols following the M, it is unclear what was intended here. Context date ca. 425-390. 88 (P 26389) Black-glaze fragment. AgoraXXI, p. 39, noted with F 131. Body fragment only. Graffito: [.. .]lixo;.

60

MARK

L.

LAWALL

89 (P 23821). Black-glaze oinochoe. AgoraXXI, p. 39, F 131, pl. 16. Complete profile preserved but very fragmentary;includes complete handle with a small bit of the rim and shoulder. 'Av8pcixo ei0 aLxcxouo;sand [.. .]XXos.The owner'sgraffito runs along the handle; the [...]XAos is on the shoulder. 90 (P 23835) Lekane. AgoraXXI, p. 39, F 132, pl. 17. Two-thirds of rim preserved, one handle, upper parts of the bowl. [&1]xoaos'A[v8]pcoxo [eis'] around the top surface of the rim; uses omicron instead of the omega in 89 above. DEPOSITS

WITH

LATE-5TH-CENTURY

WELL

SINGLE

EXAMPLES

NUMERICAL

OF GRAFFITI

C I9:9

Well C 19:9 is in the northeast room of House K in the Industrial Quarter (Young 1951, pp. 242-243, pl. 77:a). Much of the fill is dated to the late 5th century with a supplementary fill, topping up the settled initial fill, laid down in the early 4th century (suggested by a red-figure oinochoe, Agora XXX, p. 245, no. 726, pl. 77, P 18556, ca. 400). The late-5thcentury fill seems unlikely to date much beyond ca. 410. The well and its contents are thought to stem from domestic use of the building before its lower floor became a marble-worker'sbuilding in the early 4th century. From C 19:5b, floor fills associated with the phase of marbleworking, comes a water jar amphora (P 18609) with graffito TTK(monogram) KKH (AgoraXII, no. 1463). The fill and this piece could date to the late 5th century,but the closest published parallelfor the jar is AgoraXII, no. 1462, from a context starting ca. 375. It seems more likely that this small water jar also dates well into the 4th century. 91 (P 18989). Summation. Chian C/3. Lang 1956, p. 6, no. 17, pi. 1. Preserves upper part of jar only, less than half of rim, missing parts of neck and shoulder. Graffito TTfollowed by three verticals; another five verticals lower down on the shoulder on the same side of the amphora. Moderate weight of cuttings; all strokes appear fairly carefully cut. Lang reads either 8 or 7 choes and 5 kotylai, depending upon whether two or three strokes are understood to follow the T7.There is room for even more verticals after the TTin terms of the spacing of the extant strokes. In keeping with the summation practice seen elsewhere among these graffiti, the TTcould summarize the five tallies, then further decanted or added choes could have been added after the summation. Date 440-430/25.

WINE

GRAFFITI,

SELLING,

AND

REUSE

OF

ze

/

v

P 92

94

n

xxe T

X-r

El 96

99

Figure 13. Graffiti from several late5th-century deposits

97

AMPHORAS

6i

MARK

62

92 (P 18620). Chian straight-neck.

L. LAWALL

Fig. 13

AgoraXXI, p. 36, F 97, pl. 15. Preserves shoulder fragment only. Possible owner's graffito-'Avpop[axou]-on shoulder with tops of letters pointing away from the neck. Large letters, moderate to heavy cutting. For other references to Andriskos, see 89-90. Cf. 88. Form of shoulder and fabric suggest straight-neck Chian form datable anytime after ca. 430/425 to ca. 390. OTHER

DEPOSITS

CLOSED

CA.

430-400

B.C.

93 (P 16444). Volume? Chian straight-neck. G 18:1, well in the domestic area on the north slope of the Areopagus, closed ca. 400 B.C.; see AgoraXII, p. 391. Lang 1956, p. 8, no. 27, pl. 2. Preserves shoulder and part of neck. Graffiti: (1) HF monogram graffito on neck and (2) AZ (with Z written as H on its side) IIon the shoulder. Shoulder letters are moderate in depth and cleanly incised; the break allows for the possibility of one more vertical stroke between the Z and the II. Lang reads (2) as 10.5 choes and 2 kotylai. This reading, however, does not account for the use of Z instead of H (unless this is simply a mistake on the part of the inscriber). Z read as 7, plus 3 more units (choes?) could add up to the 10 indicated by the delta, though this would entail switching between numeral systems, alphabetic to acrophonic, in the same graffito (this may occur elsewhere; see 4 and 24). There is no necessary connection between the neck graffito (1) and the shoulder mark (2). Date ca. 425 or later. 94 (P 18923). Volume (with possible summary notation or weight?). Chian straight-neck. Fig. 13 B 19:7, household context in the Industrial Quarter southwest of the Agora proper, closed ca. 410. Lang 1956, p. 16, no. 66. Preserves neck and small bit of shoulder. Graffito at transition between neck and shoulder: A IIII followed a by horizontal stroke breaking off at the right side. Fairly light, uneven strokes. Lang reads as perhaps 14 drachmas or 10 choes 4 kotylai. The horizontal mark, largely ignored by Lang as "merelya concluding dash,"might shift the reading to any of the following: 14 drachmas, 1 obol (perhaps least likely since it lacks the standard F for drachma); or 10 choes, 4 kotylai, and 1 or more fraction; or 14 mnas and 1 fraction (a weight drachma?)as the empty weight of the jar. A further possibility is that the A records the amount present (4 choes), against

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

63

AMPHORAS

which the tallies are recorded as that quantity is emptied (followed by the concluding dash). Context date late 5th century. 95 (P 8432). Volume. Possibly Chian straight-neck. Well B 13:6, at the southernmost tip of the slope of the Kolonos Agoraios; late-5th- or early-4th-century context. Lang 1956, p. 9, no. 32, pl. 2. Preserves shoulder fragment only. Graffito: first line XX, second line HK.Neat but uneven strokes, ranging from lightly to moderately deep. Lang reads 2.5+ choes and 1+ kotylai. Context date late 5th or early 4th century. 96 (P 17010). Volume. Mendean.

Fig. 13

Section NN, grid 110-111/KE-KH fill notebook; p. 2375, at the southernmost tip of the excavated domestic and industrial area, late 5th century. Lang 1956, p. 12, no. 57. Preserves rim, one handle, neck, and shoulder. Graffito at the transition from neck to shoulder: TTXXXT and on second line TE. Heavily cut into the friable Mendean fabric. TE is cut more lightly and neatly as though from a different stage in the marking process. The legs of the TTare of roughly equal heightthe only time this is observed with certainty in the graffiti here. Scant traces of resin on the interior. Lang suggests either 81/3choes with another quarter-chous added to complete the filling process, or 81/4choes with the T in the first line explained by the TE in the second. In another example (AE, no. 52), the abbreviationon the lower line is in a different hand than the numerical graffito whereas 96 has nearly the same style of marking on both lines. The T is quite ambiguous, and the need to clarify such ambiguity certainly supports the quarter-chous interpretation. The neck is somewhat taller than those from R 13:4, suggesting a date in the last quarterof the 5th century. 97 (P 15053). Volume. Chian straight-neck.

Fig. 13

N 21:7, rubbish pit south of well 0 19:4, contents dated ca. 450420 (AgoraXII, p. 395), with latest datable piece an oinochoe in the manner of the Meidias Painter (AgoraXXX, p. 236, no. 657, P 15034, dated there to 410-400). Lang 1956, p. 13, no. 61. Preserves rim, neck, both handles, shoulder. I1111. Lightly Graffito near middle of neck: TP (monogram) 11111 incised. Lang (1956, p. 13, no. 61) reads as "nine tryblia of the choinix variety (i.e., three choes)." While I doubt that this is the complete

64

MARK

L.

LAWALL

capacity of the vessel, Lang's reading could be taken as a partial filling. The height of the neck places this piece late in the date range provided by the finewares in the pit, and very close to the later date suggested by Moore (AgoraXXX, p. 236, no. 657). Date last quarter 5th century. 98 (P 6126). Weight? Possible summation. Northern Greek. E 14:14, a cutting in the bedrock, disturbed in antiquity; perhaps originally a grave.The cutting included 98 as well as a nearly complete red-figure hydria in the manner of the Kleophon Painter (AgoraXXX, p. 229, no. 603, P 6053, dated ca. 430). Lang 1956, p. 15, no. 64, pl. 3. Preserves rim, one handle, neck and shoulder. Graffito ATT1111ATT(monogram) on one side of the shoulder. Another TTgraffito appears on the other side of the jar.The ATTis cut with clean but light strokes, the following tallies are even lighter and quite irregular,the final ATTis also very light and sketchy.The TTon the opposite site is heavily cut in sharp contrast to the more complex markings. Lang (1956, p. 15, no. 64) reads another ATTafter the four tallies, interpreting as 15 drachmas,with 4 obols "perhapsadded as the price of the jar."The addition of the cost of the jar is not otherwise attested. The lightness and sloppiness of the cutting of the verticals encourage reading them as four single units summed up by the TTonce the fifth was added. In no instance is a securely interpreted price mark built up in such informal tallies. The four sketchy verticals could be counting up to or down from either of the two Us; however, they seem most likely connected to the ATTjust preceding the strokes. Reading an empty weight of the jar at 15 mnas seems therefore more probable than a price reading. As for the reading of the monogram, the cuttings are in a style different from the preceding AT7,and the resulting monogram seems better read as ATT,paralleled in its sketchiness by other possible owner's marks. Date of form ca. 420, later than the date suggested for the hydria. 99 (P 33422). Possibly Chian straight-neck.

Fig. 13

Lot W 161, "ca.9/Ar," fill of the 5th century B.C. south of South Stoa I and the Mint. Preserves lower part of neck and shoulder. Graffito A. Lightly incised. Break removes area for further possible signs.

This could be simply someone's initial; however, given the prevalence of numerical graffiti in this region, it seems possible that this piece belongs in this study. Date late 5th century.

GRAFFITI,

WINE

SELLING,

INTERPRETING

AND

REUSE

OF

AMPHORAS

65

THE GRAFFITI

"The graffiti and dipinti were not meant for posterity, and the idea of our trying to understand them today would no doubt be wryly amusing to their writers."78This comment by David Jordan, reviewing Agora XXI, stands as a suitable caution and necessary addition to Eugene Vanderpool's comment cited in the preface to AgoraXXI: "It is easy to read if you know what it says."79

If the readings are open to various possibilities, as discussed earlier in this article and in the catalogue, then the broader understanding of the graffiti in terms of Athenian and Aegean social history is not made any easier. For this reason it seems appropriateto bring as much evidence as possible to bear on the interpretationof the roles of these graffiti in Aegean trade. The process followed below for interpreting the graffiti moves through three broad steps. First, I consider the graffiti in terms of various contexts: dates, findspots, amphora types involved, and the nature of the incisions themselves. Second, this contextual evidence and the contents of the graffiti themselves inform interpretation of how the graffiti may have functioned at various stages in the amphoras'use. Such stages include activity at the port, refillingjars from local wholesalers, selling wine from the amphoras, and selling other commodities as part of the reuse of the amphoras. In addition, at this stage I consider the situations in which weighing was likely to have been a part of an amphora'suse, why price marks are so rare among these graffiti, and why only in R 13:4 is there such a preponderance of graffiti on Chian amphoras.Third, and finally,I attempt to place these practices leading to the graffiti in the broadercontext of Athenian and Aegean trade and political history in the 5th century B.C. CONTEXTS

FOR ANALYSIS

DATES

78. Jordan1978, p. 92. 79. Agora XXI, p. v.

Despite the many finds of numerical and other commercial graffiti in the Agora, the 5th-century occurrences of these graffiti are surprisingly restricted in chronology. Only eight examples are datable before ca. 440/435 B.C. One early graffito (1) appearson a Chian amphora datable to ca. 500, and a later Chian fragment (7) is datable by its form to ca. 440-430. Three pieces (2, 5, and 6) are on poorly diagnostic sherds, for which either what little there is of the form or the closing date of the findspot suggests a date before ca. 440. One other mid-century graffito (3) was found on a wellpreserved northern Greek amphora of the second quarter of the century. Finally,a mushroom-rimmed, Solokha type I amphorafragment (8), probably from the southern Aegean, should be dated, on the basis of the jar's form and its findspot, not later than ca. 430. Indeed, seventy-four of the eighty-two known examples of commercial graffiti (just over 90%) are datable to the last three decades of the 5th century. The chronology of the period of frequent amphora marking is indicated both by the closing dates of deposits containing multiple examples of graffiti and by dates attributableto the amphoras themselves. Previous

66

MARK

L.

LAWALL

i studies of the finewares in deposits containing multiple graffiti, along with

my currentstudy of amphora fragments from these and many other Agora deposits and stratified fills, support the following closing dates: ca. 425 for well R 13:4; ca. 410 for wells R 13:1 and S 16:1; and ca. 400 or into the first decade of the 4th century for construction fills for the Mint and wells 0 19:4, R 13:5, and Q15:2.80 The amphoras, too, in these deposits show differences in form attributableto their morphological changes over time. R 13:4 contained fragments both of the latest stage of the Chian bulging neck form (producedca. 440-430) and earlyexamplesof the Chian straightneck form (produced ca. 430-400) (Figs. 14-15).81The later deposits contained taller and later straight-neck Chian jars (Fig. 16) with very few, if any,examples of the earlierbulging neck type. Likewise, R 13:4 contained well-preserved examples of a quite roundedjar with a short neck from the region of Mende (Figs. 17, 18), while R 13:1, R 13:5, and the other later wells contained the more angularform with taller neck of the late 5th and early 4th centuries (Figs. 19, 20).82The later wells also contained more examples of the Solokha I, mushroom-rimmed form (Fig. 21), which be80. Forthe evidenceforthe datesandreferences to published contentsof thesedeposits,see the deposit indices in AgoraIV,Agora XII, AgoraXXI, AgoraXXIII, and AgoraXXX.The datessuggestedhere,

wheretheydifferfromthe datespublishedin thesesummaries, arebased

on my currentresearch. 81. Grace 1979a, fig. 44, far right, illustratesthe latest configurationof the bulging neck form;fig. 45 shows the subsequentstraight-neckform. Lawall (1995, pp. 91-93 and 99-103) discussesthe developmentand its chronology.

Figure 14 (left). Chian C/3 amphora 25 (P 2366) from well R 13:4 Figure 15 (center).Chian early straight-neck amphora 19 (P 2372) from well R 13:4 Figure 16 (upper right). Chian late straight-neck amphora 48 (P 27513) from well S 16:1 Figure 17 (lower right). Mendean amphora 26 (P 11382) from well R 13:4

82. See Brashinskiy1976 and Eiseman and Ridgway1987 for summariesof the chronological sequenceof Mendean jars after ca. 450 B.C.; with specific reference to the Mendean chronologyof the period in question, see Lawall 1998b.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

67

AMPHORAS

I

I

I

i

Figure 18 (left). Mendean amphora P 2377, without commercial graffito, from well R 13:4 Figure 19 (center).Mendean amphora P 23864, without commercial graffito, from well Q_15:2. Ht. 0.651 m. Figure 20 (upperright). Mendean amphora 72 (P 26349) from well Q_15:2 Figure 21 (lower right). Solokha I amphora 55 (P 27526) from well S 16:1

gins to appearin Athens around 450 and is only common at the very end of the century.The graffiti, therefore, appearpredominantly on amphoras whose dates spanned the last three decades of the 5th century and entered the 4th century. As is clear from Table 1, the types of graffiti are not restricted within shorter periods in the last decades of the 5th century; all types appear in both early and late contexts. In all deposits, the volumetric notations of either simple tallies or tallies including abbreviatedunits arethe most common. Summations, weight notations, price marks, and the isolated abbreviations E/H and M/ME are far less common but do appearboth in early and later deposits. Specific idiosyncrasies among the R 13:4 graffiti also recur in later wells. For example, the practice of occasionally combining multiple-letter abbreviations with single letters or tallies is seen in both R 13:4 (25) and S 16:1 (56). In addition, the practice of aligning small tallied units along a vertical line appearstwice in R 13:4 (16, 18) and reappears in R 13:1 (35, 36, and possibly 45). Finally, the use of E instead of H for "half"may occur in R 13:4 (possibly with 13) and, with more certainty, in the later deposit 0 19:4 (64). Even if there may not have been continuity in the sellers' identities from ca. 440 to ca. 400, these similarities of practice over time suggest some familiaritywith the earlier activity. Some practices in the graffiti tend to appearin the later deposits. The most noticeable of these is the method of rendering 5 choes using rTwith a short horizontal line cutting the right leg of the letter. This practice is seen in both R 13:1 and S 16:1 (both closed ca. 410): see 41, possibly 42,

68

MARK

TABLE 1. DISTRIBUTION NEAR R 13:4 R 13:4 Total Volumetric Simple tally Abbreviated units Summation Combined with price Weight Price E/H, M/ME

10 (1) 5 2 3 (1) 2 5 3 2

OF AMPHORA

R 13:12

Mint

1

3 (2) 1

1

2 (2)

-

-

-

-

-

L.

GRAFFITI

R 13:1 12 3 8 1 -

1 (1) 1 (1) -

LAWALL

TYPES IN DEPOSITS

S 16:1

(1) (1) (1) (1)

2 (1) 1 -

019:4

Q 15:2

2

5 (2) 1 (1) 3 1 (1)

3

9 4 2

3

2 1

-

1

-1

2

R 13:5

2

1-

3 (2) 1 (1) 4

Parentheticalnumbersindicateuncertainattributionsin each group (see catalogue).Deposit Q15:2 includes two numericalgraffiti,81 and 82, that do not fall into the graffititypes listed here.This column, therefore,includes two fewer graffitithan the column for Q_15:2in Table 2. The two markings"combinedwith price"in R 13:4 include abbreviatedunits for volume and, for this reason,are also listed under"abbreviatedunits" for the deposit.

52 (modifying the idea by leaving out the second leg of the TT),and possibly 53. This may be seen as a development in the practice of placing the chous designation, X, fully inside the acrophonic numeral (as is seen in 26 and 27 with 10 choes, and in 38, which may be contemporarywith the 5chous monogram pieces, or may be slightly earlier),but there is no indication here of a sharp break in practice. FINDSPOTS

Most of these commercial graffiti are found in the southeast corner of the Agora excavations (Figs. 1-2). Deposits R 13:4 (18 examples), R 13:12 (1), Mint construction fill (3), R 13:1 (15), S 16:1 (10), R 13:5 (3), Q(15:2 (13) account for sixty-three of the seventy-four known late-5th-century examples (85%). Wells R 13:1, 4, 5, and 12 all lie just southeast of the junction between the Panathenaic Way and the road leading west toward the later Roman Agora. S 16:1 is also east of the Panathenaic Way, about 50 m southeast of the R 13 wells. Q_15:2 lies just west of the Panathenaic Way, north of the Mint building, and 40 m southwest of the R 13 wells. The fill under the Mint itself lies just across the Panathenaic Way from the R 13 area.The distances between these deposits are sufficiently small to suggest a very limited area as the source of the debris. A more distant deposit, along the southern side of the Agora, 0 19:4, contributes another four examples.Eight other late-5th- or early-4th-century graffiti arefound scattered across the Agora. In contrastto this concentrationof the commercialgraffitiin the southeast part of the Agora, other classes of graffiti (e.g., lists, owner'smarkings, or other names) are much more generally scattered throughout deposits of the Agora region.83The concentration of numerical graffiti appears even more striking in view of the fact that many other large deposits of amphoras of the same period, located elsewhere in the Agora excavations, contained few or no examples of commercial graffiti.84 This uniqueness of the spatial and temporal patterning of the 5th-

83. AgoraXXI, passim. 84. For example, deposits A-B 2122:1, B 13:5, G 16:1, M 20:3, and B 15:1, all of the late 5th century, have revealed no numerical graffiti (among inventoried and noninventoried sherds); deposits C 19:9, G 18:1, U 13:1 (early 4th century, see discussion below), and B 13:6 have very few examples of numerical graffiti.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

69

century amphora graffiti in the Agora is further emphasized by the rarity of similar graffiti outside the Agora. One very early example is found on an amphora of ca. 500 B.C. in the Kerameikos.85Similar graffiti have been published from sites along the north coast of the Black Sea, but even there, where many graffiti have been published, numerical graffiti on amphoras are not common.86Although such graffiti were clearly used outside Athens, the Athenian finds are noteworthy for the large number of examples found in one area of one site. This particular region of the Agora has been connected with wine selling in earlierpublications,especiallyby LucyTalcott andT. Leslie Shear Jr.87The unusually dense concentration of graffiti here allows various further details to be proposed as to the commercial history of this part of the Agora. Well R 13:4 was the first of the major graffiti-bearing deposits to be published. The well's many amphoras, drinking cups, a few mixing bowls, limited range of cookwares,and the many examples of graffiti on amphora fragments prompted Talcott to propose the existence of a tavern, which "flourishednear the borders of the Agora in the years around 440. Apparently it met with some disaster,which caused its abandonment and discarding of its paraphernaliasomewhere about 430."88The well's contents were the only significant evidence for such a building until excavation under the Library of Pantainos in 1970-1974 revealed modest structures datable to the end of the 5th century,continuing through the 3rd century B.C.89The better-preserved structures,mostly of the 4th century and later, lie east of R 13:4, but other, possibly 5th-century foundations appeared closer to the well (Fig. 2).90 With so little architecturalevidence or occupational debris immediately associated with R 13:4, it is difficult to distinguish between a substantive break in the wineshop's activity and periodic cleaning up of the area'strash. Susan Rotroff and John Oakley drew attention to an earthquake in 425 B.C.,which Thucydides (3.89) described in some detail, and they reasonably connect this disaster with a large number of fills around the Agora whose closing dates might fall near 425. They include R 13:4 in the list of earthquakedeposits.91The earthquake clearly created a consid85. Kerameikos IX, p. 173, E 14, with fig. 47; readsXXXXXXXXXKKKKK<

an Habove.The amphorais datableto

the early5th centuryand is likely to have come from northernGreece. Johnston 1990, p. 52, note 30, no. 129, p. 53, fig. 12 is of a similardate, perhapsslightly earlier,and northern Greek. 86. For the rarityof amphora graffitiof this sort outside Athens, see collectionsby Babinovet al. 1978, Solomonik 1984, Tolstoi 1953, and Brashinskiy1984. Particularlystriking is the scarcity(one graffito out of fiftyfive published)of numericalgraffiti

from excavationsat Cape Zyuk on the Sea of Azov. The site is quite small but included an extensivedeposit of amphoraspreciselycontemporarywith those carryingso many of the Agora amphoragraffiti;see Maslennikov 1987 for the graffitiand Abramov and Maslennikov1991 for the amphoras.A possible exceptionto the rarityof volumetricgraffitiis Gordion (see Roller 1987), though only one of the publishedvolumetricmarksis from a Greek amphorapossibly of the 5th centuryB.C. (Roller 1987, p. 67, 3B-27, fig. 49, a Mendean shoulder,broadlydatablebetween ca.

400 and 350 B.C.);the vast majority are

local(?)pithoi. 87. Talcott 1935; Shear 1975. 88. Talcott 1935, p. 497. 89. See Shear 1975, pp. 346-361 for a reporton the Classicalbuildings under the Libraryof Pantainos.

90.The better-preserved 5th- and 4th-century structuresare illustratedin Shear 1975, fig. 5. Davidson (1997a, p. 56) mentions the better-preserved structures,claiming that they resemble what a tavernshould look like; they are, however,simply rectangularrooms. 91. Rotroff and Oakley 1992, pp. 53-57.

70

MARK

L.

LAWALL

erable amount of debris,92but it does not seem to have stopped the practicesof wine selling involving graffiti.Instead, closer study of the graffiti and amphora types present in both R 13:4 and the other nearby wells indicates that the break in graffiti-writing practice occurs somewhat later. Changes in practice in the region seem to begin with the filling of well Q15:2. Like the other wells, Q15:2 contained many numerical graffiti. The amphora finds in this deposit differ, however, from those in the other graffiti-bearing wells by including more numerous Corcyran and Lesbian amphoras, two types that are rare in late-5th-century Agora deposits.93One possible explanation for this anomaly is that Q_15:2 may have been filled shortly after the arrivalof new shipments of Corcyran and Lesbian jars. These jars may have been discarded before normal distribution activities reduced their numbers near Q_15:2 to better reflect their general presence in the city's wine supply ca. 400 B.C.A sudden clearance of the shop (or shops), with many jars still on hand, may have resulted from a business failure or other catastrophe. On the other hand, there are reasons to conclude that this well was filled gradually,perhaps even over a decade or more. Preliminary study of masses of animal bone found throughout the fill of this well revealed no evidence of weathering or gnawing by scavengers, which would be expected of debris that accumulated above ground before being deposited. There also seems to be far too much bone material to be the result of a single occasion, so the trash must have been deposited periodically.94The length of time involved may be indicated by the amphora finds. Excavation of the lower depths of the well brought up fragments whose best parallelsoccur in deposits closed ca. 410. The higher levels produced fragments paralleled at the very end of the 5th and the beginning of the 4th century.95The other wells in this area do seem to have involved single mass fillings, and this difference in deposition alone might explain the anomalous presence of Lesbian and Corcyran amphoras. Alternatively, the apparentdifficulty in obtaining such jars as indicated by their rarityin late-5th-century Athens may account for their being rarelydiscarded, and the larger number of examples in Q_15:2 may represent some (unknown and short-lived) change in practice. The situation in Q_15:2 is the first indication of a change in the R 13 wineshop area.Slightly later,more certain evidence of a change in practice appearsin the finds from well U 13:1, to the east of the R 13 wells (Fig. 2, Table 2).96 Like its neighbors, U 13:1 was full of amphora sherds, and these are datable to the early years of the 4th century (slightly later than Q_15:2). Around 390 B.C., a bedrock collapse of the walls ended the use of 92. Debris that seems contemporarywith the materialin R 13:4, i.e., "earthquakedebris,"forms the vast majorityof the fill under the Mint and under South Stoa I, both buildings with constructiondates later in the 5th century. 93. For Lesbianjars, see Clinkenbeard 1982; for Corcyran/Corinthian

B, see Koehler 1978 and 1992. For the rarityof these jars in late-5th-century Athens, see Lawall 1995, p. 291. 94. Lynn Snyder,pers. comm., October 1999. I am very gratefulfor Snyder'scomments on this material, Both Grace andTalcott, in studyingthe pottery from this well, concludedthat it was a single dumped fill. Without the

faunalevidence I probablywould not have thought to reconsiderthe amphorafragmentsfor the possibility of a longer period of deposition. 95. This argumentand the material involvedwill be presentedin full in a futurework. 96. For the excavationof U 13:1 see Shear 1975, pp. 355-356.

WINE

GRAFFITI,

TABLE 2. ESTIMATED BY TYPE

NUMBERS

R 13:4 Chian SolokhaI (South Aegean) Solokha II (Peparethan) Northern Greek Mendean Outpointing Rim

26/17 (2)* 2 14 21/1

Corinthian A Corinthian B Lesbian Unknown

SELLING,

OF AMPHORAS

R 13:12

R 13:1

S 16:1

6 3

17/8 (1) 6

8/5 7/2

8 4/1

3/3 (1) 19 8/5 2

62/1 14/1

-2

1

1 -

18

2

6

AND

REUSE

OF

AND MARKED

019:4

R 13:5

3 7/1

7/1 11/1 25 12 -

-

3 16/1 8 5 1

3 2 9/1

7 3 4/2

5 1 14/1

4

AMPHORAS

71

AMPHORAS

Q 15:2 20/3 38/4 (1)

U13:1 25 52 6

9 74/2 40/2 15

24/1 22 22/1

22 10 30/2

1 12 16

1

*The first figurerepresentsthe total numberof amphoras;the numberfollowing the slash representsthe amphoraswith commercialgraffiti;and the parentheticalnumberindicatesuncertainattributionsof amphoratype.

97. Shear (1975, pp. 349-350 and

355-356) describesthis sequenceof events. His date of ca. 390 for the bedrock collapsethat puts the well out of use, which he based on the black-glaze

fragmentsin the well, is supportedby the differencesbetween the amphora

U 13:1 as a source of water, and the well was filled with amphoras and other pottery and animal bones until perhaps 380.97In its large quantities of amphoras and its location, U 13:1 resembles the other deposits discussed here. What sets U 13:1 apart from these earlier deposits, however, is the rarity of numerical graffiti. One Thasian jar, of the ca. 180 vessels represented in the fill, carried a numerical graffito: ATTfollowed by perhaps another letter or mark.98A smudgy black dipinto, TT12,appears on the neck of another jar.99This marking recalls the earlier graffito 19 from R 13:4, TT112, so it may be a numerical dipinto. On the whole, the sudden shift away from marking numerical graffiti as seen in U 13:1 should indicate a hiatus in wine selling in the R 13 area.The large number of amphoras in U 13:1, however, seems to indicate the brevity of this interruption and the subsequentcontinuation of wine selling, though with graffiti marking the amphoras not nearly so often. Numerical graffiti on amphoras only appearsporadicallyin later deposits of the 4th century and the Hellenistic period.

types in Q15:2 and those in U 13:1.

For example,U 13:1 containedthe new, cap-style toes on Chian amphoras;

these are absentfrom Q_15:2.Hence the filling of U 13:1 should have begun afterthe filling of Q15:2. 98. P 30693. 99. P 30714. 100. For Chian amphorasof this period, see Grace 1979a, figs. 4445; for the northern Greek types, see Lawall 1997 and Garlan 1988; for Peparethan, see Doulgeri-

Intzessiloglouand Garlan 1990; for South Aegean, see Kantzia 1994; for the Solokha I and II type designations, see Mantsevich 1975. All types are also discussed in Lawall 1995.

AMPHORA

TYP E S

Despite the chronological and topographical limits of these frequent finds of amphora graffiti in the Agora, there is considerably less restriction in the range of late-5th-century amphora types carrying the graffiti. Chian, Mendean, other northern Greek, Solokha I (South Aegean), Solokha II (Peparethan),and other unidentified types are all marked after firing with Table 2 summarizes estimates of the numbers of amphoras such graffiti.100 of specific types represented by the fragments found in the wells that included numerical graffiti. Table 3 shows the distribution of graffiti types over these amphora types. Two points of context that will be helpful in later explaining these numerical graffiti arise from these tables. First, there is very little direct correlationbetween type of numerical graffiti and type of amphora carrying the graffiti.There are,however, certain amphora types that tend not to

MARK

72

TABLE 3. DISTRIBUTION TYPES Volume

OF GRAFFITI

Weight

Chian

25 (5)

7 (1)

Solokha I (South Aegean) Solokha II (Peparethan) Northern Greek Mendean

2 (1) 2 (1) 4 (4) 10 3

1 (1) 1 (1) 2 2 (2) 2

Unknown

L. LAWALL

TYPES ACROSS

Price 5 (1)

MAJOR AMPHORA

E/H

M/ME

Combined

2 (2)

2

1

-

1

-

-

7 (2)

4

2 1 (1) 2 (1) -

Summation

-

-

3 (2) 1

-

2

of graffitior amphora Parenthetical numbersindicateuncertain attributions type.Not includedin thistableare81, 82, and99,whichdo not fit intothegraffititypeslistedhere.

carrynumerical graffiti: Corinthian types A and B (the latter also, or even primarily,produced along the Adriatic coast of Greece) and Lesbian.101A second point to take from these tables is that there is some correlation between the frequency of an amphora type and whether or not it tends to carry graffiti: the more common amphora types are more likely to carry graffiti. In the Mint construction fills, not included in these tables since they are a different sort of deposit, the graffiti appear on the common Chian, northern Greek, and Solokha I (South Aegean) amphora types. It may be, however, that with more amphoras of these types present to be marked,the number of marksthey carrysimply reflects their greaternumbers overall. In deposit R 13:4, however,the dominance of marked Chian jars cannot be explained in terms of the generally large number of Chian fragments in the deposit.102Northern Greek fragments are also very common, yet they carry only one graffito. In later deposits, the distribution of the graffiti better reflects the frequency of each amphora type. While earlier studies of these graffitihad perhapsoveremphasizedthe frequencyof marking Chianjarsoverall,the overrepresentationof markedChian jarsin R 13:4 requiresfurther explanation in terms of the practices resulting in the graffiti (see below). A further exception to the general pattern that graffiti types are spread over various amphora types is the case of the M/ME and E marks. The tendency for these marks from the late-Sth-century Agora to occur on Solokha I amphoras, often on or near the neck, and rarely in association with a resinated interior was noted above (pp. 17-19). Readings of meli and elaion for these marks, respectively,fit these associations between the use of the marks and the characteristicsof the amphoras.This consistency of amphora-type context for these graffiti encourages the view that they refer neither to people's names (i.e., as owner's marks) nor, in most cases, to metretes or half-metretes. Other markingsmore securelyassociatedwith volume are not as restricted in type of amphora. Other amphora markings, such as stamps, prefiring incisions, and painted marks, are more closely associated with specific amphora types.103 Such type-specific markings are best attributed to actions at the point of exportation. In the case of the numerical graffiti from Athens, however,

101. For the Corinthiantypes, see

Koehler1978and1992;cf.Whitbread 1995andPreka-Alexandri 1992.For the Lesbianamphoras, see Clinkenbeard1982.See note65 abovefora withsuch CorinthianB amphora graffitifromelsewherein theAgora. 102.It shouldbe notedas a preface to this discussionof the Chian am-

phorasin R 13:4thatthereis some doubt,as notedin the catalogueentries, as to whetherallof the piecesare,in havea fact,Chian.Chianamphoras of fabricand widerangeof appearances a dauntingrangeof morphological details.Furthermore,other securely

non-Chianamphoras carrysimilarrims

with fabricsthat could be arguedto be close to Chian in appearance.Nevertheless, despite the fact that my identification of some of these jars as Chian is somewhat uncertain,it seems important to addressissues raisedby the possibilitythat such a large portion of Chian amphorasin R 13:4 aremarked.

103.Forexample,fornorthern Greekmarkingsof the 5th century,see Lawall 1997, pp. 118-120.

GRAFFITI,

WINE

AND

SELLING,

REUSE

OF

AMPHORAS

73

the fact that amphoras from many different sources carry similar graffiti supports the assertion that the graffiti were applied at or near the spot where they were found. Had they been applied before or during shipment, it is difficult to imagine how such a concentration would appear in one area of the Agora and not simply wherever large deposits of amphoras are found. STYLE

AND

CLARITY

OF THE

GRAFFITI

The M/ME graffiti bring a further pattern to consideration beyond their appearance on Solokha I amphoras and nonresinated amphoras. These graffiti are always very carefully or deeply cut. Other single letters or isolated abbreviations,at least among those studied here, may or may not be so carefully cut. Again, some unique meaning seems to pertain to the M/ ME graffiti, distinguishing them from volumetric marks and hence militating against a reading of metretes for these abbreviations. Among the types of numerical graffiti there is no apparentcorrelation between the weight or clarity of cuttings and the type of marking. Capacity marks, whether using simple numerical signs or abbreviated units of measure,may be cut very cleanly and carefully,though more often they are lightly and somewhat sloppily cut. Markings on friable amphora fabrics are often cut quite deeply-an understandablestrategy for legibility-but even this is not a consistent pattern. Graffiti interpreted here as weight and price marks are also inconsistent in terms of the style and clarity of the graffiti. SUMMARY

In sum, the numerical and other proposed commercial graffiti are best attributed, spatially,to activity near Agora grid square R 13, in the southeast corner of the Agora excavations,and, temporally,to activity of the late 5th century B.C., from ca. 435 to 400. The intensity of the practice seems to end in the early 4th century,soon after which time, perhaps after a brief hiatus of activity, the area does continue its wine-selling function. The marks appear on a wide range of amphora types and, perhaps, especially on those that are more commonly imported to Athens in the late 5th century.Most of the graffiti types show variation in the clarity and style of the marking. Among the graffiti, only the nonnumerical, but still likely commercial, M/ME graffiti are of restricted context in terms of graffiti style and associated amphora type. These must, therefore, derive from some activity that is different from that which resulted in the more common numerical notations. GRAFFITI

AND

PRACTICE

The search for explanations for why these graffiti appear in the southeast corner of the Agora in the late 5th century B.C.involves two lines of inquiry.First, what practices related to selling wine in late-5th-century Athens created the graffiti? Second, what broader social, economic, or political phenomena at that time encouragedthe adjustmentsto practiceresulting

MARK

74

L.

LAWALL

in the apparenttemporal and spatial concentration of the graffiti?I begin with the practices, the explanations for which are explored below. The geographic restriction of the graffiti and the rarity of any direct relationships between amphora and graffiti types must be accounted for when considering the practices that prompted the graffiti. The distributional evidence is particularlyat odds with the currentexplanation of practices resulting in the graffiti-a way of confirming the capacity of amphoraswhen they were refilled for personal use. Following such an explanation we would expect a wider spread of numerical graffiti wherever amphoras appear.This is clearly not the case. To propose alternativepractices, I turn to literary and representationalevidence for wine selling for possibly relevant, graffiti-generating activities. At the same time, however, there are other activities not attested in ancient sources that we might imagine as part of the operation of a wineshop. THE

PORT

There is only scant textual evidence relating to the specific activity of bulk purchasesat Peiraieusor any other merchants'port. One point that may be drawn from literary and papyrological sources is that the concern portside was primarily with numbers of jars and not the specific quantity of contents of any one jar.104 Had there been such a concern, an individual jar would have had to be emptied for the specific volume of its contents to be checked. Then the original liquid would have been poured back in, the jar resealed, and the shipment taken away.After all that, the one or few measuredjars could not certify the accuracyof the others. The assumption on the parts of the buyers and sellers must have been that roughly the expected quantity of liquid would be present in any given jar (of a known type, and whether that jar was a half- or third-size jar, as such amphoras clearly existed).105It would not be in the seller'sinterest to gain a reputation as a merchant who sold half-empty jars. For the most part, the numerical graffiti indicating volumetric measures or even weight measures seem unlikely to have been applied during bulk purchases at the port. Had such graffiti been applied at or before the amphoras reached Peiraieus,we would not expect to find the topographical concentration of markings in the Agora that we do. The only two amphora graffiti types that may have been applied at the port or at some point before the jars'arrivalin Athens are price marks 104. This point is made by Grace (1949, pp. 175-176), who arguesfrom this that stamps must have providedthe necessaryguaranteethat the jars held the properamount.That such a guaranteewas needed seems doubtful given the rarityof stampedamphora types in general.For furtherexamples of an interestin jars, not measuresof volume, see Yardeni1994, pp. 70-72 (Elephantinecustoms document ca. 475 B.C.);Dem. AgainstLacritos10, 19, and 20, referringto shipmentsin terms

of numbersof keramiaof wine; and Larsen(1938, pp. 394-395), who providesan appropriatelycautious discussionof whether and when we should translatekeramionas referring a specific unit of measure.In this regard,it is noteworthythat references to xsp4Ioca in papyrisometimes specifytheir size in numbersof choes (e.g., PEnteux, 001, document 34rp, r, 4; P Ryl. 4, document 556rp, r, 8). In one case (P Ryl. 4, document 564rp, ctr. 17), a series of keramiaat

6 choes each on the Arsinoan standard are then convertedinto numbers of Attic metretaiof 12 choes. The 12-chous definition of an Attic metretesis a restoration. 105. The acceptableor anticipated range of variationremainsto be determined,and this range is likely to have variedby producerand time period. On more recent expectations concerningthe size of ceramicvesselsexpectationsthat were not always matched in reality-see Blitzer 1990.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

75

and volumetric marksin non-Attic units. A surprisingnumber of the price marks discussed above were read as using stater for the monetary unit. In the Athenian market, such a notation would require knowledge of which non-Athenian stater was intended and what the proper equivalent would be in Attic coinage. These graffiti, therefore, would not facilitate sales in the Agora. On the other hand, such marks may have been inscribed by the shipper either as a price tag for buyers in the port, who presumably would be accustomed to a wider range of currenciesand commercial standards, or as a reminder to the shipper of the price per jar he would have to seek for that particularshipment. Clearly, with so few price graffiti, neither scenario was especially common. The use of stater notation, however, does seem to point toward importers and activities that took place at or before reaching the port rather than Athenian (even resident alien) retail merchants in the city.106In two cases where staters are indicated, 25 (though note the uncertainty of the use of staters here) and 26, the price mark is accompanied by a presumably complete tally of the volume of contents in the jar. The initial filling of the jar by the exporter would provide a context in which to apply both notations. Tracing these graffiti to the exporter also explains why non-Attic units of money are used and why the same volumetric notation on 26 also occurs on two other Mendean jars (27 and on an example cited by Johnston found at Kommos).07 The few drachma labels could, of course, have been applied at any point in the jar'suse. On the whole, with so few price marks in this collection, it seems unwise to tie these particulargraffiti to activities in the southeast corner of the Agora. While price marks seem often to have been applied before the amphoras reached the Agora, the rarity of these markings is striking on vessels that one might assume to be the quintessential commercial containers in antiquity. I return to this question of the rarity of price marks below, after considering other possible contexts for the application of the amphora graffiti. The second group of marks that might have been applied before the jars reached Athens are the rare volumetric marks that may record nonAttic units. There are only two suggested here: 9 and 10. In both cases the number 9 is indicated by a simple tallying on Chian straight-neck jars of ca. 425. This jar type is not known with a capacity of 9 Attic choes, and even 9 Chian choes would be large, though more likely. For this reason alone, I suggest that these two graffiti may have been applied while the jars were still in Chian hands. Even if a Chian sold wine in Athens, it would be surprising to see Chian units of measure in use. 106. This statement changes my earlierview (in Lawall 1995) that these priceswere also written locally.At that time I had not yet seen Johnston'svery convincingdiscussion(1996) of these graffitiraisingthe problemof converting foreign currency.I thank the anonymousHesperia readerwho directedme to this article. 107. Johnston 1996, p. 82.

LOCAL

SUPPLIERS

Apart from purchases at the port, amphoras might also have been brought into the wineshop from a local stockpiler/supplier of either local or im-

ported wine or other products. The processes followed at such places are w. T p pt. .o o . not detalled in the anclent sources, but some evldence points toward the existence of such suppliers. Two different terms refer to places where wine might be purchased: and olvot. The former term tends to refer to places where wine xaoklsxa

76

MARK

L. LAWALL

is dispensed for drinking on the premises, often a common bar of relawine might have also been sold from such places tively low reputation;108 for home consumption. Talcott's description of the contents of well R 13:4 matches what one might expect in a bar:many drinking cups, equipment for preparing"fastfood" (a restricted range of cookwares), some plainware storage or mixing vessels, and of course the amphoras.'09Oinoi, on the other hand, functioned solely as wineshops, as opposed to places for eating, drinking, and socializing.Two representationsin vase painting of wine purchases seem to illustrate oinoi:the wine in the amphora is being tasted with a sponge by the customer rather than simply from a handy cup as must have been done in kapeleia.l10Amphoras, wineskins, and pithoi are represented in these paintings, and all of the activity is focused on the purchaseof wine-and not on other activitiesappropriatefor the kape/eia.lll While no oinoi have been identified in the Agora,112it seems possible that some of the large assemblages of amphoras turned upside down at other sites may indicate the cellars for such establishments.113Although pithos fragments have been found in wells associated with the wine-selling area described here (S 16:1 and R 13:1),114on the whole the contents of these deposits and the scanty associated architectureare not extensive enough to indicate a large oinoswhere only wine was sold. If some, or even many, of the amphoras in the R 13 kapeleia(or perhaps it was only one?) were purchaseddirectlyfrom merchants at Peiraieus, it seems equally likely that as those vessels were emptied, they were refilled from local suppliers (oinoi) who could keep a larger ready supply of im108. Davidson 1997b, pp. 393-394, with referencesand illustrativequotations. 109. Talcott 1935, p. 497. Shear (1975, p. 357), in discussingthe fill of U 13:1, not far from R 13:4, describes the mass of domestic pottery as too great for "thekitchen of a private house"and proposesinstead the presence of a tavernnearby.He adds that "perhapsin associationwith the tavern, perhapsunder separatemanagement, was a wine shop which also made heavyuse of [U 13:1] to dispose of its empty and brokenamphoras."Shear's separationof the tavernand wineshop is certainlypossible.It is also possible that both functions existed under one roof. 110. CVABelgium 2, III He, pl. 16:5 = ABV299, no. 20, Princeton PainterType B amphora.Beazley calls this a scene of the sale of oil, but the wineskin makes it more likely to be a scene from an oinos(cf. Immerwahr 1992, p. 127, suggesting that at least the side with the wineskin is a scene of wine selling). Immerwahr(1948,

p. 190; and 1992, p. 127) calls the PrincetonPainteramphoraa scene of a kapeleion, but the exclusivefocus on wine selling, especiallyin one scene showing a wineskin, seems more appropriatefor an oinos. Furthermore, Immerwahr'scomment that sales of smallerquantitiesof wine on the PrincetonPainteramphoraidentify that scene as being in a kapeleion, while supportedby the depiction of small jugs held by some of the patrons,is not supportedby the amphoraseen on the shoulderof one of the men. The second image of an oinosis on a red-figure stemless cup tondo by Douris is illustratedand discussedby Buitron 1972, pp. 102-103 (= CVAUSA 8, pp. 35-36, pl. 19:2;ARV2 445, no. 252). 111. Davidson (1997b, p. 393) identifies the Douris scene as illustrating a kapeleion.Buitron (1972, p. 103) considersthe scene to be that of a wineshop where merchantsmight buy their wine (citing Forbes1955, p. 118). Forbes,however,does not make a clear distinctionbetween kapeleiaand oinoi (he does not use the latter term). Im-

merwahr(1948, p. 190) arguesfor a "winecellar"as opposed to a kapeleion. 112.AgoraIII, p. 199, no. 660 (= Isaios 6.20), refersto an oinosnear "theposterngate."Such a gate, dating earlyin the 4th century,is found near Building Z in the Kerameikos excavationsnext to the SacredGate. While it may be tempting to see Building Z as the cuvotxaoc(tenement house) in the passage,the third period of occupationthere is later than the date of the passage(see Knigge 1991, p.93).

113. There is, however,considerable uncertaintyas to when these collections representwine cellarsand when they representdrainageor foundationlayers for their associatedbuildings.For various sides of this argument,see Koehler 1986, pp. 62 and 66 with references; Empereurand Garlan 1992, p. 213, no. 158, and 1997, p. 165, no. 18 and p. 190, no. 131; and, although dealing with later constructions,Mattioli 1998. 114. Boggess 1973, from S 16:1, p. 46, no. 66; p. 73, nos. 116-117; from R 13:1, p. 73, no. 118.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF AMPHORAS

77

ported and local wine. A fragment of Aristophanes' Heroesmight describe something like the transactions between kapeleia and oinoi. A slave is told the following: TpX? E?5TOv olvov %.popEcx xsvOv ;cap`Ov/ TCovEVaoOEvxat (367oa xal y?DG7TYpIOV, / XarCSLTaC oiGoo JauVov

06cO(pop

op?l v.

... run to the wine market taking an empty amphora from those inside and a lid and a taster, and then hire yourself out as an amphora-carrier.ll5 The passage itself may have more to do with private households' cellars in need of restocking, but the general action of taking an empty jar to the oinosmay provide a context for the application of the graffitiwe find among the kapeleia. Unlike in the port, where one imagines multiple jars being purchased with no opportunity to check the specific volume of wine in each jar, purchases at the oinos provide just the opposite situation. Individual jars are clearly the subject of attention in the representations in vase painting of these establishments.When a single jar was filled and the wine purchased, a general expectation of the size of the amphorabeing filled probablywould not have sufficed. Individualjars have quite a wide range of actual capacities. Over a batch of a hundred such jars, for example, the small jars might be expected to be balanced by the large. With a single jar, however, expectations might come up short by 1 or 2 choes out of an expected 7 or 8.116 Following Lang's reconstruction closely, the graffiti might have served to keep track of the wine being poured into the jar from a wineskin, pithos, or even another amphora.117Volumetric graffiti accounting for 7 or more choes could result from restocking the kapeleion with imported or local wine from a local supplier."l8Large units noted initially fit such activity 115. For this fragmentsee Kassel and Austin 1984, 111.2,no. 310. A visual image of a similaridea is seen in a Thasian amphorastamp (Th 11903, illustratedon the cover of Garlan 1988). An Attic red-figure oinochoe shows a man carryinga smalljug, standingfacing another figure among variousamphoras shown in stands (Boulter 1963, pl. 49, no. 13, also discussedin Koehler 1986, pp. 61-62, fig. 13). Boulter suggests (p. 131) that the man with the jug might be a merchantabout to pour out a samplefor the customer.A similarinterpretation,with the jug used to decantwine into the purchaser's containerand with the quantity decanted markeddown by the merchant,would easily reconcilethe vase paintingwith the graffiti.Koehler

(1986, p. 57, fig. 8) also illustrates a symposiastcarryinga garlanded amphoraand a smalleroinochoe or pitcher.Although clearlynot illustrating a commercialtransactionof the sort describedhere, the illustration is reminiscentof the passagefrom Aristophanesand the Thasian stamp. 116. For publishedrangesof capacitiessee Eiseman and Ridgway 1987, pp. 51-52; Brashinskiy1984, pp. 170-204; Bertucchi 1992, pp. 145151; and Ramon 1991, pp. 127-130. For examplesof very narrowvariation, see Wallace Matheson and Wallace 1982 and Koehlerand Wallace 1987. 117. For a graphicand textual representationof sales of 5 and 3 choes from wineskins,see Immerwahr1992, especiallyp. 126 and pls. 29:a and 30:a.

118. The possibilitythat oinoiwere the primarysuppliersof local vintage should not be underestimated.Identified Attic amphorasarevery rareafter ca. 475 (see Lawall 1995, pp. 37, 38, 42; cf. Grace 1953, p. 102). Local produce could be sold out of wineskins or pithoi or even reusedimportedjars, and sold into reusedimportedjars.A plentiful supplyof imports might obviate the need for a local amphoraproduction.A parallelphenomenon is proposedat Berenikein Libyawhere local amphora productionceases once imports become common; see Riley 1979, p. 120: "The dramaticdecline in the occurrenceof [local amphoraproduction]afterthe Augustanperiod is likely to be relatedto a shift in economic emphasis,with an increasedrelianceon importedproducts."

MARK

78

L. LAWALL

well, if the filler used graduallysmaller measures as the jar was filled (e.g., a 5-chous measure,followed by 1-chous, 1/2-chous,and 1-kotyle measures). If we accept Lang's connection between some of the volumetric graffiti and the act of refilling the jars, it should be emphasized that the graffiti areused by kapeloi(barkeepers),and not by the general public.The reasons for this distinction are unknown, but the difference in practice is unquestionably attested by the distribution of the Agora graffiti. A related phenomenon may be the appearanceof two volumetric graffiti on small pitchers from the Agora excavations, neither of which was accompanied by amphoragraffiti, and one of which is certainlyfrom a noncommercial context.19 While too rare to allow certainty, finds of these pitchers suggest that households may have kept on hand a measured pitcher for wine or oil shopping, whereas exchanges at the kapeleia depended on knowing the precise quantity in the refilled amphora. SELLING

WINE

Since so many graffiti are found in wells thought to belong to kapeleia,I turn next to activities, again both attested and expected, in the kapeleia themselves. These establishments, and particularlytheir proprietors,were the frequent target of jokes in Attic comedy; the barkeeper is often portrayed as an untrustworthy cheat. This reputation is so ingrained that the term kapelosis sometimes translated as "huckster,"and the adjectivalform is translated as "cheating,knavish."120 The cheating, at least as xccaTqXtx6o; in can involve the chous and portrayed comedy, kotyle measures used in The wine. two taken from Aristophanes are serving following passages (Plut. 435-436 and Thesm.346-347): XCT &OsC xoamcrMIA 'x ToV yeTO6vCov / Tuoac;xoT6 XcsLC; let? ?oaXo.u(vV?Teoa;

'Ap

oactv

Is it the barmaid from the neighboring taps who always cheats me with the kotylai? x 5e -T;

xTrxon 8LaXvtaxV?Toa

5;

t xarClc;

Too Xo6; / r T -civ xoTcoiAv to vo6vELavx

Or, being a barman or barmaid, cheats with the standardfor the chous and the kotylai ....

119. There are only two examplesso far known to me. One of the pitchers (P 18609) is from a 4th-century, noncommercialcontext, C 19:5b, not far from well C 19:9, which containedone outlying exampleof numericalamphora graffiti (91) and the shoulderfragment markedby Andriskos (see 92 and Appendix).P 18609, however, probablydates well within the 4th centuryand comes from activity significantlypostdatingthe fill of C 19:9. The second waterjar,P 30802, was found in well R 13:11 from the late 4th century.Despite the proximityof this well to the earliergraffiti-ladenwells, there are no other graffitifound in this well. Whether this pitcher comes from a strictlydomestic context or from a kapeleionwhere graffitiwere not used in transactionsis open to question. 120. LSJ,s.v. xm'nrXosand xoc7rXt0x6;.Lucian,Hermot.59, continues to characterizekapeloias cheating in the 2nd centuryA.C.: oi xdctnXor.-XspCaa Lo,votL ye oi

TcoXXoi xoaiaoXcoav-cS xat the signifiRegardlessofwhether these jibes at wine sellerswere deserved,121 cant point about these passages is the close relationship between choes and kotylai and retail sales of wine. This connection continues in the verb xoUt2iEstv. The word is often read as describing sales in kapeleia,but the term also has a wider range of meanings, from wholesale transactions that break a ship's cargo into multiple sales to the more literal translation of selling by the cup.122 Against this background of terminology and bad reputations,it seems likely that some of the amphora graffiti appearedas part of the process of "selling by the kotyle."As each kotyle was dispensed from the kapeleion's jar into the customer'scontainer,the kapelosmay have jotted a note on the

xaxoEz-cTpoUv-TC; ("barkeepers-most

of them adulteratingand cheating and giving false measures"). 121. On the reputationof kapeloi,see Kurke1989. 122. For the use of the term to describesales of cargoin parts rather than as a single batch, see Arist., Oec. 1347b8 and IG XII Suppl. 347 11lines 11-13. Davidson (1997b, p. 394) reads the latter as referringto activitiesof kapeloi,but this readingis not at all requiredby the text.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

79

jar-perhaps under the suspicious eye of the customer. Such action would result in some of the smaller volumetric notations seen in the Agora collection. Graffiti using largerunits (1- or 5-chous units instead of kotylai)123 might have resulted from customers buying wine in smallerjugs for home consumption or for drinking and dining by a larger group at the kapeleion. The graffiti read as summing up a series of smaller unit markings may also fit into this activity.Perhaps the amount ordered was marked on the jar, and then the smaller unit marks were struck as each amount was dispensed toward the total. It is possible, too, that the process worked in the opposite direction with the larger summation recording the completion of an order. Either process would result in the same arrangement of graffiti. The practice suggested here, then, would keep the graffiti in the area of the kapeleiaeven if the wine itself is destined for outside consumption. The marking of the storage vessel (i.e., the amphora in the kapeleion) rather than the receiving vessel (the customer'scup, jug, or amphora) may be indicated in a late-5th-century Thasian law: ocV-T0; ?U tOoLsoLvov pycTO, -TJv&)vWiv xuplYv EVOCL, ?CoS' OCV reS

u s on nvyttioi.

The one who buys the wine in the pithoi: the sale is legitimate if he marks the pithoi.l24

123. Cf. Kerameikos IX, p. 173, E14, fig. 47 (see above,note 85), where 9 individualchoes are tallied instead of using a TIto accountfor the first 5. 124. IG XII Suppl. 347 I lines 5-6. 125. E.g., Meijer and van Nijf 1992, no. 143. 126. Vinogradov1986; cf. Jefremov 1998, pp. 90-91. 127. Salviat1986, pp. 148-149; for examplesof volumetricgraffition pithoi, see Jefremov1998. 128. Amyx (1958, p. 168) notes that pithoi on the Attic Stelai aresold with their houses and where there areprice notations on pithoi they seem to indicate an expensivevessel (an impressionsupportedby their longterm use once purchased).The sheer size of these vessels also militates againstthe sale-and movement-of a filled pithos.

Although this inscription has been interpreted as referring to amphora stamps,125Juri Vinogradov argues that the inscription refers 1) to storage pithoi, not transport amphoras;2) to markings on the vessels after firing, not to stamping; 3) to markings on a pithos to be taken away by the buyer; and 4) to an accounting of the amount filling the pithos.126Francois Salviat argues, instead, that the action is the covering of the pithoi with some perishable material and applying a seal to this covering so that the wine cannot be tamperedwith once purchased. Salviat'sreading is supported by various later ancient sources using very similar Latin terminology, dolium signare,to describe proper sales of wine in dolia (pithoi). While Salviat's argument is persuasive, Vinogradov's reading is supported by the many volumetric graffiti found on the rims of pithoi.l27 Perhaps, roughly following Vinogradov, the law is meant to stipulate that decanted quantities must be indicated on the storage vessels. The pithoi belong to the seller,not the buyer (cf. Vinogradov). The wine being sold is "in the pithoi"; it is neither being sold into another's pithos, nor does the purchase necessarilyinvolve the pithos itself.128If these interpretations of the law are correct, then this official Thasian practice might be similar to certain practices resulting in the Athenian graffiti, marking the source container with the amounts dispensed for smaller purchases. OTHER

COMMODITIES

A further reuse of emptied imported amphoras appears to have been as containers for altogether different commodities. As noted above, the reuse of some jars to hold honey (perhaps even made in the same area on the

80

MARK

L. LAWALL

evidence of the beehive fragments from R 13:1) or oil may have required labeling the reusedjars with M/ME and E to warn customers and perhaps even a harried salesperson that the particular amphora did not contain wine. Examples of such graffiti, especially M/ME, are found with some frequency outside the southeastern areaof the excavations,so it is possible that these amphoras, once reused and labeled in this fashion, were sold to customers to take away,unlike the other markedjars discussed here. On the whole, these marks show various differences in characteristicswhen comparedwith the other graffiti.The fact that they are found at other sites as well, again on amphoras,might suggest their more widespread use with amphoras than the numerical notations and other abbreviations. MEASUREMENTS

OF WEIGHT

The weight graffiti are among the more difficult, both in terms of reading or interpretation and in terms of identifying the locus of activity.Some of the proposed weight notations may indicate the empty weight of the jar (3, 5, 7?, 14, 15, 29?, 43, 79?, 80, 98); others might refer to the weight of added contents or the gross weight of the jar with some partial filling (16, 17, 18); and some seem to refer to the weight of the jar completely filled (45?, 74).129Building on the discussions of circumstancesfor applying the price and volumetric graffiti, above, and considering what products might have been sold by weight, it is possible to propose loci for the use of these weight graffiti. On many later,Roman-period amphora dipinti from the Agora, wine, oil, and honey are all described, and so presumablysold, in terms of weight (litrai) rather than volume.'30These dipinti raise the possibility that the same practice held true earlier;however,referencesto wine and oil that are more contemporary with the late-5th-century amphora graffiti tend to speak of volumetric units or simply numbers of jars.131Papyri of various periods, including both Hellenistic and Roman documents, refer to meat, wax, glue, ochre, wool, and cloth in terms of mnas."32Aristotle also describes the use of steelyardbalances for sales of meat and dye (presumably in solid form).133Of these items, only meat is known to have been shipped, stored, and perhaps sold out of amphoras.134 Indeed, the sale of meat-presumably chopped to manageable chunks-out of amphoras would fit some of the evidence and conditions offered by the amphoraweight graffiti.An old jar might have been used to weigh out meat, and this would provide either the markings for the empty weight of the jar or the marked weight of a partially or completely filled jar.The meat would easily be removed from the jar again and given to the customer without necessarily losing any pieces adhering to the interior walls of the jar. Similarly,wine, oil, or other liquids could be sold by weight in this fashion with a minimal loss due to liquid staying behind in the jar. Honey, on the other hand, would not be an expected product for sale by weight from one amphora into another container. On analogy with the argument offered above for restocking the kapeleionfrom other merchants'stockpilesof goods, it is possible that weight graffiti indicating a completely filled vessel may have figured in the sec-

129. As above,numbersfollowed by question marksin these lists indicate uncertaintyin readingthe graffito as a weight notation. 130. AgoraXXI, pp. 77-81. 131. The degree to which the Agora graffitiand dipinti illustratea drastic shift from measuringliquids by volume to measuringby weight depends in part on how the problematic,earlygraffiti here are read;cf. Lewis 1979, p. 126, and see below, note 135. 132. For productsreferredto in terms of mnas, see, for example,from the Zenon archiveof the 3rd century B.C., Edgar 1931, nos. 38 (referringto glue), 58 (wool), 61 (wax and wool), and 120 (cloth);P Oxy.108 (A.D. 215) refersto meat;and P Oxy.2144 (3rd centuryA.C.)refersto ochre,glue, and wax. 133. Arist., Mech.849b35, 853b25. 134. G. Bass, pers. comm., October 1999; and see Williams 1979, pp. 117118, pl. 46, for fish in Punic amphoras at Corinth. A thick deposit of red pigment was found coating the interior surfacesof an amphorawith only the lower half preserved(from R 12:1, closed ca. 480 B.C.). This may be evidence that ochre too was sold from amphoras,or this may be an amphora reusedas a paint pot.

WINE

GRAFFITI,

AND

SELLING,

REUSE

OF

AMPHORAS

ondary use of that vessel in supplying the kapeleion.In such a case, the markedjar would not leave the store;its contents would be sold off gradually.The weight notation was only needed in the initial purchase from the stockpiler.Perhaps honey or processed meat was involved in such transactions since wine and oil in the late 5th century were more commonly sold by volume. The scarcityof the possible weight notations relativeto the large number of volumetricmarksmay indicate that the usualproductsof the kapeleion were not sold by weight. Alternatively, while weight could be used as a unit for sales, perhaps sales by weight were not common practice. This scarcity of weight notations might, indeed, mark an important point of distinction between Classical Aegean commerce and later Roman-period practices.l35 SCARCITY

135. Lewis (1979, p. 126) notes "a clearshift from measuringto weighing [in the Roman period]."Lang (Agora XXI, p. 64) characterizesthis difference as follows:"Tarenotations from the Greek period areboth fewer and less standardizedthan those from the Roman period." 136. See Johnston 1979, p. 33, table; andJohnston 1978b on batch price marks. 137. Johnston 1984. 138. On the importanceof individualnegotiation and bargaining in ancient,and sometimes modern, price creation,see Lowry 1987, pp. 210-211 and 237. 139. See notes 5-9 above.

OF PRICE

MARKS

If the sample of late-5th-century amphora graffiti offered here is representative of general practice,then another scarcityin the corpus is particularly striking: there are very few price graffiti. Many of the prices that do appear are likely to have been written outside Athens for earlier stages of wholesale transactions. This scarcityof price markson amphorasbecomes less surprisingwhen consideredagainstthe general natureof ceramicgraffiti.Prices on finewares rarelyreferto the individual pot. Instead, prices aregiven for lots or groups of pots."36Price graffiti are similarly rare among pithos graffiti, where, as with the amphora markings discussed here, notation of volume is much more common.137 These patterns of scarce price marks fit well with a view of Classical Aegean retail economies as involving prices createdby negotiation and influenced by any number of economic and social factors.138 While a wholesaler might mark some vessels with prices as a reminder of what price had to be gained for a successful venture, the actual prices paid later at the retail level might vary considerably.Such a view of ancient retailprices makes their study quite difficult, but it fits the extant material evidence. CHIAN

AMPHORAS

AND

GRAFFITI

As noted in the previous section, the Chian amphoras in R 13:4 carry an unusually large portion of the graffiti in that deposit. Chian amphoras were the focus of previous interest in the Agora volumetric graffiti for their possible link to the Athenian Standards Decree.139Given existing debates as to the date of the decree, it is unclear whether the marked amphoras are sufficiently close in date to the decree for the marks to be attributed to uncertainties caused by the decree. Most of the markedjars in R 13:4 date fifteen to twenty years after the earliest proposed date of the decree (ca. 449 B.C.). Furthermore,a wide range of graffiti types appearon the Chian jars in R 13:4, including volumetric notations (not all accounting for the complete capacity of the jar), but also including price marks, possible weight notations, and isolated letter abbreviations.

82

MARK

L.

LAWALL

Various possible contexts in which these graffiti may have been applied were presented above. For the few price marks and combinations of price and volume graffiti, which seem likely to have been applied before the jars reached the wine-selling area in R 13, simply the large number of Chian amphoras in R 13:4 might raise the chances of such marks occurring on them. These marks thus need not be explained in terms of the dominance of Chian jars among the R 13:4 graffiti.The graffiti related to the use of the jarsin the wine-selling area-the volumetricnotations,weight notations, and isolated abbreviations-do requirefurther consideration as to why they are so commonly found on Chian jars in this particular deposit. The often-cited high reputation and value of Chian wine may be of some importance here.140If graffiti noting partialvolume or weight, which I attribute to sales of wine in the kapeleionitself, are concentrated on one type of amphora,it is reasonableto assume there was considerable activity surroundingthat type in particular.Frequent decantings from those jars is the most likely activity.If these Chian jars were rapidly emptied due to the popularity of that particularwine, then it follows that graffiti recording complete capacity and related to their refilling at the oinoswould be more common on these jars than others. Similarly,graffiti such as weight notations and isolated abbreviations,relating to the reuse of jars for products other than wine, would be most often applied to these rapidly emptied Chian jars. SUMMARY

In the foregoing review of situations and contexts for the application of the amphora graffiti, I proposed the following links between practices and graffiti. First, the only graffiti that are at all likely to have been applied at or before the arrivalof a shipment at Peiraieus are price marks and the two volumetric notes that might be in non-Attic units. This is especially true of marks involving staters. The scarcity of price marks on amphoras and other vessels suggests that such marking was not a common practice in retail sales. Volumetric notations and perhaps some weight notations accounting for the complete capacity of the jar may be attributed to restocking the kapeleionfrom an oinosor other local stockpilerof agriculturalproducts. Such a restocking with quantities marked on the jar seems to have occurred primarily in kapeleia,not in private houses in general. Only in this way can we account for the narrow distribution of amphora graffiti around the Agora. Sales of wine at the kapeleionitself may have resulted in a wide range of markings:volumetric notations using small units or only accounting for part of the jar'scapacity;summations of simple strokes;and possibly weight measures indicating either partially filled jars or the tare weight of the jar. Sales of other commodities, such as oil, honey, or honeyed wine, might have led to the very visible use of the single abbreviations M/ME and E. It is also possible that processed meat was sold by weight and hence resulted in some of the weight notations. I couch these suggestions in the most qualified terms because such informal practices are very difficult to reconstructwith any certainty.The

140. See, e.g., Grace 1949, p. 182; 1979b, p. 122; Grace and SavvatianouPetropoulakou1970, p. 359; Lang 1956, p. 13, note 16; p. 23.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

83

proposals here are at best working hypotheses to be continually tested and refined against further contextual studies of amphoras and their graffiti. Indeed, a few more points of qualification should be noted. First, the uses of the jars as described above need not always have entailed the use of graffiti. Many amphoras and even some types of amphora (see Table 2) from the southeastern part of the Agora excavations carryno graffiti at all, and yet it seems likely that they too were involved in the processes just described. Second, we tend to have only one graffito per jar as preserved. Lang was probably correct to suggest that once the capacity of a jar was checked, the volumetric graffiti would serve as a reminder for future such uses.141 For the partialvolumetric and partialweight graffiti, proposed here to have been used in sales within the kapeleionitself, it is surprising not to have more than one set of markings for the quantity dispensed. The summation graffiti, however, might well have accounted for multiple small decantings in the kapeleion.Even if these contradictions to the models of practice proposed here eventually require the adoption of new models altogether, any new proposals will also have to take into account both the wide range of possible uses of the graffiti and the constraints of the contextual evidence. GRAFFITI, CENTURY

141. Lang 1956, p. 24. 142. In terms of the long history of this "technology"of accounting,note too that other written documents relatedto tradedate as earlyas the late 6th centuryB.C. (Wilson 1997-1998; Lombardo1988). 143. AgoraXXVII, p. 23. Similarly, Young (1951, p. 239) suggests that on House K was built by "squatters" land originallypreparedfor the constructionof the enigmatic Poros Building. Refugeesliving in huts were clearlybetter off than those living in pithaknai(perhapssmall pithoi), if one follows a literalreadingof Ar. Eq. 793.

ECONOMICS, AEGEAN

AND

HISTORY:

THE

LATE-5TH-

In the preceding section, I suggested activities that might have resulted in the application of the graffiti and considered why the graffiti clustered in the southeast corner of the Agora. None of the activities, however, is tied to the restricted chronological range of the graffiti. While a few examples of numerical amphoragraffiti can be found from the late 6th century on,142 the concentration of graffiti in the last thirty years or so of the 5th century is striking and requiresexplanation.The activities themselves do not seem inherently limited in terms of chronology (oinoi are illustrated on a 6thcenturyblack-figurevase and a 5th-century red-figurevase;cheating kapeloi appearwidely in Greek literature;and references to sealing and marking pithoi are found in law codes from the Classical period [Thasos] through later Roman times). Perhaps there are elements of the environment in which the actions took place that encouraged more frequent use of graffiti. Given that the graffiti coincide in date very closely with the Peloponnesian War, from 431 to 404 B.C., wartime conditions in Athens should offer the most probable explanation for the rise in the use of graffiti in these decades. The Periclean strategy of abandoning the Attic countryside in the face of seasonal Spartan raids is one likely cause of the changes in daily activity, changes of which these graffiti are one reflection. The overcrowding of Athens during the time the countryside was abandoned is vividly attested by both Thucydides (especially 2.17 and 2.52) and archaeological evidence. Most immediately related to the graffiti in terms of their dates and topographical distribution is a mass of scrappywalls found below the Stoa of Attalos, i.e., north of the R 13 wells. Rhys Townsend quite reasonablysuggests that these are the walls of huts for wartime refugees; they were cleared away for more substantial buildings ca. 410 B.C.143

84

MARK

L. LAWALL

In this suddenly crowded city, the earlier marketing and accounting practices may have no longer sufficed, and so the use of graffiti on amphoras, which had been rare,became more common practice. Such an increase in the complexity of economic and other organizational practices in the face of increased scale of activity has been noted in a wide range of archaeological, historical, and ethnographic contexts.144The amphora graffiti of the late 5th century would fit into this pattern very easily. While the later 5th century was in part a time of great prosperity for Athens, it also included times of considerable hardship.Most importantly for the purpose of explaining the rise in the use of graffiti is the possibility of fluctuating access to imported, seabornegoods through Peiraieus.Athenian strategy,both during the Peloponnesian War and at other times, depended very much on imported supplies. Fluctuations in the actual volume of imports are difficult to document in the archaeological record;145 however, it seems likely that Peiraieus was not a smoothly functioning harbor, especially in the last decade of the century.While imports could enter Attica from other sites (e.g., Thorikos, Sounion, Rhamnous), disturbances of peace at Peiraieuswere certainly a recognized cause of economic hardship for Athens.146 The connection between such likely, but archaeologically invisible, drops in the volume of imports and the amphoragraffiti is as follows. One practice that may have resulted in numerical graffiti was refilling amphoras from a local oinos once the original imported contents were used up. If there were periods, even as short as a few months, when new imported amphoras were in shorter supply, we would expect to see an increase in precisely this dependence on local stockpilers (oinoi). In the archaeological record, brief events taking place over many years may be obscured;nevertheless, the increased use of graffiti broadly visible across the war years could result from these periods of greater dependence on local products. Hanson, Foxhall, and others have proposed that Attic local agriculture, especially olive production, would not have been seriously disrupted by the hostilities. There is some disagreement as to the stability of grape crops during wartime, but it is probable that local oil and perhaps wine would have been availablewhen imports were in shorter supply.147 Both factors, the overcrowdingof the city and the drop in availability of imported amphoras,coincide very well with the appearanceof increased graffiti use on the Agora amphoras.It is difficult, and unnecessary,to place one factor over the other in explaining the rise in graffiti use. Both helped create an economic milieu in which the graffiti was clearly seen as necessary on a much larger scale than had ever existed before or would later; indeed the two factors never recur together in Athenian history.148 If wartime conditions help explain the general concentration of the graffiti in the last three decades of the 5th century,the specific concentration of graffiti on Chian amphoras in well R 13:4 but not in later deposits remains troubling. I proposed in the preceding section that the popularity of Chian wine provided a possible explanation for the clustering of the graffiti. Yet, this reputation, as far as we know, did not decline after ca. 425, so why do other amphora types take on a greater share of the graffiti in later deposits? As is clear from the estimated numbers of jars of each

144. Johnson 1982; van der Leeuw 1984; Feinman,Kowalewski,and Blanton 1984. 145. It is possible to studythe relativepresenceof one amphoratype comparedwith others. 146. See Garland1987, pp. 10-57 for the historicalsurveyof Peiraieus. For specific referencesto troublesat Peiraieus,see, e.g., Xen. Hell. 2.9, Lysanderclosing accessto Peiraieusin 405; 2.3.11, destructionof the Long Walls and the "wallsaroundPeiraieus" rspli Tcv rIspaLoc)in 404; (roc [LT?(s)X] 2.3.21, attackson metics orderedby Theramenes.On the importanceof a smoothly managedPeiraieusand comfortablefacilitiesfor merchants,see especiallyXen. Poroi3; and on the importanceof peace, see Poroi5. 147. See Hanson 1998, pp. 49-71 on the hardinessof cropsto resist hostile human destruction;Foxhall (1993, pp. 138-139) proposesthat vines would have been particularlyendangered (cf. Hanson 1998, pp. 223-224). 148. Interruptionsto Athenian accessto Peiraieusdo recurin Athenian history but not in conjunctionwith unusualovercrowdingof the city; see Garland1987, pp. 37-57.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF

AMPHORAS

85

major type for the wine-selling area (summarized in Table 2) and as indicated in broader studies of these and other Agora deposits, the percentage of Chian amphoras relative to other types declines considerably between ca. 425 and the end of the century.149 Despite this decline, Chian amphoras did still receive graffiti, perhaps attesting to the continued popularity of that wine, which quickly drained the fewer Chian jars in circulation. Other amphora types, however, were also being emptied, reused, and marked with graffiti. The phenomenon of these many graffiti in late-Sth-century Athens should, finally, be considered in an even broader context. Economic conditions in later-Sth-century Athens and in the Aegean more generally, down into the 4th century,include a wide range of developments of which the intensive use of graffiti on amphoras in Athens should be considered as one, very tangible, symptom. In terms of the broaderstudy of the Aegean amphora trade, the last third of the 5th century saw the emergence of consistent amphora-marking practices in major production regions on a scale not seen earlier,and this pattern of development continues through the 4th century and the Hellenistic period.150 The new monetary wealth of Athens in the second half of the 5th century has been highlighted by some scholars as contributing to a more entrepreneurial and commercialized economy, one in which, not surprisingly,the novelty of such developments and wealth received a mixed reception.l51 These changes, taken together, set up a context in which the amphora graffiti are not unexpected. Imagine an Athenian, recently displaced from the Attic countryside, going to the kapeleionwith his jury pay. He, like Dikaiopolis in the opening of Acharnians,is dismayed by the unfamiliar jostling of the markets where people he has never met before and may never see again want his obols. He reaches the kapeleionand, remembering all he has heard of the dishonesty of kapeloi,asks to have his jar filled with 21/2choes. He is relieved to see that the kapelosmarks each chous and kotyle on the side of the amphora as they are poured. "Quite an honest man!"he thinks, though he still longs for his country deme where he need not have been so apprehensive procuring wine.

149. Lawall 1995, pp. 290-292 and 1998a, p. 87, fig. 5.1. 150. From ca. 430 B.C., Mendean jars arelabeledwith small painted letters, reflectinglocal practice.Near the beginning of the 4th century, Thasos adopts a very consistent stampingprogram,and this change coincideswith the apparentexpansion of Thasian productionand exportation. For discussionof when in the 5th centurydifferentamphoraproducers use consistent markingsystems,see Lawall 1995, passim. 151. Burke(1992) places the major increasein scale and therefore

"disembeddedness" in the 4th century; however,like Crane (1992), Burkealso notes the influx of wealth, trickling down to the urbanpopulationin the later 5th century.Olson (1990) on Aristophanes'Plutusand Kurke(1989) on kapeleiaboth highlight the suspicions with which some of these noveltieswere viewed. Seaford(1998) and von Reden (1997) addressissues relatedto the ethics surrounding increasingmonetization of an economy, both acknowledgingthe broader literaryand ethnographicresearchon this question as discussedin Bloch and Parry(1989).

86

MARK

L.

LAWALL

APPENDIX AN D RISKOS, A KAPEL05 OF THE ATHENIAN A O RA? The latest of the 5th-century deposits with graffiti, Q15:2, also contained possible evidence for identifying the personnel involved in the wineshop. The deposit contained three owner'sgraffiti, each restorableas bearing the name Andriskos (88-90, 92).152As it happens, Andriskos spells xaicox; with an omega in one instance and an omicron in another. He may simply be a poor speller; however, it is notable that the inscriber(s) of the commercial graffiti here at times uses variable spellings for the same information (e.g., E or H for half). Furthermore,the same name, Andriskos, can be reconstructedfor anotherowner'sgraffito,found in C 19:9, on the shoulder of a Chian straight-neck amphora. C 19:9 is the furthest outlying of the deposits with the numerical graffiti and is slightly later than R 13:4. The fact that this name occurs four times in the same findspots as numerical graffiti may suggest a significant relationship between the name and the commercial activity implied by the amphoras and the graffiti. The name Andriskos has a widely scattered distribution and is well representedboth in Athens and in the areasproducing the amphoras discussed above. Athenian examples Name on red-figure psykter,ca. 510-500 B.C. (PAA [Suppl.] 127823) Cleruch on Lemnos, 450-430 B.C. (PAA 127825) Dedicatory inscription from Eleusis, ca. 350-300 B.C. (PAA [Suppl.] 127827) Macedonianexamples EAM, no. 86, late 2nd century A.C.,son of Neikolaos and father of Neikandros Paus. 7.13.1, son of Perseus ExamplesfromtheAegeanislands(cited in LGPNI, s.v. Andriskos) Telos, 2nd century B.C.,as father of Xenotimos (SEG XXV 857) and father of Aristandridas (SEG III 725 and IG XII.3 34, line 26) Chios, SEGXVII 388, 3rd century B.C. Naxos, RE (5), Ath. 3.78c, citing a Naxian historian Andriskos, possibly of the 4th or 3rd centuries B.C. Rhodes, I. Lindos 344.27, 47 B.C. With such widespread occurrencesof the name, there can be no certainty as to the most likely origins of the Andriskos from Q15:2. Some nonAttic dialect in the amphora graffiti and the significant presence of amphoras from both the Ionian islands and northern Greece allow for seeing Andriskos as a foreign merchant, but this is only a possibility. Andriskos could be of a long-standing Athenian family.

forthe 152. Possiblysuggestive questionof whethertheseareall inscribedby the same person,in both 89 and 90 the E of ANAPIeKO has a

loweranglebetweenthe compressed bottomtwobars.

GRAFFITI,

WINE

SELLING,

AND

REUSE

OF AMPHORAS

87

REFERENCES Abramov,A. P., and A. A. Maslennikov. 1991. "AmforyV v. do n. e. iz raskopokposeleniyana myse Zyuk" [Amphorasof the 5th centuryB.C. from the excavationof the settlement at Cape Zyuk], SovArch, pp. 234-248. Agora = TheAthenian Agora, Princeton.

III = R. Wycherley,Literaryand Epigraphical Testimonia, 1957.

IV = R. H. Howland, Greek Lamps and Their Survivals, 1958.

X = M. Lang and M. Crosby, Weights,Measures, and Tokens,1964.

XII = B. A. Sparkesand L. Talcott,Black and Plain Potteryof the 6th, 5th, and 4th Centuries B.C.,

1970. XXI = M. Lang, Graffiti and Dipinti, 1976.

XXIII = M. B. Moore and M. Z. Pease Philippides,Attic BlackFiguredPottery,1986. XXVII = R. F.Townsend, The East Side ofthe Agora: The Remains beneath the Stoa ofAttalos, 1995. XXX = M. B. Moore, Attic RedFigured and White-Ground Pottery,

1997. Amyx, D. 1958. "The Attic Stelai, Part III: Vasesand Other Containers," Hesperia27, pp. 163-310. Babinov,Yu., et al. 1978. Graffiti antichnogo Khersonesa [Graffitifrom ancient Chersonesos],Kiev. Bagnall,R. 1989. "Fourth-Century Prices:New Evidence and Further Thoughts,"ZPE 76, pp. 69-76. Barron,J. 1986. "Chiosin the Athenian at Empire,"in Chios.:A Conference the Homereion in Chios, 1984,

J. Boardmanand C. E. Vaphopoulou-Richardson,eds., Oxford, pp. 89-103. Bertucchi,G. 1992. Lesamphoreset le vin de Marseille, VT s. avantj; C.-Il

s. apresj C. (RANarb Suppl.25), Paris. Blitzer,H. 1990. "KOPQNEIKA: Storage-JarProductionandTradein the TraditionalAegean,"Hesperia 59, pp. 675-711. Bloch, M., andJ. Parry,eds. 1989. Money and the Morality of Exchange,

Cambridge. Blonde, F. 1989. "Le comblementd'un

puits public a Thasos, 3: Le ceramique,"BCH 113, pp. 481-545. Boggess, E. C. M. 1973. "The Development of the Attic Pithos" (diss. Bryn Mawr College). Boulter,C. G. 1953. "Potteryof the Mid-Fifth Centuryfrom a Well in the Athenian Agora,"Hesperia22, pp. 59-115. . 1963. "Gravesin Lenormant Street,Athens,"Hesperia 32, pp. 113-137. Brashinskiy,I. B. 1976. "Amfory Mendy," in Khudozhestvennaya kul'tura i arkheologiyaantichnogo mira [Artistic cultureand archaeol-

ogy of the ancientworld], Moscow, pp. 67-74. . 1984. Metody issledovaniya antichnoy torgovli (naprimere Severnogo Pricernomor'ya) [Methods

for the study of ancient trade (using the exampleof the northernBlack Sea)], Leningrad. Buck, C. D. 1955. The GreekDialects: Grammar, SelectedInscriptions, Glossary, Chicago. Buitron, D. M. 1972. Attic VasePainting in New England Collections,

Cambridge,Mass. Burke,E. M. 1992. "The Economy of Athens in the ClassicalEra:Some Adjustmentsto the Primitivist Model," TAPA122, pp. 199226. Burling,R. 1962. "Maximization Theories and the Study of Economic Anthropology,"American Anthropologist 64, pp. 804-821.

Camp,J. McK., II. 1977. "The Water Supplyof Ancient Athens from 3000-86 B.C." (diss. Princeton University). Clinkenbeard,B. G. 1982. "Lesbian Wine and StorageAmphoras: A ProgressReport on Identification,"Hesperia51, pp. 248-267. Collitz, H., and F. Bechtel, eds. 1905. Sammlung dergriechischenDialektInschriften 3.2, Gottingen.

Crane, G. 1992. "The Fearand Pursuit of Risk:Corinth on Athens, Sparta, and the Peloponnesians(Thucydides 1.68-71,120-121)," TAPA 122,pp.227-256. Crosby,M. 1955. "FiveComic Scenes

from Athens,"Hesperia 24, pp. 7684. and Davidson,J. N. 1997a. Courtesans Fishcakes:The Consuming Passions of ClassicalAthens, New York.

. 1997b. "ABan on Public Bars in Thasos?"CQ 47, pp. 392-395. Dean, J. E., ed. 1935. Epiphanius' Treatise on Weightsand Measures: The Syriac Version(SAOC 11), Chicago.

Doulgeri-Intzessiloglou,A., and Y. Garlan.1990. "Vin et amphores de Peparethoset d'Ikos,"BCH 114, pp. 361-389. Dupont, P. 1998. "ArchaicEast Greek TradeAmphoras,"in R. M. Cook and P. Dupont, East GreekPottery,

London, pp. 142-191. EAM = T. Rizakesand G. Touratsoglou, 'EwtypacSE '4Avw MaxcEoviac ('EAiycta, 'Eop-

?oaia, Nodra Avyxc7tac(,)A': 'Errtypaoc5v,Athens KarzcCVAoyoC 1985. Edgar, C. C. 1931. Zenon Papyri in the University of Michigan Collection,

Ann Arbor. Eiseman, C.J., and B. S. Ridgway. 1987. The Porticello Shipwreck:A Mediterranean Merchant Vesselof 415-385 B.C., College Station,Tex.

Empereur,J.-Y., andY. Garlan,eds. 1986. Recherchessur les amphores

grecques(BCH Suppl. 13), Paris. . 1992. "BulletinArcheologique: Amphores et timbresamphoriques (1987-1991)," REG 105, pp. 176220. . 1997. "BulletinArcheologique: Amphores et timbresamphoriques (1992-1996)," REG 110, pp. 161209. Feinman,F., S. Kowalewski,and R. Blanton. 1984. "Modeling Ceramic Productionand Organizational Change in the Pre-HispanicValley of Oaxaca,Mexico,"in The Many Dimensions of Pottery: Ceramicsin ArchaeologyandAnthropology, S. E.

van der Leeuw and A. C. Pritchard, eds., Amsterdam,pp. 295-333. Figueira, T. 1998. The Power of Money: Coinage and Politics in the Athenian Empire, Philadelphia. Finley,M. I. 1973. TheAncient Economy, Berkeley.

88

MARK

Forbes,R. J. 1955. Studiesin Ancient Technology3, Leiden. Foxhall,L. 1993. "Farmingand Fighting in Ancient Greece,"in War and Society in the Greek World,

J. Rich and G. Shipley,eds., London, pp. 134-145. Garlan,Y. 1985. "De l'usagepar les historiensdu materielamphorique grec," Dialogues d'histoire ancienne

11, pp. 239-255. . 1988. Vin et amphoresde Thasos

(Ecole francaised'Athenes, Sites et monuments 5), Athens. Garland,R. 1987. The Piraeusfromthe Fifth to the First Century B.C., Ithaca. Garnsey,P. 1992. "Yieldof the Land," in Agriculture in Ancient Greece. Proceedings of the Seventh International Symposium at the Swedish Institute in Athens, 16-17 May 1990,

B. Wells, ed., Stockholm,pp. 147153. Grace,V. R. 1949. "StandardPottery Containersof the Ancient Greek Studies World,"in Commemorative in Honor of TheodoreLeslie Shear (Hesperia Suppl. 8), Princeton,

pp. 175-189. . 1953. "WineJars,"in Boulter 1953, pp. 101-110. -. 1971. "SamianAmphoras," Hesperia 40, pp. 52-95. -.1979a.Amphoras and the Ancient Wine Trade (AgoraPicBk 6),

2nd ed., Princeton. -. 1979b. "ExceptionalAmphora Stamps,"in Studiesin ClassicalArt andArchaeology, G. Kopckeand M. Moore, eds., Locust Valley,N.Y., pp. 117-129. Grace,V. R., and M. SawatianouPetropoulakou.1970. "Lestimbres amphoriquesgrecs,"in Exploration archeologiquede Delos 27: L'ilot de la Maison desCome'diens, P. Bruneau,

ed., Paris,pp. 277-382. Grakov,B. N. 1968. "Zametkipo grecheskoikeramich'eskoi epigrafike"[Notes on Greek ceramic epigraphy],in Antich'naya istoriya i kultura sredizemnomor'yai prichernomorya: Sbornik statey

[Ancient history and cultureof the Mediterraneanand Black Sea: Collection of essays],V. Ph. Gaidukevichet al., eds., Leningrad, pp. 100-107.

L. LAWALL

Guarducci,M. 1970. Epigrafia greca2: Epigrafi di caratterepubblico,Rome.

Hanson, V. D. 1998. Warfareand Agriculture in Classical Greece,rev.

ed., Berkeley. Holloway,R. R. 1966. "Explorationof the Southeast Stoa in the Athenian Agora,"Hesperia 35, pp. 79-85. Hultsch, F. 1882. Griechischeund romischeMetrologie, 2nd ed., Berlin. I. Lindos= C. Blinkenberg,Lindos. Fouilles et RecherchesII: Fouilles de

Berlin 1941. LAcropole. Inscriptions, Immerwahr,H. R. 1948. "AnAthenian Wine-Shop," TAPA 79, pp. 184190. . 1992. "NewWine in Ancient Wineskins:The Evidence from Attic Vases,"Hesperia61, pp. 121132. Jefremov,N. 1998. "Einige Abkiirzungenauf der antiken Keramikin ihrem Verhaltniszu den bekanntenMagfeinheiten:Uberlegungen zu Produktionund Handel mit den Pithoi im Schwarzmeerraum,"MiinsterscheBeitrdge zur antiken Handelsgeschichte17,

pp. 71-93. Johnson, G. 1982. "Organizational Structureand ScalarStress,"in Theory and Explanation in Archaeol-

ogy,C. Renfrew,M. J. Rowlands, and B. Abbott Segraves,eds., New York,pp. 389-421. Johnston,A. W. 1978a. Rev.of M. Lang, Graffiti and Dipinti, inJHS

68, pp.218-219. . 1978b. "Listsof Contents: Attic Vases,"AJA82, pp. 222226. . 1979. Trademarkson Greek Vases,Warminster. .1982. "TwoNumericalNotes," ZPE 49, pp. 205-209. .1984. "BritannicaIII: Pithoi," BICS 31, pp. 39-51. . 1990. "Aegina,Aphaia-Tempel XIII:The StorageAmphoras,"A4, pp.37-64. . 1996. "Fifth CenturyPrices," in I vasi attici ed altre ceramichecoeve in Sicilia (Atti del convegno internazionale) II, CronCatania 30,

1991 [1996], pp. 81-87. Johnston,A. W., and R. E.Jones. 1978. "The 'SOS' Amphora,"BSA 73, pp. 103-141.

Jordan,D. 1978. Rev.of M. Lang, Graffiti and Dipinti, in ArchNews 7,

pp. 92-94. c xsapuLx6 Kantzia,C. 1994. ""Eva 9 spyac-lQpLo

aOcqpopscov Too

ipco-Cou Itaob too 40o ao. m.X.

on'v KC),"in r 'Emrc^ZovIxr yia T57V 'EA?Evii-ctx ZvuvcvuVY(7 Kepcaxjix, (OEaaacaovixY,24-27 ZCErcEfpp(oo

1991), Athens,

pp.323-354. Kassel,R., and C. Austin, eds., 1984. PoetaecomiciGraeci,Berlin. Kerameikos IX = U. Knigge,Der Siidhiigel(KerameikosIX), Berlin 1976. Knigge,U. 1991. TheAthenian Kerameikos:History, Monuments,

Excavations,Athens. Koehler,C. 1978. "CorinthianA and B TransportAmphoras"(diss. Princeton University). -. 1986. "Handlingof Greek TransportAmphoras,"in Empereur and Garlan 1986, pp. 49-67. -. 1992. "Obshayatipologiyai khronoloiyakorinfskikh transportnykhamfor"[Brief typology and chronologyof Corinthiantransportamphoras],in Grecheskieamfory: Problemy razvitiya remesla i torgovli v antichnom mire

[Greek amphoras:Problemsof the evolution of productionand commercein antiquity],S. Monakhov and V. Kats,eds., Saratov,pp. 265-279. Koehler,C., and M. Wallace. 1987. "The Hellenistic Shipwreckat Ser9e Limani,Turkey:Preliminary Report.Appendix:The Transport Amphoras:Description and Capacities,"AJA91, pp. 49-57. Kurke,L. 1989. "KAIIHAEIA and Deceit: Theognis 59-60,"AJP 110, pp.535-544. Lang, M. 1956. "NumericalNotation on Greek Vases,"Hesperia25, pp. 124. . 1974. Graffitiin the Athenian Agora(AgoraPicBk 14), Princeton. Larsen,J. A. 0. 1938. "RomanGreece," in An Economic Survey ofAncient

Rome4, T. Frank,ed., Baltimore, pp. 334-414. Lawall,M. 1995. "TransportAmphoras andTrademarks:Imports to Athens and Economic Diversity in the 5th

GRAFFITI,

Century B.C." (diss. Universityof Michigan). . 1997. "Shapeand Symbol: Regionalismin 5th-Century TransportAmphoraProductionin NortheasternGreece,"in Tradeand Production in Premonetary Greece. Production and the Craftsman (SIMA-PB 143), C. Gillis, C.

Risberg,and B. Sjoberg,eds., Jonsered,pp. 113-130. . 1998a. "Ceramicsand PositivismRevisited:GreekTransport Amphorasand History,"in Trade, Traders,and theAncient City,

H. Parkinsand C. Smith, eds., London, pp. 75-101. . 1998b. "Bolsals,Mendean Amphoras,and the Date of the PorticelloShipwreck,"IJNA27, pp. 16-23. Lewis, D. M. 1979. Rev.of M. Lang, GraffitiandDipinti, in CR 29, pp.125-126. . 1987. "The Athenian Coinage Decree,"in CoinageandAdministration in theAthenian and Persian Empires(BAR-IS 343), I. Carradice,

ed., Oxford,pp. 53-63. Lombardo,M. 1988. "Marchands, transactions6conomiques,ecriture," in Les savoirs de l'e'critureen Grece

ancienne,M. Detienne, ed., Lille, pp. 159-187. Lowry,S. T. 1987. TheArchaeology of EconomicIdeas,Durham, N.C. LGPNI = P. M. Fraserand E. Matthews,A Lexiconof Greek PersonalNames I: TheAegean Islands, Cyprus,Cyrenaica, Oxford 1987.

Liidorf,G. 1998-1999. "Leitformen der attischenGebrauchskeramik: Der Bienkorb,"Boreas21-22, pp. 41-170. Mantsevich,A. 1975. "Keramichnatara z kurganuSolokha"[Ceramic vessels from the Solokha tumulus (Ukrainian)],Arkheologiya (Kiev) 17, pp. 72-85. Martin,T. 1985. Sovereigntyand Coinagein ClassicalGreece,Princeton. Maslennikov,A. A. 1987. "Graffitii dipinti s mysa Zyuk"[Graffitiand dipinti from Cape Zyuk], Kratkie soobzheniya InstitutaArkheologii 191,

pp. 45-53. Mattingly,H. B. 1981. "Coinsand Amphoras:Chios, Samos, and

WINE

SELLING,

AND

Thasos in the Fifth Century B.C.," JHS 101, pp. 78-81. . 1993. "New Light on the Athenian StandardsDecree (ATL II, D14)," Klio 75, pp. 99-102. . 1999. "What Are the Right Dating Criteriafor Fifth-Century Attic Texts?"ZPE 126, pp. 117122. Mattioli, S. P., ed. 1998. Bonifichee drenaggi con anfore in epoca romana: Aspetti tecnici e topografici.Atti del Seminario di studi, Padova, 19-20 ottobre1995, Padova.

Meijer,F., and 0. van Nijf. 1992. Trade, Transport, and Society in the Ancient World.A Sourcebook,London.

Oikonomides,A. N. 1986. "Graffitiinscriptionsfrom the Excavationsof the Athenian Agora at Kerameikos," Horos4, pp. 43-64. . 1988. "Graffiti-inscriptions from the Excavationsof the Athenian Agora at Kerameikos, PartII,"Horos 6, pp. 39-54. Oliver,J. 1977. Rev.of M. Lang, GraffitiandDipinti, in AJP 98, pp. 209-210. Olson, S. D. 1990. "Economicsand Ideology in Aristophanes''Wealth,"' HSCP 93, pp. 223-242. PAA= J. S. Traill,PersonsofAncient Athens, Toronto 1994. Papadopoulos,J. K., and S. A. Paspalas. 1999. "Mendaianas Chalkidian Wine," Hesperia68, pp. 161-188. Patitucci,S. 1991. "Spina,"in "Rivista di epigrafiaetrusca,"M. Cristofani, ed., StEtr 57, pp. 245-259. Peacock,D. P. S., and D. F.Williams. 1986. Amphorae and the Roman Economy:An Introductory Guide,

London. Preka-Alexandri,K. 1992. "ACeramic Workshopin Figareto,Corfu,"in Les ateliers depotiers dans le monde grec aux epoquesgeometrique, archaique, et classique (BCH Suppl.

23), F. Blonde andJ. Y. Perreault, eds., Paris,pp. 41-52. Pritchett,W. K. 1956. "The Attic Stelai, PartII,"Hesperia25, pp. 178-317. Ramon,J. 1991. Las anforaspunicasde Ibiza (Trabajosdel Museo arqueol6gicode Ibiza 20), Ibiza. Reger,G. 1994. Regionalismand Change in the Economy oflndepen-

REUSE

OF

AMPHORAS

89

dent Delos, 314-167 B.C., Berkeley. Riley,J. 1979. "The Coarse Pottery from Berenike," in Excavations at Sidi Khrebish Benghazi (Berenice) 2

(LibAnt5, Suppl.),J.A. Lloyd, ed., Tripoli,pp. 91-467. Roller, L. E. 1987. Nonverbal Graffiti, Dipinti, and Stamps (Gordion

Special Studies 1), Philadelphia. Rotroff,S. I., andJ. H. Oakley.1992. Debrisfroma PublicDining Placein theAthenian Agora (Hesperia Suppl.

25), Princeton. Salviat,F. 1986. "Levin de Thasos: Amphores et sourcesecrites,"in Empereurand Garlan 1986, pp. 145-195. Samothrace11.2 = K. Lehmann, The Inscriptions on Ceramicsand Minor Objects(Samothrace 11.2), New York

1960. Schonhammer,M. 1993. "Some Thoughts on the Athenian Coinage Decree," in Proceedings of the 11th International Numismatic Congress,

T. Hackens and G. Moucharte,eds., Louvain,pp. 187-191. Seaford,R. 1998. "TragicMoney,"JHS 118, pp. 119-139. Shear,T. L., Jr. 1975. "The Athenian Agora:Excavationsof 1973-1974," Hesperia44, pp. 331-374. . 1993. "The PersianDestruction of Athens: Evidence from Agora Deposits,"Hesperia62, pp.383-482. Solomonik, E. I. 1984. Graffiti s khory Khersonesa [Graffitifrom the chora

of Chersonesos],Kiev. Talcott, L. 1935. "AtticBlack-glazed StampedWare and Other Pottery from a Fifth CenturyWell," Hesperia4, pp. 477-523. . 1936. "Vasesand Kalos-Names from an Agora Well,"Hesperia5, pp.333-354. Thompson, H. A. 1955. "Activitiesin the Athenian Agora: 1954," Hesperia24, pp. 50-71. Tod, M. N. 1911-1912. "The Greek NumeralNotation,"BSA 18, pp. 98-132. . 1926-1927. "FurtherNotes on the GreekAcrophonicNumerals," BSA 28, pp. 141-157. . 1936-1937. "The Greek AcrophonicNumerals,"BSA 37, pp.236-258.

MARK

90

ancient cities of the northernBlack Sea], Moscow. van der Leeuw, S. E. 1984. "Dust to Dust: A TransformationalView of the CeramicCycle,"in The Many Dimensions of Pottery. Ceramicsin ArchaeologyandAnthropology,

S. E. van der Leeuw and A. C. Pritchard,eds., Amsterdam, pp.707-773. Vinogradov,Ju.G. 1986. "AN TOS HIOOS SHMHNHTAI: IGXII Suppl. 347,"in Empereurand Garlan 1986, pp. 197-200. von Reden, S. 1997. "Money,Law, and Exchange:Coinage in the Greek

portAmphorae. A Petrological and ArchaeologicalStudy(British School

at Athens, Fitch Laboratory OccasionalPaper4), Athens. Williams, C. K., II. 1979. "Corinth 1978: Forum Southwest,"Hesperia 48, pp.105-144.

Mark L. Lawall

DEPARTMENT 220

DYSART

WINNIPEG,

OF MANITOBA OF CLASSICS ROAD MANITOBA

R3T

CANADA [email protected].

CA

LAWALL

Polis,"JHS 117, pp. 154-176. Wallace,M. B. 1984. "Texts,Amphoras,Coins, Standards,andTrade," AncW10, pp. 11-14. . 1986. "Progressin Measuring Amphora Capacities,"in Empereur and Garlan 1986, pp. 87-94. WallaceMatheson, P. M., and M. B. Wallace. 1982. "Some Rhodian Amphora Capacities,"Hesperia 51, pp.293-320. Whitbread,I. K. 1995. GreekTrans-

Tolstoi, I. I. 1953. Grecheskiegraffiti drevnikh gorodov Severnogo Prichernomorya [Greek graffiti from

UNIVERSITY

L.

2M8

Wilson, J.-P. 1997-1998. "The 'IlliterateTrader?"' BICS 42, pp. 2956.

Yailenko,V. P. 1980. "GraffitiLevki, Berezanii Ol'vii"[Graffitifrom Leukas,Berezan,and Olbia], VDI (2), pp. 72-99. Yardeni,A. 1994. "MaritimeTradeand RoyalAccountancyin an Erased Customs Account from 475 B.C.E. on the Ahiqar Scroll from Elephantine,"BASOR293, pp. 67-78. Young, R. 1951. "AnIndustrialQuarter of Ancient Athens,"Hesperia 20, pp.135-288. Zeest, I. 1960. Keramicheskayatara Bospora[Ceramicvessels of Bosporos] (Materialy i issledovaniya po arkheologii 83), Moscow.

HESPERIA

69,

2000

Pages 91-o03

A

PERSONAL FOUND

ATH

IN

EN/AN

LETTER THE

AGCORA

ABSTRACT 1. Fora detaileddescription of the excavation,see Shear 1973. I am grateful to T. Leslie ShearJr.for permission to publish the letter presentedbelow, and to los Zervoudakifor permission to publishhere, for the first time, photographsof the Lokrianskyphos (Figs. 5-7). I am also indebted to J. Gordon Howie for his observations about the rhetoricalfiguresin the letter, to Alison BurfordCooper and SusanI. Rotrofffor adviceabout the skyphos, and to Jaime B. Curbera,Matthew W. Dickie, and Ronald S. Stroudfor their carefulcriticismof this paperin manuscript.Craig Mauzy is responsiblefor the excellenceof the photographsof the lead tablet (Figs. 2-4). Unless I note otherwise,all ancient dates are B.C. 2. Terracottaand lead military stamps:Krolland Mitchel 1980 and Kroll 1977b; cavalryroster:Kroll 1977a;Bacchic shouts:Jordan1986. 3. In a surveyof cursetabletsfrom the Athenian Agora, I mentioned seventeentabletsfrom the crossroads well (Jordan1985a, pp. 209-210), but that was before it had been recognized that the apparentBacchic shouts and text of the inscribedpot-menderwere not curses.For the text of one of these last, see AgoraXXVIII, pp. 55-57 (SEG XLIV 226). 4. For a useful introductionto the studyof letters on lead, see Bravo 1974. At Henry 1991, p. 65, there is a good laterbibliographyof texts. See also Jordan,forthcoming.

A lead tablet found in the Athenian Agora, presentedhere for the first time, preservesthe text of a personal letter inscribed in the 4th century B.C.by a professional.It is from a boy,his non-Attic name suggesting perhapsa metic background,who complains that the master of the foundry to which he has been sent, evidently as an apprentice,is mistreatinghim. The question raised concerning the legal rights of (metic?) apprenticesis of special interest, as is the locution ofthe text, at least partof which is in the boy'sown excitedwords. Excavation in 1972 of the well beside the orthostate shrine at the crossroads near the Royal Stoa in the Athenian Agora (Fig. 1) proved to be remarkablyfruitful, especially in its small finds of terracotta and lead, for the epigraphical record of the city.1Among the terracottapieces were several stamps, presumablymilitary passes, and among the lead were not only other military stamps but also the strips that have preserved for us much of the roster of the Attic cavalry.There was also a lead tablet inscribed with what seem to be Bacchic shouts.2The few other inscribed lead pieces from the well are the written debris of private rather than civic or organized religious life: a pot-mender and fifteen curse tablets of the 4th century, and the document that I present here, a personal letter of the 4th century, addressed by one Lesis to his mother and to one Xenokles (Figs. 2-4).3 Private letters on lead constitute a small but interesting genre within Greek epigraphy.4Eight have been published in full, a ninth in part. 1 Berezan, 5th century.Ed. pr. Vinogradov 1971 (phot., ills. 1-3; drawings, ills. 4-5); see Dubois 1996, pp. 50-55, no. 26 for later bibliography.Opisthographic. Side A: Inc. 'Q IHpoayop-Y, 6 7a-TYipT0o ?maTeXXs. Side B (its writing at right angles to that of A): 'AXLAGXo8c)po TO[toMilp8tov ropa TO,l 7ocAcaI x&vaoay6pYv.

2 Berezan or Olbia, 4th century.Ed. pr. Latyschev 1904 (drawing), Wilhelm 1909 (phot., fig. 64), cf. Syll.3 1260. See Dubois 1996,

DAVID

92

R.

JORDAN

Figure1. Crossroads,shrine,and well, with RoyalStoa at upperleft andPaintedStoa at upperright. Drawing by W. B. Dinsmoor Jr.

pp. 63-66, for later bibliography.Opisthographic. Side A: Inc. : TOS ?V OLXCOL XOlpeV :. 'ApTLXCOv

3 Torone, third quarterof 4th century.Henry 1991 (phot., drawing, p. 67). Fragmentary,upper and lower edges preserved.Inc. [-c4 ]To TeyoaL XicELV.

4 Attica (Xa'xsdycp),earlier 4th century.Ed. pr. IG 111.3,pp. ii-iii (drawing, p. ii); cf. Wilhelm 1904 (phot., figs. 50-51), Syll.3 1259. Opisthographic. Side A: Inc. Mvatoisyoc I ?F?CST?X? ToS; o'(xoLI xo(ipev xca6oyLauvevsI xai)aorciT oCsco; ?poaa [x]s [??Xv]. Side B (its writing at right angles to that of A): ?eIpsv cST6v x?patlov -cioy XupLx6vO I xCooOvaoL 8? Naoctiai I N pc)pacuxXL OuitotL.

5 Attica (Pnyx), 4th century. Raubitschek 1943, pp. 10-11 (phot., drawing, p. 11). Fragmentary,upper and left-hand edges ] (suppl.J. preserved.Inc. Omo? vac. IXcxipsv xat [ytLacivv and L. Robert, Bulltp 1944, no. 90). 6 Agathe (Agde in southern France), 4th century.J. and L. Robert, BullEp 1956, no. 357. Only partly published. Apparently a private letter, with the words xpai4esvxaot6[YLocLV?tv]. 7 Emporion, 5th century.Ed. pr. Sanmarti and Santiago 1987 (phot., pl. III; drawing, p. 122); later bibliography: SEG XLII 972, Slings 1994, Lopez Garcia 1995. Only the lower edge is preserved,which shows that the letter ended xalps. 8 Emporion, late 5th century.Ed. pr. Sanmarti and Santiago 1989 (phot., pl. I; drawing, p. 37; color phot., Nieto 1997, p. 147). Fragmentary,no edge preserved. 9 Emporion, 4th century.Almagro Basch 1952, pp. 33-35, no. 21 (phot., drawing, p. 35). Fragmentary,right-hand edge preserved.5 The extant lead letters are so few, Ju. G. Vinogradov argues, because the lead, once the letters were received, was reused for other purposes or dis-

5. I knowof threeunpublished examples,fromOlbia(5thor4th century,Vinogradov1971, p. 79; cf. Bravo 1974, p. 114), Mende in the

Chalkidike(4thcentury),andAthens (4th century).

A

6. Vinogradov1971, pp. 95-96; Bravo(1974, p. 113), contra,remarks, interestingly,that the reasonwhy we find no lead letters afterthe 4th century is that by then papyrus,itself obviously easierto handle than lead, had become more generallyaffordable. 7. The one clearexceptionis 8, of which the last preservedline ends yp-. Both left- and right-hand edges of the text of 3 arelost, but its editor, A. Henry (1991), without noticing the apparentscribalhabit of not dividing words at line-ends, has offeredquite plausiblerestorationsthat do not requiresuch division.The unpublished Athenian example(see note 5 above) has word-divisions. 8. Turner1968, p. 83: "The markof the public letter-writer,making articulate the fears and only half-expressed thoughts of his clients, is to be seen everywherein the platitudesand cliches of which so many private[papyrus] letters are composed." 9. For the chronologyof these conventions,see Threatte 1980, pp. 172-176,238-258. 10. To put it somewhatmore scrupulously,these featuresmay show no more than that our scribelearnedto .write in the earlierpart of the century. The senderof the letter was fairly young, but there is no reasonwhy he could not have employeda scribe considerablyolder.In other words, letter-formsand spelling need not force us to assumethe earlierratherthan the later 4th centuryas the actualtime of writing. 11. I regretthat in my tracing(Fig. 2) I have not been able to show any of this splitting.

PERSONAL

LETTER

FOUND

IN

THE

ATHENIAN

AGORA

93

carded:the examples that we have today would be chiefly those that were never delivered.6 All of the letters with published illustrations have writing that is crisp and sure and most of them have no word-divisions at line-ends within their main texts;71 and 4 do have word-division in the addresses on their backs.To B. Bravo this style of writing without word-division suggests the work of the professional scribe. Our letter too shows no word-division. The evidence that we have, beginning a few generations later, for letterwriting on papyrus among the Greeks in Egypt shows that it was usual to employ scribes to write one's letters.8Our relatively few examples on lead do not allow any generalization for Greek-speaking areas elsewhere or in earlierperiods,but the text of our letter (see below), like those from Berezan (1), Torone (3), and from Attica (4, 5), opening as they do in the third person, may well show a formulaic distinction between scribe and sender. The writing in the letter presented here is in a good 4th-century hand, the spelling, in every word but one, impeccable. The sounds /e/ and /o/are written e and o if they result from contraction (rCepLLouv[1], ?XOeV[1], oe&vo[2]) or compensatory lengthening (6s; [2]), while only inherited diphthongs appear as ?t (6lorT?AsL:

.SVOxXAs[1],

X LXeOiL [2]; the text

has no example of a word that might be spelled with oo).9 A date in the earlier 4th century is indicated.10 The tablet is apparentlyopisthographicand is unusuallywell preserved, with all edges intact except for a slight tear near the right-hand end of Side A. The edges themselves are not flat and straight but crinkled and have traces of pinking that suggest the use of shears ratherthan a knife for cutting the lead sheet. The main text (Side A) is in four lines. Splits within the individual strokes of the letters, particularlynoticeable at the beginning of line 1, suggest the use of a splaying reed stilus (Figs. 3-4); experiment shows that fresh thin lead sheets are in fact soft enough to be inscribed with reeds. The purpose of such splits in stili evidently being to control the flow of ink, we may speculate that the scribe normally wrote not on lead but on papyrus or some other material that accepted ink, such as a potsherd, and that he used the instrument that he had at hand, a stilus that had alreadybegun to show wear. He seems to have sharpened it from time to time, at any rate: there are no splits in the strokes of QI in ANOPQIIQI in line 3 or in the strokes in line 4.11 The tablet was rolled up, with the writing on the inside (Side A, Fig. from the direction of the left-hand ends of the lines of the text; this no 2), doubt explains the slight damage at the right, which would have been the outer flap of the resulting scroll. That area of the outer side (i.e., Side B) shows, some 0.05 m from and parallel to the right-hand edge of Side A (i.e., at right angles to its writing), a row of faint scratchings, with shapes slightly larger than the letters of Side A; there may have been more. That they are neatly aligned suggests that they are deliberate, but the area is corroded, encrusted, and badly worn, this being the exposed part of the scroll, and I cannot confidently identify any one cluster of scratchings as a letter. If the marks are deliberate, they are no doubt the remains of what was once an address;indeed 1 and 4 each has an address on its back, in-

KAIeI EP gE 3 A T TI

E

XAAl k YT

RI/AvAH/

A

KAIE E P

? I

10cm

5 ,

,

,

I

Figure2. AgoraIL 1702, SideA

I

l

A

PERSONAL

LETTER

FOUND

IN

THE

ATHENIAN

AGORA

95

scribed, as here, on the exposed flap and at right angles to the writing on the other side. A final word, before we turn to the text, about the findspot, the well beside the orthostate shrine. In the Classical and Hellenistic periods the shrine evidently received so many votives-more than any other in the Agora-that they must have had to be cleared out from time to time and discarded, for the well itself had thousands of such votives, most of them suggesting a female cult. Some of the inscribed objects listed at the beginning of this article are obviously appropriate as offerings and may have come from the shrine: for example, the tablet with the Bacchic syllables, whatever its precise significance, and, if the deity is chthonian, the fifteen curse tablets.12Other inscriptions were clearly not votives: the cavalrytablets and the military stamps evidently come from the Hipparcheion, which has been assumed by C. Habicht to have been located nearby.13From the text of Lesis's letter it should, I think, be obvious that if that tablet was in any sense in situ when it was found among the votives in the well, it was so only in that the person whom Lesis had commissioned to deliver it considered it simpler to drop it down the well or into the shrine than to complete the errand.The letter was not, in other words, a votive at the shrine. This I stress only because, when I was preparing this article, I showed the text of the letter to a friend specializing in magic who, thinking of the texts of certain lead curse tablets with quasi-epistolary formulae addressing the dead,l4 at once assumed Lesis's letter to be a curse tablet, Xenokles the ghost addressed, and the "mother"the Earth herself. It is not; it is a personal letter that was never delivered. Agora inv. IL 1702

Figs. 2-4

H. 0.050, W. 0.134 m Early fourth centuryB.C. Side A 1 AoatL {I }smLTSSl 12. In all, sanctuariesof Demeter and of Demeter and Korehave yielded some forty cursetablets,from Selinous (Jamesonet al. 1993, pp. 125-131, late 6th[?]-5th century);Rhodes (Zervoudaki 1973, p. 622; cf.Jordan1985b, p. 168, 4th century);Mytilene (Curbera andJordan1998, 4th century);Knidos (DTAud1-13, Hellenistic);Morgantina (SEG XXIX 927-932, Hellenistic); and Corinth (CorinthXVIII.3, pp. 281291, Roman Imperial?). 13. See Kroll 1977a, p. 84; Habicht 1962, p. 138. 14. For a helpful discussionof such "letters"see Bravo 1987, pp. 196-200.

L i0 ESVOXXEIxaclt T- oupYt

1

.ooq tc7

2 aoTOv 6a7coX6ji?Vov v -zr xacXxEico, AXXOapo5 t6ogoscr6-zTac a0oTO?XOEV 3 xoci ?V?Opi?Oo -lT PXTl-noocvTCzL. yoc 'AvOpcTjcOL TOvO OraOCPOa,8O[L

4

7CVrO OVYpcOL

8?8?Lou. 7cpO7rxaOCxxo,tOcC- XuocAOV oaooTLYO,6Usvo5 Xuo[ECL6oc6?E jia[X]Xov.

"Lesis is sending (a letter) to Xenokles and to his mother by no means to overlook that he is perishing in the foundry but to come to his masters and find something better for him. For I have been handed over to a man thoroughly wicked; I am perishing from being whipped; I am tied up; I am treated like dirt-more and more!"

DAVID

96

R.M

R.

JORDAN

'G

Figure3. AgoraIL 1702, SideA, left end, showingeffectsof splayingof reedstilus

Figure4. AgoraIL 1702, SideA, detailof lines 2-4. The stilushas been sharpenedfor line 4. ALatS { Lt}. The name AaoLSis new and certainly not Attic. Like the masculines A6olt/Aoaicx/ADuGov from XA6o,ZEvitL/ZtocL^q/ZFevt(A)V from (poioVf2L, and so on, it is from rseyvDuJE, 8t;/OsL/OLt&ax/DJsicov no doubt XAC(= Att. 0iXco), in this case a from formed verb, evidently attested in Doric and probably that underlying the name Aqaociac,even though the apparently single instance of this last, IG V.2 439.31 (Megalopolis, 2nd century), is from Arcadia. Further, the masculine ending -Lc being preponderantlyDoric,15a Doric origin of the name AnaLScan probably be assumed.We may compare the Doric name BcoXt, from P3cOAXo[vc (= Att. 3o6X-), a near-synonym of AXj.As for the superfluous {it }, it conceivably results from the unfamiliarity of the name Alljat in Athens. ?7laT?XX?t. We may compare the use of the verb in the opening for-

15. See Masson1987,p. 246.

A PERSONAL

LETTER

FOUND

IN

THE

ATHENIAN

AGORA

97

mula of 4, as a parallel to which W. Crdnertl6 adduced Selene's lines in Aristophanes' Clouds(608-610): 'H

OVTuO( JjiV X?jVY

7rpcrcoaViv 16. Crbnert1910. 17. Cf. also the opening of 1. The preservedbeginnings of 2, 3, 5, and 6 use Mnesiergos'sand Selene'sverb but are more like the formula XOapLpLv

familiarfrom papyrusletters,with its -xotpsLv nominative-plus-dative-plus (for examplessee Exler 1923, pp. 2444, 50-56). 18. See Jordan1978, p. 93, for the articulationof the last line. 19. A readerof an earlyversionof this articlemade a suggestion:"Could Xenokles not be Lesis'smaster,who has the boy'sslave mother in his oikos? Xenokles rents out Lesis to work in a foundrywhere he is abused.Lesis writes to his master(and to his mother who may here have some influencesexual?-over Xenokles) to get him out of there.Xenokles may want to protect his investmentby finding the boy something better."As I see it, however, againstthis interestingidea is Lesis's requestthat bothXenokles and the mother should make the new arrangements with the foundry-keepers. 20. See Harrison 1968, pp. 189-193, on the relationshipof metic and prostates.The evidence on the metic at law is enhancedand complicatedby F. Costabile'srecentpreliminarypublication (1998) of an early-4th-century lead cursetablet from the Athenian Kerameikos,directedagainstthree personswith forthcominglawsuits;the curseon the third, a woman, includes "thepolemarchand his court." 21. From Roman Egypt we have a numberof apprenticeshipcontracts(see below), some in fact with penalty clausesshould the masters/teachers default,e.g., P Mich.IV 346 (Tebtunis, A.D. 13), in which a masterweaver writes, "Andif I shall not teach her,or she shall be considerednot to know what she has been taught,you will perforcehave her taught at my own expense"(lines 9-12). 22. E.g., Goodwin 1889, p. 51, ?148; Kiihner1904, pp. 55-56.

?7?7T?sL?V(pp aOcl

XoaLpLV'AOYlvoaoio7L xaTito5

jt[L0XOL;S

si-rs Ou[icLV?v EpayOxs.

The similarity suggests that the formula with which Lesis's letter opens would have been familiar in Athens.l7 An opening of a message with the verb ?LOacT?X?EV is indeed found on an earlier graffito from the Agora, with the message ooioV?0<;> I ?T?CaT?X?eI rFXcoxoi I ?; actvcoIcv 8asc6[L] (AgoraXXI, p. 9, no. B9, last quarterof 5th century).l8 6SVOXAXsL xOli Tui -YTpL0.Lesis complains to Xenokles and his mother, as we see, of mistreatment by his masters. His request that together with the masters they should "find something better for him" (line 3) implies a previous agreement. We may at the outset rule out the possibility that Xenokles is Lesis's father, for if so Lesis's mother has no legal standing in the negotiations with his 8s7zoTouain the xocAxxeov.Lesis's father, we have to conclude, is no longer on the scene or in a position to conduct transactions. In deducing the relation between Lesis and Xenokles we should begin by asking if Lesis is a slave or free. If he is a slave, Xenokles must have contracted the apprenticeshipas his owner or conceivably as his owner's representative,to whose compassion, in either case, the boy may, let us concede, be in a position to appeal.If he is a slave,though, his mother can have no legal part in the arrangement with the 8oacr6zoC. Yet Lesis writes to her as well as to Xenokles and, if the letter expresses his instructions accurately,asks them both to come to the 8canTt6oouand to make a new arrangement.It seems to me therefore unlikely that Lesis is a slave.19 I cannot determine, from what we know of Athenian law, whether Lesis and his mother are of the citizen or the metic class. His non-Attic name points to the latter,in which case Xenokles is presumablythe mother's prostates, her citizen patron, who acts on her behalf in the polemarch's court if she ever needs to undertake a legal action,20for example, if Lesis's masters should default on their part of the arrangements;21Xenokles will have been present at the negotiation, then, as prostates. I aVxTOVdCOXO;6Esvov.Spelling at the time would LuYI8a,j;7cEpLU8MV have been the same for a future participle (&7roXo6ojV?VOV, contracted from -X?0j6-) and an aorist (-X6oj-).That standardgrammars,on the other hand, give several examples of the verb CepLOpaco0with present or aorist but never future participles no doubt reflects usage, and the use of the present below probably argues against a future participle here.22In 6cu6XXoVEoul at any case, we may compare the locution jx rsp11l8siEVo7UOO[vi?VOO; Demosthenes 50.5. iv TOt XaoxsxLCt.A xapXxs?ov,its etymology notwithstanding, need not be a foundry specifically for bronzeworking;we may compare the latitude of the use of the word xoaXx:S;,which could refer to a coppersmith but also to a goldsmith (Od. 3.343), an ironworker(cf. Arist. Poet. 1461a29 xoaxxaxgTo -cob; -OvuT Rov ?pyraoji?vou;; Gen. 4.22 LXX ocXxsbg; xocXxo xaocti acOipoo-examples from LSJ).We need not assume that the

98

DAVID

R.

JORDAN

letter was actuallysent from the xocaXxeovor from anywherenear the crossroadswell. In any case, Lesis is unlikely to have worked in any of the three excavated bronze foundries in use in the Agora in the 4th century (see below, note 25). I xaot veopeaaxL TLPeX3rLov ocXo i Tp6O ?CX.Ov s;to snro6-oc; ocvto3 The word 8asnT06-o means "master," ocb-coL. usuallyin the sense of "owner," s.v. in the evidence assembled to LSJ and DiccGE. If this were according the sense here, with Lesis literally belonging to these 8arndcxToa,it would have been pointless for him to write to his mother, Xenokles, or anyone else for help. Here the word evidently has a looser sense. His situation may remind us of Lucian's,in the 2nd century of our era, whose father, not a rich man, sent him to the shop of his wife's brother, a sculptor of some reputation, to learn the art. Lucian'swords (Somn. 3) are opasE6o8eov -cT 0?t , not much different from Lesis's avOCpcj OLt. . . apo3aSe8oLalC below; conceivably the verb was a terminustechnicusfor such occasions. The uncle gave Lucian the chisel and a slab of marble and told him to strike. Lucian did, too enthusiastically,and broke the slab,whereupon the uncle beat him with a stick. Lucian went home in tears and complained to his mother, who heaped abuse on her brother, no doubt using such terms Here the similarities end, though. Lucian was able to as TCOCVU 7rOVYpo6;. go home to complain; Lesis has to write to Xenokles and to his mother. VVhatis Lesis's status vis-a-vis his masters? We have virtually no direct evidence for the institution of apprenticeship in 4th-century or earlier Athens-in fact, other than the "manumission cup" inscriptions (IG II2 1553-1578, SEG XVIII 36, all from the 320s), the present letter is the only real testimonium. I quote from Alison Burford'suseful summary: Few details are known of the apprenticing system in Greece.... Indeed we know little beyond the fact that it existed, as it had in earlier urban (and to that extent industrialized) societies. There is a reference to'competent didaskaloi'-instructors-for 'the lowest mechanics' in Xenophon's Memorabilia[4.2.2], and Plato [Meno 91D] makes a comparison between the master of rhetoric and his fee-paying pupil, and the sculptors Pheidias and Polykleitos and their fee-paying pupils; a little further on he refers to Zeuxis of Herakleia who gave painting lessons. A formal agreement of some kind was made between the parents or guardians of the prospective apprentice and the master-craftsmanwho had consented to take him on; the only surviving records of such agreements, from Hellenistic and Roman Egypt, suggest that the terms would have been of the simplest; although the political and economic organization of Egypt differed in many ways from that of the rest of the ancient world, the conditions of apprenticeship cannot in the nature of things have varied much.23 For 4th-century Attica, we may go a bit farther than this. In the "manumission cup"inscriptions a slave, once freed, receives a civic nomenclature olxovac (SEG XVIII in the form DiLaor-TaXaotr(ouvpyoc;) isl MlsX(TYrL)

23. Burford1972,pp.88-89.

A

PERSONAL

LETTER

FOUND

IN

THE

ATHENIAN

AGORA

99

36 A, lines 213-214) or'ApLaTo[ti?vY5 'i M?X(LT&) Ioixciv OXuTOTO(oCS) (lines 216-217), the mention of the profession being apparentlyoptional. The professions, all of which were necessarily learned before the manumissions, are from a wide spectrum of Attic economic life. The a7TSXV60?pOl include wool-workers,farmers,cobblers,shopkeepers,merchants, and also persons with skills that could hardly have been gained in the masters'households but would have required apprenticeship elsewhere: a gem-cutter, a goldsmith, a citharode, a flute-player, a potter, another specializing in amphoras, a xcXxe6v. In the papyrus contracts we see that owners could apprentice their slaves, guardians their wards, parents their sons and daughters to learn professions, and that there were generally two types of contracts: apprentices either worked by day and came home at night or remained with their teachers/masters until the term of instruction (which could be as long as several years) ended;24Lesis had to send a letter home, which shows that his contract was of the latter type. 'AvOpCc7oLyOcp7ocpa?8ovuaoi O'Cvu TovYpCL.In contrastto the plural 8EaTC6zatof line 2, with whom Xenokles and Lesis's mother have made the arrangement,stands the singular here, referring no doubt to the man in charge of some particularbranch of work within the aocXxelov. It is he to whom the 8osc6rotaoof the xCarkxsovwill have handed Lesis over for personal instruction.25 aGT6lyosevo; uc6AuXXDoca 8a L 7oxp7rXooaxtioua. The Old Oligarch speaks of the Athenian sanction on striking a slave or a metic ([Xen.] Ath. Pol. 1.10).The evidence is contradictory,though, for at [Dem.] 53.16 the speakertells of his neighbors' ruse of sending a young Athenian boy (cax8cptLov o cot6v) to destroy a rose-bed, "in order that if I caught him and in a fit of anger put him in bonds or struck him, assuming him to be a slave (ratocatl L kc 8oiXov elval), they might bring an indictment for assault(ypacp9vtp3scou;)"(transl.A. T Murray);presumablythe speaker is not awareof the prohibition that the Old Oligarch refers to. The matter of the ypacvcp' 3ppco;is indeed complex.26Whether the law deterred or was meant to deter corporalpunishment of young persons during apprenticeships is a question for which there is no evidence, as far as I know, to guarantee an answer.Severalcenturies later,the question of the legal rights of the apprenticewas still debated by Roman jurists. Burford27refers to an opinion of Ulpian (3rd century A.c.) in Justinian'sDigest, 9.2.5.3:

24. For the evidencefrom GraecoRoman Egypt for apprenticeshipthere, see Westermann1914; Zambon 1935, 1939. 25. If this idea is right, the letter concernsnone of the 4th-century bronze foundriesin the Athenian Agora,which are too small for such divisionsof labor;see Mattusch 1977,

pp.358-363. 26. See Fisher 1992, pp. 36-85. 27. Burford1972, p. 92.

If a teacher kills or wounds a slave during a lesson, is he liable under the lex Aquilia for having done unlawful damage?Julian writes that a man who put out a pupil's eye in the course of instruction was held liable under the lex Aquilia. There is all the more reason therefore for saying the same if he kills him. Julian also puts this case: A shoemaker,he says, struck with a last at the neck of a boy (a freeborn youngster) who was learning under him, because he had badly done what he had been teaching him, with the result that the boy's eye was knocked out. On such facts, saysJulian, the action for insult does not lie because he struck him not with intent to insult, but in order to correct and to teach him; he wonders whether there is an action for breach of contract for his services as a teacher, since a

I00

DAVID

R.

JORDAN

teacher only has the right to administer reasonable chastisement, but I have no doubt that action can be brought against him under the lexAquilia.2 As an instance of the brutality that could take place in a workshop in Greece, Burford cites a scene on a late-5th-century imitation black-figure skyphos found at Abai in Lokris (National Museum, Athens, inv. 442); the painter is conceivably relying on first-hand knowledge of life in the pottery works.29He depicts an assistant,a stack of three skyphoi with horizontal handles in his right hand, walking away to the left (Fig. 5). In the middle of the scene a seated person of undetermined gender,30the only clothed figure in the picture, holds a triangular object (a fan?)31in one hand, a kylix in the other.To the left of the stool at which the figure sits is another stack of three skyphoi, these with upturned handles. Farther to the right there is a shelf, presumably on the back wall, with a kantharos, another horizontal-handled skyphos, and two small sticks that may be paintbrushes(Fig. 6). On the floor is a potter'swheel, behind which crouches another worker, holding in his raised left hand a skyphos with upturned handles that he inspects (Figs. 7-8). Standing on the wheel, within easy reach of his right hand, is a horizontal-handled skyphos with what is evidently a paintbrushin it. The scene is obviously a pottery workshop. Above the painter who sits peacefully at the wheel we behold something gruesome. No doubt as punishment for some wrongdoing, one of the workers has been suspended face down from the ceiling. His left foot is tied against the ceiling itself; his right foot hangs lower, from a cord. His hands also hang from cords.32Another cord from the ceiling is around his neck, strangling him so badly that his tongue hangs out.33As if this were not enough, another cord, attached to his penis, is stretched tight and tied to a ring or a hook in the floor. In front of the victim and facing him stands another worker swinging a long, thin object, a leather thong or a stick, in his upraised right hand. Bliimner and all others who have discussed the figure suspended from the ceiling have assumed that he is a slave, Bliimner adducing, from Roman comedy, several examples of slaves being whipped while tied up and suspended.34(As an Attic example we have a lively description by Aristophanes, Frogs 618-822.) But is this assumption necessary?May he not be an apprentice-slave, metic, or free-here in statupupillari? The second sentence of the letter, with its first-person verbs, is in contrast to the first, with verbs in the "epistolary"third person. It is evidently intended to be read as Lesis's own words. The verbs in the last line are in an almost staccato asyndeton. We may compare the asyndeton of the SpartanHippokrates'letter that the Athenians interceptedin 410 (Xen. Hell. 1.1.23), Epp?l T-ox&XaXcMivsapo; CasCoo x'aC 7r?lv&v)nTcijvspE;offers as an example of a vivid TL and what Aristotle ncop(o|ics Xjp'1 8ap&v, peroration at the very end of his Rhetoric: 'prjxa- a0xYrx6aXTZ- tCErE xptivocs. This latter is itself an echo of the end of Lysias'sAgainst Eratosthenes: xCx6care- op6xcaTrTC?C06v0aT?ocJ. Gordon SXE"T?- tx6OCZTs. these to me to who has been kind Howie, enough point passages, concludes that the composer of the second sentence of Lesis's letter must have

28. Trans.C. Kolbertapud Mommsen, Krueger,and Watson 1985, vol. 2, p. 278. 29. Burford1972, p. 91; Bliimner 1889, pp. 150-156. Subsequent discussions:Collignon and Couve 1904, no. 1114 (phot. of one side, pl. 38); Walters 1905, pp. 135-136; Scheibler1995, p. 120; Halm-Tisserant 1998, pp. 44-45. 30. A man, accordingto Bliimner, Collignon and Couve, and Walters;the wife of the naked man at the wheel (see below), accordingto Scheibler;the female owner of the shop, accordingto Halm-Tisserant. 31. Bliimner and Waltersdescribeit as a leathersheet, with which the seated figurewas in the act of striking the departingassistant;for Collignon and Couve it is a rhyton.That it may be a fan is the suggestion of Susan Rotroff. 32. Bliimner,Collignon and Couve, and Waltersassumethat the right hand is free and is begging for mercy.There are two cords,and admittedlyneither is attachedto the right hand, but the artistryis crude,and I think that we are meant to understandboth hands to be bound. 33. Halm-Tisserant(1998, pp. 4445) alone assumesthat the victim is vomiting into the skyphosthat the man at the wheel is obliginglyholding up. The man at the wheel seems not to have envisagedthis possibility. 34. Bliimner 1889, p. 156.

A

Figures 5-7. Skyphos from Abai in Lokris, late 5th century, depicting a potter's workshop with a torture scene. Athens, National Museum inv. 442. CourtesyMuseum Figure 8. Watercolor showing potter's workshop and torture scene. CourtesyNational Museum, Athens

35. For discussion, see Arnott 1996, p. 126.

PERSONAL

LETTER

FOUND

IN

THE

ATHENIAN

AGORA

IOI

had some education in rhetoric and therefore would not be Lesis himself but presumablythe professional scribe. I myself would keep open the possibility,however, that Aristotle recommended asyndeton precisely because persons in such situations as those of Hippokrates or Lysias or our Lesis would naturally,and effectively, use it to urge attention. The last expression is otherwise known only from [lOaXtov &la[X]Xov. c t ox6acpo;) and a 1406, (IT Euripides E[ai&ov8' [.axXov 7pocg c&cpac that Photios cites s.v. from Alexis'sAtthis, (Lex., passage tai&Uov [LakXov) fr. 29 Kassel-Austin (rc;S c.vicpsL T6"rpcTrov 6 ZVS ]o0Xy, / CSETaoc Without quoting his evidence, Photios refersto Alexis's llaXov [L&aXov).35 Pezonike,Anaxilas's Horai, and Menander for the idiom. The expression evidently went out of use well before it found a firm place in the language, avrt ToOocEl though, for Photios is obliged to explain it: aiXXov &atXXov xao

XyouCOLVaveu TOb xca auv8aeou. CdAAov,or again, oUTcwO;

DAVID

102

R.

JORDAN

REFERENCES Agora XXI = M. Lang, Graffiti and

Dipinti (AgoraXXI), Princeton 1976. AgoraXXVIII = A. L. Boegehold, The Lawcourts atAthens: Sites, Buildings, Equipment, Procedure,and Testimonia (Agora XXVIII), Princeton 1995.

Almagro Basch, M. 1952. Las inscripcionesampuritanas griegas, ibericasy latinas (Monografias

ampuritanas2), Barcelona. Arnott, W. G. 1996. Alexis: The Fragments. A Commentary, Cam-

bridge. Bluimner,H. 1889. "Scenendes Handwerkes,"AM14, pp. 150-159. Bravo,B. 1974. "Unelettre sur plomb de Berezan:Colonisation et modes de contact dans le Pont,"DHA 1, pp. 110-187. . 1987. "Une tablette magique d'Olbia pontique,les morts,les heros et les demons,"in Poikila:Etudes offertesaJean-Pierre Vernant

(Recherchesd'histoireet de sciences sociales26), Paris,pp. 185-218. Burford,A. 1972. Craftsmenin Greek and Roman Society,London. Collignon, M., and L. Couve. 1904. Catalogue des vases peints du Musee National diAthenes (BEFAR 85),

Paris. Corinth XVIII.3 = N. Bookidis and R. S. Stroud, The SanctuaryofDemeterand Kore: TopographyandArchitecture (Corinth XVIII, pt. 3), Princeton

1997. Costabile,F. 1998. "Latriplicedefixio del Kerameikosdi Atene, il processo polemarchicoed un logografo attico del IV sec. a.C. Relazione preliminare,"Minima epigraphica et papyrologica1, pp. 9-54. Cronert,A. 1910. "Die beiden altesten griechischenBriefe,"RhM n.s. 65, pp.157-158. Curbera,J. B., and D. R.Jordan.1998. "CurseTabletsfrom Mytilene," Phoenix 52, pp. 31-41. DiccGE= F. R. Adrados,ed., Diccionario griego-espaniol,Madrid 1980-. DTAud = A. Audollent, Defixionum tabellae quotquot innotuerunt ..,

Paris 1904. Dubois, L. 1996. Inscriptions grecques dialectalesd'Olbia du Pont, Geneva.

Exler, F. X. 1923. The Form oftheAncient GreekLetter, Washington, D.C. Fisher, N. R. E. 1992. Hybris:A Study of the Valuesof Honour and Shame in Ancient Greece,Warminster. Goodwin, W. W. 1889. Syntaxof the Moods and Tenses of the Greek Verb,

London. Habicht, C. 1962. "Neue Inschriftenaus dem Kerameikos,"AM 76, pp. 127148. Halm-Tisserant,M. 1998. Re'aliteset imaginaire des supplicesen Grece antique (Collection d'etudes anciennes. Serie grecque 125), Paris. Harrison,A. R. W. 1968. The Law at Athens 1: The Family and Property,

Oxford. Henry,A. 1991. "ALead Letter from Torone,"ArchEph, pp. 65-70. Jameson,M. H., D. R.Jordan,and R. D. Kotansky. 1993. A lex sacrafrom Selinous (GRBM 11), Durham, N.C. Jordan,D. R. 1978. Rev. of Agora XXI, in ArchNews 7, pp. 92-94.

. 1985a. "Defixionesfrom a Well nearthe Southwest Corner of the Athenian Agora," Hesperia 54, pp. 205-255.

1985b. "ASurveyof Greek ----. Defixiones Not Included in the Special Corpora,"GRBS26, pp. 151196. . 1986. "Ululationsfrom a Well beside the PanathenaicWay,"AJA90, p. 212 (abstract). . Forthcoming. "EXXn'vtx'i ECzoXoYopccpia: IpIzCcs

in A.-Ph. ?mypoc(pLx?< pacpT-oupLe," Christidis, ed., EyxoxAortaciE'ca T7 7; ekh7qvIKx,y6'caa5,, laTopac9 fl7

Athens. Kroll,J. H. 1977a. "AnArchive of the Athenian Cavalry,"Hesperia46, pp. 83-140. . 1977b. "SomeAthenian Armor Tokens,"Hesperia46, pp. 141-146. Kroll,J. H., and F.W. Mitchel. 1980. "ClayTokens Stampedwith the Names of Athenian Military Commanders,"Hesperia49, pp. 86-96. Kiihner, R. 1904. AusfuihrlicheGrammatik dergriechischen Sprache2, Hanover.

Latyschev,V. V. 1904. "Inscriptions Found in SouthernRussia, 1901-

A PERSONAL

LETTER

FOUND

IN

1903," Izivestiya arkheologicheskoy komissii Rossiyskeyakademii nauk

[Bulletin of the Archaeological Commission of the RussianAcademy of Sciences, St. Petersburg]19, pp. 1-91 (in Russian). L6pez Garcia,A. 1995. "Nota sulla letturadi piombo da Emporion," Tyche10, pp. 101-102. Masson, 0. 1987. "Remarquesd'onomastiquecyreneenne:Quelques noms masculinsen -iS," QAL12, pp.245-248. Mattusch, C. C. 1977. "Bronze-and Ironworkingin the Area of the Athenian Agora,"Hesperia46, pp.340-379. Mommsen, Th., P. Krueger,and A. Watson. 1985. The Digest of Justinian, Philadelphia.

Nieto, X. 1997. "Le commercede cabotageet de redistribution,"in La navigation dans l'antiquite, P. Pomey,

ed., Paris,pp. 146-159. Raubitschek,A. E. 1943. "Inscriptions," in G. R. Davidson and D. B. Thompson, Small Objectsfrom the Pnyx I (Hesperia Suppl. 7,

Princeton),pp. 1-11. Sanmarti,E., and R. A. Santiago.1987. "Une lettre grecquesur plomb trouveea Emporion (fouilles 1985)," ZPE 68, pp. 119-127. . 1989. "Une nouvelleplaquette de plomb trouveea Emporion,"ZPE 77, pp.36-38.

THE

ATHENIAN

AGORA

Scheibler,I. 1995. GriechischeTopferkunst: Herstellung, Handel und Gebrauchder antiken TongefaJfe,Munich.

Shear,T. L., Jr. 1973. "The Athenian Agora:Excavationsof 1972,"Hesperia 42, pp.359-407. Slings, S. R. 1994. "Notes on the Lead Letters from Emporion,"ZPE 104, pp. 111-117. Threatte,L. L. 1980. The Grammarof AtticInscriptionsI: Phonology,Berlin. Turner,E. G. 1968. GreekPapyri. An Introduction,Princeton. Vinogradov,Ju. G. 1971. "AGreek Letter from Berezan,"VDI 118, pp. 74-100 (in Russianwith English summary). Walters,H. B. 1905. HistoryofAncient Pottery: Greek,Etruscan, and Roman I,

London. Westermann,W. L. 1914. "Apprentice Contractsand the Apprentice System in Roman Egypt,"CP 9, pp. 292-315. Wilhelm, A. 1904. "Der alteste griechischeBrief,"JOAI7, pp. 94105. . 1909. "Der Brief des Artikon," JOAI 12, pp. 118-126. Zambon, A. 1935. "Ataoxaocxxoci," Aegyptus15, pp. 3-66. .1939. "Ancorasulle &8LaOCXOCxLOA,"l Aegyptus 19, pp. 100102. Zervoudaki,I. 1973. "'ApXocaO6TYcsT xcci ArchDelt28 v[I,isLocAco8?xacvyiaoU," B, pp. 608-642.

DavidR. Jordan AMERICAN 54

SCHOOL

SOUIDIAS

106-76

STREET

ATHENS

GREECE [email protected]

OF CLASSICAL

I03

STUDIES

AT ATHENS

HESPERIA

69,

2000

Pages 105-132

ICAL

UNCANON

PORTRAITS

IMPERIAL IN

THE

ROMAN THE

CASE

EASTERN

PROVINCES OF THE

KANELLOPOULOS

EMPEROR

ABSTRACT 1. This articlegrew out of a paper originallypresentedat the Archaeological Institute of AmericaAnnual Meeting in Atlanta, Georgia,in 1994. I am indebted to many scholarsfor their contributionsand advice, especiallythose who readthis articlein earlierstages,includingAlan Shapiro, Andrew Stewart,Fred Kleiner,and the anonymousreviewersfor Hesperia. Their helpful comments and insightful suggestionscaused me to rethinkand reworkthis articlesignificantly,and I am deeply grateful.Thanks also to all who suppliedphotographs:Maria Andreadaki,VannaNiniou-Kindelis, and George Skoulas,Chania Museum; Mari Aurenhammer,Kunsthistorisches Museum, Vienna;Unal Demirer, AntalyaMuseum; Roland Etienne, Ecole francaised'Athenes;Axel Filges, Deutsches ArchaologischesInstitut, Istanbul;Ev. Giannoussaki,National ArchaeologicalMuseum, Athens; Helle W. HorsnaesandJorgen Steen Jensen, National Museum, Copenhagen;Jan Jordan,American School of Classical Studies at Athens, Agora Excavations; Ilona Trabert,StaatlicheMuseen zu Berlin-PreussischerKulturbesitz Antikensammlung,Berlin;BekirTuluk, Efes Miizesi, Seljuk;Deutsches ArchaologischesInstitut,Athens; StaatlicheAntiken Sammlungen, Munich; and William E. Metcalf, AmericanNumismatic Society. 2. Pekary1985, pp. 81-83. 3. It appearsto be Pentelic;Dontas 1975, p. 521.

The identity of the imperialportraitin the KanellopoulosMuseum in Athens has baffledmodernviewers.The portraitlacksan inscriptionandthe provenance is unknown, although it is probablyfrom the Roman East. The portraitbearsthe imperialattributeof a coronacivica yet its featuresdo not closely resemble those of any emperor.Comparison with other provincialimperial portraits in sculpture and on coins reveals that deviation from Roman canonical types is common, a finding also supportedin the ancient literature. Stylistic analysisplaces the portraitin the early2nd century,so the Kanellopoulos Emperoris probablyTrajan. In the collections of the small but impressive Kanellopoulos Museum, located on the northern slopes of the Acropolis in Athens, an imperial portrait is displayed (Figs. 1-4).1 The identity of the figure cannot be easily determined by conventional methods. The imperial status of the image is secure.The portrait, measuring 35 cm in height, is greater than life size, a common characteristicof Roman imperial portraits.2In addition, the portrait is of very high quality,both in material and carving. It was sculpted from a good block of finely grained white marble.3The eyes, now missing, were originally inlaid in a contrasting material that would certainly have added a richness to the sculpture and no doubt also provided it with a sense of liveliness, now entirely absent from the blank stare and impassive expression that currentlygreet the viewer.The only surviving attribute is a leafy crown embellished with a centraljeweled cameo, on which faint traces of a relief can be detected. The crown on the portrait guarantees the imperial status of the person represented.Exact interpretation of wreaths worn in Roman portraits is often difficult because so many types are known from literary sources, but this particularwreath is clearly intended to show a garland of multilobed leaves woven together (although the lack of detail in the execution does not allow for much texturalvariety).The crown representsoak leaves, which are multilobed, rather than laurel leaves, which have a single lobe and are the only other type of leaf regularlyused in Roman crowns. Oak leaves were used in the coronacivica, the headdress at one time awardedto

Io6

LEE

ANN

RICCARDI

Romans who saved the lives of fellow citizens in battle. From the time of Claudius onward the corona civica became a special attribute worn only by the emperor.4 The oak was sacred to Jupiter, and it grew abundantly throughout the Roman empire, making its use in the corona civica both practical and a symbol of divine sanction. Here, the presence of the central jewel may have been intended to indicate further a special category of corona civica, the gilded oak-leaf wreath given to a triumphator.5 Throughout most of the Imperial period both the corona civica and the corona triumphalis were exclusive attributes of the emperor.6 4. Glandiferi maximegeneris omnes, quibus honos apud romanosperpetuus:hinc civicae coronae,militum virtutis insigne clarissimum,iampridem vero et clementiae imperatorum,postquam civilium bellorum profano meritum coepitvideri civem non occidere("Theyarepracticallyall of the

acorn-bearingclassof oak,which is ever held in honourat Rome,becausefrom it areobtainedthe CivicWreaths,that gloriousemblem of militaryvalour,but now for a long time past also an emblem of the emperors'clemency,eversince, owing to the impietyof the civilwars, not to kill a fellow-citizenhad come to be deemedmeritorious").Plin. HN 16.2.6-3.7; trans.H. Rackham,Pliny, Natural History (Cambridge, Mass., 1945), p. 391. See also Plin. HN 16.12-

13; Gell. NA 5.6; Sen. Clem. 1.26.5. 5. Massner (1988) suggests that the presence of the central jewel on the

crownof a northernGreekportraitin the late 1st or early2nd centuryturnsthe leafywreathinto an attributeof a priest or a magistrate.Her argumentis made primarilybecauseshe finds this head's lack of conformityto any emperor's knownportraittype so troublingthat she wants to deny that its crownis an imperialattribute.For supporting evidenceshe cites threeportraitsfrom Thasos, all of which wearleafy crowns with centralmedallions,and all of which havebeen identifiedas Romanrulers (JuliusCaesar,Claudius,and Claudius/ Nero?).None of them closelyadheresto any knownportraittype.Massner

Figure 1. Kanellopoulos emperor, frontal view. Kanellopoulos Museum. CourtesyDeutsches Archaologisches Institut, Athens (neg. 83/229)

Figure 2. Kanellopoulos emperor, right profile. CourtesyDeutsches ArchaologischesInstitut, Athens (neg. 83/230)

suggests that the portraits do not represent recognizable individuals, but rather priests. This is, however, a particularly difficult argument for the Kanellopoulos

head,on which the leavesareclearly definedas oak, a materialexclusivelyused for imperialcrowns,and the medallionis impressivelylarge.Additionally,the inlaideyes and largesize of the portrait point to its imperialstatus.Moreover,the attestedorigin in northernGreeceis not secure(see note 7 below), and thereforeit may not belong geographicallywith this groupat all. 6. RE IV, 1901, cols. 1639-1640, s.v. corona(G. Haebler);Heinen 1911, p. 152, note 2; Alf6ldi 1935, pp. 10-12; Versnel1970, pp. 74-77; Maxfield 1981, pp.70-81, 97.

UNCANONICAL

Figure 3. Kanellopoulos emperor, left profile. CourtesyDeutsches ArchaologischesInstitut,Athens (neg. 83/231)

Figure 4. Kanellopoulos emperor, rearview. CourtesyDeutsches ArchaologischesInstitut, Athens (neg. 83/232)

7. The only information known

about the provenanceof the head is that Paul Kanellopoulos,who assembledthe entire collection in the museumbefore donating it to the Greek government,is reportedto have bought the piece in Thessaloniki;Dontas 1975, p. 527.

IMPERIAL

PORTRAITS

I07

Therefore, because of its crown, size, and rich inlay originally present in the eyes, the portrait can be none other than that of a Roman emperor. The face, however, does not readily conform to known portrait types of any imperial figure. The ancient viewer would have known whose image was intended by the inscription that accompanied the portrait and also perhaps from the context in which it was displayed. But today, with the inscription and context lost,7 modern viewers are left wondering who the artist intended to represent. Some Roman emperors can be immediately ruled out. The sparse use of the drill indicates that this sculpture represents an emperor whose portrait type was created during a time when coloristic effects of light and shadow (normally achieved through deeply drilled areasin the stone) were not the preferred aesthetic in imperial portraits.This deeply drilled style was first used in imperial portraits under Hadrian, and was especially admired in the portraits of the Late Antonine and Severan emperors, as exemplified in numerous images of the emperors Lucius Verus, Commodus, and Septimius Severus.8 Exuberant use of the drill is lacking on the Kanellopoulos head, however,suggesting that the portraitshould be placed either before Hadrian or after Septimius Severus. 8. Portraitsfrom these erassurvivein abundantnumbersand compilationsof them arelikewisenumerous.Among the most importantandwell-illustrated volumesareWegner 1939;McCann 1968; Poulsen1974 (see pp. 95-100, 103,105-106,109-110,130-132; nos.

76-78, 80-84, 88,90-92,97-99,128-

130;pls. 121-124,129-139,141-146, 150-153,202-207); and Fittschenand Zanker1985 (see pp. 62-98, nos. 58-85, pls. 66-104, suppl.pls. 39-66, 96, with references).

Io8

LEE

ANN

RICCARDI

Further clues are provided.The sculpture is extremelywell preserved, and its smooth, polished cheeks betray no trace of facial hair. In the postSeveran 3rd century, the soldier emperors wore beards, usually variations on the close-cropped type favored by Caracalla.9The only exceptions were child emperors, such as Gordian III. Elimination of the 3rd-century emperors as candidates pushes the later date into the early 4th century. The Kanellopoulos artist must therefore have intended to represent an emperor who ruled either before Hadrian or from Constantine's time onward. Finally, the style of the coronacivica on the Kanellopoulos portrait, carved in fairly low relief and with a centraljewel, finds its best parallelsin the 1st century A.C.in a portrait of Claudius found in Thasos, now in the Louvre,l0 and another of Augustus now in the Capitoline.l1 All three of the crowns display a similarlyundetailed treatment of the leaves,with little or no use of the drill to define them, and all bear a centraljewel. Although the broad swath of materialcovering the termination of the coronacivicaat the back of the head on the Kanellopoulos portrait is a feature not normally seen in Rome until the early 4th century,12it appears in provincial imperial portraituremore than 200 years earlier.13Since the Kanellopoulos portrait is probablyfrom the Greek East, this feature does not support the later date. George Dontas has argued that the portrait in the Kanellopoulos Museum was meant to represent an unusual and uncharacteristicallyclassicizing image of the Tetrarch Galerius, who died in 311, shortly before Constantine's accession.14This identification was accepted by James D. Breckenridge, and the piece was identified as Galerius in the exhibition TheAgeof Spirituality,shown at the Metropolitan Museum of Art in 19771978.15Dontas makes his case for the identification of the portrait on the basis of a report that the head had been purchasedin Thessaloniki, the city that served as Galerius'scapital when he was Caesar of the eastern empire under the Tetrarchy.Stylistic comparisons with a tondo portrait from the "small arch" of Galerius in Thessaloniki16 and a portrait, probably of Galerius, in Copenhagen17are included in his discussion. Neither of these arguments is compelling. Even if the Kanellopoulos portraitcame originallyfromThessaloniki,other imperialimages areknown from that city, a strategic military and trading port, closely linked to Rome in many ways. The Archaeological Museum in Thessaloniki contains several imperial portraits, including those of Augustus, Septimius Severus, and Alexander Severus. A provenance of Thessaloniki does not ensure that the portrait represents Galerius. Moreover,the stylistic comparisonsthat Dontas cites arenot close. The two portraitshe discusses as examplesof Galerius'sportraitureare similarto each other in some respects, including the rendering of the forehead, eyebrows, and eyes. Yet the Kanellopoulos head differs from both in precisely these areas.It exhibits a smooth, unlined forehead,with gently arched eyebrows.The eye sockets reveal the size of the eyes; they were not especially large, and were not emphasized by directed compositional lines to become the focal point of the face, as in the other portraits.The locks of hair are treatedwith more individuality;each lock is longer and straighterthan those

9. WiggersandWegner1971; Poulsen 1974, pp. 161-162, 166, 168173, nos. 164, 170, 173-177, pls. 262263,272-273,277-286; Wegner, Bracker,and Real 1979; Fittschen and Zanker 1985, pp. 124-127,130-143, nos. 105-106, 110-118, pls. 128-131, 135-147, suppl.pls. 89, 91-93, with references;Kleiner1992, p. 361. 10. Musee du Louvre,MA 1226. See de Kersauson1986, pp. 192-193, no. 90, with references. 11. Museo Capitolino, Stanza degli Imperatori6, no. 495. Fittschen and Zanker 1985, pp. 7-10, no. 8, pis. 9-10, with references;Boschung 1993, pp. 129-131, no. 45, pl. 38. 12. The earliestexamplesof imperialportraitswith a Roman provenancethat displaythis treatment arefrom the Constantiniandynastic group set up in Rome in the earlypart of Constantine'srule,including an over-life-size statue of Constantine himself, now in San Giovanni in Laterno,Rome (Fittschen and Zanker 1985, pp. 144-145, no. 120, pls. 149150), and the two portraitsdisplayedon the balustradeof the Campidoglio (LOrange 1933, pp. 58-67, pls. 40-41). 13. For example,on the portraitof Claudiusin the National Museum, Athens, no. 430 (Datsoule-Stavridi 1985, pp. 34-35, pls. 25-26), where the treatmentof the back of the wreath is similarto that displayedon the Kanellopouloshead. 14. Dontas 1975. 15. Breckenridge1979, pp. 13-14, no. 6. 16. Daux 1958; Vermeule1968, pp. 417-418; Riisch 1969, p. 194, no. M20, fig. 109; Makaronas1970, p. 50; Meischner 1986, pp. 223-224; Stephanidou-Tiveriou1995. 17. L'Orange1933, p. 53, no. 5, figs. 142, 144; Breckenridge1968, pp. 242-243, fig. 128; Calza 1972, p. 142, no. 54, pl. 36; Poulsen 1974, pp. 188-189, no. 196; pls. 318-319.

UNCANONICAL

IMPERIAL

PORTRAITS

Iog09

on the other portraits.Dontas actuallymakes little attempt to associate directly the features of the Kanellopoulos portraitwith those of either of his comparisons.l8 The Kanellopoulos head also differs stylistically from the Tetrarch's other images. Most identifiable portraits of Galerius were carved in a linear and schematic style, whether in stone or on coins, and he was usually shown bearded.19Dontas suggests that the Kanellopoulos portrait was carved at the end of Galerius'slife, and that Galerius had shaved his beard in order to follow the fashion set by the young and clean-shaven Constantine.20He bases this conclusion on the fact that there areboth bearded and unbeardedfigures on the Arch of Galerius in Thessaloniki (303)21and the Arch of Constantine in Rome (312-316),22 showing that the early 4th century was a time of transition. As Dontas himself points out, however, it is the younger men who are clean-shaven, while the older men retain their beards. No examples of older, unbearded men are represented on either of the arches,23or elsewhere prior to the year 312, when Constantine himself was first depicted without a beard. Another possible identity for the Kanellopoulos portraithas been proposed. H. P. L'Orange dates the portrait to the late 4th century,suggesting

of the 18. Dontas'scomparison featuresof the portraitsconsistsof a single sentence:"Bienque les pointes des cheveuxne soient pas tourneesdans le meme sens que sur les deux autres portraits(ici, de droite a gauche),je reconnaisle meme personnage:un visage large,un peu empate (davantage encoreici que sur la tete de Copenhague),des cheveuxbien peignes, retombantbas sur le front, des sourcilsqui se rejoignentsur la racine du nez et remontentaux extremites, une bouche etroite et amere"(Dontas 1975, p. 529). 19. LOrange 1984, pp. 26-28, 106109; Calza 1972, pp. 135-148. Many of Galerius'scoins are illustratedthroughout RIC VI. A few sculptedexamples of portraitsof an unbeardedGalerius do exist.The best known is in the famous porphyrygroup in Venice (Ragona 1963). Both of the junior emperorsin this group are portrayedas clean-shaven,in contrastto the bearded senior emperors.Since none of the four portraitsattemptsin any way to show the actualphysiognomyof the men they represent,the presenceor absence of facialhairwas apparentlynot intended to reflectaccuratelythe state of the Tetrarchs'faces.Most scholars agreethat beardswere used as an

iconographicsymbol to suggest the older/seniorstatusof the Augusti in contrastto the younger/juniorstatus of the Caesars. A porphyryhead from Gamzigrad (Srejovic1992-1993; 1995) is also clean-shaven,but since it may have been producedin the same studio in Egypt as the group in Venice,this is not surprising.See Kiss 1984, p. 100, where he arguesthat becauseporphyry was such a rarestone, exclusivelyfrom the Mons PorphyriesCave in Upper Egypt, few sculptorscould carvethe extremelydense and hard material.For a discussionof the provenanceof porphyry,see Gnoli 1988, pp. 124-126. The Tetrarchicporphyryportraitsmay have all been createdby Egyptian sculptorsworking near the site of the quarry.It would thereforenot be unusualfor the sculptorsto repeatthe iconographyof the group portraits,and a status-definingcharacteristicsuch as the beardmight be missing even in individualimages of the junior emperors.The Serbianportraitis dated earlyin Galerius'srise to power on the basis of portraitbusts in his imperial crown,a featurethat would only have appearedon one of the Caesarsof the firstTetrarchy(Srejovic1992-1993, p. 41), and the sculptors(who would

never have seen the emperor)might thereforehave wanted to continue the symbolismof his clean-shavenstatus. The only other unbeardedportrait of Galeriusknown to me is the portrait roundelfrom his small archin Thessaloniki.The details of this image were finished in paint (Vermeule1968, p. 418), so although no beardwas carvedon the image, one was apparently painted on afterward.Galeriusis alwaysrepresentedas beardedon his coins. 20. Dontas 1975, p. 531. Constantine'sfirst definitivelydated appearancewithout a beardon his coins was in the year 312, after Galerius was alreadydead;L'Orange1984, pp. 58-80. 21. Makaronas1970; Pond Rothman 1977; Engemann 1979; Meyer 1980. 22. Wace 1907; L'Orangeand von Gerkan 1939, pp. 22-23, pls. 6-20, 45; Giuliano 1955, pls. 30-47, 52-53; Richardson1975. 23. At least no examplesarevisible. The poor state of preservationof the archin Thessaloniki makes it impossible to make any definitive statement about the faces of many of the men representedthere.

IIO

LEE

ANN

RICCARDI

that it representsTheodosius I.24This identification is also arguedbyjutta Meischner in a series of articles citing both formal and historical evidence. The formal arguments are unconvincing, as no recognizable portraits of Theodosius in the round have survived,25and his image is known only in low relief, as, for example, in the portrait on his famous Missorium26and on his coins. These existing portraits are flat and generally schematic,27 and show virtually no correspondence to the Kanellopoulos head in either style or physiognomic detail.28Additionally, the presence of a coronacivica, as on the Kanellopoulos portrait, is unlikely so late in the 4th century, when the standardimperial headdresshad long been the jeweled diadem.29 The Kanellopoulosportraitalmost certainlyrepresentsneither Galerius nor Theodosius. As both Dontas and Breckenridge acknowledge,30however, it bears definite allusions to the standard portrait type of another emperor,the most admired general of all: Marcus Ulpius Traianus, better known as Trajan.These scholars see a deliberate attempt by Galerius to mimic the features of his hero, but could the portrait instead represent Trajanhimself? Stylistically, the image suggests a portrait type that had its origins in the early 2nd century.The tools used to carve the portrait were the point, chisel, and gouge, with little or no use of the drill. The head exhibits a strong sense of realism, yet the arrangement of the facial lines presents a clear, simple pattern created by the edges of the cheeks, brows, vertical division of the forehead above the nose, and horizontal line between the chin and the mouth. These are standard features of early-2nd-century portraiture.31The gentle facial modeling and finely finished surface con24. L'Orange1984, pp. 28, 106. His rationalefor this identificationis based mainlyon the suspectedprovenanceof Thessaloniki. 25. None, at least, that is well accepted.A few attemptsto identify Theodosius I in sculptedportraitshave been made by variousscholars,but they are all extremelyproblematic.See Kiilerich1993, p. 84, note 248, for specific examples,and pp. 94-95 for a general summaryof the problemof recognizingindividualemperorsof the Theodosian dynasty. 26. Meischner (1996) arguesthat the identity of the centralfigure on the Missorium is Theodosius II ratherthan Theodosius I, thus removingthis portraitfrom her consideration.This identificationis unlikely,however,as the emperoron the Missorium is shown between two coregentsand Theodosius II only had one. Delbrueck 1933, p. 200, pls. 94-98; L'Orange1933, pp. 67-72, pl. 171; Rumpf 1957, p. 20, pl. 16.73; Arce 1976; Kiilerich1993, p. 19.

27. Delbrueck 1933, pp. 200-202; LOrange 1933, pp. 66-77. 28. Scholarssuch as LOrange (1984), Meischner (1988, 1990, 1993), and Kiilerich(1993) maintainthat there was a classicalrevivalduringthe Theodosian era, and thus want to place this portraitat that time and see it as the work of a sculptorwith an interest in classicalforms. 29. Wreathed headdresseswere replacedas imperialinsignia in 326 by Constantine,well before the elevation of Theodosius in 379, and sculptural examplesof them areextremelyrare afterward.They disappearedentirelyin numismatic imperial portraits;

Delbrueck 1933, pp. 53-66. Meischner'sargument(1995, pp. 438-439)

Theodosius'sson Honorius wears a leafy crownon the Rothschild cameo, also perhapsmade for his wedding, is difficult to accept.Honorius'sheaddress on the cameo is certainlynot meant to representoak leaves, and seeing any kind of leaf in it is dubious,for the headdressappearsas a fairlystandard exampleof a jeweled diadem made of a double row of roughlyoval gems. For referencesand a discussionof the Rothschildcameo, see Kiilerich1993, pp. 92-94 and Meischner 1993. Typical examplesof the jeweled diadem can be seen on portraitsof the emperors roughlycontemporarywith Theodosius:Arcadius(?)(ruled395-408), in Istanbul(Inan and Alf5oldi-Rosenbaum 1979, pp. 138-139, no. 82, pl. 74.1-2),

that the corona civica on the Kanello-

orValentinian/Valens (jointlyruled

poulos portraitwas revivedto markthe occasion of Theodosius'smarriageto Galla in 387 would be more convincing if there were any literaryor representational evidencethat an oak-leaf crown was the markof a bridegroom,but none exists. Her suggestion that

364-375, Valensalone until 378) in Florence (Fittschen and Zanker 1985, pp. 158-159, no. 126, suppl.pl. 95:a, c-d). 30. Dontas 1975, pp. 523, 530-531; Breckenridge1979, p. 13. 31. AgoraI, p. 29.

UNCANONICAL

IMPERIAL

PORTRAITS

III

Figure5. Trajan,StaatlicheAntiken Sammlungen,Glyptothek,Munich, no. 335. CourtesyMuseum

32. Athens, National Museum, no. 384. Lattanzi 1968, pp. 34-35, no. 2, pl. 2:a-b; Graindor1915, p. 292, no. 2, fig. 10; IG II22021. 33. Athens, NationalMuseum, no. 392. Lattanzi1968, pp. 35-36, no. 3, pl. 3:a-b; Graindor1915, p. 304, no. 4, fig. 11. 34. All the Type I images accepted Fittschen and Zankerarelisted, by includingbibliography,in Fittschen and Zanker 1985, p. 39. They admit eight portraitsinto this category.The Kanellopoulosportraitcan be most closely associatedwith this type, althoughvery loosely.In this case, the emperoris depicted as considerably older than he is in the others. 35. Munich Glyptothek,no. 335. Gross 1940, p. 132, no. 72, pl. 32:b; Bernoulli 1891, p. 81, no. 58. 36. This trait is found on virtually every obverseportraitofTrajanin Roman Imperialcoinage. See Strack 1931, pls. 1-10; Kent 1978, pls. 74-77; BMCRE III, lxix, note 2, pis. 9-45; and RIC II, pls. 8-12.

trast markedly with the abrupt, almost mechanical carving of the hair, which can be compared to the carving on portraitsof kosmeti from Athens, especially Heliodoros, dated by inscription to A.D. 100-110/1,32 and an unidentified male from the same era.33The Kanellopoulos portrait is also likely to have come from an early-2nd-century Attic workshop, and no emperor of this period fits this iconography as well as Trajan. The facial features of the Kanellopoulos portrait present an older and more obese version of Trajan than his more familiar official images, yet certain stylistic similarities with his official portraits, especially those of his first portrait type, are clear.34His Type I portrait in Munich (Fig. 5),35 for instance, shares many featureswith the Kanellopoulos portrait,including an almost identical arrangementof thick, comma-shaped locks of hair combed forward onto the forehead and curved slightly under and to the side. Both portraits also exhibit similar broad, flat planes of the cheeks; deep furrows running from the edge of the nose to the mouth; and long, thin lips. Even their facial proportions are similar. Additionally, the Kanellopoulos portrait also exhibits a protruding ridge above the eyes, a standardTrajanic feature common to all of his portrait types and a trait especially visible in his official numismatic portraiture.36

112

LEE

ANN

RICCARDI

Despite this stylistic evidence, Dontas, Breckenridge, L'Orange, and Meischner all prefer to see other emperors in the Kanellopoulos image because its overall effect is not the same as that conveyed in standardportrait types of Trajan.Is this a valid reason to reject the identification? As shown below, provincial imperial portraits often do not closely resemble the official, canonical images of rulers made in Rome. The characteristics of this portrait point most clearly to Trajan, despite the lack of resemblance to his official types.

THE PROBLEM The study of Roman imperial portraiturehas long been dominated by the belief that images of various imperial figures can be divided into categories based upon prototypes now lost. Such prototypes are usually thought to have been created in association with a landmark event in the emperor's rule, such as the winning of a war, the acquisition of some new title, or the celebration of a decennalia,so these prototypes can be arranged chronologically. They would have been devised in imperial circles, and were intended to create official images for the reigning emperor and his family. Portraitsbased on the official models were then circulatedthroughout the empire via a process that is only imperfectly understood,37perhaps through the use of plaster or terracottamodels and certainlyin part by their prominent display on the obverses of Roman Imperial coins. Numismatic portraits, in fact, are the best surviving record of what the models must have looked like, and the identity of the emperor and often the date of issue are confirmed by the legends that accompany the portraits.The conventional scenario is that when a provincial city wanted to erect a portrait of the reigning emperor, the sculptor commissioned to create it would follow a model based on whatever prototype was currentlycirculating and produce an image as close to it as possible. Extant imperial portraits can then be analyzed and classified according to the prototype they followed. The most important characteristic that modern scholars use to categorize extant images is hairstyle,determined by coiffure patterns depicted on the coins. The facial features are considered only secondarily.The customary approach of modern scholars is to compare the labeled numismatic images on official coins with unlabeled portraits in sculpture.The chronological order of the numismatic types is also used to determine the chronological availabilityof the now-lost models, for the use of a portrait model provides at least a terminuspost quem. An important underlying assumption driving this methodology is that the overall process of erecting and commissioning an imperial portrait was subject to some form of either overt or covert central control by the emperor or his inner circle, so that adherence to whatever officially approved portrait was currently in use may be expected in all or most instances. This method has proven effective in so many cases that the identity of large numbers of anonymous portraits has been determined with assurance. Often such close correspondence to specific features can be seen, even in portraits from the far reaches of the empire, that identifications

37. Swift 1923 and Stuart 1939

references forthe remainthe standard circulationof portraitmodels. For a more recent discussionand emphasis on cities on the receivingend of this imperialpropaganda,see Rose 1997a, pp. 57-59 and especiallyRose 1997b.

UNCANONICAL

IMPERIAL

PORTRAITS

II3

can be confidently assigned. Scholars have come to rely on the prototypereplica model for identification, and rarely have different scenarios been considered.It is postulatedthat in Roman times, when a portraitwas erected in a provincial city, local artists would reproducethe latest official portrait of an emperor as well as they could. Because some sort of control by Rome over the quality of the portrait is assumed, images that deviate from official prototypes are thought to reflect the unavailabilityof a suitable model; misunderstanding by local artists;or a lack of skill in execution. Portraits that do not possess features similar enough to those of known types to be recognized may be misidentified or ignored. This focus on internally recognizable images that, however loosely, reflect a specific, centrally devised prototype has largely excluded less easily recognizable representations of emperors, that is, images that do not conform to a standard canon of features and that lack identifying inscriptions.38 But were images throughout the Roman world intended to be recognized by their appearance alone, without context or inscription? Repeatedly, scholars have been forced to acknowledge that some imperial portraits, particularly away from Rome, deviate dramatically from official types39and can be included in a replica series only by granting a great deal of creative license to the local sculptor, or by postulating special circumstances in the creation of a specific portrait. For example, Klaus Fittschen, who has worked extensively on provincial images as well as on centrally created ones, has noted the phenomenon of the "blurring"of types in some portraits of the long-ruling, popular emperor Hadrian.40Fittschen demonstrates that some of Hadrian'sportraits consist of combinations of features taken from more than one of the officially sanctioned images of the emperor.Nonetheless, Fittschen regardseach of them as a special case. He remains firmly committed to the idea that sculptors normally created their portraits with conscious intent to mimic the current prototype provided by Rome, and can offer no explanation for why others worked eclectically,

38. Rose (1997b) discussesthe evolution of scholarshipin Roman portraiturestudies from the Renaissance onward,and also notes the centralposition given historicallyto the processof identifying and cataloguing images accordingto type, as well as the historicaldisregardof context or provenance.As he points out, traditional scholarshave createda centuriesold focus on connoisseurshipand privilegedthe analysisof the images themselves,so that the function and use of the portraitsin their originalsettings were issues, until recently,only rarely considered.Rose challengesthe underlyingassumptionthat the portraitswere in some way under the control of the emperoror his inner circleby remindinghis readerson

p. 109: "One issue in particularmust be kept in mind: emperorsdid not set up portraitsof themselves;provincialcities set up portraitsof the emperorin gratitudefor or in anticipationof Imperialbenefactions."As he goes on to discuss,the direct supervisionof the installationof the portraitwas normally carriedout under local ratherthan centralauthority.Thus the appearance of the final productwas subjectto a varietyof regionalfactors,including the economic and artisticresources availableto the city,the abilitiesand trainingof the local artists,and the accessibilityof imperialportraitmodels. The result,as is clearlyvisible by examination,is that provincialportraits can be quite diverse. 39. An especiallypowerfulexample

of this deviationis visible in the imperialportraitgroup from Cyrene found in 1989 (Walker 1994). The eight statuesin this group representing membersof the Julio-Claudianfamily must all have been carvedby a local workshop and arelacking in nearlyany suggestion of Roman influence.It is clearthat the sculptorsdid not follow measuredmodels, nor attempt to give more than a vague allusionto the Roman featuresof their subjects.It is inconceivablethat these portraitscould have been approvedby anyonewhose goal was to presenta standardized appearanceof membersof the imperial family. 40. See Fittschen 1982; and especiallyFittschen 1984.

II4

LEE

ANN

RICCARDI

especially as the portraits he examines were neither all provincial nor all produced by a single workshop. More recently, Cecile Evers has addressed the phenomenon of "blended"portraits in her work on representationsof Hadrian,41agreeing in principle (although not in specific detail) with Fittschen's conclusions. Evers proposes a chronological explanation for these blended portraits. She can suggest this only by rejecting some of Fittschen's examples and revising others to reflect different official models than the ones he proposed. Thus, the portraits that Evers includes in this category are only those that supposedly used two chronologically successive models rather than two randomly selected ones. She then concludes that the sculptors were in fact clinging to features of an old type while introducing some characteristics from a new one (so that these portraits should be more properly thought of as "transitionaltypes"rather than blended ones), and she explains the whole phenomenon in terms of the personal choices made by the sculptors. Both Evers and Fittschen regard portraits with characteristics of more than one official type as relativelyisolated instances, and both assume that despite the exceptional cases of these blended types, adherence to the currently sanctioned official image from Rome was normally the goal of provincial sculptors. Paul Zanker has also investigated the variety of provincial images of several emperors, not only from the East, but from North Africa and the western provinces as well.42He, too, found in many of them a lack of conformity to type. Zanker freely acknowledges the vastness of the variety present in the provincialportraits,and he proposes a wide number of caseby-case scenarios to account for it. Yet no matter how remote the influence, Zanker, like Fittschen and Evers, usually ties the provincially produced portraitsto the sequenceof Roman prototypes,43and alwayspresumes that, one way or another, the sculptor'sintention was to re-create the visual details of an officially sanctioned model or models. By giving such a priority to the prototype, all of these scholars imply that Roman rather than civic concerns were the primaryforce determining the appearanceof the portraits. The problem of the lack of standardizationof provincialportraitswas also raised by Jale Inan and Elisabeth Rosenbaum in discussions of the portraits of Hadrian in their two extensive catalogues of Roman portraits excavated in Turkey.They state that "the portraits of Hadrian from the provinces vary considerably in type and frequently are very poor likenesses."44Their difficulties with these images can be easily seen by examining the results of their attempts to associate their portraits of Hadrian with his official types. In their first catalogue of portraits from Asia Minor, seven images identified as Hadrian are included. Among these they found only two based clearly on Hadrian's official types.45One was too badly damaged to be analyzed,46and the other four are clearly not based on imperial prototypes.47In their second catalogue of portraits from Asia Minor,48six more portraits of Hadrian were accepted. Of these, they were unable to determine a clear Roman prototype for four of them.49More than half of the thirteen portraits of Hadrian presented in these two volumes were thus apparentlycreated independently of Roman models. De-

41. Evers 1994, pp. 267-271. 42. Zanker 1983.

43. Zanker1983.See,forexample, his discussionof the bronze portrait recoveredfrom the Thames River (pp. 16-18) or the bronze portraitfrom Cilicia now in the Louvre (pp. 22-23). These both providehim with particular difficultyas he laborsto connect them with official imperialtypes of Hadrian and Nero, respectively. 44. Inan and Rosenbaum1966, 72. p. 45. From the Asklepieion in Pergamon(Inan and Rosenbaum1966, p. 70, no. 31, with references),and Miletus (Inan and Rosenbaum1966, p. 70, no. 32). 46. A relief from the Libraryof Celsus in Ephesus (Inan and Rosenbaum1966, p. 71, no. 34, with references). 47. From Perge (Inan and Rosenbaum1966, pp. 68-69, no. 29, with references);the Temple of Trajan in Pergamon(pp. 69-70, no. 30, with references);Ephesus (pp. 70-71, no. 31, with references);and Kadirli,now in Istanbul(pp. 71-72, no. 35, with references). 48. Inan and Alf6ldi-Rosenbaum 1979. 49. From the Via Tecta in Pergamon (Inan and Alfoldi-Rosenbaum1979, pp. 94-95, no. 43, with references);the Nymphaeumin Perge (pp. 95-96, no. 45, with references);SeleuciaPieria (pp. 98-99, no. 47); and Izmir (pp. 99100, no. 48).

UNCANONICAL

IMPERIAL

PORTRAITS

II"5

spite the problems that the prototype-replica scenario creates in dealing with provincialimages, however, these scholars all deny that sculptors may have had the freedom to accept, modify, or only loosely refer to current models and prototypes given to them by Rome. But what if this was the case? Provincial sculptors may have created imperial portraits by following a general verbal or written description or through a passing visual acquaintanceshipwith an image but no measured model. The pressure on provincial cities to honor and flatter the emperor was strong, for it demonstrated their loyalty to Rome and often led to important benefits. Yet access to sculptors and workshops was limited, and portraits were expensive. Not every city would have had the resources or the desire to arrange for a current model to be measured and followed, especially if the consequences of failing to do so were minimal. Since inscriptions and setting were the main determinants of portrait identity for the local population, and few viewers would have been familiar enough with imperialpropagandato recognize coiffurepatterns or exact facial proportions, it makes sense that not every civic government would ensure that sculptors had access to the latest models. It is far more likely that sculptors had a less precise description to follow-perhaps written or oral-or had a general visual familiarity with a portrait displayed in a neighboring city. While permission was sought from the emperor to allow the erection of his portrait,local officials were responsible for the process of commissioning and installing it.50In small, remote cities, a Roman official familiar with the specifics of the emperor'simage would rarelyhave been on hand to interfere, but even in more major centers where Roman representatives were probably present, attention to how closely the latest models were followed would vary according to individual interest. The sculptor of the Kanellopoulos emperor was probably trained in Athens, but the portrait itself may have been made elsewhere,where no measured model was available, and without a patron concerned with following imperial propaganda. In light of the astonishing variety present in provincial imperial portraits, this scenario may be the most workable explanation.

LITERARY EVIDENCE FOR UNCANONICAL PROVINCIAL PORTRAITS In addition to the evidence providedby extant portraitsand survivingstatue bases, ancient authors have provided important evidence for determining the authority behind the process of commissioning and erecting imperial portraits. A particularly enlightening ancient comment comes from the letters of Marcus Cornelius Fronto to his formerpupil, the emperorMarcus Aurelius.In a passageexpressinghis affection and longing for young Marcus during a period of separation, Fronto wrote:

50. Rose 1997b, pp. 110-111.

You know how in all the money-changer's bureaus,booths, bookstalls, eaves, porches, windows, anywhere and everywhere there are likenesses of you put before the public, most of them badly painted, and modeled or carved in a plain, not to say sorry,style of art. Yet at

II6

LEE

ANN

RICCARDI

the same time your image, however dissimilar it is, never meets my eyes without forcing a smile from my lips.51 This passage tells us that the images of the emperor were displayed prominently throughout the Roman world but were often badly made and poor likenesses.Yet, despite this, Fronto knew that these images were meant to representMarcus Aurelius. This knowledge was probablybased in part on the settings for the portraits,which reinforced their imperial identity. At the time that Fronto wrote the letter, however, between the years 145 and 147, Marcus was not the primaryruler,but merely heir apparent,so he was only one of several imperial family members whose portrait could have been displayedin these locations.Therefore, for Fronto to have known the images represented Marcus himself, they would probably have borne inscriptions. It is unlikely that the ancient viewer was meant to recognize the image of a particularmember of the imperial family on the basis of the appearanceof his portrait alone, out of context and without a label. The situation Fronto describes, however, does not represent quite the same scenario as that of the Kanellopoulos emperor.The artistic quality of the portraitsFronto saw was quite poor, while the Kanellopoulos emperor is the product of a highly skilled, sophisticated sculptor. Nevertheless, Fronto's testimony is important because he makes clear both the ubiquitous display of imperial portraits and the uneven quality of the images. Both of these factors imply a local ratherthan central authority.A centralized process would be much more likely to ensure both quality and consistency in the appearanceof the portraits,and overseeing the large numbers of images on displaywould strainthe resourcesof any central office. While Fronto does not speak directly to the issue, the prototype-replica scenario of portrait production is also challenged by his testimony. The different artistic, material, and economic resources of the cities and the different goals and requirements for the erection of a portrait can explain both the diversity that Fronto describes and the evidence supplied by the extant portraits themselves. If the process were under Roman control, a higher quality and greaterconsistency would surelyhave prevailed,and these portraits would have conformed more directly to currentlysanctioned official models. Another important ancient source on the question of central versus local control over a provincial imperial portrait is found in the writings of Flavius Arrianus, who served in various administrative positions under Hadrian. Shortly after he was appointed governor of Cappadocia, he embarked on a journey to inspect the Black Sea coast. His log of this voyage, which took the form of a letter addressed to his emperor,begins: At Trapezus. .. the altars are still standing, of stone so rough that the letters are not clear; and the Greek inscription was engraved with mistakes, as though written by barbarians.I decided, then, to rebuild these altars in white marble and to furnish them with inscriptions with clear letters. Your statue is successful in its posture-facing the sea-but as far as the execution goes, it does not resemble you, neither is it beautiful. Send therefore a statue worthy

51. Scis ut in omnibus argentariis mensulispergulis tabernisprotectis vestibulisfenestris usquequaqueubique imagines vestrae sint volgo propositae, male illae quidempictaepleraeque et crassa,lutea immo, minervafictae sculptaeve;quom imterim numquam tua imago tam dissimilis ad oculosmeos in itinere accidit, ut non ex ore meo excusserit

rictumosculi.Loeb ClassicalLibrary, C. R. Haines, ed., The Correspondence of Marcus CorneliusFronto I (Cambridge,

Mass., 1919), p. 207.

UNCANONICAL

IMPERIAL

PORTRAITS

II7

of bearing your name, in the same attitude; the place is wholly suitable for an eternal memorial.52

52. Et; TparcSovra- c xca oL pooiL dvSoTaiXv yS T0U TPGXOS^o, "iY), Oou suVTOOL

Yjxotlev...

xaL T(CxypoaipiaTa

xoc TouTO oUox

?YXBaq xXrecpaxTa- TCO6s 'EXX?YvLxov7tyLYpao j1ca xaL yj1apTYjVFGi(;yypaXTTaO, ola ?o

OrT0papdcpcovypacp?v."Eyoxa

o5v To6;sT? XsUxoi -pc3ovbS; L6OoU TOCryLyxpaolaca avocervalo, xa&L ?yXOC6rcDaL ?SUqU[tqO; T?OLS ypocica(7LV. 0 ,iUv yap avapta; ?T-Y]xsEV 6 (0o5 TCZp?V 6OXaxT-cL J?CL5; (7to?sLXVOuOLV yap

-vV

-cTv ?'epyaoioav OVTE OdXa-c-Tcav), 6 o656o-LOT? aXAx; xcaX6 5oGc? [?IR(PdOV aV8aWcVC-a

&tOV

a CY6V Tv an-Ci6 ?V TouTqc X7yjIGaTcLTOYap Xc&)pov ?1ovoo0V aO ?nLT6ccTo

CTOV ?L

VnjlV

aLcOVLOV.

Arr.Peripl.M. Eux. 1.1-4; trans. N. Hannestad,RomanArt and Imperial Policy(Aarhus1986), p. 201.

The statue of Hadrian that distressedArrian apparentlydid not excite the same response in the people ofTrapezus; at least, no one had been sufficiently disturbedby its lack of beauty or its lack of resemblanceto the emperor to replace the statue or prevent display.Although their satisfaction cannot be determined, the original dedicators of the monuments had presumably approvedthe poorly executed statue and flawed inscription before installation. It was the imperial representativeto the area, a man familiar with both the emperor'sactual appearanceand the emperor createdby imperial propaganda,who complained about the statue and undertook to replace the altarsand inscriptions.Arrianwas no doubt reportinghis dissatisfaction in order to highlight his own interventions on Hadrian'sbehalf and therefore encouragethe emperorto think well of him, but it is notable that Arrian chose an imperialshrine to command the attention of the artistically inclined emperor.Here is a case in which an imperial authority took over a processthat had initiallybeen carriedout by local officials;such interference was clearlyexceptional,and not a matter of course.This is solid evidence for a normal process of local ratherthan Roman control. If a Roman governor with less interest in the aesthetics of the shrine had been appointed rather than Arrian, the imperial area might have stood unchanged, with a statue that did not resemblethe emperorand with mistake-filled inscriptions.The number of other imperialshrines in the East containing statues that did not look like the intended emperor cannot be determined, of course, but this passage shows that they did exist. Arrian'scomments are also interesting because he appeals to Hadrian to send a new and better statue. Although he commissioned the new altars and inscriptions himself, he did not replace the statue, and instead asked the emperor to do so. This suggests that Arrian, accustomed to the official portrait of Hadrian, despaired of finding a local sculptor able to create an image that would satisfy him, and he appealed to Hadrian himself to send a properly approvedportrait rather than accept the kind of image provincial artists would produce. Arrian took his role as Rome's representative seriously, and wanted to have the statue in the shrine express qualities approvedby the emperor. Fronto and Arrian thereforeprovide testimony suggesting that variety and diversityin the appearanceof imperialportraitswould be normal rather than unusual, expected ratherthan surprising.Since the hundreds of cities and towns in the empire that erected portraits of emperors did so according to their own means and motives, generalized resemblance shaped by local taste rather than rigid adherence to a chronologically limited prototype would necessarily prevail. Still another revealing passage from ancient literaturethat sheds light on the process of portrait-making comes from a letter written during the time of Trajan by the younger Pliny to the scholar Vibius Severus. Pliny asked his friend to secure permission from two of his fellow townsmen for portraits of them to be made. With his request, he included cautionary statements about how the artist should create the portrait:

II8

LEE

ANN

RICCARDI

I beg, moreover,you would employ some skillful hand in this work; for if it is difficult to draw an exact likeness from life, it is much more so to preserve it in copying what is itself a copy; so I desire you would not suffer the painter you select to deviate from the latter, not even for the better.53 Pliny must have been speaking from his own visual experience when he encouraged Vibius Severus to beware of artists who might alter a portrait model, even to improve it. Born of high rank and status in the Roman world, Pliny had no doubt seen various versions of the portrait of his emperor that did not conform well to official models put out by Rome. From this passage it is clear that artists might take matters into their own hands and deviate from a model, so when he himself commissioned a portrait,he specificallyrequestedthat an artistwho would not do this be chosen. Pliny's passage confirms that not all artists could be relied upon to follow a model closely, and further supports the likelihood that no centralized control existed over the quality of an emperor'sportrait.

PROVINCIAL UNCANONICAL HADRIAN AND TRAJAN

PORTRAITS

OF

Fittschen, Evers, Zanker, and Inan and Alfildi-Rosenbaum all single out the provincial portraitureof Hadrian as being particularlysusceptible to deviation from canonical types. This is in part, no doubt, because of the large corpus of Hadrianic portraits.The widely traveledemperorfrequently made great benefactions to cities he visited, and it was potentially advantageous for a civic government to display a portrait of him prominently. Thus, cities large and small, regardlessof resourcesor access to good models or workshops, erected his image. Many of these portraitshave survived. With Hadrian, therefore, a particularlylarge body of evidence exists to document his appearancein the provinces.The diversity of these portraits and the lack of conformity to imperial types can be examined by looking at provincialportraitsof Hadrian from Greece and Asia Minor.54While many of these portraits exhibit local or regional styles of workmanship that can be seen in other works from the same area, in the portraits shown here variationsactuallypreempt the traitsof the official prototypes and are more crucial to the overall appearance of the portraits than the features that adhere to an imperial model. This implies that priority was given to the local sculptor, not the Roman prototype, in the creation of a portrait. A portrait from Perge (Figs. 6, 7),55for example, with its upwardly inclined head and dramaticallytwisted neck, mimics Hellenistic ruler iconography rather than more sober Roman propaganda.Details such as the sharply carved eyebrows that continue down the edge of the nose and the hard-edged carving overall show a classicizing tendency that may be a holdover from Trajanictimes. Facial proportions are cubic and solid, and the arrangement of the hair compares only loosely to that on any of Hadrian's official types (the closest is the Stazione Termini type). This combination of features is most likely the result of the local sculptor'sown

53. Peto autem, utpictorem quam diligentissimum adsumas. Nam cum est ardnum similitudinem effingere ex vero, turn longe difficillima est imitationis imitatio; a qua rogo ut artificem, quem elegeris ne in melius quidem sinas

aberrare.Plin. 4.28; trans.W. Melmoth, revisedby W. M. L. Hutchinson, Pliny: Letters (Cambridge, Mass.,

1957), p. 351. 54. Portraitsexaminedhere are limited to ones known to have been displayedin the Greek-speaking provincesof the Roman world, because this is the probablecontext of the Kanellopoulosemperor.I have not attempteda thorough empire-wide analysis,but my preliminaryimpression is that this situationexists all over the empire.

55. AntalyaMuseum, no. 2649. Inan and Rosenbaum1966, pp. 68-69, no. 29, pls. 19:2, 21; Fittschen and Zanker 1985, pp. 44-45; Vermeule 1968, pp. 259,392, no. 16, fig. 140; Evers 1994, pp. 84-85.

UNCANONICAL

Figure 6. Hadrian, from Perge, frontal view. Antalya Museum, no. 2649. CourtesyMuseum

Figure 8. Hadrian, from Ephesus, frontal view. Kunsthistorisches Museum, no. I 857, Vienna. Courtesy Museum (neg. III 14249)

IMPERIAL

PORTRAITS

Figure 7. Hadrian, from Perge, right profile. Courtesy Antalya Museum

Figure 9. Hadrian, from Ephesus, left profile. Courtesy KunsthistorischesMuseum, Vienna (neg. II 6743)

II9

120

LEE

ANN

RICCARDI

training and the tastes of his patrons.While the portrait retains some allusions to Hadrian'sofficial types in the arrangementof the hair and beard, it does not clearly follow any of his prototypes.56 Hadrian's portrait from Ephesus (Figs. 8, 9) also reflects a different tradition than is found in any of his Roman models.57This portrait, with its blocky,cubic facialproportions,small narroweyes, and roughly sketched locks of hair, suggests a more youthful and benign emperor than the one represented in his official types, and scholars have had to admit that it does not comfortably conform to any of them.58 56. Fittschen, in Fittschen and Zanker 1985, pp. 44-45, acknowledges this, and states that it shows a breakdown in the transmissionof official types.Yet here, as elsewhere,he implies that this is an aberration.Inan and Rosenbaum(1966, pp. 68-69) also note that it is not a replicaof any of Hadrian'sofficial types and designateit "anindividualcreationbased on a model

at leastrelatedto the'StazioneTermini' type."Theyattemptto explainthisby notingthe date(121,providedby the inscription),earlyin Hadrian'sreign, and suggest that the portraitwas

producedtoo earlyto complywith

officialpropaganda(although as they acknowledge,three of Hadrian'sofficial portraittypes had alreadybeen established by then). Evers (1994, pp. 84-85) associatesit more solidly with the StazioneTermini type, but admits that it representsa Hellenistic variantof it. 57. Found in the tower of Mithradates,now in the Kunsthistorisches Museum, Vienna, no. I 857. Wegner 1956, pp. 38-39, 116, pl. 18; Giuliano 1959; Inan and Rosenbaum1966, pp. 70-71, no. 33, pl. 20:1-2; Vermeule 1968, p. 392, fig. 139; Fittschen and Zanker 1985, p. 47; Evers 1994, pp. 9798, no. 147, ills. 24-25.

Figure 10. Hadrian, from Pergamon, frontal view. Bergama Museum, no. 783. CourtesyDeutschesArchaologischesInstitut,Istanbul(neg.66/57) Figure 11. Hadrian, from Pergamon, three-quarter view. CourtesyDeutsches Institut,Istanbul Archaologisches (neg.66/94)

58. Vermeule(1968, p. 392) calls it an "originalcreationof a local sculptorusing Hellenistic rulerportraitsas models," while Inan and Rosenbaum(1966, p. 71, afterWegner 1956, p. 38) state that "it does not slavishlycopy one of the official types but seems to be relatedmore to the type 'Chiaramonti392' than to any of the other establishedtypes."Fittschen and Zanker (1985, p. 47) include it with replicasof the "VaticanChiaramonti 392,"but acknowledgethat it has connectionsto the StazioneTermini type as well. Only Evers (1994, p. 197) has no difficultywith it as a faithfulreplicaof the Chiaramonti392 type.

UNCANONICAL

Figure 12. Hadrian, from Thasos. Thasos Museum, no. 2346. Courtesy Ecolefranraise d'Athenes(neg.59.329)

59. BergamaMuseum,no. 783. Horn andBoehringer1966, p. 474; Inan and Alf6ldi-Rosenbaum1979, pp. 94-95, no. 43, pl. 36:1-2; WegnerandUnger 1984, p. 128, pl. 7:1;FittschenandZanker1985, pp.46,51; Evers1994, pp. 150-151, no. 89. 60. Evers(1994, pp. 150-151) and Inan andAlf6ldi-Rosenbaum(1979, pp. 94-95) follow Horn and Boehringer (1966, p. 474), and believethat this is not a directreplicaof any of Hadrian'sknown types,but ratheris somewherebetweenthe StazioneTerminiand Chiaramonti392 types.Fittschen,on the otherhand,in Fittschenand Zanker1985, p. 51, sees it as a combinationof the Rollockenfrisur type (on the left side) and the Stazione Terminitype on the front and right sides, andWegner and Unger (1984, p. 128) believethat it followsthe Rollockenfrisur type exclusively.

IMPERIAL

PORTRAITS

121

Hadrian'sportrait from Pergamon (Figs. 10, 11)59presents yet a further variation. Although badly damaged, it is clear that the eyes, set close together on the face, are not typical of his other portraitsin either shape or placement. The wide, thin-lipped mouth and full fleshy cheeks are also unexpected features for the canonical imagery of this emperor.60 The portraitof Hadrian from Thasos (Fig. 12)61has also baffled scholars searching for a direct prototype. This image, carved in a crisp, dry style, exhibits so much individual variation, especially in the hair, that it cannot really be associated with a specific portrait type.62The beard, hair, large nose, wide mouth, and small eyes all create an impression that only loosely recalls standardHadrianic images. The style of the production clearly fits within the local workshop tradition, as many other sculpturesfrom Thasos exhibit these traits, but the significance of it here is that the local style overrides any attempt to follow directly a Roman model. A final example of provincial variety in the portraits of Hadrian can be seen in the survivingportraitfrom Dictynnaion in Crete, now in Chania (Figs. 13, 14).63The toss of the neck, arrangementof the beard, fullness of the chin, and simplified shape of the mouth are all inconsistent with Hadrian'sfeatures in his canonical types, and scholars cannot agree on the model that this image follows.64 Uncanonical portraitsof Trajanhave also been identified in the Greek East. Though less numerous than those of his successor, these portraits nonetheless demonstrate a marked disregard for his "official"models. In their first volume of portraits from Asia Minor, Inan and Rosenbaum include only one portrait of Trajan that they accept as following an established type.65Four portraits of Trajan are included in the second volume. Of these, they found only one that possibly follows a Roman prototype.66 They determined that the other three67were closer to Trajan'sdecennalia type than any other, but did not fit comfortably into that series. Thus, of

version of the emperor"muchmore 61. Thasos Museum,no. 2346. Rolley Hellenistic in flavor"(p. 391). Wegner and Salviat1963;ArchDelt18 (1963), Vermeule (1956, p. 95) and Datsoule-Stavridi 1966, 297; 259, 296:a, p. pls. (1970, pp. 48-50) suggest it is after the p. 55; Vermeule1968, p. 259, fig. 139; Wegner and Unger 1984, p. 140; Fittschen Chiaramonti392 type, while Fittschen, in both Fittschen and Zanker 1985, and Zanker1985, p. 46, note 5e; p. 51, note 8c; Evers 1994, pp. 187-188, no. 138. p. 50, and Fittschen 1984, p. 200, no. 32, 62. Even Evers (1994, p. 187) is pl. 64:c-d, identifies it as an exampleof the Rollockenfrisurtype. troubled by this portrait, finally deciding 65. From Pergamon(Inan and that it is a variantof the Rollockenfrisur Rosenbaum and Zanker in Fittschen 1966, p. 68, no. 28, with Fittschen, type. 1985, p. 46, note 5e, sees it as a blend of references). 66. A fragmentfrom Ephesus from the Rollockenfrisurand the Stazione Termini types. Hanghaus II, Raum SR 5 (Inan and Alfoldi-Rosenbaum1979, p. 93, no. 41, 63. ChaniaMuseum,no. 82. Foundin with references). 1913 nearan aqueductand milestone 67. From Ephesus (Inan and Alfoldibearinghis name (ICrII, no. 6). Vermeule Rosenbaum1979, pp. 91-92, no. 39, with 1968, pp. 391,443, no. 7, fig. 136. 64. Vermeule(1968, p. 443) notes that references);Istanbul(pp. 92-93, no. 40, with references);and Silifke (p. 93, the portraitbearsa "dreamylook,"and no. 42, with references). suggeststhat it representsa Cretan

122

LEE

ANN

RICCARDI

the five portraitsof Trajanthey discuss in the two volumes, only two clearly echo an official Roman model. The portrait of Trajan from Ephesus (Figs. 15-17),68 found in the cryptoporticus of Domitian's temple, bears distinctly un-Trajanicfeatures such as a wide, smiling mouth; long, narrowhead;broadcheeks;and prominent, rounded chin. It can be identified asTrajanlargelybecause its findspot suggests an imperial portrait;9 it is well over life size at 42 cm; and stylistically, it resembles an image of Trajanmore than any other emperor.The shape and carving of the eyes and pattern and linear definition of the hair bear enough similarities to Trajan'sofficial images that the identification has largely been accepted by scholars,70yet it does not appearto follow any particularmodel. An undisputed portrait of Trajan from the Greek East, found in the Temple of Trajan at Pergamon (Figs. 18, 19),71reveals yet another variation from the standardized types of Trajan'sofficial images. In this portrait, despite the disfiguring break over the right eye, it is clear that the hair is curlier and the locks more widely spaced than normal, and the proportions of the face are longer and thinner than expected. The somber look created by the heavily furrowed brow is unusual in a portrait of an emperor whose appearance is typically confident and optimistic. Yet its findspot and size (80 cm) make it difficult to see anyone else represented in this image. A final example of an untraditionalTrajanicportraitis one in the Silifke

Figure 13. Hadrian, from Dictynnaion, frontal view. Chania Museum, no. 82. CourtesyMuseum Figure 14. Hadrian, from Dictynnaion, left profile. Courtesy ChaniaMuseum

68. SeljukMuseum, no. 11/37/72. Tiirkoglu 1972, pp. 28-30, pls. 5-6; Inan and Alf6ldi-Rosenbaum1979, pp. 91-92, no. 39, pl. 33. 69. A portraitof Germanicuswas also found in this area:Scherrer1995, p. 96. 70. Fittschen and Zanker (1985), however,do not list it with any of Trajan'sportraits. 71. Pergamonmuseum,Berlin. Bluimel1933, pp. 15-16, no. R35, pl. 33; Giuliano 1959, p. 172, no. VII.3, fig. 23. Part of an acrolithicstanding statue.Accepted as a decennaliatype in Gross 1940, pp. 61, 93, 127, no. 26; Inan and Rosenbaum1966, p. 68, no. 28, pl. 17:1-2; and Fittschen and Zanker 1985, p. 42, note 1.

UNCANONICAL

Figure 15. Trajan, from Ephesus,

frontalview.Efes Miizesi, Seljuk, no. 11/37/72. CourtesyMuseum Figure16.Trajan,fromEphesus, left profile. CourtesyEfes Miizesi, Seljuk

Figure17.Trajan,fromEphesus, three-quarter view. Courtey Efes Miizesi, Seljuk

IMPERIAL

PORTRAITS

I23

124

LEE

ANN

RICCARDI

Museum, from ancient Seleucia ad Calycadnum.72Although this portrait is in poor condition, it certainly represents an emperor, as it bears the imperial attribute of a coronacivica with a centraljewel, and is colossal in size. Like many of the others, including the Kanellopoulos portrait,its identity must be established by process of elimination, and the carving style and surviving indications of the arrangement of the hair preclude all other choices but Trajan.Yet the long, thin face, fleshy cheeks, and head tilted down toward the left shoulder are characteristics not seen in any of his official types. Besides these generallyacceptedportraitsofTrajan and Hadrian, more problematic and controversialportraits are known. I include them here as possible examples of uncanonical portraits of these emperors, although conclusive evidence of their identity is lacking and scholars do not agree on whom they represent.The first of these, a portrait from the Athenian Agora (Figs. 20, 21), deviates drastically from Trajan'sconventional portraits,and its provenanceoffers no help with the identification. Some schol-

Figure 18. Trajan, from Pergamon, frontal view. Pergamonmuseum, Berlin. Courtesy StaatlicheMuseen zu Berlin-PreussischerKulturbesitz Antikensammlung(neg. SK 7336)

Figure 19. Trajan, from Pergamon, Museen right profile. CourtesyStaatliche zu Berlin-PreussischerKulturbesitz Antikensammlung(neg. SK 7337)

72. Silifke Museum, Inan and Alfoldi-Rosenbaum 1979, pp. 93-94, no. 42, pls. 35:3-4. They propose that the portrait may have been made after the decennalia type, but acknowledge that it must not have followed it closely.

UNCANONICAL

Figure 20. Trajan, from the Agora, Athens, frontal view. Agora Museum, no. S347. CourtesyAmerican School of ClassicalStudies at Athens, Agora Excavations(neg. 3-140)

Figure 21. Trajan, from the Agora, Athens, right profile. Courtesy American School of ClassicalStudies at Athens, Agora Excavations(neg. 3-141)

IMPERIAL

PORTRAITS

I25

ars do not accept it as a portrait of Trajan at all.73Yet its wreathed headdress strongly suggests that it was created as an imperial portrait, its overlife-size dimensions support that assertion, and its hairstyle, although somewhat fuller and more plastically rendered than on most images of Trajan, clearly mimics the pattern of locks seen in other portraits of the Optimus Princeps. Trajan seems the most likely candidate, despite the uncanonical appearance of the portrait. 73. Agora MIuseum,no. S347. For example,Fittschen and Zanker (1985) do not include it with the portraitsof any emperor.Harrison,in AgoraI, p. 27, no. 17, pl. 12, originallypublished it as a priest of the Flavian period. She laterrevisedher opinion,

however,and inAgoraPicBk5, fig. 11, identifies the image as Trajan.She has revealedin conversationthat she was persuadedby C. C. Vermeulethat the portraitwas an emperor,and became convincedthat it was Trajanafter a discussionwith MargareteBieber.

126

LEE

ANN

RICCARDI

Figure 22. Hadrian, from the

Olympieion,Athens, frontalview. NationalMuseum,Athens, no. 249. CourtesyMuseum

Another example of an unusual portrait that has been identified as Trajan is a bronze tondo from Ancyra, famous for its high quality and shield-portrait format. This portrait has not gained universal acceptance as a representation of the emperor,however, and most scholars who identify it as Trajansee it as posthumous.74The short, thin locks of hair,weary expression, modeling of the facial structure,and angle of the profile are all unparalleled in Trajan'sofficial types. If this is meant to be his portrait, considerable variation from his official types must be acknowledged. 74. AnkaraMuseum,no. 10345. Inan and Rosenbaum(1966, p. 15, note 2, no. 286, and p. 208, pl. 161) rejectit entirely as an imperialportrait,and identifyit as a portraitof a privatecitizen.Their reasoningis largelydue to the facial characteristics, especiallythe forehead, which they feel deviatetoo markedly from standardTrajanicfeatures. Additionally,they note that the plastic renderingof the irisesand pupilsin the

eyes forcesa Hadrianicor laterdate. They do not considerthe possibilitythat the portraitcouldbe posthumous. The portraitoriginallyadorneda wall in the imperialbouleuterion.It is identified as Trajanin Kosan 1957; Gokce 1957; Giuliano 1959, p. 146, no. VIII.1; and Budde 1965 and 1966. Budde suggestedinitiallythat Trajan's fatherwas represented,but later claimedit was Trajanhimself.

Vermeule(1965, p. 376) arguesthat Trajanwas the most important benefactorof Ancyra,and notes that on the reversesof local coins of Ancyra featuringTrajanare allusionsto his extensivebuilding programfor the city. Thereforethe portraitought to representTrajan,for he was likely to have been given an honor of this sort. Vermeuleagreesthat the portraitmight be posthumous.

UNCANONICAL

IMPERIAL

PORTRAITS

127

The portrait from the precinct of the Temple of Olympian Zeus in Athens (Fig. 22) has been associated with Hadrian, and provides another example.75The portrait conforms in some ways to the easily recognizable features of Hadrian'sofficial images. The arrangementof the beard, set of the mouth, and shape of the face recall the Stazione Termini type. Yet the Athenian portrait has severalunique qualities, such as the liberal drillwork used in the creation of the hair,the long locks on the forehead and over the ears, the straight eyebrows, and the expanse of the upper eyelids. These features create an image that does not closely imitate the model provided by Hadrian'sofficial portraiture,and the identification has been problematic.76The artist who skillfully executed this sculpture would have had ready access to portraits of the popular emperor,whose numerous images were prominently displayed throughout the empire, with many in Athens itself, and still chose not to follow any particularprototype or to imitate closely many of his canonical features.

NUMISMATIC PROVINCIAL

EVIDENCE FOR UNCANONICAL PORTRAITURE

In addition to the sculpted portraits, an extensive body of evidence for the appearance of imperial portraits in the provinces survives in the form of local coinage. Although they have been overlooked or dismissed by scholarsof Roman imperialportraiture,the obversesof Roman provincialcoins77 minted locally by various cities or provinces throughout the empire depict clearly labeled portraits of Roman emperors.78These coins, which are local products minted with little or no supervision from Rome, bear the portrait of the reigning emperor on the obverse and the ethnic of the issuing body on the reverse.In coins issued by official Roman mints, the portraits on the obverses certainly demonstrate the official imagery. On local provincial coins, however, the source of the portrait types is much less secure.This local coinage, made of bronze ratherthan preciousmetal, served as small change in local marketplacesfor widely circulating Roman silver brought to the cities by armies or Roman officials. Few of these bronze 75. National Museum, Athens, no. 249. Graindor1934, pp. 272-273, fig. 25; Weber 1954, p. 368; DatsouleStavridi1985, pp. 44-45, no. 249, fig. 42; Romiopoulou 1997, p. 80, no. 80. 76. Evers (1994, pp. 281-282), Fittschen (1984, p. 204), and Zanker (1983, p. 14, afterHekler 1912, p. xxvii) all maintainthat the portraitrepresents a privateindividualwith Hadrianic features.Vermeule(1954, p. 255) also questionedits identificationas Hadrian, but later reversedhis opinion (Vermeule 1968, pp. 40, 258) and acceptedthe portraitas a probableposthumousimage of Hadrianmade in Antonine times.

Wegner (1956, pp. 40-41, 62, 93, pl. 26:b, and again in Wegner and Unger 1984, p. 110); Datsoule-Stavridi (1985, pp. 44-45); Harrison(AgoraI, p. 39, pl. 45); and Bracker(1968, p. 77) all believe that the portraitrepresents Hadrian. 77. The awkwardand confusing term "Greekimperial"is gradually being replacedby the more appropriate designation"Romanprovincial." Butcher 1988, pp. 9-13. 78. Roman provincialcoins have recentlyreceivedmore attention,a trend that began with the publication of H. von Aulock'sprivatecollection of

these coins in the SNG Deutschland series and in Kraft 1972. Both of these works focused on the easternprovincial issues.The most recent projectinvolving these coins is the new series Roman ProvincialCoinage,only two volumes of which have thus far appeared:Burnett, Amandry,and Ripolles 1992; and Burnett,Amandry,and Carradice1999. Others are in variousstages of preparation, but the enormousvolume of local coinagesin later eras (particularlythe 3rd century)makes the cataloguingtasks for these latervolumes so much greater that none of the others areyet readyfor press.

128

LEE

ANN

RICCARDI

coins traveledbeyond their immediate areas,and thus were subjectto little, if any, scrutiny from Rome. Portraits of the emperors on the provincial coins present a bewildering variety of types, many of which would have been completely unrecognizable in Rome.79Examination of the portraits on these coins makes it clear that the image of the emperor had much greater regional variation and flexibility than standard literature on the Roman portrait usually allows.80 Although Roman provincialcoins were produced throughout the eastern empire until the reforms of Diocletian in the year 284, the vast majority of the surviving coins date from the Severan and post-Severan periods. For reasons that are yet unclear, fewer examples of local coins are known from earliercenturies of Roman rule.Thus evidence from the 2nd century is less abundant than it is from later eras. Local coins depicting Trajanand Hadrian do exist, however, and some conclusions can be drawn from the appearanceof the portraits on their obverses. Diversity in the depiction of the rulerscan be conclusivelyestablished,as has been overwhelminglydocumented on the more numerous examples from the 3rd century. Numismatic portraits of Trajanfrom the Greek world did not always show him with features similar to those seen on his official portraits. For example, a coin from Roman Crete with Trajanon the obverse (Fig. 23:a)8l depicts a remarkablyidealized and youthful emperor,with small eyes, large jaw, and smooth forehead. This portrait offers a striking contrast to the mature,small-jawed emperorwith prominent forehead ridge who was normally shown on coins struck at official imperial mints. On a coin from Smyrna (Fig. 23:b),82Trajanis shown with a most untypical square head, pointed nose, and smooth forehead. It is clear from these inscribed examples that eastern die-engravers could portray Trajan quite differently than did their counterparts in Rome. The label makes the identification unambiguous, despite the fact that the portrait is not readily recognizable as Trajanto those familiar with his official iconography. Portraits of Hadrian also appear in unusual ways on the provincial coinage. On a coin issued by Corinth (Fig. 23:c),83Hadrian is shown as slight and youthful, with small featuresand a long, thin nose, in contrast to the more robust and mature appearance on his official coins. On a Lacedaemonian coin (Fig. 23:d),84some of his features are borrowed from Roman models while others are inexplicably altered. Here, similar treatments of the hair and proportions of the face are juxtaposed with very different profiles and shapes of the heads. On the Lacedaemonian coin, the squat proportions of the head and bulbous representation of the back of the skull are quite unlike the long, thin proportions normally depicted on Roman coins. In a portrait used on coins produced in Roman Crete (Fig. 23:e),85Hadrian is shown with an uncharacteristicallysquare head, sharp chin, and linear features, creating an overall appearancethat is quite different from his standard,official images. Thus, great deviation from the official images on coins as well as in the sculpture of both Trajan and Hadrian as they were known in Rome was entirely possible in the Greek world. Although provincial coins are not directly comparable to provincial portraitsin sculpture,the categories are parallelin many ways, particularly

79. Despite recent interestin other aspectsof the Roman provincialcoins, they have normallynot been considered in portraiturestudies.This is probably due to the fact that, although early modern numismatistshad occasionally attemptedregionalstudies of the provincial coins (e.g., Imhoof-Bliimer 1897; Waddington,Babelon, and Reinach 1910; von Fritze 1913; and Bosch 1935), the sheer numberof coins was so overwhelmingthat these early attemptswere left incomplete.Additionally,the provincialcoins in the British Museum were excludedfrom RIC and barelydiscussedin BMC, as they do not fit comfortablyinto either category.Thus, earlyportraiture specialistswere either unawareof the existenceof provincialcoins or believed them useless in their own pursuits,and no traditionof utilizing them was established. 80. In my Ph.D. dissertation (Riccardi1996), I documentedthis varietyfor the yearsA.D. 235-270 in the Greek-speakingprovincesof the Roman world. Only about a tenth of the portraitson the coins that I examined(representing261 out of 2,570 dies and more than 12,000 coins) showed any reliancewhatsoeveron the portraitsthat appearedon coins producedin official mints. 81. SNG Cop581. 82. SNGAu12209. 83. SNG Cop284. 84. Grunaeur-vonHoerschelmann 1978, pl. 24.xxx.vii.19. 85. SNG Cop582.

UNCANONICAL

IMPERIAL

PORTRAITS

I29

a

b

c

d

Figure23. Coins representing TrajanandHadrian:(a)Trajan, RomanCrete;(b)Trajan,Smyrna; (c) Hadrian,Corinth;(d) Hadrian, Lacedaemonia;(e) Hadrian,Roman Crete. Scale1:1.CourtesyAmerican Numismatic Society (a-d) and the National Museum, Copenhagen (e)

86. For example, an edict from

Pergamonissued in Hadrian'stime

requiresthatcertainproductsbe onlywithbronze,others purchased only with silver, requiring customers to exchange their coin at the money-

tablesfora fee.Because changers'

moneychangers had to buy their licenses from the city, this system ensured revenue. Butcher 1988, pp. 25-26. 87. Coins were issued for many reasons, including for the occasion of civic festivals or an imperial visit, to show off monumental building projects, or to display various honors and titles awarded to the city. Macro 1980, pp. 682-684; Harl 1987, pp. 1-11.

e

in their relationship to Rome. The Romans allowed cities in the East to continue their local coinages in bronze (although they took over nearly all the minting in precious metal), but the remarkablediversity present in the imperial portraits makes clear that they did not watch the minting process very closely. Despite the fact that these coins did not circulate widely, they did fulfill unique and important functions, both economically86and in terms of civic pride,87yet the lack of conformity of the obverse portraits to accepted imperial iconography illustrates little or no supervision from Rome. I have argued above that the same situation existed with the display of the emperor'sportrait in civic centers, local agoras and theaters, and imperial cult areas. Rome was normally consulted in the initial stages of that process as well, yet the actual supervision of the installation was left up to the local magistrates, who may or may not have required a standardized version of the emperor's image. In neither case does the evidence support a general intention to adhere first and foremost to the latest portrait models from Rome, but rather it reveals a much more haphazard and loose relationship to the portrait models, whereby in some places at some times a close, up-todate model is followed, and in other places at other times, it is not. Examination of provincialportraits,whether sculpturalor numismatic,

LEE

I30

ANN

RICCARDI

shows that the local artist frequently was allowed to alter a Roman model in favor of local stylistic influences and in accordancewith his own training and skills. This brief survey of numismatic and sculptural portraits shows that the images of Hadrian and Trajanwere remarkablyvariable in the Greek East, with great flexibility in the treatment of their features. Virtually no characteristicof either emperor'sofficial iconographywas consistently represented in these images, and faithfulness to current Roman types was often of less concern than carving the portrait in a local style to fit local taste. Modern scholars who exclude any possibility of inspiration but the prototype-replica model are at a loss to identify portraits that do not fit into these categories, and portraits that lack context or setting can be easily misidentified. The Kanellopoulos emperor is a case in point. In conclusion, the evidence provided by provincialportraitsof emperboth in sculptureand on coins, and by the testimonies of Fronto,Arrian, ors, and Pliny the Younger makes clear that unconventional types must have been more abundant in the ancient world than we in the modern world can recognize. Even Trajanand Hadrian, long-ruling emperorswith wellknown official images, were representedthroughout Greece and Asia Minor in ways that deviated radically from their conventional appearances. Thus we cannot necessarily identify solely by physiognomy and hairstyle the many imperial portraits found without inscriptions throughout the Roman world. Images of the emperors were vastly flexible, influenced by far more than the types determined in Rome, and provincial artists did not always slavishlyimitate the models they were provided.The Kanellopoulos portrait in Athens need not be identified by postulating a series of exceptional circumstances.If one allows that the artistwho createdit only loosely followed an imperial model, the portrait's true identity as Trajan can be restored.

REFERENCES Agora I = E. B. Harrison, Portrait Sculpture (Agora I), Princeton 1953. AgoraPicBk 5 = E. B. Harrison, Ancient Portraitsfrom theAthenian Agora (AgoraPicBk 5), Princeton 1960.

Alfoldi, A. 1935. "Insignienund Tracht der romischenKaiser,"RM 50, pp. 1158.

Arce,J. 1976. "El missoriumde Teodosio I: Precisionesy observaciones," ArchEspArq 49,

pp. 19-139. Bernoulli,J. J. 1891. Romische Ikonographie 11.2: Von Galba bis Commodus,Berlin. Bliimel, C. 1933. Romische Bildnisse (Katalog Berlin), Berlin. Bosch, C. 1935. Die Kleinasiatischen

Muinzen der roimischenKaiserzeit II: Einzeluntersuchungen I: Bithynien,

Stuttgart. Boschung, D. 1993. Das romische Herrscherbild1.2: Die Bildnisse des Augustus, Berlin.

Bracker,J. 1968. "EinTrauerbildnis Hadrians aus Koln,"AntP 8, pp. 75-

84. Breckenridge,J. D. 1968. Likeness:A ConceptualHistory ofAncient Portraiture, Evanston, Ill.

. 1979. "Head of Galerius,"in The Age of Spirituality: Late Antique and Early Christian Art, Third to Seventh Century. Catalogue of the Exhibition at the Metropolitan Museum ofArt, November 19, 1977

through February 12, 1978, K.

Weitzmann, ed., New York,pp. 1314. Budde, L. 1965. "Imago clipeata des

KaiserTraianin Ankara,"AntP 4, pp. 103-117. . 1966. "Die Bildnisse des MarcusUlpius TraianusPater," Pantheon24, pp. 69-78. Burnett,A., M. Amandry,and I. Carradice. 1999. Roman Provincial Coinage II: From Vespasianto Domitian (A.D. 69-96), London.

Burnett,A., M. Amandry,and P. P. Ripoll&s. 1992. Roman Provincial Coinage I: From the Death of Caesar to the Death of Vitellius (44 B. C. -A.D. 69), London.

UNCANONICAL

Butcher, K. 1988. Roman Provincial Coins:An Introduction to the 'Greek Imperials,' London. Calza, R. 1972. Iconografia romana imperiale da Carausio a Giuliano (287-363 d. C.), Rome. Datsoule-Stavridi, A. 1985. Pcoducxox& TCoprpairCa oTO EOvzxo Mooacio

z,sAAOovac, Athens. . 1970. Untersuchungenzu den Kaiserportrdtsin Griechenland, Berlin.

Daux, G. 1958. "Chroniquedes fouilles en 1957,"BCH 82, pp. 756-830. de Kersauson, K. 1986. Muse'edu Louvre. Catalogue desportraits romains I: Portraits de la Republique et d'epoque Julio-Claudienne, Paris. Delbrueck, R. 1933. Spdtantike Kaiserportrdts: VonConstantinus Magnus bis zum Ende des Westreichs,Berlin.

Dontas, G. 1975. "CollectionPaul Canellopoulos(IX) portraitde Galere,"BCH 99, pp. 521-533. Engemann,J. 1979. "Akklamationsrichtung Sieger- und Besiegtenrichtungauf dem Galeriusbogenin Saloniki,"JAC22, pp. 150-160. Evers, C. 1994. Les portraits d'Hadrien: Typologieet ateliers, Gembloux.

Fittschen, K. 1982. "Hinterkopfe.Uber den wissenschaftlichenErkenntniswert von Bildnisriickseiten,"in Praestant interna. Festschriftfuir Ulrich

Hausmann,B. von Freytaggen. Loringhoff,D. Mannsperger,and F. Prayon,eds., Tiibingen, pp. 119-124. . 1984. "Eine Biiste des Kaisers Hadrianaus Milreu im Portugal: Zum Problemvon Bildnisklitterungen,"MM 25, pp. 197-207. Fittschen, K., and P. Zanker.1985. Katalog der riomischenPortrats in den Capitolinischen Museen und den anderen kommunalen Sammlungen der Stadt Rom I: Kaiser- und Prinzenbildnisse, Mainz. Giuliano, A. 1955. LArco di Constantino,

Milan. . 1959. "Laritrattisticadell'Asia Minore dall'89 a.C. al 211 d.C.," RivlstArch n.s. 8, pp. 146-201. Gnoli, R. 1988. Marmora romana, Rome. Gokce, N. 1957. Ankara Traianus portresi, Ankara.

Graindor,P. 1915. "Les cosmetes du Musee d'Athenes,"BCH 39, pp.241-401.

IMPERIAL

PORTRAITS

. 1934. Athenes sous Hadrian, Cairo. Gross,W. H. 1940. Das romische Herrscherbild11.2: Bildnisse Traians,

Berlin. Grunaeur-vonHoerschelmann,S. 1978. Die Muiinzprgung der Lakedaimonier (AMUGS 7), Berlin. Harl, K. 1987. Civic Coins and Civic Politics in the Roman East, A.D. 180-

275, Berkeley. Heinen, H. 1911. "ZurBegriindungdes romischenKaiserkultes:Chronologische Ubersichtvon 48 v. bis 14 n. Chr.,"Klio 11, pp. 129-177. Hekler, A. 1912. Greekand Roman Portraits, London.

Horn, R., and L. Boehringer.1966. "Die Ausgrabungsarbeitenzu Pergamon im Jahre 1965,"AA, pp. 459-476. Imhoof-Bluimer,F. 1897. Lydische Stadtmiinzen: Neue Untersuchungen, Leipzig.

Inan,J., and E. Alfoldi-Rosenbaum. 1979. Romische undfruihbyzantinische Portratplastik aus der Tiirkei:Neue Funde, Mainz.

I3I

Macro,A. D. 1980. "The Cities of Asia Minor under Roman Imperium," ANRWII.7.2, pp. 658-697. Makaronas,C. J. 1970. TheArch of Galerius at Thessaloniki,

Thessaloniki. Massner,A.-K. 1988. "Corona Civica, Priesterkranzoder Magistratinsignie?BildnisseThasischer Theorei?"AM103, pp. 239-250. Maxfield,V. A. 1981. The Military Decoration of the Roman Army,

Berkeley. McCann, A. M. 1968. The Portraitsof Septimius Severus,A.D. 193-211

(MA4AR 30), Rome. Meischner,J. 1986. "Die Portratkunst der ersten und zweiten Tetrarchie bis zur AlleinherrschaftKonstantins: 293 bis 324 n. Chr.,"Al, pp. 223250. . 1988. "Theodosianische Portrats,"in Ritrattoufficiale e ritrattoprivato. Atti della II Conferenza internazionale sul ritratto romano, Roma, 26-30 settembre1984,

N. Bonacasaand G. Rizza, eds., Rome, pp. 375-382. . 1990. "Das portratder Roman and Early Byzantine Portrait Theodosian Epoche I (380 bis 405 Sculpture in Asia Minor, London. n. Chr.),"JdI105, pp. 303-324. Kent,J. P. C. 1978. RomanCoins, . 1993. "Der Hochzeitkameo des London. Honorius,"A, pp. 613-619. Kiilerich, B. 1993. Late Fourth Century . 1995. "Studienzur spatantiken Classicism in the Plastic Arts: Studies in the So-called Theodosian Renaissance, 110, Kaiserikonographie,"JdI Odense. pp. 431-446. . 1996. "DasMissorium des Kiss, Z. 1984. Etudes sur leportrait Theodosius in Madrid,"JdI 111, imperial romain en Egypte, Warsaw. Kleiner,D. E. E. 1992. RomanSculpture, pp. 389-432. New Haven. Meyer, H. 1980. "Die Frieszyklenam Kosan, M. 1957. Buste en bronze de sogenanntenTriumphbogendes Galeriusin Thessaloniki,"JdI95, Trajan decouvert a Ankara, Ankara. Kraft, K. 1972. Das System der pp. 374-444. kaiserzeitlichen Muinzprcagungin Pekary,T. 1985. Das roimische Herrscherbild111.5:Das romische Kleinasien, Berlin.

Inan,J., and E. Rosenbaum.1966.

Lattanzi, E. 1968. 1 ritratti dei cosmeti nel Museo nazionale diAtene, Rome. L'Orange,H. P. 1933. Studienzur Geschichtedes spitantiken Portrats,

Oslo. . 1984. Das romischeHerrscherbild III.4: Das spdtantike Herrscherbildvon Diokletian bis zu den KonstantinSohnen, 284-361 n. Chr., Berlin.

L'Orange,H. P., and A. von Gerkan. 1939. Der spdtantike Bildschmuckdes Konstantinsbogens, Berlin.

Kaiserbildnis in Staat, Kult, und Gesellschaft,Berlin. Pond Rothman,M. S. 1977. "The Thematic Organizationof the Panel Reliefs on the Arch of Galerius," AJA81, pp. 427-454. Poulsen,V. 1974. Lesportraitsromains a la Basse-Antiquite, II: De Vespasien Copenhagen. Ragona,A. 1963. Itetrarchi deigruppi porfirei di S. Marco in Venezia,

Caltagirone.

132

LEE

Riccardi,L. A. 1996. "RomanImperial Portraiturein the EasternProvinces: A Study of the Reception and Rejectionof ImperialModels"(diss. Boston University). Richardson,L., jr. 1975. "The Date and Programof the Arch of Constantine," ArchCI 27, pp. 72-78.

Rolley,C., and F. Salviat.1963. "Une statue d'Hadriensurl'agorade Thasos,"BCH 87, pp. 548-578. Romiopoulou,K. 1997. E,Arlvopoaijxda y2vraZ TOo EOvixo6 Mooauseoo, ApXacooyYlxo

Athens. Rose, C. B. 1997a. DynasticCommemoration and Imperial Portraiture in the

Julio-ClaudianPeriod,Cambridge. . 1997b. "The ImperialImage in the EasternMediterranean,"in The Early Roman Empire in the East,

S. Alcock, ed., Oxford, pp. 108-120. Rumpf,A. 1957. Stilphasender spdtantikenKunst:Ein Versuch, Cologne. Riisch,A. 1969. "Das kaiserseitliche Portratin Makedonien,"JdI84, pp. 59-196. Scherrer,P., ed. 1995. Ephesos.Der neue Fuhrer,Vienna. SNGAul = SNG Deutschland: Sammlung

v. Aulock111.6:lonia, Berlin 1960. SNG Cop = SNG Copenhagen:The Royal Collection of Coins and Medals, Danish

NationalMuseum1.17, Copenhagen 1944.

LeeAnn Riccardi THE

COLLEGE

DEPARTMENT

P.O.

BOX

EWING,

OF NEW OF

7718 NEW JERSEY

[email protected]

JERSEY

ART

o8628

ANN

RICCARDI

Srejovic,D. 1992-1993. "APorphyry Head of a Tetrarchfrom Romuliana (Gamzigrad),"Starinar,pp. 41-47. . 1995. "ImperialHead Unearthedin Serbia,"Archaeology 48.1, p. 14. Stephanidou-Tiveriou,T. 1995. To Mtxpo T6o zoo Fcraep'oo Czy OEaoaAov(cx,,Athens. Strack, P. L. 1931. Untersuchungenzur romischenReichsprdgungdes zweiten Jahrhunderts I: Die Reichspragung zur Zeit des Traian, Stuttgart.

Stuart,M. 1939. "How Were Imperial PortraitsDistributedthroughoutthe Roman Empire?"AJA43, pp. 601617. Swift, E. H. 1923. "Imaginesin Imperial AJA27, pp. 286-301. Portraiture," Tuiirkoglu,S. 1972. Efes Harabeleri ve MuiizesiYilligi (Annual of the Ruins and Museum ofEphesus) 1.

Vermeule,C. C. 1954. Rev. of AgoraI, in AJA58, pp. 253-255. . 1965. "AGreekTheme and Its Survivals,"PAPS109, pp. 361-397. . 1966. "Hellenisticand Roman CuirassedStatues:Second Supplement,"Berytus16, pp. 49-59. . 1968. Roman ImperialArt in GreeceandAsia Minor, Cambridge,

Mass. Versnel,H. S. 1970. Triumphus:An Inquiry into the Origin, Development, and Meaning of the Roman Triumph,

Leiden.

von Fritze,H. 1913. Die antikenMuiinzen Mysiens I: Adramytion-Kisthene,

Berlin. Wace, A. J. B. 1907. "Studiesin Roman Historical Reliefs,"BSR 4, pp. 270276. Waddington,W. H., E. Babelon, and T. Reinach. 1910. Recueilgeneraldes monnaies grecquesd Asie Mineure 1.1-

4, Paris. Walker,S. 1994. "The ImperialFamily as Seen in Cyrene,"LibSt 25, pp. 167-184. Weber,H. 1954. Rev. of AgoraI, in Gnomon26, pp. 364-374. HerrWegner,M. 1939. Das ro5mische scherbild11.4: Die Herrscherbildnissein

antoninischer Zeit, Berlin. . 1956. Das romischeHerrscherbild 11.3: Hadrian, Plotina, Marciana,

Matidia, Sabina,Berlin. Wegner,M., J. Bracker,and W. Real. 1979. Das roimische Herrscherbild 111.3: GordianIII bis Carinus,Berlin. Wegner,M., and R. Unger. 1984. "Verzeichnisder Bildnissevon Hadrian,"Boreas7, pp. 105156. Wiggers, H. B., and M. Wegner. 1971. Das romischeHerrscherbildIII.1: Caracalla, Geta, Plautilla, Macrinus bis

Balbinus,Berlin. Zanker,P. 1983. ProvinzielleKaiserportrdts: Zur Rezeption der Selbstdar-

stellungdesPrinceps,Munich.

Recently Published

CATHERINE MOR6AN

The Late Bronze Age Settlement and Early Iron Age Sanctuary

566 pages, 74 plates, 102 figures (includingmore than 400 profile drawings),6 plans, 6 tables,4 appendixes IsthmiaVIII. ISBN 0-87661-938-3. December 1999. Cloth $100.00

This volume presents Late Bronze and Early Iron Age evidence (pottery, metalwork, terracottas,architectureand other constructions) from excavations conducted by the University of Chicago at the Sanctuaryof Poseidon at Isthmia between 1952 and 1989. Stylistic analysis of artifacts offers important new information on Corinthian production: Isthmia has produced the first substantial collection of Early Iron Age Corinthian terracottas,for example, as well as 8th-century human figure depictions. Functional analysis, developing established methodology for site characterization, distinguishes Late Bronze Age settlement from Early Iron Age cult activity. Thus Isthmia may be counted among the growing number of Greek shrines established during the Bronze Age/Iron Age transition, and the nature and variety of cult practices compared with those elsewhere. In its Corinthian context, Isthmia offers unique insights into 800 years of development, from Mycenaean province to Archaic polis.

MARTHA HEATH WlIENCKE

Forthcoming

The Architecture, Stratigraphy and Pottery of Lerna III

The Lerna III, or Early Helladic II, discoveries at the site of Lerna in the Argolid, from the excavations made during the 1950s, have provided the largest body of material yet available from that crucial period in the prehistory of southern Greece. The finds have been known until now chiefly from the preliminary reports published in Hesperiaby the director,J. L. Caskey. In the present volume the evidence for the stratification and architectural remains of the Lerna III phase is presented in full, with plans, sections, and photographs, including a detailed description, tentative reconstruction, and discussion of the well-known House of the Tiles. A complete catalogue of the context pottery with drawings follows, arranged chronologically according to newly established categories of shape and fabric. In following chapters the author discusses the history of the individual shapes, together with much comparativematerial,as well as painted patterns, tactile decoration, and potters' marks. It is hoped that this first detailed study of the development of EH II ceramics will be of help in future work. An appended report by P. P. Betancourt and G. H. Myer presents the petrographic analysis of a select group of sherds. Small objects including lithics, listed by location in the stratigraphicsection, will be fully presented in future monographs by E. C. Banks and C. Runnels.

In two volumes: fascicle 1: 332 pages, 108 figures,39 plans, 29 sections; fascicle2: 494 pages, 103 plates,24 plates, 3 appendixes,37 tables LernaIV. ISBN 0-87661-304-0. May 2000. Cloth $125.00

HESPERIA

SUPPLEMENTS

13 MarcusAurelius. Aspects of Civic and Cultural Policy in the East, byJames H. Oliver (1970). $15.00 14

The Political Organization ofAttica, by John S. Traill (1975). $15.00

15

The Lettering ofranAthenian Mason, by Stephen V. Tracy (1975). $15.00

16 A Sanctuary of Zeus on Mount Hymettos, by Merle K. Langdon (1976). $15.00

17 Kallias of Sphettosand the Revolt ofAthens in 286

B. C.,

by T. Leslie Shear Jr. (1978). $15.00

19 Studies in Attic Epigraphy, History, and TopographyPresented to Eugene Vanderpool(1982). $15.00 20

Studies in Athenian Architecture,Sculpture,and TopographyPresented to Homer A. Thompson(1982). $15.00

21

Excavations at Pylos in Elis, byJohn E. Coleman (1986). $25.00

22 Attic Grave Reliefs That Represent Womenin the Dress of Isis, by Elizabeth J. Walters (1988). $40.00 23

IHIellenistic Relief Moldsfrom theAthenian Agora, by Claireve Grandjouan (1989). $25.00

24

The Prepalatial Cemeteriesat Mochlos and Gournia and the House Tombsof BronzeAge Crete,byJeffrey S. Soles (1992). $35.00

25

Debrisfrom a Public Dining Place in theAthenian Agora, by Susan I. Rotroff and John H. Oakley (1992). $35.00

26

The Sanctuary ofAthena Nike in Athens. ArchitecturalStages and Chronology, by Ira S. Mark (1993). $50.00

27

Proceedingsof the EnternationalConferenceon GreekArchitectural Terracottasof the Classical and Hellenistic Periods, December12-15, 1991, edited by Nancy A. Winter (1994). $120.00

28

Studies in Archaic Corinthian VasePainting, by D. A. Amyx and Patricia Lawrence (1996). $65.00

29

TheAthenian Grain-Tax Law of374/3 B.c., by Ronald S. Stroud (1998). $35.00

H THE OF

ESPE

JOURNAL

OF THE

CL A S S I C A L

AMERICAN

STUDIES

AT VOLUME

SCHOOL ATTHENS 69:

APRIL-JUNE

'-1/

I

.(

NUMBER

2000

~~~~~~~~~~~~~~~I

American Schoolof Classical Studies at Athens 2000

2

The American School of Classical Studies at Athens is a research and teaching institution dedicated to advanced study of the archaeology,art, history, philosophy, language, and literature of Greece and the Greek world. Established in 1881 by a consortium of nine American universities, the School now serves graduate students and scholars from more than 150 affiliated colleges and universities, acting as a base for research and study in Greece. The main buildings of the School and its library are located in Athens, with administrative and publications offices in Princeton, New Jersey. As part of its mission, the School directs ongoing excavations in the Athenian Agora and at Corinth and sponsors all other American-led excavations and surveys on Greek soil. It is the official link between American archaeologists and classicists and the Archaeological Service of the Greek Ministry of Culture and, as such, is dedicated to the wise management of cultural resources and to the dissemination of knowledge of the classical world. Inquiries about membership in the School or participation in the Summer Sessions should be sent to the American School of Classical Studies at Athens, 6-8 Charlton Street, Princeton, New Jersey 08540-5232. Hesperiais published quarterlyby the American School of Classical Studies at Athens. Founded in 1932 and devoted primarily to the timely publication of reports on School-sponsored and School-directed projects, Hesperiawelcomes submissions from all scholars working in the fields of Greek archaeology,art, epigraphy,history, and literature,from earliest prehistoric times onward. Hesperiais a refereedjournal.

VOLUME

69:

APRIL-JUNE

NUMBER

2

2000

THE

JOURNAL

OF THE

OF CLASSICAL

PUB LI CAT IONS STAFF EDITOR-IN-CHIEF

DANIEL

STUDIES

AMERICAN

SCHOOL

AT ATHENS

J. PULLEN

The Prehistoric Remains of the Acropolis at Halieis: A Final Report

133

KerriCox ED ITO R, Hesperia

ALAN

W.

JOHNSTON

TraceyCullen ASSOCIATE

EDITOR

Building Z at Kommos: An 8th-Century Pottery Sequence

189

Michael Fitzgerald EDITORIAL

AsSISTANT

SuzanneAbrams PRODUCTION

MANAGER

STEPHEN

V. TRACY

Athenian Politicians and Inscriptions of the Years 307 to 302

227

SarahGeorge Figueira PRODUCTION

ASSOCIATE

Carol Ford MARKETING

MANAGER

PatriciaTanner

PUBLICATIONS COMMITTEE Carol C. Mattusch (Chairman) George Mason University Darice Birge Loyola Universityof Chicago Kevin Clinton Cornell University Jack L. Davis University of Cincinnati

JeniferNeils Case Western ReserveUniversit KathleenW. Slane Universityof Missouri-Columbi Stephen V. Tracy (ex officio)

The Ohio State University

LINDA JONES

Roccos

Back-Mantle and Peplos: The Special Costume of Greek Maidens in 4th-Century Funeraryand Votive Reliefs

235

Submissions: Manuscripts and communications should be addressed to Dr. TraceyCullen, Editor,Hesperia, American School of ClassicalStudies at Athens, 6-8 CharltonStreet,Princeton,NewJersey08540; tel. 609-683-0800; fax 609-924-0578; e-mail [email protected] photocopies of illustrationsmust be submittedin triplicate;originalartworkand photographs should not be sent unless prior arrangementsare made with the Editor. A short abstractsummarizingthe majorconclusionsof the articleshouldalso be included.Articles aresubmittedto a double-blindreviewprocessand authors are requestedto preparetheir manuscriptsaccordingly,without their name or affiliationappearing.The style for manuscriptpreparation,notes, bibliography,and other informationon submissionscan be found in the Guidelinesfor Authors published in Hesperia 62, 1993, pp. i-xvi; on the School's website (www.ascsa.org);or by writing to ASCSA Publicationsat the above address.

Copyright (? 2000 The American School of ClassicalStudies at Athens

The AmericanSchool of ClassicalStudies at Athens will not knowinglyprint in Hesperia or any of its other publicationsthe announcementor initial scholarlypresentationof anyobjectacquiredafterDecember30, 1973, by anymeans other than through an officially sanctioned excavationor survey,unless the objectwas partof a previouslyexistingcollection or was legallyexportedfrom the countryof origin.

ClassicalStudies at Athens, 6-8 Charlton Street,Princeton, New Jersey08540-5232 U.S.A. ISSN 0018-098X ISBN 87661-500-0

Subscriptions:The annualsubscriptionprice,payablein advancein dollars,is $55.00 for individuals,$65.00 for institutions, and $33.00 for students with valid I.D. Canada and other countries add $9.50 postage. We accept VISA and MasterCard.Published quarterly.Reprintsof Hesperia 1-41, Index \olume 1 (Hesperia1-10 and Supplements 1-6), and Supplements 1-12 and Supplement 18 should be orderedfrom Swets and Zeitlinger,b.v., P.O. Box 810, 2160 SZ Lisse, Netherlands. Single issues (currentand back numbers when available)of Hesperia 42 and following areavailablefor $15.00 each plus postagefromthe AmericanSchoolof ClassicalStudiesat Athens,6-8 Charlton Street,Princeton,NewJersey 08540-5232 U.S.A. OrderIndex Volume2 and Supplements 13 and following from UniversityMuseum Publications,Universityof PennsylvaniaMuseum of Archaeologyand Anthropology,33rd and SpruceStreets,Philadelphia,Pennsylvania19104 U.S.A.

Producedat EdwardsBrothers, Inc., Ann Arbor,Michigan. Design by Ellen McKie. Periodicalspostage paid at Princeton,New Jersey,and at additionalmailing offices. Postmaster:Send addresschanges to Hesperia, American School of

HESPERIA

69,

2000

TH

E

P RE

H

R IC

ISTO

Pages 133-187

REMAINS AC

A

In Memoriam Jane Hackbarth Leslie

1. The publicationof the prehistoric materialfrom the Halieis acropolisis offeredhere in anticipationof the final reporton the post-prehistoricmaterial, which is in preparation.I would like to thankThomas W. Jacobsenand Michael H. Jamesonfor entrustingthe publicationof the Halieis acropolis prehistoricmaterialto me. Jacobsen kindly turnedover to me all documentation and notes relatingto the prehistoricmaterials.I have made extensiveuse of Jacobsen'sdraft manuscript,dated to 1974, and where appropriate,I include portions here. I examinedthe materialin the Nauplion ArchaeologicalMuseum first in a preliminaryfashion in 1982 and subsequentlymore fully in 1997. Contextualinformationwas provided by JamesA. Dengate. The final drawingsof the ceramicsand some of the lithics were inked by CatherineS. Alexander.I would also like to thank ChristinaM. Dengate,JamesA. Dengate, P. Nick Kardulias,Catherine Perlks,Curtis N. Runnels, and K. D. Vitelli for their observations,help, and advice.I am also gratefulto the anonymousHesperiareviewersfor their comments. 2. Jameson 1969, 1972; Boyd and Rudolph 1978. 3.Jameson 1969, pp. 318-322; 1972, pp. 233-234.

ROPO FINAL

OF L IS

THE AT

HA

L IEI

S

REPORT

ABSTRACT Excavationsin the 1960s and 1970s on the acropolisof Halieis in the southern Argolid revealedmaterialof Final Neolithic through Early Helladic I in deposits dating to the Archaic through Classical periods. Post-prehistoric building activities have disturbedany originally in situ prehistoricdeposits. The Halieis ceramics are later than those from the nearby Franchthi Cave, but comparewell with the ceramicscollected from the surroundingregion by the Southern Argolid Survey.A single radiocarbondate derived from shell yields a marine-correcteddate range in the 4th millennium B.C. Prehistoric materials were found over a wide area of the acropolis at the ancient city of Halieis on the south shore of Porto Cheli Bay in the southern Argolid during the course of excavations begun in 1962 (Figs. 1-4).1 Excavations at Halieis were conducted by the University of Pennsylvania and Indiana University in 1962 and 1965-1976, with some small-scale investigations subsequent to 1976.2 A fortified wall enclosing a planned town with harborworks, an extramuralsanctuaryof Apollo including two temples, an altar,and a race course, and an extramuralnecropolis, all of the Archaic to Hellenistic period, were explored; there is also evidence for Geometric and Late Roman occupation of the Lower Town and harbor area. The acropolis (Fig. 4), the highest point of the fortification on the south edge, was exploredprimarilyin 1962,1965-1966, and again in 1971, under the direction of Michael H. Jameson and Charles K. Williams.3 A stone circuit wall of the late 6th or early 5th century B.C., preceded by mudbrick fortifications of the Archaic period, marksthe south edge of the acropolis, while inside (to the north) are buildings associated with the manning of the walls. A small open-air cult area with an altar and two other stone monuments lie to the northeast, while to the west a roadway leads up from the lower levels to another structureof the 5th century B.C. The acropolis currentlyrises to a height of 51 masl. The hill is part of a small ridge that separates the bay from the Argolic Gulf to the south (Fig. 2). The physical setting in the prehistoric period would have been

I34

DANIEL

J. PULLEN

Corinthian

Gulf

'6~~~~3 ~

~

Tsoungiza(Nemea

~

Zygouries

Mycenae cQ~~~~~~~~~igna,

~ CZ ~

?As Lern~ali

*Corinthigur

~ethana 1.Saroicintenrhatr

Argoli

of

GREECEZ?NE

PELOPONNESE Figure 1. Halieis in the northeastern Peloponnese

different;changes in sea level have affected the morphology of Porto Cheli Bay. Little detailed information is available for reconstructing the bay in the prehistoric period, but a generalized picture can be obtained from the extensive work on shorelines elsewhere in the southern Argolid.4 At ca. 5000 B.P. the shoreline would have stood 6-8 m below the level today, greatly reducing the size of Porto Cheli Bay and extending the Argolic Gulf coast slightly (to include, probably, the small offshore island of Khinitsa). The bay would still have been present, most likely as a narrow inlet (Fig. 5).5 The amount of low-lying land to the north of the acropolis would have been much greater than at present. The acropolis site would thus have had numerous advantages, such as good views, a fertile lowlying plain to the north, and easy access to the sea. None of the prehistoric ceramic or lithic items discovered on the acropolis can be placed in a secure prehistoric context; nevertheless, there is good reason to believe that the appearance of prehistoric materials in the otherwise historic deposits on the acropolis is not fortuitous. Rather, the building and leveling activities of the Archaic and Classical periods apparentlydisturbed remains of the prehistoric era, and sherds, obsidian, and chert items were mixed in with the later deposits. There are apparent

4. Jameson,Runnels, and van Andel 1994, pp. 200-210; van Andel 1987. 5. See Jameson,Runnels, and van Andel 1994, pp. 233-234, figs. 4.104.13 for LN-EH II shorelinereconstructionand site distribution.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I35

aaS

Figure 2. Aerial view of Porto Cheli Bay (left), Halieis (center), and

SaronicGulf (background),looking east-southeast

6. Apart from the occasionalpiece of obsidian,no indication of prehistoric activity,in the form of either pottery or structuralfeatures,was reportedfrom the excavationsbyJameson on the IndustrialTerraceto the east of the acropolis.Obsidian and prehistoric pottery,however,turnedup in the excavationsconductedby Wolf W. Rudolphin the LowerTown in 1970, 1972, and 1974. The latter materialwill not be consideredhere but will be publishedin due coursealong with the other finds from the LowerTown. 7. Karduliasand Runnels (1995) discussthe presenceof historic-period lithics at sites in the southernArgolid; see also Runnels 1982 for a general discussionof historic-periodflaked stone tools.

concentrations of prehistoric materials, including some in shallow cuttings in the bedrock, strengthening the suggestion that the prehistoric materials came from the acropolis itself and were not brought in from elsewhere. The prehistoric material was identified during the sorting of the pottery,primarilyon visual criteria.Because of the limited stratigraphic data for this material,it can best be approachedin stylistic and typological terms, much as if it derived from a surface survey. The majority of the prehistoric ceramics from the Halieis acropolis can be dated to the Final Neolithic and Early Helladic I periods; there are in addition a limited numberof items that can be dated to the Early Helladic II and Late Helladic periods.6 The lithics are more difficult to date, as there is a long tradition of the use of obsidian and chert, even into modern times.7 Most of the obsidian from Halieis, however, does have strong stylistic affinities with FN-EH lithic assemblages elsewhere in the Aegean. In this article, a brief consideration of the excavation and contexts of the prehistoric material precedes the discussion of the ceramics, lithics, and other finds. The single radiocarbon determination is discussed in the section on chronology. An assessment of the significance of the Halieis prehistoric material and the relationship of this material to that from the rest of the southern Argolid in particularand the Aegean more generally concludes this contribution.

EXCAVATION AND CONTEXTS Much of the prehistoric material has been found in the center of the hill near the northern limits of the excavated area, where the conglomerate bedrock was generally higher than elsewhere and where the stratigraphy was less clear,perhaps because of the comparativeshallowness of the over-

18~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~O

Porto Chell Bay 14

Gate 1

12

\\

-

<

1/

/~~

/1

//

//

/

WestGate

~

3

3. Halieis, site plan Figure

~

~

~ ~

~

~

~

~~~~ ,~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~ ~-modern ~~ s-horeline

~

~

~

Hermio

'

''

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~~~~~oteatGt

/O~~~~~~~~~~~~~~~~~~~~~~~~

THE

+

:51

.. +

63

.-:-:b:@4:+

tX

PREHISTORIC

.M54

::t ';, Plndmpsa 'S::X D.J ............

55

57

as

4

OF THE

REMAINS

58

4

CW.

*6-6--,

. ..

61

........ >

0 . . . . . . . .

.E

t2:F:b

.nds

~ ~ ~ ~ ~ 4~ ...... cv~~~~~~~~...C cu............ 4~~~~~~~~~~~~~~~~~~~F *4

Cw

6o

I37

I

Findlspot - *.

59

AT HALIEIS

~~~~~~~~~~~~~~~~~~~ .......

0;; Flr.... ..... -

4

ACROPOLIS

:vBuildig

4

B

4~~~~~~~~~~~~~~~~~~~~~AtrAe

cx 4

~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Figure4. Locationof major findspotson the Halieisacropolis

8. The following three paragraphs are adaptedfromJacobsen's(1974) unpublishedmanuscript. 9. See Jameson 1972, pp. 233-234. 10. Despite this apparentlack of undisturbeddeposits, statisticaltests on the ceramics,discussedbelow,indicate that the distributionof some of the ceramicclassesis not random;there may be some chronologicalor functional significanceto the distributionof ceramics.

.....

lying deposit (Figs. 5-8).8 In this areabedrock usually lay less than a meter below the modern surface, normally at depths ranging between 0.60 and 0.80 m. Very little material was found in or around the structures along the southeastern flank of the hill where the archaeological deposit was deeper. Indeed, a small sounding carried down to bedrock in 1965 in the areajust to the east of the Round Tower produced no trace whatsoever of prehistoric activity. Much prehistoric material, especially obsidian, came from the southwestern flank of the hill. In those areas where prehistoric material was found, it occurred almost invariably in contexts containing post-prehistoric remains as well. Although most commonly found on or just above bedrock, even then it was usually mixed with later material,as was alwaysthe case when it turned up in surface deposits. Excavations conducted by Williams in 1971 in the western sector of the hill produced a small amount of material in fill associated with leveling activity of the 6th century B.C. The presence of prehistoric finds in this leveling fill not only illustrates the disturbing effect of later building activity but, as Williams has suggested,9may help to explain the comparativefrequency of such finds in hollows or fissures (presumably largely of natural origin) in bedrock in the central part of the hill. In only one instance was there reason to believe that the excavateddeposit had not been contaminated by later intrusions or disturbance.'0In 1965 a small

DANIEL

I38

,'gls''sF'o' AdMi:i

i

,9

,

*

J. PULLEN

0

_~~~~

-

Al_

-

__~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i

hollow or pit ca. 1.00 m deep (maximum depth below modern surface, 1.35 m) was excavatedjust to the north of the small structure(Building A) near the center of the hill (findspot 4; see Figs. 4, 8). Although excavation was made difficult by the size of the pit, it exposed a deposit of dark earth containing a number of marine shells and a small amount of prehistoric pottery in rather fragmentary condition. A sample of the shells collected from the pit was later submitted to the Radiocarbon Laboratory at the University of Pennsylvania, and the resulting determination (see below) represents the only indication of absolute chronology from the site. Excavation produced no evidence of structuralfeatures that could be associated with the prehistoric ceramic remains.This absence, along with the considerablydisturbed stratigraphyin the deposits lying directly upon bedrock, implies that any remains of a prehistoric settlement that may have existed on the hill were obliterated by the extensive building activity that took place there during the Archaic and Classical periods. In view of this, it is important to state at the outset that almost nothing can be said about the stratigraphicdistribution of the remains discussed below. The prehistoric ceramics come from over fifty-five different lots and findspots throughout most of the excavated areas of the acropolis, especiallywhere relativelydeep excavationtook place.Altogether approximately 400 sherds were identified as "prehistoric."A few contexts, however, supplied greater quantities; indeed, the ten findspots listed in Table 1 account for four-fifths of the prehistoric ceramic material (see Fig. 4 for locations). The lithics are fewer in number but also more widely scattered than the ceramics.1"But again, a few contexts (Table 1), all of which are

Figure5. Halieisacropolis.General view lookingwest,with entranceto PortoCheli Bayin backgroundand BuildingB in foreground.

11.The scattereddistribution is due in partto a differentmannerof the findspotsof the lithic recording The obsidianwasoftennoted material. at the timeof excavation, whereasthe ceramicswereselectedin the preliminarysortingafterexcavation.

THE

PREHISTORIC

REMAINS

OF THE

i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-m1

ACROPOLIS

AT HALIEIS

I39

m

______~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~., ................ ....., __~~~~~~~~~~~~~~~~~~~ .j=__ .,,,., iS S_ |

_

|

_

S

l

|

|

M

|

i

',

.A:

::.

.,

.

.......

.................................................... ..:0-<*.'*< ,.y,

....

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... .. .. . .... .. .... . ..E -,:. i l

Figure 6. Halieis acropolis. General view looking east, with Building A in center, Building B to right, and altar area in background on the left.

_

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ... ..

_

*~~~~~~~~~~~~~~~~~~~. .,

k,

...._.;...

.>~~~~~~~~~~~~~~~

. . 2 ] ' : .. . ....

>

[|

S

|_00w3'

''

s

,

................................................................A

.

.............................

. .. ......

*u

:. .; e e#< .:S? .. .. . ,,;> ?-

m'PwE

.. -.W::,

~~~~~~~~~~~~~~~~~~ ,........1... .... M-' .

..

...

......-__A-

,@!16

0'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~:u tEri"

-

~ ~~ ~ ~ ~ ~ ~ ~~P

.... .... . .

:

;

.

2? w,

..

..

.

..

.~~~~~~~~~~~~~~~. .....

.. ....

. .

.

..

.

. .

.

.,

.,

..

..

,

.

S..

w

.

DANIEL

I40

J. PULLEN

--

~~~~~~~~iuImEbu~~~~~~~~

HSW~~~~~~~~

among those producing the majority of the ceramic material, provided nearly three-fifths of the lithic material (from a total of some 114 pieces from the acropolis).

THE POTTERY Given that the Halieis prehistoric material derives from chronologically later contexts, it is difficult to identify with certainty what the assemblage may have originally looked like. The prehistoric ceramics were selected from lots that contained mostly post-prehistoric material.Thus, there is a distinct possibility that some classes of prehistoric ceramicswere not identified and selected for study. But given the experience of the staff during the excavations and the large quantity of plain body sherds selected as "prehistoric,"this possibility of certain classes of materialbeing overlooked or underrepresentedis probably quite small.12 Added to the difficulties of characterizing the ceramic assemblage is the presence of a few certainly identifiable EH II and LH sherds.The ceramics are first discussed by class and shape, followed by a consideration of their chronological and cultural position. The majorityof the prehistoric ceramics fall into one of severalreadily recognizable classes.13These ceramic classes are based on a combination of three factors: macroscopic observations of the fabrics, painted decora-

Figure8. Halieis acropolis.Building A, lookingsouthwest.Prehistoricpit northof BuildingA at extremeright. In background,BuildingB, room2, cleft in bedrockbelowfloor. 12. The 1967 season saw the beginning of excavations,under the directionofJacobsen, at FranchthiCave; the upperlevels producedplentiful ceramicssimilarto those from Halieis. In 1972 the first season of the Southern Argolid Surveytook place, utilizing some of the same staff;Bronze Age materials were found at a numberof sites. Among the Halieis acropolismaterialhoused in the Nauplion Museum in 1997, I identified no more than a half dozen sherdsoriginallyclassifiedas "prehistoric" as post-prehistoric.These include three Geometric body sherdsand a small portion of a moldmadelamp. 13. See Lerna III, pp. 11-14, for a fullerdiscussionof the history and problemsof the classificationof prehistoricpottery.I have followed Rutter'ssystem of "classes"here.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I4I

TABLE 1. MAJOR FINDSPOTS OF PREHISTORIC CERAMICS AND LITHICS Findspot Location 1 2 3 4 5 6 7 8 9 10

Ceramics Lithics

Acropolistrench2, northeastof altararea Acropolis trench3, north of altararea AcropolisW, Mess Building destructiondebris Acropolis CS 56/57, pit in bedrocknorth of Building A Acropolis E, east of Round Tower AcropolisW, destructiondebrissouth of drainand road and north of fortificationwalls, at west end of excavatedarea Acropolis E, Building B, room 2, cleft in bedrock Acropolis B East, depressionwest of rubbleterracewall Acropolis E, CU 60/61, basket4, cleft in bedrock,southeastof altarand northeastof squaretower AcropolisE, southeastof altararea,"bothros"with Classicalpottery

73 67 42 27 6

1 2 8 0 5

58 4* 16 19 10

49 4 1 0 0

*Historical.

14. Pullen 1995, pp. 13-14. 15. Karduliasand Runnels 1995. Andesites are also found on volcanic islands such as Melos, Santorini,and Nisyros in the Aegean. They have been reportedfrom elsewhere,such asTyros, "southof Lerna"at Leonidion (Shriner and Dorais 1999, p. 45, citing Lekkas and Papanikolaou1977) and in the southernArgolid itself nearFranchthi Cave, but so far there is no evidence for andesitesfrom these latter two sites having been exploited in prehistoric times. 16. The volcanic-temperedclasses can also be called"Saronic"because their distributionseems to be primarily in regionsborderingthe SaronicGulf (personalobservation);see Mee and Taylor1997 for the only significant publicationof materialfrom such a region outside the southernArgolid. Vitelli (Franchthi8, Franchthi10) uses the term "Andesite"ware for Neolithic fabricsfrom FranchthiCave with these minerals.

tion and surface treatment, and color.The primary distinction is between those classes that have tempering particles resembling volcanic minerals and those that do not. The importance of this distinction was first recognized in the Southern Argolid Survey where, among the EH I ceramics, two of three major fabric groups had minerals identified macroscopically as of volcanic origin;14 these volcanic-mineral tempered fabrics correlated with certain shapes and decorative treatments. The volcanic minerals in the ceramics are similar to those found in andesite, a major material for grinding stones in the southern Argolid. Karduliasand Runnels conclude that the majority of Neolithic and Early Helladic grinding stones in the southern Argolid are of andesites from Aigina.15Though no petrographic studies have been made of either the Halieis or southern Argolid prehistoric ceramics,I use the term "volcanic-tempered"to distinguish this readily identifiable fabric.16 Volcanic-tempered pottery with inclusions from volcanic stones such as andesite, identifiable by the presence of black and/or gold micalike inclusions, forms one major group of ceramics.This fabric can be found in three classes based on surface treatment and color: Volcanic Red-Slipped and Burnished, Volcanic Black-Burnished, and Volcanic Plain, with the Plain and Red-Slipped classes very similar to each other. Nonvolcanic-tempered pottery forms the other major group. Several classes in the nonvolcanic group have very little overlap in surface treatments with the volcanic-tempered classes.The fabrics display a wide range of colors, types and quantities of tempering particles, and surface treatments. Each of the nonvolcanic-tempered classes is treated separately. Brown-Slipped and Dark-Faced Burnished classes: a ware that is treated very much like the volcanic ware, but lacks the typical volcanic minerals. Many of the Brown-Slipped pieces come from a single vessel (32), while a number of the Dark-Faced Burnished pieces appear to be "fryingpans."There is a very limited number of red-slipped pieces that do not have the volcanic mineral tempering of the Volcanic Red-Slipped class and they are included with the Brown-Slipped class.

DANIEL

I42

J. PULLEN

Medium Coarse: a broad class that includes the major portion of the Halieis prehistoric ceramics. Most often it has a gray core with yellow-brown or orange-brown or brown surfaces;it tends to be finer than coarse, though there is a very wide range in fineness of the fabric. Cooking Pot: a ware similar to the later EBA cooking pot class of ceramics.This ware is relatively hard-fired, medium coarse in composition, often with a reddish tinge, and sometimes with a compacted or crazed, wet-smoothed surface. In many ways this class is similar to the Medium Coarse class, but with a distinctive reddish tinge and wet-smoothed surface. Buff: a low-fired, relatively soft and crumbling ware with buff to orange surfaces and gray-brown cores; in terms of shapes it is similar to the Medium Coarse class. Compacted Red: a ware similar to the Medium Coarse class in terms of fabric, but with a distinctive red color and compacted, crackled surface. Some pieces are slipped and some are burnished. Pithos: a coarse fabric that includes many small irregularlimestone inclusions, very similar to the later EBA pithos fabric. There are a limited number of sherds that might fall into other categories, such as a few flat (closed?) bases in a hard, light-colored fabric that appears quite similar to the fabric used in later EBA jars, amphoras, and hydrias.Whether these sherds are indeed of the EH II period is difficult to tell, for while no other vessel part has been identified in this fabric, the bases are of the typical FN-EH I flat form. I have included them in the Medium Coarse class.

TABLE 2. FREQUENCY OF CERAMIC CLASSES REPRESENTED IN THE "PREHISTORIC" COLLECTION Class

Uncatalogued

VolcanicRed-Slipped and Burnished 4 VolcanicBlack-Burnished 0 VolcanicPlain 2 23 Brown-Slipped Dark-FacedBurnished 0 186 Medium Coarse 40 Cooking Pot Buff 20 3 Compacted Red Pithos 0 Subtotal Unclassified Post-FN/EHI TOTAL

278 0 6 284

Catalogued

Total

%

12 7 12 11 4 29 18 11 3 4

16 7 14 34 4 215 58 31 6 4

4.0 1.8 3.5 8.5 1.0 53.6 14.5 7.7 1.5 1.0

111 1 5 117

389 1 11 401

97.1 0.2 2.7 100%

THE

TABLE 3. DISTRIBUTION

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I43

OF CERAMIC CLASSES BY LOCATION Findspot

Class

1

2

3

4

5

6

8

9

10

Other

Total

Volcanic Brown-Slipped Dark-FacedBurnished Medium Coarse CookingPot Buff Compacted Red Pithos TOTAL*

12 10 0 33 15 2 0 0 72

1 3 0 35 12 14 1 0 66

3 0 1 24 6 6 2 0 42

0 7 0 14 3 1 1 0 26

1 0 1 2 1 1 0 0 6

13 4 1 32 5 0 0 1 56

0 1 0 13 1 0 0 0 15

1 1 0 14 3 0 0 0 19

1 0 0 4 2 2 0 0 9

5 8 1 44 10 5 2 3 78

37 34 4 215 58 31 6 4 389

*The totals for each findspot do not alwaysagreewith the totals in Table 1, as that table includes materialof periodslater than EH I. Findspot 7 is excludedhere becausethe sherdsfound therewere not identified as prehistoric.

17. Becauseof the small counts (< 5) in a numberof cells in the distributiontable (Table3), a Chisquaretest is inappropriate,as it providesonly an estimate of the true probabilityof the independenceof the variablesand is easily skewed by small counts.The Fisher'sExact-T test calculatesan exact probabilityvalue for independenceof the variables,based on the differencesbetween the data observedand the data expected (the Chi-squaretest is also based on the differencesbetween the data observed and the data expected).For purposesof the analysis,the three Volcanicclasses were lumped together.The nine principalcontextswith prehistoric potterywere used for the location variable.Some of the samplesare small, such as those for the Dark-Faced Burnished,Buff, Compacted Red, and Pithos classes,and this may have skewed the resultingcalculations.I would like to thank Glen Laird of the StatisticalConsulting Center at Florida State Universityfor conductingthe analyses. 18. The Dark-FacedBurnishedclass was eliminatedbecauseof the small sample size.

The frequenciesof classesamong the cataloguedand uncataloguedpieces are given in Table 2. Among the catalogued items presented here, the Medium Coarse class is underrepresented,while the three Volcanic classes and the Brown-Slipped and Dark-Faced Burnished classes are overrepresented. The uncatalogued items, however,are worn or very fragmentary,consist of plain body sherds, or otherwise duplicate items in the catalogue. DISTRIBUTION

OF CLASSES

Although the prehistoric ceramics were found in contexts with later pottery, and thus mixed, the distribution of the various classes is not random. Examination of the individual contexts and the distribution of the various classes shows a few anomalies (Table 3). The Medium Coarse class seems to be proportionately distributed among the different contexts. The Volcanic classes areoverrepresentedin findspots 1 and 6, but underrepresented in findspots 2 and 4. The Brown-Slipped class is overrepresented in findspots 1 and 4, but underrepresented in findspot 2 and absent from findspot 3. The Cooking Pot class is overrepresented in findspot 1 and underrepresentedin findspot 6, while the Buff class is overrepresentedin findspot 2. A Fisher's Exact-T test was applied to the distribution of the ceramic classes in the nine areasthat produced the majorityof prehistoric ceramics in order to test whether these observations are supported statistically.17In the first analysis, the Fisher's Exact-T test was conducted on the distribution of all the Volcanic (grouped), Brown-Slipped, and Dark-Faced Burnished classes.The results were significant, with a p-value of 0.0012, indicating a nonrandom distribution of those classes. In the second analysis, the Fisher's Exact-T test was conducted only on the Volcanic (grouped) and Brown-Slipped classes18in order to ascertain whether there was any difference in their distribution. These results were also significant, with a p-value of 0.0024. This indicates that the distribution of these two classes accounts for most of the nonrandomness of the overall distribution and

DANIEL

I44

J. PULLEN

that the distribution of the Dark-Faced Burnished class was of little significance. A third Fisher's Exact-T test was conducted on the Volcanic (grouped) classes and the Brown-Slipped class for findspots 1, 2, 5, 6, 8, and 9; a p-value of 0.6887 was obtained, indicating little significance to the distribution of these two classes among the six locations in question. These tests suggest the possibility that there may be some functional or chronological distinction among the contexts on the acropolis.19 There seems to be clearevidence that the two groups of ceramicclasses (Volcanic and Nonvolcanic) are mutually exclusive in findspot 3 (the area of the Mess Building) and findspot 4 (the pit to the north of Building A) and not proportionally distributed in findspots 1 and 2, areas near the altar.But it must be remembered that the number of sherds in all contexts is small, and the expected number of sherds of any one class is often not very large.We also lack the data to determine the original frequency of the prehistoric ceramics in relationship to the total number of ceramics from any one context. From findspot 4, only 26 prehistoric sherds were identified, and only 2.41 sherds of the Volcanic classes would be expected given that the Volcanic classes constitute 9.25% of the total ceramics (only 2.21 sherds of the Brown-Slipped class, at 8.5%of the total, would be expected). While we can interpret the results of the statistical tests to mean that the Volcanic classes and the Brown-Slipped class are distributed in findspots 3 and 4 in some statistically significant, nonrandom manner, the archaeologicalsignificance of these results is more difficult to determine. Most significant to my mind is the overall random distribution of most of the ceramicmaterialthroughoutthe variouscontexts on the acropolis. A Chi-square test of the independence of the distribution of the Medium Coarse class among the findspots compared to the distribution of all the other classes combined yielded a low probability of p = 0.1037 (X2= 11.9061, 7 degrees of freedom).20This would indicate that the Medium Coarse class, as well as the remaining classes grouped together, is relatively uniformly distributed among the findspots. The two most likely interpretations of the overall distribution of the ceramic classes (excluding the anomalous contexts of findspots 1-4) are that the various classes are contemporaneous with one another or that the materialhas been so thoroughly mixed that no possible functional or chronological distinctions can be drawn on the basis of location. Our current poor understanding of the FN and EH I periods and the transition between them, including their respective ceramic assemblages, offers little illumination. VOLCANIC-TEMPERED

CLASSES

The volcanic-tempered classes have a distinctive fabric characterized by inclusions of black and/or gold "mica,"in reality platy minerals typical of the volcanic stone known as andesite. As noted above, Aigina is a main source of andesite, one of the most important materialsfor grinding stones in the Aegean.21 Volcanic-mineral tempered pottery was found by the Southern Argolid Survey at a number of sites in the Hermionid region of

19. There remainsthe possibility that the nonrandomdistributionis due to some post-deposition process, whether by activitiesdating to the Archaic and Classicalperiods or those taking place in the millenniabetween the originaldeposition in the FN/EH I period and the Archaicperiod. 20. Findspot 5 was eliminated because its samplewas very small;this eliminationlessened the chance of skewing the results. 21. Runnels 1981.

THE PREHISTORIC

22. Pullen 1995, pp. 10-12. 23. The andesite-derivedminerals that appearas black and gold "mica" occurin the pottery of severalperiods from FranchthiCave (Franchthi ceramicphases 1-2 and 4, i.e., Early, Middle, and later Late Neolithic, but not the Final Neolithic [FCP 5]; Franchthi8, pp. 111-1 12, 127-130; Franchthi10, pp. 38-40). Likewise, these mineralsappearin so-called Aiginetan waresfrom the northeastern Peloponnesefrom the MH period;cf. Nordquist 1987, p. 49; 1995, pp. 46, 50-51; Zerner 1978, pp. 156-158. The overallappearanceof the fabricin question,however,as well as the shapes and the surfacetreatments,represented in both the Halieis and EH I material from the SouthernArgolid Survey,is quite distinct from that of the Franchthiceramics. 24. Severalsherdswere tested for the presenceof limestone tempering using a dilute solution of hydrochloric acid,but all resultswere negative.In the SouthernArgolid Surveypottery, the presenceof volcanic mineralsand lime is usuallymutuallyexclusive. While ninety-two sherdswere found to have lime or volcanic mineraltemper, only four had both (Pullen 1995, p. 11, table 1.1). 25. Cf. Pullen 1995, nos. 75 and 8496, respectively. 26. Cf. Pullen 1995, nos. 133-148. 27. Cf. Pullen 1995, nos. 189-191, 201-202.

REMAINS OF THE ACROPOLIS AT HALIEIS

I45

the Argolid, where Haijeis is located.22The survey pottery with volcanicmineral tempering particles has been typologically identified as EH I, although the ceramics of other prehistoric periods can also have the volcanic minerals present.23 At Halieis the volcanic-mineral tempered fabric appearsin three distinct varieties of surface treatments: red-slipped and burnished, black(slipped? and) burnished, and plain. Given the fragmentary nature of the Halicis material and the few recognizable shapes, it is difficult to be certain, but there is apparentlysome degree of overlap in shapes among the red-slipped and unslipped varieties.The black-burnished variety,however, seems different in its treatment and shapes, and most vessels of this class have thin walls and well-finished surfaces. The volcanic fabric appears in similar quantities in both evenly fired and unevenly fired biscuits, but neither one is particularly"hard"or "high"fired. In both instances the color of the core appears to be in the black to gray (5YR 2.5-5/1) or reddish brown (5YR 3/3, 4/2-4, 5/4) range. Surface colors of the unevenly fired examples tend to be lighter in color than the cores, often red to reddish brown (2.5YR 5-6/6, 5YR 3/3-4, 4/4-6, 5/6, 6/4, 7.5YR 6/4). None of the pieces could be said to be fine nor particularly coarse, though the pieces often have a gritty feel from the presence of many tiny quartz (?) inclusions.24 Surface treatments include the application of a red slip followed by burnishing, burnishing without a slip, and leaving the surfacesplain. Most pieces are well smoothed on both the exterior and the interior surfaces. Given the poor preservation of the red slip in some cases, it may be that more examples of the Plain class should be placed into the Red-Slipped and Burnished category.The slip can approachbright red in color and the burnishing sometimes produces a very lustrous surface (e.g., 10). VOLCANIC

RED-SLIPPED

AND BURNISHED

CLASS

(1-12)

The Volcanic Red-Slipped and Burnished class is more numerous than the Volcanic Black-Burnished class, but there are few recognizable shapes (Fig. 9). The identifiable open shapes in the red-slipped variety are a thick simple hemispherical bowl (1); a shallow straight-sided bowl (2); another straight-sided but deeper bowl (3); and an incurving deep bowl (4; see also 14 in the Volcanic Black-Burnished class for a similar bowl). They are similar to EH I bowls from the Southern Argolid Survey.25One jar rim, 5, has been recognized: it has a barely flaring rim.26Two handles probably come from closed vessels: 6 is a U-shaped vertical handle, rather narrow, while 7 is a broad vertical band, decorated only by the red slip and burnish. One base, 8, is slightly hollow; it is a typical shape for EH I closed shapes.27 Other examples of this class are represented by body sherds (9-10, 12). 15, included with the Black-Burnished class below, may actually be a Red-Slipped and Burnished piece that has darkened due to firing. There are a limited number of pieces found at Halieis that are red-slipped and burnished but are not in the volcanic fabric (e.g., 37 and 41); these are discussed below in the section on the Brown-Slipped class.

DANIEL

I46

J. PULLEN

\~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-

'I

c ~~~~~~5

4

p01 1

Bowl, rim

Fig. 9

P285, findspot 6. Diam. indet., but probably ca. 0.30 m. Medium even gray (5YR 4/1); many quartz inclusions; little volcanic temper. Rim thins to flat horizontal lip. Red slip and burnished interior and lip; exterior rough. FN-EH I 2

Shallow bowl, rim

Fig. 9

P098, findspot 1. Diam. ca. 0.32-0.33 m. Medium coarse uneven brown to gray-brown (5YR 3/4 to 5YR 3/2); very gritty; many quartz inclusions; volcanic black, no gold(?) inclusions. Surfaces worn; originally slipped. Shape similar to 11. EH I 3

Vessel, body

P294, findspot 3. Medium even gray-brown (5YR 4/6); volcanic black and gold inclusions. Red slip exterior, burnished. EH I 4

Bowl, rim

5cm

#O

Fig. 9

P128 + 130, findspot 1. Diam. ca. 0.19 m? Medium even gray (5YR

3/1); gritty; volcanic gold and black inclusions. Slightly incurving rim, flat lip. Red slip traces near rim where smoother; burnished. FN-EH I 5

Jar(?),rim

Fig. 9

P260, findspot 1. Diam. ca. 0.16 m. Medium uneven gray-brown to orange-brown (5YR 5/4 core to 2.5YR 6/6 surface);lots of volcanic black and gold inclusions. Flaring rim. Red slip and burnished exterior and interior. EH I 6

Vessel, handle

Fig. 9

P003, findspot 1. Medium even brown (5YR 3/3 to 4/6 at surface); volcanic black and gold inclusions. Vertical, U-shaped. Red slip and burnished exterior. EH I 7

Vessel, handle

Fig. 9

P103. Medium uneven graybrown to red-brown (5YR 4/4 to 5YR 6/4 surface);volcanic black and gold inclusions. Wide vertical ribbon, raised margins, slightly U-shaped. Red slip and burnished exterior. EH I

Figure 9. Volcanic Red-Slipped and Burnished class

THE

8

PREHISTORIC

Vessel, base

REMAINS

OF THE

Fig. 9

Pool, findspot 1. Diam. 0.06 m. Medium uneven gray to brown (5YR 4/1 to 5YR 6/4); lots of tiny and small quartz inclusions; volcanic black and gold inclusions. Slightly hollow. Red slip and burnished exterior. EH I 9

Vessel, body

P005, findspot 1. Medium even gray-brown (5YR 4/4); quartz inclusions; volcanic black and gold inclusions. Red slip and burnished exterior. Same vessel as 8?

EH I 10 Vessel, body P257, findspot 1. Th. (wall) 0.0058 m. Medium even red-brown

VOLCANIC

28. Potteryfrom the TrojanI period, particularlyfromYortanand the Sardis region, comes to mind. 29. Aspis:Touchais 1980, pp. 13, 14, fig. 4, and 15, fig. 5; Phlius:Biers 1969, p. 452, fig. 3, no. 37; pl. 115, no. 37, with diameterof ca. 0.19 (afterPhelps 1975, fig. 52, no. 6); Kephala:KeosI, pl. 31V, called by Coleman a 'jar."

BLACK-BURNISHED

ACROPOLIS

AT HALIEIS

I47

(5-2.5YR 5/4); volcanic black and gold inclusions. Red slip and burnished interior and exterior. EH I 11 Bowl or plate, rim

Fig. 9

P274, findspot 3. Diam. indet. Medium uneven gray to yellow-brown; tiny and small quartz inclusions; volcanic black and gold inclusions. Orientation not certain. Red slip and burnished. Shape similar to 2. EH I 12 Vessel, body P308, findspot 3. Medium uneven gray-brown to brown (5YR 4/1 to 5YR 4/4); lots of volcanic black and gold inclusions. Thick walled, perhaps jar neck(?) (probably not, as diameter too large at ca. 0.20 m). Red mottled to black slip exterior and interior, burnished. Secondary burning? EH I CLASS

(13-19)

The VolcanicBlack-Burnishedclassis of particularinterestfor its distinctive appearance(Fig. 10). Some pieces appearto be burntor dark-fired versionsof the Red-Slippedand Burnishedexamplesin that they have a slip and areburnished(e.g., 15, 17, 18, 19). One of the two relativelywell preservedpieces,13, the rim of a thin-walleddeepbowl,apparentlylacks a slip,but it is well burnished,with the individualverticalstrokessometimesvisible.It bearsa strikingsimilarityin its surfaceappearance to blackburnishedEB I potteryfromelsewherein theAegeanandwesternAnatolia, even in the mannerin whichthe grittyfabricappearswhen the burnished In shape,however,13 does not seem to be surfacelayerhas worn away.28 out of placeamongotherFN pieces,althoughthe slightlythickenedarea below the lip is unusual.In addition,thereis no handleor lug preserved, althoughthe extantportionrepresentsonly 10%of the originalcircumference of the vessel.Another deep bowl,but with incurvingrim, 14, lacks both a slipandburnishing.It resembles4 of the VolcanicRed-Slippedand Burnishedclass,fromthe sametrench(but differentbasket);perhapsit is merelya burntexampleof the latterclass.15 is a thick-walledbowlwith a carinatedrim,an unusualshape,andis slippedandburnishedlike the red versions.It seemsto be quitesimilarto S-profileor carinatedbowlsof FN date found,for example,at Argos;to bowlswith verticalridgesfound at Phlius; and to a bowl with pierced lug at the rim from Kephala, Kea.29 It

shouldbe noted,however,that 15 preservesonly 15%of the originalcircumference,so it mighthavehad sometype of plasticappliquethatis not preserved.None of the parallelscitedareblack,but the Phliusexamplehas a red-burnishedsurface.Bodysherd16 is probablyfromthe samevesselas 15. Otherthan 13 and15, only 17 definitelycamefroman openvessel;18

DANIEL

I48

J. PULLEN

Figure10. VolcanicBlack-Burnished class

13

0 I

5cm

I 1

14

1 I I_

15 and 19 are slipped and burnished on the exteriorbut not the interior.Complicating the picture is 42. It looks like 18 and 19 but seems to lack volcanic inclusions and is classified with the Brown-Slipped class below. 43, a deep bowl with a thin wall like 13, has a black slip and is burnished; it does not, however, have the volcanic fabric, and so it is included with the Dark-Faced Burnished class. 13 Deep bowl, rim

Fig. 10

HP210 + P240. Diam. 0.22 m. Medium even black (5YR 2.5/1); many tiny quartz inclusions; volcanic gold inclusions. Vertical rim. Burnished (possible pattern of vertical strokes). FN? 14 Bowl, rim

Fig. 10

P094, findspot 1. Diam. 0.18 m. Medium even black (5YR 3/1); volcanic black, but no gold, inclusions. Slightly incurving rim, flat lip beveled to interior. FN 15 Carinated bowl, rim

Fig. 10

P429, findspot 6. Diam. 0.18 m. Medium even black (5YR 2.5/1); volcanic gold and black inclusions. Rim curling out slightly, flat horizontal lip. Black (burnt?) slip; burnished; secondary burning. Same vessel as 16. FN

16 Vessel, body P430, findspot 6. Medium even black (5YR 2.5/1). Black-burnished, worn(?) especially apparent on exterior. Same vessel as 15. FN 17 Vessel, body P417. Medium even gray-brown (5YR 3/3 exterior to 4/4 interior); large volcanic gold and black inclusions. Black exterior, dark redbrown interior slip, burnished heavily. FN 18 Vessel, body P267, findspot 6. Medium uneven brown to black (5YR 4/4 interior to 5YR 3/1 exterior); many tiny quartz inclusions; some volcanic black, few gold inclusions; very gritty. Black slip and burnished exterior. FN?

THE

PREHISTORIC

REMAINS

OF THE

19 Vessel,body

PLAIN

CLASS

AT HALIEIS

149

Blackslip exterior(?),burnished. Sameas 18 andP269? Secondary burning? FN?

P270, findspot6. Mediumeven brown(5YR4/4); veryfew volcanic gold andblackinclusionsvisible.

VOLCANIC

ACROPOLIS

(20-31)

The unslipped variety of the Volcanic class includes both open and closed shapes, several similar to shapes of the Red-Slipped and Burnished class (Figs. 11-12). 20, a large straight-sided bowl, is like Red-Slipped and Burnished 1. Two smaller bowls have thick walls typical of vessels in the volcanic fabric: 21, with an outturned lip, and 22, shallow and slightly incurving.Three jars with flaring rims were identified: 23, 24, and 25. The large base 26 unfortunately does not preserve the bottom, but it probably was flat or at most very slightly hollow, like the Red-Slipped and Burnished base 8. Three vertical handles are flattened ovals: 27, 28, and 29. The wide, flat handle 30 (Figs. 11-12) has three columns of incised, short, curving lines, forming a herringbone pattern. One unusual plain vessel, 31, a thick-walled straight-sided bowl, has a volcanic fabric, but both the interior and exterior surfaces are burnished, giving it the appearanceof one of the Dark-Faced Burnished pieces. 20

Bowl, rim

Fig. 12

22

P268, findspot 6. Medium uneven gray to brown (5YR 4/2 to 2.5YR 6/6); volcanic black, no gold inclusions. Rim slightly thickened to exterior. Slightly reddish, self-slip on interior(?);interior wet-smoothed. FN

Bowl, rim

Fig. 12

P195. Diam. ca. 0.26 m. Medium even yellow brown (5YR 4/2 to 2.5YR 6/4); 1 or 2 black volcanic, no gold, inclusions. Splayed rim, round lip. Burnished? FN 23 Jar,rim

21

Bowl(?), rim

Fig. 12

P412, findspot 6. Two joining pieces. Medium uneven gray core to brown (5YR 6/1 to 7.5YR 6/4); volcanic gold and black inclusions. Flaring rim, flat lip beveled to exterior. Smoothed. Similar to 24 and 25 (but not same). EH I?

P423, findspot 10. Diam. ca. 0.20 m. Medium uneven gray-brown to brown (5YR 4/2 to 5YR 4/6); volcanic gold inclusions. Slightly thickened and rolled to exterior, faceted lip. Interior and rim smoothed, exterior rough. FN-EH I

Figure11. Ceramicswith incised decoration

Fig. 12

74 75

30

DANIEL

ISO

27 -

J. PULLEN

02

A_

2

__

A

A~~~~~~22

27

/20

2i

5 cm

3

30

1I 1 1

J5

Figure12. VolcanicPlain class 24 Jar,rim

Fig. 12

P413, findspot 6. Three pieces, Diam. 0.25-0.26 m. Medium even gray (5YR 5/1); gritty; volcanic black and gold inclusions. Flaring rim, round lip. Unslipped, wet-smoothed surfaces. Similar to 23 and 25 (but not same). EH I? 25 Jar,rim

Fig. 12

P414, findspot 6. Diam. indet. Medium uneven gray-brown to brown (5YR 5/1 to 5YR 5/3); volcanic temper. Flaring rim, pointed lip; orientation not certain. Self-slip? Similar to 23 and 24 (but not same). EH I? 26

Vessel,base

small quartz inclusions; volcanic black, no gold, inclusions. Flat(?) large base. Smoothed surfaces. FN 27

Vessel, handle

Fig. 12

P048, findspot 2. Medium even gray (5YR 3/3); many tiny quartz inclusions; volcanic black and few gold(?) inclusions. Vertical, flat. Unslipped. FN 28

Vessel, handle

Fig. 12

P215, findspot 6. Medium even brown (2.5YR 6/4); many tiny quartz inclusions; volcanic tiny black, no gold, inclusions. Vertical, flat. EH I

Fig. 12

P004, findspot 1. Diam. ca. 0.14 m? Medium uneven gray to orange (5YR 5/4 to 2.5YR 5/6); many tiny-

29

Vessel, handle

Fig. 12

P220, findspot 6. Medium even gray-brown (5YR 4/6 [wall], 5YR

THE

PREHISTORIC

REMAINS

OF THE

2.5/1 [handle]);manytiny quartz inclusions;volcanicblack,no gold, inclusions.Vertical,flat.Burnished. EH I 30 Vessel,handle

Figs. 11-12

HP519, findspot6. Medium coarsemottledyellow-brownto redbrown(7.5YR5/4 to 2.5YR 5/6 mottled);verygritty;volcanicblack andgold inclusions.Vertical,flat

ASSOCIATIONS

OF THE

ACROPOLIS

AT HALIEIS

I5I

wide band.Three columnsof herringboneincision. EH I 31 Bowl or basin,rim Fig. 12 P159, findspot9. Diam. indet. Mediumevengray(5YR 4/1); tinysmallquartzinclusions;veryfew gold volcanicinclusions.Thickenedrim; slightgrooveon interior;flatlip. Burnishedexteriorandinterior.

VOLCANIC

CLASSES

The Volcanic Red-Slipped and Burnished and Volcanic Plain classes discussed above can be dated to the EH I period; whether these classes are also to be found in the FN period is unclear.The distinctive fabric, as well as the surfacetreatment of red slip and burnish, was found at all EH I sites discovered in the Southern Argolid Survey of the Hermionid.30A major problem faced by the Survey was the considerable overlap of FN and EH I components at many sites: all but two sites with FN material (often in small quantities) also had EH I sherds. Nevertheless, we were able to differentiate between the two periods, in large part because of the distinctive volcanic-tempered fabric, shapes, and surface treatments of the Volcanic classes, which belong to the EH I period, and because several sites had pottery with volcanic-tempered fabric but no FN material.The volcanictempered fabric is found along the Saronic Gulf coast (e.g., Methana).31 Occasionally this ware appears outside the Hermionid in the northern Argolid, but examples are rare.In the extensive EH I phases at Tsoungiza, as well as in the Nemea Valley Survey area, no more than a dozen examples were found (personal observation); there are a few pieces of the fabric identified at Lerna, but EH I is poorly representedthere.32No mention of the fabric was made by Dousougli in her study of other sites near the Argive Plain.33 The Volcanic Black-Burnished class is problematic in that it was scarcely represented in the Southern Argolid Survey; only one piece from the FN period could be identified as Black-Burnished, a handle.34Unfortunately, the quantity of the Volcanic Black-Burnished class at Halieis is so small that it could not be tested statistically against the Volcanic Red-Slipped and Burnished class to determine whether their distributions areindeed different. Its closest parallelsarevessels from FN Kephala, FN-EH I Phlius, and the East Aegean FN-EH I. The Volcanic BlackBurnished class, then, perhaps should be dated earlier than the Volcanic Red-Slipped and Burnished and the VolcanicPlain classes,which are dated to the EH I period. 30. Pullen 1995. 31. Mee andTaylor 1997. 32. Wiencke 1989, p. 496, note 1. 33. Dousougli 1987. 34. Pullen 1995, p. 9, no. 42.

NONVOLCANIC-TEMPERED

CLASSES

The remaining classes-Brown-Slipped, Dark-Faced Burnished, Medium Coarse, Cooking Pot, Buff, Compacted Red, and Pithos-are character-

DANIEL

I52

J. PULLEN

ized by fabrics that lack the distinctive volcanic mineral tempering particles typical of the volcanic fabric. Each of these classes is discussed separately. BROWN-SLIPPED

CLASS

(32-42)

The fabrics of the Brown-Slipped and the Dark-Faced Burnished classes arevery similar.They are usually gray to gray-brown (5YR 3-4/2-4), have tiny quartz inclusions, and are usually not gritty in feel. Both varieties are burnished.When a slip is present in the Brown-Slipped class, it is usually brown in color, though red-brown (as in 36 and 41), red (37), and black slips (42) also occur; some of the Dark-Faced Burnished pieces have what appears to be a black slip, though this may be simply a result of firing. Those pieces without a slip are generally darkeron the surface. Many of the Brown-Slipped pieces appearto belong to the same vessel: 32 (Figs. 13-14). Certainly 33 and 34 have the same rough gray interior surface as the pieces of 32 and come from the same context, the pit in bedrock to the north of Building A. The precise shape of 32 is not clear, but it was a closed vessel. The large pieces of 32 seem to have a diameter of around 0.23 m (if the design was arranged in a horizontal band). If the tiny fragment of an incised band handle, 35, is part of this vessel (it was found to the northeast of the altar area, along with 36 and 37, more than 20 m from Building A), we might think of an askos or small jar with a wide vertical handle decorated with stacked continuous zigzag around the belly and grouped diagonals on the handle. All the identified Brown-

Figure13. Brown-Slippedclass

_ 32

33

32

39

33

I 38

36

THE

PREHISTORIC

REMAINS

OF

THE

ACROPOLIS

AT

HALIEIS

I53

32

/

'

=3-

,X

33

35

34 36

5cm

0 I

I

I

I _

137/ 41

39 38

1

40

Figure14. Brown-Slippedclass

35. Asea: Holmberg 1944, p. 83, fig. 84:a-b;Tsoungiza:Pullen, in preparation.

Slipped pieces have shallow curving grooves, apparently forming six or seven nested chevrons or stacked zigzags. Some pieces that do not appear to go with 32, such as 38 (found to the north of the altararea) and 36, have grooves that curve more and are spaced further apart. 39, while it has a similar fabric and decoration to 32, lacks the brown slip and so perhaps is also from a different vessel (it was found in the western part of the acropolis, far from the other pieces of Brown-Slipped ware). Askoi and askoid cups have a long history in Aegean ceramics, from the Late Neolithic (the "scoops"of the LN-FN periods) through EH II and later.The proposed Halieis askos or askoid cup from the pit north of Building A is relatedto askoidcups of FN-EH I date from the Peloponnese, like those from Asea and Tsoungiza. An additional Brown-Slipped piece decorated in a different fashion is 40, perhaps a portion of a so-called frying pan. The preserved portion is

DANIEL

I54

J. PULLEN

very slightly concave. On the decorated surface are portions of an impressed concentric circle, a slightly curving line bordered by what appears to be a row of kerbschnitt or impressed triangles, and a second curved line, perhaps without any border. Frying pans or other similarly decorated flat objects seem to date no earlier than the EH I period, according to our present knowledge, though our understanding of the FN and EH I periods is still poor.36The presence of two additional fragments of frying pans in the Dark-Faced Burnished class (see below) strengthens the EH I dating of this part of the ceramic assemblage. Two red-slipped and burnished pieces, 37, the flaring rim of a deep bowl, and 41, a vertical band handle thin but slightly concave in section and with thickened margins, and 42, a black-slipped and burnished body sherd, do not have volcanic temper in their fabrics, nor are they gritty in feel. They are not similar in fabric to the Brown-Slipped pieces above, but as they have slipped and burnished surfaces,they are included here rather than in any of the unslipped nonvolcanic classes. 32 Askos orjar(?), body

Figs. 13-14

HP626, findspot 4. Medium even dark red-brown to dark graybrown (5YR 3/2-4/2); lots of tiny quartz inclusions; nonvolcanic. Exterior slipped(?) and burnished; interior very rough, pitted, dark gray. Grooved nested chevrons. Additional pieces 33, perhaps 34, 35. FN-EH I 33 Vessel, body

Figs. 13-14

P134 + 135, findspot 4. Medium even gray-brown (5YR 4/2). Slip and burnished. Grooves. Same vessel as 32. 34 Vessel, body

Fig. 14

P262, findspot 4. Medium uneven gray-brown to red-brown (5YR 3/3-4/4, surface 5YR 6/4); tiny quartz, small irregularlimestone inclusions; nonvolcanic. Thickening to left to attachment(?), slightly flaring at one edge. Interior rough and gray.Two grooves. Same vessel as 32? FN 35

Handle

Fig. 14

P002, findspot 1. Medium even yellow-brown (5YR 5/6); tiny quartz inclusions; nonvolcanic. Broad

verticalstraphandle.Slip exterior and in grooves;interiorslightly rough.Multiplegrooves,groupedand opposedon exterior(hatched triangles?). FN 36 Vessel,body

Figs. 13-14

P077, findspot1. Medium unevengray-brownto brown(5YR 4/3 to 7.5YR3/5); tiny quartz inclusions;nonvolcanic.Red-brown slip exterior(includingin grooves). Three impressedgrooves.Probably not sameas 32. FN-EH I 37 Bowl,rim

Fig. 14

P009, findspot1. Diam. 0.15 m. Mediumunevengray-brownto brown(5YR4/6 exteriorto 5YR 3/3 interior);not gritty;nonvolcanic. Slightlyflaringrim.Red slip and burnished. FN? 38 Vessel,body

Figs. 13-14

P034, findspot2. Medium unevengrayto orange-brown(5YR 4/4 to 5YR 6/4); tiny quartzinclusions;nonvolcanic.Burnished, interiorsmoothedlike exterior.Two or moregrooves. FN?

36. Coleman (1985, p. 201) would place the mainlandexampleslate in EH I, overlappingwith the EC II period.

THE

PREHISTORIC

REMAINS

39 Vessel, body

OF THE

Figs. 13-14

41

P238, findspot 6. Medium uneven gray-brown to red-brown (5YR 3/2 to 5YR 3/4); lots of tiny quartz inclusions; nonvolcanic. Grooves. No evidence for brown slip (worn surfaces), but fabric and decoration are similar to those of 32-38. FN? 40

Frying pan(?), body

Fig. 14

P023, findspot 2. Medium even black (5YR 3/1, but red surfaces); quartz and limestone(?) inclusions; nonvolcanic. Slightly concave surface. Burnishing a little in evidence, but very smooth on both sides. Stamped circles, impressed grooves, and kerbschnitt. EH I?

DARK-FACED

37. See Coleman 1985. 38. At Tsoungiza examplesof frying pans are found with a large central impressedor incised star,arms radiatingout nearlyto the edge and with stampedspiralsin between (20112-2 and 2014-2-1). The star and spiral design, found also at Asea (Holmberg 1944, p. 86, fig. 87:a = Coleman 1985, p. 214, no. 90), is unlike the stardesign found on Attic examples,where the star has shorterarms and is confined to the inner portion of the field of decoration, e.g., from Agios Kosmas (Mylonas 1959, fig. 146, no. 195 = Coleman 1985, no. 78; Mylonas 1959, fig. 148, no. 210 = Coleman 1985, no. 79, and Mylonas 1959, fig. 149, no. 227 = Coleman 1985, no. 80). Rather,as Coleman notes, the large staris more like the design on Cycladicexamples, although these are formed solely by outline and not by multiple strokesas in the Tsoungiza and Asea examples.

BURNISHED

ACROPOLIS

AT HALIEIS

Vessel, handle

I55

Fig. 14

P213, findspot 6. Medium uneven brown to orange (5YR 5/4 to 2.5YR 5/6); many tiny and small limestone, irregularquartz(?) inclusions; nonvolcanic. Vertical(?), thick margins (U-shaped). Redbrown slip. Not true Brown-Slipped or Red-Slipped and Burnished fabric, but rather similar to those of EH II. FN-EH I 42

Vessel, body

P269, findspot 6. Medium even black-brown (5YR 3/3 interior to 5YR 2.5/1 exterior); many tiny quartz inclusions; nonvolcanic(?). Black(?) slip and burnished. Black version of Red-Slipped and Burnished, but nonvolcanic fabric;burnt exterior? FN-EH I

CLASS

(43-47)

As noted above, the fabric of the Dark-Faced Burnished class is very similar to that described for the Brown-Slipped class. Some Dark-Faced Burnished pieces lack a distinct slip, while others have a black or brown-black slip, though this may be merely a result of firing; the surfaces are generally highly burnished. Two small deep bowls, 43 with a straight rim and 44 with a slightly flaring rim, have burnished surfaces (Fig. 15). 43 has a black slip, similar to vessels in the Volcanic Black-Burnished class. The most unusual piece is 45, perhaps part of a frying pan. Given the small extant portion, it is difficult to reconstruct the form and decoration. A thick wall, probablyvertical, thins and flares toward the broken bottom; a groove emphasizes the flare.At the top, the wall bends to form what was most likely the flat upper surface.The remains of three incisions perpendicular to the edge are preserved. There is no indication of a handle or projecting flange typical of several varieties of frying pan.37The rather high wall is unusual, but the diameter of 0.18 m is quite in line with the size of frying pans. 46 is definitely part of the rim of a frying pan. A slightly convex disk 0.17 m in diameter has had added to it a low, thick wall, preserved only about 0.015 m in height; no flange is present. Both exterior and interior surfacesare burnished. On the exterior bottom are three grooves not quite perpendicularto the outer edge; they must be part of long grooves radiating from the center or perhaps forming the arms of a star pattern, as found on frying pans at Tsoungiza and Asea in EH I contexts.38Both 45 and 46 show that the edge of the disk is undecorated except for the three grooves most likely extending out from the center; similarly,these two lack decoration on the side wall. As noted above, frying pans have not been found in contexts earlier than EH I.

i56

DANIEL

_I

J. PULLEN

I 43

47

44

''l-

1

145

4

45

__?

1 1 I

47, with its slipped and burnished surface,is similar to Brown-Slipped and Dark-Faced Burnished pieces, but the shape, a cup with upswung handle, is problematic. The piece is handmade and burnt. The rectangular-sectioned handle is attached to the vessel wall, probably not too far below the rim, without a tenon piercing the wall. 43

Deep bowl(?), rim

Fig. 15

P426, findspot 10. Diam. 0.16 m. Medium even black (5YR 2.5/1); not gritty, nonvolcanic, but many tiny quartz inclusions. Vertical rim, round lip. Black slip and burnished. 44

Bowl, rim

Fig. 15

P316, findspot 3. Diam. 0.14 m. Medium uneven gray to red-brown (5YR 3/4 exterior to 5YR 3/2 interior); not gritty; nonvolcanic. Slightly flaring rim. Slip(?), burnished. Fabric similar to 43. FN? 45

Frying pan or pyxis, body

Fig. 15

P264, findspot 6. Diam. (top) 0.18 m. Medium even gray (5YR 3/1); some tiny, small quartz inclusions; nonvolcanic. Slip exterior, burnished. Three grooves perpendicular to edge. EBA?

46

Frying pan, body

Fig. 15

Findspot 5. Diam. 0.17 m. Medium uneven brown core to redbrown (5YR 5/3 to 2.5YR 4/8); tiny quartz(?)inclusions; nonvolcanic. Flat disc to which has been added a round low wall. Burnished exterior and interior. Grooves. EBA? 47

Cup, handle and body; class uncertain

Fig. 15

P461, findspot 8. Fine-medium even dark brown (5YR 4/2); handmade. High-swung handle above rim(?) (thinning part is just worn); wall thickens for upper attachment, most likely at rim. Burnt? Slip, burnished. Unusual handle form. FN? EH?

I

46

5cm I I5cm

Figure 15. Dark-FacedBurnlished class

THE

PREHISTORIC

MEDIUM

COARSE

OF THE

ACROPOLIS

AT HALIEIS

I57

(48-75)

CLASS

By farthe largestquantityof prehistoricpotteryfromHalieis canbe categorizedunderthe rubricMedium Coarse,primarilybecauseit does not fall into anyof the othercategories.Nevertheless,thereis a certainsimilarityamongthe vesselsincludedhere.By definitionthe fabricslackvolcanictemper;instead,limestoneandquartzseemto be the most common temperingagentsvisibleto the eye.The surfacecolorsclusterfairlytightly aroundlight red,red,andreddishbrown(2.5 to 5YR 4-6/4-8), while the cores(whenunevenlyfired,as is the casefor abouthalfthe pieces)tend to be reddishbrown,darkreddishbrown,or gray(e.g., 5YR 3/1-3, 4/3, 5/34), and not blackor darkgray as one might expect.Surfacetreatments vary:some kind of slip is apparenton roughlya thirdof the pieces;burnishing with or without a slip is presenton only about a fourth.Wetsmoothingis apparenton severalpieces,but withoutthe compactedsurface foundin the CookingPot classor Buff class,describedbelow. The mostcommonshapein MediumCoarseis a thick-walledspreading bowl,often fairlydeep (48-51), thoughnearlyvertical(52 and 53) or

Figure 16. Medium Coarse class: rims

K

REMAINS

_t8

1

_

49

53

~~~~~~~~~~~~54

U

* E

v

__

_

52

0

I

_

_

5cm

_

_

_ _

_

_

_

_

_5

J. PULLEN

DANIEL

I58

\\

Li6~~~

63

~~~~0

5 cm

0 I

62

~~6

I

5 666

incurving (54) walls occur (Fig. 16). Thinner-walled bowls, both shallow (55) and incurving (56), also occur; the latter has a "rolled rim" that is thought to begin in the later part of the FN period. 57 is from an open shape known as a "cheesepot,"usually dated to the FN period.39Cheesepots are also found in the Cooking Pot and Buff classes; see below. A common feature of Medium Coarse bowls is the application of ridges (51) or taenias (50, 52, 53, and 54) either at or just below the lip. Jars are not very common; only two ratherlarge-mouthed examples,58 and 59, areincluded here.These shapes find similaritiesin the FN assemblagesfrom the Southern Argolid Survey and elsewhere in the Peloponnese.40 Most of the bases recognized in Medium Coarse are of the standard flat, thick-bottomed variety that appears almost like a disk, whether on open vessels such as 60, 61, and 62, or closed vessels such as 63 (Fig. 17). Three of these flat bases have impressions of mats on their undersides:60, 64, and 65 (see below, Fig. 29). One base, 66, stands out for being a pedestal instead of the usual flat base. The foot is preserved to at least 0.03 m high; the interior of the bowl is slipped, indicating an open shape. One tubularlug, 67, was formed by vertically piercing a large lug; the piercing removed some of the vessel wall (Fig. 18). The shape this lug might have belonged to is unclear, but its similarity to the unpierced lug 68 suggests that it may have come from a bowl or other open shape. 68 is a large horizontal lug or ledge handle that angles down from the vessel wall. Along the preserved margin of the lug the edge has been flattened into facets, perhaps indicating that the lug would have been continued by a taenia band. The red-slipped and burnished interior shows that the vessel was open, probably some large bowl or basin. The bodies of Medium Coarse vessels could be decorated with ridges and taenias, as the rims are. 69 with a single ridge, 70 with two, and 71 with three preserved ridges show the common triangular-section ridge (Figs. 18-19). 72 (Fig. 20) has three rows of finger-impressed taenia bands, while 73, from a relativelylarge vessel, has two taenia bands angled to each other.

Figure17. MediumCoarseclass: bases

39. Pullen 1995, p. 8, no. 22; Renfrew 1972, p. 155. 40. SouthernArgolid Survey:Pullen 1995, pp. 8, 10, nos. 28-34; Peloponnese:Phelps 1975.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I59

Some of the most elaborate incised decoration from Halieis is found on two pieces of Medium Coarse. Both pieces, 74 and 75 (Figs. 11, 18), come from vessel walls and not handles. 75 seems to come from an open vessel, while 74 certainly seems to be from a closed vessel. From the curvatures we can reconstruct diameters of 0.10 and 0.13 m, respectively.74 is decorated with alternating rows of diagonal slashes that form an overall herringbone design. 75 has a design of two rows of diagonal slashes (in opposite directions) between a pair of incised lines. Toward one edge of the preservedportion are four lines, suggesting an end to the pattern.There is, however, no indication that this is a rim; perhaps the pattern was deliberately relegated to only a portion of the exterior as part of the design or because it would be obscured, as by a handle.

69

11 67

|

5 cm

0

4

x

70

68

-;1

75 Figure18. MediumCoarseclass: handlesanddecoratedbodies

86

Figure19. Ceramicswithapplied ridges10

_

70

71

9

J_

83

8

i6o

DANIEL

J. PULLEN

Figure20. Ceramicswith taenia decoration

78

110

j,

-

72

-7

88

0~~16_ 90

87 The majorityof the features of the Medium Coarse class are found in FN assemblages.Although the high frequency of flat to slightly hollow or concave bases, the presence of plain and impressed ridges, and the tubular lug are characteristic of the FN period, it must be stressed that some of these elements are features of other periods, including EH I. 48

Bowl,rim

Fig. 16

P152, findspot 4. Diam. 0.220.24 m? Medium uneven gray to redbrown (5YR 3/3 to 2.5YR 5/6); tiny quartz inclusions; nonvolcanic. Slightly thickened rim. Self-slip exterior(?),burnished exterior? FN 49

Bowl, rim

Bowl, rim

Fig. 16

P271, findspot6. Medium unevengray-brownto red-brown (5YR4/3 to 2.5YR 4/6); tiny limestone,smallquartz,small-large limestoneinclusions.Red surface, unslipped?Interiorsmoothed. Shallowridge. FN

Fig. 16

P301 + P290 + P312 + P326, findspot 3. Diam. 0.30 m. Medium even gray (1OYR5/2); tiny limestone, few small quartz inclusions; nonvolcanic; hard-fired; similar to EH II jar fabric. Slightly thickened rim, flat lip. FN 50

51 Bowl,rim

Fig. 16

P235, findspot 6. Diam. ca. 0.25 m. Medium even gray-brown (5YR 4/2); tiny quartz, tiny-small limestone inclusions; nonvolcanic. Piecrust rim. FN

52 Deep bowl,rim

Fig. 16

P133, findspot4. Diam. ca. 0.30 m?Mediumeven orange(2.5YR6/6) irregularsmallinclusions.Vertical rim,flat lip. Originallyslipped(?)on interiorandlip(?).Taenia;postfiring hole. FN 53 Bowl,rim

Fig. 16

P419. Diam. indet.Medium unevengray-browncoreto redorange(5YR4/6 to 2.5YR 5/8); tiny limestoneinclusions.Incurvedrim (or nearlyvertical?),slightlyflattened. Taeniabelowlip. FN-EH I

THE

54

PREHISTORIC

REMAINS

Deep bowl, rim

OF THE

Fig. 16

P251, findspot 2. Diam. indet. Medium uneven gray to orangebrown (5YR 4/3 to 2.5YR 6/6). Insloping(?), orientation not certain. Red-brown slip interior.Taenia. FN 55

Shallow bowl, rim

Fig. 16

P229, findspot 6. Diam. ca. 0.16 m? Medium uneven gray-brown to orange-brown interior surface (5YR 5/3 to 2.5YR 5/6); tiny quartz inclusions; nonvolcanic. Flat lip. Slip? FN-EH I 56

Hemispherical bowl(?), rim

Fig. 16

P445. Diam. 0.18 (interior)0.22 m (exterior). Medium even yellow-brown (7.5YR 7/4); spongy ware (vegetal temper); nonvolcanic. Slightly thickened rim, horizontal lip. Burnished, but no luster. FN or EH III? 57

Cheesepot, rim

Fig. 16

P407, findspot 6. Diam. 0.19 m. Medium even orange-brown (5YR 6/4); 1 very large inclusion (7 mm); tiny black and white quartz(?) inclusions; nonvolcanic; fabric gritty like Red-Slipped and Burnished class. Flaring rim, flat lip. FN? 59 Jar,rim

AT HALIEIS

i6i

60 Mat-impressed base Figs. 17, 29 P039, findspot 2. Diam. 0.11 m. Medium even gray-brown (5YR 5/3); tiny quartz inclusions; nonvolcanic. Flat. Mat impression: diagonal plaited, double weft over warp. FN 61

Vessel, base

Fig. 17

P047, findspot 2. Diam. 0.10 m. Medium uneven gray to brown (5YR 5/4 to 5YR 7/6); some tiny irregular small limestone(?) inclusions; nonvolcanic. Flat. FN 62

Vessel, base

Fig. 17

P299, findspot 3. Diam. 0.065 m. Medium uneven gray to orange surface (5YR 3/2 to 5YR 5/6); many tiny limestone inclusions; nonvolcanic; heavy, dense. Flat. Burnished (interior dark and looks like Volcanic classes, but no volcanic temper). FN

Fig. 16

Findspot 2. Diam. > 0.35 m? Coarse uneven gray-brown to redbrown surfaces (5YR 4/3 to 2.5YR 6/6); tiny sand, small limestone inclusions; nonvolcanic. Vertical rim. Three holes pierced (prefiring) from interior; exterior irregular,interior smoother. Worn; similar vessel to 99. FN 58 Jar,rim

ACROPOLIS

Fig. 16

P175, findspot 9. Diam. ca. 0.25 m. Coarse even gray-brown (5YR 3/3); many tiny quartz, limestone(?) inclusions; nonvolcanic. Flaring rim. Slip(?) exterior,burnished exterior. FN

63

Vessel, base

Fig. 17

P189. Diam. 0.09 m. Medium even brown-buff (5YR 6/4); many tiny limestone inclusions; resembles EH II jar fabric. Flat. Self(?)-slip. FN? EH I? 64

Mat-impressed base Figs. 17, 29

HP530. Diam. 0.12 m. Coarse uneven gray to orange (7.5YR N5/0 to 2.5YR 6/6); tiny, small, large limestone inclusions; nonvolcanic; hard-fired like EH I fruitstand. Flat. Mat impression: widely spaced warp over multiple weft, twined? FN-EH I 65 Mat-impressed base Figs. 17, 29 Findspot 10. Diam. ca. 0.140.15 m. Medium uneven gray to orange-brown (7.5YR 5/2 to 5YR 6/6); tiny and small irregularquartz, occasional large (3 mm) inclusions; nonvolcanic. Flat. Mat impression: twined, faint and worn, weft seems angled between each warp group. FN

DANIEL

I62

66

Pedestaled bowl, base

Fig. 17

P036, findspot 2. Diam. ca. 0.08 m (near foot). Medium even orangebrown (1OR 5/6); very gritty; many tiny quartz inclusions. Pedestal, slightly flaring toward foot. Slip interior bowl, wet-smoothed exterior. FN

71 Vessel, body

Vessel, handle

Fig. 18

P324, findspot 3. Medium uneven gray-brown to brown surface (5YR 3/2 to 5YR 6/4); tiny, some small quartz inclusions; nonvolcanic. Horizontal; similar to pierced lug (vertical piercing). FN

Fig. 19

P254, findspot 2. Encrusted. Medium uneven orange interior to gray-brown exterior (2.5YR 5/8 to 5YR 5/4). Red interior, self(?)-slip. At least three ridges. FN 72 Vessel, body

67

J. PULLEN

Fig. 20

P282, findspot 6. Medium uneven brown to red-orange (5YR 3/2 to 2.5YR 5/6); small-large (2 m) quartz, tiny limestone inclusions. Three taenias;wet-smoothing around taenias. FN 73 Vessel, body

68

Fig. 18

Large bowl, handle and body

P288, findspot 3. Medium even orange-brown (5YR 5/6); tiny, small, large quartz, limestone(?) inclusions; nonvolcanic. Lug handle; orientation not certain. Red slip and burnished interior. FN 69

Fig. 18

Vessel, body

P031, findspot 2. Medium even gray-brown (5YR 5/4); tiny quartz(?) inclusions. Red slip, burnished. Ridge. FN 70

Vessel, body

Figs. 18-19

P239, findspot 6. Medium uneven red-orange interior to brown exterior (2.5YR 4/8 to 5YR 5/6); many tiny, small white quartz inclusions; nonvolcanic; hard-fired. Orientation not certain. Red slip and burnished interior, exterior wetsmoothed. Ridges. FN

Th. (wall) 0.013 m. Medium even orange-brown (2.5YR 5/6); quartz(?)inclusions. Large vessel, section larger than 0.35 m diameter? Wet-smoothed interior.Two taenias, angled. FN 74 Vessel, body

Figs. 11, 18

P035, findspot 2. Diam. ca. 0.10 m? Medium uneven gray to orange interior (5YR 6/1 to 2.5YR 5/8); tiny quartz inclusions; nonvolcanic. Small vessel, ca. 0.10 m diameter. At least four columns of herringbone incision; perhaps not vertical but diagonal; exterior surface worn off FN-EH I 75 Vessel, body

Figs. 11, 18

HP504. Exterior Diam. 0.13 m? Medium uneven gray-brown interior to red-brown exterior (5YR 4/3 to 2.5YR 4/8); tiny and small irregular limestone(?) inclusions. Incised herringbone and grooves; interior compacted slightly.

EH I? COOKING

POT

CLASS

(76-93)

One distinctive group is the Cooking Pot class, so named because of its resemblance to the cooking pot fabrics of later EBA phases. Cooking pot fabrics are medium to coarse, more often unevenly fired than even, and have a reddish tinge to the fabric away from the core. The surfaces are

THE

41. Holmberg 1944, p. 55, fig. 57:b.

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I63

often wet-smoothed or even burnished, though never with a luster imparted but rather with a slight "compacted"appearance.Occasionally the compacting of the surfaceis very pronounced, creating a crazed or cracked surface (e.g., 88, 91, and 92). Slips are rare;indeed only 85, a handle, possibly had a red slip. "Self-slips,"technical slips formed by wet-smoothing the surface, do appear sometimes. The most common shapes in the Cooking Pot class arebowls andjars, as one might expect (Fig. 21). The bowls range from deep to hemispherical. 76, with a rim diameter of 0.12 m, has an irregular,slightly spreading wall; 77, just slightly larger with a rim diameter of 0.13 m, has a flaring rim. 78 (Fig. 20) and 79 are two hemispherical bowls with taenia decoration on the rim, just below the lip in the case of 79, lower down in the case of 78. A slightly larger bowl, 80 (Fig. 19), has a raised ridge instead of a taenia band below the rim. One example of a cheesepot, 81, was found in the cooking pot fabric.Two holes are preserved,both approximately0.015 m below the lip, as is often the case in this shape. Only one jar rim, 82, has been recognized: a flaring rim with lip thickened to the exterior. Other jars are attested by handles, several of which are decorated in some manner. 83 (Figs. 19, 21) is a vertical band handle, relatively flat in section, with two vertical ridges near the margins; the ridges are marked with very fine diagonal slashes. Another vertical band handle, 84, is more carefullymade, rectangularin section with raised edges. It has a single finger impression centered at the (upper?)attachment. 85 is a relatively simple handle, probablyvertical. The most elaborate piece of cooking pot is the pierced spool lug 86, probably from an open vessel, to judge by the burnished interior. A high lug projects from the exterior surface, between pairs of wide horizontal grooves with edges raised above the wall. The two sides and the center of the front of the lug are articulatedwith disks. The lug is pierced vertically, from top and bottom, forming a hole slightly smaller in diameter in the middle. The lug is reminiscent of simpler trumpet or spool lugs found in EH I contexts, for example at Asea.41 A number of cooking pot vessels were decorated with taenia bands and ridges, such as 87-89, which have single or multiple taenia bands, including some meeting at angles; 90 (Fig. 20) is unusual in that one of the three poorly made taenias has had clay smeared into the finger depressions, apparentlyto convert the taenia into a wide ridge. A further example of plastic decoration, 91 (Fig. 19), has four narrow parallel ridges preserved on the exterior of a closed vessel. These ridges may be diagonal in their orientation, if we use the comparativewall thickness for orientation of the preserved portion; one of the ridges ends simply, another makes a bend before ending. Two bases, one flat, 92, and one flat disk, 93, were recognized; the latter base has a mat impression preserved on the underside (see below, Fig. 29). For the most part, the shapes in the Cooking Pot class are nearly the same as in the Medium Coarse class, though the thick-walled spreading bowls of the Medium Coarse class are not as frequent in the Cooking Pot class. The plastic and impressed decoration is similar in the two classes.

I64

DANIEL

0X18

J. PULLEN

177

76

=tX78

t

___

=

79

80

82

81

84

83

-~~~~~~A

/85

L 92

86

0 1

I

1

1

93

5cm

- I

Figure21. CookingPot class

THE

PREHISTORIC

REMAINS

76 Deep bowl, rim

OF THE

Fig. 21

P155, findspot 4. Diam. 0.12 m. Medium uneven gray interior to brown exterior (5YR 4/3 to 5YR 6/4); tiny quartz inclusions; nonvolcanic. Vertical rim, round lip. Burnished? FN 77

Bowl, rim

Figs. 20-21

P030, findspot 2. Diam. 0.22 m. Coarse uneven gray to brown (5YR 5/1 to 2.5YR 5/4); tiny, small, and large (2 mm) quartz inclusions; nonvolcanic. Splayed rim, flat lip. Taenia with slashes. FN 79 Hemispherical bowl, rim

Fig. 21

P032, findspot 2. Diam. indet. Coarse uneven brown to orange (5YR 3/2 to 2.5YR 4/6); many tiny, small, large quartz inclusions; nonvolcanic. Round lip. Self-slip. Taenia just below lip; compacted surfaces. FN 80 Bowl, rim

Figs. 19,21

P266. Diam. ca. 0.28 m? Medium even dark gray,red surface (5YR 4/1); tiny-small limestone, quartz, and other inclusions; nonvolcanic. Flaring rim. Burnished interior. Ridge. FN 81 Cheesepot, rim

AT HALIEIS

82 Jar,rim

I65

Fig. 21

P443. Two pieces. Diam. 0.14 m. Medium even gray (5YR 4/2); tiny to small nonquartz inclusions; nonvolcanic. Slightly flaring rim, lip rounded to exterior.Interior rim especially wet-smoothed. FN-EH?

Fig. 21

P084, findspot 1. Diam. 0.13 m. Medium uneven gray-brown to orange-brown (5YR 3/3 to 5YR 6/6); tiny, small, and occasional large quartz inclusions; nonvolcanic. Slightly flaring rim, round lip. Wet-smoothed surfaces. FN 78 Hemispherical bowl, rim

ACROPOLIS

83 Vessel, handle

Figs. 19, 21

Findspot 10. Medium uneven gray-brown to red-brown (5YR 5/2 to 2.5YR 5/6); tiny quartz inclusions; nonvolcanic. Vertical.Wet-smoothed, slightly compacted. Two vertical taenias, narrowwith diagonal slashes. FN 84 Vessel, handle

Fig. 21

P241, findspot 6. Medium even gray-brown to red surface (5YR 5/1 to lOR 4/4-5/6); tiny-small limestone, some quartz(?) inclusions; nonvolcanic. Vertical, flat-sectioned with raised margins; finger impression near attachment. FN-EH I 85 Vessel, handle

Fig. 21

P258, findspot 1. Medium even red-brown (2.5YR 4/4); many tiny, some small quartz inclusions. Vertical(?),two-ridged in section. Red surface (slip?), burnished. FN 86 Vessel, body and lug Figs. 19, 21 Pill, findspot 1. Medium uneven brown to red-brown (5YR 4/4 to 2.5YR 5/8); tiny quartz inclusions; nonvolcanic. Horizontal lug with finials, pierced vertically. Self-slipped interior(?), burnished or compacted interior.Taenia(?) or three ridges continue out from lug. FN

Fig. 21

P079, findspot 1. Diam. indet. Medium uneven gray-brown to redbrown (1OYR4/2 to 2.5YR 4/4); tiny quartz, small-large limestone inclusions. Nearly vertical rim. Interior compacted, very worn exterior.Two holes pierced (prefiring) from interior. FN

87 Vessel, body

Fig. 20

P033, findspot 2. Coarse uneven black interior to red-brown exterior (5YR 3/1 to 2.5YR 3/4); small, large (6 mm) limestone, quartz(?) inclusions; nonvolcanic. Taenia. FN

i66

DANIEL

88 Vessel, body

Fig. 20

P092, findspot 1. Medium even brown (5YR 4/3); many tiny, small, and large quartz inclusions; nonvolcanic. Taenia. FN

P456, findspot 8. Medium uneven gray-brown to red-brown (5YR 4/2 to 2.5YR 5/6); tiny limestone, tiny-small quartz inclusions; nonvolcanic. Interior compacted. Two taenias at angle. FN 90 Vessel, body

Fig. 20

P256, findspot 1. Coarse mottled red-brown to black (2.5YR 5/4); 1 pebble (5 mm). Three taenias, poorly made; one has added clay to fill up finger depressions. FN 91 Vessel, body

Figs. 19,21

P024, findspot 2. Medium uneven black to red-brown surfaces (5YR 2.5/1 to 2.5YR 4/6); tiny BUFF

CLASS

pyrite(?) small quartz/limestone inclusions; nonvolcanic. Orientation not certain. Four ridges, may be diagonal if orientation by wall thickness and curvatureis accurate. FN 92 Vessel, base

89 Vessel, body

J. PULLEN

Fig. 21

P180. Diam. 0.10 m. Medium uneven black-brown to red-brown surface (5YR 3/2 to 2.5YR 6/4 surface);tiny quartz inclusions, iron pyrite spalls(?);nonvolcanic. Flat; irregularunderside or very faint mat impression? Burnished? Interior reddish, compacted. FN 93 Mat-impressed base Figs. 21, 29 Diam. ca. 0.06 m. Coarse uneven brown to red-brown (5YR 3/3 to 2.5YR 4/6); many tiny, small, and large irregularlimestone inclusions; gritty. Flat. Mat impression: small area preserved and unclear, perhaps basketry?Wetsmoothed exterior like Cooking Pot ware. FN

(94-104)

Anotherdistinctivecategoryis that calledBuff.This groupis characterized by a relativelysoft, low-fired,and crumblyware,often with compactedsurfaces,such as 95, 97, 99, 100, and 102-104. The colorsof the fabrictend to be lighter-more orangeto orange-brown-than those of the fabricsof the MediumCoarseand CookingPot categories:2.5YR 56/6-8, 5YR 4, 6-7/6 (lightred,yellowishred,andreddishyellow).Identifiableinclusionsareoften quartz,thoughothermaterials(limestone?)are also apparent.A few pieces seem to have a slip, probablya technicalor self-slipfromwet-smoothingthe surface;only 95 and 102 might have a deliberatelyslippedsurface,reddishin color. The shapesfound in the Buff class are thick-walledbowls for the most part,e.g., 94-99 (Fig. 22), though a few examplesof closedshapes, e.g., 100 (Fig. 19), also occur.Notable shapesinclude two examplesof cheesepots(95, 99), a pedestaledbowl(?)(102), and an unusualT-shaped horizontalhandleattachedbelow the rim of a bowl (94). Appliedridges andtaeniasarefoundon a numberof the closedvessels,such as 100, 101, andnumerousuncataloguedexamples.A few of the openvesselsalsohave appliedridgesand taenias,e.g., 103 and 104. Baseslike 96 and 98, as in the othercategories,areflat andthick. In terms of shapes (thick-walledbowls, cheesepots,flat bases, and pedestals)andapplieddecoration(ridgesandtaenias),the Buffclassseems

REMAINS

PREHISTORIC

THE

OF THE

AT HALIEIS

ACROPOLIS

Figure22. Buffclass

I67

_

195

L) _

l

_

96

_

98

l,) \

/_ X

1 }99

-

_

r

|

~~101

0

I

~~~~~~~102

5 cm I I I I I

to be very similar to the Medium Coarse class; perhaps it represents a poorly fired variant of the Medium Coarse class, with no functional or chronological distinction. 94

Bowl, handle and rim

Fig. 22

P025, findspot 2. Diam. ca. 0.25 m? Medium uneven gray-brown to orange-brown (5YR 3/2 to 5YR 6/6); small to large (up to 2 mm) quartz inclusions; nonvolcanic. T-lug, projecting. FN

97

Cheesepot, rim

Fig. 22

P059, findspot 2. Diam. 0.18 m. Medium uneven brown to orange (5YR 5/3 to 5YR 6/6); nonvolcanic. Slightly flattened lip. Red slip? Hole pierced below lip. FN 96

Vessel, base

Fig. 22

P068, findspot 2. Diam. ca. 0.20 m. Medium even yellow-brown (5YR 6/6); tiny quartz inclusions; nonvolcanic. Flaring rim, round lip. Slip(?), very smooth surface. FN 98

95

Bowl(?), rim

Vessel, base

Fig. 22

P296, findspot 3. Diam. indet. (ca. 0.18-0.25 m?). Medium uneven gray-brown to orange-brown (5YR 5/4 to 2.5YR 6/8); many quartz inclusions; nonvolcanic. Flat, large, and irregular.Surface mottled yellowbrown to orange. FN

Fig. 22

P066, findspot 2. Diam. 0.12 m. Medium uneven gray to orange (5YR 6/4 to 2.5YR 5/8); nonvolcanic. Flat. Surfaces smooth and slightly compacted. FN

99

Cheesepot, body

Fig. 22

P298, findspot 3. Medium uneven gray to orange-brown (5YR 5/3 to 2.5YR 6/6); quartz and limestone(?) inclusions; nonvolcanic. Thickens toward base? Parts of two

i68

DANIEL

holespreservedin thinnerportion; wet-smoothedlike CookingPot ware, but not verysmooth. FN Fig. 19

100 Vessel,body

Findspot10.Threepieces.Diam. ca. 0.28 m?Mediumeven orangebrown(5YR 7/6); tiny regularquartz, small-large(2 mm) irregularquartz andother(limestone?)inclusions; nonvolcanic.Medium-sizedvessel; estimatefor diametermaybe incorrect if pot is globularand thuslargerin anotherdirection.Interiorvery Two curvedridges. irregular. FN 101 Vessel,body andlug

Fig. 22

P431. Mediumevenbrown(5YR 6/6); tiny-large(3 mm) quartzand limestoneinclusions;nonvolcanic. Self-slip?Horizontalhigh lug with fingerimpressionson edge. FN? COMPACTED

RED

CLASS

J. PULLEN

102 Vessel, pedestal base

Fig. 22

P053, findspot 2. Medium even red-brown (2.5YR 5/6); small quartz inclusions; nonvolcanic. Pedestal. Red slip and burnished interior? Compacted interior, smoothed exterior. FN 103 Vessel, body P065, findspot 2. Th. (wall) 0.012 m. Medium even red-brown (5YR 4/6); tiny quartz inclusions; nonvolcanic. Thick vessel (1.2 cm thick). Taenia poorly made, not thick, peeling off. Same vessel as 104. FN 104 Vessel, body P249(?) + P050, P245, findspot 2. Four pieces. Medium even brown (5YR 4/6); nonvolcanic. Large vessel. Self-slip. Three taenias. Same vessel as 103. FN

(105-107)

The smallgroupof sherdstermedCompactedRed is muchlike the Medium Coarseclass in terms of fabriccomposition,but with a decidedly morereddishappearanceto the fabricandwith a heavilycompactedsurface.Some of the uncataloguedbodysherdsin this categoryexhibita slip of some sort, and a few are burnished.The compactingof the surface, which createscracksand crazingon an otherwisesmooth surface,is also foundon piecesin the CookingPot andBuff classes;the compactingmay simplyrepresentan alternativesurfacefinish and not indicatea separate classas suggestedhere. Similarly,a flatbaseof indeterminatediameter,105, a bodysherdpreservingan appliedangledtaeniaband,106, and an unusualhandle,107, would not be out of place amongthe Medium Coarseor Buff categories (Figs.20, 23).The handle107 mustbe a flat,archedtabthatrisesfromthe rimof a bowlin two places;the innermarginof the handle,wherethe hole is, was thickenedfor strength,indicatingit was not piercedbut formed thatway.A tabhandle,usuallysingle,risingabovethe rimis characteristic of somebowlsof the TroyI culture,e.g., atThermi,but only rarelyin the FN periodof the mainlandandnearbyislands,e.g.,EutresisandKephala.42 105 Vessel, base

Fig. 23

P026, findspot 2. Diam. indet. (worn). Medium uneven gray to redbrown interior surface (7.5YR 4/2 to 2.5YR 5/6); many tiny quartz, small-

large (3 mm) irregularquartz inclusions; nonvolcanic. Flat, slightly rounded at edge (probably due to wear). FN

42. Thermi: Lamb 1936, pp. 73-77, esp. 75, fig. 26, 77, fig. 27; Eutresis: Caskey and Caskey 1960, pl. 46, no. 11.28;Kephala:KeosI, p. 14.

THE

PREHISTORIC

106 Vessel, body

~

}

-

REMAINS

OF THE

Fig. 20

P250, findspot 2. Medium uneven gray-brown to orange-brown (5YR 3/1 to 2.5YR 6/8); tiny, small limestone, no quartz(?)inclusions; nonvolcanic. Self-slip exterior. Taenia, angled. FN

10

ACROPOLIS

AT HALIEIS

107 Vessel,handle

I69

Fig. 23

P191. Coarseeven orange (2.5YR6/6); 1 pebble,5 mm.Tab handle,facetedin outline,with centralcircularhole. FN

PITHOS CLASS (108-111) Four sherds characterized by their coarse fabric and resemblance to later EBA pithos fabric have been placed into the Pithos class. The fabrics are orange to red-brown (2.5YR 4-6/6, 6/8), with a gray,gray-brown,or brown (2.5YR N6/0, 5YR 6/3, 7.5YR 6/4) core. The inclusions comprise many large pieces of limestone, typical of later EBA pithoi. The shapes, though, are not those of pithoi but of medium to large bowls represented by rims, 108 and 109, body sherds, 110, and large flat bases, 111 (Figs. 20,24). 111 has the impression of a mat on the underside (Fig. 29), but unfortunately it is too worn to determine much detail. Applied taenia bands are found on these large bowls (see 108 and 110), as is often the case on examples of later pithoi and large coarse vessels.

107

5cm

0

I

I

I

I

I

Figure 23. Compacted Red class

108 Bowl, rim

Fig. 24

P284, findspot 6. Coarse uneven gray to red-brown (5YR 6/3 to 2.5YR 4/6); tiny, small, and large limestone inclusions, as in EBA pithos fabric. Thickened rim, flat lip. Taenia on exterior; interior rim: slashes. FN 109 Large incurvingbowl, rim Fig. 24 P436. Diam. 0.30 m. Coarse uneven gray-brown core to red-brown (5YR 6/3 to 2.5YR 6/6); many large limestone inclusions, as in EBA pithos fabric. Incurving rim thickened to exterior, slightly rounded lip; heavy and thick. EH I(?) or post-prehistoric?

0

55cm W

class Figure 24.Pithos

110 Vessel, body

Fig. 20

Coarseunevenbrownto orange (7.5YR6/4 to 2.5YR 6/8); smallto largelimestoneinclusions,as in EBA pithos fabric.Taenia. FN 111 Mat-impressedbase Figs. 24, 29 HP252. Diam. ca. 0.12 m? Coarseunevengrayto orange(2.5YR 6/0 to 2.5YR 6/6); manytiny,small, large,andverylargelimestone inclusions,as in EBA pithosfabric; graypatchesmaybe grog?Flat.Mat impression:diagonalplaitedmat (too vigorousa cleaningmayhaverendered it unclearandworn). FN

)

I

DANIEL

I70

J. PULLEN

Figure 25. Pottery of EH II and LH

periods

-1

113 e4,

115

114

5Ccm

0 I

POTTERY

I.I

I

OF LATER

I

PERIODS

(112-116)

Several sherds of EH II and at least one of LH date were identified among the pottery from the acropolis.The EH II sherds are mostly body sherds identified by the presence of urfirnispaint, usually black, such as 112. Two sherds were identifiable by shape; 113 is the thickened rim of a small, incurving bowl, probably painted with urfirnis on the interior, and 114 is the rim of a sauceboat, distinctive in its complex curvature,and painted in black urfirnis (Fig. 25). 115 is a large, handmade ring base in a cooking pot fabric;it has been dated to the LH period.43Another ring base, 116, may also date to the LH period, but this is not certain. 112 Vessel, body

115 Vessel,base

P141, findspot 4. Medium even orange. Black urfirnis? EH II?

P434. Coarseunevengraybrowncoreto redbrown(5YR3/3 to 2.5YR 5/6). Ringfoot, handmade. Cookingpot. LH

113 Incurved bowl, rim

Fig. 25

P219, findspot 6. Diam. 0.11 m. Medium even gray-brown (5YR 6/4); nonvolcanic. Thickened rim, pointed lip. Remnant of black urfirnis interior? EH (I-)II 114 Sauceboat, rim

Fig. 25

P178. Fine even yellow-brown (5YR 7/4). Beveled lip. Black urfirnis. EH II

116 Vessel,base

Fig. 25

Fig. 25

P433. Diam. 0.11 m. Medium unevengray-brownto red-brown (5YR4/1 to 5YR 5/4); handmade. Flat,with slightringformedby clay overlappingfromwall;5 holes preserved,drilledfrominteriorbefore firing,but conicalafterfiring.Similar to 115;Mycenaeancookingpot? LH? EH I?

THE LITHICS: CHIPPED STONE (117-230) The chippedstone artifacts,while fewerin numberthanthe ceramicartifacts,have a similardistributionin that the majoritywere found in the samegeneralcontexts,mixedwith materialof the historicalperiods.Similar problemsin identificationand analysisexist for the lithics as for the ceramics:they were found in contextsprimarilyof Archaicand Classical

43. I would like to thank Gisela Walbergfor confirmingthe LH date for this piece.

THE

OF THE

REMAINS

PREHISTORIC

ACROPOLIS

AT HALIEIS

I7I

date, or surfacedepositswithout cleardates.Thus, we may have lithics fromperiodsotherthanthe FN-EH I periodincludedamongthosefrom the acropolis,just aswe havea few ceramicsof laterprehistoric(EH II and proLH) periods.Certainelementsof the chippedstone manufacturing cess areabsentfromthe Halieis acropolis,but whetherthis is due to the methodsof recoveryor to the actualprocessremainsproblematic.Nevertheless,a carefulcomparisonof the Halieis acropolischippedstone with the chipped stone collections from the SouthernArgolid Survey and FranchthiCaveyields some conclusions(see below),and makesclearthe relativeconsistencyof the Halieisassemblage. The chippedstone from the acropolisconsistsof 107 obsidianand seven chert (flint) items (Fig. 26).44 These 114 items have been broken down by their positionin the reductivetechnologyused to manufacture chippedstone (Table4).45 Coresor corefragmentsareabsenton the Halieis acropolis,and debris (123-127), waste flakes,and tracesof visiblecortex(obsidian:117121;chert:224-225) arefew,suggestingthatthe materialwas not worked locallyon anyscale.It shouldbe notedthat some of the corticalpiecesare typologically"blades"(117-119), but becauseof the presenceof cortex, theyarecountedascorticalpieces.Therearealsoa numberof "semicrested" blades(128-136), that is, bladesthat have evidencefor core preparation on only partof theirdorsalsurface;these havebeen countedas blades. Figure26. Chippedstone

122

119

44. Jacobsenused the figuresof 115 obsidianand six chert objectsin his draftmanuscript(1974). In 1997 I found 107 obsidian and seven chert objects assignableto the acropolis;an additionalsix obsidianobjectswere noted in the recordsas missing. CatherinePerlksstudied the lithics in 1972 and again in 1997, and I am deeply gratefulto her for her observations on the Halieis material.I would also like to thank P Nick Kardulias, who is publishingthe lithics from the LowerTown of Halieis, and Curtis N. Runnels for readingand commenting on this section of the manuscript. 45. I employ here the chipped stone terminologyas used by Karduliasand Runnels (1995) for the Southern Argolid Surveyin orderto facilitate comparisons.

146

~

~~

161

~~

134

157

156

147

-~~~~~~~II , ~

132

128

C

~

158

142

160

Z

178

164

183

193

175

226-v

-@>-?

-e0-X7-@4l

208

217

220

0

221

5cm

230

DANIEL

I72

TABLE 4. COMPOSITION ASSEMBLAGE Type

Cores Corticalpieces* Crestedblades Blades Flakes Debris TOTAL * The

Obsidian % no. 0 5 1 74 22 5 107

0.0 4.4 0.9 64.9 19.3 4.4 93.9

J. PULLEN

OF CHIPPED STONE

Chert/Flint no. % 0 2 0 2 3 0 7

0.0 1.8 0.0 1.8 2.6 0.0 6.1

Total no.

%

0 7 1 76 25 5 114

0.0 6.1 0.9 66.7 21.9 4.4 100.0

corticalpieces include three blades.

The obsidian varies in color from gray-black to black. Although none of these pieces has been subjectedto scientific analysis,all resemble Melian obsidian in appearance, and most can easily be attributedto the two principal varieties recognized by Renfrew:"opaquewith a milky or pearly luster in reflected light"; and "partiallytranslucent with alternating transparent and opaque bands or striations in transmitted light."46 The seven remaining pieces (6.1%), 224-230, are of a generally reddish variety of flint or chert commonly found in the Hermionid.47Traces of cortex are preserved on two specimens and perhaps on a third. One piece may be regarded as a tool (226). 117 Blade,centralportion S103. Max.p.W. 0.86, max.p.L. 1.67, max.p.Th. 0.36 cm. Cortex. 118 Blade,centralportion

123 Debris S017.

S021, findspot6. Max.p.W. 1.08, max.p.L. 2.50, max.p.Th. 0.39 cm. Cortex.Accidentalretouch. 119 Blade,centralportion

0.60 cm. Crested blade/lame de d6gagement.

Fig. 26

S115. Max.p.W. 0.95, max.p.L. 1.67, max.p.Th. 0.34 cm. Cortex. Inverseretouchon rightedge. 120 Corticalflake,fragment S003, findspot3. 121 Waste S057, findspot7. Cortex,heavily rolled. 122 Blade,proximalportion Fig. 26 HS016, findspot2. Max.p.W. 1.20, max.p.L. 2.98, max.p.Th.

124 Debris, fragment S037. 125 Debris S043. 126 Debris S102, findspot 6. 127 Debris.

Sill, findspot6. 128 Blade

Fig. 26

S077, findspot 6. Max. p.W. 0.98, max. p.L. 3.07, max. p.Th. 0.38 cm. First blade after removal of crested blade.

46. Dixon and Renfrew1973. Renfrewvisuallyinspected the collection from Halieis in 1965 and reportedseeing nothing that could not have been from Melos. 47. It is quite possible that more than one sourceis involved here, for the color of this materialvariesfrom a dark red, almost chocolate (Munsell lOR 3/2, "duskyred"),to a lighter red (1OR 4/3, "weakred").Cf. Karduliasand Runnels 1995, pp. 77, 103-104; van Andel and Vitaliano 1987, p. 20.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I73

129 Blade, segment

138 Flake

S079, findspot 6. Max. p.W. 1.38, max. p.L. 1.40, max. p.Th. 0.50 cm. First blade after removal of crested blade.

S061, findspot 6. Max. p.W. 2.18, max. p.L. 1.98, max. p.Th. 0.27 cm. Crest preparation. 139 Flake

130 Blade, central portion S084, findspot 6. Max. p.W. 1.02, max. p.L. 1.43, max. p.Th. 0.25 cm. First blade after removal of crested blade.

HS023, findspot 8. Max. p.W. 1.39, max. p.L. 2.64, max. p.Th. 0.36 cm. First blade after removal of crested blade. Fig. 26

S054. Max. p.W. 0.99, max. p.L. 1.10, max. p.Th. 0.32 cm. First blade after removal of crested blade. Use retouch on left edge and distal end.

141 Flake S013, findspot 3. Traces of steep retouch? 142 Blade, central portion

Fig. 26

S039, findspot 6. Max. p.W. 0.93, max. p.L. 1.87, max. p.Th. 0.27 cm. Similar to 178. 143 Blade, proximal portion

133 Blade, central portion

S010.Max.p.W. 1.70, max.p.L. 2.93, max. p.Th. 0.37 cm. Flat crested, with direct and inverse retouch on left edge; first blade after removal of crested blade. 134 Blade, distal portion

140 Flake S083. Rejuvenation or crest preparation.

131 Blade

132 Blade, central portion

S069, findspot 6. Max. p.W. 1.26, max. p.L. 1.49, max. p.Th. 0.32 cm. Crest preparation.

Fig. 26

S016, findspot 3. Max. p.W. 1.62, max. p.L. 3.13, max. p.Th. 0.35 cm. Flat crested, with some direct retouch on left edge; first blade after removal of crested blade.

S050, findspot 6. Max. p.W. 1.1 cm. Worn like 160, but without retouch. 144 Blade, proximal portion S051, findspot 6. Max. p.W. 0.84, max. p.L. 1.51, max. p.Th. 0.22 cm. Use retouch on both edges. 145 Flake S063, findspot 6. Max. p.W. 1.67, max. p.L. 2.67, max. p.Th. 0.60 cm. Slight vertical retouch on left edge.

135 Blade, proximal portion S045, findspot 5. Max. p.W. 0.91, max. p.L. 1.95, max. p.Th. 0.25 cm. Use retouch right edge and accidental left edge; first blade after removal of crested blade. 136 Blade, fragment

146 Blade, distal portion missing

Fig. 26

S088. Max. p.W. 1.06, max. p.L. 3.74, max. p.Th. 0.37 cm. Ventral utilization, both edges. 147 Blade, distal portion broken

Fig. 26

S053, findspot 6. Max. p.W. 1.89, max. p.L. 1.88, max. p.Th. 0.31 cm. First blade after removal of crested blade. Possible notch.

S092. Max. p.W. 1.20, max. p.L. 3.81, max. p.Th. 0.39 cm. Utilization, both edges.

137 Rejuvenation flake, fragment

148 Blade, proximal portion missing

S002. Rejuvenation or crest preparation.

S104, findspot 6. Max. p.W. 1.35, max. p.L. 2.92, max. p.Th. 0.39 cm.

I74

DANIEL

Smallretouchon left edge;abrupt retouchon otheredgewith intentionalbreakingof blade;wearafter retouch? 149 Blade,centralportion HS013, findspot1. Max.p.W. 0.93, max.p.L. 2.75, max.p.Th. 0.25 cm. Both edgesutilized,centralridge removed. 150 Blade,proximalportion

J. PULLEN

158 Blade, proximal portion Fig. 26 S078, findspot 6. Max. p.W. 1.43, max. p.L. 2.19, max. p.Th. 0.49 cm. Heavy retouch on both edges, notched? 159 Blade S100, findspot 6. Max. p.W. 0.80,

max.p.L. 2.17, max.p.Th. 0.34 cm. Inverse retouch, notched. 160 Blade, central portion

S007, findspot3. Max.p.W. 0.90, max.p.L. 1.95, max.p.Th. 0.33 cm. Directretouchon left edge. 151 Blade,proximalportion S024.Max.p.W. 0.98, max.p.L. 1.85, max.p.Th. 0.26 cm. Direct retouchon left edge anduse on right. 152 Blade,proximaland central portion

S041. Max.p.W. 0.92, max.p.L. 2.47, max.p.Th. 0.44 cm. Some directretouchon left edge? 153 Blade,proximalportion S055, findspot6. Max.p.W. 1.20, max.p.L. 1.75, max.p.Th. 0.35 cm. Directretouchon left edge.

Fig. 26

S031. Max. p.W. 1.35, max. p.L. 2.46, rnax.p.Th. 0.45 cm. Accidental retouch from use; worn after retouch; perhaps sickle? 161 Blade, distal portion broken

Fig. 26

S110. Max. p.W. 0.82, max. p.L. 2.54, max. p.Th. 0.30 cm. Denticulated, by three inverse adjacent notches on left edge. 162 Blade, central portion S114. Max. p.W. 1.12, max. p.L. 1.79, max. p.Th. 0.29 cm. Retouch on right edge and small inverse denticulation on left edge. 163 Blade, central portion HS014, findspot 2. Max. p.W.

154 Blade,proximalportion

1.12, max.p.L. 2.58, max.p.Th. 0.35

S095.Max.p.W. 1.14, max.p.L. 2.51, max.p.Th. 0.25 cm. Direct retouchon rightedge. Splintered? Flat crested?

cm. Possible inverse retouch on both edges

155 Blade,distalportion

S042, findspot 5. Max. p.W. 1.47, max. p.L. 3.13, max. p.Th. 0.40 cm. Inverse flaking on distal end.

S106, findspot6. Max.p.W. 0.80, max.p.L. 2.32, max.p.Th. 0.34 cm. Directretouchon rightedge. 156 Blade,centralportion

Fig. 26

S067, findspot6. Max.p.W. 0.98, max.p.L. 2.84, max.p.Th. 0.27 cm. Samematerialas 178? 157 Blade

Fig. 26

S028, findspot6. Max.p.W. 0.90, max.p.L. 2.15, max.p.Th. 0.32 cm. Partiallybackedformingnotches, with use retouchon otheredge.

164 Blade

Fig. 26

165 Splintered blade, fragment S004. Max. p.W. 1.38, max. p.L. 1.55, max. p.Th. 0.22 cm. 166 Splintered blade, fragment S044, findspot 5. Max. p.W. 1.04, max. p.L. 2.48, max. p.Th. 0.48 cm. 167 Splintered blade S048, findspot 6. Max. p.W. 0.92, max. p.L. 1.89, max. p.Th. 0.57 cm.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I75

168 Splintered blade

179 Blade, central portion

S049, findspot 6. Max. p.W. 1.30, max. p.L. 2.10, max. p.Th. 0.31 cm.

S015. Max. p.W. 0.81, max. p.L. 1.35, max. p.Th. 0.20 cm.

169 Splintered blade, central portion

180 Blade, distal portion broken

S056, findspot 6. Max. p.W. 1.29, max. p.L. 1.80, max. p.Th. 0.28 cm.

S025. Max. p.W. 0.93, max. p.L. 2.65, max. p.Th. 0.35 cm.

170 Splintered blade

181 Blade, distal portion broken

S058, findspot 7. Max. p.W. 1.00, max. p.L. 1.35, max. p.Th. 0.32 cm.

S026, findspot 6. Max. p.W. 1.07, max. p.L. 2.69, max. p.Th. 0.34 cm.

171 Splintered blade S068, findspot 6. Max. p.W. 0.84, max. p.L. 1.49, max. p.Th. 0.23 cm. 172 Blade S029, findspot 6. Max. p.W. 0.87, max. p.L. 0.76, max. p.Th. 0.30 cm. No retouch. 173 Blade S020, findspot 6. Max. p.W. 1.22, max. p.L. 3.06, max. p.Th. 0.44 cm. Percussion traces on distal end. Splintered?

182 Blade, central portion S027, findspot 6. Max. p.W. 1.06, max. p.L. 1.46, max. p.Th. 0.35 cm. 183 Blade, central portion

Fig. 26

S046, findspot 7. Max. p.W. 0.88, max. p.L. 2.58, max. p.Th. 0.32 cm. 184 Blade, distal portion S052, findspot 6. Max. p.W. 0.76, max. p.L. 1.74, max. p.Th. 0.27 cm.

174 Blade, central portion S036. Max. p.W. 0.63, max. p.L. 1.27, max. p.Th. 1.04 cm. 175 Blade, distal portion missing

185 Blade, proximal portion S064, findspot 6. Max. p.W. 0.90, max. p.L. 1.67, max. p.Th. 0.38 cm.

Fig. 26 186 Blade, distal portion

S038, findspot 6. Max. p.W. 0.66, max. p.L. 4.03, max. p.Th. 0.32 cm. Pressure-flaked;especially long and thin.

S066, findspot 6. Max. p.W. 1.16, max. p.L. 1.42, max. p.Th. 0.28 cm.

176 Blade, proximal portion

187 Blade, distal portion missing

S006. Max. p.W. 0.70, max. p.L. 1.74, max. p.Th. 0.26 cm.

S070, findspot 6. Max. p.W. 0.93, max. p.L. 1.96, max. p.Th. 0.28 cm.

177 Blade, distal portion S008. Max. p.W. 0.90, max. p.L. 2.00, max. p.Th. 0.28 cm. 178 Blade, central portion

Fig. 26

S012, findspot 3. Max. p.W. 1.10, max. p.L. 1.80, max. p.Th. 0.30 cm. Same material as 156?

188 Blade, segment S073, findspot 6. Max. p.W. 1.13, max. p.L. 0.72, max. p.Th. 0.24 cm. 189 Blade, central portion S075, findspot 6. Max. p.W. 1.36, max. p.L. 3.78, max. p.Th. 0.46 cm.

I76

DANIEL

J. PULLEN

190 Blade, central portion

202 Blade, central portion

S076, findspot 6. Max. p.W. 0.87, max. p.L. 2.84, max. p.Th. 0.27 cm.

S086. Max. p.W. 1.00, max. p.L. 1.71, max. p.Th. 0.34 cm. Accidental retouch.

191 Blade, proximal portion S081. Max. p.W. 0.69, max. p.L. 1.14, max. p.Th. 0.38 cm. 192 Blade, segment

203 Blade, central portion S087. Max. p.W. 1.29, max. p.L. 2.08, max. p.Th. 0.33 cm. Accidental retouch.

S085. Max. p.W. 0.80, max. p.L. 0.80, max. p.Th. 0.15 cm.

204 Blade, distal portion

193 Blade, proximal portion Fig. 26

S089, findspot 6. Max. p.W. 1.24, max. p.L. 2.96, max. p.Th. 0.53 cm. Accidental retouch.

S090. Max. p.W. 0.95, max. p.L. 2.45, max. p.Th. 0.30 cm. 194 Blade, proximal portion S091, findspot 6. Max. p.W. 0.98, max. p.L. 1.85, max. p.Th. 0.29 cm. 195 Blade, central portion S097, findspot 6. Max. p.W. 0.90, max. p.L. 1.43, max. p.Th. 0.19 cm. 196 Blade, distal portion S099. Max. p.W. 0.79, max. p.L. 1.58, max. p.Th. 0.15 cm. 197 Blade, central portion, broken on edge S109. Max. p.W. 0.85, max. p.L. 1.60, max. p.Th. 0.32 cm. Margin not preserved (accidental?). 198 Blade, central portion S112. Max. p.W. 0.96, max. p.L. 1.58, max. p.Th. 0.22 cm. 199 Blade, central portion S0l1, findspot 6. Max. p.W. 1.20,

max.p.L. 1.75, max.p.Th. 0.34 cm. Accidental notch. 200 Blade S065, findspot 6. Max. p.W. 0.60, max. p.L. 2.02, max. p.Th. 0.30 cm. Accidental retouch.

205 Blade, proximal portion S113. Max. p.W. 1.36, max. p.L. 2.45, max. p.Th. 0.43 cm. Accidental utilization on both edges. 206 Blade, central portion. S094. Max. p.W. 0.76, max. p.L. 1.82, max. p.Th. 0.18 cm. Bulb just broken away. 207 Blade, central portion SOOl.Max. p.W. 0.96, max. p.L. 2.53, max. p.Th. 0.22 cm. No bulb preserved. 208 Blade, central portion

S030, findspot 5. Max. p.W. 0.90, max. p.L. 2.01, max. p.Th. 0.27 cm. Use retouch on both edges. 209 Flake S011. Max. p.W. 1.89, max. p.L. 2.76, max. p.Th. 0.37 cm. 210 Flake S014. Max. p.W. 1.42, max. p.L. 1.94, max. p.Th. 0.33 cm. 211 Flake S022, findspot 3. 212 Flake

S034. 201 Blade, central portion S072, findspot 6. Max. p.W. 0.97, max. p.L. 2.02, max. p.Th. 0.31 cm. Accidental retouch.

Fig. 26

213 Flake

S040.

PREHISTORIC

THE

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I77

223 Flake or chip

214 Flake, fragment S062, findspot 6. Max. p.W. 1.24, max. p.L. 1.65, max. p.Th. 0.33 cm.

S035. 224 Flake S060, findspot 6. Max. p.W. 2.7, max. p.L. 4.1 cm. Red flint. Cortex.

215 Flake, fragment S071, findspot 6.

225 Debris 216 Flake, fragment S082. Max. p.W. 1.34, max. p.L. 0.98, max. p.Th. 0.33 cm.

S047. Red flint. Cortex. Accidental retouch. 226 End scraper

217 Flake

Fig. 26

S093.

Fig. 26

S105. Max. p.W. 1.01, max. p.L. 1.12, max. p.Th. 0.26 cm. Beige flint. Nonsymmetrical edge.

218 Flake 227 Blade

S096.

HS388. Red flint. 219 Flake 228 Blade, distal portion broken

S108. 220 Flake

Fig. 26

S074, findspot 6. Max. p.W. 1.45, max. p.L. 2.35, max. p.Th. 0.38 cm. Inverse distal retouch (fresh from trampling?). 221 Flake, fragment

Fig. 26

S009, findspot 3. Max. p.Th. 0.58 cm. Retouched (accidental; fresh from trampling?). 222 Flake or chip S023, findspot 6.

TYPO

S107, findspot 6. Max. p.W. 1.11, max. p.L. 2.92, max. p.Th. 0.48 cm. Chocolate flint. 229 Flake S032. Max. p.W. 2.16, max. p.L. 2.37, max. p.Th. 0.60 cm. Red flint. Inverse retouch of bulb. 230 Flake

Fig. 26

S033, findspot 5. Max. p.W. 1.66, max. p.L. 2.12, max. p.Th. 0.48 cm. Red flint. Direct retouch on right edge. Rolled? Cortex?

LOGY

Although it is possible that more than one industry is represented here, the preserved remains include a relatively limited number of types. The assemblage is clearly dominated by evidence of blade production. Of the total number of pieces of obsidian and flint (114), at least 82 (71.9%) may be regarded as typical blades or blade segments (117-119, 122, 128-136, 142-144,146-208, 226-228). Only a fraction of these have their percussion bulbs intact, yet there is sufficient evidence to indicate that both pressure flaking and percussion were employed in their production. The majority of pieces from Halieis are incomplete blades or segments of blades, whether intentional or accidental (Table 5). Often neither the proximal end (that end preserving the bulb of percussion) nor the distal end is preserved. A large portion of the preserved segments have lengths 1.5 to 2.5 times the maximum width (46 out of 77, or 60%);alto-

I78

DANIEL

J. PULLEN

gether 65, or nearly 85%, have lengths 1.5 or more times the maximum width. These figures perhaps reflect the preservationmore than any intentionally desired size of segment. No evidence for truncation (the deliberate shortening of a blade or segment) was noted. Most of the blades or blade segments found in our collection are of the simple parallel-sided varietywithout secondaryworking. Although no complete blades have been preserved, the largest surviving segments usually exhibit ventral surfaceswith little longitudinal curvature.Dorsal surfaces generally give the appearance of blades with a triangular section, trapezoidal section, or, less commonly, semicrested (128-136) of roughly triangular section. The semicrested form is more common than the fully crested form (lame a crete), represented on the acropolis by only one example (122). A certain uniformity is attributableto this assemblage on the basis of its generally microscale character.This is indicated not only by the small size of most of the flakes but also by the maximum widths and thicknesses of the measurableblades and blade segments.The low coefficients of variation (Table 6) indicate consistency in the widths and thicknesses of the blades.48The mean width and coefficient of variationfor the Halieis blades fit well with those of Neolithic blades found by the Southern Argolid Survey, though the Halieis blades are thicker.49The Bronze Age blades from the Southern Argolid Survey (overwhelmingly of EH date) are narrower than the Halieis blades. Nevertheless, the Halieis blades are as consistent in their widths as those blades from surrounding areas. The crested and semicrested blades are significant, as they provide the best evidence for working of obsidian at Halieis itself. A crested blade, one with latitudinal flake scars on either side of a central dorsal ridge or crest (axes of flake scars perpendicularto the axis of the blade), is the first blade removed after the preparationof the core.50The next blade removed is semicrested and has latitudinal flake scars along only one facet of the dorsal face; the other dorsal face has the usual one longitudinal scar from the removal of the initial, crested blade. Four flakes (137-140) may have resulted from preparation of the crest or rejuvenation of the core.

TABLE 6. DIMENSIONS OF OBSIDIAN BLADES AND BLADE SEGMENTS FROM HALIEIS AND COMPARISONS WITH OTHER SAMPLES Site orRegion* x Halieis Southern Argolid, Neolithic Southern Argolid, Bronze Age Lerna III (EH II) Kephala, Neolithic

1.03 1.04

Width(cm) SD CV 0.29

0.86

0.29 0.24

0.96 1.44

0.24 0.37

n

x

28.2 27.9

77 35

0.33

27.9 25.0 25.7

Thickness(cm) SD CV n 77

308

0.28 0.26

0.12 0.08 0.08

36.4 28.6 30.8

35 308

318 128

0.26 0.39

0.07 0.16

26.9 41.0

318 127

*Figuresfor sites and regions other than Halieis taken from Kardulias1992, p. 440, table 6. x mean, SD = standarddeviation,CV = coefficient of variation,CV = 100 x SD/x

=

TABLE 5. BLADE SEGMENTS Preserved Portion

Proximalend only Proximaland medial portion Medial portion/segment Medial and distal portion Distal portion

Number

15 8 33 1 7

48. The distinctionbetween "blades" and "bladelets,"based on widths greater or less than a certainarbitrarymeasurement, has not been made here. Bordaz (1970, p. 51) advocates1.25 cm for the cutoff point, while Perles has utilized 1.1 cm as the line of demarcationat Franchthi(Franchthi3, p. 36, table III), but for purposesof descriptiononly. Whichever dimension one uses, it is clearthat the Halieis acropolisblades, like those of so many Neolithic and Early Bronze Age lithic assemblagesin the Aegean, are relativelynarrow.Cf. Kardulias1992, p. 440, table 6, for comparativedimensions (a version of this chartis reproducedin Kardulias and Runnels 1995, p. 98, table 5.15, but the dimensions are given there in tenths of millimeters,not millimetersas per the caption). 49. Kardulias1992, p. 440, table 6. 50. This preparationof the core can take place duringthe removalof blades if the core needs to be rejuvenatedfor furtherblade production.

THE PREHISTORIC

51. Accidental notches as in 199 or accidentalretouchas in 118, 200-205, and 220-221 can be producedby a varietyof methods, especiallyduring and after excavation. 52. No silica gloss or matting (dulling of the obsidiansurfaceby contactwith plant material)was observedon 160 or on any other chipped stone from the Halieis acropolis. 53. A possibilityraisedby Perlks (pers.comm.). But as she and Kardulias (pers.comm.) note, a Bronze Age date is not out of the question. 54. Cf. Karduliasand Runnels 1995.

REMAINS OF THE ACROPOLIS AT HALIEIS

I79

While these crested and semicrested blades and flakes attest some working of obsidian at Halieis, they do not necessarilyprovide evidence of core formation. Some obsidian blades and flakes give evidence of utilization in the form of light marginalwear, due to the relativelybrittle quality of obsidian (141-149).5 143 and 148, though, exhibit wear similar to that of the denticulated blade 160 and perhaps could be labeled as sickles. Other blades (23, or 30% of all obsidian blades) provide evidence of purposeful secondary working. In a number of cases, this takes the form of simple marginal retouch (usually quite delicate) on one edge, perhaps intended to serve as light blunting (150-155). Yet blunting may also have been effected by means of burin blows, that is, by striking a narrowlongitudinal flake from one edge of a blade in order to create a steeper surface there, as is the case with 156. In some instances, the marginal retouch is extensive, in order to produce notched or denticulated blades. Examples of the former are 157159, while examples of the latter are illustrated by 160-162. Inverse retouch is also observable on a number of blades. 119 is of interest in this respect since it shows evidence of light marginal retouch on the same edge as the inverse retouch; it also retains some cortex. One blade, 163, has possible inverse retouch on both edges, while another blade, 164, has inverse flaking on its distal end. A number of blades are splintered:165-171. The denticulated blade 160, as well as perhaps also 143 and 148, which are worn in a similar manner though without the same retouch, can be identified as a sickle blade, but none of the other retouched obsidian blades and flakes are identifiable as true tool types.52 Flakes of obsidian, that is, pieces that do not conform to the blade form, are not plentiful: only fifteen were recognized (209-223), of which only two (220-221) have any retouch, though only that on 220 may be deliberate. Among the chert pieces, in addition to the two cortical pieces 224225, aretwo unretouchedblades (227-228) and two retouched flakes (229230). A small specimen of red flint, 226, is of some interest.It is the snapped distal end of a small blade or bladelet with asymmetrical steep retouch at the end. Although it may give the appearanceof a small end-scraper,this artifact was possibly meant to serve a purpose similar to that of a small trapeze or geometric microlith. As such, it may be earlier than the general FN-EH I characterof the acropolis lithic collection.53 The chipped stone assemblage from the acropolis at Halieis is typical of a small FN-EH site in the Hermionid.54 While some sites in the Hermionid were extensively involved in obsidian procurement and processing into cores, other sites, such as Halieis, were involved only in the later part of the process, that of blade production from cores prepared elsewhere, while still others seem to have been merely recipients of blades produced elsewhere. As has been noted for sites of the Final Neolithic and Early Bronze Age, blades were a major part of the stone tool production, and the high frequency of blades in the Halieis acropolis assemblage fits this model. What is noticeable about the Halieis assemblage, compared to the rest of the Hermionid, is the low percentage of tools of various types. Other than the three possible sickiles,direct retouch and utilization are

I80

DANIEL

J. PULLEN

found on a limited number of blades and flakes, but not in the manner of any recognizable tool type. Runnels, in examining the lithics from the Lower Town at Halieis, has noted the rarityof cores, debris from core reduction, and tools in contrast to the high frequency of blades and flakes in that collection;55that is, the lithics appear to have been worked elsewhere. Runnels argues that while the assemblagesfrom the Lower Town date to the Geometric, Archaic, and Classical periods in which the various pieces were found, the individualpieces may actuallybe reused items from earliertimes. He notes that if the lithics were indeed "kickups"from earlier levels, or had been part of the earth used for mudbricks, then prehistoric pottery would have been found in the Lower Town; such is not the case. The presence of lithics, generally "unspecializedor retouched blades, bladelets, and flakes," in floor deposits and mixed with domestic debris suggests to Runnels that the lithics from the Lower Town represent deliberate selection and use by the inhabitants.56Though the acropolis lithic assemblage would appearto be similarto that from the Lower Town, the circumstancesof the findspots are different. Rather than being a functional part of the Archaic and Classical deposits as in the Lower Town, the acropolis lithics are truly kickups, found with ceramics of the appropriateperiod. One might suggest that if the Lower Town lithics are indeed collected from elsewhere, the acropolis would be a likely source, especially as the prehistoric levels there were disturbed by building activities. The few lithics from the Industrial Terrace are more likely to be similar to those from the Lower Town, since no prehistoric pottery was identified there.

GROUND AND POLISHED STONE (231-235) Four ground and polished stone objects found on the acropolis have been associated with the prehistoric period, largely on the basis of their morphology (Figs. 27-28). It is important to recognize, however, that none of their contexts permits any certainty of dating. These objects include two steatite "buttons"or "whorls"of Mycenaean (LH III) date (231 and 232), a pierced pebble bead or pendant (233), and a felsite hammerstone (235).57 The pierced pebble bead is a rather simple ornament, one that would not be out of place in a Neolithic or earlier context, as they are not very common in EBA contexts. One such pebble, though incompletely pierced, was discovered by the Southern Argolid Survey at a FN site near Franchthi Cave; another incompletely pierced example was found in the recent excavations at Tsoungiza, but not in a good context.58 The hammerstone need not be of prehistoric date; indeed, one of the major contributions of the Southern Argolid Survey was the demonstration that handstones or hammerstones have been found at sites of all periods from the Neolithic to the Roman period, with little chronological distinction possible.59 Given the importance of milling activities in an agriculturalcommunity or a self-sustaining military outpost, it would not be surprisingif the hammerstone in question dated to the Classical period of occupation on the acropolis.60)

55. Runnels 1982, pp. 366-369. 56. Runnels 1982, p. 371. 57. R. Giegengack of the Department of Geology at the Universityof Pennsylvaniaoriginallyidentified the stones included here. 58. Naturalpebble,incompletely pierced,from site C29 of the Southern Argolid Survey;on the basis of comparandaof Upper Mesolithic date from FranchthiCave, Karduliasand Runnels (1995, p. 136, fig. 116) suggest that this ornamentmay be a recycled object,but a similarnaturalpebble, also incompletelypierced,from late MH fill at Tsoungiza that includes EN, EH IIII, and MH material(Nemea Valley ArchaeologicalProjectinventory number430-8-1 [Pullen, in preparation]) would suggest it need not be so early. 59. Karduliasand Runnels 1995, p. 121. 60. Similarly,a fine-grainedquartz sandstonedisk, 234, is more likely to be of Classicaldate than prehistoric,given the lack of similarobjectsin prehistoric contexts.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

i8i

AT HALIEIS

233

232

231235

234

Figure27. Groundstone objects

231

0 I

Figure28. Groundstone objects

I

234

233

232 I

I

I

5cm I

None of the ground stone from the acropolis at Halieis is necessarily dated to the FN-EH I period. The two steatite conical buttons add to the meager evidence for possible Mycenaean activity at Halieis. 231 Whorl

Figs.27-28

HS025. Diam. 2.6, H. 1.6 cm. Steatite. LH III 232 Whorl

Figs. 27-28

HS045. H. 1.8 cm. Steatiteor alabaster.Fragmentary. LH III 233 Bead

Figs. 27-28

HS041, findspot3. Max. dim.

1.6,Th. 0.2 cm. Limestone,tan microcrystalline. Piercedpebble. 234 Stone disk

Figs. 27-28

HS038. Diam. 3.1,Th. 0.9 cm. Sandstone,fine grained,quartz sandstone,iron-oxidecementing agent. Post-prehistoric? 235 Grinding(hand) stone

Fig. 27

HS043, findspot7. L. 6, diam. 5.5 cm. Felsiteor rhyolite.

MAT IMPRESSIONS Six flat baseshavemat impressionson theirundersides,but only five are presentedhere(Fig.29); the sixthis too smallandtoo poorlypreservedto allow more than the recognitionof an impression.A majordifficultyin dealingwith mat impressionsis that they representonly one face of the textileand they rarelypreservethe edge or selvedge,so the identification of warp and weft are difficult.For purposesof description,I have arbitrarilyassignedthe term"warp"(the vertical,fixed element)to that ele-

DANI

I82

EL J . PULLEN

. .

#

'-|i:?4'w

: .? t:_ jgg.*

j< *'4ibi Y.i., ' .....

,:

W

*

............................................

_

_

L'.

>;.

oX

.............

......

E

l _

-,"0 !

l

g

_

.

_

i SIF .t

..

_

|

F

_

l

gw s".

,$

%"Ar'.'; *S_ 14a._

i

*-'$=

i

l

I

'

.:

|

_

s

l * __

_

l

*

_

|

I 1

|

w

|

I| 811 a

_

*_*-

|

i11

*

_

W

_ L

-I

s

!

-

|

_

-l_

t1 1

_

!

i

|

W

z

_

*

|

.

....

j _ : W

_

--i

l

_

_

.

E

|

_

_

_

_ 2

*

|

i _

_

I_

.

..........

SX

B:

?X.

I

g

_

NS

C\s

,,

l

i

_

64

60

-

.9-.S_

,;|.,?,

w@

..:''

$

|

1

1

r

|

__ r

l5deY

1 1

1

1

r

e

F et

. s { .sX :

.2::

*F.4 $i

|

r

2

2^-&ilwS$

x :.it. . _

_

'

-

B

r

I_

r

_

_

.. *.,

.x_ if_

nn

7D

,._ .

._

_

*ti_F _

.

._

_

bases FiFre 29.Mat-impressed ment that seems to be more often in the background,and "weft"(the horizontal, flexible element) to that which appearsto be interwoven and more often in the foreground. One impression, 93, may be of basketry utilizing wide flat material, though the area preserved is small and unclear.The remaining four impressions seem to be impressions of mats, using cordage, in either a plaited or twined technique. No remains of finely woven cloth6"or fabric were detected. The term "plaiting"is sometimes applied to the technique used to form mats, to distinguish it from "weaving"as applied to the production of cloth.62Weaving involves the use of a loom or frame to provide tension for the warp elements; in plaiting there is no need for a frame, as the warp is often self-supporting. The most common techniques used to produce mats in the FN and EH I periods are the simple or diagonal plaiting techniques.63In a simple plait, the same groupings of warp elements are used for all weft elements. In diagonal plaiting, the warp elements are regrouped in various patterns. Twining involves two weft elements: one goes over and the other goes under the same warp element, then the two weft elements are twisted around each other before the next warp element. This last technique is used more for producing mats and basketry of an open or loose weave than for producing cloth. Twining often produces weft rows where the twisting of the weft over the warp produces a distinctive angled appearancein each segment of the weave, much like a single row of herringbone masonry.

61. I use the term "cloth"following Barber(1991, p. 5), who distinguishes between cloth, defined as "large,thin sheets of materialmade from fiber," usuallywith the aid of a loom, and mats and baskets,defined as relativelystiff and self-shaped,made from materials that do not requirea loom for interweaving. 62. Barber1991, p. 5. 63. The majorityof the Tsoungiza impressionsare of plaited mats, often with a definite diagonalpattern, indicating a shift in the groupingof warp elements as the weft is plaited or woven through.

THE

64. Though only one mat impression at Tsoungiza is dated to the FN period,it should be pointed out that the total numberof FN ceramicsis fairlysmall. 65. Talioti:Weisshaar1990, pp. 2-4; Kephala:CarringtonSmith 1977. 66. This section is paraphrasedfrom Jacobsen'smanuscript(1974), updated with referenceto Judith Shackleton's study (Franchthi4). None of the recoveredmarineshell was found in the Halieis storeroomin Nauplion in 1982 or 1997, so all identificationsand conclusionspresentedhere must remain unconfirmed.The animalbone and shell from the Archaic and Classical deposits on the acropolisare to be publishedby David Reese in the forthcomingHalieis I. 67. Confirmationof these identificationswas providedby Nicholas J. Shackletonof CambridgeUniversity upon examinationof photographsof selected specimensfrom the site.

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I83

111 and 60 show the diagonal plaited technique. In 60 the weft elements are doubled and several warp elements grouped to create a strong diagonal pattern; 111, unfortunately,is not very clear,so we cannot determine whether there was multiplication of the weft elements. The impression on 64 seems to indicate a widely spaced weft over multiple warp; though the forwardweft elements are not particularlyangled as in a twining technique, the wide spacing of the weft elements suggests that this might be the technique involved. There is some indication too that this mat might have been circular.65 might also be twined, as the weft seems angled between each warp group.This impression is faint and worn, but it seems to be of a fairly tightly interwoven mat, which probably indicates that it was not twined. Mat impressions, while found on the bases of vessels of many periods, are common in the EH I period of the northeast Peloponnese. Of the twenty-five mat impressions preserved at Tsoungiza, sixteen date to the EH I period.64Weisshaar has recently commented on the high frequency of mat impressions on the bases of vessels of the Talioti, or late EH I, phase, while on Kea, a large number of mat impressions were found on the EN pottery at Kephala.65That six mat impressions appearamong the small number of sherds (ca. 400) from the Halieis acropolis gives an impression of their popularity during the FN-EH I period. Mats were probably not made specifically for use in pottery manufacturing; rather, the presence of several impressions of torn mats in the Tsoungiza corpus suggests that the mats were recycled into pottery manufacture.We might imagine househ-oldswith mats covering floors and other surfaces as well as basketry and open-plaited or twined containers such as nets, none ofwhich is usuallydirectlypreservedin the archaeologicalrecord. Though meager, this evidence of mat impressions from the acropolis does give us a small window into the domestic economy at the site and helps place Halieis among other sites of the FN-EH I period.

MOLLUSCA Both animal bone and shell were noted during the course of the excavations on the acropolis, but the only class of organic remains in association with the prehistoric pottery was the shells of both marine and terrestrial molluscs (Fig. 30).66 Land snails and severalvarieties of marine shell were especially numerous in the deepest deposits on the hill, particularlyin the pits or fissures in bedrock, where evidence of prehistoric activity was also most frequently encountered. Findspot 4, the pit north of Building A, was a seemingly uncontaminated deposit of prehistoricpottery (see above) and contained a substantialquantity of marine molluscs, most notably bivalves. A sample of these was submitted to the University of Pennsylvania for radiocarbon analysis, and the result is discussed below. Several species of marine molluscs were noted among the preserved collection: Cerastoderma(Cardium)edule L. (cockle), Murex trunculusL., L.67Although CerithiumvulgatumBrug., and perhaps Spondylusgaederopus accurate statistics are not available,it is fair to say that Cerastodermaand

DANIEL

I84

J. PULLEN

Figure30. Selectionof molluscs

f*1***

Z*

0

3cm

Murex were the most numerous species represented.The recent publication of the Franchthi Cave molluscs by Shackleton provides the most indepth study of this class of archaeological remains.68She suggests that and Cerithiumare found in shallow waters, as well as Murex, Cerastoderma, deeper, and are relatively easy to collect. Spondylus,while found in deeper waters and more difficult to collect, are often found washed ashore, and it is probablyby that method that many if not most were collected. All four species are readily edible. It is likely that most of the preserved shells represent the remains of meals; none shows clear-cut indications of working for a tool or ornament. Thus, although other evidence relating to the diet and the economy of the prehistoric inhabitants of the Halieis acropolis is not available, these modest remains suggest that shell collecting was at least part of their subsistence base.

ABSOLUTE CHRONOLOGY The sole indication of the absolute date of the prehistoric remains from the acropolis is the radiocarbon analysis of a group of marine shells, predominantly Cerastoderma(Cardium)eduleL.,69from findspot 4, which, as discussed above, contained potsherds of the Brown-Slipped and Medium Coarse dasses. The result is a date of 5102 ? 72 B.P., or,when calibratedin accordancewith the bidecadal calibration curve of Stuiver and Reimer on a 2-sigma basis, 4037-4016 B.C., 4006-3750 B.C., 3750-3712 B.C.; with the marine-corrected option, a date range of 3909-3367 B.C. is derived.70 In the most recent assessment of Early Bronze Age chronology, Manning makes a valiant attempt to deal with the meager data for the FN and EH I periods in particular.71He supports a date of ca. 3100-3000 B.C. for the beginning of the EB I period in the Aegean. He argues that while certainly "partof the LN/FN period must have occurred during the first

68. See Franchthi4. 69. The samplewas collected in 1965 and consisted of a total of thirtysix shells, twenty-eight of which were (Cardium) of the species Cerastoderma eduleL. 70. For the uncalibrateddate (P1397), see Lawn 1971; for calibration curvesand marine-correcteddate, see Stuiverand Reimer 1993 and Manning 1995, pp. 169-170. 71. Manning 1995, pp. 168-170.

THE

PREHISTORIC

REMAINS

OF THE

ACROPOLIS

AT HALIEIS

I85

half of the fourth millennium B.C." (with the Halieis date contributing modestly to this argument), there are in addition a few dates, however problematic, that support a mid- to late-4th-millennium B.C. date for the FN period, thus closing the apparent 4th-millennium "gap"in radiocarbon dating between dates associated with the later Neolithic and dates associated with the Early Bronze Age.72Therefore, though perhaps better than no date at all, the above radiocarbondate can be accepted as no more than the most general indication of the absolute chronology of the prehistoric occupation at Halieis.

CONCLUSIONS

72. Manning 1995, p. 170. 73. For the SouthernArgolid Survey,see Pullen 1995. For Franchthi, see Franchthi10, p. 99. 74. See Jameson,Runnels,and van Andel 1994 for a more complete pictureof the geographicaland historicalsetting of the Hermionid and its connectionsto the rest of Greece and the Aegean.

The assemblage of prehistoric materials from the Halieis acropolis spans the period from the Final Neolithic through Early Helladic I. Given that the prehistoric materials are found in contexts mixed with material of later date, it is not surprisingthat these prehistoricceramics and lithics are chronologically mixed themselves. Despite the overlap in shapes among the various classes (e.g., similar shapes in the Volcanic Red-Slipped and Burnished and the Volcanic Black-Burnished classes) that might lead to the conclusion that those classes were contemporaneous, there are enough features to conclude that the preserved prehistoric ceramic assemblage from the acropolis includes material from both the FN and the EH I periods as defined elsewhere. The frying pans and the Volcanic Red-Slipped and Burnished class are strong indications of an EH I date, while the BrownSlipped and Black-Burnished classes and the cheesepots are strong indications of a FN date. Attempts at discerning some meaningful distribution of the variousclassesof prehistoricceramicswere not entirelysuccessful: the best candidate is the small pit north of Building A from which came most of the examples of the Brown-Slipped class as well as a small quantity of marine shells that provided the single radiocarbon determination for the prehistoric material. The FN material from Halieis, like that collected by the Southern Argolid Survey, has little in common with the FN assemblage from Franchthi Cave. Vitelli suggests that the FN material from the Survey postdates the stratified deposits at Franchthi Cave.73 The Halieis acropolis is apparently one of several small FN-EH I sites situated on low hills throughout the Hermionid; another site with some FN and considerably more EH I material was found by the Southern Argolid Survey (A33) nearby on the Nisi Kheliou to the west, across from the entrance to the Porto Cheli Bay. The hilltop location of both sites provided some degree of security,while at the same time affording good vantage points for viewing the surrounding sea, the nearby islands such as Spetses and Dhokos, and furtherlands, including the Peloponnese, across the Argolic Gulf.74Numerous resources in addition to the sea and its sheltering bay are found in the surrounding area,including gently rolling hills and well-watered lowlands. The large number of FN sites in the Hermionid represents a fundamental change in settlement patterns from the earlier(perhapsintermittent) habitationdocumented at EranchthiCave

i86

DANIEL

J. PULLEN

to a number of open-air and cave sites. This expansion of settlement continues into the Early Bronze Age.75 While a careful study of the ceramics can distinguish the pottery of the FN period from that of the EH I period, perhaps what is more important is the evidence for continuity between the later FN and EH I at Halieis and in the southern Argolid; this continuity stands in contrast to the apparentdifferencesbetween the earlierFN as representedat Franchthi Cave and the later FN as representedby Halieis and the southern Argolid material. The prehistoric remains on the Halieis acropolis, then, look not back to the Neolithic but forward to the Early Bronze Age.

REFERENCES Barber,E. J. W. 1991. Prehistoric Textiles: The Development of Cloth in the Neolithic and BronzeAges,

Princeton. Biers,W. R. 1969. "Excavationsat Phlius, 1924:The Prehistoric Deposits,"Hesperia 38, pp. 443-458. Bordaz,J. 1970. Tools of the Old and New StoneAge, Garden City, N.Y. Boyd,T. D., andW. W. Rudolph. 1978. "Excavationsat Porto Cheli and Vicinity,PreliminaryReportIV: The LowerTown of Halieis, 19701977,"Hesperia 47, pp. 333-355. CarringtonSmith,J. 1977. "Cloth and Mat Impressions,"in Keos I, pp. 114-127. Caskey,J.L., and E. G. Caskey.1960. "The Earliest Settlements at Eutresis:SupplementaryExcavations, 1958," Hesperia 29, pp. 126167. Coleman,J. E. 1985. "'FryingPans'of the Early Bronze Age Aegean,"AJA 89, pp. 191-219. Dixon,J. E., and C. Renfrew.1973. "The Sourceof the Franchthi Obsidians,"in T. W. Jacobsen, "Excavationsin the FranchthiCave, 1969-1971, Part 1,"Hesperia 42, pp. 82-85. Dousougli, A. 1987. "Makrovouni, KefalariMagoula,Talioti: Bemerkungenzu den Stufen FH I und II in der Argolis,"PZ 62, pp. 162-220. Franchthi = Excavations at Franchthi Cave, Greece,Bloomington.

2 = T. H. van Andel and S. B. Sutton, Landscape and People of the Franchthi Region (Franchthi 2), 1987.

3 = C. Perles,Les industries lithiquestaille'esdeFranchthi (Argolide,Grece)1: Pr6sentation generaleet industriespal6olithiques (Franchthi3), 1987. 4 = J. C. Shackleton,Marine MolluscanRemainsfromFranchthi Cave(Franchthi4), 1988. 8 = K. D. Vitelli, Franchthi and NeolithicPottery1: Classification CeramicsPhases1 and2 (Franchthi 8), 1993. 10 = K. D. Vitelli, Franchthi NeolithicPottery2: TheLater NeolithicPhases3-5 (Franchthi10), 1999. Holmberg, E. J. 1944. The Swedish ExcavationsatAsea in Arcadia (SkrRom 4.11), Lund. Jacobsen,T. W. 1974. "The Prehistoric Remains,"unpublishedmanuscript. Jameson,M. H. 1969. "Excavationsat Porto Cheli and Vicinity,Preliminary ReportI: Halieis, 1962-1968," Hesperia 38, pp. 311-342. 1972. "Excavationsat Porto Cheli, Halieis,"ArchDelt27, Chronika,pp. 233-236. Jameson,M. H., C. N. Runnels,and T. H. van Andel. 1994. A Greek TheSouthernArgolid Countryside: from Prehistoryto thePresentDay, Stanford. Kardulias,P. N. 1992. "The Ecology of Bronze Age Flaked Stone Tool Productionin SouthernGreece: Evidence from Agios Stephanosand the SouthernArgolid,"AJA96, pp. 421-442. Kardulias,P. N., and C. N. Runnels. 1995. "The Lithic Artifacts:Flaked

75. Jameson,Runnels, and van Andel 1994; Wiencke 1989.

THE

PREHISTORIC

REMAINS

OF THE

Stone and Other Nonflaked Lithics," in Runnels et al., eds., pp. 74-139. KeosI = J. E. Coleman, Kephala,a Late NeolithicSettlementand Cemetery (KeosI), Princeton 1977. Lamb,W. 1936. Excavationsat Thermi in Lesbos,Cambridge. Lawn, B. 1971. "Universityof Pennsylvania RadiocarbonDates XIV," Radiocarbon13, pp. 363-377. Lekkas,S., and D. Papanikolaou.1977. "On the Phyllite Problemin des Peloponnesus,"Annalesgeologiques pays hellMniques 29, pp. 395-410. LernaIII = J. B. Rutter,ThePotteryof LernaIV(Lerna III), Princeton 1995. Manning, S. W. 1995. TheAbsolute Chronology of theAegeanEarlyBronze and Age:Archaeology, Radiocarbon, History(Monographsin Mediterranean Archaeology 1), Sheffield. Mee, C., and G. Taylor.1997. "Prehistoric Methana,"in A Roughand Rocky Place:TheLandscapeand Settlement Historyof theMethanaPeninsula, Greece(LiverpoolMonographsin Archaeologyand Oriental Studies), C. Mee and H. Forbes,eds., Liverpool,pp. 42-56. Mylonas, G. E. 1959. AghiosKosmas:An EarlyBronzeAgeSettlementand CemeteryinAttica, Princeton. Nordquist,G. C. 1987. A Middle HelladicVillage:Asinein theArgolid (Acta UniversitatisUpsaliensis, Boreas:Uppsala Studies in Ancient Mediterraneanand Near Eastern Civilizations16), Uppsala. . 1995. "The Pottery of the Early Helladic III and Middle Helladic Periods,"in Runnels et al., eds., pp. 43-51. Phelps, W. W. 1975. "The Neolithic Pottery Sequencein Southern Greece"(diss. Universityof London). Pullen, D. J. 1995. "The Pottery of the Neolithic, Early Helladic I, and Early Helladic II Periods,"in Runnels et al., eds., pp. 6-42.

DanielJ. Pullen THE FLORIDA

STATE UNIVERSITY

DEPARTMENT

OF CLASSICS

TALLAHASSEE,

FLORIDA

[email protected]

32306

ACROPOLIS

AT HALIEIS

I87

. In preparation. The Early Bronze Age Settlement on Tsoungiza Hill, Ancient Nemea. Renfrew, C. 1972. The Emergence of Civilisation: The Cycladesand the Aegean in the ThirdMillennium B.C.,

London. Runnels, C. N. 1981. "ADiachronic Study and Economic Analysis of Millstones from the Argolid, Greece" (diss. IndianaUniversity). . 1982. "Flaked-StoneArtifacts in Greece duringthe Historical Period,"JFA9, pp. 363-373. Runnels, C. N., D.J. Pullen, and S. Langdon, eds. 1995. = Art fact and Assemblage: The Findsfrom a Regional Survey of the SouthernArgolid, Greece 1: The Prehistoric and Early Iron Age Pottery and the LithicArtifacts,

Stanford. Shriner,C., and M. J. Dorais. 1999. "A ComparativeElectron Microprobe Study of Lerna III and IV Ceramics and Local Clay-Rich Sediments," Archaeometry 41, pp. 25-49.

Stuiver,M., and P.J. Reimer.1993. "Extended14C Data Base and Revised CALIB 3.0 14CAge CalibrationProgram,"Radiocarbon 35, pp.215-230. Touchais,G. 1980. "Laceramique neolithique de l'Aspis,"in Etudes Argiennes(BCH Suppl. 6), Athens, pp.1-40. van Andel, T. H. 1987. "The Adjacent Sea," in Franchthi 2, pp. 31-54.

van Andel, T. H., and C. J. Vitaliano. 1987. "Waterand Other Resources," in Franchthi2, pp. 17-20. Weisshaar,H.-J. 1990. "Die Keramik von Talioti," in Tiryns: Forschungen

undBerichteXI, Mainz, pp. 1-34. Wiencke, M. H. 1989. "Changein Early Helladic II,"AJA93, pp. 495-509. Zerner,C. W. 1978. "The Beginning of the Middle Helladic Period at Lerna" (diss. Universityof Cincinnati).

HESPERIA Pages

69,

2000

DU

ILW

NC

Z

AT

189-226

KtOM/v\O AN

8TH-CENTURY

POTTERY

SEQUENCE

ABSTRACT Building Z at Kommos is built into one of the Late Minoan shipsheds of Building P; its major phase of use is in the Middle Geometric period. The potteryfrom the building, including earlierIronAge materialand some from post-use deposition, is fully published here. The totality gives the opportunity of seeing the various contemporaryproductsof the Mesara in fuller detail than is providedby other published sets of material.The purpose of the building is unclear,but it appearsto have served a subsidiaryfunction to that of the temple to the north. The excavations in 1992 at Kommos, South Crete, included the thorough investigation of the third gallery of the Late Minoan IIIA Building P.1 This large structure lies at the southern foot of the seaside hill on which the Bronze Age settlement, and later the temple and associated structures of the Iron Age, were located.A sounding in 1980 (trench36B) had sampled the specific area, through deep sand, and in 1985 an area further west, seaward, had been investigated (trenches 65A1 and 2; here the stratigraphy was poorly preserved,due largely to erosion). Middle Geometric material from 36B was interpreted as dumped fill from the temple area, such as was found in quantity a little to the north. It was not wholly surprising, however, that in 1992 excavators in trench 82A came down on a habitation area of the Geometric period, immediately dubbed Building Z in the 1. This is the second of a planned series of three supplementaryarticles on the Iron Age pottery from excavations at Kommos,under the direction of Joseph W. Shaw,Universityof Toronto.The bulk of the materialis now included in KommosIV (Callaghan andJohnston 2000), that from the store building Qhas alreadybeen published (Johnston 1993), while a third will deal with a rangeof materialof general ceramicinterest.As ever,I am indebted

to Prof. Shaw for materialand financial support,as also to the British Academy and UniversityCollege London (Facultyof Social and Historical Sciences, GraduateSchool, and Institute of Archaeology)for grantsto assist my work. I am gratefulto Jenny Doole, who made the drawings(save for 14, 64, and 98, drawnby Nikki Holmes), and to TaylorDabney for the photographsof some recalcitrantmaterial.The prin-

cipal trenchmasterswereJos6e Sabourin(81B), LaraTabac(82), and Gordon Nixon (83C). It should be noted that what I term here, for reasonsof abbreviation, "BuildingZ" (or "mainphase"in Fig. l:b) is called "BuildingZ, phase II" in KommosIV, with phase I being my "earlierstructures"(Fig. l:a). All measurementsare in centimetersunless otherwise specified.

I90

ALAN

W. JOHNSTON

sequential series (Fig. 1). A preliminary report has alreadypresented the architecturalevidence and touched on the characterof the finds, ceramic and other.2The aim of this article is to present the ceramic material recovered from within and immediately adjacent to Building Z, which gives a good cross-section of the appearance,and to a certain extent development, of local pottery styles ca. 800-725 B.C. This is the period in which the new temple B was built to the north; the pottery from Z very usefully supplements that from the sacral area.3Finds from the earlier Iron Age levels (Fig. l:a) are very scrappy,but the material is almost wholly discrete from that recovered from the main phase of use above; in particular,the "typefossils" of the two periods, the earlier bell-cup and the later flat-based, one-handled, black-glazed cup, are stratigraphicallydistinct. The subtitle of this article reflects the main period of use, although indeed the earlier phase is also treated below. The study is in a sense a supplement to the broadertreatment in KommosIV of both the sanctuaryfinds and Building Z, and I am therefore sparing in recounting the details published there, as in reviewing the range of contacts between Kommos and other areas suggested by the pottery. Regrettably,as is the case for much of the material from the southern area of the site, time has not been kind to the ceramic remains, which consist mainly of small, worn, and often isolated sherds. Nonetheless, the material is of interest because of its relatively well stratified context. The picture that we have of the later Geometric pottery of the Mesara is the result of study by a number of scholars, who will be cited as appropriate below; the pottery from Building Z provides a significant check on their conclusions. The building concerned (Fig. l:b) was inserted into the shipshed P3 the construction, from nearby debris, of a cross-wall and a form of by buttress wall up against the still-standing north wall of P3; the latter was presumablybuilt to lessen the span of the roof, from 5.90 to 4.70 m.4 The west end of the structurewas not preserved and so its length remains uncertain, though it is likely that it terminated at the still-standing east end of Building P.The base of the newly constructed walls is at 3.95 masl. The builders not only employed stones fallen from the Minoan structure but cleared the area for use at the same time. One large stone in the new buttress wall, with its lower edge at 4.30 m, may have acted as a threshold, though it is ratherhigh above the original floor; there are no other signs of an entry to the structure,though naturallyone may have been situated at the lost western end. The building was excavatedin trench 82A to the west and trench 82B to the east; at the same time trench 81B cleared the area to the east of the rear wall, with trench 83C being dug further east, at the east end of the Minoan gallery.

LOWER LEVELS The lowest levels containing post-Minoan pottery, albeit with a large admixture of Minoan, were at approximately 3.75 m and stratigraphically below the building level of Z, associated with some minor architectural

2. Plansof the broaderareaare includedin ShawandShaw1993. 3. I do not questionthebasicuseof this areafor cult purposes,though I note the remarksabout usage of the buildings made by Viviers 1994, p. 245. He perhapsunderestimatesthe number of figurinesand miniaturevases from the whole generalarea,and also the significanceof the originalstatue base in Temple C, as well as the inscriptions mentioning deities, but he undoubtedly is right to stressthe communalnature of activitiesin the buildings. 4. Shaw and Shaw 1993, p. 175.

BUILDING

Z AT

KOMMOS

z

I9I

hearth

3. 85

bench

v

bottom 3.84

bench

bottom 3.84

4.04

W

(

24.06

1

v.4.10

a

l. 12.65

b 0 joseph

Figure 1. Building Z at Kommos: (a) earlier structures and (b) main phase

1

2

w.show-giuliono

3 bionco

4 1992

5M.

W.

ALAN

I92

JOHNSTON

TABLE 1. SCHEMATIC LISTING OF AMOUNTS OF POTTERY BY DEPTH PAILS, ARRANGED APPROXIMATELY Pail UPPER

36B/2a

BUILDING

West Sherds

65A2/26

Pail

Central Sherds Grams

Pail

East Sherds

Grams

LEVELS

ca. 150

6,900

82A/18 22 23 23A 19 24 20 11 12 7 13 14 82B/b45 46 47 48 49

81 122 152 239 149 302 139 104 149 244 152 62 510 161 221 148 289

2,050 925 900 1,440 4,510 1,605 4,850 3,080 1,875 4,750 2,715 1,420 5,860 1,450 3,160 4,990 7,380

81B/50D 50E 61 83C/64 61

331 129 146 99 333

4,990 2,980 2,250 2,860 5,110

26,100 9,150 8,700 550 6,100 7,900 11,430

82A/15 15A 16 16A 10 17 21 25 26 27 29 30 31 32 33 82B/50 52 53 (54

171 92 109 145 72 89 98 353 198 28 77 162 349 79

4,030 6,102 3,290 2,850 1,995

81B/64 65 66 (67 83C/66

256 199 54 186 155

8,040 4,805 1,505 4,520) 3,865

667 408 139 67

2,410 5,110 3,930 1,530 2,590 3,425 6,420 1,230 2,410 21,250 10,970 3,270 1,240)

82A/28 34 36 37 38 82B/54 55

25 133 7 106 57 67 110

1,265 3,790 200 1,985 2,205 1,240 1,785

81B/70 71

185 226

3,520 3,240

Z

65A2/22C 23 65A1/12 14d 40 36B/5 ca. 60 20 ca. 230 21 ca. 340

LOWER

Grams

FROM RELEVANT

100

1,100

LEVELS

-

4,850

4.50-4.35 m. b Pails 41-44 above4.60 m. CAt4.48-4.23 m, mostly higher than Building Z levels to the east. dAt 4.09-3.73 m, reacheslevels that predateBuilding Z levels to the east. a

BUILDING

Z AT

KOMMOS

I93

features.5 In trenches 82 and 81B the material is of Late Protogeometric style, scanty though it is (Figs. 2-3). The presence of several pedestal feet is a good general pointer to the date; the kind of ribbed pedestal seen in Z above this level (e.g., 57) is of a very different type. More than half of the vessel was preserved in only two cases, the amphora 1 and the very small bell-skyphos 7; the rest of the catalogued pieces are rarely more than single fragments. The material from all trenches is taken together. The total amount and weight of pottery from the lower levels are listed in Table 1 and catalogued here from west to east. 1

5. Shaw and Shaw 1993, p. 177, where the architecturalhistory of the areais described. 6. E.g., Coldstreamand MacDonald 1997, p. 233, placing their El, with a less elaborateprofile than 2, late in PG.

Belly-handled amphora

Fig. 3

C9585 (82A/34 and 37, with a fragment of handle from 82B/55) and C10694 (82A/34). P.H. 35.5, original H. ca. 55-58, Diam. 33, Diam. (rim) ca. 13.5. Fine buff fabric, 7.5YR 7/6, with rosier core; many, mainly small, dark inclusions; creamy surface, possibly a slip. Many fragments preserved;most join, giving a profile from lip to mid-body; one handle is preserved in two joining fragments, but there is no handle attachment preserved on any body fragment; it is therefore placed as high as possible in the profile drawing. The possibility that this was a hydria cannot be totally ruled out since a vertical handle could be accommodated, though the frieze decoration makes it extremely unlikely.The foot, C10694, is from the same pail as much of the amphora and has the same fabric and decorative finish; it almost certainly belongs. Only its small size (Diam. 3.7 at the narrowest point) leaves room for doubt, and it is therefore placed in a "neutral"position in Fig. 3. The top of the lip is not preserved, though the wall is very thin where the preserved outturn at the top of the neck is found, evidence for a small, upright rim. Horizontal strap handles are set low on body. The foot, if belonging, has a very narrowwaist before flaring out into a pedestal, only partly preserved. The decoration is in typical dull brown paint, now much worn in places. Band on inside of rim. Top of neck painted outside. On neck: threeline zigzag between single lines; below, checkerboard,four "squares"

high, but irregularlypainted; at base, pendant crosshatched triangles; band over turn of shoulder. On shoulder, eight sets of five concentric circles, one of them given much more space than the rest (but not because of the proximity to any handle). Four lines below, above a frieze of quadruple zigzags (though some sections have three or five lines); below that, upright crosshatched triangles between lines. A single line on lower wall fragments (if belonging); outside of handle painted. On the foot, two bands preserved below a painted area. The scheme of the decoration, as well as the pedestal foot, suggests manufacture in the Late Protogeometric or, at latest, Protogeometric B period. The syntax is reminiscent of a probable PGB hydria from Knossos, KNC, T107.185 (fig. 112); a LPG

kraterfromTekketomb D 5 (KNC, fig. 57) has similar motifs.

2

Amphora (or hydria), rim and neck

Fig. 2

C10692 (81B/70 and 71, the two lowest Iron Age levels east of Z). P.H. 4.7, Diam. (rim) 18. Semifine light tan clay, 7.5YR 7/4, with many dark and some white inclusions. Two joining fragments of rim and neck. Flaring neck and low vertical, concave lip. Band on inside of rim, another on outside of rim and top of neck. The relatively developed lip, with concave face, suggests that the date of these lowest levels should be chronologically advanced, perhaps beyond the LPG style of accompanying material, although the relatively straight line to the neck could be taken as a counterbalancing earlier feature.6

ALAN

194

W. JOHNSTON

/ 2 4

3o

7

5

3

Krater,foot

Fig. 2

C9584 (82A/28, the lowest IA level in the western part of Z). H. (foot) 4.3, Diam. (foot) 7.0. Fine buff clay, 5YR 7/6. Single fragment of the pedestal foot of a krater.Very worn and chipped, the edge of the foot very nearly preserved at just one point. Heavy foot and thick walls; a noteworthy sinking at center of underside. Dull, dark paint on inside; no traces of paint on outside, possibly wholly worn? For the type of foot, see also 4. 4

Krater(?),base

Fig. 2

C9759 (81B/70). P.H. 6.2, Diam. (foot) 6.6. Medium coarse pink clay, 5YR 8/4. Five joining fragments of a base, probably of a krater.Low conical foot with depression at center of underside. No traces of any decoration, though the surface is generally in good condition. 5

Krater,body

Fig.

C10100 (81B/71). P.H. 8.0. Medium coarse buff-brown clay, 7.5YR 7/4, with some large white inclusions; cream surface.

2

Three joining fragments of upper wall. Ridge preserved near top of wall; it and the wall above are painted in dull brown glaze. On wall, to right, part of the outer circle of a set of concentric circles; three verticals to left, with part of a diagonal line to their left, most likely part of a chevron motif forming the central decorative panel on the wall. The shoulder is at a relatively shallow angle. The range of motifs used is basic, though precise parallels may be lacking. 6

Krater,rim

Fig. 3

C10453 (65A2/26, lowest pail with any IA pottery west of Z, top at 3.65 m). Diam. (rim) ca. 44. Semifine light brown clay, 5YR 8/4 to 7/4, with many varied inclusions. Two joining fragments. Slightly thickened outcurving rim with ridge below. Once wholly painted. There are possible body fragments in 82A/37 and a body sherd with concentric circles from 82B/55, all in very similar fabric. The profile points to a MPGLPG date.

Figure 2. Material from levels earlier than the construction of Building Z. Scale 1:3

BUILDING

Z AT

KOMMOS

I95

,

1

I

I lk1:

flt

15

t6

N'

PI~~~~~~~~~? .

.

J

12

r

9

.* 14

Figure 3. Large pots from Building Z and below. Scale ca. 1:7 7

Bell-skyphos

Fig. 2

C9763 (81B/71, with joining fragment from 70). P.H. 12, Diam. (rim) 10. Fine pinkish brown clay, 5YR 7/4. Nine fragments, several joining, of rim and body of a bellskyphos. Slim-bodied skyphos, with both handles at least partly preserved;flaring rim; base lost. Much worn in parts. Dull dark paint. Lower part of body reserved outside; the rest painted, probably dipped. Curvature of the walls, height of handles, and extent of glaze indicate a date in LPG. 8

Closed vessel

C10693 (82A/37). P.H. 5.6. Fine beige clay, 10YR 6/4, with some inclusions. Single sherd used as a "label."Worn fragment; a single hole cut at top and at least the top right side deliberately cut to shape; the other edges seem more like

accidental breaks. An oddity; if some form of identificatory tag, one must suppose it was not marked in any specific "textual" way. 9

Cooking pot

Fig. 3

C10077 (82A/38). P.H. 11.5, Diam. (rim) 33. Semicoarse clay with a variety of nonmicaceous inclusions; red-brown surface (5YR 5/6), gray in core. Thick fabric. Large fragment, composed of four joining fragments of rim, wall, and handle. Small horizontal rim with slight ridge below; deep bowl. Horizontal, upturned handle attached close under rim. Traces of burning inside. The context is probably early IA. Though most identifiable material in the pail is Minoan, one worn shoulder sherd seems to be from a jug of broadly PG date.

I96

ALAN

W.

JOHNSTON

A few other pieces should be mentioned: a body fragment from 82B/ with a simple band, is very close in fabric to 117; from the same pail 55, another fragment is akin to 106. In 82B/54 are plain body fragments of a distinctive yellow-surfaced closed vase, fragments of which appear also in pails 52 and 53 above, and in 81B/66, at a similar level to 52. From 82B/37 come neck and handle fragments of an amphora similar to 2 and to 129; body fragments of dipped bell-skyphoi come from 82A/34 and 37. None of the black-glazed cups so common in Building Z above were found in these levels. There is little here that would appear to date later than LPG; 1 and perhaps also 2 are possible PGB pieces. One may tentatively ascribe the material to a date before ca. 825 B.C. The material is clearly mixed and scrappyand shows that there was little occupation of the areabetween the abandonment in LM IIIB and the construction of Building Z, and that what there was should be placed largely in the later part of that period.

BUILDING

Z

No joins have been found between material from the lower levels discussed above and that from Building Z (3.95 m and above); as just noted, a few examples of sherds from the lower levels may belong to pots in Z, e.g., 117, 129, though only in the case of the former is the link strong. Clearly Minoan pottery is extremely rare, certainly not "sporadic,"above the level 3.95 m; a few LM III pieces are in the "transitional"pail 82B/54, and more, including two conical cups, in the equally low 82A/33, at the first floor of Z. Similarly, to the east of Z, Minoan pottery is tolerably frequent at a lower level (81B/70, 71), slight in the "transitional"pail 67, and vestigial from 4.00 to 4.50 m. It then becomes a little more noticeable in higher levels. On the other hand, joins between material from the floor at 3.95 m and that from higher levels are frequent and widespread;some pieces, notably 12, 13, and 14, were widely strewn after breakage. Only one actual join has been located with material to the west (41), although nonjoining fragments of the same pots can be confidently identified in other cases;41 is an interesting case, linking not only Building Z with the west but also with material in the temple dump to the north. I reserve opinion here on whether there was a second floor in Z at ca. 4.30 m, on a level with the lower edge of the presumed threshold block in the north buttress wall. First we should examine the pottery found within the levels between 3.95 and 4.30 m. While by far the greatest concentration of pottery was found in a pail (82B/50) at the level of the possible upper floor and near the rear,east wall of Z, at least ten pots (eight catalogued) from this pail join sherds from lower pails; one may well suggest that the material representsclearanceof accumulateddebris to the back of the building or, perhaps less likely in view of the joins at lower levels, material fallen from wooden shelves on the back wall. At the same time, a relativelylarge amount of sherds was found in trench 36B in the presumed western part of the structure.

BUILDING

Z AT

KOMMOS

I97

The pails with material from this level are listed in Table 1, from west to east (in approximate descending stratigraphic order per trench). PHOENICIAN

POTTERY

10 Amphora sherd(s) C7855 (65A1/14). See Bikai 2000, p. 19. The precise level of the find within the pail cannot easily be correlatedwith Building Z; it could belong to the earlier horizon, predating Z. 11 Flask, neck and handle

Fig. 4

C10695 (82A/26). P.H. 4.6. Nearly fine, but gritty, light brown

clay, 7.5YR 6/6, with white inclusions. Two joining fragments of neck and handle. Neck heavily ridged inside. Surface extremely worn, any possible decoration now lost. With respect to the three other flasks known from the temple dumps, this piece is larger than C3134 (noted by Bikai); close in size to C3078, which is not ridged inside (Bikai 2000, p. 16); and similar in fabric to the thicker-walled C8050 (p. 17).

From the rest of the material I extracted thirteen body sherds and one lip sherd that could possibly come from Phoenician amphoras, though the lip is high, 3.8 cm, and a little everted. None has the rather sandy texture of the normal amphora type. AMPHORAS

12 Amphora

AND

HYDRIAS

Fig. 3

C9609 (82A/15, 16A, 24,26, 29, 30, 31, 33; 82B/49, 50, 52; 81B/ 64, 66, 67; 65A2/22). A possible fragment from 83C/67. C10475 (82A/5), the foot. H. ca. 48, max. Diam. ca. 32, Diam. (rim) 10, Diam. (foot) 9.6. Semicoarse, highly micaceous, gritty, and porous clay; pale buff surface;core variously fired, generally purple-brown. One hundred and two fragments of one or more amphoras, from all parts save lower neck and lower wall. Irregularity of potting, painting, and firing makes attribution of individual fragments very difficult. Some fragments moderately worn, others wholly abraded (especially those from 81B and 83C, making any final judgment on their pertinence impossible). The only joins between pails are between 82A/30 and 82B/50; but also from those two pails come three fragments (not joining others) with the ends of vertical bands beneath the handle area, indicating that there is more than one piece in this highly distinctive fabric.The foot cannot easily be joined with the major body section

as restored on paper (as made clear in Fig. 3), though it has every appearance of pertaining to an amphora of this type, save that the red-orange surface has a match on only one, assured,body fragment. Slightly everted lip; neck apparently curving smoothly into shoulder; stout ring foot. Much of one broad, vertical handle preserved, with two shallow grooves down the outside. Decoration in dull paint, variously fired chestnut to black. Top of neck and lip painted; neck decoration uncertain. Two very irregular bands at top of shoulder.Two sets of eight concentric circles, with central dot, on the shoulder; below them, a band above one (or possibly two) lines. Irregulardouble serpentine line on belly, above three bands. Bands down the grooves on the handles and around the handle roots, extending down onto the belly. No traces of paint on worn foot. Clearly an import, probably Cycladic; the amount of mica seems excessive for any East Greek ware, as is the coarseness of the material, even perhaps for Milesian (cf. KNC, p. 405, T285.89).

ALAN

I98

$1

W. JOHNSTON

r` 11

17

16 19 nW

.

.

._

.:

-

I .)

_

~ ~~ ~ ~ ~~~~-

__A^

r

22 18 13 Amphora (or hydria?),body C10042 (81B/67; 82A/18, 19, 22, 23, 33, 42, 45). No useful dimension measurable.Fine pinkish buff, very micaceous clay,with cream surface.Thirty fragments, some joining (but no joins between pails). Full rounded body. Surface very worn, some fragments totally abraded.Decoration in dull brown: concentric circles on shoulder, with one set largely preserved, above triple

band; single band set apart below these and an undulating band on the belly; double band on lower body. More than one amphora may be represented here, but the fabric appears the same throughout, including that of the sherds from the lower levels, 82A/33 and 81B/67. Again an import, probably from the Cyclades, but the clay far finer than in 12. A piece that appears close, a hydria from Kavousi, is thought by Tsipopoulou to be an import from

Figure 4. Building Z. Phoenician (11) and large closed vessels. Scale 1:3

BUILDING

Euboea (Tsipopoulou 1990, p. 140, fig. 4; she kindly informs me that RichardJones would support that origin after analysis of the clay). 14 Amphora

Fig. 3

C9555 (82A/3, 7, 8, 13, 16A, 19, 23, 31; 82B/49, 50,52, perhaps 54). Diam. (rim) 22.2. Hard-fired light brown fabric, 7.5YR 6/2 to 6/4; some dark inclusions; surface slipped, with greenish tinge. About fifty fragments, many joining, mostly from upper parts, giving virtually complete upper profile. Surface generally much worn. Lip not joining but certainly belonging, flaring with vertical concave rim. A flat base may belong. Vertical flattened handles set high on shoulder. Paint fired dark gray.Broad band on inside of lip. Outside of rim painted. Band of outlined solid lozenges on lower neck, with triple line above and single broad band below at turn of shoulder.Top of shoulder reserved;in handle zone, between triple line above and band below, frieze of double axes, filled with double crescents. Banding on body below. The foot, if belonging, is extremely worn. However this amphora is categorized, it would appear to be among the latest pieces in the building itself. Stylistically it would seem to be of local production and perhaps to be placed at the end of MG or early LG. Neither ornamental pattern, outlined leaf or double ax, is common, though the former is found at Knossos in LG (Fortetsa,pattern 12c); for the approach, cf. Levi 1927-1929, fig. 352a. 15 Amphora, neck and upper body

Fig. 3

C9627 (82A/7, 10, 15A, 26). P.H. 25.5, Diam. 31. Medium coarse salmon-pink clay, 7.5YR 7/6, more yellow in parts, 2.5YR 6/6, with many large red-brown inclusions. Fifty-one fragments of neck and upper body. Narrow neck, with short strap handles (parts of both preserved);lip not assuredly preserved-the highest neck fragment seems to have a worn, not a finished, top. Surface worn to very worn, flaking in places. Paint

Z AT

KOMMOS

I99

fired dull brown. Inside of neck reserved, as far as is preserved. Band outside at top of shoulder. Broad band above two narrow bands at waist; on shoulder, pairs of large sets often concentric circles. Band down outside of handle, continuing onto body, and band of glaze around lower handle root. There are further probable body sherds from 81B/67. 16 Amphora, rim, neck, and body

Fig. 4

C10715 (82B/53). H. (neck) 14, Diam. (rim) 14. Fine clay with many dark inclusions; fired various colors, surface generally cream, but with pink patches. Forty-one fragments, many joining. Simply molded vertical rim, with marked convexity. Paint fired orange-brown to brown. Broad band at rim inside; upper part of outside of rim painted. Two bands on neck at level of handle attachment, band at turn of shoulder, and verticals flanking handle attachment. Three bands above painted area on a lower body fragment. 17 Amphora, rim and neck

Fig. 4

C10699 (82A/7, 31). P.H. 12.7, Diam. (rim) 22. Micaceous and rather coarse red-brown clay, 5YR 5/6. Three nonjoining fragments of rim and neck. Rather rounded rim, lightly concave on inside; handle root preserved on one fragment but its placement on neck uncertain. Plain. An imported storage amphora, perhaps from East Greece. 18 Hydria, rim and neck

Fig. 4

C3208 (36B/19,20). P.H. 15.5, Diam. (rim) 18. Nearly fine, buffbrown clay, 7.5YR 8/4, with grayblue inclusions; slipped. Sixteen fragments (enough of the diameter is preserved to ensure that there was only one vertical handle), giving rim and neck profile; see Callaghan and Johnston 2000, no. 260. Plain; the date is LG, while the context is akin to that of 23 below, which is a comparable piece.

Figure 5. Amphora(?) 21 OTHER

LARGE

CLOSED

SHAPES

19 Amphora(?), shoulder

Fig. 4

C9616 (82A/32). P.H. 11.5, Th. (wall) 1.1. Coarse red-brown clay, 7.5YR 7/6, with large gray inclusions; paler buff surface.Two joining fragments of shoulder, probably of an amphora. Slight ridge at top of shoulder. Dark paint, much worn. Band over ridge; parts of two sets of 14 concentric circles preserved on shoulder below. 20

Amphora or hydria, rim and neck

C9756 (83C/66). P.H. 9.6, H. (lip) 2. Buff clay, 7.5YR 7/4, with some dark inclusions; cream surface, perhaps slipped. Two nonjoining fragments of rim and neck. Flaring neck with "Cretan"vertical rim. Very worn. Band near rim inside. Outside of rim and top of neck painted; two lines below, then frieze of hatched leaves. Similar to 14, especially in fabric, but smaller. 21

Amphora(?), shoulder

Fig. 5

C10486 (82B/50). 6.5 x 7.4. Fine pink-buff clay, 7.5YR 7/6, with blue-gray core. Fragment of shoulder of a closed vase. Much worn. Probably from an amphora. Part of a broad band on lower part; ten long verticals above, pendant from a horizontal band ("fringe"ornament).

200

ALAN

v'/

\

W.

JOHNSTON

( 23

24

K

25

27 26 L

1...

t^

--1

.

l

-

I

f

28

37

/''+

33 35

Figure6. BuildingZ. Closedvessels. Scale 1:3

BUILDING

22

Z AT

201

KOMMOS

Amphora or hydria, rim and neck

Fig. 4

C9586 (82A/23,23A, 15). P.H. 7.9., Diam. (rim) 16.2. Medium coarse pale pink clay, 7.5YR 8/4, with many dark inclusions. Four joining (pails 23A and 15) and one nonjoining (23) fragment of lip and neck. Everted rim with flat outer face. Paint dark,but very worn. Rim wholly painted. Part of single horizontal zigzag preserved on an otherwise reserved neck. A conservative form of decoration; the profile helps little in placing the piece, though the fairly sharp outswing of the lip points to MG-LG rather than earlier. 23

Amphora or hydria, rim

Fig. 6

C3244 (36B/21). P.H. 4, Diam. (rim) 20. Fine, low-fired, pink-buff clay, 7.5YR 7/6. Single fragment of rim. Hollow lip offset from vertical neck. Dark paint. Band near rim inside; outside painted, as far as preserved. Probably LG; cf. Coldstream 1972, p. 92, fig. 11, Fl. It was specifically noted that this piece was found at a level immediately above the main dump of pottery in this sector and so can be regarded as potentially later than the main period of use. 24

Amphora or hydria, shoulder

Amphora or hydria, shoulder

26

Amphora or hydria, neck and rim

Fig. 6

C9688 (81B/64, outside Z). P.H. Diam. (rim) 24. Fine pink clay, 14, 5YR 7/4 to 7.6, with light brown slip. Some large yellow-brown inclusions. Three joining fragments of neck and rim. Broad, flat-topped, flaring rim with ridge at top of neck. Extremely worn, traces of burning on rim. Outside of rim and upper part of neck painted (inside wholly worn). Parts of two sets of six pendant concentric semicircles preserved below. Cf. Fortetsa,no. 140, pl. 10, for the profile; PGB to EG. 27

Amphora or hydria, rim, neck, and shoulder

Fig. 6

C9556 (82A/25). Diam. (rim) 14. Powdery semicoarse red-tan clay, 5YR 7/4, with some red and dark inclusions. Eleven fragments, eight joining, of rim, neck, and shoulder. Everted "mushroom"lip. Decoration in red-fired glaze, very worn. Rim wholly painted. On neck, central motif of stacked chevrons between pairs of verticals; lanceolate verticals to either side (not clear in Fig. 6). A band preserved on a small body sherd.

Fig. 6 28

C10481 (82B/52). P.H. 8.2. Fine clay with many inclusions, pink-buff in core, 5YR 7/6, gray-blue near surface. Five joining fragments of shoulder. Part of a set of at least seven concentric semicircles preserved, on a broad band; two narrowerbands and a glazed area below. 25

Part of a set of seven concentric semicircles, with central dot, preserved above a painted band.

Fig. 6

C10478 (82B/54). P.H. 6.4. Semifine pale buff clay, 10YR 8/4, with many inclusions, mostly bluegray. Single fragment of shoulder.

Amphora, lip and upper neck

Fig. 6

C10704 (82A/16). P.H. 4.5, Diam. (rim) 26. Fine pink-buff clay, 5YR 7/6 to 6/6, with some inclusions and pale buff surface.Three joining fragments of lip and upper neck. Horizontal rim to flaring lip, ridged below. Upper part of inside of lip glazed; top surface reserved;outside wholly glazed. An unusually elaborate molding. 29

Amphora(?), body

C10482 (82A/17 and 82B/52). H. of largest fragment 6. Semicoarse rose-pink clay,2.5YR 6/4, with many

202

ALAN

W.

JOHNSTON

Figure 7 (left). Amphora(?) 30 dark inclusions and creamy yellow slip. Two joining and two loose fragments of wall, perhaps of an amphora. Part of a large, crosshatched horizontal zigzag, in worn dull paint, on all sherds. A much simplified version of neck decoration such as KNC, pl. 48, Gl.

cal outer face; attachment for massive vertical handle preserved on a rim fragment. Variouslyworn. Dull brown paint. Broad band on inside of rim; close-set verticals on outer face; neck painted as far as preserved;band down handle.

30 Amphora(?), neck

C10480 (82B/50). P.H. 3.2, Diam. (rim) 18. Fine pinkish buff clay,5YR 7/4. Lightweight rim, flaring as frequently on such pots, but with broad flat top. Partly burnt and very worn. Upper part of inside painted; one set of twelve rather thick bars preserved on top of rim. Outside painted as far as preserved.

Fig. 7

C9582 (82A/21). P.W.11.6. Fine buff clay, 5YR 7/6. Two joining fragments of cylindrical neck, orientation uncertain. Worn. Streaky dark brown paint. Two bands above (or below) a broad net pattern, not fully preserved, presumably at center of neck. 31 Amphora(?), body C7043 (36B/21 and 65A2/23). P.H. 10.7. Fine ware, red-buff, 5YR 6/6. Three joining fragments of belly; see Callaghan and Johnston 2000, no. 261. Fully painted outside except for two reserved horizontal bands decorated with groups of nearly vertical lines. 32 Amphora or hydria, lip, neck, and handle

Fig. 8

C9687 (82B/50 and 52). Diam. (rim) 13. Fine pinkish brown clay, 5YR 7/6, with grayer core. Five nonjoining fragments of lip, neck, and handle. Flaring rim, with verti-

33 Amphora or hydria, rim

34 Amphora or hydria, shoulder, handle root

Fig. 6

Fig. 9

C9613 (82A/30 and 52). Max. p. dim. 10.4. Medium coarse buff clay, 7.5YR 8/4, with large white and black inclusions. Two nonjoining fragments of shoulder and lower handle root. Thickwalled amphora with decoration in dull dark paint, now worn: band on belly, parts of two sets of nine concentric semicircles above, though continued well over the band below; band from handle down onto body. A small rim fragment in 82A/29 may belong.

Figure 8 (right). Amphora or hydria 32

BUILDING

(

Z AT

KOMMOS

203

38

Figure 9 (above,left). Amphora or hydria 34 35 Amphora(?), body Figures 10, 11 (right;below, left). Amphora 38

Fig. 6

C9610 (82A/30). Max. p. dim. 5.1. Fine gray-pink clay,2.5YR 6/2, with redder core. Thick-walled fragment of closed vase, probably from the shoulder. Burnt after breakage.Two bands below a frieze of sets of double circles with fill of hatched leaves. The spirit, if not the letter, of the decoration is similar to that on a LG hydria from Kavousi; Coldstream 1968, pl. 57:m.

37 Amphora(?), rim

Fig. 6

C10484 (82B/50). P.H. 3.7, Diam. (rim) 24. Nearly coarse, orange-brown fabric, 5YR 6/8, gray in core, with much mica and other inclusions. Two joining fragments. Strongly everted rim, lightly molded on outside; handle attachment close below. Traces of dark paint on top of lip. The paint and type of handle point away from this being a cooking pot, as suggested by the clay. Import.

36 Amphora(?), body C10691 (82A/31). Max. p. dim. 7.3. Fine hard-fired purplish brown clay,2.5YR 6/4, blue-gray in core. Fragment of lower shoulder. Extremely worn decoration, a set of standing concentric semicircles above a band.

38 Amphora, body

Figs. 10, 11

C9699 (82A/30 and 50). W. of largest sherd 11.5. Medium coarse, nearly white clay, 10YR 8/2, with darkercore, 5YR 7/4; brown inclusions. Eleven fragments, five joining, from the waist and upper wall. Decoration in very worn paint: band above three lines on waist, supporting sets of nine concentric semicircles on shoulder.

Also worth noting for its size is a rim and neck fragment from 82B/50, from a plain amphora, very worn, with a typical Cretan rim, Diam. (rim) 21, and strap handles 5.4 x 2.3.

I

ALAN

204

W.

JOHNSTON

A

39

42 40

I I e I

4

46 48

45

,,

I

-

1,

It, I I

II

^

',1." \.~~~~~~~~~~~ I

52

,

53

55 56

SMALLER

CLOSED

39 Amphoriskos, body and handles

Fig. 12

C10085 (82A29). / P.H. (6.7 C10085v~~ Diam. (foot) 3.7. Fine pink-red clay, 5YR 7/6. Twelve joining fragments of body and handles. Low disk foot and globular body, faceted. Vertical strap

handles.Plain.

40 Jug, rim and neck

Figure 12. Building Z. Smaller jugs and bowls. Scale 1:3

SHAPES

Fig. 12

C9629 (82A/7 and 13). Diam. (rim) 14. Fine reddish yellow clay, 7.5YR 7/8. Nine nonjoining fragments. Slightly everted, flat-topped rim. Moderately to extremely worn. Streaky red-brown paint. Bars on top of rim (probably in sets rather than continuous). Inside probably had

band near top of lip (not included in the drawing); banding on outside. Lip fragments of very similar jugs come from 82B/50 (slightly thicker lip; assured band, but broader than the possible traces on 40) and 82B/52 (wholly painted inside), with one of body fragments probably from bodyfragmentsprobablyfromone of these from 83C/64 to the east. Hence the inclusion of this more fully preserved example, from a higher level, in this section. This is the shape of jug that must have accompanied the pedestaled kraters (see below). Cf. Fortetsa,no. 928. Oinochoe, rim and body C8051 (63A/48 and 50; 65A2/ 22; 82B/52). P.H. 14.5, Diam. (rim) 41

BUILDING

Figure 13. Oinochoe(?) 47

Z AT

KOMMOS

205

ca. 13. Fine orange-buff clay, 5YR 7/6. Mainly joining fragments from rim to lower belly. Callaghan and Johnston 2000, no. 54. Slightly concave neck with flaring rim, and strap handle joining the rim. Very worn, once dark, paint. The neck and rim are decorated with horizontal bands enclosing a horizontal squiggle. The shoulder bears a hatched "cloud"pattern (Fortetsa, pattern lln). Below is a broad band flanked by horizontal stripes, and the lower belly is decorated with a running wave pattern. Horizontal bars on handle. The fragment from 82B/52 has clay and decoration identical to those from trench 63A, even if it does not join; however, the join between 63A and 65A2 is assured, an important, if isolated, link between Building Z and the temple area.The style of 41 seems hardly later than EG, which primafacie suggests that the jug originally was placed in the sanctuary area and the sherds were later transferredto Z. Cf. KNC, T107.144, but also the MG T283.14.

C2987 (36B/5). Rest. H. 14.5, Diam. (rim) 17. Pale buff surface, more pink in core, 5YR 7/6; many small black inclusions, a few larger white ones. Thirteen fragments, giving rim and lower body. The shape is similar to that of a giant cup. Dipped in black paint at least twice on the upper body.

42

46 Juglet

Oinochoe, base and body

Fig. 12

C9681 (82B/50 and 52). P.H. 8.3, Diam. (foot) 6. Fine hard graybrown clay, 7.5YR 7/2 to 7/4. Two sets of two joining fragments of base, wall, and shoulder. Small oinochoe with flat base and full body; upper parts lost (though fragments in pail 52 may belong). Worn. Band at base of wall and another at waist; perhaps originally three sets of five concentric semicircles on shoulder. 43

Oinochoe, shoulder

C9686 (82B/50). Max. p. dim. 6.6. Fine pink clay, 7.5YR 7/6, with blue-gray core. Five joining and one other sherd of shoulder with nipple. Very heavily worn in part. Decoration in once dark paint. Band around nipple; to its left a vertical band of multiple zigzag; below, a worn area including curving lines. The same pail contained several standard amphora or hydria rims and

a good range of fragments of larger vases, all with a blue core to the biscuit; 43 is the sole catalogued example. 44

Oinochoe(?), strap handle

C10102 (65A2/22). P.L. 4.5, W. 3.8. Fine buff clay, 10YR 7/4. Upper part of strap handle. Very flat handle rising slightly above rim; size suggests it is not from a drinking vessel. Decoration in dull, dark paint; inside of neck painted, as far as preserved; sides of handles painted; nine bars preserved on outside, the lowest broader than the rest. 45 Jug or cup, rim and body

Fig. 12

Fig. 12

C9553 (82A/16). P.H. 4, Diam. (foot) 3. Fine clay, fired red-brown, 5YR 7/6, to pale buff with a few dark brown and one large white inclusion. Lacking neck, lip, and handle. Round bodied, with flat base; scar of handle attachment at broadest part of belly. Worn and chipped. Plain. 47

Oinochoe(?), body sherd

Fig. 13

C9768 (65A2/23). 3.6 x 2.6. Fine buff clay, 7.5YR 8/4 in core. Single fragment of wall of a closed vase. Inside reserved;outside, part of a band of horizontal, multiple zigzag above a triple band; below, part of a single billet and a further band. Very similar to 54, though the triple bands are thicker than there. Lack of paint on the inside disassociates these (and 50) from the kraterswith similar zigzag pattern.

206

48

ALAN

Oinochoe, neck/handle

Fig. 12

C9558 (82A/26). P.H. 4.5, p. Diam. 9.5. Fine buff clay, 5YR 7/4, with red-brown surface. Single fragment of neck and handle. Full upper body, narrow neck, and round, flaring rim; slight ridge on shoulder at base of neck. Strap handle. Very worn, but the outside once wholly, if thinly, painted. 49

Small jug, shoulder

Fig. 14

C10039 (82A/25). Max. p. dim. 4.2, Th. 0.2-0.5. Fine pale tan clay, hard fired, 5YR 7/4. Two joining fragments. Rounded body with turn of neck just preserved. Dull dark brown paint. Lower shoulder glazed; pendant triangles between double lines on upper part; band of diagonals at base of neck. For the juxtaposition of motifs, cf. KNC, T292.139 and for broad triangles, Tomb D, 13 (PGB). 50

Oinochoe(?), shoulder

Fig. 15

C9615 (82A/33). Max. p. dim. 4.2. Fine buff clay, 10YR 7/3, more pink in core. Single fragment. Burnt and much worn. Quadruple horizontal zigzag between bands; further paint below, reserved above. 51 Jug(?), handle/body

Fig. 16

C10086 (82A/16).W. 3.1, Th. 0.9. Fine pale buff-orange clay, 7.5YR 7/6. Single fragment of strap handle and wall. Much worn. Dull red-brown paint; bars across the handle. 52 Aryballos or oinochoe, shoulder

Fig. 12

C10084 (82A/31). Max. p. dim. 5.2. Fine pinkish buff clay, 10YR 8/3. Two joining fragments. Much worn. Decoration in dull brown: three lines below; above, partly preserved patterned panel, with solid, outlined triangle flanked by elaborate and asymmetrical ladder pattern. The decoration more complex than on a LPG lekythos, KNC, T218.67 (fig. 130), but from a similar pedigree.

53

W.

JOHNSTON

Short-necked jug, rim

Fig. 12

C9578 (82A/16A). P.H. 3.4, Diam. (rim) 12. Fine buff clay, 7.5 YR 8/4. Single fragment of a plain, flaring rim. Worn dark paint; inside reserved save for band at top; outside, a frieze of linked circles framed by two bands above and a band and a glazed area below. The circles are painted partly individually, partly as intertwined wavy lines. Common enough in LG, the motif has MG attestations. Stampolidis 1996, p. 48, fig. 90 and p. 64, fig. 122 (with incomplete reference to Tsipopoulou's doctoral thesis, pp. 203-204, non vidi; to be published in the TAHA monograph series; see Tsipopoulou, in press); Coldstream 1968, pl. 57:b. 54

Oinochoe, shoulder

Fig. 17

C9656 (82A/26). P.H. 3.3. Fine buff clay, 5YR 7/3. Single fragment. Decoration in worn brown paint. Two horizontal zigzags preserved at top; three lines below, from the lowest of which two thick verticals are pendant, part of a lower frieze. Possibly from the same jug as 50, but it is not burnt and the lines are thinner. 55 Trefoil oinochoe, rim, neck, and shoulder

Figure 14. Small jug 49

Figure 15. Oinochoe(?) 50

Fig. 12

C10078 (82A/29). P.H. 3.1. Fine orange-buff clay, 5YR 7/6. Single fragment, handle lost. Worn; painted on inside of mouth and on all of outside with dull dark

Figure 16. Jug(?) 51

glaze. 56

Olpe or oinochoe, body and base

Fig. 12

C10079 (82A/30). P.H. 9.5, Diam. (foot) 5. Fine orange-tan clay, 5YR 5/6. Five joining fragments of simple, flat base and wall. Worn, dull, red-brown paint. Underside, and band at base of wall, reserved;three broad bands of paint preserved above.

Figure 17. Oinochoe 54

BUILDING

Z AT

KOMMOS

207

I,

59

_'A

.

-~~~~~~~~~~~~~~~~~~~~~~~~~~~~---

62

60

`1

j

71 Figure18. BuildingZ. Kraters. Scale 1:3

t

76 KRATERS STIRRUP-HANDLED

AND

RELATED

57

Krater,pedestal C7046 (36B/2 and 5; 65A1/12

and 65A2/22). P.H. ca. 6, Diam. at base ca. 22. Fine ware red-tan, 2.5YR 6/6. Eight fragments, mending to five; see Callaghan and Johnston 2000, no. 263. Fully painted outside in the red-chestnut glaze typical of this workshop. A fragment of rim akin to that of 79 may belong. The shape and horizontal ribbing on the exterior are features borrowed from other parts of the Greek world, which at Knossos appear confined to the EG and MG period. Cf. Coldstream 1972, p. 79 and KNC, p. 375, arguing an Attic derivation. There are Ionian models too; see SamosV, pls. 7 and 14 (800750 B.C.). The number of broadly Cycladic or Ionian imports at Kommos in this period bids us

beware of using Atticizing Knossos as the sole possible source. 58

Krater,pedestal

C7048 (36B/5). P.H. 5.3, Diam. at top ca. 12. Fine ware, orange-buff, 7.5YR 6/6. Three fragments of stem, with part of floor of bowl; see Callaghan and Johnston 2000, no. 265. Fully painted on exterior and on interior of the bowl. Ribbed.

Relatedto 57. 59

Krater,pedestal

Fig. 18

C7813 (65A1/12). P.H. 3, Diam. (foot) 20. Light red clay with small silvery micaceous inclusions. Fragment of foot. Glaze fired redbrown. Painted on outside, save for a reserved band. Certainly does not belong to 76 or 80.

208

ALAN

W.

JOHNSTON

Of/ 72

-.1I~

74

64

/

70 Figure 19. Building Z. Kraters. Scales 1:6 (64, 72) and 1:4 (70, 74)

Figure 20 (left). Stirrup-krater 65 Figure 21 (right). Krater 66

BUILDING

60

Z AT

KOMMOS

Krater(s?),pedestal

209

Fig. 18

C9580 (82A/17). Diam. (foot) ca. 20. Fine buff clay, 5YR 7/6. Three nonjoining fragments of edge of foot, stem, and top of pedestal of probably two kraters (the restored drawing highlights the difficulty of fitting the stem fragment at the angle required by the other two). Narrow resting surface. Glaze fired red-brown, worn. Wholly painted outside and on bowl inside. Not same vessel(s) as 57 or 58, but possibly belongs with 113. 61

Krater,pedestal

C9722 (82B/50). Diam. (foot) ca. 18. Fine light red clay, 5YR 7/6. Fragment of foot. Narrow resting surface;first rib of stem just preserved above. Paint fired dark brown. Three reserved bands on top of foot. Close in fabric to 58. 62

Krater,pedestal

Fig. 18

C10716 (82B/52). P.H. 2.9, Diam. (foot) ca. 19. Fine micaceous pinkish beige clay, 7.5YR 7/6. Fragment of edge of pedestal. No trace of decoration preserved on very worn surface. Fabric similar to 59, but not the same piece. The clay here does suggest a non-Cretan origin. 63

Krater,pedestal

C9748 (81B/67). Max. p. dim. 8. Fine light red clay, 10YR 6/6. Single fragment, part of base and lowest rib of stem. Extremely worn; glaze fired red-brown. Fabric similar to 60, 79, and 113, but thicker. 64

Krater,rim and body

Fig. 19

C9766 (82A/10 and 31; 82B/50, 52, 53). Diam. (rim) ca. 40. Fine light purplish red clay, 5YR 7/4 in core. One fragment of rim, ten of body (five joining), and four joining of handle. Some lower body fragments and one rather flat sherd of a pedestal base may or may not belong (one of these from pail 31, which, stratigraphically speaking, lies both horizontally and vertically between

the rest). Much worn. Inside wholly painted. Sets of bars on top of rim (no set fully preserved). Four bands on outside of lip. Triple vertical zigzag between sets of four verticals at edge of panel. To center, part of a meander pattern above a strip of triple horizontal zigzags, with triple line between. Lower part of panel extremely worn. Bars across strap handle. Though the firing is rather different from that of 65, it is plausible that the two, together with 66, come from the same krater.The outside of the rim would have been fully glazed between the "warts"and decorated with four bands on the rest of the circumference; a "wart"would then have been placed just beyond the break to the right of the rim fragment of 64. The fabric is the same as that of skyphoi 85 and 86. 65

Stirrup-krater,rim

Fig. 20

C9762 (65A2/22). P.H. 7.5. Fine buff clay, 5YR 7/4. Two joining fragments of rim and shoulder.Thick, everted rim, 2.4 high. Generally worn. A "wart"preserved on rim to right, the face decorated with an eightpointed star. Paint fired generally medium to dark red-brown. Inside once painted, now very worn. Top of lip reserved, with three bars preserved. On outside of lip, to left of wart, four horizontals; fully painted to right (presumably near the handle join). On shoulder, five verticals to right, triple vertical zigzag in center, and a single vertical line preserved to left. Cf. 64 above. 66

Krater,rim and handle

Fig. 21

C9764 (65A2/22 and 23). Diam. (rim) ca. 40, Diam. (handle) 2.2, W. (strap element) 3.8. Fine red-buff clay, 5YR 7/4. Five joining fragments of handle and three of rim of perhaps one krater.Thick everted rim, 2.2 high. Paint unevenly fired. Inside painted; top of rim reserved, with one set of six bars preserved. Outside of rim and shoulder painted, though very worn. Outside of loop handle painted; on the strap X between sets of six horizontals. Associable with 64, 65, and 67.

ALAN

2IO

W.

JOHNSTON

Figure 22. Krater 67 67

Krater,rim and handle

Fig. 22

C9630 (82A/10; 82B/50, 52). P.W. 13. Fine red-buff clay, 5YR 7/4 in core. Eight fragments, severaljoining. Varying wear on the fragments. Thick rim, 2.4 high and 1.4 thick; one "wart"preserved, presumably flanking a handle attachment. Round, horizontal handle, rising high and once joined to rim by a strap. Rim painted, save the top, on which sets of bars are painted (five in the one fully preserved set), and the face of the "wart,"on which is an eightpointed star. Similar to, but not from the same piece, as 64-66. 68

Krater,pedestal and bowl

C9765 (65A2/23). P.H. 3.7, Th. (wall) 2. Fine buff clay, 10YR 7/3 in core, 5YR 7/6 on surface. Fragment of pedestal with part of floor of bowl. Extremely heavily worn; outside may have been ribbed and painted; some paint left on floor of bowl. 69

very dark:two stripes preserved on strap;thin bars between bands, on round handle. 70

Krater,rim

Fig. 19

C10087 (82A/26). P.H. 6 .5l Diam. (rim) 34. Fine light brown clay, 7.5YR 7/4. Four joining fragrrients of rim and shoulder. Plain verticoal rim above sloping shoulder. Dull, sstreaky, dark brown paint. Inside whol ly painted. Top of lip reserved, writhone set of five bars preserved. Out,side glazed down to reserved band at lowest preserved point. 71

Krater,rim

Fig. 18

C7918 (65A2/23). P.H. 5 '7) Diam. (rim) 23.3. Pale red cla' yyI 2.5YR 6/4, with some mica. Two joining rim and shoulder fragrnents. Vertical offset rim, flat-toppec fired streaky medium brown. Sx. ix bars bars preserved on top of reserved lilp.Four irregularlyset reserved lines orn outside. Parts of hatched mearider hooks on shoulder below.

Krater,stirrup handle

C9767 (65A2/22). Diam. (handle) 1.8. Fine orange-buff clay, 7.5YR 6/6. Single fragment. Handle from a comparatively small krater; most of the round handle and part of the strap preserved. Paint fired

72

Krater

Fig. 19

C9700 (82B/50 and 52, and ldsed fragments from 65A2/22 not i included in drawing). P.H. 10.2, Diam. (rim) 38, Diam. (foot) 12.8, W. (stro yy10YR handles) 3.4. Fine light red cLaty lOYR

The very fragmentary nature of this material makes any assessment of the total number of pieces represented hazardous. There are at least six feet of different diameter.

6/6. Eight fragments, including two sets of three, of rim, upper wall, and handle, and thirteen fragments of lower wall and foot. Open-bowled kraterwith sharply everted rim and wide strap handles (but not stirruphandled). Slightly flaring ring foot, 1.3 high. One fragment somewhat burnt. Paint fired dark chocolate brown on the outside, orange-red in. Inside painted, save for top of rim. Top and bottom of rim outside reserved, also the area under the handle; wavy line between bands on outside of handle. 73

Krater,body

C9680 (82B/50). Max. p. dim. 7.5. Fine light red clay, 10YR 6/6. Two joining fragments of body, with handle attachment. Much worn, especially the edges. Dark brown paint; reserved in area behind handle, painted below. Almost certainly from same krater as 72. 74

Krater,rim and shoulder Fig. 19

C10631 (65A2/22). P.H. 3.3, Diam. (rim) 34. Fine buff-brown clay, 5YR 6/4. Two joining fragments of rim and shoulder. Painted, as far as preserved, save for top of rim, which is reserved with sets of six bars. Probably from a stirrup-handled krater.

BUILDING

Z AT

Figure 23 (left). Krater 75

OTHER

TYPES

Figure 24 (right). Krater(?) 77

75

211

KOMMOS

Krater, rim and

Figs. 23, 33

body C9723 (82A/10 and 82B/50). P.H. 15.5, Diam. (rim) 40, Th. 1.4. Hard-fired medium coarse buff-pink clay, 5YR 7/3, with some large dark inclusions and creamy slip. Five joining and four further fragments of rim and upper wall. Thick-walled kraterwith ledge rim and ridge at top of wall. Extremely worn. Broad band near rim inside. Traces of paint on outside of rim. Set of at least eleven concentric circles on wall, with central cross. 76

Krater,rim and body

Fig. 18

C8054 (65A2/24; the pail spans the broad depth range 3.56 to 4.46 m). P.H. 11.5, Diam. (rim) ca. 35. Near fine, light red clay, 2.5YR 6/4, with large dark red and white inclusions. Three joining fragments of rim and wall. Flat-topped flaring rim; slight molding outside at top of wall. Rim wholly painted, save for top. On upper wall, part of central strip of decoration preserved, flanked on right by triple vertical line and part of a hatched feature. Band of crosshatched hour-glass ornament above one of wholly painted butterfly set in doubly outlined rectangle; further hourglass band below. For hatched hourglass patterns in central strip, see KNC, fig. 57:5, and more chaotically, fig. 59:1. Probably no later than PGB.

77

Krater(?),body

Fig. 24

C10104 (65A2/22). Max. p. dim. 9.2, Th. 1.6. Semicoarse yellowbuff clay, 7.5YR 8/4, with many inclusions. Single fragment of wall of large vase, probably a krater.Rather worn paint: part of a set of concentric circles preserved. 78

Krater(?),body

C10106 (65A2/22). Max. p. dim. 5.9. Fine buff clay, 5YR 7/4, with paler buff surface. Single fragment of shoulder, with handle scar. Slightly curving rather thin-walled fragment of a large vase. Large handle scar,with two sections of a group of concentric circles in fine brown lines beside it. 79 Belly-handled jar C7040 (36B/21; a foot fragment from 65A2/22 may belong). P.H. ca. 14, Diam. (rim) ca. 15. Fine ware, orange-buff, 5YR 6/8. Twelve fragments, mending to eight, of rim and wall, with much of one horizontal handle. See Callaghan and Johnston 2000, no. 262. The exterior is painted except for two reserved horizontals on the upper shoulder and a low frieze between the handles, which is decorated with close-set verticals. Bars, perhaps continuous, on top of lip. The fabric is close to that of several of the stirrup-kraters.

ALAN

2I2

W.

JOHNSTON

SKYPHOI

80

Skyphos

Fig. 25

C9729 (82B/50,52,53). H. 8.8, Diam. (rim) 14. Fine pale red clay, <.< -inn \ ^-n * r?~~' lighter than 10R 6/6. Thirteen fragments from all parts of skyphos, with joins between pails. Low, concave rim, high shoulder, and straight lower profile; flat base. Paint variously fired, dark brown to orange-red. Wholly painted, save for underside and narrow band at top of rim inside, which has sets of probably five verticals on it. A very worn fragment of rim from 82B/54 may belong, but is difficult to assess. The angular shape is not unparalleled in MG and LG. Possibly an import, though not Attic. 81

Skyphos, rim

Fig. 25

(82A /27 P.H. PH. 2.6, C10702 (82A/27). ). CD.am. 2i. brown Diam. (rim) ca. 14. Fine light clay,5YR 5/8r. Single frag-ment fragmen ofA /8. Sing-le >Y 5 Clay, ' i i.. i concave lip. Streaky. rim, tall, slightly *. . . .,1' red-brown rpaint. Painted inside, save /,,, < i. ,~~'there i for band below lip; three bands on .~~~~ -. . 1 -r outside of lip, with further painted belo area below. ~Crete ~~area 82 Skyp sr5rim Skyphos, J82

Fig. 25 F1968,

C9611 (82A/30). P.H. 3.6, Diam. (rim) ca. 12. Fine light brown clay, 7.5YR 7/4, with fine flecks of golden mica. Single fragment of short and straight offset rim and wall. Very worn, especially inside. Probably once had reserved band inside near lip. Reserved band on outside of rim. Below, part of panel with its boundof small dots ing line to right: 0? row _. .~~~6.4, above three lines.iPainted,,below, as far a pred far as preserved.r Import. Impo~~~rt. ~Low, .handle. 83 rim~ Skyphos~ 8 83 Skyphos, rimr SyhsriPainted C3243 (36B/21). P.H. 4.4, Diam. (rim) ca. 14. Fine ware, light red-brown, 7.5YR 6/4. Rim and shoulder fragment; see Callaghan and Johnston 2000, no. 256. Fully painted except for a reserved band at rim inside and out. A MG form, cf. Coldstream 1972, p. 82, fig. 7, D28. Several other

similar rim fragments were rec from this and adjacent pails. 844

handle Skyphos, rim and and handle Skyphos, rim

C7041 (36B/21). P.H. 4.2 Diam. (rim) 21. Fine ware, ora buff, 5YR 7/6. Nine fragments shoulder, and both handles of; skyphos; see Callaghan and Jo] 2000, no. 258. Paint variously i red to dark brown. Painted ins except for a reserved band nea] rim; lip painted outside, with F of the decoration of the handle preserved below, a few vertical visible. The type is probably as 8c below. The rim on these pieces relatively short and often stron offset, cf. Fortetsa,no. 494, pl. form of decoration is difficult to parallel;cross-and-vertical a . . foud in iry z nation is found in * subsidiary ^ ji and Argive Geometric, sta: tri nd the o ski .ve p vertical pattern1S foundon slq tere (orn 19 po l 5 Ck C (Courbmn1966, C , o pi.56, . oup7 bt ahn aand C59; p. 378, note but n< 3, te im An. b btee the minimal lin links between ArE on pp. 530-536); Samos vides LG examples (Coldstrea pl. 64:a; SamosV, p. 92, r These pieces would seem to be work of a single, perhaps local painter, and therefore it is wor noting the minor variations in and detail of decoration in the material.

pink-buff clay, 5YR 7/4. Fragment of rim and body. Low, everted rim and full body. Paint fired dark brown inside, red outside, much worn. Inside glazed, save for reserved band at top. Outside, rim painted, panel in handle zone; parts of at least six verticals discernible, the rightmost, incompletely preserved, ending below as a diagonal. Possibly from same piece as 85. 87

Skyphos, rim

Fig. 25

C9577 (82A/15, 16A, 33; 82B/50). P.H. 8.1, Diam. (rim) 20.2. Fine red-brown clay, 7.5YR 7/6. Fourteen, mostly joining, fragments of lip and body wall. A foot fragment from 82B/50, reserved underneath, with slightly raised disk foot, may belong. Low, everted rim. Paint variously fired, dark brown to orange, inside and out. Worn to very worn. Reserved band on inside of rim with sets of six verticals on it. Two reserved lines on outside of lip. Panel in handle zone decorated with alternating X and sets of five or six verticals. Reserved area below handles. Reserved band below handle zone, not continuous under handles. 88

Skyphos, rim and body

Fig. 25

C10040 (82A/25). P.H. 8.3, Diam. (rim) 18. Fine clay,variously fired, pale orange-brown to orangered, 5YR 7/8. Seven fragments, four joining, of rim and body wall. Low, lightly offset lip with full body; thick walls. Glaze fired red-brown to black. Inside painted, save for band near top of rim, with traces of vertical bars at one point. Outside, reserved band on rim. Frieze in handle zone with alternating X and sets of six verticals; two bands below.

Skyphos, rim and handle C9739 (81B/66 east of Z 6. .(im 156 Faine pi Diam. 15.6. r me pnr re ciay, (rim) r6/4.Fo. r joir 10YR 6/4. Fourjoninr clay, ' r 1 and j of wall, han' fragments lip, everted rim and large, roi Decoration fired dark 1 . . ._ inside, save for reserve at rim with (continuous?) friez 89 Skyphos(?), foot verticals. Outside, rim painted Fig. 25 in handle zone with eight vertiiLtali C10703 (82A/15, 16, 16A). and part of an X; wall reserved under Iunder P.H. 3.2, Diam. (foot) 7.5. Fine handle. 85

pale-brownclay,7.5YR 6/6. Five

86

Fig. 25

Skyphos, rim C 1 (81B/65, outside Z). PH. 4.6, Diam. (rim) 17.8. Fii ne

joining fragments of lower parts of a skyphos? Edges and surfaces worn to very worn. Low ring foot. Wholly glazed inside and out.

Z AT

BUILDING

KOMMOS

213

,

L~~~~~~~~~~~~~~~~~ 85 80 I

v=-

I

.

81

82

82

86

I

I.

^

89 87

I 92

88

2 93 91 Figure25. BuildingZ. Skyphoi. Scale1:3

90

ALAN

214

The type of foot suggests a date before LG. The shape could perhaps be a pyxis; cf. KNC, T292.77. 90

Skyphos(?), body

Skyphos, rim

JOHNSTON

dark paint; painted inside and outside save for a narrow band near lip. Probably an import. A similar rim fragment comes from 82A/32.

Fig. 25

C10696 (82A/10). P.H. 6. Fine beige clay, 7.5YR 7/6. Fragment of wall of a skyphos?Wall of deepbowled open vase, with turn to lip just preserved. Inside reserved, save for band at top; outside painted down to lower wall. The context dates to the 8th century,while the shape is that of a PG bell-skyphos; the decorative scheme is not, however, found on other bell-skyphoi. 91

W.

Fig. 25

C10485 (82B/50). P.H. 3.9, Diam. (rim) 10. Fine pink-buff clay, 10YR 7/6. Fragment of rim and bowl of a skyphos. Low, everted rim. Very

92

Skyphos, rim

Fig. 25

C10706 (81B/67). P.H. 2.4, Diam. (rim) ca. 13. Fine pale beige clay,5YR 7/4. Single fragment of rim and shoulder. Painted inside, save for narrow band near lip. Two reserved bands on lip outside, rest glazed.

93 Skyphos,rim Fig. 25 C10698 (82B/50). P.H.2.3, Diam. (rim) 13.2, Th. (wall) 0.25. Fine pink-buff clay, 5YR 7/4. Two nonjoining fragments of rim of a skyphos (rather than cup). Finewalled, with low, concave rim, offset from the shoulder by a groove. Probably once wholly painted.7

KANTHAROI

94

Kantharos

Fig. 26

C10041 (82A/25). H. with handles ca. 12; Diam. (foot) 5.2. Fine light buff-brown clay, 5YR 7/6. Thirteen fragments, a few joining, of foot, body, and handles. The preserved handle fragments cannot easily be accommodated to a single handle; the restored rim remains conjectural.Flat foot. Extremely worn surface;traces of paint remain on inside, underneath the foot, and on the outside of the handles; probably once wholly glazed. 95

Kantharos

C7039 (36B/21). P.H. 7.3, Diam. (rim) ca. 13. Fine ware, light buff-brown (7.5YR 7/6). Twenty

small sherds from rim to belly, with parts of both handles; see Callaghan and Johnston 2000, no. 259. Unpainted. The S profile and thick, narrow handle are early features in cups, where the type survives into the early LG period. Cf. Coldstream 1972, p. 83, fig. 8, E3.8 96

Kantharos(?)

Fig. 26

C10705 (82A/10). Diam. (foot) 7.4, W. (handle) 2.8. Fine micaceous red-tan clay, 2.5YR 6/8. Eight fragments, mending to six, of an open vase with at least one vertical strap handle; low ring foot and flat floor. Underside reserved, rest streakily painted red-brown. Import.

CUPS

There is considerablevariety in the treatment of lip and bowl of the many black-glazed cups from the building; however,the fabricis normally rather heavy and the foot almost always flat or slightly concave (e.g., from 82A/ 48); one foot, in 82A/30, whose material otherwise is solidly MG, has a slight groove around the outer part of the underside, while others have a slightly raised disk foot (e.g., from 81B/65). High and low rims appear in the same pails (e.g., 82A/50 and 52), the latter far more common, though

7. Most pails have isolated skyphos fragments:from 82B/50, fifteen feet from cups or skyphoi,twelve painted, three reservedunderneath;from 82A/ 16-16A, joining fragmentsof a darkglazed skyphosbase with ring foot; from 82A/26, severalrims and a ring foot; and from 82A/32, a foot similarto that from 82A/26 (undersideand lower wall reserved)and a rim similarto 91. 8. From 82A/25 there is also a rim and wall fragmentof a thin-walled, straight-sided,painted kantharos, lightly ribbedhorizontally.

Z AT

BUILDING

-

KOMMOS

2I5

z-r II ,

\ \--

\\

II I t ,I

I) 99 - --~~1;----I

-

94

98

102 96

101 100 Figure26. BuildingZ. Kantharoi and cups. Scale 1:3

three vessels in 82A/16 have a high rim. The handle tends to be attached to the top of the rim, neither clearly inside nor outside. The examples given below are among the better-preserved pieces. Comparable is the range of shapes found in room G at Phaistos (Rocchetti 1974-1975, pp. 261262). 97

Cup, rim

C7045 (36B/2). P.H. 3, Diam. (rim) ca. 12. Fine ware, 7.5YR 7/6. Profile rim to shoulder; see Callaghan and Johnston 2000, no. 257. Fully painted. MG. For the type, cf. Fortetsa, no. 1032, pl. 62; Coldstream 1972, p. 95, no. D105.

98 Cup Fig. 26 C9721 (82B/50 and53). H. 9,

Diam. (foot) 4.9. Fine pinkish brown clay, 5YR 7/3 to 7/4. Twenty-one fragments, many joining. Tall, slightly concave lip; narrow body and flat base, with pronounced pimple at center inside. Handle lost. Very worn outside. Once wholly painted. 99

Cup

Fig. 26

C7919 (65A2/23). P.H. 3.7, Diam. (rim) ca. 12. Fine orange-buff clay, 5YR 7/6. Two joining fragments

2I6

ALAN

of rim and shoulder. Low slightly concave lip; three slight grooves at top of shoulder; rounded body, with fairly thick walls. Once wholly painted. 100 Black-glazed cup

Fig. 2i6

C9581 (82A/17). P.H. 6.8, Diam. (rim) 13.5, W. (handle) 1.6. Fine buff clay, 5YR 8/4, rather porous. Fifteen fragments, some joining, of rim, wall, and handle. Fairly straight and low offset rim; flattened handle. Paint orangebrown, much worn. Once wholly painted.

OTHER

W.

101 Black-glazed cup

JOHNSTON

Fig. 26

C10717 (82A/25). P.H. 9, Diam. (rim) ca. 15.5. Fine pale, greenish buff clay,2.5YR 7/2; soft fabric, perhaps affected by burial environment. Seven fragments, mending to two, of rim, handle, and body. Nearly vertical lip, simply offset from shoulder. Very worn, but once wholly painted. 102 Cup, rim

Fig. 26

C10483 (82B/54). P.H. 5.2, Diam. (rim) 10. Fine, rather porous, pale buff clay, 10YR 8/3. Fragment of rim and wall. Plain everted rim and rounded bowl. Dipped in dark paint.

SHAPES

103 Uncertain shape

Fig. 28

C10710 (82A/26 and 27). P.H. 3.7, Diam. 20. Fine buff-brown clay, 7.5YR 6/6. Uncertain shape, perhaps to be turned round, as a lid; four worn fragments, three joining. Triple ribbing on outside of rim; wall turns out a little at lowest-preserved point. Dark paint. Inside plain; band over outside of rim, reserved area at break. 104 Uncertain shape

Fig. 27

C9679 (82B/50).Max.p. dim. 7.2. Fine pink clay, 5YR 6/4. Single fragment from rather elaborate piece. Painted outside in mottled chocolatered glaze. The piece suggests an amphora neck, but a swelling top left is not normally found near handle attachments, and there are parts of

two holes preserved, top left and right, which run diagonally or obliquely through the fabric. 105 Chytra

Fig. 28

C9554 (82A/16). H. 10.3, Diam. (rim) 12, Diam. (foot) 4.4. Fine buff clay, 5YR 5/3, with dark gray surface.Twenty-six fragments of rim, body, handle, and base of a small chytra. Delicate form, with offset rim; flat base. The piece is unusual; though the clay is of fine consistency, with little temper, the surface is porous and unpolished. If this had been a black-glazed piece, it might have been dated late in the 8th century.

Extremely worn fragments of a lid, or possibly kalathos (C11248), come from 82A/26. COOKING

WARE

106 Cooking pot, rim

Fig.. 28

C9698 (82B/52 and 53).

P.H. 5.8. Coarselight brownclay, 7.5YR 6/4, with wide variety of inclusions. Four nonjoining fragments of rim and wall. Deep bowl with simple everted rim; part of a

lug handle, close to rim, preserved on one fragment. Rim warped. A piece of similar fabric from

a lowerlevel (82B/55) has part

of a small, round, vertical handle at the same point as the lug here.

Figure 27. Uncertain shape 104

BUILDING

Z AT

KOMMOS

2I7

103

105

1

j

___-

I

I

106

,,

107

114 111 115

I

118

116

Figure 28. Various shapes from Building Z and later material. Scale 1:3

119

I 120

2I8

ALAN

107 Cooking pot, rim

Fig., 28

C10479 (82B/52). H. of largest fragment 5.4, Diam. (rim) ca. 20. Coarse ware, full red-brown in core, 5YR 5/6, with a little mica. Two nonjoining fragments of rim

W.

JOHNSTON

of a shallow cooking pot. Ledge rim, slightly concave with light molding below; the start of the outturn of a spout preserved on one fragment.

In 82B/50 there was much cooking ware of various shapes but no complete profile. From 82A/16A came one leg and much of the lower part of a tripod cooking pot with a flat floor. Otherwise, such ware was very sporadic and fragmentary. Fragments of a pithos with herringbone incised decoration on at least one raised band came from several pails, 82A/26, 27,

32, 52, and 53.

UPPER

LEVELS

The levels above ca. 4.30 m show consistent evidence of later dating, after the solidly MG to earliest LG terminus ante quem of the material considered so far. Later material is particularly evident at the east end of the gallery at this level. Added separately to the catalogue are a few pieces of interest from the rather scrappy surface levels above ca. 4.60 m, where burning and limpets, plus cooking-ware sherds, reflect regular 7th-century open-air activity. The highest-preserved point of the walls of Z lies at 4.86 m; this level is fairly uniform, as if it were the product of intentional leveling. The pails of relevance are listed in Table 1. I list the material below by section from east to west, and from lower levels upward. 108 Transport amphora, handle

Fig. 29

C9697 (83C/64). P.L. 12.7, Diam. at break 3.2. Slightly micaceous, rather coarse, clay,gray in core, rose-pink in "halo"around core (10YR 6/4) and orange on surface (5YR 7/6). Five fragments of handle, virtually round in section; lower part of handle, near shoulder join, is ridged outside. Surface worn. The shape is typically Lesbian, and this piece seems to be a very early example of a red-fired variety of Lesbian amphora, though the clay has less mica than one would expect. Dupont (1998, pp. 160-161) lists examples of such early material.The handle does not seem to accord with the "Phocaean"variety of such jars as published by Ozygit 1994, pp. 88-90. The pail concerned is from a level which, a little to the west, in and immediately outside Z, is below that of the transition from pure MG to

later material;this piece is stratigraphicallyintriguing. 109 Amphora, rim/neck

Fig. 30

C9671 (81B/61). Diam (rim) ca. 15. Fine brownish pink clay, 5YR 7/4, slipped creamy white. Two joining fragments of neck and rim of a thick-walled amphora; offset at base of lip; rest lost. Inside reserved, as far as preserved. Decoration in dark paint: rim and upper part of neck painted; frieze of linked Xs; painted below.

Figure 29. Transport amphora 108

110 Hydria(?), handle, body, and foot C9651 (81B/50D, 56A, 61; 83C/59 and 61). Max. p. dim. 35; Diam. (foot) 12.3. Micaceous graybrown to pink clay, 7.5YR 6/2 to 5YR 7/6. Gray surface. Over eighty fragments of foot, body, and one handle probably of a hydria. Closed vase with ring foot and full body; low

Figure 30. Amphora 109

BUILDING

Z AT

KOMMOS

2I9

and thin floor. Root of a horizontal handle on one fragment. Surface generally worn. Variegated firing makes definition of decoration difficult, compounded by uncertainties whether there are one or two pots here. Band on outside of foot and on lower wall; three bands below waist; parts of a sinuous band preserved in zone above, perhaps of "mustache"or "scroll"shape. Perhaps East Greek. The stratigraphy does not allow a close dating. 111 Amphora(?), neck

Fig. 28

C9650 (81B/50E). P.H. 9.7, Diam. (neck) ca. 13. Semicoarse buffpink clay, 5YR 7/4, with redder core; gray and red inclusions. Four joining fragments of neck. Broad black bands partly preserved at top and bottom, with broad vertical band connecting them. Added white decoration on the glaze: concentric circles between pairs of lines on lower band, stacked chevrons between pairs of lines on vertical band, and two lines preserved on upper band. LG or Early Orientalizing. White-on-dark is rarely found preserved at Kommos. For early,MG, examples of the technique, see the black-glazed cup from Eleutherna, Stampolidis 1990, p. 389, fig. 17, and pithos lids and cups from Knossos, KNC, pp. 330 and 389. 112 Krater(?),body

Figure 31. Krater(?)112

C9746 (83C/61). Max. p. dim. 6.4. Fine red clay, 10R 6/6. Single fragment of body wall. Decoration extremely worn. Double concentric circles linked tangentially to another motif; two bands below, painted area above. Orientalizing. 113 Krater,pedestal

Figure 32. Krater 113

Fig. 31

Fig. 32

C9667 (82B/49). Max. p. dim. 4.5. Fine buff clay, 5YR 6/6, with some mica. Single fragment of pedestal. Parts of three ribs of stem preserved and part of top of foot. Extremely worn; remains of red-brown paint on outside. Perhaps from the same vase as 60.

114 Kalathos

Fig. 28

C9668 (82B/49). H. 3.2, Diam. 8.5. Fine buff clay, 7.5YR 8/6. Two joining fragments of a small kalathos. String-cut base, thin walls with slightly everted lip; one small excrescence preserved on top of lip. Plain. The type may be of Cretan pedigree, though might be termed a miniature; cf. the MG-LG piece, Coldstream 1972, p. 96, no. 110. 115 Cup

Fig. 28

C10487 (82B/49). H. ca. 11, Diam. (rim) 12, Diam. (foot) 4. Fine pale beige clay, 7.5YR 7/6. Nineteen fragments, mending to twelve. Profile not completely preserved. Fine build; slightly raised base; thick handle. Once wholly painted. The shape and build suggest a LG date or a little later. 116 Jug, body and foot

Fig. 28

C10690 (82B/46, 47,48). P.H. 20.4, Diam. (foot) 7. Fine light graybuff clay, 10YR 7/4. Twenty-one fragments, mending to ten. Complete profile, save for upper parts and handle. Very fine vessel. Dull brown paint: band on outside of foot and another on lower wall; band above three lines at waist; crosshatched triangles, painted a little irregularly, at top of shoulder; two bands preserved on lower neck. LG or Early Orientalizing. Cf. Fortetsa,no. 704, pl. 45, for the general decorative system and shape. 117 Amphora, neck and shoulder

Fig. 33

C10080 (82B/45, 47, 48). Diam. 32. Nearly fine salmon-pink clay, 7.5YR 8/5, with paler buff surface.Twenty-one fragments, mending to thirteen. Large, fullbodied amphora with round handles; slight inset at base of neck, of which 5 cm is preserved. Dull brown glaze, terriblyworn. Band at base of neck and parts of two broad bands preserved on lower shoulder, supporting sets of six concentric semicircles.

ALAN

220

W. JOHNSTON

75 i

-10

EM, 117

129

The piece seems to be a LPG survival;there is a further possible fragment, from a lower level, in 82B/55 (see above p. 196). 118 Uncertain shape, body

Fig. 28

C10688 (82B/47). P.H. 5.6. Fine buff clay, 10YR 7/6, more pink in core, with many dark inclusions. Single fragment of wall of closed vase. Very worn paint: part of a hatched zigzag, a diagonal line, and a small circularmotif preserved. 119 Uncertain shape, body

Fig. 28

C10700 (82B/47). P.H. 6.6. Fine pink-tan clay, 5YR 6/6, with red inclusions. Fragment, probably of lower wall, of a closed vase. Inside and edges extremely worn. Part of a band of solid double axes and sets of at least five verticals preserved between bands. LG? The motif was spread, no doubt via Attic MG, to many areas of the Greek world. The clay appears Cretan. 120 Krateriskosor pyxis

Fig. 28

C10714 (82B/45 and 47). P.H. 9.6, Diam. (rim) 17.8. Semicoarse pinkish buff clay, 7.5YR 7/6, with many inclusions. Thirteen fragments, some joining. Much burnt. Upper

parts of some form of large bowl; sharply modeled, everted rim with ridge below. Upper parts painted outside; perhaps a frieze of double concentric circles below (Fig. 28 shows the only set well preserved, on two fragments); two lines and a band preserved on the lower parts. 7th century.An unusual shape, with crisp molding and hard, but nearly coarse, clay; vessel seems to have been used for cooking. 121 Jug(?), body C10689 (82B/45). 4.5 x 4.1,Th. 0.6. Fine buff-pink clay, 5YR 6/4. Two joining fragments of wall perhaps of a jug. Broad band of dull dark paint with a set of five concentric circles (plus central dot) in added white on it. Another rare example of white-ondark decoration, though not closely datable. 122 Cup or kantharos, rim and handle

Fig. 34

C9576 (82A/20). P.H. 7.6 (with handle), Diam. (rim) 14. Fine light red clay,2.5YR 5/6, with much small mica. Eight fragments, mostly joining, of rim, wall, and handle. Deep with straight wall, flaring slightly at plain lip; one heavy,vertical strap handle preserved.Wholly painted in dark glaze, save for area behind handle.

Figure 33. Larger pieces from Building Z and upper levels. Scale 1:6

BUILDING

Z AT

KOMMOS

221

122 123

/

126

124

125

132

n 131 I

n0 127

128 133

136 137 Figure 34. Material from upper levels. Scale 1:3

ALAN

222

W.

JOHNSTON

of some of the stirrup-kratersand skyphoi. An 8th-century date would therefore be indicated.

Cf. Fortetsa,no. 1057, pl. 73, though our piece may be an import. LG? 123 Cup

C10713 (82A/23A and 20). H. 10.2, Diam. (rim) 14.3. Fine salmon-pink clay, 5YR 7/8. Twentythree fragments, mending to five. Unusual transition of "collared" shoulder to lip. Fairly fine fabric. Wholly painted; orange inside, dark,but streaky outside. Perhaps later than MG. 124 Jug or aryballos,body

Fig. 34

C10708 (82A/20). P.H. 3.2. Fine orange-buff clay, 7.5YR 6/6, with pale buff surface. Single fragment of shoulder. Small jug or aryballos,with turn of neck just preserved.Worn. Parts of two highly irregularcrosshatched triangles preserved. LG? Cf. Fortetsa,p. 171, no. 41, "mainlyLPG, except the aryballoi which are LG." 125 Jug, body and foot

Fig. 34

C10709 (82A/14 and 24). P.H. 13.2, Diam. (foot) 6. Fine pinkish beige clay, 5YR 6/6, with some large inclusions. Twenty joining fragments of foot and body. Flat-based jug with very narrow neck; handle not preserved. Worn paint: band at base and two on lower wall; four close-set bands on shoulder. 126 Skyphos, rim

131 Cup, rim and body

Fig. 34

Figs. 34, 35

C10707 (82A/23). P.H. 1.5, Diam. (rim) ca. 16. Fine buff clay, 7.5YR 6/6. Two joining fragments of rim and shoulder.Top of low rim completely worn. Narrow reserved band on outside of lip; top of frieze preserved below. The presumed height of the frieze suggests that the motif consists of sigmas, not chevrons; chevron skyphoi, of whatever origin, are not otherwise attested at the site. The decoration here floats in the field, and, together with the short, low lip, indicates a LG date.

Figure 35. Skyphos 126 127 Cup

Fig. 34

C10697 (82A/7,13,14). H. 9.1, Diam. (rim) 12.7. Fine red-brown clay,5YR 5/6. Twelve fragments, mending to three, of a cup, about half-preserved. Rather heavy,with low everted lip and flat base. Once wholly painted. The build clearly belongs to the MG tradition. 128 Cup

Fig. 34

C10037 (82A/13). H. with handle 12.5, Diam. 15.6, Diam. (foot) 7. Fine, rather soft, pink-buff clay, 5YR 7/6 to 7.5YR 7/6. Twenty fragments, including two joining sets. Rather heavy build, with slightly concave base. Painted red-brown inside and out. Similar to 127. 129 Amphora, rim/neck

Fig. 33

C9628 (82A/11 and 13). P.H. 16.8, Diam. (rim) 18. Medium coarse reddish yellow clay, 5YR 7/6. Five fragments, four joining, of rim and neck. Broad neck, flaring to rim with vertical outer face, considerably hollowed on inside. Lip painted; lower neck painted, as far as preserved; horizontal zigzag at mid-neck. Cf. 22 above. 130 Closed vessel C10050 (82A/7 and 13). Diam. (foot) 5. Fine orange-red clay, paler surface,2.5YR 6/6. Eight fragments, two (from both pails) joining, of lower parts of a small closed vase. Worn, noticeably on breaks. Heavy wheel marks inside. Flat base and slim body. Painted streakily in dull red-brown glaze. The fabric is comparable to that

Fig. 34

C10038 (82A/12). P.H. 10, Diam. (rim) 11.2. Fine buff clay, 5YR 7/4. About forty fragments of upper parts of a cup. Globular body with low, nearly vertical lip. Sturdy strap handle. Wholly painted. The shape is typologically early, but the fine build would be unusual in MG. 132 Kantharos(?),rim and body

Fig. 34

C10051 (82A/11). P.H. 7.5, Diam. (rim) 12. Fine pink-buff, micaceous clay, 7.5YR 8/6. Five fragments, three joining, of an open vase, perhaps a kantharos. Straight wall, with slightly flaring lip. Very worn; once wholly painted in brown-fired glaze, save for narrow band near rim inside. An import; reminiscent of the developed Chian chalice and also Samian cups (cf. Furtwangler 1980, p. 209, fig. 16.11/1). 133 Cup, rim, body, and handle

Fig. 34

C10701 (82A/11). P.H. 6.4, Diam. (rim) 14. Semicoarse beigebrown clay, 7.5YR 6/6, more red in core, with much included matter; cream surface. Four fragments, two joining, of lip, wall, and handle. Upper parts,with heavy strap handle; fairly low and straight lip. Traces of paint remaining; probably once wholly painted (or dipped). This is another cup of MG tradition at a high level in Building Z; cf. 127. 134 Krater,pedestal C7047 (36B/2). Diam. (base) ca. 20. Fine ware, beige-buff, 5YR 6/6. Six fragments of pedestal; Callaghan and Johnston 2000, no. 264. Painted on the exterior, save for three reserved horizontals. Type and date similar to material from lower levels, see 57ff.

BUILDING

Z AT

KOMMOS

223

To be added to the catalogued material above are the following pieces of clear 7th-century date: 83C/61: body fragments of a Chian amphora, Protocorinthian kotyle handle, Orientalizing amphora sherd. 81B/50E: Protocorinthian linear kotyle sherd. 81B/61: amphora foot in coarse red clay, with joining fragments from 82B/51 and 83C/59. 82B/49: several fine-walled black-glazed cup fragments and one foot with a groove near the outer edge, comparable to that in 82A/30 (p. 214 above), but more crisply modeled. 82B/48: several black-glazed cup feet with concave underside and perhaps a body fragment from a Samian amphora. 82B/47: a very worn body fragment, perhaps from a Chian amphora. 82B/45: two Lesbian amphora body fragments; a fine juglet base, with a groove at the edge of the underside and a spiral of glaze at the center; and a chytra of fine fabric with tall lip. 82B/41: lowest appearance of Protocorinthian (worn kotyle fragments) within the area of Z, together with clearly later 7thcentury black-glazed cup feet. Fragments of a pithos with an impressed circle frieze on a raised band were found in both

82B/41 and 45. 82A/19: little clearly later than MG, though the handle of a jug with a wavy line down it probably is so. Cooking ware is scarce in this area, but this pail has part of a flat-floored tripod bowl (other sherds in 82A/13 and 18). 82A/23 and 23A: scraps of fine-walled black-glazed cups.9 From the surface levels, three pieces are of particular interest: 135 Amphora, rim and neck Fig. 36

Figure 36. Amphora 135

9. One pail that is from these levels but seems to have only 8th-century materialis 82A/20 (above4.27 m); there aremany bases of cups and skyphoi,and largervases, all flat (save for a probablepedestaledkraterfloor fragment).The pail also contains rim fragmentsof "blue-core"clay from a vase representedby sherdsin 82A/15 and 15A.

C9670 (83C/59). P.H. 7.5, Diam. (rim) ca. 14. Micaceous graybrown clay,with lighter surface, 5YR 5/1. Six joining fragments of neck and rim. Rather square but simple lip; slight ridge on neck below. Plain. Lesbian transport amphora of very early type; see Dupont 1998, pp. 159-160, for early parallels. 136 Open vessel, body and foot

Fig. 34

C10711 (82B/42). P.H. 8, Diam. (foot) 7.8. Micaceous fine tan clay, 7.5YR 6/6. Three joining fragments of lower part of an open vase. Flat foot and steep walls. Rather roughly finished inside. Wholly painted, save for underside. The paint inside precludes the most obvious inference

that this is a closed vase, such as an olpe, though one may note the rough finish of the inside of East Greek pyxides; Kerschner 1997, col. 185, with note 67. 137 Amphora or Fig. 34 (part) hydria, body and foot C10712 (82B/41). Diam. (foot) ca. 14. Fine micaceous tan clay, 7.5YR 6/6, with gray core and pale beige surface. Six fragments, joining to three, of foot and body. Tall, splaying ring foot. Banded decoration in dull red-brown glaze: band on outside of foot (not extending onto wall); double band on lower body; wavy band, ca. 1.0 thick on small body fragment. Typical of a range of East Greek amphoras found on the site; 7th century.

ALAN

224

GENERAL

W.

JOHNSTON

REMARKS

As noted in the introduction, we have here a good cross-section of local pottery styles of ca. 800-725 B.C., with finds from the earlier Iron Age levels being very poorly preserved and almost wholly discrete from the material from the phase of use above. Stratigraphically,the pottery distribution goes some way to confirma second floor was laid at ca. 4.35 m, above the large deposit in that ing 82B/50, although some sherds from pots found mainly below this floor do appear in higher levels. Certainly, stylistically later pottery only begins to be found at this point, with some sherds clearly showing Late Geometric to Early Orientalizing trends; there is an admixture of earlier material, with a little Minoan pottery, but amounts are very small until we reach a level where the area seems to have been once more open land, into which Bronze Age residue drifted sometime in the 7th century.A more precise date for the end of the period of use of Building Z is less easy to establish, largely because of the chronologically recalcitrant nature of the pottery found,10though there are also difficulties in distinguishing which sherds are from the final phase of use of Z and which are from later use of the area. The pottery that can be surely attributed to the main period of use, however, can be definitively isolated and attributed to the Middle Geometric period, probably late in it. There are few obvious exceptions, of which 41, primarily found to the north, and 14, which is one of the few tolerablywell preservedlarge vessels, are the clearest examples, bracketing either side of MG. Much of the material from within the building is preserved in small fragments, with no joining sherds found outside; the erosion to the west hampers closer argument,but it would seem that a residue of sherds from pots once used in the building remained there, and that we might think of a period of use in very rough terms of about two generations.11Another possible way of considering the chronology is to reflect on the lack of Corinthian pottery in Building Z; does this indicate a "preThapsos" date, therefore earlierthan apparentlycomparablematerial from the temple area?An unhelpful, but fair, answer is that Corinthian LG pottery is so rare at Kommos that we should not expect a piece to appear in the restricted corpus of Building Z.12 There is in fact no Corinthian pottery at all in Building Z, and the amount of other imported material is small; Phoenician pottery is scant (though not absent), and a number of micaceous vases, mostly of closed shape, will have had a Euboean, Cycladic, or East Greek origin; they are more frequently encountered in the upper levels. It is probable that some kraters and skyphoi are imports from elsewhere in Crete. The range of shapes is wide and variety of decoration within each shape also considerable, to the extent that it is rarethat closely similar vases are found. There are clear exceptions in the very fragmentary amphoras with concentric semicircles on the shoulder (the majority of which are included in the catalogue above),the black-glazed cups, and the "symposiumsets"ofkrater, skyphos, and (more rarely)jug; the import 12 and the later 110 also conceal more than one piece. Overall in Building Z the number of decorated

10. It is largelylocal and without strong diagnosticfeatures.For Knossos, cf. Coldstreamand MacDonald 1997, p. 237: "closedating of seventh century sherdsis rarelypossible." 11. FurtherMG materialwas found, apparentlyredeposited,in disturbed levels in Gallery 4 to the south (trench 86F); equallyfragmentary,this material presentsno joins with materialfrom Z nor with the temple deposits.These sherdswill be treatedmore fully in my third reporton the Iron Age pottery, but they are noted here as the only other concentratedset of contemporary materialexcavatedin the vicinity of Z. 12. This materialis treatedin CallaghanandJohnston 2000, p. 312, n. 12; the amount is very restricted.

BUILDING

13.The attemptto differentiate betweensymposiumandothervessels has alwaysseemed to me to be largely misplaced,since most allegedly diagnosticmaterialwould be suitable for a rangeof social activity.Certainly the whole corpusof any materialhas to be examinedas a unit for confident conclusionsto be drawn,and in the case of Z, it may be only the sheer amount of pottery and the relative elaboratenessof much of it that point to a "higher"social use such as in a symposiumat a festival. 14. See Shaw and Shaw 1993, p. 177, and KommosIV, pp. 25-26. 15. For the dating and use of Building Q see Johnston 1993, esp. pp. 374-376, and for the architecture, see Kommos IV, pp. 14-36. 16. For a useful statementof the possibilities,see Vink 1996, even if a little overgeneralon the difficultiesof dating the pottery.

Z AT

KOMMOS

225

amphoraswould originally have been substantial,but larger semicoarse or coarse pithoi appear in a very restricted number of fragments. In many respects, the character of the period-of-use pottery can be matched closely with that from the temple area deposits, though the virtual lack of joins between the two areas does not primafacie suggest that Building Z acted simply as a storage facility for festival equipment. In both areasvessels that could not be described as such equipment are rare.l3 There are indeed differences in the assemblages, for example, no trace of figurines in Z, more pedestaled kratersin Z, and a greater variety of cups and skyphoi in the temple area,but it might be impertinent in view of the incomplete recovery of the material there (because of overlying features) and the highly fragmentary nature of Z to describe these differences as significant, except perhaps with respect to the kraters (adding those mentioned in note 11 above), which do point to sympotic activity being more frequent than, or at least of a different characterfrom, whatever such activity was carried out in the temple area. The closer definition of such activity,whether reflecting a Cretan form of syssition or more "mainland" style of male commensality, is scarcely possible on available evidence. Nonceramic finds are very few, most noticeably a set of four iron points from a relatively high level (82A/8), two probable spindle whorls (from different areas), and some fishhooks (also from a high level, 82A/48), together with a few worked stones-a mortar and ten stone fishing weights. Substantial traces of hearths, together with bones, limpets, and pumice, complete the picture of food preparationactually in the building,14 though we have noted that cooking ware is of no great prominence. There are some pointers here to Building Z being a form of andreion, certainly not a cooking-free andreion of later type, but naturallythere are dangers in imposing later developments on 8th-century evidence. The building would appearto have been abandoned by about the end of the 8th century B.C.; some 7th-century sherds were found at a level slightly below the highest remains of the walls of Building Z, but associated finds point to the area being open at the time. The material here includes imported amphoras, and it may be tempting to suggest, with respect to the function of Z, that it was replaced by the long, snaky Building Qa little to the west-northwest, where 7th-century amphoras constitute so much of the material recovered, and evidence of use as any form of syssition is lacking.5 The presence of these amphora sherds above Z could, however, be misleading, since there would certainly appear to be a hiatus between the end of the use of Z and the installation of Q possibly a hiatus of some considerable extent that would remain even if we favored a radical downdating of Late Geometric and Subgeometric material.'6Yet one factor that begs an explanation remains: the architecture of Q_seems, in its extraordinarylength and without contemporary parallel, to mimic a gallery of P. Could it not be modeled on Building Z, a reworking of one of the galleries, thereby implying that Z was still at least visible when Q was built? While it should be noted that Building Qwas a longer building than Z could have been, it may be that Q took as its model another of the galleries that may have remained visible, possibly still in use, well down into the 7th century.This does not of course nullify any argument that the

ALAN

226

W.

JOHNSTON

functions of Qand Z were similar, but to support it, at least on current evidence, an explanation would be required concerning the location, or indeed locations, of those functions through much of the 7th century. There are no comparable corpora of material published from individual buildings in central Crete, a situation that one hopes can be rectified in the near future. It is to be regretted that the finds from similarly dated rooms at Phaistos are even more fragmentary than this set from Kommos, while those from Gortyn are naturally to be associated more with the sanctuary material at our site.17This publication is therefore intended as an initial exploration rather than the final word.

17. Phaistos:Rocchetti 1974-1975, where there is a broaderchronological spreadfrom each room;parallelsexist for many of the Kommos types at a generallevel, but few with respectto specific cases. Gortyn: Rizza and Scrinari1968. In neitherpublication arethere publishedmore than a few profiles of the material.

REFERENCES Bikai, P. 2000. "PhoenicianCeramics from the Greek Sanctuary,"in Kommos IV, pp. 302-312. Callaghan,P.J., and A. W. Johnston. 2000. "The Potteryfrom the Greek Temples at Kommos,"in Kommos IV, pp. 210-301. Coldstream,J. N. 1968. GreekGeometric Pottery,London. . 1972. "Knossos1951-61: Protogeometricand Geometric Potteryfrom the Town,"BSA 67, pp. 63-98. Coldstream,J. N., and C. F. MacDonald. 1997. "Knossos:Area of South-west Houses, Early Hellenic Occupation,"BSA 92, pp. 191-245. Courbin,P. 1966. La ceramique geometrique de lArgolide, Paris.

Dupont, P. 1998. "ArchaicEast Greek TradeAmphoras,"in R. M. Cook and P. Dupont, East GreekPottery, London, pp. 142-191. Fortetsa= J. K. Brock,Fortetsa: Early GreekTombsnear Knossos (BSA

Suppl.2), Cambridge1957. Furtwangler,A. 1980. "Heraionvon Samos:Grabungenim Siidtemenos 1977,1I: Schicht- und Baubefund, Keramik,"AM95, pp. 149-224.

Alan W.Johnston INSTITUTE

OF ARCHAEOLOGY COLLEGE

UNIVERSITY 31-34

GORDON

LONDON

WCI

LONDON

SQUARE

oPY

ENGLAND

ALAN.JOHNSTON@)UCL.AC.UK

Johnston,A. W. 1993. "Potteryfrom Archaic Building Q at Kommos," Hesperia62, pp. 339-382. Kerschner,M. 1997. "Ein stratifizierter Opferkomplexdes 7. Jh.s v. Chr. aus den Artemision von Ephesos,"OJb 66, Beiblatt,cols. 84-226. KNC = J. N. Coldstreamand H. W. Catling, eds., KnossosNortb Cemetery:Early GreekTombs (BSA

Suppl.28), London 1996. KommosIV = J. W. Shaw and M. C. Shaw, The GreekSanctuary, Princeton2000. Levi, D. 1927-1929. "Arkades: Una citta cretese all'albadella civiltaellenica,"ASAtene10-12, pp. 1-710. Ozygit, 0. 1994. "The City Walls of Phokaia,"REA 96, pp. 77-109. Rizza, G., and V. SantaMaria Scrinari. 1968. Il santuario sull'acropolidi Gortina I, Rome.

Rocchetti, L. 1974-1975. "Laceramica dell'abitatogeometrico di Fest6s a occidente del palazzo minoico," ASAtene33-36, pp. 169-300. Samos V = H. Walter, Friihe samische Gefcisse:Chronologie undLandschaftsstileostgriecbiscbenGefisse (Samos V), Bonn 1968.

Shaw,J. W., and M. C. Shaw. 1993. "Excavationsat Kommos (Crete) during 1986-1992," Hesperia 62, pp. 129-190. Stampolidis,N. 1990. "Eleuthernaon Crete:An Interim Report on the Geometric-ArchaicCemetery," BSA 85, pp. 369-403. . 1996. Avr'r otva-Reprisals, Rethymnon. Tsipopoulou,M. 1990. "Ceramicadei periodi subminoico,geometrico,ed orientalizzantein Creta orientale," Seminari1990 (C.N.R. Istituto per gli Studi micenei ed egeo-anatolici, Rome), pp. 137-164. . In press. TwoyUV)ix'CXr, reW 1pw-TOYs?cP1prp(x6 UEjwK%jtx xat AvvacroriSOoua xepaxYWlx TzY1 CnYdSHIEptOX E-?oxpY-rX (Athens). Vink, M. C. 1996. "The Archaic Period in Greece:Another Dark Age?"in Debating Dark Ages (Caeculus 3),

M. C. Vink et al., eds., Groningen, pp. 1-18. Viviers,D. 1994. "Lacite de Dattalla et l'expansionterritorialede Lyktos en Crete centrale,"BCH 118, pp. 229259.

HESPERIA Pages

69,

2000

ATH

PO

E NIAN

227-233

AN

D

T HE

For Christian Habicht

R IPTION

I NSC

Y[EAR

LITI

S

3 07

CIAN O

S TO

S

30

F 2

ABSTRACT The leading politicians of the years 307/6 to 302/1 B.C. in Athens were Demosthenes' nephew, Demochares, son of Laches, of Leukonoion, and Stratokles,son of Euthydemos,from Diomeia. With the ousterof Demetrios of Phaleron in the year 307, they and their fellow politicians ostentatiously proclaimedrenewedfreedom.As partoftheir democraticprogram,they published many decreeson stone. They also took carethat the inscribersof these measuresused blank spaces or line-initial position to give their names visual prominence. Clearly,they expectedtheir fellow citizens to notice-that is, to read-their names.

1. A version of this paperwas deliveredin Chicago on December 29, 1997, at the inauguralpapersession of the American Society of Greek and Latin Epigraphy(ASGLE), a joint colloquiumof the American Philological Association and the Archaeological Institute of America.That session honored my colleagueand friend ChristianHabicht. 2. On the proceduresof a cutter,see Tracy1975, pp. 95-120, esp. 115-119. 3. For specifics,see below,p. 229 and note 29. 4. Tracy 1995, pp. 36-51, esp. 37-41.

It is the purpose of this note to document briefly a curious case in which prominent Athenian political figures appearto have exercised a direct influence on the layout of inscriptions.1The layout of the text on the stone was a matter normally,and of necessity, left to the discretion of the lettercutters, that is to say,those who actually did the inscribing.2 Inscriptions basically reflect political acts. Attic decrees are a manifestation of actions taken by the Athenian council and public assembly.The motivation behind the passage of a decree, in most cases, cannot be known by us in any detail. Moreover, the decision about which measures were to be inscribed and set up in a public place was no doubt often a matter of much partisan wrangling. Again, it is unclear what rules applied. Not inscribing may also be a political act. During the ten-year period from 317 to 307 B.C., when Demetrios of Phaleron was Cassander's spokesman in Athens, few decrees were inscribed.3This situation has been taken to indicate the autocraticnature of the government at that time; it certainly made Demetrios vulnerableto attack on the issue. I have argued elsewhere, however, that this should be interpreted not as an anti-democratic act but an economic one.4 Demetrios of Phaleron circumscribedthe funds available for inscribing stelai,just as he limited private expenditures for lavish grave monuments; thus, few texts were inscribed in a more or less permanent form on stone. In any case, a lack of inscriptions provides no evidence that the assembly was less active than usual. When the other Demetrios, the

228

STEPHEN

V. TRACY

son of King Antigonos, took Athens in 307 and expelled the Macedonian garrison, he claimed to have freed Athens and restored the democracy. Demetrios of Phaleron was allowed to withdraw to Thebes. The Athenians responded with an outburst of enthusiasm at the ouster of the hated Macedonian garrisonfrom Piraeusand by lavishing praiseupon Demetrios and his father King Antigonos as "Saviorgods."5 The leading politicians of the ensuing new order were Demochares, son of Laches, of the deme Leukonoion' and Stratokles,son of Euthydemos, from Diomeia.7 Demochares was a nephew of Demosthenes and followed his uncle's policy of outspoken opposition to the Macedonians. During the late summer of the year 322 B.C., when he was only in his late twenties, he opposed the demand by the victorious Macedonian general Antipater to condemn Demosthenes, Hypereides, and other Athenian leaders after the defeat at Krannon.AHe was very active in rearming and refortifying the city in the year 307 and thereafter,following the ouster of Demetrios of Phaleron;9indeed, it was almost certainly he (the name is restored)who proposed in the year 307/6 the inscription published as IG 112 463, a decree to rebuild the long walls. In doing this, he surely had in mind the leadership of his uncle just thirty years earlier in refortifying the city after the defeat at Chaironeia.'0During the year 307 or 306, Demochares also supported a certain Sophokles, who proposed and secured the passage of a law stipulating that no one be allowed to head up one of the philosophical schools unless approved by the council and the assembly of Athens. Although ostensibly a general law, it was aimed at Aristotle's school, the Peripatos.Aristotle and his successor,Theophrastos, both had had extensive connections with Macedonia and were no doubt considered by some to be dangerously "pro-Macedonian."Sophokles' law was soon found unconstitutional and he was fined, despite Demochares' plea for the defense in which he vigorously attacked Aristotle."' Demochares' dislike of the Macedonians and admiration for his uncle Demosthenes lasted all his life. Indeed, one of his final public acts was to move a decree in the year 280/79 granting posthumous honors-a statue and the right to eat in the public dining room-to Demosthenes.12 The other majorpolitical figure of the time was Stratoklesof Diomeia. He came from a wealthy family; both his grandfatherand father served as trierarchseveral times.13 He was almost an exact contemporary of Demochares. We first hear of him in the year 324, when he joined in the prosecution of Demosthenes at the time of the Harpalos affair.14He proposed twenty-eight decrees that we know of in the years 307 and after, more than any other known politician.15Among these is IG 112 457, the famous decree of early 306 B.c. honoring Lykourgos, the great Athenian financier, for his championing of the democracy.16This seems a clear attempt on his part to recall the great days of the democracy of the 330s and to align himself programmaticallywith the policies of Lykourgos. He soon, however, became a strong adherent of Demetrios Poliorketes and was criticized for his flattery of the king by Demochares.17This led to Demochares' exile early in the year 303.18 Demochares and Stratokleswere not, of course, the only political figures active at this time; there were others (see the names of the orators!

5. Plut. Demetr.10. 6. PA 3716. 7. PA 12938. 8. [Plut.] Vit.X orat.847d with Plut. Dem. 28.2. 9. [Plut.] Vit.Xorat. 851d. 10. On Demosthenes' role as tsexcXv eirttzX,Uq, see Dem. 18.118, Aeschin. 3.17, and [Plut.] Vit.Xorat. 845f-846a, 851a. 11. Ath. 13.610f; Diog. Laert.5.38. On the whole affair,see Habicht 1988, pp. 7-10. 12. [Plut.] Vit.Xorat. 850f-851c. 13. IG112 1612, lines 136-140, 271278, 317-319, 320-321, 322-323. For more on the family,see Davies 1971, pp. 494-495. 14. Din. 1.1, 20-21. 15. For a list, Dinsmoor 1931, pp. 13-14, with the additionsnoted by Habicht (1977, p. 39, note 15). There is anotherdecreeof his now publishedin Koumanoudis1986. Osborne (19811983, II, p. 137) has identified no. D62 in his collection as yet another. 16. This inscriptionis quite fragmentary.[Plut.] Vit.Xorat. 852 also providesa text of this measure. 17. Ath. 6.252f-253d. 18. Plut. Demetr.24; [Plut.] Vit.X orat.851e.

ATHENIAN

19. PA5326. During the year304 he proposedthe fragmentaryhonorary decree,Hesperia5, 1936, p. 201, for a supporterof the democracy(line 15) and in the next yearIG 112 483, a measurepraisinga Rhodiandoctor, most probablyfor his servicesin the four-yearwar againstCassander. 20. PA 9450; Davies 1971, p. 425. In about 302 he proposedIG 112 506, a decreepraisinga certainMenoites for, among other things, his aid in the LamianWar. 21. He is most probablythe son of Nikostratos,son of Pythodoros,of Acharnai(PA11026). For the family, see Davies 1971, pp. 481-484. He is restoredas the oratorin line 9 of Agora XVI, no. 107, a fragmentaryhonorary decreeof the year 307/6. 22. Davies 1971, pp. 161-165. 23. Lys. 13.13, 31; 30.14. 24. Xen. Hell. 6.3.2, 7. 25. Dem. 18.114; IG 112 1496, lines 22-25. 26. Arr.Anab.1.10.4. 27. Habicht 1995, pp. 76-88; this is translatedin the English version as "Democracywithout Full Freedom" (Habicht 1997, pp. 67-81). 28. (Lykourgos) 11XtV] ixet?RAV 7Ot6v] spav

x6tL6,votlov

[UZ?0p

-rob Ti'v

ELVXL Lt Xac A

7cXG)qL V[XqxCV-L

Ocycov])6]O?vo0 (IG JJ2 457, lines 16-

17). These fulsome words of Stratokles praisingLykourgosearlyin the year 306 providea good idea of the political rhetoricof the time. 29. For a discussionof other inscriptionsthat have been assignedto these yearsor may be, see Tracy 1995, p. 36, note 2. 30. Fifteen arelisted in Tracy 1995, p. 40, note 21. We may now add Agora I 4953 (AgoraXVI, no. 107A). 31. No such prominenceis granted to the secretaryor to the chairmanof the proedroi,the other two officials routinelynamed in the bodies of preambles.The archon'sname has naturalprominence,for it normally comes first in the dating formulaC -rob aezVO~ &,PxovTo~.

POLITICIANS

AND

INSCRIPTIONS

229

politicians listed in Table 1), such as Euboulides of Eleusis,19Lysikratesof Melite,20Pythodoros ofAcharnai,21and Diotimos of Euonymon. Lamentably, we know little directly about these individuals beyond the decrees they proposed. Diotimos of Euonymon is representative.He put forward IG 112467, a decree of the year 306/5, honoring a certain Timosthenes for his actions against the Macedonian general Antipater at the time of the Lamian War. This is all we know of him, but his family is a well-known one with a history of supporting the democracy.22His great-grandfather's brother was a general in the Peloponnesian War and was executed by the Thirty Tyrants in 405/4.23 His grandfather'scousin was remembered for a strongly anti-Spartan speech.24His uncle contributed shields to rearmthe people after the battle at Chaironeia2' and was one of those followers of Lykourgos and Demosthenes whom Alexander demanded be handed over to him in autumn of the year 335.26

The record is clear.These men tended to look to the past and sought to align themselves, through proposals put forward in the assembly,with prominent individuals who had led the city in opposition to Macedon. Whatever their political differences-Demochares will not, for example, have appreciated orforgottenStratokles'prosecution of his uncle Demosthenes in the Harpalos affair-in the first years following the ouster of Demetrios of Phaleronthey collaboratedin presentinga strongpublic stance that was pro-democratic and anti-Macedonian. Their cohesiveness was undoubtedly abetted by Cassander's efforts to regain control of Athens during the years 307 to 304 in the conflict known as the Four Years'War. The two led what Christian Habicht in his recent book on Hellenistic Athens has felicitously described as "die halbfreie Demokratie."27 Whatever else they did, they made a special point of issuing many decrees on stone and obviously intended these inscriptions to publicize their political program: the ostentatious claim of a renewed freedom.28 The contrastbetween their practice and that under Demetrios of Phaleron is most striking. Only two inscriptions, IG 112450 and 453, can certainly be assigned to the years 317 to 307, the ten-year period of Demetrios.29 Sixteen inscriptions can now be attributed to the year 307/6 alone, the first year of the new regime.30These politicians clearly considered decrees published on stone not only as visible signs of democratic action, but also as a clearway of differentiating themselves from the previous regime. They unquestionably expected their fellow citizens to take note. All this is more or less well known, at least to a select few. What has not heretofore been noticed is that an unusually high number of the texts of the years 307 to 302 emphasize the orator, that is, the politician who This is accomplishedeither by placing blank spaces proposed the measure.31 before the speaker'sname or by according his name first position in the line. Large letters, which might also have achieved the desired effect, were not used; they are deployed in Attic decrees only in headings, never in the body of a text. Moreover, Attic inscriptions normally have no spaces between clauses and none between words; their inscribed faces present to the viewer a sea of letters all run together. Thus, blank space(s) and line-initial position will have enabled a readerto pick out quite easily the name of the speaker.That was the purpose and it is a matter of some consequence.

STEPHEN

230

TABLE

1. INSCRIPTIONS

Reference

Date

SPEAKER

V. TRACY

OF THE YEARS 307/6 TO 302/1 Means

Orator/Politician

Cutter/Hand

EMPHASIZED

IG 112 1262 Cutter IG 112498 Cutter

AgoraXVI, no. 107 IG 112 358 Hondius 1925, pp. 39-40 IG 112 455 IG 112 463 IG112 467 AgoraXVI, no. 113 IG112 471 IG112 796 IG 112 554 IG112 482 Horos4, 1986, pp. 19-20 Hesperia7, 1938, p.297, no.22 IG 1J2 486 IG 112 496 IG 1J2 498 Hesperia1, 1932, p. 45, no. IV AgoraXVI, no. 123 IG 1J2 504

[Anaxikrates]307/6' [Anaxikrates]307/6 Anaxikrates307/6

space[s]before `80o0v [PythodorosAcharneus] ----nos Kydathenaieus line-initial line-initial [Str]atokles[Diomeeus]

[Anaxikrates]307/6 [Anaxikrates]307/6 Koroibos306/5 Koroibos306/5 Koroibos306/5 [Euxenippos]305/4b ca. 305 Pherekles304/3 [Pherekles]304/3

line-initial line-initial line-initial blank space[s] 1 blank space line-initial 2 blank spaces line-initial 2 blank spaces

IG 112 1262 Cutter [StratoklesDiomeeus] [DemocharesLeukonoieus] Diotimos [Euonymeus] son of KephisogenesAcharneus StratoklesDiomeeus ----Phileas Palleneus -------eus [Kallaides [Xypetaion]

[Pherekles]304/3

1 blank space

Stratokles[Diome]eus

IG 112478 Cutter

Pherekles304/3C [L]eostr[atos]303/2 [Leostratos]303/2 Nikokles 302/1

4 spacesbefore `80oiv 1 blank space 6 blank spaces 3 blank spaces

[Stratokles]Diomeeus [Stratokles]Diomeeus PhilostratosKephisieus Strat[oklesDiomeeus]

IG 112 495 Cutter IG 112 1262 Cutter IG 112 498 Cutter IG 112478 Cutter

Nikokles 302/1 [Nikokles] 302/1

IGC 12 374

307-303/2d

2 blank spaces line-initial line-initial 1 blank space line-initial line-initial

Memnon Aphidnaios ---------eus [----k]leous Gargettios Str[atoklesDiomeeus] [-------G]argettios Lysikrates[Meliteus]

AgoraXVI, no. 110 307-301 ca. 303 IG112 595 ca. 302 IG112 506

No

EMPHASIS

457 IG 1J2 460 Hesperia5,1936, p.201 IG 1J2 483 Horos4, 1986, pp. 11-12 IG 112 489 IG 112 493 IG 112 494 IG 112 495 IG112 499 IG 112 500 IG112 503 IG 112 505 IG 112 559 + 568 (addenda p. 662) IG112

IG 112 1262 Cutter

IG 112478 Cutter

ON SPEAKER

[Anaxikrates]307/6 [Ana]xikrates307/6 Euxeni[ppos] 305/4

Strat[oklesDiomeeus] [Stratokles]Diomeeus [Euboulides]E[leusinios]

IG 112 1262 Cutter

Pherekles304/3 [Pherekles]304/3'

EuboulidesEleusinios Stratokles[Diomeeus]

IG 112 498 Cutter IG 112 495 Cutter

Leostratos303/2 Leostratos303/2 Leostratos303/2 Leostratos303/2f [Nikokles] 302/1 Nikokles 302/1 Nikokles 302/1 Nikokles 302/1 ca. 303

Charisos [Sphettios] KalliasHermeios [A]ristok--StratoklesDiomeeus [StratoklesDiom]eieus Memnon Aphidnaios [Stratokles]Diomeeus Euphiletos Kephisieus Strato[klesDiomeeus]

IG 112498 Cutter

IG 112495 Cutter IG 112 650 Cutter IG 112 1262 Cutter

established tedate. bOn thedate,Meritt1936,pp.201-203. CPassedon the same dayas the previousinscription. dThe date aDw1963,pp.5-0 486. fPassedon the samedayas 7 (1938), p. 297,no. 22, andIG11I2 is thatof Osborne1981-1983,no.D50. epassedon the samedayasHesperia 496. IG11I2

AND

POLITICIANS

ATHENIAN

23I

INSCRIPTIONS

The preamble of IG 112 471, an honorary decree of the year 306/5, provides a good example of the use of a blank space placed before the speaker'sname. The text that follows repeats that of J. Kirchner from the second edition of IG. 'ETt Kopot3ov ''pXov-coqSt

-

S OEvET8o0 8&X6-rqq Tcpv-cxvstcaS sELrILocq

5

OEsoysE-Covo-

S 'Pc[tvo0cmtoqSyQ'XYpoctcSvSV VXLMsMoUVL7XtVo SVSLX%aL Sx XOGrTXL vr3oM1vuOL, SVocTSa

si -it; 7PD-cCave6axe%%X?G6a s4s@cpL4sv P1610 OLC7cogfluOLcovo$MxpaocoxvoTV apooi8poov [S]

Gct opucposapoL.

t8o,uv

E`o8sv

TJ.i-

ED6Oup-rocp-oxX;k

[oU ALoysssb

Wlsv

sETCi' [o]-

When a cutter chose to place the speaker first in his own line, that is, to begin the line with his name, he usually could accomplish this only by leaving some blank space at the end of the previous line. IG 112 455, the top right part of a decree that has been convincingly restored as proposed by Stratokles near the beginning of the year 307/6, displays this arrangement in lines 6 to 7. The inscriber,the IG 112 1262 Cutter,32has left nine letter-spaces

blank at the end of line 6 in order to begin line 7 with

Stratokles'name.33The text is Kirchner'sfrom the second edition of IG. I7coo%]v- vacat (7) -sat ['ETcV AAvacExpa'-oou 6`P(ovTo;6cL ALo,uAuomac NoOL3bwcoo L [-c0o; asuTiopca;7CPD-ravseoc;,

5

'a v[sseu 'YPy6a-acsUsv Ms-ranyse-cvivo; SVa] [ov, TP6-se -cit 7CPDuvsLa; eXXX?G6a,v -cJiv 7P]o8iPCov C[s 27.]b. Io; xaovv .............

-cit PoUAi~ xaoct -rcJ8,IyO I[ vacat(9) [-px-c-oxki; E6Oo8uyou ALoyssbesclsv sC] &'8o68N,o[7cp6oospoo E`o8sv

Not all texts of these years give the speaker emphasis. IG 112 460, also of the year 307/6, also proposed by Stratokles, and also by the IGC12 1262 Cutter, provides an example of a text whose layout gives no emphasis to the name of the speaker.Here the clause of ratification has been omitted and Stratokles'name, revealed by the stoichedonorder,straddled lines 6 to 7 with no spaces left before it. Again the text is that of Kirchner: ap)exov;osET rN ]. , L Aoua;x NoOnc7co][v Atoys]tV~sb 'pyto[y-susv 'E cpNfoRvo; s]-

['EsCAt 'Aval]xp'-ro[o;

[L3oq 8s]%x-cl7CPu[Tocvsvoc

32. See Tracy 1995, pp. 136-147, for a descriptionof this cutter'shand and a list of his inscriptions. 33. The rationalefor the vacat of seven spacesin line 1 is unclear.

[v6ct-c] kLas-civoo,

5

ev'[-ac -cit

a

vsx

dxxl]-

[?XathxocJiv TCopoe8o[v snpsv........... . . .

v 'A?Rco~csxiz%sv 4cc)auycx6pot

[iS E]6Oua8N,ouA\o,u[ssbSWcasv.]

pcaToxkx]-

232

STEPHEN

V. TRACY

This latter case, with no special emphasis on the speaker's name, is the norm. There is also evidence, however, in the years before and after the period studied here, for a tendency to set off with blank spaces either the speaker's name or the clause of ratification (E-o4sv -cii Poo;kL xocxc-rCo -xtoL), which, when included, immediately precedes his name. This practice occurs roughly a quarterof the time. For example, of forty preambles published in the second edition of volume II of IG and dated to the years 336 to 308, ten emphasize the speaker.34 The proportion after the year 300 is coincidentally the same-four of the sixteen inscriptions dated to the years 299 to 280.35There was, thus, some inclination on the part of cutters to guide the reader'seye past the preamble toward the beginning of the decree proper. In sharp contrast, nearly two-thirds of the inscriptions from the years 307/6 to 302/1 give emphasis to the name of the orator/politician (23 out of 37, or 62%;see Table 1). Clearly the politicians of these years took care that their names were given prominence ratheroften in the texts inscribed on stone and set up publicly. How, precisely, they accomplished this is difficult to say.It is hard to imagine a Demochares or a Stratokles visiting a cutter'sworkshop to make his wishes known on the subject. It is probably the case that the matter was called to the attention of the secretaries in the assembly and they, or their emissaries, conveyed it, more or less strongly, to the cutters at the time when they delivered the copies of the measures to be inscribed on stone. There was clearly no absolute requirement to give prominence to the name of the politician; the cutter was, after all, the one who was primarily responsible for the layout. The information provided in Table 1 reveals that not every measure proposed by a given politician, nor every one inscribed by a given cutter,36nor even all those passed at the same session, emphasized the speaker.This lack of consistency suggests that the mechanism was informal, perhaps a note of reminder attached to the measure to be inscribed. There was apparently,as one might expect, greater onus on cutters to do this initially, when enthusiasm was high for the new regime, than in later years when it had waned and the threat of invasion by Cassander no longer provided cohesion. In any case, 5 of 7 texts give prominence to the orator'sname in the year 307/6, 3 of 3 in 306/5, 1 of 2 in 305/4, and 4 of 6 in 304/3; but this ratio falls to 2 of 6 in 303/2 and 3 of 7 in 302/1. That politicians took the trouble to have their names prominently featured on inscribed copies of measures set up in public is of more than passing interest. It is, first of all, an indication, it bears repeating, that these Athenian politicians saw inscriptions as visible symbols that could be used to support their political claims of renewed freedom. Second, it is apparent that they anticipated people reading their names; clearly, they both wanted and expected to receive credit for measures that they had put forward. Perhaps,indeed, this is evidence that people often referredto a measure by the name of the man who proposed it, for example, "Stratokles' decree."This point has interesting implications for the level and extent of literacy in Athens at this time. At the very least, we may conclude that by the late 4th century B.C. most, if not all, citizens residing in the city used writing and "found literacy indispensable."37

34. IG I2 348, 349 (s-o&sv), 350,

358, 360, 363 (s-o&sv), 368, 372, 374, and 448 (I). 35. IG 112661, 662 (sdo&sv),665 (s-o&sv), and 666. 36. Apart from the IG 112478 Cutter,who will be discussedin a study of 3rd-centuryinscribersnow in preparation,these cuttershave been presentedand studied in Tracy 1995. 37. This last phraseis from Harris 1989, p. 115. Harristakes a conservative view of the generallevel of literacy in Attica, regardingit, in fact, as quite low (p. 94 and note 135, for example).

ATHENIAN

POLITICIANS

AND

INSCRIPTIONS

233

REFERENCES AgoraXVI = A. G. Woodhead,Inscriptions:TheDecrees(TheAthenian AgoraXVI), Princeton 1997. Davies,J. K. 1971.AthenianPropertied Families.600-300 B.C., Oxford. Dinsmoor,W. B. 1931. TheArchonsof Athensin theHellenisticAge, Cambridge,Mass. Dow, S. 1963. "ThreeAthenian Decrees,"HSCP 67, pp. 55-75. Habicht, C. 1977. "Athenisches Ehrendekretvom Jahredes Koroibos(306/5) fur einen koniglichen Offizier,"AJAH2, pp. 37-39. .1988. "HellenisticAthens and Her Philosophers,"David Magie Lecture,Princeton (reprintedin C. Habicht,Athen in hellenistischer Munich Zeit: GesammelteAufsatze, 1994, pp. 231-247,357-358). der .1995. Athen:Die Geschichte Stadtin hellenistischer Zeit, Munich.

1997. Athensfrom Alexander to Antony, Cambridge,Mass. Harris,W. V. 1989.Ancient Literacy, Cambridge,Mass. Hondius,J. J. E. 1925. Novae InscriptionesAtticae, Leiden. Koumanoudis,S. N. 1986. "(c&ciU? A&xcLO'ou Horos 4, Kupvjvoclog," pp. 11-18. Meritt, B. D. 1936. "The Seventh Metonic Cycle,"Hesperia 5, pp. 201-205. Osborne,M. J. 1981-1983. Naturalization in Athens I-IV, Brussels. PA = J.Kirchner,ProsopographiaAttica, Berlin 1901-1903. Tracy,S. V. 1975. The Lettering of an Athenian Mason (Hesperia Suppl. 15), Princeton. 1995. Athenian Democracy in Transition:Attic Letter-Cutters of 340 to 290 B. c., Berkeley.

StephenV Tracy OHIO

UNIVERSITY

STATE

OF GREEK

DEPARTMENT 4I4

UNIVERSITY

COLUMBUS,

OHIO

[email protected]

HALL 432IO

AND

LATIN

HESPERIA

Pages

69,

2000

235-26S

AND DACK<-AANTL[ PE THE

PLO$ SPECIAL

MAIDENS FUNERARY

COSTUME

OF GREEK

IN 4TH-CENTURY AND

VOTIVE

RELIEFS

ABSTRACT The distinctivecostumeof back-mantleandpeplosappearsmost often on standingfemaleson LateClassicalfuneraryreliefs.Thesemaidensareintentionallyset apartfromotherfemalesin groupsceneson gravereliefs aswell scenesonvotivereliefs.A decreeof 422/1 B.C.providesthe asin processional earliestexampleof the costume,wornbyAthena,whomthe maidensappear to emulate.Mythologicalmaidensapproaching marriage,suchas Hebe and this The monuments also wear costume. suggestthatthe maiden Deianeira, statusdressedin thiscostumeoccupieda specialplacewithin of marriageable familiesandsocietyin ClassicalGreece. In the ancientworld,costumewas an all-importantindicatorof statusand socialstanding,for clothingsignifiedunofficialaswell as officialmembership in a group.1Distinctivegarmentsinformus not only aboutthe characteristicsof individualfiguresbut alsoaboutthe relationshipsamongfigures.The specialcostume-back-mantle and peplos-that identifiesand characterizesAthenianmaidensin the Late Classicalperiodis the focus of this study.The consistencyin the typeof figuresshownin this costume, as well as the high qualityof the monumentson which they aredepicted, indicatesthat these youngwomenwere easilyidentifiableand important to society.They are the parthenoicelebratedin myth and cult, girlsjust pasttheirchildhoodand on the thresholdof marriage. Maidensin back-mantleand peplosappearon sixtyfunerarymonuments and eight votive reliefs.The monumentsdate from an important 1. It is with great pleasurethat I thank the American Philosophical Society for researchgrants to work on this materialand the American School of ClassicalStudies at Athens for assistanceand cooperation.I also thank the anonymousreviewersfor Hesperia for their useful observations.Appreciation is due to many museum staff memberswho enabledme to view the

sculptures,often under difficult circumstances,at the National Museum and the AcropolisMuseum, Athens; ArchaeologicalMuseum, Piraeus; British Museum, London; Metropolitan Museum of Art, New York;and Musee du Louvre,Paris.I thank Carol Lawton and Olga Palagiafor their helpful comments afterreadinga draft of this work, Nancy Winter for research

assistance,and especiallyEvelyn B. Harrison,who inspiredme to finish it. Although this study is based on my 1986 New YorkUniversitydissertation for the Institute of Fine Arts, "The Shoulder-PinnedBack-Mantle in Greek and Roman Sculpture,"there is a large shift in emphasishere and much new materialis presented.

236

LINDA

JONES

ROCCOS

but not well-defined period of Greek art, the middle decades of the 4th century (ca. 370 to 310 B.C.). Nearly all of the monuments come from Attica-Athens, Brauron,Eleusis, and Piraeus or from placeswith strong ties to Athens. Identifying examples of the costume is difficult; early photographs of sculpture often display only a full frontal view with flat lighting that obscures the back-mantle. Moreover, the garment was not considered distinctive or significant by earlierresearchers.Not only Margarete Bieber's works on Greek costume2 but also more recent works, such as those by Georges Losfeld, Elsa Gullberg,and AnastasiaPekridou-Gorecki,3 only briefly mention the costume considered here. Studies show that when we isolate the distinguishing characteristics of specific garments, we can learn much about social and economic roles in ancient Greece. Elizabeth Walters has demonstrated that women who were initiates in the cult of Isis associated themselves with her by wearing a garment of Egyptian type.4Although it is not known whether the maidens in back-mantle and peplos representedany particularcult, they appear to associate themselves with the maiden goddesses Athena and Artemis, who are also represented in this costume.5 Christiane Sourvinou-Inwood and, recently, Karen Stears suggest that the young women who wear the back-mantle and peplos are teenagers.6 Other scholars have explored the ritual and cultic significance of certain garments. Margaret Miller's study of the ependytesshows that it was worn as a status symbol at festivals.7For example, a young maiden shown as a kanephorosin a ritualprocession on a red-figure kraterby the Kleophon Painter wears an ependytes.8Nancy Serwint identifies the exomisas a dress worn by female athletes in initiation rites at the Heraia in Olympia.9Evelyn Harrison suggests that a figure wearing a shoulder-cord over a chiton may be Themis, but Olga Palagia identifies the same figure as Demokrateia or Agathe Tyche.10 Some works focus on the identifying nature of specific garments; the dress of the Archaic korai, for example, is explored by Harrison, Brunilde Ridgway, and Judith Schaeffer.1"Karin Polaschek studies the wrapped himation worn by men, and A. Geddes investigates the clothing worn by Athenian men.12Harrison notes that groups of horsemen on the Parthenon frieze can be distinguished by garments that identify them as members of Kleisthenic Attic tribal units.13 Studies such as these, as well as recent works on the low-belted chiton by Hannelore Winkler or the thickly rolled himation by Axel Filges,14also illustrate how garments characterize the wearer.For example, youthful goddesses such as Flora and Nymphs wear the low-belted chiton, while more mature figures like Persephone wear the himation wrapped tightly across the breast. Young women in back-mantle and peplos appear most often among the standing females on Late Classical grave reliefs published first by Alexander Conzet5 and in the two- or three-figure groups described recently by Christoph Clairmont.16Some of these maidens are also included in studies of funerary monuments by Hans Diepolder and Knud Friis Johansen (stelai),BernhardSchmaltz (stelai and lekythoi), and Gerit Kokula (loutrophoroi).17Two recent studies of 4th-century grave stelai, by Andreas Scholl on the "'Bildfeldstelen''18 and Johannes Bergemann on the naiskos

2. Bieber (1928) discussescostume garmentby garmentwith examplesof modern reconstructions(peplos, pp. 77-82; himation,pp. 82-90); Bieber and Eckstein 1967 is a briefer study by chronologicalperiod (Classical period, pp. 32-34), with examplestaken mostly from ancient sculpture.For recent bibliographyon Greek dress,see Losfeld 1991, pp. 370-399; PekridouGorecki 1989, pp. 138-154, for notes as there is no bibliography;and Lee 1999, pp. 558-596. I am also currently preparingan annotatedbibliography. 3. Losfeld (1991) considersthe textualevidence for dress;he includes a useful list of 336 terms relatedto Greek dress,pp. 327-339, but none seem to applyto the back-mantle.See also Gullbergand Astr6m 1970 for extremelybrief descriptions;PekridouGorecki 1989 for some useful drawings; and Losfeld 1994 for examplesin art, but very few drawings. 4. Walters 1988. 5. For Athena, see LIMC II, 1984, p. 977, nos. 220-230, pls. 729-730, s.v. Athena (P. Demargne).For Artemis, see LIMC II, 1984, p. 636, nos. 125133, pl. 454, s.v.Artemis (L. Kahil). 6. Sourvinou-Inwood1988; Stears 1995. 7. Miller 1989. 8. Ferrara44894 T57C:ARV2 1143, no. 1;Addenda2334; LIMC II, 1984, p. 220, no. 303, pl. 208, s.v.Apollon (W. Lambrinoudakis);Miller 1997, p. 159, fig. 68. 9. Serwint 1993. 10. Harrison1977; Palagia1982; 1994. See Pekridou-Gorecki1989, p. 97 for a diagramof this costume. 11. Harrison1991; Ridgway1997; Schaeffer1975. See also Richter 1968, pp. 6-13, for the costume. 12. Polaschek1969; Geddes 1987. 13. Harrison 1984, 1, pp. 230-233; 1989, p. 49. 14. Winkler 1996; Filges 1997. 15. Conze 1922, IV, nos. 803-907. 16. Clairmont 1993, Introduction, pp. 32-33. 17. Diepolder 1931; FriisJohansen 1951; Schmaltz 1970. See also Schmaltz 1983 and Kokula1984. 18. Scholl (1996, p. 121) includes only five panel stelai (picturepanels

BACK-MANTLE

AND

PEPLOS

237

stelai,19include several of these maidens in their catalogues but have little to say about the costume. In studies of votive reliefs, particularlyby Ulrich Hausmann and GerhardNeumann,20 the maidens rarelyreceive more than a note. In general studies of 4th-century monuments, very little attention is paid to funeraryand votive reliefs.21These reliefs, however, depict valuable images of the world of ordinary citizens in the Classical period. The back-mantle and peplos set the wearer apart from other females in processional scenes on votive reliefs and in group scenes on grave reliefs. The maiden wearing this costume was important enough to her family to merit her own funerarymonument. Stears points out that in Conze's corpus of grave reliefs, 168 monuments depict men alone, while 176 depict women alone,22twenty to thirty of which portray a solitary young woman wearing the back-mantle and peplos. Some fragmentarystelai may also be part of this group. Despite their frequent appearanceand distinctive costume, these maidens have not previously been studied as a group.

THE COSTUME BAC K-MANTLE

The shoulder-pinned back-mantle is the unifying and distinctive identifying characteristic of all the maidens who wear it over the Attic peplos, which has the belt on top of the overfold (see below for the traditional distinction made between the "Attic"and "Argive"peploi).23Only one person in any given group wears the back-mantle and peplos, which mark her as special.It is this combination of garments,back-mantle and Attic peplos, rather than the back-mantle alone, that distinguishes this costume. The back-mantle first appearsin the late 5th century B.C., as the costume of the maiden goddesses Athena and Artemis, chiefly on Attic monuments. The mantle pinned on the shoulders may have been inspired by the with low relief images set into a rectangularpanel) with maidens in back-mantle,out of 528 catalogued examples(pp. 223-364). He places the maidensin his discussionof children, pp. 114-124, but devotes only one page to them and calls the back-mantlea or back-veil. "Ruckenschleier," 19. Bergemann(1997, pp. 85-86) lists only twenty-seven naiskos stelai (high relief figureswith side antae and pediment or entablature)with maidens in back-mantle,out of 865 catalogued examples(pp. 158-179). He devotes a very brief space to the maidens. 20. Hausmann 1948; 1960; Neumann 1979. There is no corpusof Attic votive reliefs as yet, and no recent work on the Asklepios votive reliefs that depict the girls in back-mantle.

Olga Palagiais preparinga study of the votive reliefs from the Athenian Acropolis and Carol Lawton a study of the votive reliefs from the Athenian Agora. 21. Ridgway1997: funeraryreliefs, pp. 157-192, girls with back-mantle, pp. 169-179; votive reliefs,pp. 193236. Boardman(1995, pp. 114-142) discussesvery few of either type of relief. 22. Stears 1995, p. 113. 23. Three maidenswithout a backmantle who appearon fragmentary stelai may also be consideredbecauseof their similaritiesto other monuments and becauseClairmont(1993, VI, p. 128) includes them in his list of maidensin back-mantle:a stele in London, British Museum 1915.4-16.1

(Clairmont1993,111,no. 3.414a; Bergemann1997, p. 171, no. 500) like the Mantua stele; the fragmentarystele Piraeus 1778 (Clairmont 1993, 1, no. 1.307; Bergemann1997, p. 178, no. 751) like Silenis;and Hagnostrate on Athens, National Museum 1863 (Clairmont 1993, 1, no. 1.431; Bergemann1997, pp. 66, 174, no. 619, pl. 117:4) becauseher mantle could have been added in paint as suggested by Karouzou1979 for some votive reliefs.For color,see also Schmaltz 1970, pp. 60-75; 1983, pp.71-81; Kokula1984, p. 359; Scholl 1996, pp. 185-200. Other maidens on fragmentary stelai could have worn the backmantle as well, such as Clairmont 1993, I, nos. 1.294 (Bergemann1997, p. 173, no. 574) and 1.310.

238

LINDA

JONES

ROCCOS

Eastern method of wearing the Oriental kandys;24a late-5th-century calyx krater in Berlin shows it hanging from Andromeda's shoulders.25The popularityof Oriental paraphernaliain the later 5th centuryhas been noted by Miller, who interprets the adoption of such foreign luxury items as a display of elite status. Miller also notes the recent trend "towardsrecognition of the importance placed on clothing as indicators of status by the Greeks."26 The ancient name of the back-mantle is uncertain. Losfeld cites several types of mantles, none of which seem to be the same as the pinned back-mantle.27Clairmont's choice of the term "mantlet"is particularlyunsatisfactory.The term implies a small size, but many of the maidens'backmantles extend below their knees and have an overfold.28 The term epiporpema(literally, "pinned on") seems not to refer specifically to garments worn by women; it could as well refer to a male chlamys pinned in front.29Whatever its originalGreek name, the pinned back-mantleis clearly recognizable as a garment reserved for maidens. The pinned back-mantle is the same rectangularwoolen cloth common to all Greek mantles but folded with one side shorter than the other, ratherlike a peplos. It is also, like the peplos, pinned on the shoulders.The back-mantle falls from the shouldersdown the back to just below the knees, with an overfold about halfway down. It is best observed on a freestanding figure such as the statue of a maiden in New York (Fig. 1).3? The earliest example of the pinned back-mantle with belted peplos is worn by Athena on the Rheitos Bridge decree of 422/1 B.C. in Eleusis (Fig. 2).31Athena wears the back-mantle over an Attic peplos, with a small

| ..

I

-!

91

_|

E~~'p

I l 1I

24. For the kandys,see Miller 1997, p. 166, where she points out that in Persiathe kandysis worn hanging from the shoulders,but in Greece is normally worn like a jacket, as the attendanton the Princetonstele wears it, and as Myttion on the stele in the J. Paul Getty Museum wears it: Miller 1997, fig. 91 and Kingsley 1975; see also Knauer1978. 25. Berlin 3237: ARV2 1336; LIMC I, 1981, p. 416, no. 21, s.v.Aithiopes (F. SnowdenJr.) and p. 776, no. 8, s.v. AndromedaI (K. Schauenburg);LIMC VI, 1992, p. 7, no. 5, pl. 8, s.v. Kepheus I (K. Schauenburg);Miller 1997, fig. 82; see also Linders 1984. 26. Miller 1997, p. 183. See also Bieber and Eckstein 1967, pp. 20-21. 27. Losfeld 1994, pp. 207-214. 28. Clairmont1993, Introduction, p. 32. His index listing for "mantlet" (1993, VI, p. 129) is useful and the following 52 monumentsare included here: 1.256, 1.267, 1.268, 1.280, 1.294, 1.310,1.312,1.321a, 1.329,1.359, 1.382, 1.428, 1.433, 1.459, 1.774, 1.783,1.814, 1.827, 1.839, 1.840,

1.862,1.883, 1.932,1.938,1.943, 2.292a, 2.334, 2.344a, 2.345a, 2.357c, 2.362c, 2.383b, 2.395c, 2.417a, 2.421, 2.434b, 2.436,2.470,2.825,3.337, 3.339a, 3.340a, 3.387a, 3.394b, 3.413a, 3.543, 3.860, 3.870,3.880,4.381, 4.420,4.830. 29. Pollux (10.190) calls it the garmentof the kitharode;see also LSJ, "garmentbuckledover the shoulders, cloak, mantle,part of the dressof the musician.

30. New York,Metropolitan Museum of Art 44.11.2: Clairmont 1993, I, p. 513, no. 1.971; Roccos 1995, p. 663, fig. 23; Bergemann1997, p. 177, no. 703. I thankJoan Mertens for providingaccess to study and measure this monument. For similarstatues,see below. 31. Rheitos Bridge decree,Eleusis Museum 5093: IG 12 81; IG IP79; LIMC II, 1984, p. 1013, no. 606, pl. 763, s.v.Athena (P. Demargne); Boardman1985, fig. 178; LIMC IV, 1988, p. 881, no. 446, s.v. Demeter (L. Beschi);Meyer 1989, p. 266, no. A5; Mangold 1993, p. 19, pl. 2:1;

Figure 1 Stteo ~ ~. mie _ Meroolta York Muemo .1.2 Corts Museum

e r

BACK-MANTLE

AND

PEPLOS

239

Figure2. RheitosBridgedecree. EleusisMuseum5093. Courtesy Deutsches Arch,iologischesInstitut,Athens

collar aegis and an Attic helmet. Since there are other 5th-century relief images and three later freestanding versions of this type, these representations may reflect an original statue of Athena created ca. 425 B.C., shortly before the Rheitos Bridge decree.32 The Erechtheion caryatids, dated soon after the Rheitos Bridge decree, also wear the pinned back-mantle,3 but coupled with the so-called Argive peplos worn by the maidens on the east frieze of the Parthenon. In contrast to the Attic peplos, the Argive peplos is belted underthe overfold, obscuring the belt. A girl who wears the same back-mantle and peplos as the caryatidsappearsat about this time on an Attic white-ground lekythos by the Woman Painter in Karlsruhe.34 In addition to Athena, Artemis is often shown wearing the back-mantle on votive reliefs, particularlyfrom Brauron.3sThis costume is appropriate Lawton 1995, p. 82, no. 3, pl. 2. Clinton (1992, pp. 75-76) identifies the male figurein the dexiosismotif with Athena as Eumolpos, founderof the EleusinianMysteries;Athena, as "founder,"or Archegetis,then may wear the back-mantle,as suggestedin Roccos 1991, p. 407. 32. Roccos 1986. For a later statue of Athena with a back-mantle,see Acropolis 1336: Brouskari1974, p. 21, figs. 8-9; Roccos 1991, p. 399, pl. 112:a. For Athena with back-mantle

on votive reliefs,see Mangold 1993, pp. 19-21; Lawton 1995, pp. 41-42. 33. Ridgway1981, figs. 82-83; Boardman1985, fig. 125; Scholl 1995, pp. 196-212. The caryatid'sbackmantle can be seen best on the Hadrianiccopies at Tivoli, where more of the edges along the sides are preserved: Scholl 1995, p. 200, figs. 14:a-c. Several maidens in this costume,which Clairmont consideredin his corpus,are not discussedhere, see Clairmont 1993, IIII, nos. 1.152, 1.967,2.207, 2.334b,

3.340; that costume with back-mantle and Argive peplos is best seen as a variantof the more popularbackmantle and Attic peplos costume. 34. Karlsruhe,Badisches Landesmuseum B1528: ARV2 1372, no. 17; Addenda2370; Scholl 1995, p. 207, fig. 18. 35. BrauronMuseum 1157: LIMC II, 1984, p. 668, no. 621, pl. 495, s.v. Artemis (L. Kahil). BrauronMuseum 1182: LIMC II, 1984, p. 658, no. 463, pl. 483, s.v.Artemis (L. Kahil).

240

LINDA

JONES

ROCCOS

Figure 3. Bodle relief. Athens,

NationalMuseum 1473. Courtesy Museum

........

for Artemis, the maiden goddess with a special relationship to young girls and unmarried young women; the Brauronian Arkteia was a festival for very young girls.36 Older girls dedicated their belts (and toys) to Artemis Lysizonos before marriage,on a red-figure lekythos in Syracuse a maiden wearing a peplos unties her belt before Artemis. Tullia Linders studied the inventories of dedications to Artemis Brauroniaon the Acropolis and found that the overwhelming majonitywere women'sgarments,as Pausanias (1.23.7) had recorded.38 A type of pinned back-mantle without an overfold can be seen on images of Apollo Patroos in Athens and Apollo Kitharoidos in Delphi.39The pinned back-mantle over the belted peplos appears not only in Attica but in places influenced by Athenian art: Cyrene in North Africa and Gortyn on Crete.40 36. Sourvinou-Inwood1988, p. 15. 37. Van Straten1981, pp. 90-91, who cites Anth. Pal. 6.280, and fig. 30, red-figurelekythosby the Achilles Painter,Syracuse21186: ARV2993, no. 80;Addenda2312; LIMC II, 1984, p. 676, no. 721a, pl. 504, s.v.Artemis (L. Kahil). 38. Linders 1972, pp. 2-3, 11-13. 39. Apollo Patroos,Athens, Agora S 2154: LIMC II, 1984, p. 204, no. 145,

pl. 195, s.v.Apollon (0. Palagia); Stewart 1990, fig. 512; Boardman1995, fig. 30. Apollo, Delphi Museum 11876: LIMC II, 1984, p. 204, no. 145f, s.v. Apollon (0. Palagia);Boardman1995, fig. 14:2.The reasonsfor Apollo wearing a version of the back-mantle have not been filly explained;perhaps it servedas a connection to Athens, as well as a contemporarymusician's festivalcostume.

40. Artemis, Cyrene 14377: LIMC II, 1984, p. 636, no. 127, pl. 454, s.v. Artemis (L. Kahil).Apollo, Cyrene 14095 and 14097 (now lost): LIMC II, 1984, p. 204, no. 145e, h, k, s.v.Apollon (0. Palagia).Artemis from Gortyn, HerakleionMuseum 39: Kabus-Jahn 1972, p. 91, figs. 12-15. Apollo from Gortyn, HerakleionMuseum 326: LIMC II, 1984, p. 204, no. 146c, s.v. Apollon (0. Palagia).

BACK-MANTLE

AND

PEPLOS

24I

Two figuresof Athenaillustratedifferentmethodsof representingthe pinnedback-mantlein the ClassicalandLateClassicalperiods.The backmantleon Athenain the RheitosBridgedecreefromthe 5th centuryhas zigzag side folds. Shortlyafter340 B.C., the back-mantleis shownwith curvilinearside folds, as on figuresof Athena on documentand votive reliefs,suchas the Boule reliefin Athens (Fig. 3), as well as on laterstatHouse Athena" ues.4'I havenamedthe 4th-centurytype the "Areopagus afterthe findspotof a well-preservedRomanversion.This type recallsin a deliberatelyclassicizingmannerthe dressof two 5th-centuryexamples: theback-mantleof the RheitosAthenaandthe largeaegisof the Parthenos. In style,the AreopagusHouse Athenais close to some of the latermaidens with back-mantle,suchasTheophile(20, Fig. 4). Contemporarywith the AreopagusHouse Athenatype is a statueof Themis or Demokrateia in the AthenianAgora,aboutwhich Palagiawrites:"itsclassicismis conThe AreopagusHouse trivedand heraldsthe end of the Classicalera."42 Athenais a similarlyretrospectivecreationof the Lykourganperiodin the third-quarterof the 4th centuryB.C., the era characterizedby Fordyce Mitchel as recreatingthe Perikleanera in a self-consciouslydeliberate manner.43It is in this period that most of the images of maidens in back-

Figure 4 Stele Theophile (20). Athens, National Museum 1305.

mantleandpeplosoccur. or epiblemaA morecommontypeof mantle-the shoulder-mantle, is simplythrownoverthe shouldersratherthanpinnedon.The shouldermantleis generallywornwith the Argivepeplosby morematronlyfigures: Herain 5th- and4th-centurydocumentreliefs;"Leto on a votivereliefin Athens andas a Romanstatuecarryingthe infantsApollo andArtemis;45 Demeter on a documentreliefand as a Romanstatuein the Capitoline Museum;46and the statue of Eirene with Ploutos in Munich.47It is some-

CourtesyMuseum

timespulledoverthe headandis generallyshorterin backthanthe shoulder-pinnedback-mantle.Biebercombinesthese mantlesinto a singlecategory,called"theshoulder-back-mantle."48 In the LateArchaicandEarlyClassicalperiods,the shoulder-mantle waswornby certainyoungwomen,perhapsto signifya traditionalfestival

41. Boule relief,Athens, National Museum 1473: Palagia1982, p. 109, pl. 36:d;LIMC III, 1986, p. 146, no. 1, pl. 125, s.v. Boule (V. Komninos); Meyer 1989, pp. 303-304, no. A136, pl. 41:1; Roccos 1991, esp. p. 409, no. 4, pl. 109 and pp. 408-410 for a list of twelve examples;Lawton 1995, p. 143, no. 143, pl. 76. For the votive reliefs, see Mangold 1993. 42. Athens, Agora S 2370: Palagia 1982, p. 108, pls. 29-30; with it she comparesthe better dated reliefs, Athens, National Museum 1335 (pl. 31:c, ca. 330-325) and 1476 (pl. 33:b, 331/0). See also LIMC III, 1986, p. 373, no. 8, s.v. Demokratia (0. Alexandri-Tzahou);Stewart 1990, fig. 575; Ridgway1990, pp. 54-56, pl. 29; Todisco 1993, pl. 156;

Boardman1995, fig. 51; Ridgway1997, p. 339. See also above,note 10, for other identificationsof this figureby Harrison and Palagia. 43. Mitchel 1970. 44. Hera on AcropolisMuseum 1333 (403/2): Meyer 1989, p. 273, no. A26, pl. 10:1; Lawton 1995, p. 88, no. 12, pl. 7; LIMC II, 1984, p. 1013, no. 607, pl. 763, s.v.Athena (P.Demargne); LIMC IV, 1988, p. 691, no. 296, s.v. Hera (A. Kossatz-Deiss-mann);and on Athens, EpigraphicalMuseum 7862 (399/8): Meyer 1989, p. 273, no. A27, pl. 10:2; Lawton 1995, p. 89, no. 13, pl. 7; LIMC IV, 1988, p. 691, no. 297, pl. 421, s.v. Hera (P. Demargne). 45. Leto on Athens, National Museum 1389: LIMC II, 1984, p. 265, no. 657, pl. 237, s.v.Apollon

(C. Daumas);Leto, Rome, Museo Torlonia68: LIMC II, 1984, p. 424, no. 439, pl. 337, s.v.Apollon/Apollo (E. Simon);LIMC VI, 1992, p. 259, no. 25, pl. 132, s.v. Leto (L. Kahil); Todisco 1993, pl. 209. 46. Demeter on the Rheitos Bridge decree:see above,note 31. Demeter, Rome, Museo Capitolino 642: LIMC IV, 1988, p. 852, no. 55, pl. 566, s.v. Demeter (L. Beschi). 47. Eirene,Munich, Glyptothek 219: LIMC III, 1986, p. 703, no. 8, pl. 541, s.v. Eirene (E. Simon); Stewart 1990, figs. 485-486; Todisco 1993, pl. 96; Boardman1995, fig. 24. 48. Bieber 1977, pp. 104-117. For dedicationsof epiblemataon the Acropolis, see Linders 1972, p. 13.

242

LINDA

JONES

ROCCOS

garment. An unusual Archaic kore from Attica known as the "BerlinGoddess"wears a long shoulder-mantle.49A few of the Archaic korai from the Acropolis also wear their mantles this way rather than in the more usual diagonal form.50The shoulder-mantle is worn with the Argive peplos by the maidens in the Parthenon east frieze.5'A veil or mantle over the Attic peplos is uncommon but appears on a mid-5th-century marble statuette of a young woman in Athens.52Similarly,the kanephoroson the Kleophon Painter's red-figure krater in Ferrarawears the long shoulder-mantle; in an earlier study I have referredto this mantle as a "festivalmantle."53 In vase painting,mythological maidenswear the shoulder-pinnedbackmantle. On an Attic red-figure kraterin Karlsruhe,Hebe wears the backmantle.54Hebe also wears this mantle on a red-figure bell krater in the Villa Giulia, where she raises the sides of the back-mantle with both of her hands, a gesture unknown in the sculpture of maidens.55Deianeira on a red-figure pelike by the Meidias Painter in New York makes a similar gesture.56Clearly, these two young women are soon to be brides. Only a few of the maidens on the funerary monuments, such as Plangon on the loutrophoros in Eleusis (46, Fig. 5), raise one side of the mantle with their These maidhand, in the unveiling gesture of brides for the anakalypteria.57 ens may have been closer to their wedding day when they died. Like the Erechtheion caryatids,most of the 4th-century maidens hold the edge of the mantle with their lowered hand. Semni Karouzou suggests that the motif of holding the edge of the mantle came from a famous source, possibly the Erechtheion caryatidsthemselves.58Holding the edge of a garment seems to us a rather feminine mannerism, but it may have been a sign of youth in antiquity. Youths and boys on grave stelai often hold one edge of their mantle, as Aristion and Stephanos do on two 4thcentury stelai in Athens.59The youngTriptolemos on the Great Eleusinian Relief in Athens also holds the edge of his mantle with one hand, but in a more naturalistic manner.60 49. Berlin 1800: Richter 1968, pp. 39-40, no. 42, figs. 129-140; LIMCII, 1984, p. 15, no. 54, pl. 9, s.v. Aphrodite (A. Delivorrias). 50. Richter 1968, pp. 8-9. Acropolis 671: Richter 1968, pp. 70-71, no. 111, figs. 341-342. Acropolis 688: Richter 1968, pp. 102-103, no. 184, figs. 587590. See also Losfeld 1994, pp. 212214, for koraiwith their mantles pinned on both shoulders,hanging down in front as well as in back more like a peplos, as on Acropolis 673 and 678. 51. Brommer1975, pl. 186; Boardman1985, fig. 96:15; Stewart 1990, fig. 346. 52. Athens, National Museum 1848: Tolle-Kastenbein 1980, p. 54, no. 9d, pls. 40-41. 53. See above,note 8; and Roccos

1995, p. 649, fig. 3. 54. Karlsruhe,BadischesLandesmuseum259 (B 36):ARV2 1315, no. 1; CVA,Karlsruhe1, pls. 22:4 and 23 (320-321). Athena too wears the backmantle on this vase;see also LIMC I, 1981, p. 499, no. 12, pl. 378, s.v. Alexandros(R. Hampe);LIMC II, 1984, p. 992, no. 412, s.v.Athena (P. Demargne);LIMC IV,1988, p. 126, no. 2, s.v. Eutychia (H. A. Shapiro); LIMC VI, 1992, p. 71, no. 1, pl. 35, s.v. KlymeneVI (A. Kossatz-Deissmann). For Hebe, see Laurens1987; Hebe may also be wearing the back-mantle on LouvreG508 (Laurens,pp. 68-69, fig. 16) and Paris,Musee Rodin (Laurens,fig. 17). 55. Rome, Villa Giulia 2382: ARV2 1339, no. 4;Addenda2 367. 56. New York,MetropolitanMu-

I

I

Figure 5. Loutrophoros, Plangon (46). Eleusis Museum 5098. Courtesy Deutsches ArchaiologischesInstitut, Athens

seum of Art 37.11.23: ARV2 1313.7; Addenda2362; LIMC IV, 1988, p. 835, no. 1681, pl. 559, s.v. Herakles

(J. Boardman). 57. Oakley 1982; Oakley and Sinos 1993, pp. 25-26; Reeder 1995, pp. 123129. 58. Karouzou1979, p. 113, note 15. 59. Aristion'sstele, Athens, National Museum 4487: Clairmont 1993, 1, no. 1.855; Boardman1995, fig. 126; Bergemann1997, p. 167, no. 337, pl. 91:4. Stephanos'sstele, Athens, National Museum 2578: Clairmont 1993, , no. 1.214. 60. Athens, National Museum 126: Boardman1985, fig. 144; LIMC IV, 1988, p. 875, no. 375, pl. 588, s.v. Demeter (L. Beschi); Clinton 1992, figs. 1-2; I thank 0. Palagiafor reminding me of this figure.

AND

BACK-MANTLE

..... .. ....

....

.. .....

....

...

... ....

........ .. .... .... ..... ..... ....

.. ....

....

... ..

... ....

. . . .. ....

. .. ...

.. ..............

....

.....

... .. . .......

........

......

...

... . ..

.

:. ::.

:. - . ::: - . :. - :. . - ::s - . :. -: :: -: : -- . :-::. ....... .... . . . . ....... .. .......... .. ....

... ..... ..... .......... ...... ............

..

....

. .......... .........

.

.... . ...

--:

. . :: - . :f -::. : . ::.

PEPLOS

::: :n :; ! ::: : ...

...

... ....

243

..

..

....... ..

.... .....

.. . . . ......

, = :,::;, :f;:n. ... ..... ...fs;:.....

.... .............

.

....

....

. .. ... .. ..

.. ....... zf'

............. . .. . .... . ....

------------

.......... ---- ---- - ---- ---------

QT

.... ......

Ut, f

W 1.

..... .........1

.......... .... ..... m v:a

VZ

.....

.... ... .

... ....

.... ........

.. .... .........

... ... . ... ...... .

...

...

.

--us-

.. .

........ ..

Kir-

. ... .... ...

........

U !, N z t f

.... .....

....

Ef

s

s;--.::t.

.........

.... .................

................

..........

MU4s n4

s

sfi

.10

....

..

.. ....

..

.. . ...

.. ...

...

Figure 6. Votive relief from Athens, Asklepieion (61). Athens, National

Museum1333. CourtesyMuseum

61. Athens, KerameikosMuseum XV, 8166, 8185, 13443: Kerameikos p. 120, no. 361, pl. 64; other examples of a girl with back-mantleappearin that volume:p. 46, no. 139, pl. 38 (a girl dressedlike the Erechtheion caryatids);p. 119, no. 357, pl. 63 (belted peplos);p. 120, no. 362, pl. 64 (fragment).

Sometimes maidens hold the sides of their mantles with two hands, as do all of those shown on votive reliefs (see, e.g., 61, Fig. 6). Yet only a few of the maidens on grave reliefs, those who hold no objects,grasp the mantle with both hands; see, for example, 59, the maiden on a relief base in Athens (Fig. 7). The motif of holding the mantle with both hands seems to occur later rather than earlier,after the mid-4th century. In other media, terracotta statuettes often portray a maiden wearing the Attic peplos with back-mantle. One of the finest examples of a girl in belted peplos with crossbandswho holds her back-mantle with both hands comes from the Athenian Kerameikos (Fig. 8).61 Depictions of girls wearing back-mantle and peplos on plastic vases are very similar to those in monumental sculptures.62Nike, too, wears a peplos and back-mantle that seem to blend in with her wings on a fine plastic lekythos in Athens.63The maiden in back-mantle and peplos also appears in 5th- and 4th-century terracottastatuettes as a kanephoros,carrying a basket on her head. A late4th-century statuette in Karlsruhe may have come from Athens, for it is statuette found in the Athequite similarto fragmentsof a small kanephoros 62. Paris,LouvreMNC 630: Trumpf-Lyritzaki1969, p. 8, no. 11. Brussels,Musee RoyalA892: CVA, Malibu 3, pl. 1 (135), no. 6a-d; Trumpf-Lyritzaki1969, p. 9, no. 16. Piraeus,ArchaeologicalMuseum 692: Trumpf-Lyritzaki1969, p. 8, no. 12. A furtherexample,Trumpf-Lyritzaki 1969, p. 9, no. 15, is now lost.

63. Athens, National Museum 2076: Trumpf-Lyritzaki1969, p. 7, no. 10, pl. 3:a-b; LIMC VI, 1992, p. 861, no. 124, pl. 571, s.v. Nike (A. GoulakiVoutira).See also Athens, Kerameikos T 482: Trumpf-Lyritzaki1969, p. 9, no. 13, pl. 4:a;and one in the market, p. 9, no. 14.

LINDA

244

JONES

ROCCOS

7. Reliefbase (59). Athens, ~~~~~~~~~~~~~~~~~~~Figure

-

First Ephoreia(FethiyeDjami). -

Courtesy Deutsches Archaologisches

Institut, Athens

.

.

?.

l

l

e

<

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i, 5

4The back-mantle is worn only by mortal maidens and maiden nian Agora.M

goddesseson votivereliefsandvotiveterracottastatuettes. PEPLOS The belted peplos seen on nearly all the maidens is that worn by Athena, especially the Parthenos,6 and also by other maiden goddesses such as Artemis and Persephone.66i peplos first appearsin the Early Classical period on the SevereStyle figure"Angelitos'sAthena"and on the "Mourning Athena"stele, both in the Acropolis Museum.6 Severe Style figures nearly alwayswear the peplos, either Attic or Argive; the style that takes its name largelyfrom this garmenthas been thoroughlyinvestigatedby RenateTdlleKastenbein.6 Mireille Lee presents the literary and archaeological evi64. Karlsruhe,Badisches LandesmuseumB2662: Schiirmann1989, pp. 40-41, no. 86, pl. 18. Athens, Agora T 101 andT 104:Thompson 1954, p. 106, no.4, pl.20 and p. 107, no.9, pl. 22; Nicholls 1995, p. 423, note 72: the similarityin type to the figure in the Agora mold T 4064 (p. 478, no. 11) should identify that figure as a kanephoros also;the costume is like that on 5th-century depictionsof Athena; see above,notes 31-32. 65. For the meaning of "peplos"as referringspecificallyto the Parthenos dress from the mid-4th centuryon, see Mansfield 1985. See also Bieber 1928,

pp. 34-37, pls. 4-5; 1977, pp. 89-93, figs. 386-408; Pekridou-Gorecki1989, pp. 80-81, figs. 52-53; Barber1992; Lee 1999, pp. 218-238. 66. Artemis, see above,note 35; Bieber 1977, pp. 89-90, figs. 386-390. Persephone,Eleusis Museum 64: Mylonas 1961, fig. 75; Kanta 1979, pp. 62-63, fig. 18; Neumann (1979, p. 59, pl. 35) determinedthat the figure should representPersephone,not Demeter. Neumann, however,refersto that mantle as a "Ruckenmantel," which is usuallyreservedfor the pinned back-mantleconsideredhere. See Ridgway1981, fig. 96, and Boardman

1985, p. 179, fig. 137, however,who both referto the statue as Demeter, albeit in quotes or with a question mark. 67. Angelitos'sAthena, Acropolis 140: LIMC II, 1984, p. 972, no. 144, pl. 720, s.v.Athena (P.Demargne). Mourning Athena stele, Acropolis 695: LIMC II, 1984, p. 1015, no. 625, pl. 765, s.v.Athena (P. Demargne). 68. T6lle-Kastenbein 1980; see also Ridgway1977, pp. 8-9. This peplos is sometimes called the Dorian chiton, to contrastthe heavywoolen garment with the voluminousIonian linen chiton.

BACK-MANTLE

R

Figure8. Terracottastatuette. Athens, KerameikosMuseum8166, 8185, 13443. CourtesyDeutsches ArchaologischesInstitut, Athens

AND

PEPLOS

245

dence for the peplos in the Early Classical period and concludes that the peplophoroshad special meaning as a sign of Hellenic identity.69Lee also provides an exhaustive study of the scholarship on the peplos from the Renaissance to the present day, as well as a list of the relevant ancient literary sources. The peplos in its various forms was the dominant mode of dress for women from the Early Classical through the High Classical periods, from about 480 to 400 B.C. At the end of the 5th century and the beginning of the 4th, the thin clinging chiton and himation had a period of immense popularity.Subsequently,there was a revivalof interest in the peplos, which can be observed early in the second quarter of the 4th century on the statue of Eirene, ca. 370 B.C.70 This is the period when the maidens in back-mantle and peplos make their first appearance. The term "Attic peplos" has become a conventional shorthand term referringto the peplos with a belt on top of the overfold; it is retained here for its brevity as well as for its associations with Attic art. The misuse of the term apoptygmafor overfold is regrettable; Dorothy Thompson has pointed out that the term seldom occurs and refers to various parts of drapery,not the entire overfold.71 The Attic peplos could also be called the maiden'speplos in contrast to the "Argivepeplos,"or women's peplos-the peplos with the belt obscuredby the overfold and worn chiefly by Demeter, Eirene, and Hera.72A third form of peplos, without any belt at all, is often called the "Laconianpeplos"and is worn by young girls or women at home.73 Young women wear the peplos (with or without the back-mantle) in their role as attendants to the bride on red-figure and white-ground vases of the later 5th century,particularlythose in wedding or funeraryiconography.The bride wears a chiton and himation, and her attendant usually wears a peplos. In Joan Reilly's catalogue of white-ground lekythoi, nearly half the attendants or maids wear the Attic peplos.74In the 4th century, the Attic peplos is nearly always reserved for maidens, whether portrayed in monumental sculptures,terracottas,or vases. CR0

SSBANDS

The peplos worn by maidens in 4th-century funerarymonuments is often enhanced by the addition of crossbandsworn over the breast on top of the overfold (Fig. 4). A medallion or amulet is sometimes affixed to the center of the cross, an object referredto by Higgins as aperiamma.75 None of the maidens in the votive reliefs wear crossbands, however.

69. Lee 1999, esp. pp. 11-56. 70. See above,note 47. See also Clairmont1993, Introduction,p. 17, for a prototypeca. 370; and Ridgway 1997, p. 339. 71. Thompson (1944, p. 198) explainsthat in the accountsfor the golden Nikai on the Acropolis,the apoptygma is the part of the peplos between the "thorax,"or throatpart

(that above the belt), and the "skele,"or leg part (that below the overfold).See also Harris 1995, pp. 131-132, translatedthere as "draperyfolds,"and Lee 1999, pp. 50-54, for varioususes of apoptygma.For a recent mention of apoptygmafor overfold,see Ridgway 1997, p. 336, and the PerseusProject Encyclopedia, s.v.apoptygma. 72. For Demeter, Eirene, and Hera,

see above,notes 44, 46, and 47. 73. Bieber 1977, pp. 84-86, figs. 321-331; Pekridou-Gorecki1989, pp. 78-79, fig. 51:a-e. See two grave reliefswith young girls in this peplos, Berlin 1482 and New York27.45: Boardman1985, figs. 51-52. 74. Reilly 1989, p. 416. 75. Higgins 1980, p. 167.

246

LINDA

JONES

ROCCOS

Crossbands seem to have various meanings in Greek art.They may be a sign of the wearer'smarriageablestate and may signify the potential fertility of the virginal maidens, orparthenoi. Crossbands in this role seem to have nothing to do with a supporting function, for on a red-figure lebes stand in Athens, the female dancers in skirts wear crossbands over their bare breasts.76Perhaps the crossbands were, like the belt, removed before marriageand dedicated to Artemis. The unusual marble statue of Artemis Kindyas found with the Piraeus bronzes wears crossbands over a large mantle covering her bound hands.77On a late-4th-century pelike in Athens, both the seated bride-to-be and her attendant holding the kanoun wear crossbands.This scene must be the preparationfor the wedding, not the day after,as Karouzou has claimed, for marriedwomen never wear the crossbands.78 Crossbands could also have served a purely practical function in securing a garment. Shoulder-cords are a similar constraint when worn by charioteers,young girls, and youthful goddesses.79Crossbands as a restraint are worn with a short chiton by active females such as Furies,80Lyssa,8" Iris,82Artemis,83and Amazons.4 Crossbands are also worn with a peplos by several of the Nikai on the Nike Temple parapet and by two Roman statues of Nikai in Berlin that have been connected with the statue of Zeus at Olympia.5 In the 4th century and later, crossbands are worn by Athena on Panathenaic amphoras,86on Hellenistic coins,87 and in the Eleusinian scene on a 4th-century relief hydria from Cumae now in the Hermitage.88 76. Metzger 1942-1943, pl. 13. 77. Artemis on a late-5th-century votive relief from Sparta,Museum 468: LIMC II, 1984, p. 267, no. 679, pl. 238, s.v.Apollon (W. Lambrinoudakis); Artemis on the votive relief,London, British Musuem 816: LIMC II, 1984, p. 687, no. 882, pl. 512, s.v.Artemis (L. Kahil).For Artemis Kindyas,see Jucker1967; LIMC II, 1984, no. 4, s.v. Artemis Kindyas(R. Fleischer). 78. Athens, National Museum 1718: CVA,Athens 2, pp. 17-18, pls. 32-33. 79. See above,note 10. 80. Furieson Naples 82324: LIMC III, 1986, p. 827, no. 4, pl. 595, s.v. Erinyes (H. Sarian);Ruvo,Jatta 1094: LIMC III, p. 828, no. 8, pl. 596; Karlsruhe,BadischesLandesmuseum B4: LIMC III, p. 828, no. 11, pl. 596; Paestum,National Museum 4794: Trendall1987, pp. 109,142, pl. 62:a; LIMC III, 1986, p. 833, no. 64, pl. 600, s.v. Erinyes(H. Sarian). 81. Lyssa on G6teborg krater: LIMC VI, 1992, p. 325, no. 4, pl. 167, s.v. Lyssa (A. Kossatz-Deissmann); see also LIMC VI, p. 325, no. 6

(Cremona23: RVAp,p. 263, no. 27a) and no. 8 (London, British Museum F 271: RVAp,p. 415, no. 5), with pl. 167. 82. Iris on Attic red-figurebellkrater,Perugia:LIMC V, 1990, p. 747, no. 60, pl. 489, s.v.Iris I (A. KossatzDeissmann);Apulian red-figurevolute krater,Geneva, privatecollection: LIMC V, p. 750, no. 100, pl. 494; Apulian red-figurevolute krater,St. Petersburg,Hermitage B 1717/St. 424: LIMC V, p. 756, no. 156, pl. 498; Apulian red-figurevolute krater,Ruvo, Jatta31: LIMC IV, 1988, p. 705, no. 391, pl. 427, s.v. Hera (A. KossatzDeissmann). 83. Artemis on a South Italianvase, London, BritishMuseum F 159: RVAp, p. 208, no. 104. Mary Sturgeonkindly broughtto my attention the Roman marblevariationof a 5th-century bronze sculpturein Corinth, which she interpretsas Artemis wearing crossbands over a short chiton: Sturgeon 1995. 84. Amazons on Delphi tholos metope:LIMC 1,1981, p. 593, no. 100,

pl. 453, s.v.Amazones (P. Devambez). Amazons on Apulian red-figurevolute krater,BrusselsMusee RoyalA1018: RVAp,p. 35, no. 1, pl. 10; LIMC I, 1981, p. 610, no. 369, pl. 486, s.v. Amazones (P. Devambez). 85. Nike parapet,Athens, Acropolis Museum 994: Brouskari1974, p. 160, fig. 343; LIMC VI, 1992, p. 865, no. 160, pl. 575, s.v. Nike (A. GoulakiVoutira);Simon 1997. For the socalled Pheidian Nikai, Berlin K181 and K182:Thompson 1944, p. 191, fig. 9. 86. Panathenaicamphora,Paris, LouvreN 3163:ABViKp. 415, no. 3; LIMC II, 1984, p. 971, no. 141, pl. 720, s.v.Athena (P Demargne). 87. Athena on a staterof Lysimachos(297-281 B.C.): LIMC VI, 1992, p. 895, no. 720, pl. 605, s.v.Nike (A. Goulaki-Voutira). 88. Relief hydriafrom Cumae, St. Petersburg,Hermitage 51659: LIMC IV, 1988, p. 878, no. 405, pl. 593, s.v. Demeter (L. Beschi); Clinton 1992, pp. 78-81, ill. 9, figs. 17-19.

BACK-MANTLE

89. Helios on Apulian red-figure oinochoe in Zurich:LIMC V, 1990, p. 1013, no. 78, pl. 636, s.v. Helios (N. Yalouris).Orpheuson an Apulian red-figurevolute krater,Naples, Museo Nazionale, Santangelo709: RVAp, p. 533, no. 284, pl. 196; LIMC IV, 1988, p. 387, no. 154, pl. 224, s.v. Hades (R. Lindner).Parison Apulian red-figurevases in Geneva (Sciclounoff), Mattinata (Sansone 685), and Melbourne (Geddes):LIMC IV, 1988, pp. 519-520, nos. 103, 104, pl. 311 and p. 535, no. 197, pl. 328, s.v. Helene (L. Kahil). 90. Hades on an Apulian red-figure kraterin Malibu,J. Paul Getty Museum 77.AE.13: RVAp,p. 863, no. 17, pl. 323:3; CVA,Malibu 3, pp. 7-8, pls. 133-135; LIMC VI, 1992, p. 385, no. 125, s.v. Hades (R. Lindner),and LIMC VI, p. 727, no. 15, pl. 444, s.v. Herakleidai(M. Schmidt). Kreonon an Apulian red-figurevolute krater, Munich, StaatlicheAntikensammlungen 3296: RVAp,p. 533, no. 283, pl. 195; LIMC VI, 1992, pp. 123-124, no. 17, pl. 54, s.v. KreousaII (G. Berger-Doer),and LIMC VI, p. 391, no. 29, pl. 197, s.v.Medeia (M. Schmidt). 91. Sicilian red-figurecalyxkrater, Caltagirone:Trendall1967, p. 590, no. 29, pl. 228:5-6; LIMC 1, 1981, p. 778, no. 23, pl. 628, s.v.AndromedaI (K. Schauenburg). 92. Miller 1997, pp. 153-187. 93. Eleusis Museum 5104: Mylonas 1961, fig. 56; Kanta 1979, p. 102, fig. 47. 94. Helbig4IV, pp. 181-182, no. 3217 (H. von Steuben). 95. Cyrene C14191: LIMC VIII, 1997, p. 241, no. 14, pl. 168, s.v. Victoria (R. Vollkommer). 96. Athens, National Museum 695: Vorster1983, p. 351, no. 59. 97. Brommer1975, pp. 186-188, pls. 166-169.

AND

PEPLOS

247

In mythological scenes on vases from the end of the 5th century on, crossbands are part of the festival costume worn by an Oriental king or other notable person. On 4th-century vase paintings, crossbands appear on figures who often have Eastern associations: Helios, Orpheus, and Paris.89Figures wearing crossbands,usually over long-sleeved chitons, appear most often in scenes based on theatrical representations and include Hades, Kreon, and others.90Andromeda wears crossbands in South Italian vase painting.9"In these examples,the crossbandsmay indicate a luxury item like the ependytesand parasols noted by Miller in her work on Oriental paraphernaliain Athens.92 Toward the end of the Hellenistic period, the caryatidsfrom the Inner Propylaea at Eleusis wear crossbands.93The crossband costume continues into the Roman period for young females; the caryatids now in the Villa Albani, who held cistae upon their heads, also wear crossbands over the peplos as well as the back-mantle.94A Roman Victory from Cyrene wears crossbands over an Attic belted peplos.95 Nearly half the maidens on funerary monuments wear crossbands, twenty-seven out of sixty, and they occur more often on the later monuments. Crossbands may have served all three ftinctions discussed abovea sign of fertility for marriageableyoung women; a restraining device, either literally or symbolically; and a status symbol borrowed from the East. Most of the maidens in large groups wear the crossbands, for example, Eukoline (1, Fig. 9) and the girl on the Mantua stele (36). A few very young girls, still children and certainly too young for marriage, wear crossbandsover a chiton but without the back-mantle, as on a statue from the Ilissos areadedicated to the Eileithyiai, goddesses of childbirth.96None of the maidens on loutrophoroi wear the crossbands, nor do maidens in cult scenes on votive reliefs. Older females never wear them, and it may simply be that the maidens wearing the crossbands are among the youngest. One can also speculate that maidens without crossbands were those alreadychosen for marriagewho had given up their childhood symbol. CHITON

The maiden'schiton is the typical Classical garment, representedwith four to six buttons on the arms indicating the loosely gathered arm sections. A few maidens like Mynnion (6, Fig. 10) wear a simpler chiton with short arm sections. Most of the maidens on funerary monuments wear the chiton under the peplos, but none of the maidens in the cult scenes represented on votive reliefs (61-68) wear the chiton, nor do those on the two relief bases (59, 60). There seems to be no clear reason why some maidens wear a chiton under their peplos, and others do not. Neither the date of the monument nor the apparent age of the maiden seems to provide any criteria for the absence or presence of the chiton. In general, more of the older girls representedon later monuments wear chitons under the peplos. In the series of maidens in peplos on the east frieze of the Parthenon, only those coming from the south frieze procession, from the left of the central scene, wear a chiton under a peplos.97Maidens coming from the north frieze procession, from the right side, do not wear a chiton under

LINDA

248

JONES

ROCCOS

Figure9. Stele, Eukoline(1). Athens, KerameikosMuseum8754. CourtesyDeutsches Archaologisches Institut, Athens

e~~~~~~~~~~~~~~~~~~~~~~MA -v--_----

)iW

~~_

|f:4L I

i |R

[

ad4S'0 _

4,

t

i

_

wh

i

o

!

Ti|i

D8:t

the peplos. This difference in dress has been explained in various ways. C. J. Herington had earlier proposed that the two processions led to different sacrifices:the northern one for Athena Polias, religious in nature, and the southern one for Athena Parthenos, a civic rite.98Harrison recently proposed that the north procession with its loosely arrangedtribal units of four depicted an earlier time period, and the south procession depicted contemporary times with Kleisthenes' ten tribal divisions represented by differencesin garments worn by ranksof riders.99Thus, Harrison suggests that the chiton was worn under the peplos in contemporaryAthens, by the maidens in the south procession, while the peplos was worn alone in earlier times. On Classical Athenian grave monuments, most women wear a chiton under the peplos or himation,100possibly also representing contemporary styles. Those wearing only the peplos may indeed refer to an earlier era, representing families with more historical ties to Athens. Maidens who wear no chitons in the cult scenes on votive reliefs are surely taking part in a traditional sacrificialritual.

98. Herington 1955. 99. Harrison 1984, I and II; 1989. 100. According to Bieber and Eckstein (1967, p. 19), the chiton in Homer'stime is worn only by men. The word is Semitic in origin, and Carians wore it in Herodotus'stime (5.88); see alsoAgoraXI, p. 53, note 31. Janet Grossmanhas also observedthis phenomenon in the gravestelai from the Athenian Agora, which she is publishing.

BACK-MANTLE

AND

PEPLOS

249

Figure 10. Stele, Mynnion (6). Athens, National Museum 763. CourtesyMuseum ww~~~~~~~~~~~gsisiast::::<.:

_

_

_

tA

* IIII 111

~~~~~~

-.

_

CHRONOLOGY

101. Clairmont 1993, Introduction, pp. 12-18. I follow his chronologyonly very loosely since I date several monuments differently. 102. For the years360-330, see Frel 1969, pp. 35-40, nos. 212, 218, 273, 274,282.

Althoughthe maidenswith back-mantleandpeplosretaina similarclasca.370 B.C. untilthe latestmonusicizingimagefromtheirfirstappearance mentsca. 310, the styleof the figureandof the garmentdoes change.It is possibleto distinguishsome of the earlyfigureslike Stratyllis(9, Fig. 11) from later ones such as Theophile (20, Fig. 4). It is far more difficult, however,to arrangethe series of monumentsinto chronologicalorder; perhaps,as Clairmonthas also suggested,it is betterto avoidsucha rigid in onlytwogroups: structure.10' The monumentsdiscussedherearearranged earlierandlater(see below,pp. 253-257). Severalworkson differenttypesof monumentsprovideguidelinesfor datingthe maidens.Jiri FreldiscussesAttic workshopsthat includeseveral of the stelai;his broadchronologicaldivisions(400-360, 360-330, and end of century)areusefulfor datingthe stelaiwith maidens,and he Schmaltz'sstudyof marblelekythoi includesseveralin his middlegroup.102

LINDA

250

?M4SB .-

JONES

ROCCOS

_

Figure11. Stele, Stratyllis(9). Athens, NationalMuseum3691. CourtesyMuseum

4-111l *e

.

| .#>'8r

Z

includes some of the maidens,'103and Kokula provides some dates in her dates4th-century fu_nerary studyof marbleloutrophoroi.'04UrsulaVedder monuments and includes some newly published stelai.1'5In studying 4th century chronology,Timotheos Lygkopoulos calls the New York maiden (Fig. 1) the "last grave monument."106 He breaks up the century into five groups that accord well with the chronology of the stelai depicting maidens, which fall into his last three groups: III, 370-350; IV, 350-330 V 330-306. Scholl includes only five stelai and vessels with maidens in his 103. Schmaltz 1970, p. 126, no. A6 (Vienna), 350-330; p. 130, no. A122 (Piraeus,ArchaeologicalMuseum 1529), 370-350; p. 135, no. A179 (Athens, National Museum 1098), 370-350. 104. Kokula1984, p. 185, no. H2 (Archestrate),370-360; p. 186, no. H6 (Plangon), 360-350; pp. 187-188,

no. H10 (Musee Rodin 44), ca. 370; p. 189, no. H16 (Athens market),340320; pp. 193-194, no. H37 (Philhoumene, with loutrophorosin the pediment), 350-340. 105. Vedder1985, p. 29, no. F34 (Silenis), 360-350; p. 30, no. A4 (acroterion),360-350; p. 33, no. F2 (Kleariste),340-330; p. 33, no. F12

(Theophile), 330-320; see especially p. 33, note 76, with a list of maidens (a-m) from Silenis, 360-350, to the New Yorkmaiden, ca. 310. 106. New Yorkmaiden:Lygkopoulos 1983, quotationon p. 68; see also pp. 76-77, no. 39, and fig. 60.

BACK-MANTLE

AND

PEPLOS

25I

work on the "Bildfeldstelen"and dates them between 360 and 340.107 A recent work on the naiskos stelai by Bergemann places severalmonuments with maidens in the latest two groups, dating to 360-330 and 330-300.1o8 The maidens in this costume seem to appearfor only a few years after 317, when Demetrios of Phaleron issued an anti-luxury decree, which may not have stopped the production of all funerary monuments immediately.109 No maidens wearing back-mantle and peplos appear on Hellenistic grave stelai.110 Since all the maidens wear the same belted Attic peplos, changes in the placement of the belt on the overfold may provide a framework for a chronology.The system of proportions that results is similar to the changes in the ratio of head to body observed on 5th- and 4th-century sculptures. The body structureis more useful to us here since many of the maidens are preserved without heads. First, from the Early Classical period until the late 5th century, the part of the peplos overfold below the belt is significantly shorter than the part above the belt. Such proportions produce a rather heavy looking, columnar figure. This top-heavy division of the overfold can be seen clearly on "Angelitos'sAthena" and the "Mourning Athena" stele.1"'None of the maidens discussed here have these proportions. Second, from the late 5th century until ca. 350, the overfold divisions are nearly equal, for the belt roughly bisects the overfold. We can see this scheme on the statue of Persephone from Eleusis (ca. 410) and on Athena in the document relief of 409/8.112Some of the earlier maidens on stelai and lekythoi from 370 to 340 show this equal division of the overfold (see, e.g., Stratyllis, 9, Fig. 11). Third, noticeable changes appearsoon after the mid-4th century,when the section of the overfold below the belt becomes significantly longer than that above it. Such figures have a slim, attenuated appearance,which is exaggerated in the Hellenistic era. These proportions occur on Athena in the Boule stele (Fig. 3).113Most of the maidens show this bottom-heavy division of the overfold (see, e.g., Kleariste, 18, and Theophile, 20, Fig. 4). Lawton has observed similar changes in the peplos overfold of figures on document reliefs.114 In addition to proportions, another important element of the maidens' dress gives us a guide to chronology: the edges of the back-mantle are represented in two ways. Prior to the mid-4th century, the sides of the mantle are generally curved in a naturalistic manner when held out with one hand, as on an acroterion in Athens (5, Fig. 12), a scheme perhaps 107. Scholl (1996, pp. 69-70) includes most of the maiden stelai in his "PotamonGroup,"360-340 B.C., nos. 145, 482, and 484. I believe that nos. 482 and 484 (here nos. 36-37) in fact date to after 340 B.C. 108. Bergemann1997, pp. 158-179. His group dated to 390-360 includes only one maiden stele, no. 231, Paris, LouvreMa 4556. The "360-330"group includes eleven maiden stelai:nos. 268, 271,293,333,368,380,431,436,466,

468, 580. The "330-300"group includes seven maiden stelai:nos. 596, 618, 626, 628, 644, 650, 661; the group dating to the second half of the 4th centuryincludes three maiden stelai: nos. 773, 795, 816. Other stelai that may representmaidens are nos. 500 and 619. The statuesof maidens in backmantle and peplos-nos. 642, 646, 703, and 815-all occur in the latest groups. 109. Ridgway1990, pp. 30-31; 1997, p. 157.

110. Schmidt 1991. 111. See above,note 67. 112. Persephone,see above,note 66. 409/8 relief,Paris,LouvreMa 831: Boardman1985, fig. 179; Meyer 1989, p. 270, no. A16; Lawton 1995, p. 86, no. 4, pl. 5. 113. Boule relief,see above,note 41; for other reliefs of this period, see above,note 42. 114. Lawton 1995, p. 69.

252

LINDA

JONES

ROCCOS

Figure12. Acroterion(5). Athens, NationalMuseum744. Courtesy Museum

(59, Fig. 7). Certain features of the maidens can be compared with those of other sculpturesto arriveat a date. Around the middle of the 4th century,heavier, bulkier cloth resembling that on the Mausoleum sculptures appears on some maidens."' The relief carving is generally deeper thereafter,as it is in the document relief of 347/6 in Athens and some stelai, like a stele in Mantua (36).116 After mid-century,finer draperyfolds appearin the peplos, as on Kleariste (18). The static frontal poses of some of the later highwaisted maidens like Theophile (20, Fig. 4) resemble that of Athena on the 295/4 document relief and seem to reflect the monumental images of the 5th century."17 Major changes in art as well as in philosophy and politics are often noted for 4th-century Athens, the Late Classicalperiod."18Andrew Stewart claims that the statue of Eirene, ca. 370 B.C., represented"the first 'official' neo-classicism in Greek (and Western) art.""'9 This period is indeed distinct from the preceding, strictly Classical period, as so many have observed. The strongly classicizing trends in the figures of the maidens imply that a revivalof interestin the prime of ClassicalAthens was intentional.

115. Ridgway 1997, pp. 117-135. 116. 347/6 B.C. relief,Athens, National Museum 1471: Meyer 1989, p. 290, no. A88, pl. 28:1; Lawton 1995, p. 98, no. 35, pl. 18. 117.295/4 B.C. relief,Athens, National Museum 4063 + 2307: Meyer 1989, p. 312, no. A169, pl. 45:2; Lawton 1995, p. 109, no. 59, pl. 31. 118. Brown 1973; Ridgway1990, pp. 13-71. 119. Stewart 1990, p. 174.

BACK-MANTLE

AND

PEPLOS

253

THE MONUMENTS120 EARLY

STELAI

(370-340)

1 Athens, Kerameikos 8754, Fig. 9. IG 112 9203; Clairmont 1993, IV, p. 95, no. 4.420; Traill 1994, 7, p. 328, no. 437180; Bergemann 1997, p. 164, no. 271. Three figures with girl: chiton, crossbands,

bird,dog. Inscription:'Ovij,oFLV'OvIzoPo; A[f]oc3Lo;I

120. Clairmont(1993) and Conze (1893-1922) providereferencesand illustrationsfor most of the funerary monuments.In addition,see Hausmann 1948 and LIMC II, 1984, pp. 873, 881, s.v.Asklepios (B. Holtzmann) for the votive reliefs (referredto below as LIMC II); several more are unpublishedbut visible in museums.All images of the girl show her in back-mantleand belted peplos; chiton and crossbandsare noted as well as other attributessuch as dolls, birds, and dogs.

Uproovo` NtxoGZop6rcZEivxoXLvj. 2 Athens, Kerameikos P667/4864. Clairmont 1993, l, p. 371, no. 1.433; Bergemann 1997, p. 164, no. 268. Fragment of girl: chiton. 3 Athens, National Museum 113. Conze 1893-1922, no. 1236. Fragment of girl: chiton, crossbands. 4 Athens, National Museum 145. Clairmont 1993, Ill, p. 166, no. 3.340a; Bergemann 1997, p. 178, no. 805. Two figures with girl: chiton. 5 Athens, National Museum 744, acroterion, Fig. 12. Conze 18931922, no. 852. Girl: crossbands. 6 Athens, National Museum 763, Fig. 10. IG 1125273; Clairmont 1993,1I, p. 520, no. 2.421; Osborne and Byrne 1996, p. 322, no. 3; Bergemann 1997, p. 165, no. 293. Woman with girl: chiton. Inscription on architrave:M6vvrtovXoupsopa&oToo 'Ayvo[u]cPou. 7 Athens, National Museum 1017. IG 112 11007; Clairmont 1993, l, p. 455, no. 1.827; Scholl 1996, p. 264, no. 145; Traill 1994,4, p. 290, no. 277550; Osborne and Byrne 1996, p. 95, no. 17. Girl: chiton, crossbands,bird(?). Inscription: FXuXp?oc. 8 Athens, National Museum 2775. Clairmont 1993, l, p. 309, no. 1.312; Bergemann 1997, p. 174, no. 626. Fragment of girl: chiton, doll. 9 Athens, National Museum 3691, Fig. 11. IG 1126582; Clairmont 1993,1I, p. 553, no. 2.436; Osborne and Byrne 1996, p. 407, no. 2; Bergemann 1997, p. 166, no. 333. Seated man holds hands with girl: chiton. Inscription on architrave,left: K-qptaoxpvoorFocvxxovo; K 68oc0vousk. Inscription on architrave,right: ZTpocToX?t K-qpvYoxpLTo. 10 Berlin, Staatliche Museum 1492 K 40 (from Attica). IG 112 8421; Clairmont 1993, 1, p. 470, no. 1.862; Bergemann 1997, p. 169, no. 431. Attendant with girl: crossbands.Inscription on architrave: Zkvl; Muvaxoo Bootwoc. 11 Eretria Museum 630. Clairmont 1993, 1, p. 281, no. 1.256. Fragment of girl: chiton. 12 Istanbul, Archaeological Museum 99. Clairmont 1993, 1, p. 291, no. 1.280; Bergemann 1997, p. 169, no. 466. Girl: chiton, crossbands,bird. 13 Paris, Musee du Louvre Ma 4556. Clairmont 1993, 1, p. 316, no. 1.329; Bergemann 1997, p. 164, no. 231. Fragment of girl: chiton, doll.

254

LINDA

JONES

ROCCOS

14 Unknown, once private. Conze 1893-1922, no. 906; Clairmont 1993,1 , p. 500, no. 1.943; Bergemann 1997, p. 166, no. 368. Attendant and woman with girl: chiton. 15 Unknown. Conze 1893-1922, no. 1227. Fragment of girl: chiton. 16 Unknown. Conze 1893-1922, no. 1235. Fragment of girl: chiton. 17 Unknown. Conze 1893-1922, no. 1242. Fragment of girl: chiton. LATE

STELAI

(340-310)

18 Athens, Kerameikos P279. IG 112 11854; Clairmont 1993, l, p. 494, no. 1.932; Osborne and Byrne 1996, p. 263, no. 7; Bergemann 1997, p. 174, no. 596. Attendant with girl: chiton. Inscription on architrave:K? socpoYT 'Enouv6?TVo. 19 Athens, National Museum 1023. Clairmont 1993, IV, p. 147, no. 4.830; Scholl 1996, p. 265, no. 150. Five figures (two seated) with girl: chiton. 20 Athens, National Museum 1305, Fig. 4. IG 112 11660; Clairmont 1993,1 , p. 450, no. 1.814; Osborne and Byrne 1996, p. 221, no. 20; Bergemann 1997, p. 173, no. 618. Attendant with girl: chiton, crossbands.Inscription: Oeo)tpR'o. 21 Athens, National Museum 2107. Clairmont 1993, l, p. 313, no. 1.321a; Bergemann 1997, p. 178, no. 795. Fragment of girl: chiton, crossbands,bird, dog. 22 Athens, National Museum 2885. Clairmont 1993, Ill, p. 381, no. 3.453; Bergemann 1997, p. 174, no. 628. Two figures with girl: chiton, crossbands. 23 Athens, National Museum 4900, from Rhamnous. Clairmont 1993,1 , p. 350, no. 1.382; Bergemann 1997, p. 175, no. 661. Fragment of girl. 24 Bignor Park, Hawkins. Clairmont 1993, 1, p. 335, no. 1.359; Bergemann 1997, p. 169, no. 436. Fragment of girl: chiton, crossbands. 25 Brauron Museum 84. Clairmont 1993, 1, p. 285, no. 1.267; Bergemann 1997, p. 175, no. 644. Fragment of girl: chiton, crossbands. 26 Brauron Museum BE 847. Clairmont 1993,1, p.285, no.1.268; Bergemann 1997, p. 178, no. 815. Fragment of girl: chiton, crossbands. 27 BrauronMuseum. Clairmont 1993, 1, p. 378, no. 1.459; Bergemann 1997, p. 175, no. 646. Fragment of girl. 28 Brauron Museum. Clairmont 1993, I, p. 440, no. 1.783; Bergemann 1997, p. 178, no. 816. Fragment of attendant and girl: chiton, crossbands. 29 Brauron Museum. Clairmont 1993,I1, p. 306, no. 2.334. Seated woman with girl. 30 Broadlands, Broom Hall. Clairmont 1993,I1, p. 365, no. 2.357c; Scholl 1996, p. 321, no. 368. Seated woman holds hands with girl: crossbands. 31 Unknown, oncc in Chalandri. Conzc 1893-1922, no. 881;

BACK-MANTLE

AND

255

PEPLOS

Clairmont 1993, Il, p. 597, no. 2.470; Bergemann 1997, p. 167, no. 380. Attendant with girl: chiton, crossbands. 32 Eleusis Museum. Clairmont 1993, l, p. 458, no. 1.840; Bergemann 1997, p. 175, no. 650. Fragment of attendant with girl: doll. 33 Istanbul, Archaeological Museum E265. Clairmont 1993, l, p. 458, no. 1.839. Fragment of attendant with box, back-mantle of maiden. 34 Istanbul, Archaeological Museum E692. Clairmont 1993, l, p. 482, no. 1.883; Bergemann 1997, p. 169, no. 468. Attendant with girl: chiton, crossbands,bird. 35 Laon, Musee Municipale 37.1193. Clairmont 1993, Il, p. 711, no. 2.825; Bergemann 1997, p. 179, no. 839. Attendant and woman with girl: chiton. 36 Mantua, Ducal Palace. Clairmont 1993, Ill, p. 292, no. 3.394b; Bergemann 1997, p. 171, no. 512. Two figures with girl: chiton, crossbands. 37 Marathon Museum. Clairmont 1993,I11, p. 275, no. 3.387a. Three figures with girl. 38 Paris, Musee Rodin 32. Clairmont 1993, 1, p. 366, no. 1.428; Scholl 1996, p. 351, no. 482. Attendant with girl: crossbands, bird, dog. 39 Princeton, Art Museum 204. Clairmont 1993, 1, p. 439, no. 1.775; Scholl 1996, p. 352, no. 484. Attendant with girl: crossbands,bird. 40 Unknown, once Brummer Collection. Clairmont 1993, 1, p. 308, no. 1.310; Bergemann 1997, p. 173, no. 574. Girl: chiton. Inscription: MsZE(xx. 41 Unknown, once Melchett Collection. Clairmont 1993, I1, p. 493, no. 3.880; Bergemann 1997, p. 173, no. 580. Man, woman, and attendant with girl: chiton. 42 Unknown. Clairmont 1993,I11, p. 322, no. 3.413a; Traill 1994, 4, p. 291, no. 277595. Three figures with girl. Inscription: FXux6pa Atovoutcax}Hlpx?[x6o]. 43 Unknown. Clairmont 1993, 1, p. 300, no. 1.294; Bergemann 1997, p. 177, no. 725. Fragment of girl: chiton, crossbands. 44 Unknown. Conze 1893-1922, no. 1243; Bergemann 1997, p. 178, no. 773. Fragment of girl. EARLY LOUTROPHOROI

(370-340)

45 Athens, Agora I 4516. Clairmont 1993, I1, p. 486, no. 3.870. Three figures with girl. Fragmentaryinscription. 46 Eleusis Museum 5098, Fig. 5. Clairmont 1993, I1, p. 480, no. 3.860. Attendant, girl holds hands with seated woman. Inscription: Hl?XcyyWv. 47 Mariemont, Musee Royal B 18. Clairmont 1993, 11, p. 162, no. 3.337. Man, scated woman holds hands with girl: chiton, crossbands.

256

LINDA

EARLY

LEKYTHOI

JONES

ROCCOS

(370-340)

48 Athens,NationalMuseum1098. Clairmont1993, Il, p. 378, no. 2.362c. Man holdshandswith girl:crossbands. 49 Athens,NationalMuseum.Clairmont1993, Il, p. 481, no. 2.395c.Youthwith girl:chiton.Inscription:Ntxo8Rpo6,. 50 Athens,RomanAgora 1073. Clairmont1993, IV, p. 80, no. 4.381. Three figures with girl. Inscription: XpovDov. 51 Brauron Museum. Clairmont 1993, II, p. 442, no. 2.383b. Seated woman holds hands with girl: chiton, crossbands. 52 BrauronMuseum. Clairmont 1993, II, p. 552, no. 2.434b. Seated woman holds hands with girl: crossbands. 53 Milan, private collection. Clairmont 1993, I, p. 498, no. 1.938. Attendant with girl: chiton, crossbands. 54 Piraeus, Archaeological Museum 1529. Clairmont 1993, II, p. 234, no. 2.292a. Woman with girl: crossbands.Inscription: Mvqo6cpYlT. 55 Piraeus, Archaeological Museum 3027. Clairmont 1993, II, p. 512, no. 2.417a. Seated woman holds hands with girl. Inscription: (DQq. LATE

LEKYTHOI

(340-31O)

56 Piraeus, Archaeological Museum 3365. Clairmont 1993, Il, p. 330, no. 2.344a. Youth holds hands with girl: chiton, crossbands. Inscription: Ntxo8Rp6oo. 57 Piraeus, Archaeological Museum 3367. Clairmont 1993, II, p. 332, no. 2.345a. Youth holds hands with girl: chiton. 58 Vienna, Kunsthistorisches Museum 886. IG 112 5475; Clairmont 1993, III, p. 164, no. 3.339a. Attendant, seated woman holds hands with girl: crossbands. Inscription over seated figure, not the maiden: A[r]VooopcT XocvrT6cA3ou 'A?vxtos. LATE

BASES

(340-310)

59 Athens, First Ephoreia (Fethiye Djami), Fig. 7. Clairmont 1993, I, pp. 10-13, fig. 10. Thirteen figures on three sides, girl frontal. 60 Athens, Second Ephoreia 2346. Schmaltz 1978. Fragment of figure with girl. LATE

VOTIVE

RELIEFS

(340-310)

61 Athens, National Museum 1333, from Athens Asklepicion, Fig. 6. LIMC II, p. 873, no. 66, pl. 639. 62 Athens, National Museum 1367, from Athens Asklepicion. Hausmann 1948, p. 143, no. A58. 63 Athens, National Museum 1377, from Athens Asklepicion. LIMC II, p. 881, no. 201, pl. 650. 64 Athcns, National Muscum 1426, from Epidaurus. LIMC II, p. 881, no. 204, pl. 650.

BACK-MANTLE

AND

PEPLOS

257

65 Athens, National Museum 1429, from Piraeus Asklepicion. Hausmann 1948, p. 142, no. A9. 66 Eleusis Museum 5057. Kanta 1979, p. 71. 67 Paris, Musee du Louvre Ma 755. LIMC II, p. 873, no. 64, pl. 639. 68 Poligyros Museum 300. Uunpublished.

121. Kokula1984. 122. FriisJohansen 1951, p. 60. For Hagnostrate,see above,note 23. 123. Clairmont 1993,11, nos. 2.344a and 2.345a. 124. Schmaltz 1978, p. 84. For 5thcenturyrelief bases, see Kosmopoulou 1998.

The maidens in back-mantle and peplos appear on nearly all types of funerary sculpture: stelai and their acroteria, relief vessels, and bases of vessels. The maidens also appear on votive reliefs from Athens and Epidaurus,chiefly those dedicated to Asklepios. As noted above, nearly all the monuments, both funeraryand votive, come from Attica, particularly from Athens, Brauron, Eleusis, and Piraeus. Many of the extant funerarymonuments are simple one- or two-figure stelai, but they areoccasionallyquite large,such as 9, depicting Stratyllis (Fig. 11). Often the maidens are portrayed with a larger figure, presumably an adult. Mynnion (6) may be portrayed with her mother (Fig. 10); Stratylliswith her father;or, on some lekythoi, a brother appears.Inscriptions rarelyclarify the relationship, however, even when a man or woman is named. Three- or four-figure groups as on the stele in Mantua (36) appear to portraythe maiden with her parents. Several monuments include a variety of figures in a large family grouping, for example, the three figures with Eukoline (1, Fig. 9) on the stele in the Kerameikos. Such family groups occur on large and small stelai, as well as on vessels. The largest groups include three to five figures of varying types, as if created for specific families. In all groups the maiden appears to be younger than the adults since she is shown as shorter, yet she is taller than the children sometimes included. All the lekythoi and loutrophoroi represent the maiden with one or more figures, such as Plangon in Eleusis (46, Fig. 5), who is presumably with membersof her family.Kokula'sclaim that all loutrophoroiwere meant for maidens121is supportedby FriisJohansen'sobservationthat Hagnostrate on Athens NM 1863 stands beside her loutrophoros because she died a maiden.122 Several lekythoi depict a maiden with another figure. Far more lekythoi (eleven) than loutrophoroi(three) depict the maiden.Two lekythoi, 56 and 57, found together in the Piraeus, have the same two figures represented. Clairmont, who claims that many of these memorials were intended to honor the girl, thinks that both vessels were for the maiden. It is also possible, however, that both were intended for the youth, or even one for the maiden, the other for the youth.123 Without clarifying inscriptions it is impossible to determine the honoree. Figural decoration on reliefbases of marblevessels is uncommon. One fragment in the Piraeus Museum shows a maiden and the hand of a second figure reaching out to her.The large base in Athens (59, Fig. 7), found in 1954 on the south slope of the Acropolis, is particularly impressive. Schmaltz suggests that the youth who is portrayedon all three sides of the base with various other figures is the recipient of this monument,124 but signs of cxtrcmc gricf in thc gcsturc of two figurcs ncar the maidcn may

258

LINDA

JONES

ROCCOS

indicate that she too has died, perhapsbefore the youth. Here the maiden is clearly separatedfrom her family by a stele renderedin relief.The stele appearsin profile like that on the funerarylekythos of Myrrhine-a thin slab thought to separate the living from the dead, the mortal from the immortal.'25

The acroterion in Athens (5, Fig. 12) is unique among extant figured acroteria.The combination of figures within a floral ornament occurs more often in South Italian art than in Attic. On a red-figure Campanian hydria in New York, a young girl holding a phiale is surrounded by a floral ornament much like the scene on the acroterion.126 Several beautiful freestanding statues of the maiden in back-mantle and peplos attest her importance in ancient Greek society (e.g., Fig. 1). Although these figures are not considered here, they provide additional cvidence for the costume, often including back views. A statue in the Athenian Agora, which is missing both sides of the back-mantle, is stiff and frontal rather like a caryatid, although nothing is preserved of the arms to indicate the hand positions.127 Unless the figures were found in a funerary context, such as a maiden and her attendant in Athens,128 it is difficult to tell whether the monuments were funerary or votive. On votive reliefs, for the most part dedicated to Asklepios and Hygicia, the maiden usuallyappearsprominentlyin the front plane among a large family of worshippers.'29Unusual is the shrine relief to Asklepios, 63, where the maiden, although placed in the front plane, is somewhat lost among a large family group. On some votive reliefs in Athens the maiden is placed conspicuously in the front and turns away from the group in a rather self-conscious manner, for example on 61 (Fig. 6) and 62. Occasionally a child or smaller attendant appears,wearing clothing different from the maiden. Although several females in the gatherings wear the ubiquitous chiton and wrapped himation, only one female on each reliefwearsthe maiden'scostume.The votive reliefsshow most clearly that this girl is singled out as special not only by her costume but also by her position in the forefront of the group and by her open pose,130 holding her back-mantle with both hands. There are names of only a few maidens inscribed on stelai and vessels. In many inscriptions it is not clear to which female the name refers. Even fewer provide an ethnic or demotic for clues to their families or place of origin. The names are not uncommon for Greek women and appearon other grave monuments listed by Osborne and Byrne: Glykera (36 times), Plangon (17 times), and Theophile (25 times).13' A few names are quitS unusual: Stratyllis (2 times) and Mynnion (7 times). Traill's work on!Athenian names includes only a few of the maidens in the eight volumes published so far,but there are far more examples of each name: Eukoline (44) and Glykera (46).132Overall, more vessels than stelai have inscriptions, but those inscriptions do not necessarily name the maidens. Only when the name is directly overhead can we be sure that it belongs to the maiden. A few inscriptions indicate the father's name or place of origin: Stratyllis Kephisokritou of Kydathenaion (9), Mynnion Chairestratou of Hagnous (6), and Klcariste Epainetou (18). There are also a

125. Athens, National Museum 4485: FriisJohansen 1951, pp. 160161, fig. 82; Clairmont 1979; Boardman1985, fig. 154. For tomb markersrepresentedon stelai, see Ridgway 1997, p. 178, note 5; Clairmont 1993, Introduction,p. 98. 126. New York,Metropolitan Museum of Art 06.1021.230: Trendall 1967, p. 411, no. 342, pl. 165:3. 127. Athens, Agora S 339: Shear 1935, pp. 372-374, figs. 1-3. 128. Alexandri 1969;AR 19701971, pp. 4-5, fig. 3. 129. Hausmann 1948; LIMC II, 1984, p. 873, nos. 64, 66, pl. 639 and p. 881, nos. 201,204, pl. 650, s.v. Asklepios (B. Holtzmann). Few votive reliefs come from Epidauros,however, and none from the sanctuaryon Cos; see LIMC II, 1984, pp. 891-892. For the Asklepios cult in Athens, see Larson 1995, pp. 61-64. 130. For the girl opening the mantle herself,cf. Harrison 1979 on Apollo's cloak, and Neumann 1979 on the statue of Persephone(also known as Demeter) at Eleusis, note 66 above. 131. Osborne and Byrne 1996. 132. Traill 1994.

BACK-MANTLE

133. Stears 1995, pp. 118-121. 134. Beaumont 1994, esp. pp. 8693. 135. For the dedicationsto Artemis on the Acropolis,see Linders 1972. 136. Woysch-Meautis 1982, pp. 110-124, nos. 65-255. 137. Clairmont 1993, VI, p. 102. 138. Schmidt 1977, pp. 114-128; Reilly 1997. 139. Briimmer1985. 140. Clairmont1970, p. 30, pl. 15, regardingMunich, Glyptothek 491 (Mnesarete):Ridgway1997, pl. 39. See Beaumont 1994, p. 88. 141. Vorster1983, p. 330, nos. 1 and 2; Beaumont 1994. 142. Athens, National Museum 3624: Boardman1985, fig. 151; Stewart1990, fig. 477; Ridgway1997, pl.38.

AND

PEPLOS

259

few girls of non-Athenian origin: Silenis of Bocotia (10) and Eukoline from Lesbos (1); these girls emulate the Athenian maidens in dress, pose, and monument type. The maidens in back-mantle and peplos seem to belong to a single age group, about eleven to fifteen years old. The youngest-looking figure is Mynnion (6, Fig. 10), the oldest in appearance, Stratyllis (9, Fig. 11). Stears has recently noted that these teenage girls fall in the middle of age groups portrayed on grave reliefs (baby, child, teenager, married woman, old woman).'33Lesley Beaumont in her exploration of the iconography of childhood in Athenian art notes that relativeratherthan specific ages were most commonly represented.34 Greek art of the 4th century makes distinctions in age that had not previously been attempted, and the maidens in particularseem to represent quite accuratelythe varying stages of adolescent growth. Motifs are used to distinguish younger from older: the young girl holds a bird or a doll, the older girl holds her mantle up with one hand in a bridal gesture (e.g., Fig. 5). These 4th-century maidens are between two worlds, no longer children but not yet adults. Dedications by women of dolls and toys as well as veils and belts parallel the differences seen in the maidens.135 In her study of animals in funerarycontexts, Daphne Woysch-Meautis found that birds appearmost often, especially with children.36 Of the 201 examples of birds catalogued by Woysch-Meautis on grave monuments, six birds appear with maidens (1, 7, 12, 21, 38, 39). Of the eighty-seven dogs catalogued, three accompany maidens (1, 21,38). Both bird and dog appear on the same three stelai, as, for example, with Eukoline (Fig. 9). There are no horses, hares, goats, or felines with the maidens. Interestingly, on vessels, none of the maidens have animals. Of the twenty-eight examples of figures holding dolls listed by Clairmont, only three are maidens (8, 13, 32).37 The type of doll is usually a simple female figure with stump arms and legs, more an anatomical study than a toy.138None of the maidens on vessels hold dolls. The maiden in back-mantle appears quite often with an attendant. On sixteen funerary monuments-thirteen stelai (e.g., Fig. 4) and three vessels (e.g., Fig. 5)-the maiden is attended by a small girl with a box,

presumablyforjewelry(10, 14, 18,20,28,31,32,33,34,35,38,39,41,46, 53, 58).13 Just as on rcd-figure vases maidenscarrychests to the bride, these smaller girls on grave stelai carrychests or boxes for the maiden, who will never be a bride. Identifying these attendants as sister,servant,or slave is problematic. Clairmont believes that the girls were sisters rather than servants, but Beaumont notes that smaller size usually denotes inferior status (but see above, where smaller size is discussed as an indication of youth).140 The attendants are never named. They stand quietly holding an object, as a friend or a servant would. In some cases the attendant's hair is cut short, perhaps in mourning, and she wears a simple chiton with shoulder-cord and long overfold similar to that worn by younger children on Athens NM 693 and 694.141 On some monuments attendants wear the same long-sleeved chiton and head covering worn by attendants on other 4th-century stelai, such as the attendant to Hegeso in Athens.'42 Miller

260

LINDA

JONES

ROCCOS

has recently pointed out that the long-sleeved chiton does not necessarily mark one as a slave or foreigner, since women are shown wearing it in the late 5th century, and girls wear it in the 4th century.143 - When the maiden appears as part of a family group of three or four, either named or unnamed, it is unclearwho is the deceased. FriisJohansen explored this unsolved question, concluding that the ambiguity is purposeful, that we are not meant to distinguish the living from the dead.144 Sally Humphreys points out that the dead are shown as still alive, thus making it nearly impossible to distinguish the two states.145Several monuments indicate signs of grief in the tender gesture of a mother's hand toward the maiden's face, as on Mynnion's stele (6, Fig. 10), or a father's hand held to his head as on the Melchett stele (41).146Already noted were the quite melodramatic gestures on the Acropolis base (59, Fig. 7), and we may imagine that the object of the grief-the maiden-has died. The dexiosis(hand-clasp) gesture is associated with adults more often than with children, but some of the maidens are shown in this pose. The gesture has been studied by Friis Johansen, Glenys Davies, and Elizabeth Pemberton.147 Most scholars agree that it is a gesture of unity, by which a newly deceased person mayjoin one previously deceased. Maidens on vessels, especially on some of the earlier examples (46, 47, 48, 51, 52, 55, 56, 57, 58), usually clasp the hand of a woman, as does Plangon (46, Fig. 5). Stratyllis is the only maiden in back-mantle on grave stelai who holds the hand of another figure (9, Fig. 11). Very few of the maidens hold the edge of their mantle up next to their heads, as brides and marriedwomen often do with a veil. On some monuments, like the loutrophoros depicting Plangon (46, Fig. 5), the maiden holds her back-mantle with one raised hand and is accompanied by a small attendant with a box (see also 18,33,41). These maidens were possibly the closest to marriage.John Oakley and Rebecca Sinos have pointed out the increasing number of women's scenes on late-5th- and 4th-century Attic red-figure vases that relate to wedding iconography.148'There are,however, very few examples of the girl in back-mantle and peplos in these scenes, one example being a red-figure lebes in Athens on which a girl holds one end of her back-mantle with her lowered left hand and the upper edge with her raised right hand.149 On the grave reliefs, the maiden is nearly alwaysrepresentedas shorter than the adults who accompany her. Mynnion (6, Fig. 10), for example, is shorter than the woman with her, and Stratyllis (9, Fig. 11) stands next to her seated father,yet their heads are close in height. Eukoline (1, Fig. 9) is much shorter than the other individuals in her group. This difference in height is not shown as clearly on the vessels, which are also generally earlier. On votive reliefs, the maiden tends to be much shorter than other figures, except for children. She is always much taller than her small attendant (see, e.g., 20 and 46, Figs. 4-5). If there are other children in a group scene on funeraryor votive reliefs, they are smallerand presumablyyounger than the maiden.150 Stears, as mentioned above, places the maidens in a middle group between children and marriedwomen.151 Signs of age for girls represented at thc BrauronianArktcia arc classificd by Sourvinou-Jnwood in vcry gcn-

143. Miller 1997, pp. 160-161. 144. FriisJohansen 1951, pp. 28-36. 145. Humphreys 1993, pp. 148168. 146. Humphreys(1993, p. 108) points out that 4th-century memorials show more expressionof feelings for those who died young. See also Neumann 1979, pp. 108-115, figs. 50, 53; Bergemann1997, pp. 52-53. 147. FriisJohansen 1951, p. 29; Davies 1985; Pemberton1989; Stears 1995, p. 126; Scholl 1996, pp. 164-167; Bergemann1997, pp. 61-62. 148. Oakley and Sinos 1993. 149. Athens, National Museum 15851: Roccos 1995, p. 662, fig. 21. 150. Clairmont(1993, Introduction, pp. 19-29) has very little to say about the ages of the maidens. 151. Stears 1995, pp. 118-123.

BACK-MANTLE

AND

26I

PEPLOS

eral iconographical types, usually identified as a younger and an older figure.152Her "agesigns"may be used only as a rough guide to determine the ages of the maidens in back-mantle and peplos. If we use her criteriataller stature, a proportionally smaller head than body, and budding breasts-to indicate older females, then small stature,a proportionallylarge head, and a flat chest should indicate younger girls. Most of the maidens, like Theophile (20, Fig. 4), show signs of budding breasts,which SourvinouInwood notes on the oldest arktoi (bears). Some maidens are relatively full-breasted, like the girl shown on 59 (Fig. 7) and Stratyllis (9, Fig. 11). Only a few seem to be as flat-chested as Mynnion (6, Fig. 10). When shown with an adult, however, the maiden is usually shown with smaller breasts, like Eukoline (1, Fig. 9). Variations in breast size may indicate a close observation of the variations in female adolescence on the part of 4th-century sculptors. As a group, the maidens in back-mantle and peplos appearto be older than the girls who are arktoi(five to ten years of age) and arrhephoroi(seven to eleven). The signs of size, body type, and pose best represent the stage of the kanephoros(aged eleven to fifteen) as defined by Sourvinou-Inwood. Recognition in ancient Greece of the physical changes on the path to adulthood granted significance to each stage along the way.Just as boys underwent their ephebic rituals, girls experienced their own transitions, from (grinder)to kanephoros (basket-bearer).Aristophanes'Lysistrata arrhephoros notes the roles she has played:'53 cT'

ST9

Y>

TC-, /

G

o

o ?x?0Vou TOy xpox&rcov pxTo; Bpoop&vtot;. xocXi~ 'xou0' o&c0o =Toc~ ~TCOT XaXocvrq6POOv xa6ov 6ppxOV.

x'

As soon as I was seven years old, I was an Arrephoros; then I was a Grinder; when I was ten, at the Brauronia, I shed my saffron gown as one of the Foundress'sBears; and I was also once a basket-bearer,a beautiful girl, wearing a string of dried figs.

152. Sourvinou-Inwood1988, pp.33-38,55-57. 153. Ar. Lys. 641-647; trans.A. Sommerstein.See also Neils 1992, at the pp. 23-24, for the kanephoros Panathenaia. 154. Geagan (1994, pp. 167-169) discussesthe dedicatorymonuments for kanephoroi. 155. Athens pelike:see note 78. 156. Sourvinou-Inwood1988, p. 95, note 256. See also Van Straten 1981, p. 84, for the figure as a servant carryinga cista.

Being a kanephorosis the last in the series of roles played only by selected aristocraticyoung women. It is the stage just preceding marriage, represented by an iconographical type similar to the bride.154For example, on a pelike in Athens, the same peplos and crossbands are worn by both bride and attendant."55According to Sourvinou-Inwood, the kanephorosis represented with full breasts,but she is slightly shorter than any adult female with her. She is not to be confused with the cista-bearers,who appear in the back row of groups on some votive reliefs, as Sourvinou-Inwood was the ideal proclaims.'56Although the beautiful and virginal kanephoros totype of well-born young women in Classical Greece, all young women could be represented as parthenoi at the moment between childhood and adulthood. They were portrayed as such in 4th-century Athens wearing thc spccial costumc of back-mantle and peplos.

262

LINDA

JONES

ROCCOS

CONCLUSIONS Maidenswho wearthe back-mantleandpeplosin the 4th centuryB.C. can no longer remain anonymous among so many "standing females" on funerary monuments. Their distinctive costume sets them apart from the multitude of women wearing the chiton with himation. Although there are only sixty-eight monuments representingthese maidens out of a much largertotal corpus, the monuments are of high quality.The maiden is usually portrayed alone in a heroizing manner or with a small slave or servant girl. Less often she is shown within a family group, which may reflect the loss to family and society of a future mother. Athena and Artemis also wear this costume in the late 5th and 4th centuries B.C., and an association of the maidens with these goddesses was most likely intended. In a similar fashion, devotees of Isis dressed in the same garment as the goddess to identify themselves with her; the devotees were not necessarilypriestesses, but worshippers."57 Likewise, the maidens in back-mantle and peplos are not priestesses, but important in their own right. The maiden's size, pose, and unique costume emphasize her importance within a group. She appearsyounger than adults in group scenes but older than the child attendants often accompanying them. She is usually shown in the forefront of a group and is always standing. She often has a smaller attendant, as do older women on grave stelai. Finally, her costume of back-mantle and peplos distinguishes her from other females. She represents the ideal maiden, the postpubescentparthenoson the brink of marriage, arrayedin her festival costume. The maidens who wear this distinctive and easily identifiable costume occupied a special place within their families and wider society of Classical Greece.'58They were of an age to be married, and their deaths meant the loss of future offspring, a loss not only to their families but to the entire culture. Votive reliefs that represent the maidens remind us of the young women's participation in ritual. Splendid funerary monuments underline the maidens' importance to their own families, who erected their memorials, and to Classical Greek civilization as a whole.

157. Walters 1988, pp. 56-57. 158. See Larson 1995 and Lyons 1997 for recentworks on the heroizing of females in ancient Athens; girls and young women in particularwere greatly esteemed.

BACK-MANTLE

AND

PEPLOS

263

REFERENCES Agora XI = E. Harrison,Archaic and Archaistic Sculpture (Agora XI),

Princeton 1965. Alexandri,0. 1969. "FamilyBurial Terracenear the Road to the Academy,"AAA 2, pp. 257-266. Barber,E. J. W. 1992. "The Peplos of Athena,"in Neils 1992, pp. 103117. Beaumont,L. 1994. "Constructinga Methodology for the Interpretation of Childhood Age in Classical Athenian Iconography," ArchaeologicalReviewfrom Cambridge 13:2, pp. 81-96. Bergemann,J. 1997. Demos und Thanatos: Untersuchungenzum Wertsystemder Polis im Spiegel der attischen Grabreliefsdes 4. Jahrhunderts v. Chr. und zur Funktion dergleichzeitigen Grabbauten,Munich. Bieber,M. 1928. GriechischeKleidung,

Berlin. .1977. Ancient Copies: Contributions to the History of Greekand Roman Art, New York.

Bieber,M., and F. Eckstein. 1967. Entwicklungsgeschichteder griechischen Tracht von der vorgriechischen Zeit bis zum Ausgang der Antike, 2nd ed., Berlin. Boardman,J.1985. GreekSculpture: The Classical Period, London. 1995. GreekSculpture:The Late Classical Period and Sculpture in Colonies and Overseas, London. Brommer,F. 1977. Die Parthenonfries: Katalog und Untersuchung,Mainz. Brouskari,M. 1974. The Acropolis Museum, Athens. Brown, B. R. 1973. Anticlassicism in GreekSculpture of the Fourth Century B. c., New York.

Briimmer,E. 1985. "Griechische Truhenbehalter,"JdI 100, pp. 1-168. Clairmont,C. W. 1970. Gravestone and Epigram: GreekMemorialsfrom the Archaic and Classical Periods, Mainz.

. 1979. "The Lekythos of Myrrhine,"in Kopckeand Moore 1979, pp. 103-110. . 1993. ClassicalAttic Tombstones, Introduction,I-Vil, Kilchberg.

Clinton, K. 1992. Myth and Cult: TheIconography of theEleusinianMysteries,Stockholm. Conze, A. 1893-1922. Die attischenGrabreliefs I-IV, Berlin. Davies, G. 1985. "The Significance of the HandshakeMotif in ClassicalFuneraryArt," AJA 89, pp. 627-640. Diepolder,H. 1931. Die attischen des5. und4. Grabreliefs v. Chr.,Berlin. Jahrhunderts Filges, A. 1997. Standbilder jugendlicherGottinnen: Klassischeundfriihhellenistiche Gewandstatuenmit Brustwulst und ibrekaiserzeitliche Rezeption,Cologne. Frel,J. 1969. Lessculpteurs attiques anonymes,430-300, Prague. FriisJohansen,K. 1951. The Attic GraveReliefsof the Classical Period,Copenhagen. Geagen, D. 1994. "Childrenin Athenian DedicatoryMonuments,"Boeotiaantiqua4, pp. 163-173. Geddes, A. G. 1987. "Ragsto Riches:The Costume of Athenian Men in the Fifth Century,"CQ 37, pp. 307331. Gullberg,E., and P. Astrom. 1970.TheThreadofAriadne:A StudyofAncientGreekDress, Goteborg. Harris,D. 1995. TheTreasures of the Parthenonand theErechtheion,Oxford. Harrison,E. B. 1977. "The Shoulder-Cordof Themis,"in FestschriftuirFrankBrommer, U. Hockmann and A. Krug, eds., Mainz, pp. 155-161. .1979. "Apollo'sCloak," in Kopckeand Moore 1979, pp. 91-98. .1984. "Time in the ParthenonFrieze,"in Basel: Parthenon-Kongress ReferateundBerichte,4. bis 8. April1 982, E. Berger,ed., Mainz, I, pp. 230-234; II, pp. 416-418.

. 1989. "HellenicIdentity and Athenian Identity in the Fifth CenturyB.C.," in Cultural Dfferentiationand CulturalIdentityin the VisualArts (Studies in the History of Art 27), S. J. Barnes and W. S. Melion, eds., Washington, D.C., pp. 41-61. 1991. "The Dress of the Archaic Greek Korai,"in New Perspectives in Early GreekArt (Studies in the History of Art 32), D. Buitron-Oliver,ed., Washington, D.C., pp. 217-239. Hausmann,U. 1948. Kunst und Heiltum:Untersuchungen zu den griechischenAsklepiosreliefs, Potsdam. 1960.Griechische Weibreliefs, Berlin. Herington, C. J. 1955. AthenaParthenos andAthenaPolias.A Studyin the Religionof PericleanAthens, Manchester. Higgins, R. A. 1980. Greekand Roman Jewellery, 2nd ed., Berkeley. Humphreys,S. C. 1993. The Family, Women,and Death: Comparative Studies, 2nd ed., Ann Arbor. Jucker,I. 1967. "ArtemisKindyas,"in Gestaltund Geschichte: Festschrift KarlSchefold(AntK-BH4), M. Rohde-Liegle, H. A. Cahn, and H. C. Ackermann,eds., Bern, pp. 133-145. Kabus-Jahn,R. 1972. Die Grimanische in Venedig(AntP 11), Figurengruppe Berlin. Kanta,K. 1979.Eleusis:Myth,Mysteries,History,Museum,Athens. Karouzou,S. 1979. "Bemalteattische Weihreliefs,"in Kopcke and Moore 1979, pp. 111-116. XV = B. Vierneisel-Schlorb, Kerameikos I: DiefigurlichenTerrakotten bisspathellenistisch Spatmykenisch (Kerameikos XV), Munich1997. Kingsley,B. 1975. "The Stele of Myttion," GettyMusJ2, pp. 7-16. Knauer,E. 1978. "Towardsa History of the Sleeved Coat: A Study of the Impact of an Ancient Eastern Garment on the West,"Expedition 21:1, pp. 18-36. Kokula,G. 1984. Marmorlutrophoren (AM-BH 10), Berlin.

264

Kopcke,G., and M. B. Moore, eds. 1979.Studiesin ClassicalArtand A Tributeto Peter Archaeology. Heinrichvon Blanckenhagen, Locust Valley,N.Y. Kosmopoulou,A. 1998. "AFunerary Base from Kallithea:New Light on Fifth-Century Eschatalogy,"AJA 102, pp.531-545. Larson,J. 1995. GreekHeroineCults, Madison. Laurens,A.-F. 1987. "Identification d'Hebe? Le nom, l'un et le in Imageset societeen multiple," Greceancienne:L'iconographie comme methoded'analyse,C. Berard, C. Bron, and A. Pomari,eds., Lausanne,pp. 59-72. Lawton, C. A. 1995. Attic Document Reliefs:Artand Politicsin Ancient Athens,Oxford. Lee, M. M. 1999. "The Myth of the ClassicalPeplos"(diss. Bryn Mawr College). Linders,T. 1972.Studiesin the Treasure RecordsofArtemisBrauroniaFound inAthens,Stockholm. . 1984. "The kandys in Greece and Persia,"OpAth15, pp. 107-114. Losfeld,G. 1991.Essaisurle costume grec,Paris. . 1994.L'artgrecet le vedtement, Paris. Lygkopoulos,T. 1983. Untersuchungen zur Chronologie derPlastikdes4. Jhs. v. Chr.,Bonn. Lyons,D. 1997.Genderand Immortality:Heroinesin AncientGreekMyth and Cult,Princeton. Mangold, M. 1993. Athenatypenauf attischenWeibreliefs des5. und4. Jhs. v. Chr.,Bern. Mansfield,J. 1985. TheRobeofAthena and thePanathenaicPeplos(diss. Universityof California,Berkeley). Metzger, H. 1942-1943. "Leb&s gamikos a figuresrouges du Mus6e BCH 46-47, pp.228d'Ath&nes," 247. Meyer,M. 1989. Die griechische Urkundenreliefs (AM-BH 13), Berlin. Miller,M. E. 1989. "The ependytesin ClassicalAthens,"Hesperia 58, pp. 313-329. . 1997.Athensand Persiain the Ffith CenturyB. C., Cambridge. Mitchel, F.W. 1970. Lykourgan Athens:

LINDA

JONES

ROCCOS

338-322 B.C. (Lecturesin Memory of Louise Taft Semple,2nd series), Cincinnati.

Mylonas, G. E. 1961. Eleusis and the Eleusinian Mysteries, Princeton. Neils, J., ed. 1992. Goddessand Polis. The Panathenaic Festival in Ancient Athens, Princeton. Neumann, G. 1979. Probleme des griechischen Weibreliefs,Tiibingen.

Nicholls, R. V. 1995. "The SteleGoddess Workshop:Terracottas fromWell U 13:1 in the Athenian Agora,"Hesperia 64, pp. 405-492. Oakley,J.H. 1982. "The Anakalypteria,"AA,pp. 113-118. Oakley,J.H., and R. H. Sinos. 1993. The Weddingin GreekArt, Madison. Osborne,M. J, and S. G. Byrne. 1996. Lexicon of GreekPersonal Names II: Attica, Oxford. Palagia,0. 1982. "AColossal Statue of a Personificationfrom the Agora of Athens,"Hesperia 51, pp. 99-113. 1994. "No Demokrateia,"in The Archaeology ofAthens andAttica under the Democracy,W. D. E.

Coulson, ed., Oxford, pp. 113-122. Pekridou-Gorecki,A. 1989. Mode im antiken Griechenland:Textile Fertigung und Kleidung, Munich.

Pemberton,E. G. 1989. "The Dexiosis on Attic Gravestones,"MeditArch 2, pp. 45-50. Polaschek,K. 1969. Untersuchungenzu griechischenMantelstatuen: Der Himationtypus mitArmschlinge,

Berlin. Reeder,E. D. 1995. Pandora: Womenin Classical Greece,Baltimore. Reilly,J. 1989. "ManyBrides:'Mistress and Maid' on Athenian Lekythoi," Hesperia 58, pp. 411-444. . 1997. "Nakedand Limbless: Learningabout the Feminine Body in Ancient Athens,"in Naked Truths: Women,Sexuality, and Gender in ClassicalArt and Archaeology, A. 0. Koloski-Ostrowand

C. L. Lyons, eds., London, pp. 154173. Richter,G. M. A. 1968. Korai, Princeton. Ridgway,B. S. 1977. The Severe Style in GreekSculpture, Princeton. 1981. Fifth-Century Styles in GreekSculpture,Princeton.

. 1990. Hellenistic SculptureI, Madison. . 1997. Fourth-Century Styles in GreekSculpture,Madison. Roccos, L. J. 1986. "Introducingthe 'Rheitoi Athena,"'AJA90, p. 208 (abstract). . 1991. "Athenafrom a House on the Areopagus,"Hesperia 60, pp. 397-410. .1995. "The Kanephorosand Her FestivalMantle in Greek Art," AJA99, pp. 641-666. Schaeffer,J. A. 1975. "The Costume of the Korai:A Reinterpretation," CSCA8, pp. 241-256. Schmaltz,B. 1970. Untersuchungenzu den attischen Marmorlekythen, Berlin. . 1978. "Zu einer attischer Grabmalbasesdes 4. Jahrhundertsv. Chr.,"AM 93, pp. 83-97. . 1983. GriechischeGrabreliefs, Darmstadt. Schmidt, R. 1977. Die Darstellung von Kinderspielzeug und Kinderspiel in dergriechischeKunst, Vienna. Schmidt, S. 1991. Hellenistische Grabreliefs. Typologischeund chronologische Beobachtungen, Cologne.

Scholl, A. 1995. "XOH(DOPOI:Zur Deutung der Korenhalledes Erechtheion,"JdI110, pp. 179-212. . 1996. Die attischen Bildfeldstelen des 4. Jhs. v. Chr: Untersuchungen zu den kleinformatigen Grabreliefsim spitklassischenAthen,

Berlin. Schuirmann,W. 1989. Katalog der antiken Terrakottenim Badischen Landesmuseum Karlsrube, Goteborg.

Serwint, N. 1993. "The Female Athletic Costume at the Heraia and PrenuptialInitiation Rites,"4AJA 97, pp. 403-422. Shear,T. L. 1935. "The Sculpture Found in 1933,"Hesperia 4, pp. 371-420. Simon, E. 1997. "AnInterpretationof the Nike Temple Parapet,"in The Interpretation ofArchitectural Sculpture in Greeceand Rome

(Studies in the History of Art 49), D. Buitron-Oliver,ed., Washington, D.C., pp. 127-143. Sourvinou-Inwood,C. 1988. Studies in Girls' Transitions. Aspects of the Arkteia andAge Representation in Attic Iconography,Athens.

BACK-MANTLE

Stears,K. 1995. "DeadWomen's Society: ConstructingFemale Gender in ClassicalAthenian FunerarySculpture,"in Time, Traditionand Societyin Greek Bridgingthe "Great Archaeology. Divide,"N. Spencer,ed., London, pp. 109-131. An Stewart,A. F. 1990. GreekSculpture. Exploration,New Haven. Sturgeon,M. C. 1995. "The Corinth Amazon: Formationof a Roman 99, ClassicalSculpture,"4AJA pp. 483-505. Thompson, D. B. 1944. "The Golden Nikai Reconsidered,"Hesperia 13, pp. 173-209. . 1954. "ThreeCenturiesof HellenisticTerracottas:I, B and C," Hesperia 23, pp. 72-107. Todisco,L. 1993.SculturagrecadelIV secolo:Maestrie scuoledi statuariatra edellenismo,Milan. classicitai Tolle-Kastenbein,R. 1980. Peplosfiguren: Friihklassische Originale,Mainz. Traill,J. S. 1994-. PersonsofAncient Athens 1-8, Toronto. Trendall,A. D. 1967. TheRed-Figured

COLLEGE

OF STATEN

iL-io9D

2800

VICTORY

STATEN

ISLAND,

Van Straten,F. 1981. "Giftsfor the Gods,"in Faith, Hope, and Worship. Aspects of Religious Mentality in the Ancient World,H. S. Versnel,ed., Leiden, pp. 65-151. Vedder,U. 1985. Untersuchungenzur plastischen Ausstattung attischer Grabanlagen des 4. Jhr. v. Chr.,

Frankfurt. Vorster,C. 1983. GriechischeKinderstatuen, Cologne. Walters,E. 1988. Attic Grave Reliefs That Represent Women in the Dress of Isis (Hesperia Suppl.22), Princeton. Winkler, H. 1996. Die tiefe Guirtung: Ein verkanntes Motiv dergriechischen Frauenbekleidung, Rheinfelden. Woysch-Meautis,D. 1982. La representation des animaux et des eHtres fabuleux sur les monumentsfuneraires grecs:De l'epoquearchaYquea lafin du IVe siecle av. j - C., Lausanne.

LJRSI@CUNYVM

ISLAND

BOULEVARD NEW .CUNY.

YORK EDU

PEPLOS

VasesofLucania, Campania, and Sicily, Oxford. . 1987. The Red-Figured Vasesof Paestum, London. Trumpf-Lyritzaki,M. 1969. Griechische Figurenvasen des reichenStils und der spdten Klassik, Bonn.

Linda Jones Roccos LIBRARY

AND

I03I4

265

6LORIA S. MERKER

JustPublished

The Sanctuary of Demeter and Kore: Terracotta Figurines of the Classical, Hellenis tic, and Roman Periods

This volume publishes the terracotta figurines of the Classical, Hellenistic, and Roman periods from the Sanctuary of Demeter and Kore on Acrocorinth. The coroplastic finds from this site, numbering in total about 24,000 figurines and fragments, greatly enrich the known body of Corinthian figurines not only in number but also in the addition of many entirely new types and styles. Working far beyond the output of the Potters' Quarter workshops, the Corinthian coroplasts arerevealed as inventive, often highly adept in technique, and attuned to stylistic developments in the plastic arts in general. Evidence of terracotta figurines and small bronze ones, especially in the 4th and early 3rd centuries B.C., through the use of shared models, provides a glimpse of the mostly lost bronze production of that period. The figurines are also important because they help to explicate the meaning and conduct of the cult of Demeter and Kore in Corinth. The figural coroplastic art provides additional data on the deities and heroes recognized at the sanctuary,the age and gender of the participants in the rituals, the offerings they brought, and the nature of their cultic activities. Beyond these data, the figurines are examined for what they may reveal through their imagery of the underlying ideas of the cult, how the deities were perceived,why they were approached,and how the cult functioned as a part of Corinthian society.

394 pages, 79 plates

Corinth XVIII,iv ISBN 0-87661-184-6 May 2000. Cloth $100.00

HESPERIA

SUPPLEMENTS

13 Marcus Aurelius:Aspects of Civic and Cultural Policy in the East,

by James H. Oliver (1970). $15.00 14

The Political Organization ofAttica, byJohn S. Trail (1975). $15.00

15

The Lettering of anAthenian Mason, by Stephen V. Tracy(1975). $15.00

16 A Sanctuary of Zeus on Mount Hymettos, by Merle K. Langdon (1976).

$15.00 17 Kallias of Sphettosand the Revolt ofAthens in 286 byT. Leslie ShearJr.(1978). $15.00

B.C.,

19 Studies in Attic Epigraphy, History, and TopographyPresented to Eugene Vanderpool (1982).$15.00 20 Studies in Athenian Architecture,Sculpture,and TopographyPresented to HomerA. Thompson(1982). $15.00 21 Excavations at Pylos in Elis, by John E. Coleman (1986). $25.00 22 Attic Grave Reliefs That Represent Womenin the Dress of Isis, by ElizabethJ. Walters (1988). $40.00 23 Hellenistic Relief Moldsfrom theAthenian Agora, by ClaireveGrandjouan (1989). $25.00 24 The Prepalatial Cemeteriesat Mochlos and Gournia and the House Tombsof BronzeAge Crete,byJeffrey S. Soles (1992). $35.00 25

Debrisfrom a Public Dining Place in theAthenian Agora,

by Susan I. Rotroff andJohn H. Oakley (1992). $35.00 26 The Sanctuary ofAthena Nike inAthens:Architectural Stages and Chronology, by Ira S. Mark (1993). $50.00 27

Proceedingsof the International Conferenceon GreekArchitectural Terracottasof the Classical and Hellenistic Periods, December 12-15, 1991,

edited by Nancy A. Winter (1994). $120.00 28 Studies in Archaic Corinthian VasePainting, by D. A. Amyx and Patricia Lawrence(1996). $65.00 29 TheAthenianGrain-TaxLaw of 374/3 B.C., by RonaldS. Stroud(1998). $35.00

H THE OF

E$PER

JOURNAL

OF THE

CL A S S I C A L

AMERICAN

STUDIES

AT VOLUME

AM .p

SCHOOL ATTHENS 69:

NUMBER 2000

gJULY-SEPTEMBER

Clasia Stde 2000

tAhns

3

The American School of Classical Studies at Athens is a research and teaching institution dedicated to advanced study of the archaeology,art, history, philosophy, language, and literature of Greece and the Greek world. Established in 1881 by a consortium of nine American universities, the School now serves graduate students and scholars from more than 150 affiliated colleges and universities, acting as a base for research and study in Greece. The main buildings of the School and its library are located in Athens, with administrative and publications offices in Princeton, New Jersey. As part of its mission, the School directs ongoing excavations in the Athenian Agora and at Corinth and sponsors all other American-led excavations and surveys on Greek soil. It is the official link between American archaeologists and classicists and the Archaeological Service of the Greek Ministry of Culture and, as such, is dedicated to the wise management of cultural resources and to the dissemination of knowledge of the classical world. Inquiries about membership in the School or participation in the Summer Sessions should be sent to the American School of Classical Studies at Athens, 6-8 Charlton Street, Princeton, New Jersey 08540-5232. Hesperiais published quarterlyby the American School of Classical Studies at Athens. Founded in 1932 and devoted primarily to the timely publication of reports on School-sponsored and School-directed projects, Hesperiawelcomes submissions from all scholars working in the fields of Greek archaeology,art, epigraphy,history, and literature,from earliest prehistoric times onward. Hesperiais a refereedjournal.

THE

JOURNAL

OF THE STUDIES

OF CLASSICAL

PUBLICATIONS STAFF

EVELYN

AMERICAN

SCHOOL

AT ATHENS

HARRISON

B.

Eumolpos Arrives in Eleusis

267

EDITOR-IN-CHIEF

KerriCox EDITOR,

NATALIA

Hesperia

TraceyCullen EDITOR

ASSOCIATE

Michael Fitzgerald

Late Hellenistic Pottery in Athens: A New Deposit and FurtherThoughts on the Association of Pottery and Societal Change

293

ASSISTANT

EDITORIAL

SuzanneAbrams PRODUCTION

MANAGER

SarahGeorge Figueira CREATIVE

VOGEIKOFF-BROGAN

MICHAEL

D.

DIXON

A New Latin and Greek Inscription from Corinth

335

COORDINATOR

JordanPeled MARKETING

MANAGER

PatriciaTanner

PUBLICATIONS COMMITTEE Carol C. Mattusch (Chairman) George Mason University Darice Birge Loyola Universityof Chicago Jack L. Davis Universityof Cincinnati JeniferNeils Case Western ReserveUniversity James P. Sickinger Florida State University KathleenW. Slane Universityof Missouri-Columbia Stephen V. Tracy (ex officio)

The Ohio State University

JOHN AND

BENNET, FARIBA

JACK

L. DAVIS,

ZARINEBAF-SHAHR

Pylos Regional Archaeological Project, Part III: Sir William Gell's Itinerary in the Pylia and Regional Landscapes in the Morea in the Second Ottoman Period

343

Submissions: Manuscripts and communications should be addressed to Dr. TraceyCullen, Editor,Hesperia, American School of ClassicalStudies at Athens, 6-8 CharltonStreet,Princeton,NewJersey08540; tel. 609-683-0800; fax 609-924-0578; e-mail [email protected] photocopies of illustrationsmust be submittedin triplicate;originalartworkand photographs should not be sent unless prior arrangementsare made with the Editor. A short abstractsummarizingthe majorconclusionsof the articleshould also be included.Articles aresubmittedto a double-blindreviewprocessand authors arerequestedto preparetheir manuscriptsaccordingly,without their name or affiliationappearing.The style for manuscriptpreparation,notes, bibliography,and other informationon submissionscan be found in the Guidelinesfor Authors published in Hesperia 62, 1993, pp. i-xvi; on the School's website (www.ascsa.org);or by writing to ASCSA Publicationsat the above address. The AmericanSchool of ClassicalStudies at Athens will not knowinglyprint in Hesperia or any of its other publicationsthe announcementor initial scholarlypresentationof anyobjectacquiredafterDecember30, 1973, by anymeans other than through an officially sanctioned excavationor survey,unless the objectwas partof a previouslyexisting collectionor was legally exportedfrom the countryof origin. Subscriptions:The annualsubscriptionprice,payablein advancein dollars,is $55.00 for individuals,$65.00 for institutions, and $33.00 for studentswith valid I.D. Canada and other countries add $9.50 postage. We accept VISA and MasterCard.Published quarterly.Reprintsof Hesperia 1-41, Index Volume 1 (Hesperia1-10 and Supplements 1-6), and Supplements 1-12 and Supplement 18 should be orderedfrom Swets and Zeitlinger,b.v., P.O. Box 810, 2160 SZ Lisse, Netherlands. Single issues (currentand back numbers when available)of Hesperia 42 and following areavailablefor $15.00 each plus postagefromthe AmericanSchoolof ClassicalStudiesat Athens, 6-8 Charlton Street,Princeton,NewJersey 08540-5232 U.S.A. OrderIndex Volume2 and Supplements 13 and following from UniversityMuseum Publications,Universityof PennsylvaniaMuseum of Archaeologyand Anthropology,33rd and Spruce Streets,Philadelphia,Pennsylvania19104 U.S.A.

Copyright ?) 2000 The American School of ClassicalStudies at Athens Producedat EdwardsBrothers, Inc., Ann Arbor,Michigan. Design by Ellen McKie. Cover photographcourtesy Deutsches Archaologisches Institut,Athens. Periodicalspostage paid at Princeton,New Jersey,and at additionalmailing offices. Postmaster. Send addresschanges to Hesperia, American School of

ClassicalStudies at Athens, 6-8 Charlton Street,Princeton, New Jersey08540-5232 U.S.A. ISSN 0018-098X ISBN 87661-500-0

hesper ia 69, 2000 Pa ge s 2 6 7 – 29 1

EUMOLPOS ARRIVES IN ELEUSIS

A B S T RAC T The boy between Demeter and Kore in the Great Eleusinian Relief is commonly called Triptolemos, but Ploutos and Demophon have recently been proposed instead. Here I suggest that he is Eumolpos, receiving from Demeter a tainia depicted in paint. In attitude he resembles another son of Poseidon: Theseus, arriving in Athens to claim his inheritance. Other representations of Eumolpos are identified, youthful on the Parthenon west frieze and on a votive relief from the Athenian Eleusinion, and as a strong young man on the Parthenon west pediment. The Great Eleusinian Relief may commemorate a late-5th-century b.c. historical event.

In his richly informative book on Eleusinian iconography, Kevin Clinton points out that the identity of the youth in the “Great Eleusinian Relief ” in Athens (Fig. 1) has been a center of controversy ever since the relief was discovered in 1859: “A controversy that plays itself out against a backdrop of widespread belief that the figure is Triptolemos.”1 At present this remains the most favored identification, but Clinton advances serious arguments against it and proposes instead to understand the boy as Ploutos, handing up a bunch of wheat to Demeter rather than receiving one from her.2 1. I owe thanks to three anonymous Hesperia reviewers for queries and suggestions that helped to clarify my thoughts and my expression. Elizabeth Milleker discussed problems in front of the relief in the Metropolitan Museum of Art, which combines a cast of the original with fragments of a Roman copy, and was especially helpful in obtaining photographs. Others to whom I am grateful for photographs and/or permissions are Hans Rupprecht Goette of the Deutsches Archäologisches Institut, Athens;

Jasper Gaunt of the Institute of Fine Arts, New York University; D. Widmer, Basel; Skulpturhalle, Basel; the late John Travlos; and the late Frank Brommer. Clinton 1992, p. 39; see also pp. 48–49. Simon (1985, p. 117) repeats from earlier editions of the same work the mistaken statement that the relief was found “innerhalb des heiligen Bezirks von Eleusis.” As the original findspot next to the Church of St. Zacharias helped to foster the identification of the subject as Triptolemos

(Clinton 1992, pp. 48–49 with full citations), so Simon’s acceptance of a provenance inside the sanctuary led her to consider the work a sacred icon rather than a votive offering. In her most recent article, Simon corrects the information on the provenance of the relief (1998, p. 373), but maintains her hypothesis that it stood originally inside the Telesterion (pp. 377–378). 2. Clinton 1992, pp. 38–55, 83–84, 96, 103–104, 133–135; LIMC VII, 1994, pp. 416–420, s.v. Ploutos (K. Clinton).

268

e v e ly n b . h a r r i s o n Figure 1. Great Eleusinian Relief, Athens, National Museum 126. Courtesy Deutsches Archäologisches Institut, Athens

Both Triptolemos and Ploutos have standard attributes that they are very rarely without: the winged chariot for Triptolemos and the cornucopia for Ploutos. Triptolemos is ready to ride long distances to spread his gifts abroad; Ploutos comes up from the fertile earth wherever the seed has taken root. Neither is specifically related to travel on foot. Before deciding that the boy in the relief must be either Triptolemos or Ploutos and searching for arguments to explain his deviation from the standard iconography of one or the other, it might be preferable to look for another possibility. Elements that deserve more consideration are the positions of the hands of the figures, the placement of the attributes assumed to have been added in metal or in paint, the stance and implied character of the boy, and the significance of his footwear.3 The right hand of Demeter and that of the boy are closer together

3. Simon (1998) considers some of these questions and offers a suggestion that is, as far as I know, entirely new. She identifies the boy as Demophon, the nursling of Demeter in the Homeric Hymn. She makes a number of useful observations on the figures of the relief, which I am happy to accept. These observations do not, however, persuade me to abandon my own interpretation (the present article was largely written before hers appeared), though I have profited from some of her insights.

eumolpos arrives in eleusis

Figure 2. Detail of fragmentary Roman copy of the Great Eleusinian Relief incorporated in a cast of the original. Hands of Demeter and boy. New York, Metropolitan Museum of Art, Rogers Fund, 1914 (14.130.9). Photo S. Houston, courtesy Museum

4. See the numerous examples in the illustrations in Peschlow-Bindokat 1972 and Clinton 1992. 5. Clinton (1992, pp. 39–40) argues that in spite of the absence of drilled holes for attaching any object in the boy’s right hand, he might have held “a large number of stalks bunched together, so tightly pressed against his thumb and fingers that they could not slip out.” If, however, the sculptor intended to depict a tightly grasped bundle of stalks, such as is represented in the figure on the Lakrateides relief (Clinton 1992, figs. 5–6) and in the Hierophant dedication (Clinton 1992, fig. 55; here Fig. 10), the boy’s four fingers should curl together and the thumb should turn inward to touch or rest on top of the index finger. This is the normal hand position for grasping a bunch of twigs held by priests or other participants in a sacrificial procession (cf. Parthenon North X 41). In none of these examples is the bunch

269

than in most representations of objects being offered by one person to another (Fig. 2). In particular, this proximity is not convenient for depicting the transfer of a bunch of wheat. In vase paintings, the stalks spread out freely from the hand that grasps them and the heads nod, either symmetrically or in one direction. The length of the stalks and the weight of the heads proclaim the wealth of the harvest.4 The proponents of Triptolemos would restore the stalks held in Demeter’s hand, above which they should fan out to display the ripened heads against the background, while the boy is not yet fully grasping them. In fact, there is not enough background space above her hand for the grain to spread out properly, especially since there are no holes to attach a bunch of metal wheat that could have overlapped the scepter. The relation of the hands is even less favorable for Ploutos showing the grain to Demeter, since the boy’s hand is open as if reaching for something not yet fully grasped (Fig. 2).5 It is more probable that whatever was portrayed was painted on the marble. Demeter’s right index finger is lifted free, and her right thumb is hidden behind the other fingers. She must have held something between her thumb and one or more fingers, probably the middle one. The boy’s thumb, now broken off, was widely separated from his index finger, which has left its trace against the background. Though he was not grasping any object, something could have extended below Demeter’s hand over his open palm. The easiest thing to depict convincingly in paint would be a long fillet (tainia). Parallels from vase painting suggest that the goddess held it doubled, with a loop visible above her hand. Compare the representation on a Paestan neck-amphora by Python, made around the middle of the 4th century b.c. (Fig. 3).6 The orientation of stalks or twigs being offered to another person. Rather it seems to be held as an attribute of the one who carries it. Vase painting does not differ from sculpture in this kind of representation. Cf. Athena holding the suppliant twigs of the Seven against Thebes on the volute krater in Ferrara by the Painter of Bologna 279, ARV 2 612, no. 1. This vase also shows the gestures of the Seven greeting each other. The right hand of the boy on the Eleusinian Relief differs from these examples in that the index finger does not curl around together with the other fingers. The back of it adhered to the relief background as far as the second joint, where it was bent and is broken off. The thumb, whose base projects in high relief from the shallow concavity of the palm, rises vertically at such an angle that its tip cannot possibly have made contact with any of the fingers. That the three curled fingers were not

tightly curled is evident from the spacing of the two folds of skin at the base of the little finger. There would have been no need for the sculptor to model these folds if they were to be obscured by a cluster of stems. Finally, if the boy held an attached object, it would be strange if the sculptor who used drilled holes to attach metal additions elsewhere in the relief did not use them here. Simon’s suggestion (1998, pp. 376– 377) that Demeter and the boy are simply addressing each other is not demonstrably impossible, but it does not explain why Demeter’s thumb is hidden or why the boy’s hand is turned toward the viewer rather than the goddess. 6. Trendall 1987, pl. 111:a. Once New York Market, now in a private collection in New Jersey. For the date of Python, see the chronological chart in Trendall 1989, p. 271.

270

e v e ly n b . h a r r i s o n

of the scene is opposite that of the relief, with the goddess on the right and the boy on the left, and we can see exactly how the ribbon is held and how the ends fall. A tall female (probably Aphrodite, though she wears the same “Argive” peplos and shoulder mantle as Demeter in the relief ) holds the fillet up so high that the boy Eros, shorter than the boy on the relief, has to reach up for it. He stretches up both hands, with thumbs spread outward, and one end of the ribbon falls across his left palm as we must imagine one end of the ribbon in the Eleusinian relief to fall across the open right palm of Demeter’s protégé. The raised forearm of the latter is in low relief against the background and his wrist so positioned that a fillet could fall almost vertically over it and end against the flat background without being overlapped by his cloak. The other end, perhaps shorter as on the vase, will have been partly visible in the narrow space between the two hands and the overfall of Demeter’s peplos and can have emerged more fully farther down. It has generally been assumed that Kore, whose hand is visible above the boy’s head (Fig. 1), is crowning him with a wreath. A drilled hole in front of his forehead could have served to attach a wreath of metal. Lambert Schneider denied that there is a corresponding hole at the back of the boy’s head and concluded that he wore no wreath, but since wreaths are worn in most representations of participants in the Eleusinian Mysteries, be they mythological or contemporary individuals, a wreath is to be expected.7 Kore’s hand does not appear to be grasping the wreath, however; it is held in an open position above the crown of the boy’s head. The idea that she is anointing him with perfumed oil has been recently advanced by Simon, with supporting observations by Erika Zwierlein-Diehl and John Boardman. The latter remarked that the smooth hair on top of the boy’s head looks wet.8 Since the ends of Kore’s fingers and most of her thumb are broken off, one cannot refute the suggestion that she held a small perfume vase in her hand, but if so, the manner of its attachment is not clear. There is ample background space for an aryballos to be painted hanging on the wall above the boy.9 This would make it clear that Kore’s action is one of smoothing the oil on his hair (perhaps the vertical position of her hand is intended to leave it visible above the wreath to a viewer from below). Her action serves to establish a feeling of intimacy between Kore and the boy in this scene, while Demeter’s attitude is more formal. 7. Schneider 1973, pp. 105–106. His idea that the hole in front of the boy’s forehead was made to attach a forehead knot in metal is unconvincing. Given the profile view, the knot, if there were one, should have been carved together with the background. If some damage or oversight prevented this, it would have been reasonable to make a marble patch, but there is no dressing for this on the stone. Simon (1998, p. 376)

accepts the hair knot, however, and does not restore a wreath. For a 5thcentury example of a metal wreath attached only to the front of the head, see below (Fig. 13). 8. Simon 1998, pp. 381–383, note 48. 9. For a hanging aryballos painted on marble, see the late-5th- or early4th-century tombstone in the Kerameikos: Robertson 1959, pp. 153, 156.

Figure 3. Paestan neck-amphora by Python. Aphrodite and Eros. Photo D. Widmer, Basel

eumolpos arrives in eleusis

271

Figure 4. Red-figure cup by the Briseis Painter. Tondo, Theseus received by Amphitrite. New York, Metropolitan Museum of Art. Purchase, Joseph Pulitzer Bequest, 1953 (53.11.4) and gift of Emily Dickinson Blake Vermeule, 1970 (1970.46). Courtesy Museum

10. Schwarz 1987, p. 196. 11. See Neils 1987 for discussion of these adventures and their illustration in art. They are first depicted in Attic vase painting of the late 6th century and become especially popular in the 5th. 12. The story is told in Bacchylides’ Third Dithyramb, Bacchyl. 17. For discussion, see Neils 1987, p. 10. 13. See Neils 1987, pl. X, figs. 46 and 47; pl. XII, figs. 55 and 56. 14. ARV 2 406, no. 7; von Bothmer 1987, pp. 52–53.

The stance of the boy and the character that it implies are essential clues to his identity. All agree that he is on familiar terms with the goddesses and accepted as their protégé, but he seems to be presenting himself formally for their acceptance like a soldier reporting to his captain. Schwarz expresses his attitude well: “stoltz und ehrfürchtig zugleich.”10 Attic vase painting offers a striking parallel in the representation of the young Theseus arriving in Athens after he has proved his paternity by passing through a series of trials. He first had to establish himself as the true son and heir of Aegeus, the King of Athens, by growing strong enough to raise the stone under which his father had hidden the tokens of his identity, a sword and a pair of sandals. With these he made his way on foot from Troizen to Athens, destroying a number of wicked and formidable opponents along the route.11 A separate trial of a very different kind established him as the son of Poseidon. He had to dive into the sea and retrieve a ring that Minos had thrown overboard. He not only recovered the ring but also brought back a crown that Amphitrite gave him along with a purple cloak.12 On first arriving in Athens from Troizen, Theseus in some vase paintings greets his human father, Aegeus. Sometimes Poseidon looks on.13 In a second version, which seems to represent his arrival after he has slain the Minotaur and brought home the rescued Athenian children, he greets Athena directly, while the grateful mothers surround him, offering fillets. The scene appears on a red-figure cup by the Briseis Painter in New York, made around 480–475 b.c. (Figs. 4–7).14 The cup also shows Theseus in

272

e v e ly n b . h a r r i s o n

Figure 5. Side A of cup, Theseus at the palace of Poseidon prepares to leave. Courtesy Metropolitan Museum of Art, New York

Figure 6. Side B of cup, Theseus arrives in Athens. Courtesy Metropolitan Museum of Art, New York

eumolpos arrives in eleusis

273

Figure 7. Detail of side A of cup. Courtesy Metropolitan Museum of Art, New York

15. Bacchyl. 17.105–106. 16. The famous cup by Onesimos in the Louvre, ARV 2 318, no. 1, earlier than that of the Briseis Painter, depicts in its tondo a very young Theseus, supported on the hands of a small Triton, greeting Amphitrite enthroned in her chamber under the sea. Her left hand held “a large white wreath, now largely disappeared” (Neils 1987, pp. 60–61). Athena, mistress of Athens, with her owl in her hand, stands tall in the center of the tondo, overlapped by the hands of Theseus and Amphitrite. Neils points out that Athena, like Theseus, is a guest of Amphitrite. She inclines her head graciously to greet the mistress of the sea.

the tondo, being presented with a fillet by Amphitrite (Fig. 4). He stretches out his open hands below the raised hand of the goddess while his swaying stance adjusts to the underwater currents. On the outside of the cup he appears outside Poseidon’s palace, gently steadied by a large Triton of kingly aspect, while Poseidon bids him farewell and Nereids flank the scene (Fig. 5), much as the Athenian women do on the other side of the cup, where he presents himself to Athena (Fig. 6). Under the sea, Theseus wears his sword in both scenes, but, appropriately, he does not wear sandals. As he gestures in reverence toward Poseidon and in farewell to Amphitrite, he is wearing a wreath of flowers, evidently the wreath of roses mentioned by Bacchylides.15 This wreath projects beyond his forehead in the same way that is implied by the attachment hole in front of the forehead of the Eleusinian boy. In the undersea scenes of the Briseis Painter, Theseus is younger than in the scenes of his arrival in Athens, either before or after the Cretan adventure. This may seem to contradict mythological chronology, but that is not the point. Theseus under the sea is the nursling of Poseidon and Amphitrite, carried by the sea creatures in a realm in which he cannot walk.16 On land, by contrast, when he presents himself as heir to the throne of Athens he has learned to make his own way, guided and protected but not transported by the power of Athena. It is as a young prince who has proven his strength and intelligence on purely human terms that he presents himself to the goddess of Athens as ready and worthy to serve her. The central figure of the Great Eleusinian Relief presents himself to the Two Goddesses of Eleusis in the same manner. It is easiest to understand him as another son of Poseidon who is also a descendant of the royal

274

e v e ly n b . h a r r i s o n

family of Athens, to wit, Eumolpos, the founder of the Eleusinian Mysteries and eponym of the genos of the Eumolpidai, from whom the Hierophant, the principal priest of the Mysteries, was chosen. The genealogy that made Eumolpos the son of Poseidon also made him part Thracian: he is the son of Chione, daughter of the Erechtheid Oreithyia, who was carried off from Athens by Boreas, the North Wind. Apollodoros (3.15.4) says that when Chione bore Eumolpos to Poseidon without her father’s knowledge, she threw him into the sea to hide the fact. Poseidon then saved him and carried him to Ethiopia and gave him to Benthesikyme, daughter of Poseidon and Amphitrite, to rear. Most sources take the son of Chione and Poseidon to be the Eumolpos who came as an ally of Eleusis against Athens. Even this strains genealogical logic, since he is the greatgrandson of Erechtheus, but in art relationships are often stressed at the expense of chronology. On the skyphos by Makron in London,17 the swan beside Eumolpos and the scepter in his hand mark him as the first Hierophant, the swan representing his beautiful voice and the scepter his divinely sanctioned authority (Fig. 8). At the same time the images of Poseidon with his dolphin and Amphitrite standing beside him recall the divine paternity of Eumolpos and his childhood immersion in the sea (Fig. 9). Eumolpos, like other important figures in the Eleusinian sphere, is depicted at various ages, and his identifiable representations are not as numerous as those of Triptolemos and Ploutos. Nevertheless, Clinton has

Figure 8. Skyphos by Makron. Eumolpos with swan. London, British Museum E 140. Courtesy Museum

Figure 9. Skyphos by Makron. Poseidon with dolphin. London, British Museum E 140. Courtesy Museum

17. ARV 2 459, no. 3; Clinton 1992, p. 138, “Eumolpos in Attic Vase Painting,” no. 1, with bibliography, pls. 193–194, discussion p. 75; Agora XXXI, pp. 53–55, pls. 20–21.

eumolpos arrives in eleusis

275

Figure 10. Votive relief dedicated to Demeter and Kore by a Hierophant. Antonine period. Athens, temporarily stored in the Agora Museum. Courtesy J. Travlos

18. Clinton 1992, p. 139, “Eumolpos in Sculpture,” no. 1, with bibliography, discussion p. 75. The marble relief was found in excavations by John Travlos in 1959 north of the Olympieion precinct. It has never received full publication, but Eugene Vanderpool gave a good brief account of its finding, with a plan of the site and a photograph of the relief: Vanderpool 1960, p. 268, ill. 1 and pl. 73, fig. 17. 19. The strophion worn by the Hierophant in the Roman relief also occurs on figures of boys connected with Eleusinian ritual, likewise of the Roman period. The crown consisting of a fillet wound around a circlet is discussed by Krug 1968 as Type 12 II, pp. 45–47, 104–106 and Appendix III. She concludes that vase painting does not represent the details of this crown and is therefore not useful for chronology. Since the circlet is completely covered by the ribbon, it is the latter that carries its significance, as in the case of the Hellenistic diadem. Scepter and strophion as symbols of religious authority correspond to the “scepter and ribbons of the god” in Il. 1.14, 28.

succeeded in establishing his Classical iconography based on that of his descendant, the Hierophant as depicted on a classicizing votive relief of the Roman period, made for dedication by a Hierophant from the deme of Hagnous (Fig. 10). This can be taken in conjunction with labeled representations of Eumolpos on vases to distinguish him from other prominent Eleusinian persons. The Hierophant on the relief displays all his official regalia. He approaches the Two Goddesses in a walking pose, with his right foot advanced, and holds out his scepter in front of him. Kore’s left arm overlaps the scepter and the knuckles of his right hand, emphasizing his closeness to the divinities. On his head he wears a strophion and a wreath of myrtle. The strophion consists of a long fillet wound continuously around a circlet of some kind, overlapping so that only the cloth is visible. The wreath rests above the strophion. The Hierophant wears a long-sleeved tunic that is short enough to reveal the tops of his boots. Over it he wears a draped himation. The boots seem to be elaborately decorated.18 Long sleeves and boots are worn by many Eleusinian gods and heroes in Attic vase painting of the late 5th and 4th centuries b.c., as are wreaths and fillets.19 Clinton concludes that the distinctive attribute of Eumolpos is the scepter, which can serve to set him apart from other long-haired youths, such as Eubouleus and Iakchos, in the populous assemblies of Eleusinian gods and heroes on 4th-century Attic vases. Since the protégé of the Two Goddesses on the great marble relief is still a boy, he is not yet entitled to a scepter of his own. Like the boy Theseus in his arrival scenes, Eumolpos has the scepter in his future, not

276

e v e ly n b . h a r r i s o n Figure 11. Neck-amphora by the Niobid Painter. Youth (Eumolpos?) offered wheat by a goddess (Kore?). London, British Museum E 274. Courtesy Museum

in his hand. At the same time, the way in which Demeter holds out her own scepter in front of her, so that it is overlapped by the arm and body of the boy, conveys the transmission of her power to him in the same way as does the spear of Athena in the arrival of Theseus on the Briseis Painter’s cup, where it overlaps his sandaled right foot (Fig. 6). All three figures in the Great Eleusinian Relief wear sandals, but those of the boy are different from those of the goddesses. Known as network sandals, they often appear in Classical vase paintings as worn over a sock that rises partway up the leg and is held in place by straps wrapped around the leg.20 The Briseis Painter’s arriving Theseus wears such sandals, as do other mythical travelers. They also appear on grave reliefs from Boeotia and the Cyclades.21 Theseus and Peirithoos stuck in the Underworld wear them on the Nekyia krater in New York, while Herakles stands barefoot beside them.22 Hermes attending the rising Persephone reveals the tops of such sandals as he comes up out of the ground.23 Hunters and riders also wear network sandals. In any case, it seems to be generally agreed that a boy wearing sandals on both feet cannot be identified as a pa›w éf’ •st¤aw muhye¤w, since he is not barefoot as an initiate ought to be.24 As mentioned above, the ages at which Eumolpos is depicted in art differ. Like other gods and heroes, he tends to be shown bearded in Late Archaic and Classical representations and to lose his beard and become younger as time goes on. The one vase painting in which Eumolpos is identified by inscription, the skyphos by Makron in London, shows him as a mature bearded man with long black hair. He is seated and holds a scepter (Fig. 8).25 A standing youthful male on a neck-amphora by the

20. Morrow 1985, p. 58, pls. 41–42, and see index under “network sandals.” 21. Despinis 1967, pp. 79–81, pl. 35. 22. ARV 2 1086, no. 1; Nekyia Painter: Richter and Hall 1936, no. 135. 23. ARV 2 1012, no. 1; Persephone Painter: Richter and Hall 1936, no. 124. 24. Clinton 1992, p. 49. 25. See note 17, above.

eumolpos arrives in eleusis

277

Figure 12. Relief of the Rheitoi Bridge decree. Eleusis, Archaeological Museum 5093. Courtesy Deutsches Archäologisches Institut, Athens

26. ARV 2 604, no. 53; Clinton 1992, p. 139, no. 12, with bibliography. 27. Clinton 1992, pp. 75–76 and p. 140, no. 2, with bibliography, fig. 60. See also Lawton 1995, p. 54 and pp. 82– 83, no. 3, pl. 2. Lawton refers to Clinton 1992 in her text (p. 54, note 113) for evidence of Triptolemos’s “divine status at Eleusis,” but she maintains the identification of the male figure as Triptolemos without mentioning Clinton’s discussion of the figure. 28. Agora inv. S 1119: Agora XXXI, p. 217, no. 2, pl. 36; LIMC IV.1, 1988, p. 875, no. 377, s.v. Demeter (L. Beschi). Both Beschi and Miles identify the young male as Triptolemos. 29. Close examination of the relief reveals some details not mentioned by Miles (Agora XXXI) that may be relevant to the discussion of the youth’s identity. His head is only a little smaller than that of Kore, as his height was only a little less than hers. The approximate height of Kore’s head, chin to crown, is ca. 0.07 m, that of the boy’s ca. 0.065 m. The relief appears to have been intentionally mutilated. There are tiny point marks in the area of Kore’s right eye and similar marks below and behind the right eye of the boy. The mouths and chins of both figures are broken off, but the line of intersection of the face with the neck is clearly preserved on both. The locks on top of the boy’s head are simply carved with long strokes; one curls up against the background behind the two holes drilled for attaching a wreath. There is no corresponding hole at the back of the head, but an indentation suggests that the wreath overlapped the large locks above and behind his ear. Similar locks cling to the back of his neck below this. It seems clear that Kore did not wear a bridal crown in addition to her veil. Her front hair is battered, but one can see that it overlapped the right ear, revealing only the lobe. It appears to have been combed forward above the forehead. There may have been curls, now broken away.

Niobid Painter in London holds a scepter in his left hand and pours a libation with his right to a goddess who offers him a bundle of wheat (Fig. 11). Clinton remarks, “The scepter and the wheat suggest importance in the cult and remind us of the Hierophant in the Olympieion relief. The figure therefore seems to be Eumolpos.”26 The male figure who greets Athena on the stele of the Rheitoi Bridge decree of 422/1 (IG I3 79) in Eleusis (Fig. 12) apparently held a scepter, rendered in paint as was Athena’s spear. Clinton calls him “most likely Eumolpos.”27 He is draped in a himation without a chiton, like the youth on the Niobid Painter’s neck-amphora. He is a little shorter than the Two Goddesses, but nearly the same height as Athena. He is evidently beardless, though all the faces are damaged. His hair is cut shorter than that of the boy of the Great Eleusinian Relief, but not as short as that of an ephebe on a fragment of a late-5th-century votive relief found in 1939 in the Athenian Agora (S 1119). There is an interesting correspondence between the Hierophant relief (Fig. 10), probably made in the Early Antonine period, and Agora S 1119, mentioned by L. Beschi in LIMC but fully published only recently by M. Miles (Fig. 13).28 Though it was found in the demolition of a modern house outside the City Eleusinion to the west, Miles is surely right to associate it with that sanctuary. At the broken left edge of the relief there is preserved a raised left forearm and hand grasping a scepter whose shaft overlaps the forearm and an adjacent bit of hanging drapery. The hand overlaps the ovolo beneath the taenia that crowned the relief. The tip of the scepter has disappeared in the break.29

278

e v e ly n b . h a r r i s o n Figure 13. Fragment of a votive relief (Agora S 1119) showing Demeter, Kore, and a youth (Eumolpos?). Courtesy American School of Classical Studies at Athens, Agora Excavations

The angle of the arm and scepter is very like that in the figure of Demeter in the Hierophant relief (Fig. 10). The Demeter of the Agora votive must have been enthroned like her, with Kore standing in front of her. Kore has a veil over her head in both reliefs, and her head is inclined in both toward the male figure who stands facing her. Whereas Kore in the Hierophant relief holds a downturned torch by her side,30 in the late-5th-century relief she holds a lighted torch in front of her. In place of the Hierophant, the male figure next to her is a boyish youth, probably an ephebe, for he wears a chlamys. Rather than standing stiffly like the Hierophant, he inclines his head toward the goddess. This creates a feeling of familiar intimacy between the boy and Kore such as we sense in the Great Eleusinian Relief, where she lays her hand on his head. On the Hierophant relief, however, Kore inclines her head toward the Hierophant in the same way, and a special relationship between them is established by the manner in which her left arm overlaps his right hand and scepter.31 It may well be that the Two Goddesses on the Hierophant relief, with their strikingly classicizing dress and faces, are copied from a famous prototype, perhaps a painting, and that this also influenced the sculptor of the Agora relief. The similar composition strongly suggests that the young male on Agora S 1119 is Eumolpos. He evidently wore a wreath, as is indicated by the two drilled holes in the short hair above his forehead. Though nothing survives of any attributes he may have held, his lowered right shoulder and raised left one would fit the action depicted on the Niobid Painter’s neck-amphora (Fig. 11), pouring a libation with the right hand while grasping an upright scepter with the left.32 The boy on the Great Eleusinian Relief makes a gesture with his mantle that could be read either as taking it off or putting it on. While one end remains draped over his right shoulder, he grasps the other end with his lowered left hand. Simon, who identifies the boy as Demophon, explains that his act of robing must have followed a ritual of disrobing, namely the bathing in the sea on the second day of the Mysteries that was known as Halade Mystai. Demophon in his role as King of Eleusis would stand for all the mystai. The boy would present himself to the goddesses freshly

30. As no flames are indicated in relief, the question arises whether this torch may symbolize the darkness that precedes the reappearance of Kore in the Eleusinian ritual. See Clinton 1992, pp. 84–90. Because we do not know the sanctuary for which the relief was intended, no certainty is possible. 31. That the feeling of intimacy exists between the boy and Kore rather than between the boy and Demeter may be an additional argument against Simon’s (1998) identification of the boy as Demophon. 32. See above, note 26. Both Clinton (1992) and Schwarz (1987) recognize that this representation by the Niobid Painter is unusual. If the goddess offering the wheat is Demeter, we cannot invoke the vase as a parallel for the Agora relief with as much confidence as we might if we could call her Kore. Both goddesses offer grain; both can wear crowns and long back hair without shoulder locks, but two aspects suggest that the goddess on this amphora is Kore. First, the Niobid Painter seems to have taken pleasure in assimilating the young man and the goddess facing him to Apollo and Artemis on the opposite side. The laurel wreath and sideburns on the young man echo those of the youthful Apollo, and the crown and long back hair of the goddess offering wheat echo those of the maiden Artemis. Second is the similarity of the back hair of this goddess to the figure labeled “Kore” on the volute krater in Stanford (Clinton 1992, fig. 12; Raubitschek and Raubitschek 1982, pl. 15). Schwarz (1987, p. 121) says only “wohl Demeter” but she is sure the youth is Triptolemos. Clinton (1992, p. 139, no. 12) says simply “Demeter” and of the youth, “apparently Eumolpos.”

eumolpos arrives in eleusis

Figure 14. Parthenon west frieze figure 30 (Eumolpos?) from a cast. Photo D. Widmer, courtesy Skulpturhalle, Basel

33. Simon 1998, pp. 380–381. 34. For a valuable succinct account and full references, see Kearns 1989, p. 163. The story that Chione threw her child Eumolpos into the sea and that Poseidon saved him and took him to Ethiopia for Benthesikyme to rear is most fully told in Apollod. 3.15.4 (together with a string of misdeeds and exiles not to be found in most sources), but the core myth seems to be no later than the time of Euripides’ Erechtheus. In addition to the references in Kearns, see Collard et al. 1995, pp. 156–157, frag. 349, text, translations, and commentary. For the declaration of Zeus that a descendant of the son of Chione and Poseidon will found the Mysteries, see Collard et al. 1995, pp. 174–175, lines 102–114 and commentary, pp. 193–194, on these lines. For the story of another divine ancestor who was thrown into the sea by his mother, see Il. 18.393–405, where Hephaistos relates how his mother, wanting to hide him, threw him out of Olympos, and how Eurynome and Thetis saved him and hid him in a cave by the River of Ocean. One could imagine the Ethiopian hiding place of Eumolpos to

279

purified from his sea-bath. Simon connects the anointing of the boy’s hair by Kore as an appropriate ritual following the bath.33 All this would be suitable for Eumolpos in a double sense. As the son of Poseidon and the Erechtheid Chione, daughter of Oreithyia, saved from the sea and reared by Benthesikyme, the daughter of Poseidon and Amphitrite, Eumolpos grew up to become a military defender of Eleusis and in some versions the founder of the Mysteries.34 I propose that Eumolpos is twice represented on the west end of the Parthenon, once in the south corner of the pediment and again at a lower level in the south end of the frieze. In the frieze, the act of dressing is impersonated by a nude and apparently barefoot young male (West XVI, 30; Fig. 14).35 He holds up his mantle toward his left side with both hands hidden under the cloth. Ian Jenkins notes that his arms seem to point the way around the corner to the south side of the frieze and so to express a connection with the south cavalcade. The first group of riders on the south side is distinguished by elements of Thracian dress and probably represents the tribe of Hippothontis, whose eponymous hero, Hippothoon, had a sanctuary at Eleusis.36 On the frieze, Eumolpos would be shown as the ancestor of the genos of the Eumolpidai, the clan to which the Hierophant belonged.37 Next to him (West XV, 29) is a figure so much like Hermes in his dress and attitude that he ought to be Keryx, the son of Hermes and ancestor of the genos of the Kerykes, the clan that shared with the Eumolpidai the administration of the Sanctuary of Eleusis.38 West 29, with his left foot propped have also been a cave by the River of Ocean, in its southernmost reaches. 35. Most fully described and illustrated by Brommer 1977, p. 23, pls. 5, 45, 46, and 116:2. Carrey was at a particular disadvantage in trying to draw this figure because of the heavy shadow cast by the ceiling beam above its head. Some of this also affected his vision of the neighboring figure, West 29. His drawing indicates the existence of this shadow over the two figures. Irregularities in the surface of 30 led him to believe the figure was a chitonclad female. He also omitted the petasos and chlamys of 29. The Elgin cast (see Jenkins 1990, p. 113, pl. 19) remains the best record of the head, though the right side of the face was already damaged and filled out in the cast. The short curls on top of the head are still clearly visible in the cast and are not unlike those of the boy in the Agora relief. For further information see Berger and GislerHuwiler 1996, pp. 55–56 (Text), pls. 34– 37 (Plates). 36. See Kron 1976, pp. 177–187; Kearns 1989, p. 173, for Hippothoon; Jenkins 1994, p. 111, for the pose of West 30 as pointing to a connection with the

south frieze; Brommer 1977, pl. 46:4, for a good corner view. 37. In Harrison 1984, p. 234, I suggested that the figures in the west frieze belong to early times and may stand for gene, as the north frieze seems to represent the time of the four tribes and the south frieze recent times, with the ten tribes of Kleisthenes. 38. Clinton 1974, p. 8. See Kearns 1989, p. 177, for Keryx’s somewhat varied genealogies. The one proclaimed by the Kerykes themselves made him the son of Hermes by one of the daughters of Kekrops (Paus. 1.38.3). This specifically Athenian genealogy is given little attention in LIMC VI, 1992, pp. 36–38, s.v. Keryx (E. Simon). Resemblance to Hermes is not treated as an element in the iconography of Keryx and so there is no occasion to mention either in approval or rejection my tentative identification (1984, pp. 234 and 417, note 61) of West 29 as Keryx. By taking the Dadouchos rather than the Hierokeryx as the iconographical prototype of the ancestor of the Kerykes, Simon discusses as possible representations of Keryx a variety of torch-bearing males to whom Clinton 1992 has given other names.

280

e v e ly n b . h a r r i s o n

on a large rock, is in the act of tying his sandal. His right foot is still bare; no sandal sole is indicated. Attributes that West 29 shares with Hermes are his chlamys, his petasos, and the act of tying his sandal. He does not carry a kerykeion, and there are no wings on his shoes or hat. The figures of the west frieze may all be heroes, but none of them is a god.39 Because of the indications that the Parthenon frieze as a whole has as its subject the worship of the gods in joyous festivals by the Athenians, we may guess that these two ancestors of the clans that administer the Sanctuary of Eleusis together represent the Festival of the Eleusinia. I would take the rock on which Keryx rests his foot not simply as a mounting block, but as a topographical sign, locating the figure on the sacred hill of Eleusis. The rock is irregularly shaped and has a bumpy surface like that of another such rock that appears in the northern half of the west frieze. On this another rider (West VI, 12), also facing south, props up his left foot and ties his sandal. His right foot is already shod. He differs from West XV, 29 in wearing a helmet. West VI, 11, directly behind him, is also helmeted and wears a cuirass. Elements of defensive armor seem appropriately associated with the Acropolis, the citadel of Athens.40 Thus the rock on which West VI, 12 rests his foot might be that of the Acropolis, and he might be Boutes, the eponym of the Eteoboutadai, who had an altar in the Erechtheion (Paus. 1.26.5).41 The human figures in the west frieze number thirty, and this number is by no means so securely linked to the gene as are the numbers ten and four to the Kleisthenic and pre-Kleisthenic tribes. It does figure, however, in a fragment of the Aristotelian Constitution of the Athenians, according to which “in olden days” (pala¤) the whole citizen body was divided into 360 gene, consisting of thirty men each, one genos for every day of the year.42 Thus a depiction of thirty heroes, each representing a genos, could be tied into the yearly cycle that seems to dominate the frieze as a whole. Eumolpos in this context would take his place as an Athenian citizen, like his descendant, the Hierophant. In between the rocks that represent Eleusis and Athens is a partly preserved segment of sloping terrain against which a magnificent hero with flying chlamys braces his foot while he exerts all his strength to control a splendid rearing horse (West VIII, 15). This hero had metal (presumably golden) sandal straps, attested by drilled holes for their attachment, three in the right leg and one in the left.43 I have suggested elsewhere that this hero is Theseus as King, adult and bearded, and that his act of bringing the spirited horse under control symbolizes the synoikismos, the union of all the previously autonomous towns of Attica to form the single Athenian state. Since festival activities are a principal subject of the frieze, this figure might also stand for the festival of the Synoikia on the 16th of Hekatombaion, twelve days before the Panathenaia.44 This opens up the possibility that other festivals, as unidentifiable to us as the majority of the genos heroes of the west frieze, might be represented by these figures. The Parthenon pediments, in contrast to the frieze, are more concerned with what the gods do for the Athenians than with what the Athenians do for the gods. West pediment V (Figs. 15–16), the figure that I have earlier proposed as Eumolpos, personifies the Mysteries and as such

39. Kardara (1961, pp. 151–152, pl. 7) suggested that West 23 was Hermes himself, but this has not been generally accepted. The rule seems to hold that no person is shown more than once on the frieze, though the same god or hero may appear in other sections of the Parthenon sculptures: metopes, pediments, and statue base. 40. I have suggested (Harrison 1984, pp. 232–233) that the three groups of riders on the south frieze who wear elements of defensive armor may represent the three tribes whose eponymous heroes are associated with the Acropolis: Kekropis, Erechtheis, and Pandionis. 41. On the elaborate 4th-century hydria from Cumae in St. Petersburg known as the “Regina Vasorum,” Athena sits on an irregular rock that evidently stands for the Acropolis (Clinton 1992, pp. 78–79, ill. 9). It appears to have a cave in its lower part. 42. Aristotle frag. 385 Rose. 43. Harrison 1984, p. 234. See Brommer 1977, p. 13, for a description of the holes, their interpretation as sandal straps, and the observation that nowhere else on the frieze were sandal straps made separately in this way. See Brommer 1977, pl. 23, and, for an even clearer depiction of the holes, Robertson and Frantz 1975, pl. 11. 44. Harrison 1984, p. 234. See Parke 1977, pp. 30–32; Deubner 1932, pp. 37–38; Simon 1983, p. 50; Figueira 1984, pp. 465–466. N. Robertson (1992, pp. 32–43) gives an explanation of the ultimate origin of the festival that is too complex to be useful in interpreting a sculptural representation of the 5th century b.c. If Parthenon West 15 alludes to Theseus and the Synoikia, its version will have been closer to Plutarch’s than to Robertson’s.

eumolpos arrives in eleusis

Figure 15. Parthenon west pediment V (Eumolpos?). Frontal view. After Brommer 1963, pl. 128:1

Figure 16. Parthenon west pediment V (Eumolpos?). Side view. After Brommer 1963, pl. 128:2

45. Harrison 1967a, p. 9, note 55. For a cult of Poseidon Pater, outside the entrance of the sanctuary at Eleusis, see Paus. 1.38.6; Clinton 1992, p. 116. 46. Spaeth (1991) identifies the seated female figures in the southern half of the west pediment with members of the royal family of Eleusis instead of that of Athens, but in doing

281

is the gift of Poseidon who fathered him.45 He is linked to the autochthonous Athenians by the female line of two generations of Erechtheids, who produced heroes by union with male divinities. These divine fathers played a special role in Athens. The first of the Erechtheids to the right of the central divinities in the west pediment (Fig. 17) is Oreithyia, the bride of the North Wind.46 She is recognized by the fact that she holds infant twins, P and R, Kalais and Zetes, and the union with Boreas that engendered them is signified by her wind-fluttered mantle and chiton hem. Without the favor and the power of the North Wind, neither the destruction of the Persian fleet at Cape Artemision nor the success of the Delian League (which had become an Empire by the time work was begun on the Parthenon sculptures) would have been possible. After Oreithyia and her children comes another daughter of Erechtheus who was equally important to the Athenian self-image in the 5th century. Kreousa, T, became the mother by Apollo of Ion, S, eponym of the Ionians and the key to the myth that Athens was the mother city of all the Ionians.47 By the same token Apollo became Patroos, father of the Athenians and the other Ionians. this she must deny iconographical validity to the wind-ruffled drapery of West Q and play down the importance of her twins. Concerning the proposal of Weidauer and Krauskopf (1993) to substitute Erechtheus and his daughters for Kekrops and his daughters in the northern half of the pediment, see Harrison 1997, p. 124, note 29.

47. The suggestion in LIMC IV, 1988, pp. 56–59, s.v. Eumolpos (L. Weidauer) under “Deutung vermutet” (Eumolpos 2, p. 56), that West T is Chione and West S Eumolpos violates the generational sequence of the Erechtheids, inserting the granddaughter and great-grandson of Erechtheus between his daughters and their offspring.

282

e v e ly n b . h a r r i s o n

The next figure, U, seems to have been maidenly and not an ancestress. I still accept Rhys Carpenter’s identification of the lower part of a seated female, Acropolis 1363, as the figure depicted in Carrey’s drawing (Fig. 17).48 I would also accept the statuette from Eleusis in the Athens National Museum, NM 201 (Fig. 18), to which I was able to join the upper torso from the storeroom at Eleusis, as copied at a small scale from the original Parthenon U, with scarcely more deviation from the prototype than we find in the group NM 200, imitated from West B–C, Kekrops and his daughter.49 These, along with a fragment that may conform to the lower part of the reclining woman, W, shown by Carrey,50 are the only figures among the Eleusis pedimental statuettes that are reasonably faithful quotations of Parthenon prototypes. The other pieces vary widely in style and quality of workmanship. Athens NM 202, the woman with a female child in her lap (Fig. 18), which probably occupied the place to the right of NM 201, is portrayed with the high girding and heavy peplos with folds stretched tight over full breasts that is typical of later 4th-century figures of mother goddesses.51 She could represent an Eleusinian heroine with her daughter, as Ruth Lindner has suggested;52 alternatively, she could be Demeter with Kore, if the context were suitable, but that interpretation poses difficulties, since Kore is portrayed in the center of the pediment being carried off by Hades. Assuming that the figure to the right of U on the Parthenon, already missing from the pediment in Carrey’s time, was like statuette NM 202 in being a mother with a child in her lap, regardless of whom the child may have represented in the Eleusinian context, we can easily imagine that the position U* was occupied by Acropolis 888, a fragment of the lower part of a woman seated on a smoothly rounded object with a bell-shaped profile that might best be explained as a hearth or household altar made of clay.53 Olga Palagia objects that Acropolis 1363 is unique among seated female figures in the Parthenon pediments in having a plinth, and Liselotte Weidauer and Ingrid Krauskopf object to the juxtaposition of two related figures with such different seats.54 These seeming anomalies are easily ex48. Carpenter 1932. See below, note 54. 49. See Harrison 1967a, figs. 1–16; for detailed description see pp. 1–9. For detailed description of all the figures assigned to the small Eleusis pediment, see Lindner 1982. She compares the fragments with the Parthenon figures that seem to have inspired them and makes clear their deviations from the prototypes. She also suggests identifications for most of the figures, though some are more tentative than others. These do not directly affect the identifications of the Parthenon figures. For further speculation on Eleusinian identities, see Weidauer and Krauskopf 1993.

50. Lindner (1982, pp. 380–381, no. 14, figs. 53–54) does not exclude the possibility that this figure was modeled on W. In any case, as she points out, it is not a closer copy of W than the Eleusis group modeled on B–C is of its Parthenon prototype. 51. Harrison 1967a, figs. 9–11; Lindner 1982, figs. 49–50. 52. Lindner 1982. 53. In its smoothly rounded shape, lacking the irregularities of natural rock, it resembles the seat of east pediment K; clearest in Palagia 1993, fig. 44. See also Brommer 1963, pls. 50–51; Harrison 1967b, p. 46. 54. Palagia 1993, pp. 50–51;

Weidauer and Krauskopf 1993, p. 10. In questioning the identification of Acropolis 1363 with west pediment U, Palagia writes, “Carrey drew U as a seated woman in peplos and himation covering her legs, leaning heavily to her proper left.” Palagia evidently rejects, though she does not mention it, Carpenter’s explanation that the figure has been knocked out of alignment by the falling cornice block that also damaged the left knee. When one corrects the vertical axis of the displaced figure in the drawing, one becomes aware of the edge of a chiton overfall on the proper right side. U was not drawn wearing a peplos.

eumolpos arrives in eleusis

283

Figure 17. Parthenon west pediment, south corner ( J. Carrey). Photo courtesy F. Brommer

Figure 18. Statuettes from a small pediment at Eleusis. Athens, National Museum 201 (left) in a cast joined with an original fragment in Eleusis; Athens NM 202 in cast. Photo S. Meletzis, courtesy Eleusis Museum

284

e v e ly n b . h a r r i s o n

plained if we retain the earlier interpretations of U as the maiden daughter of Erechtheus whom he sacrificed to save the city of Athens at the time of the war with Eleusis and of U* as Praxithea, the wife of Erechtheus and mother of the sacrificed maiden.55 A tradition that goes back to the Atthidographer Phanodemos in the second half of the 4th century b.c. places the sacrifice of the daughters of Erechtheus “on the Pagos called Hyakinthos.”56 It is unlikely that the term pagos would be applied to the Hill of the Nymphs as a whole, as was suggested by Machteld Mellink and subsequently by Miriam Ervin,57 but it would well suit the rocky extension of that hill at its eastern foot, separated from the main mass of the hill by a low saddle, as the Areopagus is separated from the Acropolis.58 The rocky seat of the maiden U with a low plinthlike extension of the terrain under her feet would mark this locale and help to identify the figure.59 The smoothly rounded seat of Acropolis 888, which is retained in the Eleusis statuette NM 202 although its human figure seems to have been altered, resembles the little that survives of the seat of Parthenon east pediment K, who is best identified as Hestia. It should represent a clay altar or hearth (perhaps emulating some prehistoric form). The mother, U*, should be seen as inside the palace, where Praxithea belonged as Queen.60 If the mother is Praxithea, the child on her lap was probably an infant. The small size of this child would explain why no trace of it appears on the fragment Acropolis 888.61 55. See Palagia 1993, p. 50, and p. 58, note 193, for proposers of these identifications, which she does not accept. 56. Phanodemos, FGrHist 325, F 4; Jacoby, FGrHist IIIB, Suppl. I, pp. 178– 180. Photios, Suda, s.v. pary°noi. For a convenient collection of testimonia on the same myth, see Austin 1968, pp. 22–23. 57. Mellink 1943, pp. 56–59; Ervin 1959, esp. pp. 152–159. 58. Kron 1979, pp. 67–72. 59. Note that the “plinth” is only under the feet of NM 201, as it is of Acropolis 1363. On the sides and back, the carved rock surface extends to the pediment floor. Pittakis, who excavated Acropolis 1363 west of the Erechtheion in 1860, recognized the stylistic similarity of the piece to sculptures of the Parthenon pediments. He proposed that it represented Athena Polias seated on the rock of the Acropolis, which the rock under U resembles in its projections and indentations. The rock of the Hill of the Nymphs is similar to that of the Acropolis, with caves at its base (see Kron 1979, p. 67). For the transcription of Pittakis’s report, see Despinis 1982,

pp. 61–62. Despinis stresses the importance of the fact that Pittakis was the first to associate the statue stylistically and iconographically with the Parthenon pediments. 60. See Harrison 1967a, figs. 9, 16– 17. In the excellent new Acropolis Museum picture book (Trianti 1998, p. 269, pl. 270), Acropolis 888 is described as sitting on a rock, but the front part of the fragment, next to the drapery that falls against the seat, is evenly dressed with a small-toothed implement, whereas the back part, which would never have been visible to spectators, shows some shallow indentation and coarser smoothing. These tool marks are clearer here than in any previous publication that I have seen. The fine claw chisel work is remarkably similar to that on the background of a 5th-century Thespian gravestone in New York that was reused in the Roman period (see Harrison 1990, pp. 173–174, fig. 11:b). This suggests that Acropolis 888, like other west pediment sculptures, had undergone rehabilitation after some damage. The irregular surface of the back part of the seat, which would have been

invisible, will have been roughly smoothed over, while the front part will have been carefully dressed down to remove stains and weathering, still preserving its bell-shaped profile. Palagia (1993, p. 51) remarks on the “pristine surface” but does not attribute it to recutting. She suggests that a figure of Praxithea occupied position U and one of Erechtheus the position U*. She casts doubt on both 1363 and 888 as Parthenon fragments and reads V and W as Ilissos and a Water Nymph (Appendix, p. 61). 61. The traditions about the number of the Erechtheids, their names, and who was sacrificed and who survived, are impossible to reconcile in any consistent way. In the Erechtheus of Euripides it is implied that all the sisters died, whether by sacrifice or by suicide. In the Ion, however, Kreousa says that she escaped because she was a newborn baby in her mother’s arms (Ion, line 280). A newborn would not yet count as a member of the family in the first few precarious days of its life, but the story is still inconsistent with the plot of the Erechtheus. We cannot, therefore, interpret the Parthenon sculptures using the

eumolpos arrives in eleusis

text of either tragedy. It would be easy to restore a newborn suckling in the arms of the woman represented in Acropolis 888, however. The baby would have been very small, and the mother would have held it on her left side, close to her heart. It would have left no trace on the surviving fragment. This child should not be called Kreousa, for, as we have seen, T with her half-grown boy S is almost certainly Kreousa, but it is not at all impossible that the Parthenon image influenced Euripides when he came to write the Ion. 62. Harrison 1967a, p. 9. This suggestion was noted as uncertain by Gratia Berger-Doer, LIMC I, 1981, pp. 469–470, pl. 363, s.v. Aktaios I. 63. Brommer 1963, pp. 31, 55. 64. Palagia (1993, p. 51) comments on “his powerful musculature.” See her pl. 117 and Brommer 1963, pls.127– 128. 65. Harrison 1979. See the Boston cup, pl. 25, for Apollo opening his cloak while both hands remain covered. Eumolpos, West 30 of the Parthenon frieze, has both hands covered while preparing to put on his cloak.

285

The male figure West V (Figs. 15–16) shares with West A, the reclining male in the left corner of the west pediment, the appearance of one emerging out of the water. I have suggested elsewhere that A is Aktaios, who preceded Kekrops in some mythical lists of Attic kings and gave his name to the land of Attica.62 The fall of drapery over the left arm and shoulder of Aktaios as he raises himself up on a low shelf of land suggests the fall of water from a body rising up out of the sea. The cloak of V creates a somewhat similar effect, though the figure is kneeling, rather than reclining as A does. Brommer remarked that the right leg of V, like A’s left, seems to sink into the floor of the pediment.63 The feet of those legs on both are enveloped in drapery. Figure V of the west pediment was drawn by Carrey with its right upper arm raised. Though the lower arm and hand were already missing when Carrey saw the figure, the forearm probably extended forward, to judge from his right shoulder, which leans forward, not back like that of Poseidon in the center of the pediment. If this raised hand held a staff or scepter, its lower end may have been anchored in a cutting near the front of the cornice by V’s right knee. The whole front part of block 6 is now missing, so this detail cannot be verified. The very thick edge of V’s cloak where it is broken away from his back suggests that a wind blew it back from his hand in an enveloping shell of cloth whose color would have framed the figure against the background. This mass of cloth would have supported the forearm and the hand. Kekrops (B) in the opposite corner of the pediment must also have held a scepter, for his right arm too is raised. He was entitled to it as a king and as a judge. Eumolpos in the pediment was a grown man, though beardless, as we know from Carrey’s drawing.64 He too is entitled to the scepter, as founder and ruler of the Mysteries. His scepter would express the agreement whereby Eleusis, though becoming a part of the Athenian state, retained control in all matters pertaining to the Mysteries (Paus. 1.38.3). The broad shoulders, heavily developed thighs, and well-muscled torso of V befit a son of Poseidon grown to manhood. The similar angles of the arms, shoulders, and upper torso of V and those of Poseidon (M) as drawn by Carrey reinforce the impression that the two are related. Figure West V thus declared its identity by quoting three others in the same pediment: A (Aktaios) in the use of falling drapery to mimic the flow of water from a body emerging from the sea, B (Kekrops) in the scepter of a ruler and a judge, and M (Poseidon) in the physical resemblance of two powerful bodies in a similar action. It remains to discuss the cloak of the boy Eumolpos in the Great Eleusinian Relief and the manner of its draping. Like the cloak of west pediment V, it covers his back but is open in front. This may be meant to recall his emergence from the sea-bath of his childhood into the upper world, as his sandals recall a journey over land. One is also reminded in a certain way of the images of Apollo opening his cloak, which I have suggested may refer to his prophetic revelations.65 The Hierophant of the Eleusinian Mysteries is by his very title one who reveals holy things. He must also keep secret those matters that are not to be revealed to the uninitiated. It is not clear in the image itself whether the boy’s gesture of

286

e v e ly n b . h a r r i s o n

grasping the gathered folds of the himation is a prelude to taking it off or wrapping it around him. Perhaps it is both, an expression of the dual powers that the goddesses confer on him.66 There is no general agreement on the date of the Great Eleusinian Relief, but it seems clearly to be somewhat apart from the main line of development of Attic relief sculpture as we know it from the reasonably well-dated series of architectural sculptures in Athens from the Parthenon frieze down to the monument of Dexileos. The stiffness of the poses may well be due, as many have suggested, to a desire to render a sacred image with appropriate solemnity. The very conscientious carving of the finer folds in the drapery and the almost metallic rendering of the hair give a faintly archaistic impression. Tobias Dohrn suggested a date in the last decade of the 5th century B.C.67 Brunilde Ridgway considered this to be too late,68 but the hypothesis of a Boeotian or central Greek sculptor, or an Attic sculptor who had worked in Boeotia, an idea advanced by Schuchhardt, or a “Middle Greek” sculptor as proposed by Schneider69 and favored by Ridgway, would explain its failure to conform to official Attic works of that period, such as the Erechtheion frieze or the western half of the south side of the Nike Parapet. Reliefs from Thespiai show similar characteristics. The boy on the grave stele reused in the Roman period as the monument of Agathokles (Fig. 19)70 has been cited in favor of a Boeotian connection. This is reinforced by a comparison of the lower portion of Persephone’s chiton and the stiff sandal sole of her left foot with the gravestone of a seated woman in New York (Fig. 20).71 Although no provenance is recorded for the latter, the type belongs to a series found in Thespiai that extends down into the 4th century, and the signs of reworking in the Roman period are similar, as is the non-Attic marble. In the Great Eleusinian Relief the unusually low relief in proportion to the size of the panel implies a greater reliance on paint to clarify the actions of the figures and the pattern of the whole than we have in most Attic reliefs. We are reminded of the fine series of engraved and painted Boeotian gravestones of warriors from the late 5th and early 4th centuries.72 Among unquestionably Attic monuments of high quality, the best parallel for the low relief of elements adjacent to the background and the reliance on color to unify the silhouette is the funerary stele of Hegeso, dated by common consent in the last decade of the 5th century or the very beginning of the 4th.73 The right hand of Hegeso, like the right hand of Demeter in the Eleusinian relief, is holding up something that has hardly any thickness; the tips of the thumb and forefinger nearly meet. The base of the thumb and its junction with the wrist are carved in extremely simple forms, as though their basic shapes were visible, but not their details. This recalls the tainia held by Demeter and about to be grasped by Eumolpos. The arguments of Jürgen Thimme in favor of seeing the object offered to Hegeso as a funerary tainia and not an actual necklace are strengthened by this resemblance.74 The stele of Hegeso is surely by an Attic artist. The beautiful echoing curves of the lower drapery of Hegeso and of her maid link the two figures in much the same way as the two daughters of Pelias are linked in the

66. Simon’s direct reference (1998, pp. 380–381) to dressing after the seabath of Halade Mystai is made with the assumption that the boy is Demophon, serving as a representative of all the mystai in the preparatory rite of purification. This interpretation, however, lacks the broader allusion to the whole mystic experience that a youthful Eumolpos would embody in his rebirth from the sea. 67. Dohrn 1957, pp. 40–48. 68. Ridgway 1981, p. 140, note 23: Dohrn “dated the monument far too late.” 69. Schuchhardt 1958, p. 487; Schneider 1973, p. 107. 70. Dohrn 1957, pp. 40, 42–44; Ridgway 1981, p. 140; Harrison 1990, pp. 171–172, fig. 8. 71. Richter 1954, pp. 50–51, no. 75, pl. 61; Harrison 1990, pp. 173–174. 72. Pfuhl 1923, pls. 259–260. 73. Athens, NM 3624. The relief has been reproduced in a number of different photographs, but the one that best illustrates the relation of the relief to the background is in Lullies and Hirmer 1957, pl. 185 (pl. 187 in 1960 ed., pl. 182 in 1979). See also the detail in Knigge 1991, p. 133, fig. 128. The most recent discussion of the date, IG I3 1289 (1994), affirms that almost all except Dohrn and Robertson date the stele to the 5th century (ca. 410–400?). Knigge (1991, p. 134) also considers Hegeso’s stele to be from the late 5th century: “the excavations showed that the late 5th C. B.C. relief of Hegeso was put beside the mid-4th B.C. anthemion stele of Koroibos after the latter had been set up.” 74. Thimme 1964, pp. 18–19.

eumolpos arrives in eleusis

Figure 19. 5th-century b.c. gravestone of a youth from Thespiai. Athens, National Museum 742. Photo Hirmer

Figure 20. Grave stele of a seated woman, probably from Thespiai. New York, Metropolitan Museum of Art, Rogers Fund, 1908 (08.258.42). Courtesy Museum

75. Harrison, forthcoming.

287

famous Three-Figure Relief. I have given arguments elsewhere for dating the latter in the last decade of the 5th century, though these reliefs too have often been dated earlier.75 The Great Eleusinian Relief is also a threefigure relief, and it has in common with the Peliad relief the presence of a commanding, vertically posed, and relatively static figure at the left, faced by two less formally standing figures of different ages. If the composition lacks the rhythmic beauty and implicit drama of the Attic reliefs to which it can be compared, it is because its purpose is wholly different: to embody and confirm existing values, not to warn against their loss. If we accept, and I think we must, that the Eleusinian relief was a votive offering and not a sacred icon to be set up within the inner sanctuary, its sheer size finds parallels among votive reliefs of the last years of the 5th century rather than in earlier works. Best documented, though recon-

288

e v e ly n b . h a r r i s o n

structed from many fragments, is the extraordinary “Monument of Telemachos,” consisting of a votive pinax ca. 0.70 m wide set on a pillar of such a height that the whole measured over 2 m. Inscriptions on the pillar recorded that Telemachos was the first founder of the Asklepieion in Athens.76 The pinax was sculptured on both sides with images celebrating the gods of the sanctuary that Telemachos built.77 Though the foundation must have taken place before 419/8 b.c., the actual monument is placed around the end of the century. Its purpose seems to have been to document Telemachos’s claim as first founder against any other possible claimants. The beautiful relief set up by Xenokrateia in a small sanctuary near Phaleron measures 0.57 m in height and 1.05 m in width (Fig. 21). It was mounted on a poros pillar 2.12 m high. It is exceptional in the size of the pinax, the great number of figures represented, and the quality of the design and execution. The style is purely Attic, suggesting a date in the last decade of the 5th century. Xenokrateia dedicates the relief to Kephisos and the divinities who share the altar with him. It is assumed that she also dedicated the altar.78 These impressive votive offerings displayed the generosity of their donors by depicting a large number of small figures. On the other hand, Pausanias (9.11.6) records that Thrasyboulos and the Athenians with him who destroyed the tyranny of the Thirty dedicated colossal images of Athena and Herakles on a Pentelic marble relief in the Herakleion at Thebes (because they had set out from Thebes on their return to Athens). The stele was the work of Alkamenes.

Figure 21. Votive relief of Xenokrateia. Athens, National Museum 2756. Courtesy Deutsches Archäologisches Institut, Athens

76. For the historical background of the dedication, see Aleshire 1989, pp. 7, 11, 34. 77. For the reliefs on the stele, see Beschi 1967–1968 and 1982. A reconstruction made in the Acropolis Center in Athens with the help of casts from the original fragments in Athens and London and fragments of a contemporary copy now in Padova and Verona has a height of well over 2.5 m. This is only an approximation since the height of the base is not preserved. 78. Athens, NM 2756: Karouzou 1968, p. 57; Guarducci 1974, pp. 57– 66, pls. 8–12; Edwards 1985, pp. 310– 338.

eumolpos arrives in eleusis

79. Plut. Alc. 33–34. For commentary, see Clinton 1974, pp. 15–16. 80. The date of the procession in 407 b.c. can serve only as a terminus post quem for the execution and dedication of the relief. An anonymous Hesperia reviewer has kindly called my attention to the fact that IG I3 386/387 shows that no serious activity was taking place in the sanctuary in 408/7 b.c. If one looks at the history of the years from 407 to the end of the century, it would seem that ca. 400 b.c. would be the earliest likely date for the resumption of normal life at Eleusis. Alkibiades had taken refuge in Phrygia, where he was murdered in 404 b.c. at the instigation of the Thirty Tyrants, in league with Lysander. After the battle in that same year in which Thrasyboulos triumphed over the Thirty and Kritias was killed, the Herald of the mystai, Kleokritos, is said to have recalled in an eloquent plea for reconciliation, “We never did you any harm, but we have shared with you in the most solemn rites and sacrifices and the most splendid festivals” (Xen. Hell. 2.4.20, trans. C. L. Brownson, Loeb). Alkibiades had not slain tyrants as Thrasyboulos did, but through his having himself been slain by the Tyrants’ friends, his memory may have acquired by the turn of the century some of the aura of Harmodios and Aristogeiton, whose images now appeared on the shield of Athena on the Panathenaic amphoras. 81. A letter of Cicero to Atticus in 67 b.c. attests to the Roman interest in the old traditions of Eleusis. He asks Atticus to send him, at the request of Thyillus, a copy of the EÈmolpid«n Pãtria (Letters to Atticus 1.9.2). See Clinton 1974, p. 93. Cicero himself had been initiated, as had Atticus.

289

These examples would suggest for the Great Eleusinian Relief a donor connected with a special event in the history of the Eleusinian cult. Once we accept the date near the end of the 5th century that is favored by the parallels, an event comes readily to mind. From the time when the Spartans seized and fortified Dekeleia in 413/2 until the triumphant return of Alkibiades to Athens in 408/7, the Athenians had not been able to hold their procession by land from Athens to Eleusis at the time of the Mysteries, though the Mysteries continued to be celebrated. Plutarch records that during this time the celebrants were forced to travel by sea and so to omit sacrifices and dances and other rites performed during the march. Thus the festival lost much of its splendor. For the festival of 407 b.c., Alkibiades undertook to provide military protection for a land procession, which, thanks to his intervention, was carried out in undisturbed solemnity. Previously the Demos had voted that the Eumolpidai and the Kerykes should revoke the curses that they had laid upon Alkibiades when he was accused of profaning the Mysteries.79 This triumphant return of the customary land procession is the kind of occasion that could well have been memorialized by a colossal votive relief. It was a kind of refoundation of the cult as its original founder would have wanted it to be. The Great Eleusinian Relief would have been a private dedication, perhaps by the Hierophant himself. The monument was large enough to be impressive but economical in its use of low relief and reliance on painted detail. Like the procession of Alkibiades, it emphasized the solemnity of the ritual rather than the enthusiasm of the participants.80 The fact that the relief was copied in Roman times for display in Italy might suggest that it was valued for its historical associations rather than simply as an appealing work of art from Classical Athens.81

R EF EREN C ES Agora XXXI = M. M. Miles, The City Eleusinion, Princeton 1998. Aleshire, S. B. 1989. The Athenian Asklepieion: The People, Their Dedications, and the Inventories, Amsterdam. Austin, C. 1968. Nova fragmenta Euripidea in papyris reperta (Kleine Texte 187), Berlin. Berger, E., and M. Gisler-Huwiler. 1996. Der Parthenon in Basel: Dokumentation zum Fries, Mainz. Beschi, L. 1967–1968. “Il Monumento di Telemachos, fondatore dell’Asklepieion ateniese,” ASAtene, n.s. 45–46, pp. 381–436. ———. 1982. “Il Rilievo di Telemachos ricompletato,” AAA 15, pp. 31–43. Brommer, F. 1963. Die Skulpturen der Parthenon-Giebel, Mainz.

———. 1977. Der Parthenonfries: Katalog und Untersuchung, Mainz. Carpenter, R. 1932. “New Material for the West Pediment of the Parthenon,” Hesperia 1, pp. 1–30. Clinton, K. 1974. The Sacred Officials of the Eleusinian Mysteries (TAPA 64.3), Philadelphia. ———. 1992. Myth and Cult: The Iconography of the Eleusinian Mysteries (SkrAth 8.11), Stockholm. Collard, C., M. J. Kropp, and K. H. Lee. 1995. Euripides: Selected Fragmentary Plays I, Warminster. Despinis, G. 1967. “Kykladische Grabstelen des 5./4. Jh. v. Chr.,” AntP 7, pp. 77–86. ———. 1982. Paryen≈ neia (BiblioyÆkh t∞w §n ÉAyÆnaiw ÉArxaiologik∞w ÑEtaire¤aw 97), Athens.

290 Deubner, L. 1932. Attische Feste, Berlin (repr. Hildesheim 1969). Dohrn, T. 1957. Attische Plastik vom Tode des Phidias bis zum Wirken der grossen Meister des IV. Jahrhunderts v. Chr., Krefeld. Edwards, C. M. 1985. “Greek Votive Reliefs to Pan and the Nymphs” (diss. New York University). Ervin, M. 1959. “Geraistai Nymphai Genethliai and the Hill of the Nymphs,” Plãtvn 1, pp. 146–159. Figueira, T. 1984. “The Ten Archontes of 579/8 at Athens,” Hesperia 53, pp. 447–473. Guarducci, M. 1974. “L’offerta di Xenokrateia nel santuario di Cefiso al Falero,” in FÒrow: Tribute to Benjamin Dean Meritt, Locust Valley, N.Y., pp. 55–66. Harrison, E. B. 1967a. “U and Her Neighbors in the West Pediment of the Parthenon,” in Essays in the History of Art Presented to Rudolf Wittkower, London, pp. 1–9. ———. 1967b. “Athena and Athens in the East Pediment of the Parthenon,” AJA 71, pp. 27–58. ———. 1979. “Apollo’s Cloak,” in Studies in Classical Art and Archaeology: A Tribute to Peter Heinrich von Blanckenhagen, G. Kopcke and M. B. Moore, eds., Locust Valley, N.Y., pp. 91–98. ———. 1984. “Time in the Parthenon Frieze,” in Parthenon-Kongress Basel, E. Berger, ed., Mainz, pp. 230–234, 416–418. ———. 1990. “Repair, Reuse, and Reworking of Ancient Greek Sculpture,” in Marble: Art Historical and Scientific Perspectives in Ancient Sculpture, M. True and J. Podany, eds., Malibu, pp. 163–184. ———. 1997. “The Glories of the Athenians: Observations on the Program of the Frieze of the Temple of Athena Nike,” in The Interpretation of Architectural Sculpture in Greece and Rome (Studies in the History of Art 49, CASVA Symposium Papers 29), D. Buitron-Oliver, ed. Washington, D.C., pp. 109–125. ———. Forthcoming. “The Aged Pelias in the Erechtheion Frieze and the Meaning of the Three Figure Reliefs,” in Essays in Honor of

e v e ly n b . h a r r i s o n Dietrich von Bothmer, A. Clark and J. Gaunt, eds. Jenkins, I. 1990. “Acquisition and Supply of Casts of the Parthenon Sculptures by the British Museum, 1835–1939,” BSA 85, pp. 89–114, pls. 10–20. ———. 1994. The Parthenon Frieze, London. Kardara, C. 1961. “Glauk«piw: ÑO érxa›ow naÚw ka‹ tÚ y°ma toË zvfÒrou toË Paryen«now,” ArchEph, pp. 61–158.

Karouzou, S. 1968. National Archaeological Museum. Collection of Sculpture: A Catalogue, Athens. Kearns, E. 1989. The Heroes of Attica (BICS Suppl. 57), London. Knigge, U. 1991. The Athenian Kerameikos: History—Monuments—Excavations, Athens. Kron, U. 1976. Die zehn attischen Phylenheroen (AM-BH 5), Berlin. ———. 1979. “Demos, Pnyx, und Nymphenhügel,” AM 94, pp. 49–75. Krug, A. 1968. “Binden in der griechischen Kunst: Untersuchungen zur Typologie, 6.–1. Jahrh. v. Chr.” (diss. Mainz), Hösel. Lawton, C. 1995. Attic Document Reliefs: Art and Politics in Ancient Athens, Oxford. Lindner, R. 1982. “Die Giebelgruppe von Eleusis mit dem Raub der Persephone,” JdI 97, pp. 303–400. Lullies, R., and M. Hirmer. 1957. Greek Sculpture, London. Mellink, M. 1943. “Hyakinthos” (diss. Utrecht). Morrow, K. 1985. Greek Footwear and the Dating of Sculpture, Madison. Neils, J. 1987. The Youthful Deeds of Theseus, Rome. Palagia, O. 1993. The Pediments of the Parthenon (Monumenta graeca et romana 7), Leiden. Parke, H. W. 1977. Festivals of the Athenians, London. Peschlow-Bindokat, A. 1972. “Demeter und Persephone in der attischen Kunst des 6. bis 4. Jahrhunderts,” JdI 87, pp. 60–157. Pfuhl, E. 1923. Malerei und Zeichnung der Griechen, Munich. Raubitschek, I. K., and A. E. Raubitschek. 1982. “The Mission of Triptolemos,” in Studies in Athenian Architecture, Sculpture, and Topography (Hesperia Suppl. 20), Princeton, pp. 109–117.

eumolpos arrives in eleusis Richter, G. M. A. R. 1954. Catalogue of Greek Sculptures in the Metropolitan Museum of Art, Cambridge, Mass. Richter, G. M. A. R., and L. Hall. 1936. Red-Figured Athenian Vases in the Metropolitan Museum of Art, New Haven. Ridgway, B. S. 1981. Fifth Century Styles in Greek Sculpture, Princeton. Robertson, M. 1959. Greek Painting, Geneva. Robertson, M., and A. Frantz. 1975. The Parthenon Frieze, New York. Robertson, N. 1992. Festivals and Legends: The Formation of Greek Cities in the Light of Public Ritual, Toronto. Schneider, L. 1973. “Der Grosse Eleusinische Relief und seine Kopien,” AntP 12, pp. 103–122. Schuchhardt, W.-H. 1958. Review of Richter 1954, in Gnomon 35, pp. 481– 495. Schwarz, G. 1987. Triptolemos: Ikonographie einer Agrar- und Mysteriengottheit (GrazBeitr Suppl. 2), Horn.

Evelyn B. Harrison N e w Yor k U n i v e r s i t y institute of fine arts 1 east 7 8th street ne w york, ne w york 10021

Simon, E. 1983. Festivals of Attica: An Archaeological Commentary, Madison. ———. 1985. Die Götter der Griechen, 3rd ed., Munich. ———. 1998. “Neues zum grossen Relief von Eleusis,” AA, pp. 373–387. Spaeth, B. 1991. “Athenians and Eleusinians in the West Pediment of the Parthenon,” Hesperia 60, pp. 331–362. Thimme, J. 1964. “Die Stele der Hegeso als Zeugnis des attischen Grabkults,” AntK 7, pp. 16–29. Trendall, A. D. 1987. The Red-Figured Vases of Paestum, Rome. ———. 1989. Red Figure Vases of South Italy and Sicily: A Handbook, London. Trianti, I. 1998. TÚ Mouse›o ÉAkropÒlevw, Athens. Vanderpool, E. 1960. “News Letter from Greece,” AJA 64, pp. 265–271,pls. 65–67. von Bothmer, D. 1987. Greek Vase Painting, 2nd ed., New York. Weidauer, L., and I. Krauskopf. 1993. “Urkönige in Athen und Eleusis,” JdI 108, pp. 1–16, pls. 1–4.

291

HESPERIA Pages

69,

2000

293-333

LATE HELLETNI IN

POTTERY

A

NEW

DEPOSIT

THOUGHTS POTTERY

TIC

ATHENS AND FURTHER

ON THE ASSOCIATION AND

SOCIETAL

OF

CHANGE

ABSTRACT

1. I am gratefulto the Ev AOvoL; EtocLpsaoand the A' ApXxLokooyLtx Eqopz?a

HpoYa-coptx6v xat KXaaLxcv

ApXtoo-c-cv for permissionto study and publish the Hellenistic pottery from the South Slope for my doctoral thesis (Vogeikoff 1993); the deposit presentedhere forms part of that material.I also thank Maro Kyrkou, one of the excavatorsof the site, for her support;conservatorNancy Lazaroufor her help at variousstages of my project; Theodora Kakarougafor undertaking the majorityof the drawings;Michalis Tiverios of the Universityof Thessalonikifor advice;and my husband,Tom Brogan,Director of the INSTAP Study Center for East Crete, for help with my English.The submittedversion of this paper benefited greatlyfrom the valuable comments of the two anonymous Hesperia reviewers. 2. Brief preliminaryreportsare publishedin Prakt and Ergon. See Prakt 1955, pp. 36-52; Prakt 1956, pp. 262265; Prakt1957, pp. 23-26; Prakt 1959, pp. 5-7; Ergon1955, pp. 5-14; Ergon 1956, pp. 5-10; Ergon1957, pp. 5-12; Ergon 1959, pp. 157-161.

In addition to presenting an Attic ceramic deposit of the Late Hellenistic period, the authorconnects the increaseof imported pottery and local imitations in assemblagesof the late 2nd and early1st centuries B.C. with contemporarychanges in Athenian society.During this period the emerging political elite ofAthens developedstrongcommercialinterestsand foreign contacts that may have led to changes in dining practices (e.g., the introduction of foreign metal shapes). Emulation of these cosmopolitan practices may have encouraged importation of foreign pottery-presumably skeuomorphs of metal vessels-and providedthe impetus for Attic ceramicimitation of imported metalware. The aim of the present paper is twofold: to contribute to the study of Attic pottery in the Late Hellenistic period by adding one more deposit to the rather small published corpus from this period and, most importantly, to view Attic Hellenistic pottery in its wider historical context. I am particularly concerned with the significance of pottery as an indicator of social and economic change in Late Hellenistic Athens.1

A LATE HELLENISTIC

DEPOSIT

FROM THE

SOUTH SLOPE OF THE ACROPOLIS CONTEXT

The materialunder discussion comes from a cistern located to the south of the eastern flank of the Odeion of Herodes Atticus (Fig. 1). The area south of the Odeion was excavated by the Greek Archaeological Society between 1955 and 1960 under the supervisionof the directorof the Acropolis Museum, I. Miliades. South Slope Cistern 14 (hereafter, C14) was excavated in 1956, and the material found there is currently stored at the mosque next to the Roman Agora.

NATALIA

294

VOGEIKOFF-BROGAN

$2~~~~~~~~~~~~~~~:

ID~~I

IN~~~~~~~~~~~I

i!~~~~~~~~~~~~~~~~~~~~~~~~~~~~t

;-N

0~~~~ V~~~~~~~~~~~~

Figure 1. Plan of excavations on the South Slope of the Acropolis, showing location of Cistern 14 (C14)

LATE

3. Upper diameter0.95 m, lower diameter1.27 m. 4. The dimensionsof the bench are H. 0.50 m; L. 1.00 m; and W. 0.35 m. 5. Prakt 1956, pl. 129:b. For a similarcase, see two red-figurepelikai in Thompson'sGroup B (Thompson 1934, pp. 333-334,427-428, B 1-2, figs. 13:a-b). The red-figurepyxis and the nonceramicobjectsarenot included in the presentarticlebecause they were not assignedto me for publication. 6. For parallels,see Siebert 1976, p. 817, figs. 29-30. 7. The terracottasare discussedin Appendix 2. 8. 1956-NAA-173 (unpublished). 9. 1956-NAF-1 (unpublished). 10. 1956-NAA-102 (unpublished). 11. 1956-NAM-41 (unpublished). 12. 1956-NAF-56 (unpublished). 13. One of the knucklebonesis inscribedAHTO. 14. Sling bullets have also been found in a Sullan deposit on the North Slope of the Acropolis;see Parsons 1943, pp. 240-241, note 136. 15. Grace and SavvatianouPetropoulakou1970, p. 305, E 15. 16. The notationsbelong to Manto Oikonomidou,formerdirectorof the Athens NumismaticMuseum. Inventorynumbers:1956-NAN-7, 1956-NAN-11, 1956-NAN-26,1956NAN-27, 1956-NAN-32.

HELLENISTIC

POTTERY

IN

ATHENS

295

South Slope C14 is a flask-shaped cistern measuring 5.50 m in depth. From the northwest side a tunnel 4.95 m in length extends its capacity.3 Footholes along the walls were probably made for cleaning and maintenance of the cistern and at the bottom a bench was built along the southwest side.4 Finally, a thick coat of hydraulic cement lines the cistern, tunnel, and bench. The cistern was originally excavated in a series of "baskets"(ophu&cs). When the excavators determined that the cistern contained a single fill, these baskets were combined and the undiagnostic pottery was discarded for lack of storage space. Subsequent analysis of the pottery, however, has indicated that C14 contained not only a large deposit of Late Hellenistic material, but also some Roman pottery dating to the first half of the 3rd century A.C. The presence of a small amount of Roman pottery in the deposit is likely to be the result of disturbance that occurred when a Roman pipe was installed along the east side of the cistern at a depth of 0.80 m beneath the surface. C14 also yielded a large red-figure pyxis from the late 5th century B.C., very likely an heirloom.5 In addition to the pottery, the deposit included fragments of tiles, iron objects, bronze bosses,6 a few terracottas, lead weights,8 clay loomweights, stone weights, a Hermaic stele together with its base,9a marble phiale with three spouts,10the thumb of a life-size marble statue,11a small marble altar,'2several knucklebones (some with color),'3 and two sling bullets.14 DATE

OF THE

DEPOSIT

The contents of C14 included sixteen legible stamped amphorahandlesfifteen Knidian examples and one Rhodian. Almost all of the Knidian handles belong to the VI B period (97-88 B.C.)of the "duoviri";the two exceptions belong to the VI A period (108-98 B.C.). The Rhodian handle preservesthe name of the maker,AycO`fooXoc, and is dated to the second half of the 2nd century.15The evidence from the stamped amphora handles is very useful because it provides a terminus post quem (92 B.C.) for the date of the deposit. The stamped amphora handles are discussed below in Appendix 1. Numismatic evidence may serve as a further guide for the date. The excavation notebooks record bronze coins in the fill of the cistern. The coins have not yet been located in the storerooms of the Athens Numismatic Museum; however, the same notebooks record some of the coins with the notation Mithridates." If these coins belong to the special issue of 87/86 B.C., distinguished by the characteristic star between two crescents, the missing coins would provide another terminus post quem (86 B.C.) for the filling of C14. The proposed dating thus falls within the period of Sulla's attack on Athens. According to S. I. Rotroff, the identification of a deposit as Sullan destruction debris is based on two distinctive pieces of evidence: stamped Knidianamphorahandles of period VI B (97-88 B.C.);and Athenian bronze coins depicting a starbetween two crescents on the reverse,minted shortly before Sulla's attack on Athens in 86 B.C. Most of the so-called Sullan

NATALIA

296

VOGEIKOFF-BROGAN

deposits, however, also contain material that can be dated several decades later than Sulla'sattack, indicating that the cleanup operations were a slow process. The deposits that do not contain any material that can be dated later than 75 B.C.are characterized as "pure"Sullan by Rotroff.17 The study of the pottery from C14 suggests that it is a Sullan deposit, though not a pure one since it also yielded a few pieces of pottery that date to the second quarterof the 1st century B.C. (i.e., a Mottled Oliver lid [54] and a skyphos made of Pergamene sigillata [37]), suggesting that either the debris did not find its way to the cistern immediately after Sulla'sattack or the cistern remained open for some time after the material was thrown in. OF THE

DISCUSSION

POTTERY

The pottery has been organized first by fabric, thus separating the fine ware from the coarseware, and then by function, following G. R. Edwards and Rotroff."8I also have taken an unusual step in my presentation by grouping local and imported vessels of the same function. Thus, for example, Knidian two-handled cups are presented together with their Attic imitations, under Drinking Vessels, and lagynoi appeartogether with Attic oinochoai, under Pouring Vessels. This approach is warranted by the considerable number of imports present in the deposit. Rotroff, in justifying her approach of juxtaposing vessels of disparate form but similar function, has argued that this method better reveals the culturalsystem in which these vessels once functioned together."9I believe that this approach is also applicable to vessels of similar function but different provenience. The presentation below includes the entire collection of Hellenistic pottery found in C14. The entries include the inventory number in parentheses, the preserveddimensions,20a short description moving from rim to base, a Munsell reading of the fabric, and a brief description of the quality of the gloss (or slip).To avoid inaccurateidentifications,it has been thought wise not to name the inclusions in the fabrics. TAB LEWARE VESSELS

FOR

FOOD

Plate (1-7) C14 yielded a few fragments of plates in a fabricknown as Eastern Sigillata A ware (1-4). Although there is no consensus about either the place of origin or the date of introduction of this type of pottery, many scholars, most recently K.W. Slane, have strongly arguedfor a Syro-Palestinian origin and a beginning date ca. 150 B.C.21 In fact, Slane's work has resolved the chronological debate. The Eastern Sigillata A plates from C14 are broad with a low base, flat resting surface, flat floor, and an upturned rim. The slip is dull and varies from orange-red (2) to darkred (4), and has been applied by double dipping (note the streak on the floors of 2 and 4). One of the plates (2)

17.AgoraXXIX, pp. 35-36. 18. CorinthVII, iii;AgoraXXIX. For a discussionof the advantagesand disadvantagesof this approach,see AgoraXXIX, pp. 5-7. 19.AgoraXXIX, p. 6. 20. Standardabbreviationsareused for dimensions,e.g., max. p.H. = maximumpreservedheight, Diam. = diameter,and so on. 21. Slane 1997, pp. 272-274 (with referencesto earlierbibliography).

LATE

HELLENISTIC

POTTERY

IN

297

ATHENS

Figure2. Plateswith upturnedrim, EasternSigillataA (1-4). Scale1:3 3

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ........... ^S .. ....... .......... ........... .. ... .

':

T

'

',;

:

::

'

'

:

':'

'

'

... .......;.... . ....

\

..\

~~~~~~~... .

.

' ..........

..

.........

.

...

..

..

. .

carrieson the floor five rouletted (not clearlydefined) circles and a stamped rosette at the center. Plates 1-4 recall those of Tel Anafa type 13c, the majority of which were found in deposits closed ca. 80 B.C.22 Two large black-gloss bases (5-6) with red stacking circles and rouletting on the floor should be restored as plates either with offset or upturned rim. Examples with offset rim have been found in Thompson's Groups D and E, as well as in Sullan debris.23Plates with upturned rim also occur in contexts associated with the Sullan attack.24Both types of plates are of particular interest because their shapes represent a foreign influence in the Attic ceramic repertory of the Hellenistic period.25Both forms are encountered in the black-gloss predecessor (BSP) of Eastern Sigillata A and in Italian Campana B. According to Slane, the broad geographical occurrence of the plates with offset rim suggests that they copied metal prototypes.26 The shape of plate 7 finds no parallels in the Attic repertory. It is probably imported, although the fabric could be characterized as Attic.

PLATE WITH UPTURNED 1

22. Slane 1997, pp. 260-261, 288289, pls. 7-9. 23. Thompson 1934, p. 370, D 1, figs. 55, 116; p. 395, E 22-26, fig. 83. AgoraXXIX, p. 154. 24. AgoraXXIX, p. 155. 25. On the relationshipof these forms to Italianprototypes,see Agora XXIX, pp. 154-155. 26. Slane 1997, p. 283.

(1956-NAK-483) Fig. 2 Eastern Sigillata A. Import.

Max. p.H. 0.025 m. Part of rim and floor. Vertical rim, flat floor. Fabric:7.5YR 8/4. Gloss: dull red. 2

(1956-NAK-486) Fig. 2 Eastern Sigillata A. Import.

Max. p.H. 0.013 m, Diam. base 0.082 m. Base and part of floor. Ring base

RIM with flat resting surface, flat floor. Stamped rosette surrounded by rouletting. Streak from double dipping. Fabric:7.5YR 8/4. Gloss: dull red. 3

(1956-NAK-501) Fig. 2 Eastern Sigillata A. Import.

Max. p.H. 0.035 m. Part of rim and floor. Vertical rim, flat floor. Fabric:7.5YR 8/4. Gloss: dull red.

NATALIA

298

VOGEIKOFF-BROGAN

7

Figure3. Otherplates,blackgloss (5-7). Scale1:3 4

(1956-NAK-482) Fig. 2 Eastern Sigillata A. Import. Max. p.H. 0.04 m.

OTHER 5

PLATES

(1956-NAK-480) Attic.

Fig. 3

Max. p.H. 0.02 m, Diam. base 0.07 m. Base and part of floor. Ring base, flat floor. Rouletting on floor. Fabric:5YR 7/8. Gloss: dull black, red stacking circle and underside. 6

Fullprofile.Slightlyevertedrim, flat floor,ringbasewith flat resting surface.Streakfromdoubledipping. Fabric:lOYR8/4. Gloss:dull red.

(1956-NAK-485) Attic.

Fig. 3

Max. p.H. 0.025 m, Diam. base 0.07 m. Base and part of floor. Ring base

with groovedrestingsurface,flat floor with rouletting. Fabric:5YR 7/6. Gloss:dullblack, redstackingcircleandunderside. 7

(1956-NAK-458)Attic? Fig. 3

Max.dim. 0.066 m, est. Diam. rim 0.22 m. Partof rim and sidewall.Slightly profiledrim.Incisedinscriptionnear rim: POAHC.

Fabric:7.5YR7/6. Gloss:dull black,partial.

Bowl (8-11) There were two bowls with outturned rim and angular profile (8-9) in C14. Fragment 10 probablybelongs either to an echinus or a footed hemispherical bowl. Fine rouletting, a slip of good quality, and a brown stacking circle characterizethis particularpiece, which might not be Attic. The shape, decoration, and quality of the slip on bowl 11 are also alien to Attic pottery of this period. The closest parallel is a bowl from the Athenian Agora, which Rotroff also classifies as an import.27The spiral decoration on the center of the bowl recalls an example from Delos that J.-P. Morel has classified as Italic.28

27. AgoraXXIX, p. 420, no. 1739, fig. 103, pl. 137. 28. Morel 1986, p. 479, no. 125, fig. 20.

LATE

HELLENISTIC

POTTERY

IN

ATHENS

9

8

~

299

10

Figure 4. Bowls with outturned rim (8-9); echinus or footed hemispherical (10); carinated with riled rim (11). Scale 1:3

BOWL WITH OUTTURNED 8

(1956-NAK-489)Attic. Fig. 4

H. 0.045 m. Outturnedrim,carinated sidewall,heavyringbasewith slightly groovedrestingsurface,and nipple underside. Fabric:7.5YR 7/2. Gloss:black metallic,flaked,mottledaroundbase.

RIM 9

(1956-NAK-488)Attic. Fig. 4

H. 0.045 m. Outturnedrim,carinated sidewall,ringbase. Fabric:5YR 7/8. Gloss:dull red, basewithoutslip.

ECHINUS OR FOOTED HEMISPHERICAL

BOWL

(1956-NAK-487) restingsurface.Curvedsidewall. Fig. 4 Rouletting. Import? blackwith brown , .Max,;. 5 b. istacking circle. p,.,H. 0039 Diami m, sGloss: Max m.

10

Tall ringbasewith grooved

CARINATED BOWL WITH RILLED RIM 11

(1956-NAK-438, 528) Import.

Fig. 4

Diam. base0.064 m. Two nonjoiningfragments preserving base,partof body,and rim.Thickenedrimwith pairof grooveson top. Carinatedbodywith VESSELS

FOR

concaveupperpart.Ringbasewith groovedrestingsurface.Rouletting, fourpalmettes,andspiraldecoration at centerof floor. Fabric:7.5YR 7/4. Gloss:dull n s nflo bott,bom. botm

DRINK

Vessels for drink have been subdivided into three categories: containers (West Slope Amphora, 12-15); pouring vessels (Small Oinochoe, 16-19; Round-Mouth Juglet, 20; and Lagynos, 21-26); and drinking vessels (Moldmade Bowl, 27-33; Two-Handled Cup, 34-36; and Skyphos, 37).

300

NATALIA

VOGEIKOFF-BROGAN

Containers: WestSlope Amphora (12-15) Of the two definite West Slope amphorasfound in C14, 12 is earlierthan 13. The latter has features typical of late-2nd-/early-lst-century B.C. amphoras, such as the molded rim, the angle between shoulder and body, the simplified twist of the handles, the extensive use of incision, and the replacement of the plastic masks by spurs.29On the neck, the large ivy leaves find parallels in Rotroff's Large Leaf Group 2, which is dated between 120 and 86 B.C. on the basisofwell-dated depositsfromthe Athenian Agora.30 Amphora 12, on the other hand, recalls a group of amphoras from Agora deposits dating to the middle of the 2nd century B.C.31 Because of the poor state of preservation of fragments 14-15, their shape is not certain.They probablybelong to small amphoras with parallels in the Agora.32

WEST SLOPE AMPHORA 12 (1956-NAK-225)Attic. Fig. 5 H. 0.22 m. pile; rstore romm ofof rim, fullments. fragments. Missingsmallpart belly,andone mask.Outcurvedrim. Tall,wide, flaringneck.Upperhandle attachmentsflankedby knobs.Pairof scrapedgroovesatjunctionof neck with shoulder.Twistedhandleswith reliefmasksat base.Scrapedgroove atjunctionof shoulderwith lower body.Ringbasewith flat resting surface. Fabric:5YR 7/4. Gloss:metallic black,flaked.

stems),andwhite dot rosettes. Slightlytwistedhandleswith vestigial masksat base.Shoulderdecorated with latticeflankedby checkerboard. Fabric:2.5YR 6/6. Gloss:dull back. black. 14 (1956-NAK-511)Attic. Fig. 6 Max.p.H. 0.055 m. Partof neckand shoulder, handle.White thicklines alternate with tan debasedornamenton shoulder.Pairof white verticallines, pairof horizontaltan lines on neck. Fabric:7.5YR7/4. Gloss:dull brownishblack.

13 (1956-NAK-429)Attic. Fig. 6 Max. p.H. 0.095, est. Diam. rim 0.12 m. Rim,neck,shoulder,one handle, and smallpartof body.Moldedrim. Neck decoratedwith invertedegg anddart(incisedegg, painteddart), an ivygarland(leavesin thinnedclay with some addedwhite;incised

15 (1956-NAK-490)Attic. Fig. 6 Max.p.H. 0.05 m. Partof neckandshoulder. Debasedwhite andtan alternating ornamentson shoulder.White dots on baseof neck. Fabric:7.5YR7/4. Gloss:dull black.

29. See Rotroff 1991, nos. 98 from Large Leaf Group 1 (fig. 22, pl. 41), 106 from Large Leaf Group 4 (fig. 29, pl. 44), and 117-119 from the Gaudy Amphora Group (figs. 30-32, pls. 45-

46). Pouring Vessels:Small Oinochoe(16-19), Round-Mouth Juglet (20), Lagynos (21-26) Small oinochoai with trefoil mouth, slender neck, and globular or pearshaped body are popular throughout the 2nd century B.C. Parallels appear in Agora deposits of the earlyand late 2nd centuryB.C.33 and in the Peiraieus Cistern, the closing date of which has recently been placed ca. 140 B.C.34 The absence of late features such as double grooves on the shoulder and a broad foot suggests a date around the middle of the 2nd century B.C. for

30. Rotroff 1991, pp. 89-90, pls. 39-40. 31. AgoraXXIX, p. 122, nos. 427429, 431, 432, pl. 42. 32. AgoraXXIX, p. 123, no. 451, fig. 33, pl. 45. 33. AgoraXXIX, pp. 125-127. 34. AgoraXXIX, pp. 33-34; Metzger 1973, p. 55, no. 63, fig. 4, p1. 10.

LATE

HELLENISTIC

POTTERY

IN

ATHENS

30I

~~~~~~~~~~~~~~~~~~~. ........\

Figure5. West Slopeamphora(12). Scale 1:3

221,

15

14

13- ... ...::..1 ~~~~~~~~~~~.::.:..... ..'.

.'' ,..

-,-..4

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~ ::t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _

....

..

Figure6. West Slope amphoras (13-15).

Scale 1:3

35. AgoraXXIX, no. 490, fig. 36, pl. 48. 36. AgoraXXIX, pp. 132-133, esp. note 61.

the three small oinochoai 16-18. Earlier oinochoai are usually taller and larger. Fragment 19 carries West Slope decoration and strongly recalls a pear-shaped small oinochoe from the Athenian Agora from a Late Hellenistic context.35 Small juglets with a round mouth are quite popular in Attic deposits of the second half of the 2nd century and the first half of the 1st century B.c. The example from C14 (20) features a double handle that splits into spurs on the rim. It is also characterized by a scraped groove below the rim. The treatment of the handle suggests the influence of prototypes in metal.36

302

NATALIA

VOGEIKOFF-BROGAN

A fair number of white-ground lagynoi were found in C14 (21-26). With the exception of 21, the rest areextremelyfragmentary.White-ground lagynoi are commonly understood to have flourished from the second half of the 2nd century to the second quarterof the 1st century B.C., but there is increasing evidence to push the introduction of the type back to the first half of the 2nd century B.c.37The paucity of white-ground lagynoi (three examples) at Corinth, however, suggests that the practice of importing them was just beginning at the time of Corinth's destruction in 146 B.c.38 The center of production of white-ground pottery is unknown, although it is usually assumed to lie somewhere along the west coast of Asia Minor. J. Schafer has suggested Pergamon for the following reasons: the excavations at Pergamon yielded a variety of shapes including lagynoi that are decorated in the white-ground technique; and two small white-ground fragments (unfortunately now lost) of the late 2nd century B.C. were discovered in the debris of a pottery kiln near the Gymnasium at Pergamon, demonstrating that the city did indeed produce white-ground pottery.39 Pitane, the port of Pergamon, has also been suggested as a possible center of production on the basis of the large number of white-ground lagynoi found there.40J.W. Hayes has recently attributeda series of whiteground lagynoi from Paphos to Cypriot workshops on the basis of their fabric.41There are,however,other centers of production because the whiteground lagynoi from Thompson's Group E are not related, in terms of fabric,to either the Pergamene or the Cypriot series, and must come from yet another center.42 Lagynos 21 from South Slope C14 displays an angular profile, with upper and lower body of the same height, and belongs to Shape 2 in Westholm'sclassificationof Cypriot lagynoi.43According to Rotroff, Shape 2 in the Athenian Agora is largely associated with Sullan contexts.4421 also finds parallelsin the Antikythera shipwreck.45In addition to 21, fragments from at least four other lagynoi were found in C14. Three sherds inventoried under the same number (22) probablybelong to the same vessel. Depictions of musical instruments, not uncommon on lagynoi, are preserved in dark slip on the white surface. Four other fragments (23) from another lagynos are decorated with large brownish-black leaves. Finally, fragments from a base (24) and a neck (25) complete the catalogue of white-ground lagynoi from C14. In addition to white-ground lagynoi, the deposit also yielded a redslipped lagynos with West Slope decoration (26). The clay is red, hard, and slightly micaceous, and the slip reddish brown with black spots. Most of the decoration was done by incision, not the hasty and careless incision of the Late Hellenistic Attic workshops, but a fine and delicate technique. A similar black-gloss example comes from a Sullan deposit in the Athenian Agora.46The finesse of the incision on 26 recalls Pergamene West Slope pottery.47

37. AgoraXXIX, pp. 227-228. 38. CorinthVII, iii, p. 50, note 34; Broneer 1947, p. 240, pl. LVIII:12.See also Broneer 1935, pp. 71-72, fig. 16. 39. PFII, p. 110. 40.AvPIX,p. 122. 41. PaphosIII, pp. 18-22, esp. series 4 and 6. 42. PFII, p. 111. 43. Vessbergand Westholm 1956, p. 59. 44. AgoraXXIX, p. 228. 45. Edwards1965, p. 21, no. 10; ArchEph1902, pl. H:8. 46. AgoraXXIX, p. 222, no. 1676, fig. 100, pl. 133. 47. In my dissertation(Vogeikoff 1993), I attributed26 to the Rhodian workshop,but I now believe that it is closer to the Pergameneworkshops.For the use of incision on PergameneWest Slope pottery,see Behr 1988, p. 106; and Patsiada1990, p. 134.

LATE

HELLENISTIC

IN

POTTERY

ATHENS

303

18 .......'..|l .,,,*'!. ?

16;

w

~~~~~~~~~~~17

l19

Figure7. Smalloinochoai(16 19); round-mouthjuglet (20). Scale1:3

SMALL OINOCHOE 16

(1956-NAK-7) Attic.

Fig. 7

H. 0.115 m, Diam. base 0.038 m, max. Diam. 0.075 m. Full profile. Missing handle and part of rim. Trefoil mouth, thin neck, globular body, ring base. Fabric:5YR 6/4. Gloss: dull black, mottled around base.

18

Fig. 7

H. 0.12 m, Diam. base 0.048 m, max. Diam. 0.079 m. Full profile. Missing part of trefoil mouth, and handle. Trefoil mouth, thin neck, round body, ring base. Same shape as 16. Fabric:5YR 6/4. Gloss: dull black, mottled around base.

ROUND-MOUTH 20

Fig. 7

Max. p.H. 0.101 m. Missing mouth and handle. Thin neck, pear-shaped body, and ring base. Fabric:5YR 6/4. Gloss: brownish black, flaky, mottled around base. 19

17 (1956-NAK-3) Attic.

(1956-NAK-6) Attic.

(1956-NAK-516) Attic. Max. p.H. 0.05 m.

Fig. 7

Part of neck and shoulder. Pair of scraped grooves at junction of neck with shoulder. Alternation of white and yellow teardrop lines on neck, white zigzag line on shoulder. Fabric:5YR 7/6. Gloss: dull black.

JUGLET

(1956-NAK-4) Attic.

Fig. 7

Max. p.H. 0.057 m, Diam. rim 0.044 m, Diam. base 0.036 m. Fully preserved. Outturned rim, round body, flat bottom. Thin incised line at junction of rim with body.

Double rolled handle ending at two spurs on rim. black.

304

NATALIA

VOGEI

KOFF-

B ROGAN

LAGYNOS 21 (1956-NAK-224) White-ground.Import.

Fig. 8

Max. p.H.0.10m,iam.bFabric: Max. 0.105 m, Diam. base p.H. 0.12 m. Missingrim,neck,andhandle. Convexshoulder.Carinatedbody. Ringbasewith angularexterior; beveledrestingsurface;slightlyconvex underside.Tracesof reddecorationon shoulder. Fabric:7.5YR6/6. Gloss:offwhite. 22 (1956-NAK-447) White-ground.Import.

Fig. 8

Four nonjoining fragments from

shoulder,decoratedwith largebrown ivyleavesandlight brownstems. Fabric:7.5YR 6/4, micaceous.

Gloss:off-white.

White-ground. Import.

Fig. 8

Max.p.H. 0.073 m. Fragmentpreservingrim,largest partof neck,andbeginningof handle. Rolledrim,tubularneck,and strap handle. Fabric:5YR 7/8. Gloss:off-white. 26 (1956-NAK-436)

Max.p.H. a) 0.06;b) 0.049;c) 0.04; d) 0.04 m.

24 (1956-NAK-551)

25 (1956-NAK-531) White-ground.Import.

Fig. 8

Max.p.H. a) 0.043;b) 0.071;c) 0.049 m. Three nonjoiningfragmentsfrom neckandbody.Body decoratedwith imagesof stringedmusicalinstrumentsin brownandyellowslip. Fabric:7.5YR6/4, micaceous. Gloss:off-white. 23 (1956-NAK-448) White-ground.Import.

Fragmentpreservingsmallpart of baseandbody.Low ringbase. 7.5YR 7/6. Gloss:offwhite.

Fig. 8

Fig. 9

West Slope.Import. Max.p.H. 0.104 m. Threejoiningfragmentspreserving partof neckand shoulder.Neck: necklacewith spearheadpendants attachedto incisedbandby incised zigzagline;two incisedties at ends. Ridgeatjunctionof neckwith body. Convexshoulderdividedinto two registersby pairof grooves.Another pairof groovesatjunctionof shoulder with body.Shoulder,upperregister: incisedrunningspiral;lowerregister: pendant with spearhead pendants

attachedto incisedbandby incised zigzagline;incisedbow with ties and trailing tips below pendants. White

four-leafrosetteto the left of necklace. Fabric:2.5YR 5/6, verylittle

mica. Gloss: reddish brown (2.5YR 4/6), black at places.

Max.p.H. 0.015 m. Drinking Vessels:Moldmade Bowl (27-33), Two-Handled Cup (3436), Skyphos (37) The debris from C14 yielded fragments from seven moldmade bowls, most of which, with two exceptions (one imbricate bowl [27] and one figured [281), belong to the long petal type. The small number of moldmade bowls in the deposit is rather surprising, since they occur in large numbers in Agora deposits associated with the sack of Athens by Sulla.48 Of the two-handled cups, two (34-35) are considered local versions of the so-called Knidian bowl, while the third (36) is actually an import from somewhere in the areaof the Dodecanese and coastal Caria, possibly Knidos. The "Knidian bowl" is found frequently in Attic deposits of the

48. AgoraXXII, p. 36. 49. For Knidianbowls and imitations, as well as bibliography,see Agora XXIX, pp. 119,233-234. 50. PF VI, pl. 7:13, S 3; PF II, p. 68, E 79, E 85; PF VII, pls. 6-7. 51. PFVI, p. 65, S 3. For an opposing view, see PF VII, p. 23; and review of PF VII by S. I. Rotroff in Gnomon68, 1996, pp. 356-361.

LATE

HELLENISTIC

POTTERY

...

IN

ATHENS

305

.

....

<1

2224~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...

...

....._

*

.

'

!{

.i'

'.

!

lt

$|

ti

'

'

'

....... ..

!:................................................................

rFigure8. White-groundlagynoi (21-25).Scale1:3 b

,

-

_-

_

.-

.. ..........

..

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..........

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ..................:.. .:_ .. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... ,

.....

..

.

.

.

*

.. .. . .... ::i..}..

*.

.... :W.::

*;F,:;s,'$igure2

Wes

9.' Re-sipe decoratio.;

ts ... Slope .

' layos

with

Scal 1:3 ....26...................

52. PFVI, pl. 7:13, S 3. 53. Perniceand Winter 1901, pl. 10; see also Gehring 1980. The "HildesheimTreasure"is thought to have been hidden duringthe third

quarterof the 1stcenturyA.C., butthe fact that some of the pieces betray heavy use and some have been reworkedsuggests that the oldest items in the treasureshould be dated in the 1st centuryB.C.

second half of the 2nd century through the first half of the 1st century B.C. and later.49 Fragment 37, an example of Pergamene sigillata, preserves the horizontal thumb-rest of a vertical handle from a distinctive type of skyphos that was popular at Pergamon.50The evidence from the cisterns in the city of Pergamon suggests that production of the shape began some time in the second quarter of the 1st century B.C. and continued until the beginning of the 1st century A.C.x5 The shape is often decorated with applied reliefs.52It is certain that metalware was the source of inspiration for this type of vessel. The Hildesheim Silver Treasure provides us with possible metal prototypes for vessels like 37.5337 representsone of the latest pieces of pottery thrown into C14.

306

NATALIA

VOGEIKOFF-BROGAN

MOLDMADE BOWL Fig. 10

27 (1956-NAK-435) Attic. Imbricate.

H. 0.07 m, est. Diam.rim0.15 m. Full profile.Line filledwith miltosbeneathrim.Bandof single spirals,rowof downwardpalmettes and ovolopatternon rim.Imbricate fems on calyxandwall.Medallion: eight-petalrosettesurroundedby two ridges. Fabric:5YR 7/6. Gloss:brownish black,misfiredat spots. Fig. 10

28 (1956-NAK-517) Attic. Figured.

Max.p.H. 0.031 m. Partof wall.Remainsof tree with pairof doublevoluteson top. Frontlegs of goat(?). Fabric:5YR 7/6. Gloss:dull black. Fig. 10

29 (1956-NAK-428) Attic. Long petal.

Max.p.H. 0.08 m, Diam. rim 0.15 m. Full profile.Scrapedgroove belowrim.Wall:pointedpetals alternatingwithjeweling,stylized floweron top.Medallion:nine-petal rosettesurroundedby two ndges. Rust Rust to bowl. adhenrng adhering Fabric:7.5YR 6/6. Gloss:dull black,brownmedallion.

TWO-HANDLED

30 (1956-NAK-544a) Attic. Long petal.

Fig. 10

Max.p.H. 0.057 m. Smallfragment.Partof rimand wall. Scrapedgroovebelowrim.Wall: horizontaljeweling,contiguouslong petals. Fabric:7.5YR 7/6. Gloss:dull black. 31 (1956-NAK-544b) Attic. Long petal.

Fig. 10

Max.p.H. 0.055 m. Smallfragment.Partof wall and ridgeof medallion.Contiguouslong petalson wall. Fabric:SYR7/6. Gloss:dullblack. 32 (1956-NAK-544d) Attic. Long petal. Max.p.H. 0.04 m.

Fig. 10

Smallfragment.Partof wall. Long petalsseparatedby lines of jeweling. Fabric:2.5YR 6/8. Gloss:red. 33 (1956-NAK-510a+b) Attic. Long petal.

Fig. 10

Max.pjHg 0.06 m. Twojoiningfragmentspreserving partof wall.Contiguouslong petals. Fabric:7.5YR7/6. Gloss:dull brown.

CUP

34 (1956-NAK-538)Attic. Fig. 11 Max.p.H. 0.03 m. One handleandpartof rim. horizontalhandle Straightrim,horlwontal nanaSimilar Stumedht upward. turned Fabric:7.5YR7/2. Gloss:black metallic. metallic.

~27

35 (1956-NAK-507)Attic. Fig. 11 Max.p.H. 0.03a1m. One andle andpartof m. to 34. Fabric:7.5YR7/2. Gloss: bakotie e nie outside,redinside. ~~~~~~black

2293

Figure10. Moldmadebowls(27-33). Scale 1:3

9;l

LATE

HELLENISTIC

POTTERY

36 (1956-NAK-521) Import.

34

36

Fig. 11

Max.p.H. 0.03 m, est. Diam. rim 0.16 m.

IN

ATHENS

307

One handleand smallpartof rim. Straightrim,horizontalrecurved handle. Fabric:5YR 7/4, mica(?).

SKYPHOS

:

:!_ Fiur

_

;

37 (1956-NAK-552) Fig. 12 Pergamenesigillata.Import. ..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... Max.p.L. 0.035 m.

....

Tw-ade

.+

:

cups

.......

Ati

FOOD

.~~~~~~~~~~~~~~~~~~~~~~~~~. .._- ..... . Fiur 12s Skpo

(3)

cl

Horizontalpartof handle (thumb-rest)carryingreliefdecoration of doublevolutes. Fabric:2.5YR 6/8.

:

54. Alexandri1969, pp. 65-67, pl. 34:c; CorinthVII, iii, pp. 119-120, nos. 646-647, pl. 61; and Leonard 1973, ill. 1. For a discussionof types, see Le Roy 1961, pp. 474-479. 55. Leonard 1973, p. 22, ill. 1:2. 56. Didelot, forthcoming. 57. For the closest parallel,see Mayence 1905, p. 392, fig. 48. 58. Vogeikoff 1994, pl. 15:e. 59. Mayence 1905, p. 391, fig. 43. 60. Le Roy 1961, p. 480, fig. 6; p. 489, fig. 16. 61. Zacharou-Loutrari1989, p. 46, fig. 41. 62. Kapitan1980, figs. 1, 3. 63. For a list of findspots,see Le Roy 1961, p. 478, note 1; Scheffer 1981, p. 84, note 227; and Rahmani 1984. 64. For a detailed study of braziers, see Didelot 1990, to be publishedin the Delosseries. 65. CorinthVII, iii, p. 119, note 3; Scheffer1981, p. 87, note 252; and Le Roy 1961, p. 499, fig. 29. 66. Le Roy 1961, pp. 484-488,498500. 67. Fraser1972, p. 287, note 285.

STOVES

PREPARATION AND

HEATING

DEVICES

Brazier (38-41) Of the two nearly intact braziers found in C14, 38 is related to the low type with parallels from Athens and Corinth and in the Bodrum Museum.54Low braziers are sometimes equipped with a projecting tray used perhaps for resting spits.s5 The shape of 38, however,is unparalleled.Moreover, there is a thick layer of added unbaked clay on the firebowl, a feature that also appears in some braziers from Delos. This peculiarity is perhaps an indication that the brazier was also used as an oven.56The unusual shape of the firebowl, which is open on one side, probably facilitated the removal of the lopas that had been placed straight on the layer of clay.The bearded heads on the supports are not well enough defined to be attributed with certainty to one of Mayence's types.57 Brazier 39 belongs to a type that the French call "rechauds 'apied eleve."58The supports on the rim resemble Mayence's type IVD, but those of brazier 39 are slightly different.59The relief head above the inverted heart-shaped vent recalls the Silenes in the lower frieze of a lavish brazier from Delos.60Finally,the heart-shapedvent finds parallelsin braziersfrom Chios61and from Marsala in Sicily.62 Braziersin general have a wide distribution, including the coastal sites of Asia Minor and Syro-Palestine, Egypt, the Cyclades, Sicily, and South Italy. Examples also occur in Athens and Corinth, but braziers are rather rareon mainland Greece.63Some were locally produced, but tall ones such as 39 appearto have been imported from an unknown center.64Delos and Egypt have been suggested, among other places, but without strong evidence. Although fragments of braziersare abundant in both places, molds are completely lacking from Delos, and only two examples are said to have been found in Egypt (now in Athens).65On this evidence, C. Le Roy has suggested Alexandria as a center of manufacture and Delos as a place of distribution.66Conversely, P. M. Fraser excludes an Egyptian origin, in favor of Rhodes.67 Neutron activation analysis of a number of braziers from Israel has shown that almost none of them could have been made in Israel or any of the coastal sites in western and southern Anatolia; the researchersare more

308

NATALIA

VOGEIKOFF-BROGAN

inclined to accept an Aegean origin for their samples.68 0. Didelot has also come to the same conclusion. Because of the volcanic nature of the inclusions in the clay of most of the braziers, she places the area of production somewhere between Kos, Mindos, and Knidos.69 Fragment 40 belongs to a brazier similar to 39, while 41 is unparalleled.

BRAZIER 38

Fig. 13

(1956-NAA-180) Import.

H. H. 0.26 Diam. base 0.23 m, m, m, 0.26 Diam.base023m, 0.074 m' r. L. L(upper)of eachlug 0.074

ofelug lm(pe Almostfullypresered.Uopen shaped firebowl with pierced bottom (one central and three lateral holes). Thick layer of clay covers bottom of

firebowl,leavingonly centralhole

o Sfirebowl clole dopen.id fcfrebowlga"tray." slpee downward cag ac"try th witdecorated supporting dcouuraed lugds. bearded heads. Two upturned handles. Low stand flaring at bottom. Trapeoidalvent Trapezoidal vent att bacside.where backside.ofsad Fabric:2.5YR 5/6, micaceous, crystalline

39

inclusions.Farc25Y56,maeos

Fig. 14

(1956-NAA-179) Import.

H. 0.62 m, Diam. rim (int.) 0.27 m, L. (upper) of each lug 0.082 m.

Restored in many places. Tall, on stand. Down-curving rim. Deep firebowl pierced with five perforations; additional perforations are associated with pot-mending. Three supporting lugs in the shape of bearded heads. Two upturned rope handles. Cylindrical stand flaring out at bottom. Backside of stand carries two vents, partly restored:trapezoidal vent with convex sides (above), inverted heart-shaped vent (below). Rim decorated with ovolo pattern over dentil frieze. Bearded VESSELS

USED

WITH

heads: hair locks arranged symmetrically, parting at center; raised eyebrows; snub nose; beard formed with locks, flanked by the heavy wit four forlcs locks of the long mustache; slightly mouth. Exterior of firebowl ored with triglyph a l edwishbong pa ls.rSand with wned ofdels. thnf

crwe'yrwo

enis

hnfiz

with recessed rectangles (occasionally, triangularopenings replace the rectangles) and ovolo band. Stand with moldmade garlands adicsdrbos ieesha wherectwo rlans meeta bad

two garlandsmeet at backside

Fabc. crystalline inclusions. 40

(1956-NAK-506) Import.

Max.p.H. 0.092 m.

Fragment from tall brazier on stand preserving part of triglyph, dentil and ovolo frieze below supporting lugs. Fabric:2.5YR 5/6, micaceous, 41

(1956-NAK-513)

Fig. 14

Import? Max. p.H. 0.095 m. Single fragment with two garlands in relief and remains of triangularair vents(?). Fabric: lOYR 6/4. Remains of red slip.

HEAT

Fragment 42 belongs to a baking dish and is characterized by a thin flat bottom, a thicker sidewall, and a round rim. The bottom of the dish is rough on the exterior and blackened by soot on the interior only, an indication that the dish was not used over fire, but in the oven. Parallels are to be found in Thompson's Group E and in the Late Hellenistic assemblage from Tel Anafa.71 Macroscopically, the fabric of 42 recalls that of the baking dish in Thompson's Group E.

68. Gunneweg and Perlman1984. 69. Didelot 1997, esp. p. 381. The macroscopicexaminationof the fabric of both 38 and 39 recallsthe fabric descriptionof Didelot's majorcenter of production. 70. According to Didelot, this is an importantmeasurementin distinguishing the differentgenerationsof molds. See Didelot 1997, p. 376, note 1. 71. Thompson 1934, E 139, fig. 105; Berlin 1997, p. 110, no. PW 298, pl.34.

LATE

.

....

.

.

.

.

.

..

.

.

.

....

..

HELLENISTIC

IN

POTTERY

ATHENS

309

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... .

~~~~~~~~~~~~~~~~~~~~~~~~~~. .. . .. ...

.

'

t ............... 't:U

1

e

;

z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ . ...

?

~~~~~~~~~~~~~~~~ *..,_B '.,,; 4,.................................................. i E .. .. _ ...

'.~~~~~~~~~~. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.;-"': ..| .t ..... ... ':t'"'t':""'''''

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,t ..

:.g .S ............... .' ,'',: _.

_

ir'::

.

.

..

.

1

M.

r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ '; .... ....... ,SP1'=a-U.|z.-

..

310

NATALIA

VOGEIKOFF-BROGAN

Figure14. Braziers(39, 41). Scalesca. 1:4 (39) and 1:3 (41)

QF

t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... ...

_

_~~~~~~~~~~~3

= bZ

rRX_~~~~~%w

39

4

The shape has a wide distribution in the Mediterranean, but never appearsin large quantities. Petrological analysis of this type of baking dish from Tel Anafa suggests a production center near a volcanic source. Slane has suggested Pergamon because the distribution of the shape indicates its popularity in the area.Moreover, this type of baking dish is found together with early Pergamene fine wares, and later with 9,andarli ware (the baking dish is characterizedby longevity and continues well into the 2nd century A.C.).72

Although holes on the floor of fragment 43 suggest a possible identification as a strainer,the thickness of its sidewalls indicates a heat-resistant function. It could have served as a cooking stand for a flat-bottomed

72. Slane 1986, p. 312, note 76, no. 91, fig. 15.

LATE

Figure

15.

Baking

dish

(42);

POTTERY

HELLENISTIC

IN

3II

ATHENS

cooking

stand(43);mortar(44). Scale1:3 dish This type of cooking stand was especially popular in Italy. One example was found in the Hellenistic assemblage of Tel Anafa. A. Berlin thinks that its manufacturewas motivated by the acquisition of the orlo bifido pans from Italy.75

COOKING STAND

BAKING DISH 42 (1956-NAK-539) Import.

Fig. 15

Max.p.H. 0.032 m. Partof rim and floor.Slightly thickenedrim,obliquesidewall,flat floor. Fabric:2.5YR 6/8, micaceous with crystallineinclusions. NOT

VESSELS

FOR USE

WITH

43 (1956-NAK-273b) Max.p.H. 0.012 m.

Fig. 15

Twojoiningfragmentsfrom bottomand sidewall.Vertical sidewall,flatbottomwith perforations. Fabric:2.5YR 6/6, micaceous.

HEAT

This category is represented only by the spout of a mortar (44).76

MORTAR 44 (1956-NAK-500) Import.

Fig. 15

(2.5YR6/8) coreandinclusions.

Max.p.L. 0.05 m.

73. Berlin 1993, esp. pp. 40-41. 74. On cooking standsin general, see Scheffer 1981. 75. Berlin 1997, p. 122, PW 340, pls. 37, 82. 76. For the most recent discussion on the function of the mortaria,see Berlin 1997, pp. 123-126. 77. AgoraXXIX, pp. 189-190, nos. 1235-1238, fig. 76, pls. 91-92. 78. Rotroff 1991, p. 89. 79. AgoraXXIX, p. 192, no. 1259, fig. 78, pl. 93.

OTHER COVERED

Spoutwith flukedend. Fabric:lOYR8/4 with red

PURPOSES TOILET

VESSELS

South Slope C14 yielded one pyxis lid, one fragment from a handleless lekanis, and nine examples of so-called reversible lids. The shape of the pyxis lid (45) finds parallelsin pyxides from deposits of the late 2nd/early 1st century B.C.in the Athenian Agora.77Moreover, the decoration can be attributed to Rotroff's Large Leaf Group 2 (ca. 120-86 B.C.), which is characterizedby big heart-shaped ivy leaves with incised stems.78 Fragment 46 belongs to a large handleless lekanis and recalls examples from the Athenian Agora found in Sullan deposits.79According to Rotroff,

3I2

NATALIA

VOGEIKOFF-BROGAN

the large type of lekanis first appearsin the yearsimmediately before Sulla's destruction and continues throughout the 1st century B.C., with the later examples slipped red. The reversiblelids from C14 vary in size, from 0.11 m to 0.35 m in diameter.The shape and decorationof lid 55 recallthat of a Mottled Oliver lid dated immediately after the Sullan destruction.A0 The spearhead necklace and the bead-and-reel borders of 52 are comparableto the decoration of a reversiblelid from the Large Leaf Group 3*81 Lid 54 belongs to the Mottled Oliver type whose production begins sometime in the second quarterof the 1st century B.c.82 Its presence in the deposit is indicative of the closing date of C14. Lids 45 and 50 can be attributed to Rotroff's Large Leaf Group 2.83 53 recalls Rotroff's Large Leaf Group 1, which is distinguished by an undulating white line, big heart-shaped leaves, but no stems.84The Large Leaf Groups are commonly found in deposits dated ca. 120-86 B.C. The decoration of lid 47 has no Attic parallels. On the contrary,the closest parallels are found in Samaria in Palestine.85One of the two examples at Samaria is dated to the 2nd century B.C. on the basis of its context (area Qy, south of the Round Tower).86According to the excavators, the West Slope pottery of the 3rd century B.C. found at Samariawas imported from Attica, but the West Slope pottery of the 2nd century is nonAttic and "must have come from another center than Athens."87A close examination of the West Slope pottery found at Samaria suggests that some of it might have originated in Pergamon.88If we also consider that Pergamon is one of the few places outside Athens where reversiblelids are very popular,89then there is a strong possibility that 47 represents a Pergamene import.

PYXIS LID

80. AgoraXXIX, pp. 193, 446, no.

45 (1956-NAK-433) Attic.

Fig. 16

H. 0.083 m, Diam. 0.18 m. Partof dome and cylinder misng ar()ofsrpe rovs aroundcentralareaof dome.Convex domewith scrapedgrooveat base. Paintedgrooveat outeredge of upper surfaceof flange.Cylinderflaringout

with scrapedgroovenearbottom. Centerof dome decoratedwith teardropasterisk.Dome: ivy leaf garlandwith incisedstemsandwhite dot rosettesinterspersed. Flange decoratedwith white diagonallines. der:ted-and-re deoratin. Fabri: 5Yre74 Glosraown ish black.

81. Rotroff 1991, pp. 89-90, no. 101, fig. 24, pl. 43. 82. AgoraXXIX, pp. 193-194. 83. Rotroff 1991, p. 89. 84. Rotroff 1991, p. 91, no. 89, fig. 20, pl. 37. 85. Samaria-Sebaste III, p. 243, fig.

47:7; Reisneret al. 1924,11, pl. 73:d 1. 86. For the date of the context, see Samaria-Sebaste III, pp. 236, 244. 87. Samaria-Sebaste III, pp. 238, 243. 88. Cf. Samaria-Sebaste III, p. 243,

LEKANIS HANDLELESS

1285, fig. 80, pl. 97.

LEKANIS

46 (1956-NAK-543) Attic. Max.p.H. 0.038 m.

Fig. 17

Partof rim and sidewall.Inverted rimwith externalflange. Fabric:5YR 7/4. Gloss:dullblack.

fig. 47:9-10, with kratersfrom Pergamon(Behr 1988, pp. 155-158, esp. no. 71; and Patsiada1990, nos. 113114); or Samaria-Sebaste III, p. 243, fig. 47:7, with kantharosfrom Pergamon (Behr 1988, p. 118, no. 9, fig. 3). 89. AgoraXXIX, p. 192, esp. note 16.

LATE

HELLENISTIC

POTTERY

IN

ATHENS

3I3

Figure 16. Pyxis lid (45). Scale1:3

. . . . . . ... ..

.. .........j..

.....

..........

...

. :. . . .' :X_1

S ~~~~~~~~~~~~~~~~~~~~~~~~

Figure 17. Handleless lekanis (46); reversible lids (47-49). Scale1:3

{

._

..._

'I ,,E

:

..... ..

.._

47

...

a

..4

46

49

~48 REVERSIBLE

47

47

LID

(1956-NAK-434) Import.

Fig. 17

Max. p.H. 0.085m, Diam. rim 0.25 m. Rim, part of wall and knob missing. Slightly concave wall with two scraped grooves at bottom and top. Rim slightly inverted. Knob: teardrop asterisk with alternating tan and white rays.Wall: bows and garlands;bead-and-reel between scraped grooves, Fabric: sa:- Y 7/6 4YRm c

Gloss: dul brownish black, red at places. 48

(1956-NAK-540a) Attic. Fig. 17

Max. p.H. 0.02 m, est. Diam. rim 0.30 m. Small part of rim and wall preserved. Concave wall with pair of scraped grooves at junction of wall with rim. Wall: spiral painted white. Pseudo bead-and-reel on vertical face of rim. 7/6 d

3I4

49

NATALIA

(1956-NAK-441) Attic. Fig. 17 Max. p.H. 0.055 m, max. p.L.

0.067 m. m. 0.067

r-

*

, wall.Tan d Smlpart par of ofwl.Tndlhn jumping over spirals painted in white. Pair of scraped grooves above dolphins. Fabric:5YR 7/6. Gloss: brownish black. 50

(1956-NAK-442) Attic. Fig. 18

H. 0.14 m, Diam. rim 0.35 m. rim H.r0.14of wallDam. Parts and rim0.35om. restored. Knob with groove on top of raised edge. Slightly concave wall with two scraped grooves at top and bottom. Slightly inverted rim. Knob: teardrop asterisk with alternating tan and white rays.Wall: ivy garland with incised stems on sidewall. Bead-andreel between grooves at top and bottom. Fabric:5YR 7/6. Gloss: metallic, red at places. 51

(1956-NAK-271) Attic. Fig. 18

Max. p.H. 0.077 m, Diam. knob 0.122 m. Rim and parts of wall missing. Knob convex to outside, with scraped groove on top of raised edge; slightly convex center. Convex wall with wheelmade grooves at lower part. Part of wall decorated with incised fishbone pattern. Fabric:2.5YR 5/6. Gloss: dull black, mottled around knob, part of knob without slip, center of floor red. 52

(1956-NAK-548) Attic. Fig. 18 Max. p.H. 0.09 m. Three fragments (two joining)

VESSELS

FOR

PERFUME

AND

VOGEIKOFF-BROGAN

preserving part of wall. Two scraped grooves at lower part of body. Wall: necklace with spearhead pendants originating from incised zigzag line; white dots at tips of pendants.Beadand-reel between grooves. Fabric:5YR 7/6. Gloss: dull brown. 53

(1956-NAK-514) Attic. Fig. 18

Max. dim. 0.04 m. Part of wall. Ivy garland with white ribbon but without leaf stems. ht o oetsitrpre bete ivy leaves. Fabric:5YR 7/6. Gloss: metallic bacoide, red inside. 54

(1956-NAK-481) Attic. Fig. 19 Mottled Oliver.

Max. p.H. 0.051 m, Diam. knob 0.11 m. Knob and upper part of wall preserved. High flaring knob with groove on top of raised edge. Two scraped grooves at upper part of body. Knob: tan star(?).Bead-and-reel border between scraped grooves. Fabric:5YR 7/6. Gloss: dull red. (1956-NAK-439) Attic. Fig.19 5 H. 0.045 m, Diam. 0.11 m. Missing small part of rim and body. Knob with groove on top of raised edge. Slightly convex wall with two scraped grooves at top and bottom. Vertical rim. Knob: cross. Wall: bead-and-reel (white reel has vanished). Pair of white diagonal lines between grooves at lower part. Fabric: 7.5YR 7/4. Gloss: dull brownish black.

OIL

56 belongs to a side-pouring lekythos, a shape that, according to Rotroff, appears in the last quarter of the 2nd century B.C.90The shape finds parallels in Thompson's Group E and the pottery from the Maison des Comediens at Delos.91 The treatment of the rim indicates that the vessel was used for pouring oil. Fragments 57-58 belong to two small thin-walled aryballoi, with gray core and reddish-yellow slip. Their shape strongly recalls that of aryballoi of the so-called blister ware found in Well X in the South Stoa at Corinth and at the Sanctuary of Demeter and Kore on Acrocorinth.92 The blister-

90.AgoraXXIX, pp. 170-171. 91. Thompson 1934, p. 417, E 1333, fig. 103; Grace and SavvatianouPetropoulakou1970, D 183, pl. 147. 92. CorinthVII, iii, pp. 145-147, 228, no. 775, pls. 35, 64; CorinthXVIII, i, pp. 53-54, no. 476, pl. 48.

LATE

POTTERY

HELLENISTIC

IN

ATHENS

3I5

Figure18. Reversiblelids (50-53). Scale1:3

52

50

:54|||

. 51

53

........................

Figure 19. Reversible lids (54-55).

1:3A

Scale

54

~~~~~~~55

3I6

NATALIA

VOGEIKOFF-BROGAN

ware aryballos is a shape with a long history of production at Corinth, starting in the 5th century and continuing until 146 B.C.93 Both 57 and 58 should be treated as Corinthian imports. Their survivalin an Attic deposit of the Late Hellenistic period is noteworthy. Of the unguentaria found in C14, the best preserved (61) displays a tall solid foot, a somewhat bulging body, and a tall neck; the rest, although less well preserved, should also be restored with the same features. Some still preservetheir decoration of white lines around the neck, shoulder,and body. Similar unguentaria occur in contexts of the 2nd and 1st century B.C., in the Kerameikos, Pylos, and Pergamon.94Their place of origin is uncertain. Fragment 63 belongs to a small amphoriskos probably used as an unguentarium.

SIDE-POURING

LEKYTHOS

flaringrim and smallpinchedspout at rightanglewith straphandle.Pair Max. p.H. 0.068 m (with of scrapedgrooveson shoulder. handle handle). 5YR 6/6. Gloss:black, neck,handle handle,and andsmallFabric: small Mouth,neck flaked. partof shoulder.Deep mouthwith

56 (1956-NAK-484)Attic. Fig.20

SQUAT ARYBALLOS 57 (1956-NAK-497) Import.

Fig. 20

Max.p.H. 0.06 m. Rim,neck,handle,andpartof shoulder.Flaringrim,conicalneck, roundshoulder.Double rolledhandle frombelowrim to shoulder.

Fabric:7.5R N5. Gloss: yellowish. 58 (1956-NAK-499)Import. Max.p.H.n0.045m. Rim,neck,andpartof shoulder. Shape,fabric,andgloss sameas 57.

UNGUENTARIUM 59 (1956-NAK-503) Import.

Fig. 21

Max.p.H. 0.117 m. Partof neck,body,andpartof stem. Slightlyswellingbody.Solid stem.Tracesof burningon neck.Two white lines aroundbody. Fabric:graywith redcore. 60 (1956-NAK-493) Import.

61 (1956-NAK-430) Import.

Fig. 21

H. 0.15 m. Missingpartof rim.Tall thin slightlyflaringneck,bulgingbody, solid stem.White lines aroundneck, betweenbodyand neckand around body. Fabric:graywith white inclusions.

Fig. 21

Max.p.H. 0.095 m. Base,stem,andlowerbody.Tall solid stem. Fabric:gray.

62 (1956-NAK-495)

Fig. 21

Import. Max.p.H. 0.08 m. Partof neckandbody.Bulging body. Fabric:gray.

93. CorinthXVIII, i, p. 54, note 177. The fabricof 57 and 58 is also similar to a blister-warefilter vase found in a 2nd-century B.C. deposit at Corinth;see Romano 1994, p. 78, no. 40. IX, pp. 187-188, E 94. Kerameikos 105, E 108, pls. 97:1, 98:1; Kaltsas 1990, p. 18 (AE MifY1735), fig. 21, pl. 23:a;p. 12 (AE MIIY1793), fig. 21, pl. 1746), fig. 25, pl. 14:c;p. 12 (AE MYJT 16:a;and PFI, p. 100, C 1 and C 2, fig. 7. For additionalbibliographyon unguentaria,see Anderson-Stojanovic 1987.

LATE

HELLENISTIC

POTTERY

IN

ATHENS

3I7

Figure20. Side-pouringlekythos (56);squat aryballos (57). Scale 1:3

.7-Ni 56

57

61

60

59

Fgure 21. Unguentaria (59-62); amphoriskos(63). 63

Scale1:3

''

J6~~2

AMPHORISKOS 63 (1956-NAK-535) Attic.

Fig. 21

p.H.0.078 m,Diam.rim

0.03Ma3x. m. Rim, neck,shoulder,partof body, one handle, and beginning

of second.Outcurvedrim,neck flaringout at top andbottom, slightcarinationatjunctionof shoulder with body. Neck slipped

inside. Fabric:SYR 7/6, micaceous.

3I8

NATALIA

VOGEI

KOFF-

B ROGAN

LAMPS

WH EELMADE

Most of the lampsfromC14werewheelmade(64-72).A numberof nozzles (64-67) belong to lampsof Howland'stypes 35A or 37B and are dated within the last quarterof the 2nd centuryandthe firstdecadesof the 1st centuryB.C.95Nozzles 68 and 69 are similarand contemporary, but becauseof theirunslippedexteriortheyareclassifiedundertype35A.96C14 also containedan exampleof a type 35B lampwith five nozzles(70), and one (71) close to type 3997 but without the ridge and concentricgroove around the rim.98 Lamp 72 warrants special mention because of the peculiarity of its shape. It recalls a type of lamp with completely coveredbody and an added filling hole on its top. Such lamps are known in limited numbers from contexts in the Agora and the Kerameikos,dated to the second half of the 4th century B.c.99Two more examples of the same type found in Pella and Priene have been dated to the 3rd century B.C. on stylistic grounds."'072 must be later in date, however, to judge from the incipient flukes of its small nozzle. A date in the 2nd century B.C. is proposed here. Finally, the gray fabric of 72 suggests an origin on the west coast of Asia Minor, perhaps Knidos. MOLDMADE

The two moldmade lamps (73-74) belong to types that are dated slightly earlier than the previous lamps, but still within the second half of the 2nd century B.c. 73 definitely belongs to type 48B since it carries the characteristic broken-barredalpha in high relief at the bottom. R. H. Howland has dated the type to the second half of the 2nd century B.C. on the basis of another example found in an Agora fill (17:5, upper fill) that was deposited toward the end of the 2nd century B.C.10CI. Scheibler suggests an earlierdate (170-130 B.C.),'02 but the appearanceof the type in contexts of the late 2nd/early 1st century B.C. lends support to Howland's chronology. Lamp 74 recalls type 48A. The use of a worn mold probably explains why the characteristicgrooves and ridges are missing from the nozzle of 74. Type 48A is long-lived since it has been found both in earlier (second half of 3rd/first quarterof 2nd century B.C.) and later (third quarterof 2nd century B.C.) deposits.'03Scheibler has distinguished a "prototype,"however, which she dates to the second half of the 3rd century B.C.,104 and a "maintype,"which she dates between 150 and 90 B.C.10574 recalls lamps of the later main type.'06

LAMPS 64 (1956-NAK-541a) Attic.

Fig. 22

Max.p.L. 0.045 m. Fragmentpreservinglong,

pointednozzle.Flukesand oval wick-hole.Type35A or 37B. Fabric:7.5YR 7/6. Gloss:dull brownishblack.

95. AgoralIV, pp. 109-110, pls. 17, 43; pp. 118-120, pls. 18, 19, 44. KerameikosXI, pp. 55-56 (SSL 3), pls. 52-54; pp. 57-59 (KRL), pls. 56-59.

AgoraXXIX, Appendix 3, pp. 503-504 (an adjustmentof the chronologyof Hellenistic lamps by Rotroff). 96.AgoraIV, pp. 110-111, pl. 43; AgoraXXIX, p. 503. 97. AgoraIV, pp. 111-112, esp. no. 478, pl. 43; AgoraXXIX, pp. 503-504. 98. AgoraIV, p. 124, nos. 515-516, pl. 45. 99. AgoraIV, pp. 84-85, type 26B; Kerameikos XI, pp. 43-44, nos. 229232. 100. Drougou 1992, pp. 58-59, no. 108, fig. 8, pl. 26; Priene,p. 449, no. 161, fig. 556. 101. AgoraIV, p. 161, no. 637. For the dating of Agora deposit D 17:5, see AgoraXXIX, p. 443. 102. Kerameikos XI, pp. 66, 192. 103.AgoraIV, pp. 158-160, pl. 48; AgoraXXII, pp. 102-103. 104. Kerameikos XI, nos. 384, 390; AgoraIV, no. 621. 105. Kerameikos XI, pp. 66-68, 70. 106. Kerameikos XI, no. 416.

LATE

65

HELLENISTIC

POTTERY

(1956-NAK-542a) Attic.

Fig. 22

Max.p Fragment preserving long, pointed nozzle and small part of body. Flukes and oval wick-hole. Type 35A or 37B. Fabric: 7.5YR 7/6. Gloss: lustrous brownish black. 66

(1956-NAK-534) Attic.

Fig. 22

Max. p.L. 0.04 m. Fragment preserving preserving long, Fragment pointed nozzle. Flukes and oval wick-hole. Type 35A or 37B. Fabric: 7.5YR 7/6. Gloss: brown.

IN

deeply depressed top, sloping sidewall, ring base. Type 35B. Fabric:7.5YR 7/6, micaceous. Gloss: dull black. 71

(1956-NAK-533) Attic.

Fig. 22

around nozzle. Similar to type 39. Fabric:5YR 6/8. Gloss: black, misfired. 72

(1956-NAA-37) Import.

Fig. 22

Max. p.H. 0.05 m, Diam. base

Max. p.L. 0.043 m. Fragmentpreserglong, Fragent resevin pointed nozzle. Flukes and oval wick-hole. Type 35A or 37B. Fabric: 7.5YR 7/6. Gloss: brown. 68

(1956-NAA-5) Attic.

Max. p.H. 0.027 m, Diam. base 0.034 m, max. p.L. 0.08 m. Intact except for minor chip on rim. Short nozzle with flukes, plain rim with curving sides, slightly convex disc, flat base. Traces of fire

(1956-NAK-537) Attic.

67

3I9

ATHENS

Fig. 22

Max. p.L. 0.07 m. Fragment preserving nozzle and small part of body. Nozzle long, pointed, tubular,and rounded on top. Type 35A'. Gloss: no 7.5YR 7/6. Glos: Fabric: Fabri: 7.5R7/6 gloss on exterior;black gloss on interior, (1956-NAK-532) Fig. 22 69A(t956-NAK-532)cFig..22 Attic.

69

0.04 m, max. p.L. 0.07 m. Short nozzle with incipient flukes, filling hole on edge of shoulder, central tubular "handle,"flat shoulder slightly convex, almost vertical sidewall, flat bottom ending in mastoid projection. Fabric:gray.Gloss: black, flaky. (1956-NAA-4) Attic. Max. p.H. 0.033 m, Diam. base 0.041 m, max. p.L. 0.09 m. Intact except for handle and Itc xetfrhnln upper part of nozzle. Rays on disc, aroundfillinghole, cornucopia i ridge r a rounded sidewall, ring base. Brokenbarred alpha at bottom. Type 48B. Fabric:2YR 6/6. Gloss: dull reddish brown. 73

74 (1956-NAA-285) Max. p.L. 0.06 m. Fig. 22 Attic. Fragment preserving nozzle, Max. p.H. 0.027 m, max. p.L. long, tubular,and blunted. Type 35A'. OO.10M m. Fabric: 7.5YR 7/6. Gloss: no Most gloss on exterior;red gloss onMotflwepatmsig Long nozzle without decoration, ilossonexterior. interior. concave band and ridge around filling 70

(1956-NAA-1) Attic.

Fig. 22

Max. p.H. 0.04 m, Diam. base 0.05 m, max. p.L. 0.13 m. Intact except for tips of nozzles and vertical handle. Five nozzles,

hole, rounded sidewall, rays on top, cornucopia, ring base. Similar to type 48A. Fabric:7.5YR 7/4, micaceous. Gloss: brown, flaky.

NATALIA

320

VOGEIKOFF-BROGAN

Figure22. Wheelmadelamps(64 66, 68-72); moldmadelamp(74). Scale1:3

8

~~~~~~~~~~~~~~~~~~~~~6 66

70

S

0

70

72

71

72

74

LATE

HELLENISTIC

INCENSE

BURNERS

POTTERY

IN

ATHENS

32I

(THYMIATERIA)

C14 yieldeda largepartof the coverof an elaboratethymiaterion(75) and fragmentsof threesimilarlids (76-78). 75 is a bell-shapedcoverwith four verticalhandlesarounda knob in the shape of an egg or pomegranate, A hen or some other type of attachspringingout of acanthusleaves.107 ment shouldbe restoredon top of the knob.Reliefrosettes,masksof Satyrs, and figuresof women decoratethe zone below the handles.Both heart-shapedand roundopeningsservedas outletsfor the smoke of the burningincense.Largethymiateriaarebetterknownin metal.But there arealsoa few elaborateexamplesin terracotta,like the one foundin Byrsa in a contextassociatedwith the destructionof Carthagein 146 B.C.108 In the case of 75, both the friezewith the interspersedreliefmasksand rosettesandthe shapeof the knobrecallCenturipeware.109 Rotroffhasposited a similarconnectionfor an importedpyxis lid from the Athenian AgorabearingreliefdecorationandcolorsthatalsorecallCenturipeware.110 On the basisof clay,Rotroffhas attributedthe Agorapieceto a non-Attic workshop.75, on the other hand, couldbe Attic. The fabriclooks Attic and the Satyr'smaskis a consistentlypopularmotif for the decorationof Attic vesselsof the Hellenisticperiod. It is not clearwhetherfragments76-78 arelids fromthymiateriaor pyxides.All threeexamplescarrysimpleegglikefinials,but the acanthus leavesare absent.The finialsof 77 and 78 are brokenat the top.111 The finialof 76 endsin a small,coiledpeak.76 alsodiffersfromthe othertwo in the positionof the handles,which areset closerto the finialandincline. LIDS 107. The motif of egglike finials springingout of acanthusleaves appearson certainexamplesof Centuripeanware.This polychromatic warewas producedin Centuripein Sicily duringthe 3rd and 2nd centuries B.C., decoratedwith scenes of nuptial character.See Wintermeyer1975, pp. 152-169. 108. Carrie1979. 109. On Centuripeware, see Joly 1980; Wintermeyer1975. 110. AgoraXXIX, p. 232, no. 1565, pl. 123. 111. 77 and 78 also recallexamples of importedwhite-ground pyxis lids from the Athenian Agora;see AgoraXXIX, nos. 1559, 1561, figs. 94-95, pl. 122. Also see Hayes 1992, p. 165, no. 206. Not enough is preservedfrom 77 and 78, however,to excludethe possibilitythat they carriedopenings and belonged to thymiaterialids.

75 (1956-NAK-276)Attic. Fig. 23

76 (1956-NAK-505)Attic. Fig. 24

Max.p.H. 0.25 m, est. Diam. 0.365 m. Partlyrestored.Bell-shapedwith fourhandles.Knobin shapeof pomegranatespringingfrom acanthusleaves;hole for attachment. Two twistedropehandles;two others ridged.Impressionsof fingersat base of handles.Decorationin three zones:1) roundsmallholes and heart-shapedopenings;2) masks, femaleheads,rosettesin relief,round openings;3) lowerzone mostly restored. Fabric:5YR 6/4. Tracesof white slip.

Max. p.H. 0.075 m. Knob,one handle,beginningof second. Fabric:5YR 5/6, micaceouswith inclusions.Tracesof red slip. 77 (1956-NAK-504) Attic. Fig. 24 Max. p.H. 0.065 Knbrin shap of oranate. Fabric:lOYR6/4. Tracesof white? 78 (1956-NAK-496)Attic. Fig. 24 Max. p.H. 0.085 m. Pomegranate knobandpartof lid. Fabric:5YR 7/4. White slip.

322

NATALIA

VOGEIKOFF-BROGAN

...........~~~7

77

Fiue2.Liso5

hmitrao

pyxides (76-78). Scale 1:3~~~~~~~~~~~~~~~...

LATE

HELLENISTIC

UNIDENTIFIED

Figure 23 (opposite page). Thymiaterion lid (75). Scale 1:3

POTTERY

SHAPES

IN

AND

ATHENS

FUNCTIONS

This sectionincludestwo piecesof potterywhose shapecannotbe confidentlyrestored.79 probablybelongsto some kindof a pitcher.Its shoulderis decoratedwith a brown-filledcircleanda brownlooplikeline against a tan background.80 belongs to a large closed shapewith a small ring base.Its decorationis unparalleledandconsistsof ribbonsandalternating wide and narrowpetalsdrawnin a brownoutlineon an off-white background.One of the fragmentsthatconstitute80, however,preservesdecorationverysimilarto that of a sherdfromThompson'sGroupD.11280 is probablythe side-productof a centermakingwhite-groundlagynoi.113 79 (1956-NAK-465) Import.

Fig. 25

Max.p.H. 0.097 m. Partof neckand shoulder. Shoulderdecoratedwith brown-filled circleand a brown,looplikeline against a tanbackground.the a tan against background. Fabric:5YR 7/4, micaceous. 80 (1956-NAK-272) Import.

Fig. 25

Max.p.H. 0.157 m, Diam. base 0.067 m.

AN

112. Thompson 1934, p. 390, D 69. 113. 1 have recentlyseen several hydriaiin East Crete with similarfloral decorationon white backgroundand an emphasison hook-shaped ornaments like those between the petals of 80. 114. Maro Kyrkou(pers.comm.), who is responsiblefor its publication. 115. Imports:1-4, 11, 21-26, 3640,42,44,47,57-62,72,79,80. Imitations:5, 6,34, 35. 116. AgoraXXIX, pp. 222-223 and graph 10.

323

OVERVIEW

OF THE

Fourfragments(threejoining) frombaseandlowersidewall.Convex sidewall,smallringbase.Largeand smallroundedpetalsin brownoutline with centralrib alternating.Light browngarlandinterspersedbetween lowerends of the petals.Hookshapedornamentatjunctionof small andlargepetals.Nonjoiningfragment decoratedwith swagpaintedas a seriesof overlappingdots andtied at the end;light browngarland. Fabric:2.5YR 5/6, micaceous.

DEPOSIT

It is difficultto reconstructthe processby which South Slope C14 came into existence.C14 is the resultof cleanupoperationsof randomdebris followingSulla'sattack.The cisternwas not sealedimmediately,but remainedopenforsomedecades.Most of the materialmusthaveoriginated fromnearbyhouses.Remainsof severalHellenistichouseswereexcavated in the area.The absenceof cookingandplainpotterymight be explained by the fact that the depositwas culledby the excavators,who probably savedonlywhattheythoughtwasdatable.That someof the materialmight havecome froma nearbysanctuaryis alsopossible.The Sanctuaryof the Nymphe,some tens of metersto the south of C14, is certainlythe most likelycandidate.It is not clearyet when the sanctuarywas abandoned,but it yieldedpotterythat can be dated to the 2nd centuryB.c.114Until the latteris fullypublished,however,anyassociationof depositC14 with the Sanctuaryof the Nymphe remainsspeculative.Moreover,the sanctuary wasfound"packed" withfragmentsfromblack-andred-figureloutrophoroi. If some of the potteryin C14 did indeed come from the sanctuary,one wouldhaveexpectedalsoto findfragmentsof loutrophoroiin the deposit, which is not the case. The analysisof the potteryshows that South Slope C14 includeda considerableamountof importedpotteryas well as Attic imitationsof foreignforms.115 The Agoradepositsalsoconfirmthis increasein imports and imitationsduringthe last decadesof the 2nd centuryB.C.116

324

NATALIA

VOGEIKOFF-BROGAN

Figure25. Pitcher(?)(79);large white-groundpitcher(?)with floral decoration(80). Scale 1:3

_ What then is the origin of these imports and what is the source for the imitations by the Athenian potters? As Rotroff has noted, the evidence from the Agora indicates two trends. The first reveals an interest in forms and pottery from eastern Mediterranean sources, including coastal sites in Asia Minor and Syro-Palestine.117 C14, in line with the Agora deposits, has yielded Eastem SigillataA plates,white-ground lagynoi, twohandled Knidian bowls, and other imports from those areas. The second trend in the Agora suggests an interest in Italian forms."8 The plate with offset or upturned rim and the bowl with vertical upper wall find parallelsin Italy.The number of actual imports, however,is small. According to Rotroff, only about twenty pieces of Italian fine pottery have been identified in the Agora from all periods, sixteen of which come from deposits dating after the middle of the 2nd century B.C.; if Italian pottery was not imported to Athens "in such numbers as to have overwhelmed the local tradition,"we must assume that the Athenian potters are very likely to have drawn their inspiration from imported metal prototypes.)"9A certain skeuomorphism is probably true for other types of pottery from the same period. M. Vickers has recendy argued that the different kinds of red-gloss pottery, including Eastern SigillUataA, from around the Mediterranean were skeuomorphs of gold vessels.'20One wonders whether this could be

80

.... ..... ...

-~

~~~~~~~~~~~~~~~~~~~~~~~ ....??

}

117. AgoraXXX, pp. 221-223. 118. Rotroff 1997, p. 99. 119. Rotroff 1997, p. 98. 120. Vickers 1994, p. 248; Vickers and Gill 1994, pp. 179-180.

LATE

HELLENISTIC

POTTERY

IN

ATHENS

325

true for the graywares produced along the coast of Asia Minor. The hard, thin fabric, as well as the form, of the Knidian two-handled bowls evokes metal prototypes.The well-known appliquekraterfromThompson's Group E is certainly a skeuomorph of a silver one.'2' In the following section, I would like to suggest that the appearance of an increased amount of imported pottery and imitations in the Attic deposits of the second half of the 2nd and the early 1st centuries B.C. reflects contemporary changes in Athenian society.122

121. Thompson 1934, pp. 423-426, E 153, fig. 111,pl. III. 122. In recentyears,a numberof scholarshave arguedthat pottery itself is insensitive to sociopolitical changes

while others have counteredthat pottery can, in fact, reflectsocietal change (see Adams 1979. I thank Susan I. Rotrofffor drawingmy attentionto this articlea few years ago. See also Rotroff 1997, esp. pp. 111113). It is not my intention to arguefor or againsteitherview, and then adopt the resultas a generalstatement. Instead,I maintainthat the dynamics of ceramicproduction,consumption, and change arecomplex and fluid, differingfrom one place and period to the next. In this view, the relationship between materialculture-pottery in this case-and society must be discoveredeach time, and not presumed to exist (see Hodder 1987, pp. 1-2). 123. It is misleading,however,to think that Rhodian commercecame to an abruptend afterthe emergenceof Delos as a free port. See Berthold 1984, p. 219: "Rhodes'position did not collapseovernightin the 160's,but ratherweakenedbit by bit in the course of the following decades." 124. Rauh 1993, p. 43. 125. On the foreign residentsof Delos, see Day 1942, pp. 50-119, and Hatzfeld 1919, pp. 31-37, 82-84. 126. On the position of Delos in Mediterraneantradeafter 166 B.C., see Rauh 1993. 127. On foreignersin Athens and Peiraieus,see Baslez 1988, esp. p. 139, note 4; Garland1987, pp. 58-72, 101138. On Romansin Athens and Delos, see Will 1997, pp. 122-123. 128. Agora XXIX, p. 223. 129. Day 1942, pp. 78-79. 130. Roussel 1916, pp. 85-87.

LATE HELLENISTIC POTTERY FROM ATHENS IN HISTORICAL CONTEXT During the 2nd centuryB.C., a series of events broadlyaffected the economy and society of Athens. Shortly before the middle of the century, in 166 B.C., the Romans awardedAthens the island of Delos for the city's help in the struggle against the Macedonians. The Athenians received the island with the stipulation that it would become a free port. This provision is the first clear indication of Rome's intention to promote her own commercial interests in the East. This move caused an immediate decline in the position of Rhodes in tradenetworks.123 The subsequentdestructionsof Corinth and Carthage in 146 B.C. further damaged age-old traditional trade routes. Finally, the formation of the Roman province of Asia in 133 B.C. ensured the development of the east-west axis in Mediterranean trade.'24By the 2nd century B.C., Italy had become the most important purchaserof Hellenistic goods while Delos attained its greatest prosperity as the most important center for transit trade in the Aegean. In a short time, the island was flooded by foreign traders,Romans and Syro-Palestinians among others.125At the same time, the role of Athens as the administrator of Delos allowed it a major role in the island's prosperity and cosmopolitanism.126 Consequently, it is very possible that Athens and Peiraieus attracteda large number of foreign residents at this time, although this alone does not explain the presence of imported pottery and local imitations of foreign forms.127Athens and particularlyPeiraieus had always been inhabited by a considerable number of metics, but the presence of imported pottery in the archaeologicalrecord of the previous centuries is scarce. Other attempts to explain the phenomenon include a possible dissatisfaction of Athenian consumers with their local production or possibly an increase in the number of trading vessels calling at Peiraieus.128With respect to the latter, it is very likely that it was not strictly local business that attractedforeign merchants to Peiraieus,but instead transit tradewith mainland Greece.129Moreover, on the basis of the poor representation of Greeks from the mainland in the epigraphicrecordsfrom Delos, P. Roussel has argued that mainland Greeks preferredAthens to Delos for conducting business with foreign merchants.130 The explanation for the increased number of imports and imitations in Attic deposits of the second half of the 2nd century B.C. probablylies in a combination of factors. Apart from a possible increase in the number of foreigners residing in town and the development of Peiraieus as a transit

326

NATALIA

VOGEIKOFF-BROGAN

center,the phenomenon might also suggest a change in the social behavior of the Athenians. In an article on commodities and the politics of value, A. Appadurai argues that politics of demand make the following distinction between merchants and political elites: "whereas merchants tend to be the social representativesof new commodities and strange tastes, political elites tend to be the custodians of establishedtastes and sumptuarycustoms."'33Could a modified version of this statement be applicable to our case? We are fortunate to have an inscription (IG 112 2336) recording contributions for seven consecutive years-from 103/2 to 97/6 B.c.-for the last Pythais, a religious procession to Delphi. The inscription is very important because it provides a list of the principal officials in Athens during those years.Most interesting, the leading political figures in Athens at the time were not the nine archons, but people who had previously held a Lks?oycrcj Ao, 'Er ''E4rcoptou, seriesof Delian offices (i.e., 'ErcLVs?oTh and 6oCTrtYIV8'oPoaov Tpacaoc -iv 'v AV . According to S.V. Tracy, "the most influential men in Athens ca. 100 [B.C.] were people from families which had extensive commercial interests on Delos.' The picture that emerges from IG 112 2336 suggests an important sociopolitical change in Athens of the late 2nd century B.C. In the years before, commerce in Athens was largely left to the metics and, generally, to people not of the highest social status.'33Although there was never a strong opposition or a class struggle between the trading class and those of a more traditional and conservativekind, it is obvious that merchantswere not part of the political elite.134In the 2nd century B.C., however, the political center of the polis moved from the astu to Peiraieus,andThemistocles' plan to attach the astu to the sea was finally achieved after more than 300 years of resistance.'3' From about the middle of the 2nd century B.C., the city's political elite began to include people with strong commercial interests,136who were used to interacting with foreigners on a daily basis, more likely to accept the unusual, and in a position to influence the lifestyle of the Athenians as a whole. Vickers, in a slightly different context, refers to the "trickle-down"effect, a term used in economics "to describe the way in which elite fashions can influence taste lower down the social scale.'37 If we were to follow Vickers and Gill, the nouveaux riches would have used metalware,not pottery, during their dining or other practices.138The imported pottery and the Attic imitations-most of which copy metalware-were probably meant for a social class of lower status (i.e., petite bourgeoisie?). In addition to the trickle-down effect, it is quite clear that we also have an emulation of the elite by the nonelite,"9 whereby people of a lower social status "attempt to realize their aspirations towards higher status by modifying their behaviour,their dress and the kind of goods they purchase."'140The desire of this lower class to emulate the cosmopolitan practices of the new prevailing elite in Athens would have encouraged importation of foreign pottery and provided Athenian potters with impetus to deviate from tried patterns and adopt new forms inspired by metalware.'41

131. Appadurai1986, p. 33. 132. Tracy 1982, pp. 167-168. 133. Garland1987, p. 88. 134. Von Reden 1995; Mosse 1983, p.58. 135. Plut. Them.19.2-4. 136. Tracy 1982, pp. 167-168. 137. Vickers 1994, p. 235, note 47. 138. Vickers and Gill 1994. 139. Dusinberre 1999, esp. p. 97. 140. Miller 1987, p. 136. 141. On the conservatismof potters, see Rotroff 1997, p. 98.

APPENDIX 1 THE STAMPED AMPHORA HAN DLES

KNIDOS 1

KT 1531. Duoviri corresponding eponym `Ep[icov (KT 1158). See above, 2. Period VI B. Cf. Grace and Savvatianou-Petropoulakou 1970, p. 351, E 187; Grace 1985, p. 35.

(1956-NAK-281) -with [A]v8pov ['A]vap&ct)v IoctToc10

Ei'roXe4IOu Cluster KT 1598. Duoviri corresponding with eponym Eo'ppcxyopoc. Period ~ VI B. Cf. Grace 1985, p. 35.

5

2

Kv8LT[cov] KT 606. Companion type is believed to be E6`3o)XogKapvs6c8o (KT 1579). See below, 6. Period VI B. Cf Grace and

'

(1956-NAK-280) [Kvt]CV

[CK710"Epp. cluster cluster covo~

[HcX] 6Ep toVO

[HloXOvveog]

(1956-NAK-244) 'E, Em HStcGVOO 9VTOU ~~~~~~Ep[io(poculovO

KT 1158. Eponym corresponding with duoviri 'AptcrGoT3ooXog and MeXoXvrocg(KT 1531). See below, 3. Period VI B. Cf. Grace and Savvatianou-Petropoulakou 1970, p. 350, E 183.

Savvatianou-Petropoulakou 1970, p. 346, E 163; Grace 1985, p. 35.

3

KT 1579. The companion type is to be Hltcvog (KT 606). See above, 5. Period VI B. Cf. Grace and Savvatianou-Petropoulakou 1970, p. 346, E 163; Grace 1985, p. 35.

6

KEfpv3oooher

'

(1956-NAK-282) -believed LAv]ap6v cluster To[foOXoo] *,A]pt [Mex6ckvTc]

KT 1531. Duoviri corresponding with eponym `Ep[icov (KT 1158). See above, 2. Period VI B. Cf Grace and Savvatianou-Petropoulakou 1970, p. 351, E 187. 4

(1956-NAK-289) , ['Av]8apc?[v] [A]ptG cluster Tofo6Aot) [M]sX6cvTc~~{~]

(1956-NAK-283)

herm ~~~~~~~~~Ev`Pt3Xov

7

(1956-NAK-245) KxXXtSO ALovoGCoU KvVo&v] amphora

[LI] c76tpoctog HovXoXp?oS KT 1820. Period VI A. Cf. Grace 1985, p. 35; Jefremow 1995, pp. 153, 159, 165.

NATALIA

328

8

(1956-NAK-243) K[ockXXt]dp-86 A[ko]vt)cFoo K[vw8]6covamphora ~IVcTc6]o-vpcx-vo 'I[770](J-Cpoc?go Hl[ovX6cx]p)ll[og]

(1956-NAK-246) 'Ay-qato[Xtg]'Ep,6pcXVTo0

(1956-NAK-276)

KT 1501. The companion type is believed to be 'Av8po[i `vq (KT 650). Period VI B. Cf Grace and Savvatianou-Petropoulakou1970, p. 346, E 166; Grace 1985, p. 35.

FE&7t 'Aytc(4[o]

13

See 7, above. 9

12

VOGEIKOFF-BROGAN

(1956-NAK-275)

8(a[[,u1tvAyox[0o] [x]Xi g] Kv[L]86[ov] (retr.) KT 36. The companion type is believed to be 'ApvoT6&x[io (KT 1515). Period VI B. AL6oyvJTog Cf Grace and SavvatianouPetropoulakou 1970, p. 352, E 193; Grace 1985, p. 35. 10

'E7' 'Apt[j-voxp6c-vso] [Atooo%xoop]8uc Kvw8tov (retr.) bull's head KT 483. The companion type is believed to be MvcaTr-'I6ccCco)v (KT 1611). See below, 15. Period VI B. Cf. Grace and SavvatianouPetropoulakou 1970, p. 342, E 138; Grace 1985, p. 35.

(1956-NAK-278) [FE]T 'AptcrT[o] 864[o] 'Ayoc[0o xXix KvO3tov] (retr.)

14 (1956-NAK-279) 'E71 'A[p]c4[Troxp]6cT [sos] Atocrxotvp0a& KvO3tov (retr.) bull's head KT 483. See 13, above.

See 9, above. 11 (1956-NAK-277) 'Ay-qu6n[oxte] 'Epio6pcxvTo herm? KT 1501? The companion type is believed to be 'Av8povuivq (KT 650). Period VI B. Cf. Grace and Savvatianou-Petropoulakou1970, p. 346, E 166; Grace 1985, p. 35.

15

(1956-NAK-273) 'AvapJv'I6cocov [MDoTr] bull's head with fillet KT 1611. The companion type

is believed to be 'Apvatoxp6ctNg (KT 483). See above, 13-14. Period VI B. Cf. Grace and SavvatianouPetropoulakou 1970, p. 342, E 141; Grace 1985, p. 35.

RHODES 16 (1956-NAK-274) 'Aycxooo cluster? o

PeriodV (146-108 B.C.). Cf. PF XI, p. 79, no. 1; Grace and Savvatianou-Petropoulakou1970, p. 305, E 15.

APPENDIX 2 TE RRACOTTAS

South Slope C14 yielded a very small number of terracottas,which on the basis of fabric and style appear to be contemporary with the pottery: 1) a wreathed female head including part of the bust, 1956-NAUI-37;2) a lower left leg and foot, 1956-NAUI-43;3) the lower part of a seated female figure, 1956-NAUI-50;and 4) a female head broken across the neck, 1956-NAnI78 (Fig. 26).142 In addition, C14 yielded a small incense burner, 1956-NAII-29, also known as an arula, with its four sides decorated in relief: a female figure crowning a trophy; Apollo and Leto; Poseidon and Amymone; and Dionysos with Ariadne and a Satyr (Fig. 26).143This same imagery appears on moldmade bowls from Athens, Corinth, and Argos.144Although these arulaearewidely distributed in the Mediterranean, neither their origin nor function is known. Some scholars have suggested a South Italian origin because of the large number of arulaefound inTaras.145 Others have suggested an Attic origin on the basis of the popularity of the imagery of the arulae on Attic moldmade bowls.146 Our arula is made of a soft, dark gray fabric that by itself suggests a non-Attic origin. It is also the product of a fresh mold. In Athens, these arulaeusually appearin contexts of the 3rd and early 2nd century B.C.147 If the arula from South Slope C14 is not another heirloom, its occurrencein such a late context might suggest a remarkablelongevity for the type.148

142. 1) 1956-NAII-37, p.H. 0.10 m; 2) 1956-NAII-43, p.H. 0.03 m; 3) 1956-NAII-50, p.H. 0.07; 4) 1956NAII-78,p.H. 0.05 m. 143. H. 0.10 m; andW. 0.07 m. 144. AgoraXXII, pp. 20-21; Corinth VII, iii, nos. 807, 810, pl. 37; and Siebert 1978, p. 240. 145. Wuilleumier 1929; Siebert 1978,p.240. 146. Thompson 1962, p. 259; Agora

XXII, pp. 20-21; and VafopoulouRichardson1982. 147. On the chronologyof the arulae, see Massa 1992, pp. 70-78. 148. Vafopoulou-Richardson(1982, p. 229) refersto an unpublishedarula in the Boston Museum of Fine Arts that in the label of the illustrationis

datedto the 1stcenturyA.C.; however, Vafopoulou-Richardsondoes not discuss its late date.

330

NATALIA

....... ....I- . .. ..

.. ..

... ...

VOGEIKOFF-BROGAN

...... .

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

. . ... '. ''.' . ;:.' 1,ii_i; . . ..

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ...... .,22& . *;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .i,

..

-.

..

.

..

^

...

'

.'

' ..

..

r:..... ..

.

-:-

..

..~

~~~~ ~~~~ ~~~~ ~~~~ ~~~~ ~~~~~~~~~~~ .

.

.

:

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. :..:. .;........ ....

t!A

'

i

=

:f11

::

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

dl1ati ,'5 iS ~ ~~

.~

-0

~

~~~~~~~~~~15-A-7

~~~~~~~~~~~~~~~~~~

s~~~ 26 ercta ~~~~~~~~~~igur uprrw 43

...~~~~~~~~~~~~~~nes bu er_ or

an

aua(orves

8

n

LATE

HELLENISTIC

POTTERY

IN

ATHENS

33I

REFERENCES Adams,W. Y. 1979. "On the Argument from Ceramicsto History: A Challenge Based on Evidence from Medieval Nubia,"CurrAnthr20, pp.727-744. Agora= TheAthenianAgora,Princeton. IV = R. H. Howland, Greek Lampsand TheirSurvivals,1958. V = H. S. Robinson,Potteryof theRomanPeriod:Chronology, 1959. XXII = S. I. Rotroff,Hellenistic Pottery:Athenianand Imported MoldmadeBowls, 1982. XXIX = S. I. Rotroff,Hellenistic Pottery:Athenianand Imported TableWareand Related Wheelmade Material,1997. Alexandri,0. 1969. "F Epoproc Kc.xocx6v Apx(xLor-ovw AOivw'v. 6,"ArchDelt23, 086; Kotpooox6coxn 1968, B' [1969], pp. 33-109. Anderson-Stojanovic',V. R. 1987. "The Chronology and Function of CeramicUnguentaria,"AJA91, pp.105-122. Appadurai,A. 1986. "Introduction: Commodities and the Politics of Value,"in TheSocialLifeof Things: in CulturalPerspective, Commodities A. Appadurai,ed., Cambridge,pp. 3-63. AvP IX = E. Boehringerand F. Krauss, Das Temenosfuir denHerrscherkult: PrinzessinnenPalais (Altertuimer von PergamonIX), Berlin 1937. Baslez, M.-F. 1988. "Les communautes d'orientaux,"in L'etrangerdansle mondegrec,R. Lonis, ed., Nancy,pp. 139-158. Behr,D. 1988. "Neue Ergebnissezur pergamenischenWestabhangkeramik,"IstMitt 38, pp. 97-178. Berlin,A. 1993. "ItalianCooking Vessels and Cuisine fromTel Anafa,"IEJ 43, pp. 35-44. . 1997. "The Plain Wares,"in TelAnafa11.1:TheHellenisticand RomanPottery(JRASuppl. 10), A. Berlin, K. W. Slane, and S. C. Herbert,eds., Ann Arbor,pp. ix244. Berthold, R. M. 1984. Rhodesin the HellenisticAge,Ithaca. Broneer,0. 1935. "Excavationsin Corinth, 1934,"AJA39, pp. 53-75.

11947. "Investigationsat Corinth, 1946-1947," Hesperia 16, pp.233-247. Carrie,J.-M. 1979. "Un brale-parfiums trouvea Carthage,"in Byrsa. Mission arch6ologiquefranfaisea Carthage I (CEFR 41), S. Lancel and P. Gros,

eds., Rome, pp. 311-331. Corinth VII, iii = G. R. Edwards, Corinthian Hellenistic Pottery,

Princeton 1975. Corinth XVIII, i = E. G. Pemberton, The Sanctuary of Demeter and Kore. The GreekPottery, Princeton 1989. Day,J. 1942. An Economic History of Athens under Roman Domination,

New York. Didelot, 0. 1990. "Recherchessur les rechaudshellenistiquesde Delos" (diss. Universitede StrasbourgII). .1997. "Rechaudsd'epoque hell6nistique:La diffusion des signatures,"in Le moulage en terrecuite dans l'antiquitM Creation et production derivJe,fabrication et difusion. Actes du XVIIIe Colloquedu Centre de RecherchesArch&ologiquesLille 11 (7-8 d&cembre 1995), A.

Muller,ed., Villeneuve d'Ascq,pp. 376-395.

Forthcoming."Rechaudsa foyer ouvert de la Maison des Sceauxde Delos," in E' Er1a7uovixY Zovavoc)r70 yca ERUvtorxtC ?Kpayitx, Xavt,

Tnv

6-13 ArptcA'oo1997. Drougou, S. 1992. AvaucxycprHocx, 1957-1964: -Ot 7rvAtvotA?vwvot,

Athens. Dusinberre,E. R. M. 1999. "Satrapal Sardis:Achaemenid Bowls in an Achaemenid Capital,"AJA103, pp. 73-102. Edwards,G. R. 1965. "The Hellenistic Pottery,"in G. D. Weinberget al., TheAntikythera Shi7pwreckReconsidered (TAPS, n.s. 55, part 3), Philadel-

phia, pp. 18-27. Fraser,P. M. 1972. PtolemaicAlexandria, Oxford. Garland,R. 1987. The Piraeusfrom the Fifth to the First Century B.C., Ithaca. Gehrig, U. 1980. Hildesheimer Silberschatz aus dem Antikenmuseum (Bilderheft der Staatlichen Museen

332

NATALIA

PreussischerKulturbesitz 4), 2nd ed., Berlin.

Grace,V. R. 1985. "The Middle Stoa Dated by Amphora Stamps," Hesperia 54, pp. 1-54. Grace,V. R., and M. SavvatianouPetropoulakou.1970. "Lestimbres amphoriquesgrecs,"in Exploration arch6ologiquede De'los27: L'ilot de la Maison des Come'diens,P. Bruneauet al., eds., Paris,pp. 277-382. Gunneweg,J., and I. Perlman.1984. "HellenisticBraziersfrom Israel: Results of PotteryAnalysis,"IEJ 34, pp.232-238. Hatzfeld,J. 1919. Les trafiquants italiens dans l'Orient hellenique (BEFAR 115), Paris. Hayes,J. W. 1992. Greekand GreekStyle Painted and Plain Pottery in the Royal Ontario Museum, Toronto.

Hodder,I. 1987. "The Contributionof the Long Term,"in Archaeology as Long-Term History (New Directions in Archaeology),I. Hodder, ed., Cambridge,pp. 1-8. Hoff and Rotroff = M. C. Hoff and S. I. Rotroff,eds., The Romanization of Athens: Proceedingsof an International ConferenceHeld at Lincoln, Nebraska (April1996) (Oxbow

Monograph 94), Oxford 1997. Jefremow,N. 1995. Die Amp horen-

VOGEIKOFF-BROGAN

Le Roy,C. 1961. "Rechaudsdeliens," BCH 85, pp. 474-500. Massa, M. 1992. La ceramicaellenistica con decorazionea rilievo della Bottega di Efestia (Monografiedella Scuola

archeologicadi Atene e delle missioni italianein Oriente 5), Rome. Mayence,F. 1905. "Fouillesde Delos: Les rechaudsen terre-cuite,"BCH 29, pp.373-404. Metzger, I. R. 1973. "Piraeus-Zisterne," ArchDelt 26,1971, A' [1973], pp. 41-94. Miller D. 1987. Material Culture and Mass Consumption, Oxford. Morel,J.-P. 1986. "Ceramiquesa vernis noir d'Italietrouveesa Delos," BCH 110, pp.461-493. Mosse, C. 1983. "The 'World of the Emporium' in the PrivateSpeeches of Demosthenes,"in Trade in the Ancient Economy, P. Garnsey,K. Hopkins, and C. R. Whittaker,eds., Berkeley,pp. 53-63. Paphos III = J. W. Hayes, The Hellenistic and Roman Pottery (Paphos III), Nicosia 1991. Parsons,A. W. 1943. "Klepsydraand the Paved Court of the Pythion," Hesperia 12, pp. 191-267.

stempel des hellenistischenKnidos

Patsiada, V. 1990. "K?p%MXN IcOV 6IV7Ou vrIg 'Atu-ctxg KXaT6o;'omi P6ao,"ArchDelt 38,1983, A' [1990],

(Quellen und Forschungenzur antikenWelt 19), Munich. Joly,E. 1980. "Teorievecchie e nuove sulla ceramicapolicromadi

pp.105-210. Pernice,E., and F. Winter. 1901. Der Hildesheimer Silberfund, Berlin. PF = PergamenischeForschungen, Berlin.

Centuripe," in 4'Mc5X&ip v: Miscellanea di Studi classici in onore di Eugenio Manni IV, Rome, pp.

1241-1254. Kaltsas,N. 1990. "A0o -ocxxV6 v?xpo01ct?c-rg

C

flODXou,"ArchDelt

38, 1983, A' [19901,pp. 1-77. Kapitan,G. 1980. "ThreeTerracotta Braziersfrom the Sea off Sicily," IJNA 9, pp. 127-131. Kerameikos = Kerameikos:Ergebnisse der Ausgrabungen, Berlin. IX = U. Knigge,Der Siidhzugel,

1976. XI = I. Scheibler,Griechische Lampen, 1976. Leonard,R. M. 1973. "Braziersin the Bodrum Museum,"4AJA 77, pp. 2125.

I = Pergamon: Gesammelte Aufsatze, 1972. II = J. Schafer,Hellenistische Keramik aus Pergamon, 1968.

VI = C. Meyer-Schlichtmann, Die pergamenische Sigillata aus der Stadtgrabung von Pergamon:Mitte 2. Jh. v. Chr.-Mitte 2. Jh. n. Chr., 1988. VII = G. Huibner,Die Applikenkeramik von Pergamon: Eine Bildespracheim Dienst des Herrscherkultes,1993. XI = C. BorkerandJ. Burow,Die hellenistischenAmphorenstempelaus Pergamon, 1998. Priene = T. Wiegand and H. Schrader, eds., Priene. Ergebnisse der Ausgrabungen und Untersuchungenin den Jahren 1895-1898, Berlin 1904.

Rahmani,L. Y. 1984. "Hellenistic BrazierFragmentsfrom Israel,"IEJ 34, pp.224-231. Rauh,N. K. 1993. TheSacredBondsof Commerce. Religion,Economy,and TradeSocietyat HellenisticRoman Delos, 166-87B.C., Amsterdam. Reisner,G. A., C. S. Fisher,and D. G. Lyon.1924.HarvardExcavationsat Samaria,1908-1910, 2 vols., Cambridge,Mass. Romano,I. B. 1994. "AHellenistic Deposit from Corinth:Evidence for Interim Period Activity,"Hesperia 63, pp.57-104. Rotroff,S. I. 1991. "AtticWest Slope Vase Painting,"Hesperia 60, pp. 59102. .1997. "FromGreek to Roman in Athenian Ceramics,"in Hoff and Rotroff,pp. 97-116. Roussel,P. 1916. De'los,colonie athenienne(BEFAR111), Paris. Samaria-Sebaste III = J. W. Crowfoot, G. M. Crowfoot, and K. M. Kenyon,The ObjectsfromSamaria (Samaria-Sebaste: Reportsof the Work of theJoint Expeditionin 1931-1933 and of theBritishExpeditionin 1935 III), London 1957. Scheffer,C. 1981.Cookingand Cooking Stands in Italy, 1400-400

B.C.

(Acquarossa 11.1),Stockholm. Siebert,G. 1976. "Rapportssur les travauxde l'Ecole francaiseen 1975: Delos," BCH 100, pp. 799828. .1978. Recherches surlesateliers de bolsa reliefsdu Peloponnese a l'eoque hellenistique(BEFAR233), Athens. Slane, K. W. 1986. "TwoDeposits from the Early Roman CellarBuilding, Corinth,"Hesperia 55, pp. 271-318. 1997. "The Fine Wares,"in Tel Anafa II.1:TheHellenisticand RomanPottery(JRASuppl.10),A. Berlin, K. W. Slane, and S. C. Herbert,eds., Ann Arbor,pp. 247406. Thompson, D. B. 1962, "Three Centuriesof Hellenistic Terracottas, IIC: The SatyrCistern,"Hesperia 31, pp.244-262. Thompson, H. A. 1934. "Two Centuriesof Hellenistic Pottery," Hesperia 3, pp. 311-480.

LATE

HELLENISTIC

Tracy,S. V. 1982. I G. 1122336: Contributors of First Fruitsfor the Pythass (Beitragezur klassischen

Philologie 139), Meisenheim am Glan. Vafopoulou-Richardson,C. E. 1982. "AnUnpublishedArula in the Ashmolean Museum:A Minor Contributionto Hellenistic Chronology,"JHS102, pp. 229-232. Vessberg,O., and A. Westholm. 1956.

1994.

E'

POTTERY

vvjrntx

I?pxSLx

ocm -! No6Ltoc KXL-c r-vr

in F E7rtuv7yovtx'c Axpo67okqg," ZoVd

aT ytx -7v EArj7vtuvx'

Kepaczix. XpovoAoyrivca Z0vo%Ax-EpyauC5Tpxa,

OeuuaAov(oxc, 24-27 Zoeu3pplOo

1991, Athens, pp. 39-45. Von Reden, S. 1995. "The Piraeus-A World Apart,"GaR42, pp. 2437. The Hellenistic and Roman Periods in Will, E. L. 1997. "ShippingAmphoras Cyprus (SwCyprusExp IV.3), as Indicatorsof Economic Stockholm. Romanizationin Athens,"in Hoff Vickers,M. 1994. "Nabataea,India, and Rotroff,pp. 117-133. Gaul, and Carthage:Reflectionson Wintermeyer,U. 1975. "Die polyHellenistic and Roman Gold chrome Reliefkeramikaus Vesselsand Red-Gloss Pottery,"4AJA Centuripe,"JdI90, pp. 136-241. 98,pp.231-248. Wuilleumier,P. 1929. "Bruile-parfums en terrecuite,"MEFR 46, pp. 68Vickers,M., and D. Gill. 1994. Artful Crafts:Ancient GreekSilverware and 72. Pottery, Oxford. Zacharou-Loutrari,A. 1989. "Ancient Chios,"in Chios:HistoryandArt, A. Vogeikoff,N. 1993. "HellenisticPottery from the South Slope of the Zacharou-Loutrari,V. Pena, andT. Athenian Akropolis"(diss. Bryn T. Mandala,eds., Chios, pp. 6-49. Mawr College).

Natalia Vogeikoff-Brogan AMERICAN

54

SOUIDIAS

SCHOOL STREET

IO6 -76 ATHENS GREECE [email protected]

OF CLASSICAL

STUDIES

AT ATHENS

IN

ATHENS

333

HESPERIA

69,

2000

A

N[W

G

R[[KEI<

LATIN

Pages 335-342

FRO/v

I N

S

AND

C RI

PTlON

CORINTH

ABSTRACT Presentedhere is the editioprinceps of a recently discoveredinscription from excavationsat Corinth conductedby the American School of ClassicalStudies at Athens. The text preserves four fragmentarylines inscribed in Latin and two in Greek. Reference is made to a sacerdosof Proserpinaas well as a stoa and temple of Pluto. The possibility that these monuments were originally located either at Corinth or in the Isthmian Sanctuaryof Poseidon is explored here. Lacking any means to date the inscription with precision, I suggest a tentative date in the 2nd centuryA.C. This brief note presents the text of an important, although fragmentary, inscription recently discoveredin the excavationsat Corinth by the American School of Classical Studies.' The text raises a number of interesting topographical and historical questions concerning both Corinth and Isthmia. The fragmentaryLatin and Greek inscription is on a stele discovered on May 6, 1997, in debris of Frankish date immediately east of unit 6, southeast of Temple E. The stele is of white marble, with the right, top, and bottom edges preserved.The stone is broken on the left and on the top right corner.The top of the stele is smoothly dressed; the bottom is sawn. The back is roughly finished on the left-hand side, perhaps quarryfaced. A slightly recessed area along the right side of the back surface is executed with diagonal point strokes. A hook clamp cutting, which may have been for a secondary purpose, is preserved on the bottom of the right side of the stele. The inscribed face is lightly polished. The deeply cut 1. I would like to expressmy deepest gratitudeto Charles K. Williams II, Director Emeritusof the Corinth Excavations,both for permissionto publish the inscriptionand for his encouragementand suggestionsduring the preparationof this article.Thanks are also due to Nancy Bookidis, Assistant Director,for her assistancein

the Corinth Museum. Ronald S. Stroud graciouslyread draftsat both earlyand late stages;his comments and advice have been invaluable.I would also like to thank Elizabeth R. Gebhardfor sharingwith me her knowledgeof the Isthmian sanctuary.David R. Jordan reada draft and providedmany useful comments.Both he and Daniel J.

Geagan kindly allowed me to examine the Isthmian inscriptionsassignedto them. The comments of the anonymous Hesperia readerssaved me from numerousembarrassmentsand helped improvethe article.The photographfor Figure 1 was taken by Ino Ioannidou and Lenio Bartzioti.

336

MICHAEL

Z

D.

DIXON

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...

~

|

E

Figure1. Latin and Greekinscrp on stele found at Corinth.Scale1:2. Photo I. oannidouand L. Bartzioti.

A"

~ ~ ~

.....~~~~~~ ..

~

..

...

letters close to the right edge are well preservedwhile those on the left, in lines 5-6, are heavily worn, possibly from later reuse. Traces of cement, presumably also the result of secondary use, are preserved along the bottom right corner and within the break in the upper right corner. Corinth inv. 1 .97-1

Fig. 1

H.0.272mW 0223m Th m LH 004005 m (lines 13), 0.02 m (line 4), 0.015-0.022 m (lines 5-6), interlinear spacing 0.008 m (lines 1-4); 0.005 m (lines 5-6) saec. IIp.

1

4

[- [- -

] .AE *C *H**] SAC]ERDOS

P]ROSERPINAE [--[---]MARMOREA [- - ] vacat (0.035 m) ct]oav xXL vaovflXou[--[tcovoc -c] - a pocxoc rYi,ux-aAR ?] vacat (0.02 m) [-

A NEW

LATIN

AND

GREEK

INSCRIPTION

FROM

CORINTH

337

COMMENTARY Line 1: Beforethe A is a slighttraceof a verticalstrokewith a serifat the bottomof theline,whichcouldpossiblybelongto l,N, or H.The interpuncts between the E and the C in line 1 and at the end of line 4 are roughly triangular;both are positionedslightly above the midpointof the line. After the C thereis alsothe traceof a verticalstrokeat the bottomof the line and a horizontalendingin a serif,which can only be an Eor L.There is spacefor only one letterafterthis trace.If the cutteruseda ligaturehere as he did in line 3, then we must admitthe possibilityof morethan one letter.We shouldexpectthe nameof the sacerdos here,but unfortunately, not enoughof this line is preservedto speculatewhat the namemayhave been.2

2. Cf SEG XXXVII 559 no. 4, a bilingualLatin and Greek text from Kassandreiain which A. KoqvtqpL'Xo; Topocv-tvosis honored for dedicatinga gymnasiumwith its itpocxoc,uu-oCxo to the city with his own money. 3. Cf. CIL III 547 from Eleusis: Cere[ri] I [et Proserpi]nae. 4. For examplesof marblerevetment dedicatedat Corinth, see CorinthVIII, iii, nos. 170, 322, 340.

Line 2: Perhapsa continuationof the nameor anothertitle. Line 3: A slighttraceof the serifat the bottomof the diagonalof the Ris visibleat the break.This traceis notvisibleon the photograph,but can be seen clearlyboth on the stone and a squeeze.The last threelettersare cut in ligature,apparentlybecausethe cutterbeganto runout of spacefor the nameof the deityat the end of the line. Fora similaruse of A and N in ligature,see CorinthVIII, iii, no. 152, line 6. If the stele was originallymuchwider,we might expectthe nameof anotherdeity.In this case one could restoreCeres in conjunctionwith or possiblyPluto. Proserpina3 Line 4: There is an interpunctafterMARMOREA. Latin inscriptions fromCorintharenot usuallypunctuatedin this wayat the end of the text except in the case of an abbreviation,which we do not have here.The Latin inscriptionperhapscontinuedto anotherline and finished there, thusexplainingthe vacaton the rightof the stone.The possibilitythatthe inscriptioncontinuedon anotherblock to the right also cannot be excluded. MARMOREA couldreferto anynumberof architectural elementsor to decoration,suchas revetment. Line 5: At the left, beforethe alpha,is a traceof a strokecurvingto the upperright.This couldbe an omicronor,less likely,a theta. We shouldexpecta verbbeforethe directobject.Possibilitiesinclude or xz7x6cxa?v, found on IG IV 203 (lines 19 and 23) from ?z70t6?v Isthmiain relationto 7cpocxoc4tocrocdedicatedin the templesof Demeter and Koreandin the Plutoneion,respectively. Line 6:The rightverticalstroke,the righthalf of the horizontal,and partof the left verticalstrokeof what canonlybe a pi arepreservedbefore the rho.The verticalof the rho extends0.011 m belowthe line andcurves slightlyto the left at the bottom.The finaltwo lettersareAH.Thereis no spacefor furtherletterson this line, thus these two lettersmaybe an abbreviation.If AHis not an abbreviation,then we must once againadmit the possibilitythat the inscriptionextendedonto anotherblock to the right,but this seemsunlikelyas thereis no clampon the top of the stone. if the sawnbottomis the resultof a secondaryuse, then the Furthermore, existenceof a seventhline couldcertainlybe assumed.A hypotheticalseventh line on the extantstele,however,is highlyunlikelyconsideringthat a

338

MICHAEL

D.

DIXON

space of only 0.02 m is uninscribed at the right edge of the stone below line 6. The last two letters, AH, could be an abbreviation for the monetary unit denarius, denarii.This is paralleled in JEphesosno. 27 (lines 111-112, 324,362-363,530-531). In each of these cases, the word TcpoCX6c,qL,%is followed by the name of a divinity in the genitive case, the abbreviationof AH, and then a numerical amount. Another possible abbreviation, especially if the inscription continued onto another block, is Al('utpog). The original width of the stone cannot be determined with any accuracy. All that we can say with certainty is that the stele had to be wide enough to accommodate the three missing letters [SAC]on the left side of line 2. Since calculations of the minimum number of letters missing at the left side of the stone do not yield a satisfactory text, it would appear that the stone was wider than just the three letter spaces needed to fill out [SAC]ERDOSof line 2. THE

DATE

OF THE

INSCRIPTION

To propose a date for this inscription, we must examine the relationship between the Latin and Greek texts, since we have no prosopographicalor historical evidence to serve as the basis for assigning a date. The text is not sufficiently preserved to argue that this is a true bilingual inscription with a literal translation of the Latin into Greek. It is also possible that the two texts are not contemporaneous, for the Greek may have been inscribed after the Latin. The number of bilingual or Latin and Greek texts found at Corinth is too small to improve our interpretation of the inscription.5All that can be said with certaintyis that Greek inscriptions arerarein Roman Corinth in the period before Hadrian. J. H. Kent, in fact, maintains that until the time of Hadrian, nearly all official inscriptions at Corinth were published in Latin.6 Only after this time does the percentage of official texts inscribed in Greek increase in number.At best we can conclude that the primacy of the Latin text suggests a date roughly around the time of Hadrian, or slightly earlier,and that the presence of the lunate sigma and 5. Only five inscriptionsfrom Corinth (CorinthVIII, i, nos. 71, 130; CorinthVIII, iii, nos. 276, 306, 342) have both Greek and Latin on the same stone. Only one of these (CorinthVIII, iii, no. 276), which was found in Solomos and not Corinth itself, is a literaltranslationfrom the one languageto the other. 6. CorinthVIII, iii, pp. 18-19. Kent notes here that of 104 inscriptions datableprior to Hadrian'sreign, only three (CorinthVIII, i, nos. 14, 19, 70) were inscribedin Greek, two of which are lists of victors from the Isthmian games.To this numbershould now be added a fourth publishedby Martin

(1977, no. 1, pp. 178-179 = SEG XXVI 394). Geagan (1975 = SEG XLV 234) also publishedanotherGreek inscription from Corinth predatingHadrian'sreign; it recordsan officialletter of Trajanto a synod at Isthmia.Another list of Isthmianvictors has been published since the appearanceof CorinthVIII, iii (Biers and Geagan 1970) that is datable to A.D. 127, duringHadrian'sreign.A total of only six Greek texts from Corinth can be dated priorto the reign of Hadrian,four of which deal with the Isthmiangames and a fifth with an Isthmian synod.This adds further weight to the case that our text should be dated close to the Hadrianicperiod.

A NEW

LATIN

AND

GREEK

INSCRIPTION

FROM

CORINTH

339

a capital mu in the Greek text points to a date as early as the reign of Hadrian. Close parallels for the Latin letter forms, which might assist us in dating the inscription, do not exist at Corinth.7The presence of a lunate sigma within a Greek text has generally been used to date Greek inscriptions from Corinth to the second half of the 2nd century A.C., but it can be earlier.8The use of the lunate sigma is also frequently associated with the lunate epsilon, cursive (minuscule/lower case) mu, and cursive omega. Of these letters we have only mu preserved on the stone, but not in its cursive form. Several published inscriptions from Corinth on which both a lunate sigma and a capital mu appear have been dated between the reigns of Hadrian and Marcus Aurelius.9The evidence of the Latin and Greek texts together, therefore, suggests a Hadrianic date for this inscription, but does not exclude the possibility that it is post-Hadrianic. POSSIBLE

ASSOCIATION

ISTHMIAN

SANCTUARY

WITH

THE

OF POSEIDON

in line 6 suggests one possible interpretation The use of TCpOCxOCqt'OCtOC

of the text. This word, paralleled elsewhere, refers to adornments in buildings such as temples, stoas, gymnasia, and theaters.10The most useful parallel is IG IV 203 from Isthmia, now in Verona, a catalogue of P. Licinius Priscus luventianus's dedications at the Sanctuary of Poseidon at Isthmia. Priscus and his family arewell known from Corinth and Isthmia as generous benefactors.1"Among Priscus'sdedications recorded on this stele are the 7cpocxocvt'octoc in temples of Demeter and Kore located in the Sacred Glen (lines 16-18), his restoration of the Plutoneion (lines 20-21), and his erection of a stoa next to the stadium (line 24):12 &v ocvtn I vocobv v xoceAtov6aot I xoca'Ap-,uttog abv totg Fv

v tov ts I 7sptf3o?,ov trCg Esp&cgvoc q xaOCt tog

I\ cpog v xaoct Kop-

7. Severalinscriptionsdo show similaritiesin letter forms, but unfortunately,they arenot datableby any criteria other than their letter forms. For example,see Corinth VIII, ii, no. 103, and Corinth VIII, iii, no. 222, where the A, E, M,O, R, and S all show similar characteristics,including serifs.The editorsoffer no dates for these inscriptions. Cf. also Corinth VIII, ii, no. 12, and Corinth VIII, iii, no. 237, which are both dated to the second quarterof the 2nd centuryA.C. Although their letter forms do not exactlyparallelthose on our inscription,they do sharesome similarities. 8. See Corinth VIII, iii, p. 36. For exampleswithin the Hadrianicperiod, see Corinth VIII, iii, nos. 124, 139.

9. CorinthVIII, i, no. 86; Corinth VIII, iii, nos. 139, 223, 269. 10. CIG 3080 (Teos); IG IV 203 (Isthmia);SEG XI 923 (Gytheion); JEphesosnos. 10, 27, 3214+3334 (SEG XXXIII 946) and SEG XXVIII 866 (Ephesos);SEG XXXV 744 (Kalindoia); SEGXLI 910 (Aphrodisias);TAMII, i 396 (Patara);SEGXXXVII 559 (Kassandreia);[Longinus] Subl.43.3. 11. For epigraphicalreferencesto Priscus,see IG IV 202,203; ICGXIV 2543; CorinthVIII, i, no. 105; Corinth VIII, ii, no. 70; CorinthVIII, iii, nos. 199-201, 306; Clement 1976, p. 230 (SEG XXVI 410). For discussionsof Priscusand the inscriptions,see Broneer 1939 (SEGXI 51); Robert 1940; Robert 1966, pp. 754-755; Ritti 1981, pp. 30-

31, no. 6 (SEGXXXIX 340); Puech 1983, pp. 35-41 (SEG XXXV 264); IsthmiaIV, pp. 10-11 (SEG XXXVII 263); Geagan-1989 (SEGXXXIX 340); Jordan(1994, pp. 115-116, note 7 = SEG XLIV 303), who "tentatively associates"anotherinscriptionpreviously publishedby Clement (1974, pp. 110-111 = SEG XXIX 339) with Priscus'sbuildingprogram;Pierart1998, pp. 97-100; for his prosopography,see PIR2V, i, p. 55, no. 232. 12. For a discussionof the temples located in the SacredGlen, see Isthmia II, pp. 113-116. IG IV 203 also records a numberof other monumentsthat were either constructedor restoredby Priscus.

340

MICHAEL

D.

DIXON

aXUTo

C JVLaxGmv xoat7po0xoUN,xc%tav xaotd pov6a[ot] I 'I ay[ao sv xod. ox t vaxo[bU] Iti Evstcpt'g xaoctti v xao TO 1At cGs FOV]It-CVsLOV xao tCaL &va 3a6 xad t-a KOpq , sv v xaod xa?atotog xRsoktalu avVIVfxtfxc670 G3 va

tCJv iatcov

TCSG%aXSFXGSV vv Gbv

I to<;

O aVt6g

xsxOC,LPC)qLS)VOL;

wTe(p ayopavo%xaI

xao

I tv

0o'xot;

GtOav v xoat

T

-co

TcpoIoG1XoN,VoxG7v

G toX vv

v &CviO-Xsv. V13

A second fragmentary inscription, SEG XXXIX 340, also relates to the activities of Priscus at Isthmia.14Recorded on this stele is the response from a proconsul, possibly the governor of Achaia, regarding Priscus'srequest to purchase the ruins of the stoa of Regulus in order to transform it into fifty oikoifor competitors at the Isthmian games.15The completion of this project is only one of the dedications for which Priscus was honored in IG IV 203. The dedication of 7cpocxocvtvoc-rncin temples of Demeter and Kore and the restoration of the Plutoneion, together with the reference to a stoa on the Isthmian stele (IG IV 203), raise the possibility that this new Corinthian inscription is another record of Priscus'sbenefactions at the Isthmian sanctuary.This new inscription is the only literary or epigraphic testimony for temples dedicated to these deities at Corinth, except for a reference to Demeter and Kore on Acrocorinth by Pausanias,16but, as we have seen, temples of Demeter and Kore and a Plutoneion at Isthmia are attested. Severaldifficulties areimmediately apparentin associating the inscription with Priscus'sbuilding program at Isthmia. First, Priscus'sprogram is generally dated to the second half of the 2nd century A.C.17 Such a date is most likely too late for this inscription. Second, there seems to be no way to construe Priscus's name with the surviving letters in line 1. Furthermore, we would not expect the title of sacerdos.Priscus is known to have held the title of ocpytspsg 8t&crtMou when he was honored for his benefactions at Isthmia on IG IV 203.18 We should expect archiereus,a simple transliterationof the Greek into Latin, as Priscus is referredto in Corinth 13. IG IV 203, lines 14-27. The text is based on Geagan'sinterpretation(1989, p. 350), with a few modificationsbased on the text of Ritti (1981, p. 31, no. 6) and on the photographsprovidedby both Broneer(1939, p. 187) and Ritti (1981, p. 30). Ritti bracketedthe last two letters in lines 17 and 19, and the last three in line 18, which were readboth in IG IV 203 and by Geagan. Ritti did not include the vacats in his edition, which were noted by Geagan, nor did he bracketthe last two letters in line 20. Both photographsclearlyshow the vacats and indicate that the final two letters of line 20 no longer surviveon the stone, and thus I bracketthem here.

14. Geagan 1989 = SEGXXXIX 340. This inscription,which Geagan refersto as Stele B, consists of ten joining fragmentsfound at Corinth and Isthmia.The Corinthianfragmentwas originallypublishedby Broneer(1939). For this inscription,see also Corinth VIII, iii, no. 306. The Isthmian fragmentwas first reportedby Broneer (1955, p. 124), who concludedthat the stele was originallyset up at Isthmia; see also the comments of Geagan (1989, p. 349). 15. For the possible identificationof the proconsulwith the governorof Achaia, see Geagan 1989, pp. 354-355. For a discussionof the oikoiand IG IV

203, see Jordanand Spawforth1982, pp. 67-68 (SEG XXXII 364), and Jordan1994, p. 115. 16. Paus.2.4.7. For a detailed discussion of Demeter and Kore at Corinth, see CorinthXVIII, iii, pp. 1-11. 17. For the date, see Geagan 1989, pp. 358-360; Gebhard 1993, pp. 89-93; and Gebhard,Hemans, and Hayes 1998, pp. 438-444. 18. IG IV 203, lines 4-5; see also IG IV 202; IG XIV 2543; CorinthVIII, ii, pp. 53-54, no. 70; and CorinthVIII, iii, no. 199. Priscusis also attested simply as Lepe0c:CorinthVIII, iii, no. 201, a statue base on which is inscribed 'Ioo3svtxcxvoc I ispe6c.

A NEW

LATIN

AND

GREEK

INSCRIPTION

FROM

CORINTH

34I

VIII, ii, no. 70, and CorinthVIII, iii, no. 199. These discrepancies,therefore, indicate that Priscus'sbuilding program and this inscription should not be associated. The final issue that must be addressed is the original location of the stele. The inscription is probably too early to be associated with Priscus's building program at the Isthmian sanctuary,and the monuments mentioned in it might better be sought at Corinth itself This is, however, not conclusive evidence for the original location of the stele. Its findspot in Corinth, as has alreadybeen seen in the case of SEG XXXIX 340, is not proof that the inscription was originally set up there.19It is certainly possible that the stone was originally set up at Isthmia and later moved to Corinth.

CONCLUSION

19. This is also the case with Corinth VIII, iii, no. 201, a statuebase

in honor of Priscusfound in New Corinth,which Kent believed originatedat Isthmia, not Corinth. I found no associatedor joining fragmentsamong the unpublished inscriptionsfrom Isthmia.

The inscription on this stele, if it refers to buildings at Corinth, may shed important new light on the topography and monuments there. It is unfortunate that the name of the dedicant is not preserved, but it records what may be the first testimony of a temple of Pluto and a stoa possibly associated with it at Corinth. It seems unlikely that this inscription should be associated with the building program of P. Licinius Priscus luventianus at Isthmia. The fact that the monuments mentioned on this new inscription are also attested there does, however, raise various other possibilities: 1) that the stone refers to monuments at Isthmia and was originally set up there; 2) that the stone refers to monuments at Isthmia, but was originally set up at Corinth; or 3) that the stone refers to buildings at Corinth and was originally set up there. The last of these, if true, would also suggest that either a local citizen or a prominent Roman other than Priscus was responsible for the construction or restoration of the monuments mentioned on the stone. More evidence, however, must be forthcoming to place any conclusion beyond dispute.

REFERENCES Biers,W. R., and D. J. Geagan. 1970. "ANew List of Victors in the Caesareaat Isthmia,"Hesperia 39, pp. 79-93. Broneer,0. 1939. "AnOfficial Rescript from Corinth,"Hesperia 8, pp. 181190. 1955. "Excavationsat Isthmia, 1954"Hesperia 24, pp. 110-141. Clement, P. A. 1974. "Isthmia Excavations,"ArchDelt 26, 1971, B' 1 [1974], pp. 100-111. 1976. "IsthmiaExcavations," ArchDelt 27, 1972, B' 1 [1976], pp.224-231.

Corinth = Corinth: Results of the Excavations Conductedby the American School of Classical Studies atAthens. VIII, i = B. D. Meritt, Greek Inscriptions, 1896-1927, Cambridge,

Mass., 1931. VIII, ii = A. B. West, Latin Inscriptions, 1896-1926, Cambridge, Mass., 1931. VIII, iii = J. H. Kent, The Inscriptions, 1926-1950, Princeton 1960. XVIII, iii = N. Bookidis and R. S. Stroud, The Sanctuary of Demeter and Kore: TopographyandArchitecture,

Princeton 1997.

342

MICHAEL

Geagan, D. J. 1975. "ALetter of Trajan to a Synod at Isthmia,"Hesperia 44, pp.396-401. .1989. "The Isthmian Dossier of P. Licinius PriscusJuventianus," Hesperia58, pp. 349-360. Gebhard,E. R. 1993. "The Isthmian Games and the Sanctuaryof Poseidon in the Early Empire,"in The Corinthiain theRomanPeriod (JRASuppl. 8), T. E. Gregory,ed., Ann Arbor,pp. 78-94. Gebhard,E. R., F. Hemans, andJ. W. Hayes. 1998. "Universityof Chicago Excavationsat Isthmia, 1989:111," Hesperia 67, pp. 405-456. Isthmia= Isthmia:Excavationsbythe Universityof Chicagounderthe Auspicesof theAmericanSchoolof ClassicalStudiesatAthens,Princeton. II = 0. Broneer,Topography and 1973. Architecture, IV = M. C. Sturgeon,SculptureI: 1952-1967, 1987. Jordan,D. R. 1994. "InscribedLead Tabletsfrom the Games in the

Michael D. Dixon OHIO

STATE

UNIVERSITY

DEPARTMENT

io6

DULLES

230

WEST

OF HISTORY HALL AVENUE

I7TH

COLUMBUS, DIXON.7I@OSU.

OHIO

432IO

EDU

D.

DIXON

Sanctuaryof Poseidon,"Hesperia 63, pp. 111-126. Jordan,D. R., and A. J. Spawforth. 1982. "ANew Document from the Isthmian Games,"Hesperia 51, pp. 65-68. Martin,T. R. 1977. "Inscriptionsat Corinth,"Hesperia 46, pp. 178-198. Pierart,M. 1998. "Pantheonet hellenisationdans la colonie romainede Corinthe:La 'red6couverte'du culte de Palaimon a l'Isthme,"Kernos 11, pp. 85-109. Puech, B. 1983. "Grands-pr6treset Helladarquesd'Achaie,"REA 85, pp. 15-43. Ritti,T. 1981.Iscrizionie rilievigrecinel MuseoMaffeianodi Verona,Rome. Robert, L. 1940. "Un edifice du sanctuairede l'Isthme dans une inscriptionde Corinthe,"Hellenica 1, pp. 43-53. . 1966. "Inscriptionsde I'antiquiteet du bas-empirea Corinthe,"REG 79, pp. 733-770.

HESPERIA

69,

PY

2000

LO

S

R E G

I O

NAL

Pages 343-380

A

RC

H

ALEOLOCGI CAL

PROJECT, SIR

WILLIAM

IN THE

PYLIA

PART

111

GELL'S AND

REGIONAL

LANDSCAPES

IN THE

THE

OTTOMAN

SECOND

ITINERARY

MOREA

IN

PERIOD

ABSTRACT This article previewsthe study of the Second Ottoman period (1715-1821) by members of the Pylos Regional Archaeological Project. By closely comparing Sir William Gell's apparently dispassionate descriptions of the Navarino Bay areawith other documentaryand archaeologicaldata,we suggest that reconstruction of settlement and land use relying solely on Gell's descriptions can result in misrepresentation.This conclusion has implications for modern Greek social and economic history, since the image that Gell sketches appearsto supporta commonlyheld belief that settlement during Ottoman occupationwas concentratedin more mountainousareas,while the lowlandswere largelydevoid of permanenthabitation. Since the earliest days of regional archaeological surveys in Greece, the study of the periods of Ottoman occupation has proven frustrating.' For most earlier periods, artifact typologies-based on chipped stone or ceramics-are sufficiently well understood to allow surface remains found by means of intensive survey to be dated, sometimes with great precision. Pottery from the times of Turkish domination, on the other hand, is relatively poorly understood, particularlyfor the 18th and early 19th centuries, and can generally be assigned only to broad chronological parameters. 1. We would like to thank Pierre MacKayfor permissionto quote from his unpublishedtranslationof Evliya ?Ielebi,and Mary Lee Schmidt of the Lloyd Libraryand Museum in Cincinnati for furnishingus with photographicnegativesof maps included in the Atlas volume preparedby the Expedition scientifiquede Moree. Peter Topping offered much encouragement at an earlystage in this project.We are

also gratefulto Siriol Davies for making availableto us in advanceof publication the resultsof her researchesin Venice,conductedon behalf of the Pylos RegionalArchaeologicalProject (PRAP). We thank Susan E. Alcock, directorof historicalstudies for PRAP, for her encouragement.SharonGerstel, who has studied ByzantineMessenia for PRAP, helped us obtain copies of publicationsfrom the collections of

Dumbarton Oaks.The constant support of Debi Harlan has been indispensable to the successof our endeavors.We thank the BaUbakanlik Archives in Istanbulfor their generosityin permitting Zarinebaf-Shahrto study the Ottoman cadastralregisterTapu Tahrir (hereafter,TI) 880. Finally,the article has benefited from the adviceof Hesperia's anonymousreferees.

344

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Recently this situation has begun to ameliorate,but artifactsremain difficult to date,2 and the chronology of churches can sometimes be the most important source of evidence for patterns of settlement.3 Archaeologists have thus been compelled to place great emphasis on written texts in attempting to reconstruct settlement and land use patterns. However, the number and nature of such textual sources vary widely from one part of the Aegean to another, depending largely on the way in which an areawas integrated within the power structure of the Ottoman state. In lieu of access to official Ottoman administrative documents, archaeologists, as well as historians, have often relied heavily on the accounts of western travelers. For the final century and a half of Ottoman rule, these can be extraordinarilyrich, especially for the Greek islands, but also for many parts of mainland Greece.4 It has become increasingly clear,however, that even the most apparentlystraightforwardaccounts can hardly be read at face value, without consideration of the circumstancesin which their narrativeswere constructed. Reevaluation of their texts can, in turn, have significant consequences for how we view the economic and social history of a region. This report is intended to preview the approach to the study of the Second Ottoman period (1715-1821) followed by members of the Pylos Regional Archaeological Project (PRAP), as we have begun to addressthe issues outlined in the preceding paragraph.5It was clear from the beginning of the project that it would be essential to make a concerted effort not only to collect artifacts of the recent past, but also to obtain information from Turkish archives that would be relevant to the history of settlement and land use in western Messenia. We have vigorously directed investigations since 1994 toward the achievement of this goal, and results are now being preparedfor publicationin a monograph.6As our researchprogressed, it became increasingly obvious that there existed a critical disjuncturebetween information recorded about the Pylos area by travelers,on the one hand, and the records of Ottoman administrators and the results of archaeological investigations, on the other. Travelers seemed to be viewing western Messenia through the filter of their own experiences.The notion that travelers'accounts are not undistorted windows into past landscapes is, of course,well understood by others who have recently approachedtheir analysis in a Greek context. Susan Sutton has, for example, observed that "although [travel narratives]made claims of visual legitimacy, of describing what was tangibly there to be seen, they were nevertheless remarkably myopic.

In this report we focus on a specific account of a Peloponnesian landscape in the early 19th century,that of the areanorth of the Bay of Navarino by Sir William Gell, generally considered to be one of the most accurate and careful of authors who traveled to Greece in the decades immediately prior to the Greek Revolution. We have chosen this account because it is tightly tied to certain features of the landscape and includes sufficient detail about travel times to facilitate the identification more or less exactly of his route through that landscape.There is considerablevalue in presenting a detailed account of Gell's route, in and of itself, with reference to the contemporary (1997) landscape. This is in part because the landscape is

2. Forbes1997;Jameson,Runnels, and van Andel 1994. For recent studies of post-Byzantine pottery,see Vroom 1998 and Hahn 1997. 3. E.g., Koukoulis1997. 4. For the Greek islands,see Wagstaff 1982; Bennet and Voutsaki 1991. 5. PRAP has exploredarchaeological remainsin an areaof approximately 100 km2centeredon the Englianos Ridge and the Bronze Age Palaceof Nestor. A multidisciplinaryteam of earth scientists,historians,art historians, and archaeologistshas studied the history of settlement and land use in southwest Messenia, north of the Bay of Navarino,from earliestprehistoric times until the presentday.See Davis et al. 1997; Zangger et al. 1997; Davis 1998. 6. By F. Zarinebaf-Shahr,J. Bennet, andJ. L. Davis, provisionallyentitled The Morea in the Wakeof the Ottoman Reconquest (A.D. 1 715):A Case Study from the Southwestern Peloponnese.

7. Sutton, in press.

SIR

8. See FrangakisandWagstaff 1987. The quotationis from Frangakis-Syrett and Wagstaff 1992, p. 439.

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

345

increasingly subject to development and may no longer be available for this kind of study in ten or more years. Also, Gell's account, because of its precise relation of times and places, gives the impression of offering an authoritativeand authentic view of the quality of the landscape in terms of settlement and the types of activities that were carried out in it in his day. We present a collection of other 18th- and early-19th-century travelers' accounts, as well as additional sources, to set against Gell's. The travelers'accounts tend to concur with Gell's picture of a rather desolate landscape, while the other sources suggest this may have been an impression colored by traditions of travel narrative in the region and by Gell's own prejudices regarding Greece under the Ottoman Empire. Rereading his accounts in light of the documentary evidence has led us to believe that Gell has somewhat distorted the picture of settlement and agriculturalactivity in the lowlands around the Bay of Navarino. Archival information from Ottoman sources, as well as archaeological discoveries made by PRAP, suggests that Gell either ignored or was oblivious to some of the agriculturalactivities and evidence for ruralsettlement that he might have observed and recorded in the course of his journey through the Pylia. Any reconstructionof patterns of settlement and land use that would rely on his descriptions alone, or, for that matter,on the narrativesof western travelersby themselves, would result in serious misrepresentations of reality. This conclusion in turn has significant implications for the social and economic history of modern Greece, since the image that Gell sketches would, on first examination, appear to support the commonly held belief that "settlement and population in Greece were concentrated in the more mountainous areas during the Ottoman occupation, whilst the lowland was largelydevoid of permanenthabitation,"a picturethat Frangakis-Syrett and Wagstaff have contested in two recent articles.8Their arguments are convincing, insofar as they concern settlement distribution at the macroregional level, and in the Pylia there is ample evidence that lowlands were settled and exploited as significant components in agriculturalproduction systems of the Ottoman period. Although the coasts of the Pylia do appear to have been relatively underpopulated in comparison with nearby uplands, they were hardly desolate, and major centers of Greek settlement were not far distant. A second goal of the article is, therefore, to examine the relationship between settlement location and elevation at a more detailed level in this spatially and temporally restricted landscape from the early 18th to the mid-19th century,using the sources outlined above and data derived from PRAP's archaeological research.The picture sketched here provides a preview of the much fuller treatment alluded to above, in the context of the publication of an Ottoman tax register dating to the year after the Ottoman reconquest of the Morea. The conclusions reached in this articlewould seem to demand greater sophistication in the analysis of travel literature as well as the dedication to archival research of more resources than have been generally invested by regional studies projects. Potential new archivalsources of information for the history of Ottoman Greece have often been ignored by archaeologists. Semi-independent communities (koinotites)of Greek or Albanian-

346

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

speaking Orthodox Christians in many of the Cycladic islands, the Sporades, and the islands of the Saronic Gulf maintained local governments and paid taxes to the Porte in fulfillment of communal obligations.9 Ottoman administratorsrarelyvisited and local archiveswere recorded in Greek; some of the latter have in part been preserved, as for the island of Keos.10Unlike many of these Greek islands, the Morea, or Peloponnese, was fully incorporated into the Ottoman Empire as a province, or eyalet. In contrast to the islands, the Morea was governed and taxed directly by authorities ultimately designated by the central government in Istanbul. Turkish records of transactions with the villages and cities of the Peloponnese were kept in the archives of Istanbul. Access to these has been difficult until recently, and few archaeological teams could master the historical, linguistic, and epigraphical skills requiredto read and interpret the documents curated there. But high-quality textual information can, in fact, be obtained, and archaeologists have lately come to realize that such new unpublished data are readily accessible not only from the archivesof the Ottoman state in Istanbul"1but also from those of Venice.12 Neglect of critical archival sources for the study of post-Byzantine Greece reflects in part a general apathy among archaeologists concerning research and instruction in the recent history of Greece, a tendency particularlyacute outside Greece itself Certainlyin western Europe and North America, a preponderance of interest has focused on the history and archaeology of the Classical and pre-Classical periods. But recently there have been signs that attitudes are beginning to change. The archaeologies of Frankish,Venetian, and Ottoman Greece, long marginalized as shameful chapters in the history of Hellenism, are at last attracting the attention that they deserve. The year 1998 marked a watershed in this respect, with the convening in May of a two-day conference on the archaeology and history of post-Classical Greece, organized by the British School of Archaeology at Athens and the Ionian University of Corfu."3 Such an increase in interest in later Greek studies, and in archival research in particular,is of significance to archaeologists engaged in regional studies in the Peloponnese. It is of special importance for the history of the 18th century, a time period for which detailed descriptions of many localities areremarkablyscarcein sourcesthat have traditionallybeen tapped by scholars. Many of the so-called "New Wave" of surface surveys in Greece14 have been situated in areas distant from Ottoman centers of power; the accounts of foreign visitors to these places, if they exist at all, are generally impoverished. Even the richest narrativesof most European travelers to Greece tended to concentrate on urban centers and almost never supply the information one would like to have concerning ruralhinterlands. In regions for which travelers'accounts are plentiful, an evenly distributed spatial coverage of areas chosen for archaeological investigation is never provided and, of course, the accounts are often subject to the inherent biases discussed in this article.1"Rich Venetian administrative documents16exist for the Peloponnese in the period of Venetian domination between 1685 and 1715, but there are no comparable published bureaucraticdocuments for the next hundred years. In addition, there can be a sometimes dramatic mismatch between archaeological and archivalin-

9. On the independentcommunities of the later Ottoman Empire,see Kontoyioryis1982; Koukou 1980; Davis 1991. 10. See Cherry,Davis, and Mantzourani1991; Davis 1991. 11. Regional archaeologicalstudies in Greece andTurkeyother than our own have sponsoredarchivalresearch; for example,see Kiel 1997; Nixon, Moody, and Price,in press;Bintliff 1999; and Robinson, in press.Other recentpublicationsof Ottoman documents,although not commissioned by archaeologicalprojects,are nonetheless of greatpotentialvalue to archaeologists:e.g., Kiel 1987; Kiel and Sauerwein1994; Kiel 1990; Valta1989. 12. For Venice too the situationis also graduallyimproving,as archaeologists come to realizethe relevanceof unpublisheddocumentsto their work. For example,see Topping 1976; Davies 1994. Pioneering researchwas conductedin Messenia by Peter Topping;see Topping 1972. 13. Jointly convenedby D. Tsoungarakis,J. L. Bintliff, and E. Angelomati-Tsoungaraki,this gatheringfeaturedpapersthat built significantlyon the foundationslaid by Lock and Sanders1996 and were concernedwith a broaderchronological range. 14. See Cherry 1994. 15. See also Sutton, in press. 16. E.g., Topping 1972; 1976.

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

347

formation, even in periods for which documentary evidence is plentiful. In some areas, such as Aigion,"7extremely detailed maps of settlements and fields exist for the early 18th century,but no regional studies projects have yet been initiated in the region. New archivalresearchcan assist regional studies projects in achieving their stated objective of integrating fully the evidence of text with that of material culture.The Ottoman archives in particularoffer the possibility of acquiring complete, spatially continuous descriptions of the regions to be examined by our projects, accounts that recorded all places of settlement and agriculturalexploitation. Once this information is in hand, we can begin to develop archaeologicalstrategies that earmarkthese locations for special investigation. Starting with the knowledge that certain places were occupied at a particulartime, it will be possible to compare patterns of settlement and land use reconstructedfrom documentary evidence with the material remains collected by surveys.In such a manner, it might perhaps be possible to improve our understanding of the material culture of the early modern period, especially by studying small sites that were the locus of settlement for only a short duration.

SIR WILLIAM GELL AND OTHER WESTERN TRAVELERS NORTH OF THE BAY OF NAVARINO William Leake, who traveledin Greece in 1805-1807 and 1810, describes the western coast of Messenia (Fig. 1) that lies between the modern town of Pylos (formerly Ottoman Anavarin and Greek Navarino) and the Classical town of Koryphasion, on the one hand, and Kyparissia (formerly Arcadia), on the other, with a brief and despondent note: There is no portion of the Peloponnese less noticed by ancient authors than the part of Messenia lying between Coryfasion and Cyparissia,though its length is not less than twenty miles.18

17. See Dokos and Panagopoulos 1994. 18. Leake 1830, p. 425; Wagstaff,in press.

The neglect of the ancients was echoed in the accounts of modern travelersto the area in Leake's own time-the last years of the Ottoman occupation of Greece-including those of his occasional traveling companion, Sir William Gell. Few foreigners traveled between Pylos and Kyparissiain the Second Ottoman period and, of those who did, still fewer have left us descriptions of either human or natural landscapes as they then appeared.The accounts that do exist, however, depict a countryside largely deserted by its former Christian population. Gell, in particular,in two published works, would seem to support such a conclusion. Were the narrativesof his journey around the Bay of Navarino accepted uncritically, it would be easy to surmise that the coastal areas around the bay were relatively desolate. Indeed, in the course of a journey of more than five hours he reported that he did not see a single individual other than the members of his own party. Such a conclusion would, however, be entirely mistaken. Much of the

348

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Figure 1. Map of the

V.

.1

Peloponneseandwestern

PatrasO

Messenia.

R. J. Robertson

Co

ron!

meihoni

-4

10

I 0

Kypanssia

50 sk

J

areathrough which Gell traveledlay in the area studied by PRAP. We are, therefore, in a position to marshalboth archaeologicalevidence and documentary information (from Ottoman archives in Istanbul here published for the first time) to show how impressions gained from first readings of Gell's accounts of his journey can be corrected and moderated. The result is a nuanced picture of human settlement in the area as it existed in the time of the second Ottoman occupation of the Peloponnese (1715-1829), and one that is richer than any previously available to scholars. It thus permits a reconstruction of patterns of settlement in the coastal areas of Pylos that is sensitive both to long-term change and to drastic short-term fluctuations. Gell's accounts of the geography around the Bay of Navarino appear characteristicallyfrilland pedantic.9 Two works recount his travelsin 1804 in the Peloponnese.20The first, published in 1817, is his Itinerary of the Morea:Being a Descriptionof the Routes of That Peninsula, a bare list of routes between major towns, with notes on sights seen along the way, and their distances,expressedin minutes elapsed in transitfrom the point where he had previously stopped. Six years later, in 1823, he presented his full reminiscences in Narrative ofa Journeyin theMorea, a literaryjournal that, among other things, adds considerable detail to the descriptions in his Itinerary. A complete account of his trip through the area later investigated by PRAP can thus best be gleaned from an interlinear compilation of the two texts (excerpts from the Narrative in italics).

19. Some of Gell's notebooks are now curatedby the British School of Archaeologyat Athens: see Woodward and Austin 1925-1926. Notebook 1 contains materialconcerning"Attica, Boeotia, Phokis, etc."that was later recastin Gell 1819. On Gell's life, and his researchon behalf of the Society of Dilettanti, see also Ridgway 1996 (emphasison Gell's activitiesin Italy); Wroth 1921-1922; and Simopoulos 1985, pp. 120-143 (emphasison Gell's activitiesin Greece). 20. Gell 1817; 1823. For the date of his travels,see Gell 1823, p. 4.

SIR

WILLIAM

NAVARINO, FROM

1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

GELL

OR NEOKASTRO,

NAVARINO

14.

15.

21. This is Gell's title for his route 25 in Gell 1817, pp. 51-52, coveredin Gell 1823, pp. 59-62. Note.sfor this itineraryare not containedin the documentspreservedin the archivesof the British School at Athens. We are gratefulto the archivist,Anne Sackett, and to her assistant,Malcolm Nicolson, for investigatingthis matterfor us. 22. Each place (station)where Gell makes an observationis numbered.The time elapsed since the previousstation is recorded(in minutes), followed in parenthesesby the total elapsedtime (also in minutes) since leaving Navarino.

16. 17. 18.

TO THE

IN

THE

PYLIA

349

TO GARGAGLIANO21 RIVER

KOURBEH

27 (27) Havingproceededby a bad roadfromthe fountain in the Greekvillageof Navarino,1.a ruinedtower,r. a fountain. The port1.22 The trackrunsalong the easternshoreof theportfor sometime, afterwhich it descendsinto an alluvialplain, 10 (37) In the plain. 3 (40) A ruined bridge, and little river. 2 (42) Another stream. 3 (45) Ascend. 3 (48) A plain. Mt. Tabolachi, or Pilaw, r. a conic hill. 9 (57) A ruined bridge, with deep water. Beshli. Two small brooks, and one after. R. a pretty glen. 1 (58) A bridge, and rice field, in a marshy plain. 12 (70) The hills from r. approach the water. Rice grounds and shrubby hills. 8 (78) A river.On r. bank a garden. Turn NE. Church of Agio Nicolo r. 7 (85) RiverKourbeh.

FROM THE RIVER

12. 13.

S ITINERARY

KOURBEH

TO THE ROMANOU

RIVER

5 (90) The bridge. 10 (100) Hills close to the road on r. Under the rook of old Navarino, or Pylos, is a white house, and 1. of it the village Petrachorio. leaving the little villages of Petrachorioand Leuka on the left, 20 (120) Boidochilia, at the north end of Pylos, 1.A hut or place called Geophyre, perhaps a bridge. A grove of olives. and Gephyraeand Lisaki on little knollson the right. 19 (139) An eminence approaches the road r.The plain extends on r. See the villages, or tchifliks of Osman Aga and Haslan Aga. The country-houseand cypressof a certainOsmanAga are soonafterseen on the right, betweenwhich and the village of Haslan Aga 4(143) A platanus and aqueduct. 2 (145) A well, house, and hill called Lirachi. 19(164) R. a wooded valley, and bridge over the Romanos. is a pretty woodedvalley wateredby the river Romanus,which is crossedby a bridge.Theseresidences generally consistin a tower,overlookingthe humbledwellings of thepeasants, who cultivatethe soil, whichforma quadrangleinclosingthe house. The territorywas at that time cultivatedwith lupins,and hadjust produceda cropof Indian cornin theplains. The aspectof the countrywas howeverneitherfertilenor inviting, and muchof it was neglected.

BENNET,

350

FROM THE ROMANOU

19. 20. 21. 22.

23. 24.

25. 26. 27. 28.

29. 30.

DAVIS,

RIVER

AND

ZARINEBAF-SHAHR

TO GARGALIANI

On the top of the oppositebank. Descend.Heath andtrees. A water-course.Wild rocksand treesr. and1. BrisomeroNerro,and a woody dell. wepassedthe In thenextdellintowhichwe descended, Brusomavo bya bridge,and weredelightedbya groveorthicketof a beautifulshrubbery oneachsideof ourpath. arbutus,whichformed 12 (222) A water-course,or aqueduct.Arbutus. 28 (250) A summit,aftera steepand dangerousdescent. Weascendedfrom hencebya steepanddJf culttrackto a summitwhichat lengthafforded afine view oftheseaandProte; 7 (257) A cultivatedhollow. 18 (275) A semicircular valley. 23 (298) A summit. 5 (303) See Gargagliano. arrivedat andaftera longanduninterestingjourney in theeveningofthelastdayofJanuary,without Gargagliano havingmeta singleindividualon theroad. 9 (312) Chapelof St. Nicholo. 3 (315) Gargagliano,a largeGreekvillageoverlookinga plain,andProteto the west.The housesaregood, and the situation,which is lofty,is muchembellishedby manycypresses. 19 (183) 18 (201) 7 (208) 2 (210)

Totaltime elapsed:5 hoursand 15 minutes(315 minutes). MAPPING

GELL'S

RoUTE

Gell'sroutebetweenNavarinoandGargalianiin partfollowsthe courseof modern(1997)highways,butoftendeviatesfromthem(Fig.2). It is, nonetheless,possibleto map the route that he followedquite precisely,since Gell suppliesus with informationaboutthe total length (in minutes)of eachof his itinerariesandthe timebetweeneachstationwherehe stopped to makeobservations.By dividingthe total distancetraveledby the total minutesin transit,we can determineapproximately his rateof travelper minuteandthereforeestimatethe distancesbetweenstations.In this way, the positionof eachstationcanbe determinedwith reasonableaccuracy.23 Locationsof stationsestimatedin this fashioncan,of course,be only approximatesinceGell'sactualtravelspeedwill havevarieddependingon the natureof the terrainthroughwhichhe wastraveling.He, in fact,makes this point himselfwhen he describesthe terriblehardshipsof traveling acrossa Greekplainon horseback: of a Greekroadovera disNothing can equalthe impracticability trictof pointedlimestone-rocksperpetuallyappearingat the surface, exceptthat acrossthe succeedingvalleyor plain,when it hasbeen soakedby the autumnalrains;and the shortherbagebeginning to springup in the winterrendersit necessaryfor the travellerto

23. In addition to Gell's,three other publishedjournalsshed substantiallight on the topographyof the areabetween the Bay of Navarino and Gargalianiin the early 19th century:A. L. Castellan'saccount of a journey from Navarinoto Filiatra (Castellan1808, pp. 90-100); Bory de Saint Vincent'srelationof his travels (1829) from Navarinoto the French militarycamp at Yialova (Bory 1836, pp. 135-139) and to Gargaliani (pp. 165-169); and Pouqueville's narrativeof a journeyfrom Gargaliani to Navarino(Pouqueville1826-1827, VI, pp. 26-27). Each in part followed the same route as Gell, and their accountsat times clarifydetails that are obscurein his.

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

351

Figure 2. Map of Gell's route, showing modem place-names. R. J. Robertson

ir*dv

q* c

faboand dc

dphl*gei

attend to his own involuntary agitations, while the luggage-horse, after a thousand slips, and as many recoveries, almost invariablyputs a stop to further progress for a short time, by receiving a desperate fall after a slide of several feet and a succession of unavailing struggles.24

24. Gel 1823, pp. 59-60. See also Gell 1817, pp. ix-x.

Gell's journey from Navarino to Gargaliani has been divided into three segments, the beginning and end points of which can be located on the ground without dispute. This method allows the position of each of his observation stations to be determined with greater accuracy,since the terrain over which Gell was traveling in each segment differed: the first segment, very marshy;the second, drier and flatter;the third, quite steep. It is to be expected, therefore,that a minute of traveltime in one segment would not have resulted in the same amount of forwardprogress in another.The empirical results of this method are encouraging: almost all of Gell's stations can be located, and knowing their situation allows several difficulties in the interpretation of his text, introduced by ambiguities in his prose, to be resolved.

352

THE

BENNET,

SETTLEMENT

DAVIS,

AND

ZARINEBAF-SHAHR

AT NAVARINO

Gell'sitinerarybeginsat Navarino,the moderntown of Pylos.In his day, the settlementat Navarinoconsistedof two parts,the fort of Neokastro occupiedby Muslims,and a largelyOrthodoxGreeksuburboutsideit.25 Gell himselfstayedin the Greekquarteras the guest of a wealthyGreek merchantof the town, GiorgiosKonomopoli[sic], and paid only a brief visit to the walledTurkishcitadel.26 It is clearthatthis Greeksuburbat Navarinowasnot a recentfoundation. Maps compiled at the time of the Venetian occupation of the Peloponnese(1685-1715)indicatethe sametwo-partstructureof the communitythat Gell describes.The most usefulfor understandingthe geographyof the settlementis entitled"Piantadi NavarinoNovo relevatasoto GriniProrGeneraliin Morea,"reproil comadodel Illmoet EccmoSigrFraco ducedby KevinAndrewsas plateXI in his CastlesoftheMorea,wherethe suburb(varochi)is indicatedimmediatelyeast of the fortressandnorthof the line of an aqueductthat ran south of the modernroad to Modon (Methoni).27The topographic relationships represented in this map are

Andrews'plate XI is confirmedby a close-up view of the settlement.28 particularlyvaluablefor the interpretationof Gell'stext in that it shows the locationof the publicfountainthat is the firstlandmarkrecordedby Gell andthe "goldenmilestone"for his measurements. This "Fontanadel Basa"("Fountainof the Pasha")is the onlyfountain labeledon Andrews'plateXI, andit is situatedsomedistanceto the north of the varochi,in a placethat todaywould lie to one side of the modern centralsquare(plateia)of Pylos, nearthe locationwhere the roadnorth from Navarinobegins its ascent along the easternedge of the Bay of Navarino.The fountainwas a prominentlandmarkand attractedthe attention of other travelers.For example,it is describedby Bory de Saint AcVincentof the FrenchExpeditionde Moree,andis alsoillustrated.29 cording to him, there were no wells or springs inside the fortressof Neokastroandthe aqueductthathadpreviouslyfed its cisternswas dilapidated.He foundwells nearthe fountain,which he speculatedwas Venetian in date,althoughit was uninscribed.Blouet also discussedthe locaenplanches surunemauvaisejetee tionof the fountain:"La,nousabordames

25. FrancoisPouqueville,the French diplomat,capturedby piratesin 1800 and taken to Navarino,describedboth fort and suburb(Pouqueville1820-1821, V, p. 122): "The districtof Navarino... counts within the walls of its modern capitalsix hundredTurksand one hundredand thirty Greekswho inhabit the varochi."In the second edition of his work (1826-1827, VI, p. 73, note 1), he attachesa table of the population is his word) of the "canton" ("cadastre" of Navarinlisted byfamily. For Navarin itself, 142 families arelisted. It is clear

from the total numberof familieslisted for the "canton"(1019), however,that Pouquevillemust have made an error and meant to recordindividuals, since that populationfar exceeds the figure of 447 he gives for the Ottoman poll tax (cizye) of the kaza, or district (18201821, V, p. 15). In theory,the poll-tax figure (i.e., the numberof non-Muslim adult males) ought to approximatethe numberof families.It is clear,therefore, that the figure of 130 individualsis more likely to be correct.This correctionis assumedfor all figuresbased

on Pouqueville's1815 table. On Pouqueville'scaptureby pirates,see Lair 1902; 1904. 26. Gell 1823, pp. 18-21. 27. Andrews 1953, p. 244, pl. XI in Appendix C. Partsof the aqueductstill exist.The suburbwas in existence still earlierin the 17th century,before the Venetianconquest (see the account of Evliya C,elebi,below, note 35). 28. Andrews 1953, p. 244, pl. XIII (top) in Appendix C. The original drawinghas no Italian title or legend. 29. Bory 1836, p. 135.

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

353

qui nous conduisit a une espece de place pratiquee dans un renfoncement des montagnes.... A gauche de cette place, sous l'escarpementd'un chemin montant, nous remarquamesun fontaine venetienne."30Earlier, Castellan had noted and illustrated this same fountain at the entrance of the village, hollowed with care under a rock, with benches and a vast reservoir; he imagined that its water, almost dried up, had been borne by the aqueduct.3" By the time of Gell's arrivalat Pylos, the Greek community appearsto have expandedbeyond the boundariesindicated on Venetian maps:a "small cluster of Greek houses" had been built at the skala (port) of Neokastro, near the fountain and the plateia of modern Pylos. His Narrative (p. 7) also speaks of a customshouse, perhaps identical to the "Lazareta"or quarantine house marked on a Venetian map.32Konomopoli's house, a new construction, was 50 yards distant from the landing, about a half-kilometer from the fort, and "situatedat the foot of a hill, sloping to the west of the port.... The windows of these rooms command, toward the west, a very pretty view of the port."33 Such an expansion of the suburb in the course of the 18th century may in part reflect a general increase in international commerce in the Morea at that time, an idea taken up again below. The shift does not appear to have occurredbefore the Ottoman reconquest of the Peloponnese. Venetian maps do not show structuresin the area of the modern harbor of Pylos or of its plateia. Nor can the existence of buildings there be deduced from the text of a full Ottoman survey completed in Januaryof 1716 and containing an apparentlyexhaustive inventory of properties in the suburb (Turkish varzs) and a list of its non-Muslim residents.34The picture presented by TT880 is in the main supportedby the account of Evliya 9elebi, who visited Navarino in the 17th century, before the Venetian conquest, and made several references to the community surrounding the fort of 30. Blouet 1831-1838, p. 2, and pl. 2, fig. I. 31. Castellan 1808, p. 84. The FrenchofficerMangeart also passed a fountainin 1828, before leaving Pylos along the main roadfor the French militarycamp,but he says that it was located at a distanceof one league (4 km) from Neokastro;this seems too distant to be the same fountain as that recordedby Castellanand Bory.See Mangeart 1830, pp. 354-355: "Aune lieue environa l'est de Navarin,on voit une sourcequ'on assureetre celle que Bacchusfit jaillir de la terreen la frappantde son thyrse.C'est a cette fontaine que commence l'aqu6ducqui transportaitses eaux dans la place. Les restes de cet hydragogue,construiten pierre,sont encore dignes de remarque. For the Frenchcamp, see

Duheaume 1833, p. 22; its location is indicatedon a map insertedin the front of Mangeart 1830. 32. This customshousemust have been, in any case,very nearthe present teloneion (customshouse)of Pylos. Bory (1836, p. 52) seems to confirm the location, since in 1829 he noted the presenceof wooden shackson the shore in the place where, opposite a wharf, the Turkshad their customshouse. 33. Leake,who visited Pylos in 1805, the year after Gell's tour,also stayedat the skalaof Neokastrowith Yioryios Oikonomopoulos,who "hasall the tradeof Neokastroin his hands, and is agent for some of the European nations.His house and magazines, which stand on the water side three or four hundredyardsbelow the fort, very naturallyexcite the cupidityof the poor Turksof the town, who are starvingby

the effects of their pride and idleness" (Leake 1830, p. 399, note 2). Admiral Codrington'smap, drawnat the time of the Battle of Navarino(1828), shows a clusterof buildings at the landing place,probablyincluding Oikonomopoulos'shouse (see Fokas 1927). Pouqueville(1826-1827, VI, p. 73, note 1) includes in the "cadastre du canton de Navarin"a place called "Calivia"with a populationof twenty heads of households.This entry may referto the houses by the landing place. 34. The document in question is registeredas an unpublishedmanuscript tax register(TT 880 [A.H. 1128/ A.D. 1716], pp. 78-100) in the BaUbakanlik Archives,Istanbul.In it, only males who paid the ispence, a head-taxlevied on non-Muslim adult men (see Inalclk 1994, p. xlviii), are recorded.

354

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Neokastro.35In 1716, all houses seem to have been near the gate of the fort, in approximatelythe areawhere the suburbappearson Venetianmaps.36 A century later, Gell noted the remains of this "miserableGreek village"at the gate of the fortress.37 A second shift in the structure of the settlement at Navarino, more radical than the first, occurred in the course of the Greek Revolution; a process was then set in motion that ultimately resulted in the formation of the modern town of Pylos. This stage in the development of Navarino is well documented, since there fortunately exists a rich collection of firsthand descriptions of the area in the later 1820s, in particular memoirs published by members of the French expeditionary force that occupied the Morea from 1828 to 1834 and the reports of the contemporaneous "Expedition scientifique."38 The fortress of Navarino fell without a shot in the fall of 1828.39The earliest French visitors to Navarino, however,found a community still torn by war, sundered by violence following years of brutality at the hands of Egyptian troops: Le lendemain de notre arrivee,nous descendimes a terre, ou m'attendait le plus affreux spectacle que j'aie vu de ma vie. Au milieu de quelques baraques de bois construites sur le rivage, en dehors de la ville, dont il ne restait que des ruines, circulaient, haves et deguenilles, des hommes, des femmes, des enfants, qui n'avaientplus rien d'humain dans les traits:les uns sans nez, d'autres sans oreilles, tous plus ou moins couverts de cicatrices;mais ce qui nous emut au dernier point, ce fut un petit enfant de quatre ou cinq ans que son

35. Folio 267b, 1-5: "Therearetwo gates . .. the other is on the landward

side and opens southeastward.This is the greatgate to the outer suburb"; Folio 267b, 30-268a, 15: "On the other side of the road that runs in front of the castle gate, in the gardenbehind the shop of Cerah Ali 9,elebi there is a coffee-tree,which producesbeyond measureeveryyear"and "Outsidethe castle to the southeast,on a wide, level plateau,there aretwo hundredhouses, all with gardenslike the gardensof Irem, two-storey masonrystructures roofed with tile. Most of these houses are Greek, and there are no Armenians or Jews.There is one neighborhood mosque, one inn for voyagersand fifteen shops for merchants,but no bath nor any other public benefaction. There is, however,one inhabited cloister of poor devotees of God that ought to be visited."We are gratefulto PierreA. MacKayfor providingus with his originaltranslationof Evliya'svisit

to Messenia (basedon the manuscript TopkapiSarayi.Bagdat K6okii308). The communityof reaya (peasants) recordedin TT 880 is notablysmaller than the 200 houses one might expect from Evliya'saccount;Evliya'sfigures, on the other hand, are often exaggerated and suspect (cf.Jameson,Runnels, and van Andel 1994, pp. 128-129; Faroqhi1999, pp. 160-161). 36. The expansionin the areaof the suburbtowardthe harborin the course of the 18th centurydid not resultin a significantincreasein population.TT 880 recordsthirty non-Muslim households.Pouqueville,a centurylater, listed 142 families (1826-1827, VI, p. 73, note 1), almost certainlyindividuals. In contrast,estimatesof the population of the fortressin the early 19th century varywidely.Pouqueville's600 Turks (1820-1821, V, p. 122) are difficult to reconcilewith Leake'sestimate (1830, p. 400, note 2) of 300 Muslim families. At the time of its captureby the French

in 1828 (on which see Duheaume 1833, pp. 17-28), the fort held 400 Arabs, 70 Turkishcannoneers,and 60 PeloponnesianTurks(see also Bessan 1833). Baltas (1997, p. 148) says that 500 armedmen and 234 women, children,and old people surrenderedin 1821 to the Greeks. 37. Gell 1823, p. 19. 38. For a generalhistory of French militaryinvolvementin the Morea in the earlyyearsof the modern Greek state, see Pellion 1855. The immediate purposeof this (successful)campaign was to compel the forces of Ali Pasha of Egypt, commandedby his son IbrahimPasha,to departfrom the Peloponnese.Driault (1930) describes Ibrahim'scampaigns(1824-1828) in detail and publishesrelevantFrench consularcorrespondence.For a summaryof Frenchmemoirs,see Simopoulos 1984, chapters20-21, 24. 39. Duheaume 1833, p. 28.

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

355

frere conduisait par la main; je m'approchai:il avait les yeux creves. Les Turcs et les egyptiens n'avaientepargne personne dans cette guerre.40 When the Egyptians departed, conditions improved rapidly.Commercial concerns were quickly reestablished and refugees returned. The French officer Mangeart described the village later in 1829: Quand je quittai Patras, on y comptait plus de soixante cafes;j'en remarquaiune quinzaine au plus 'aNavarin; dans un seul se trouvait un billard. I1y avait cependent quelques magasins assez bien fournis de choses les plus necessaires 'ala vie. La plupart appartenaient 'ades marchands francais. Ils etaient, comme 'aPatras,batis sur le bord de la mer, hors des murs de l'ancienne ville, et 'al'endroit meme ou etaient campees, six mois auparavant,les troupes d'Ibrahim.Je n'y remarquaipoint, comme 'aPatras, de ces nombreuses familles grecques, ou l'on comptait tant de jeunes femmes plus interessantes encore par leurs infortunes que par leur beaute. Le voisinage des iles loniennes, ou toutes ces familles romeliotes se trouvaient refugiees, les avait fait refluer 'aPatras,plutot que dans la partie meridionale du Peloponese.4'

40. Amaury-Duval 1885, pp. 80-81. On the dreadfulcondition of Navarino while still occupiedby Ibrahim, immediatelyfollowing the Battle of Navarinoin earlyFebruaryof 1828, see also firsthandaccountsby Miller 1828, p. 178, and by Duheaume 1833, p. 25, who visited the Egyptian camp in Septemberof that same year and found conditions deplorable.Not only the Greekshad suffered;with regardto Ibrahim'stroops he wrote:"Veritable spectresambulans,ils souffrentsans se plaindre." 41. Mangeart 1830, pp. 352-353. The map insertedin the front of Mangeart 1830 shows preciselywhere the Egyptian troops had camped at Navarino,as does Codrington'smap (see Fokas 1927). 42. Bory 1836, pp. 52-54, 132. 43. For the ravine,see Baltas 1997, p. 13, pl. 3. Other Greek families lived in cavesin a place where the road to Modon (modernMethoni) circumnavigated a cavernousravine.

Like Mangeart, Bory of the Expedition scientifique found the community at Navarino in recovery.42Having abandoned the town altogether a year earlier,what was left of the Greek population had returned to take refuge in plank shanties, or in caves and excavations in a ravine that descended to the plank wharf of the skala.43Other wood shacks had been built by foreigners on the shore where the Turkish customshouse had been located, opposite the wharf, and merchants were constructing a temporary village there. Artillery stores, a military hospital, a customshouse, and a police station were established in this new suburb.Before the onset of construction,Turkish and Greek cemeteries in the areahad been cleaned out; many corpses had been found exposed. The fortress of Neokastro lay in ruins but was quickly being repaired. Edgar Quinet found a flurryof activity there, by Greeks and French alike: Le 4 [mars],je touche terre entre dix et onze heures. Nous montons 'aNavarin par un chemin forme de roc et d'une terre argileuse. Le sol, du reste, nu et brule. Petite porte de village. Les maisons ruinees, les rues en decombres. Des soldats s'y sont forme des abris.Je les ai entendus chanter. Le bruit des tambours et de la musique guerriere au milieu de ces amas de pierres. La mosquee criblee de boulets sert de magasin de farines.Je porte mes lettres 'aM. Sau[ ]. Chambre d'un Turc. En le quittant, nous sortons de la ville. Spectacle extraordinairedes Grecs meles avec les soldats. Comment ils travaillent.Le glacis se releve. On leur a donne des tentes. Un grand nombre de familles se sont forme des abris avec les pierres des decombres. Ils etendent de gros morceaux de roc sur des peaux, des toiles, des draps qui leur servent de toit. Quelques-uns sont encore

356

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

refugies plus au haut de la montagne dans de petites grottes o'ul'eau degoutte. Beaucoup occupent d'ailleursdes baraques de bois ou ils vendent des dattes, du riz, des liquers, d'excellents citrons. Leur bon accord avec les Francais.44 Bory's account appears more dispassionate but conveys the same impression: dans l'enceinte . . . vivaient seulement, lorsque j'en partis, le commandant de la place, les intendans militaires, le payeur,la garnison, en un mot, ce qui tenait a l'armee;la population commercante, ou les habitants du pays, etaient descendus au rivage, et s'y construisaient une ville provisoire.45 In summary,it is clear that three distinct phases in the development of the settlement at Navarino can be defined. At the time of the Ottoman reconquest of the Morea in 1715, Greek and Muslim residents were concentrated at the southern end of the modern village, either in or just outside the main gate of the fortress of Neokastro. By the early 19th century,nonMuslim merchants had established themselves near the port, and this shift in the focus of the community might be seen, at least in part, as a reflection of the incorporation of Greece into larger Mediterranean economic systems in the 18th century.This was Gell's Navarino. After enormous disruptions in settlement in the course of the Greek Revolution, including total abandonmentof the town by its Christianpopulation,it was refounded under French sponsorship in 1829. The focus of this new community, renamed Pylos, remainedthe port. In contrast,the fortressof Navarino ceased to play a central role in the life of the town, as the new nation-state of Greece assigned it various nonmilitary duties.46

44. Aeschimann andTucoo-Chala 1984, p. 316. This volume contains Quinet's previouslypublishedand highly literaryLa Grecemoderne,as well as the more useful (for present purposes)and previouslyunpublished journal,quoted here, on which La Grece modernewas based. 45. Bory 1836, p. 52. Other accountsconcur,e.g., Anderson 1830, p. 121: "The Frenchgeneralfound its houses nothing but a heap of infectious ruins.We found the fortress still occupiedby the Frenchtroops, who were repairingsome of its batteries.The few Greeksin the place live in a village of about 100 temporary wooden houses, chiefly shops, northeast of the castle."Bory (1836, p. 133), however,says that the commandant

of the place and its paymasterlived in the village, at the foot of the citadel. 46. Visiting not long afterthe Revolution,the American consul, Perdicaris,found Neokastrocommandedby two philhellenes (not "nativeGreeks":Perdicaris1845, p. 195), but the fortresssoon lost its defensivefunctions.It was first transformedinto a high-security prison, although travelerscontinued to visit, e.g., Belle (1881). Ultimately, Neokastrobecame an archaeological park.GeneralMaison'sbarracksare in part a museum,and the keep of the castle,or IShisar,servesas headquartersfor the Superintendency of UnderwaterAntiquities of Greece.

SIR

WILLIAM

FROM

THE

KoURB

EH

GELL

S ITINERARY

FOUNTAIN

IN

THE

OF NAVARINO

357

PYLIA

TO THE

RIVER

The firststageof Gell'sjourneynorthfromthis town of Navarinobegan at the fountainin the Greeksuburbof Navarino(Fig. 2, station 1) and ended at the riverKourbeh(station11), a distanceof approximately five kilometers.In his daythe main roaddid not at firstfollow the courseof the modern asphalt highway that leads northeast from Pylos.47 Indeed,

afterclimbingthroughseveralswitchbackturnsfromthe moderncentral plateiaof Pylos,it is stillpossibleto see the bed of the olderroadway,now completelyovergrownwith weeds and scrub,downslopefromthe asphalt road.In the locationcalledMiden, the olderroadnearlyjoins the modern highway,but soon veerssharplyawayagaintowardthe sea. Next it descendedinto a smallcoastalplain(Gell'sstation2), labeledHaritouon the 1:50,000mapof Pylosproducedby the cartographicserviceof the Greek army.Betweenthe fountainand Haritou,the roadappearsto havebeen paved(i.e., a kalderimi): tracesof boththe pavementandthe retainingwall for the roadbedare still visible (Fig. 3).48In the smallplain of Haritou, Gell crosseda little riveranda stream,then ascendedthe hill labeledPsili Rahion the Greek1:50,000map,not farfromthe modernchapelof Ayios Vasileios,descendingagaininto the valleyof the XeriasRiver(the plain that he notes at station6). The XeriasRiverwas consistentlycalledthe PesiliRiverby travelersof the early19th century,a namethat mustderive from the TurkishtoponymBe?li,or "five-fold,"doubtlessa referenceto the manyrivuletsthat flow into the Bay of Navarinohere (Fig. 4). Gell specificallymentionsfour of these (his stations7 and 8); three modern bridgesacrossstreamsherewerenoted in 1997.49 The map preparedby the Expeditionscientifiquesupportsthe preceding interpretationof Gell's route:for example,it does not show the switchbackturnsthat aresucha prominentfeatureof the modernroadas

47. The approximatecourseof the road traveledby Gell is indicatedon a map included in the Atlas of the Expedition scientifiquede Moree, pls. 111.3and 111.5;see Fig. 5 and further discussionbelow. Other useful maps showing the road include one prepared by the British Admiral Codrington (see Fokas 1927) and the map insertedin the front of Mangeart 1830. See also a descriptionof the journeyby Perdicaris (1845, p. 208): "wecoasted along the shores of the magnificentbay to the north of the town, and in our way to the khan at the other end of it, we noticed some remnantsof the TurcoEgyptianfleet. . . right of search[for cannons,etc.] has been sold by the governmentto a companyof mer-

chants.... Beyond the khan we left the lagoon and the heights of ancient Pylos to our left."A small rectangleis drawn immediatelysouth of Yialovaon the Atlas of the Expedition scientifique,pl. 111.5;is this Perdicaris'skhan? 48. Castellan(1808, p. 90) explicitly states that the roadwas "constructedof large slabs of irregularform, and wellpreserved."Bory speaksof "Venetian paving"afterthe road crossesa small ravineby means of a bridge (1836, p. 136). 49. Bory (1836, p. 137) mentions only two streams,each crossedby a steep, single-archedstone bridge of Turko-Venetianstyle;see alsoAtlas, pl. 111.5.Be?li appearsas KaC4uog MTctXtAon modern 1:5000 maps of

the area,but the riveritself is called the Xerias.In TT 880, p. 89, Be*li is one of three mazra'asin this area:Rudiye (Gr. Po60c), Melis (Gr. Megart.), and Yufirior Be*li.The alternativename, Yufiri,must derivefrom Greek yeopvpoc/yscp6pt,"bridge."A mazra'a or mezra'ahere and elsewherein TT 880 is a "largefarmwith no permanent settlement,"perhaps"originallya desertedvillage or land reclaimedby a nearbyvillage"(see Inalcik 1994, p. xliv; also Encyclopedia oflslam, 2nd ed., 6, pp. 959-961). We were able to verify (1997) from inhabitantsof the nearby village of Pyla that Po8t6c and MeMoot lay adjacentto one another, south of the Xerias River,nearits mouth.

358

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Figure3. View of kalderimi near Miden

|~~~~

-

-~g*

hi i

_.,..,.,....

.......

*

.

.

..

'.

.

...

: ...K. .....ni Tavolakij. PylaJ

ruine towr" Thi toe

mus haebe

oae

ttepeetacao

Be*li

it descends into the Haritou plain (Fig. 5). The official account of the Expe6ditionscientifique can furthermore be employed to amplify Gell's narrative, since Bory describes his trip in great detail. In the first instance,

he reports that he climbed the slopes at the eastern end of the Bay of Navarino, following the course of an aqueduct that brought water to Pylos from the spring at Koubes, near the modern village of Handrinou; the

road crossed the aqueduct, then recrossed it near the ruins of an old tower about half a league (2 kin) from the town of Navarino (compare Gell's "1.a

Figure 4. View of

valley

SIR

WILLIAM

GELL

S ITINERARY

IN

Figure5. Excerptfrommapsof the areaof the Bayof Navarino.Atas of

THE

PYLIA

359

..

the Exp6ditionscientifique,pis. 111.3and 111.5

~~~laSr!

4.

p

logical site of Miden Vigla(tower"), on a bluff overlookingthe Bay of Navarino;todaythe locationis markedby a trigonometricmarkerand is indicatedon the 1:50,000 map.50It is from Vigla that Bory firstviewed the valleyof the PesiliRiver(whichhe also callsthe "Navarinitza"). Here he sawa Frenchmarinehospital,andfromherehe descendedpasta small ) with a nicheinto the plainof dryfountain(compareGell's"r.a fountain"

the Pesili,wherehe foundthe ruinsof a largehouse,similarto a French farm.51

Fromthe plainof Pesili,Gell firstviewsMt. Tabolachi(Tavolaki),or Mt. Pilaw,the prominentconicalhill, todaycalledProfitisIlias,that rises immediatelysouth of the moderntown of Handrinou;it is a prominent landmarkin the distantlandscape,which would not havebeen visibleto Gell beforehe descended(Fig. 6).2 He notices no villagesas he passes 50. Archaeologicalremainshave been observed.See McDonald and Hope Simpson 1961, p. 244, no. 63. 51. Castellan(1808, p. 90) also mentions springs("fontaines")"de distance en distance"at the edge of the road.The ruins,which Bory identifies as belonging to a ruinedvillage called Zonchio or Zunkio, may be those of the mazra'aof Rudiye (TT 880, p. 88). 52. Leake (1830, p. 398) describes Mt. Tavolakias follows:"We descend at

9:18 into an elevatedvale at the foot of Lykodhemoon its northernside. Tjaban is a quarterof a mile on the left, and two miles in the same direction,on the side of the mountain,Kambasi;the countryon the right is coveredwith oaks. At 9:48 we pass Sulinari,situated to the left in a retiredhollow at the foot of Mt. Lykodhemo.The three lastmentioned villages belong to the district of Mothoni. The valley terminates at the peakedhill, calledTavolaki,

which we leave on the left: on the right is the hill of Kondozoni."See also Leake 1830, pl. 5, where the prominence is mapped.On the map of the Expedition scientifique(Atlas),it is marked"Pic de Koumbes"(see also Fig. 5). Gell (1823, p. 4) remarkson Mt. Pilaw as a landmarkfor ships approachingthe harborof New Navarino.Pouqueville(1826-1827, VI, p. 26) gives its Turkishname as Pilaftepe, "rice-hill."

360

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

.EM~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. : . n..

...

..:.:.

..........

:

i:000:: *

~~~~~~~~~

t~~ ~~yla

u7... ......... ............

~

Lyko,ditnoS,` ~~~~~~~~~~~~~~~~~~~Tavolaki ... . ..C.

~

.............;

.

.....

. . ..

.

;

. . ..

. . .

..~~~~~~~~~~~~~~~~~~~

.. \........... .. .. . . . ......... .......

. .

... \..... ... ... ...................

through this valley,a factor that contributes to the overallpicture of desertion that he paints throughout his journey from Pylos to Gargaliani. But it is clear both from the Ottoman records and from the accounts of the Expedition scientifique that two villages (viz., Pyla and Zaimoglou, or Zaimogli) existed nearby,in the hills that border this valley immediately to the northeast, no more than two or three kilometers from Gell's route. Because of their proximityto Neokastro and Ibrahim'scamp at Yialova, both Pyla and Zaimoglou suffered greatly in the Greek Revolution. Bory says that the town and environs of Pyla were entirely deserted; he counted fifteen destroyed houses.53Zaimoglou was in the same condition. After departing Pyla, Bory writes: Nous laissames a droite Zaimogli, qui fut un joli hameau situe sur le contrefort forme par le confluent d'un ravin lateral avec le vallon de Pezili. Ses pentes portaient les verdoyantes traces de jardins abandonnes; on distinguait au milieu les ruines d'une maison d'importance, qu'on nous dit avoir ete celle d'un riche Turc.54 It is absolutelyclearfrom Bory'sdescriptionof his journeythrough Zaimogli that this village is to be identified with the present village of Balodimaika, apparently refounded, subsequent to its desertion, by members of the Balodimas clan, a family name that remains dominant today.55Such an identification is further supported by Blouet, another member of the Expedition scientifique, who visited the ruins of the village and recorded the distance in minutes between it and Pyla.56Finally, the location of the village (as "Zaimoglou")is fixed on the map published in the Atlas of the Expedition scientifique (Fig. 5). In contrast to its condition in 1829, Pyla in 1716 was populated. It was a fiftlik, had been registered for the state, and was a dependency of Anavarin (Navarino).Thirteen sharecroppers(ortakciyan)arerecordedwith their property, all but one of them bearing recognizably Greek names.57 The entry in TT 880 is helpful for interpreting Gell's text since, in com-

Figure6. View towardMt. Tavolaki fromthe Be?livalley

53. Bory 1836, p. 179. Puillon de Boblayeand Virlet (1833-1834, p. 85) reporta populationof twenty families at Pyla.Their data arebased in part on a census initiated in 1829 by Kapodistrias,the presidentof Greece. An averagefamily size of 4.75 is estimated.Frangakis-Syrettand Wagstaff (1992, pp. 439-441) discuss how the Expedition'sdatawere gatheredand evaluatetheir accuracy. This appearsto be only half its populationin 1815 (see Pouqueville 1826-1827, VI, p. 73). 54. Bory 1836, p. 191. 55. The village of Zaimoglou/ Zaimogli appearsalso in Venetian census recordsas a possession of Navarino,but has been confusedby Sauerwein(1969) with a village of the same name in the territoryof Koroni. In TT 880, the place is called Zaimzade (equivalentto TurkishZaimoglou). Puillon de Boblaye and Virlet (18331834, p. 85) list its populationas fifteen families, much reducedfrom its levels in 1815 (see Pouqueville1826-1827, VI, p. 73: 38 individuals). 56. Blouet 1831-1838, pp. 5-7. 57. TT880, p. 91.

SIR

58. Furthersupportfor this supposition is given by the three mazra'as recordedon TT 880 (above,note 49). These estates arelikely to have been cultivatedboth before and after 1716. Gell describesa churchof Ayios Nikolaos at station 10 (see Fig. 7), apparentlythe same one mentioned by Pouqueville(1826-1827, VI, p. 26). 59. TT 880, p. 97, lists a fUftlikof Tup&inimmediatelyafter KurdBey. 60. See Andrews 1953, pl. VII; Panayiotopoulos1987, pp. 226,262. 61. Locatelli 1691, pp. 218, 222. It may be relevantthat Turkishyali can mean "mansionby the shore,"although yalk plus ova should mean "coastal plain,"an apt descriptionof the vicinity. Although the place-nameYialovais Slavicin appearance,the fact that it is apparentlynot attestedpriorto the 19th century,as well as its accentuation (Yialova,not Yial6va),is perhapsbest explainedif it is a relictTurkishname for this location. 62. Molin 1900, p. 438. Venetian sourcesalso mention a Turkishraid on the shore of the village of Curbeiin 1698 (Siriol Davies, pers. comm.). 63. TT 880, p. 97. Dimensions are given in arpun= 0.758 m (Inalclk1983, p. 340). For purposesof taxationthis fiftlikwas recordedwith the mazra'aof Yufiri,presumablythat which elsewhere is called the mazra'aof Yufirior BeUli(see above,note 49). 64. Bory 1836, p. 138. 65. Castellanis similarlysilent. 66. Bory 1836, p. 138. We have not yet been able to locate a copy of Smyth'smap. 67. Bory 1836, pp. 138-139.

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

36I

mon with most entries, it lists the boundaries of the f~ftlik.Two of these were Be*li and Ustna Nikula [ston Ayio Nikola]. From this information, it seems that at least some of the agricultural activities that Gell noticed both in the valley of the Pesili River and at Yialova, near station 10 of his Itinerary,were based in this village.58The rice production he reports is not, however, recorded in the Ottoman document and was perhaps introduced to the Navarino area later than 1716. At station 9, Gell states that hills approached the Bay of Navarino from the right. Here, he must be near the ridge labeled Kanonia on the modern 1:50,000 map, Tupcin in the Ottoman document.59There had once been a substantial settlement in the areabetween the Pesili River and the river north of modern Yialova. A Venetian map, dating prior to the Ottoman reconquestof 1715, depicts a "villacorbei"midway between these rivers,while Venetian censuses of 1689 and 1700 give populations for the settlement of 68 and 54, respectively.60Locatelli, in describing the Venetian conquest of Navarino in 1686, referred to a "Palazzo de Cipressi di Curbei"and to the "Villa Cucurnara,poco distante dal Palazzo del sudetto Curbei,"suggesting that, prior to the Venetian occupation, there had been a mansion of a prominent individual there.6'A report in 1692 to the Most Serene Republic by Tadio Gradenigo recommended "il luoco di Curbei" as a suitable place for a factory to make biscuits for the fleet because a stream capable of driving grain mills had its outlet into the bay at this point.62

But with the reconquest of the Morea by the Ottomans, the community appearsto have been deserted. TT880 records an unpopulated fftlik called Kurd Bey, certainly the same settlement.63Abandoned structures included: 3 attached lower rooms, ceiling in ruin, walls standing: L. 35 x W. 12. 7 attached lower rooms, ceiling in ruin, walls standing: L. 50 x W. 15. 8 attached lower rooms: L. 52 x W. 11. 1 lower room, ceiling in ruin, walls standing: L. 11 x W. 7. 1 lower room, ceiling in ruin, walls standing: L. 9 x W. 6. 1 lower room, ceiling in ruin, walls standing: L. 15 x W. 9. Two water mills were in ruin, perhaps the remains of the Venetian factory. A century later Bory noticed the ruins of a destroyed mill and an associated canal at Ayios Spyridon along the road north of Yialova.64 At the location of the modern village of Yialova, Gell mentions no houses at all, and it is not clear that this place was settled between 1716 and the early 19th century.65Near the mouth of the nearby river,Bory saw traces of walls that he wanted to identify with a village of Yialova that he presumed to have existed there prior to the Greek War of Independence. But in the actual location marked Yialova on Captain Smyth's map of 1823, preparedfor the British Admiralty, Bory found only huts of fishermen.66The troops of Ibrahim had camped at Yialova and were replacedby the bivouacs of the French expeditionary force.67It seems indisputable

362

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

that the river at Yialova, now the Gouvalari River (Fig. 2), is that called Kourbehby Gell. Bory described its mouth soon after he arrivedin Yialova as 15 feet wide and needing to be forded. (Apparently the bridge that had existed in Gell's day" was a casualty of war.) While Bory himself is not confident that this is the river Kourbeh and imagines some confusion in earlier travelers'accounts between Kourbeh and Khombey/Koumbes (the spring near Handrinou that feeds the aqueduct to Navarino), his doubts must be misplaced.69Captain Smyth labels the riverat Yialova as Kourbeh, and Leake also names it thus on his map of Messenia.70 FROM

THE

RIVER

KOURBEH

TO THE

ROMANOU

RIVER

In the second segment of his journey, Gell for the first time remarks on evidence of settlement, although even then, as in his total journey of five and a quarterhours, he reports that he did not encounter a single person. This section of his account is laden with interpretativeproblems, although the general course of his route is itself completely clear. On reaching station 13, Gell must have been in the place today occupied by the hamlet of Yiannoupoulaika, on the outskirts of modern Yialova. This hamlet, like Balodimaika, appears to have been established after the Greek Revolution,7' probably by a single extended family that gave the settlement its name. From here Gell saw, in the Itinerary (p. 51), a "white house" under the "rook"(archaic,"rock")of Navarino and, to the left of it, the village of Petrachorio.The account in the Narrative (p. 61), however, differs, as he speaks of "leavingthe little villages of Petrachorio and Leuka on the left." The discrepancies can be satisfactorilyresolved. In the first place, the location of Petrachorio, modern Petrohori, is clear (Fig. 7). The place is also mentioned in TT 880, although it was uninhabited and therefore recorded as a mazra'a,and said to be cultivated (see below). By the time by the reaya(i.e., "peasants")of Hasan Aga jftlik f of the Greek Revolution, a permanent settlement was established at Petrohori, and Bory notes that before the Egyptian invasion this village contained fifteen households.72 Gell's Leuka is more problematic and one might be tempted to speculate that, in the course of revising his notes for publication in the Narrative, he had in confusion transformed the "white house" of his Itinerary into a town. But it seems more probable that the opposite transformation occurred.What are his guides likely to have told him? If they described to him a village called Leuka, did he at first, because of his knowledge of 68. Gell 1817, p. 51. 69. Bory 1836, pp. 138-139. The Frenchconsistentlycall the areaaround modern Handrinou Koumbey/ Koumbes(Fig. 5); this same areais called KurtBey by Perdicaris(1845), who traveledto the areanot much later. 70. Leake 1830, pl. 5. 71. The Atlas of the Expedition scientifiqueplaces a settlement labeled

here (Fig. 5). "Gouvalovoros" Its populationis given as eight families (Puillon de Boblayeand Virlet 18331834, p. 85), but it is not mentioned by either Bory or Gell. A slightly earliercensus compiled in February of 1830 by K. Ramfos,the Greek temporaryadministratorof the fortress of Navarino,recordsa total population of 49; that document spells the name in Februaryand as "Vouvoulogoroi"

in October,1830 as "Gouvalovora" (Loukatos 1984, pp. 211-212, note 1, and p. 219, note 1). 72. Bory 1836, p. 159; see also Puillon de Boblaye and Virlet 18331834, p. 85, where its populationis given as 15 families.Loukatos (1984, p. 219, note 1) lists 56 individuals. Pouquevillereported26 individualsin 1815 (1826-1827, VI, p. 73, note 1).

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

363

PYLIA

~~~~~~~~~~~~~~~~~~~~~~~~~~ i~~~~~~~~~~~~~~~~~~~~~~~~~ ss~~~~~~~~~~~~~~~~~~~~~~~~ ...........

s*:: i3. ,: ................ ;s

sI

E

:i

,#v

.''

s

:.

.,

,.

. ...... ...

..

ava<'<.c N-. '. rino a

....

..

, ............ m '.S ,.'. . ............. ..S. ................ ... .....

....

:...........

:

::

.:

:::i

......

?x,

r

''

..... ... ..... ............ "

.

1."''

...... ...... .............. z ..... ....d

"

:

[V

okoilija|

............ . ............. . !;.~ '

'.

'.

''."

Wt

okoiliA .Void

'

'.''

''

''''S'lg~~~~~~~~~~~~~~~~~~~~ ,.R,

,.

Petrohori

...

....~~~~~~~~~~~~~~

.... , g, ..........

g~~~~~~~~~~~~~~~~~~~~

....... . ........

s/s2- 1 .. .......... .3' s.l-,.''

.

.

...~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....

_sWr,'t.W.ist,., /... ,.' ..... ...

e.g.P

p||li

Figure7. View of the areanorthof the Bayof Navarino,including Petrohori,fromnearElaiofyto

73. It is obvious that Gel employed accountsof observationsby others in addition to his own when composing his publishedworks:see Woodward and Austin 1925-1926, p. 69. 74. Bory 1836, pp. 158-159. He claims that there were 15 hearthshere before the expeditionof Ibrahim. Pouqueville(1826-1827, VI, p. 73, note 1) reported17 individuals. 75. TT 880, p. 97. The toponym Tavernais listed in Georgacasand McDonald 1967, no. 7659.108, in the

...

A4F - j8_.,=__

.....

.........

classicalGreek, imagine that they were referringto a white building rather than a village? Did he then realize and correct this mistake when writing the Narrative?73It seems in fact that such a place did exist, although the toponym has today vanished from this particulararea. Bory also speaks of a place called Leukos or Lesko. Only an oven remained there in his day; the village had first been destroyed by an earthquake,then devastated by Ibrahim.74TT 880, moreover,provides indisputable proof of the accuracy of Gell's Narrative. It records a fiftlik of Lefku orTavarne that consisted in 1716 only of fields and a damaged tower; the place was uninhabited and was farmed by reayaresident at Osman Aga (see below).75Furthermore, the document also lists a mazra'aof Rum Bag or Lefku. It is clear that this Lefku is near Petrohori because the yields of the two properties are to be counted together.76A final confirmation comes from the map reproduced in the Atlas published by the Expedition scientifique. On it, a Leukos is located immediately north of the Osmanaga Lagoon (Fig. 5), halfway between the modern asphalt road and Petrohori, in the arealabeled Barakou on the modern Greek 1:50,000 map (Fig. 7).77 For Gell, looking northwest from Yiannoupoulaika, Leuka/Leukos would indeed have lain to the right of Petrohori.78

vicinity of the modern village of Koryfasio.Although not indicated as Tavernaon any contemporarymap,we determinedfrom interviewinglocal inhabitantsthat the name is appliedto an areanearthe churchof Ayios Nikolaos that lies on the modern Hora-Pylos road,south of a gasoline station (currentlyBritish Petroleum). On the modern 1:50,000 map a placename Z3xcpv(a appearsat this location, probablya version or corruptionof the same name.

76. TT880, p. 81, and p. 362 above. 77. Curiously,Leukos is omitted from the unpublishedcopy of this map producedby General Pelet for George Finlay,now in the libraryof the British School at Athens. 78. Gell (1823, p. 28) also mentions this place on the occasion of his visit to Old Navarino.After commenting on VoidokoiliaBay he remarks:"Ruins probablyexist on the hills near the villages of Petrachorioand Leukos, on the north-east."

364

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Lykodimos Osman Aga

Pisaski

... . . ......

A

Gell's observations from station 14 are equally perplexing. The location of Voidokoilia is indisputable. He sees this small coastal inlet to his left, perhaps because he has gained a bit of elevation above the alluvial plain north of the Osmanaga Lagoon. But what are Geophyre/Gephyrae and Lisaki? Gell himself appearsconfused. In the Itinerary(p. 51), he only notes Geophyre at station 14, and wonders if it is a hut, a place, or a bridge, while in the Narrative (p. 61), he simply says that it and Lisaki lie right of the road on small knolls. In the Itinerary (p. 52), he mentions "a ... hill called Lirachi" (perhaps the same name as Lisaki), but at station 17, six minutes later,afterhe has passed Osman Aga. It seems most plausible that Gell has conflated the two observationsin his Narrative.Neither Geophyre nor Lisaki/Liraki is mentioned by the Expedition scientifique, but both appearin TT880 in contexts that suggest they were located near modern Koryfasio.Gephyrae, visible to Gell from station 14, should be identified with the mazra'aof Yufiri (also known as Rum Baglari); Lisaki probably hides behind the name Azake that appears on TT 880, and should be identified with the areacurrentlyknown as Lezaki (perhaps through reinterpretation as "El Azaki"?).On the modern Greek 1:50,000 map, Lezaki lies above and to the east of modern Koryfasio, in an elevated location. It may only have been visible to Gell from a point farther north than station 14, perhaps only from station 17, where it is mentioned in the Itinerary.79 Descriptions at station 15 are much less opaque. The eminence that approaches the road is almost certainly the southern end of the ridge now called Beylerbey, after which there does open to the right of the road, in

.. ......

Figure8. View of OsmanAga and PisaskifromnearTragana

79. Furtherconfirmationof the location of Azake/Lezaki is the fact that it is one of the toponyms that define the boundaryof the sfitlik of Osman Aga (TT 880, p. 85), while Azake itself is describedas bounded by KiiuiikPisaski and Osman Aga (TT 880, p. 82). Pouqueville(1826-1827, VI, p. 26) also refersto a Mt. Lyrakiin the vicinity of Osman Aga: "Unevalle marecageuse,et le mont Lyrakiservent a signalercet espace,qui est baigne par une riviereappelee Romanos."It is possible,however,that the mountain referredto as "Lyraki"is distinct from the settlement Lezaki, and is perhaps an alternativename for one of the mountainsknown as Lykodimos(cf. Fig. 8) and Maglavas,each of which dominatesthe region, depending on one's point of view.

WILLIAM

SIR

S ITINERARY

GELL

IN

365

PYLIA

THE

Hasan Aga

/

_S

.. ;^ -.^ , | - i i i lsi l | ......... _

| .g=N:ar

. ..

_ __ rze[s

_>_

..._ . _z_z_

;;_

j

* _X____

_ w]_^ | s

_ *

*

_

_

l

_L_l

-|

_Ai

rs

-

|

1

141

|

6

iW^

-

1

-

i

_ |-I-_

*

*

| -

_

.

l

-

*

s

|

.

|

l

*

-| _

.

* | _E

_ -l - 1_ :.

l

|

-

l

| l *

|

|

I 81_

_,

_

__r_

J

w -aa lI_ | | -l_ P

E

E

l

_

E E __

|l l

*

_:;

ZiiJ

*

|

P I __

| I_

the direction of modern Koryfasio, an extensive alluvial plain. From here Gell would have gained his first view (Fig. 8) of the village of Koryfasio (Osman Aga) and, to the northwest of it, on the opposite side of the Romanou River (the modern Selas), a glimpse of the prominent hillock of Hasan Aga (his Haslan Aga). A few minutes later,but before reaching the river,he notices a plane tree, an aqueduct, and then a well, a house, and the hill of"Lirachi." It is clear that the settlements of Osman Aga and Hasan Aga were well populated in the early 18th century and that they remained so in Gell's day. Gell therefore barely missed reaching two significant population centers, whose residents engaged in extensive agriculturalactivities in the areasthrough which he passed.Archaeological investigationsby PRAP also provide insights into the nature of the settlement at Hasan Aga in the Second Ottoman period. HASAN

80. Davis et al. 1997, pp. 481-482; Alcock 1998b.

| | l __

I

* _ | SI l _

....

Figure9. View of HasanAga from the KatoEnglianosridge

l_

I I __

___

B B

|-1 w _ -

-

W

.

* l * I _11 * I| | _

|

|

_1

__

*-

i

||| _- I1__ 1_ | - 1 | | |

l

1

11* I-|

* |

|

lw

|

__ _

_

|

_*

|

|

l

S

|

|

_.| |-I

!_ ! -

_ * - | | - l-I m _ l s_ | |*l l | _i_ |_ *| || _-* I

l

|

|

_ _* _l

|

|_ |-1 ,

_ ||

*

,

u

*

I

_

Z

|

|

N

ACA

Members of PRAP first became aware of Hasan Aga in the course of fieldwalking in the area (Fig. 9). The toponym, which appears on neither the modern 1:5000 nor 1:50,000 maps of the area,was supplied to us by a local resident. Subsequently, the site was examined in detail by PRAP, its architecturaland other remainsplotted, and a collection of artifactsmade.80 One of the most striking features of the site today is a house with a twostory tower in one corner. Although probably a 20th-century construction, it may have imitated earlier structureson the site; TT 880 refers to a

366

BENNET,

AND

ZARINEBAF-SHAHR

Figure10. BridgenearRomanou

|

-|g;

L

DAVIS,

.:>'.*,._

^

_

5 ,

tAl'

.

V~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~W Ea

i

t9Ee:3

tower and Gell in his Narrative (pp. 61-62; above, p. 349) also alludes to one. In 1716 nine sharecropperslived there; slightly earlier Venetian censuses gave figures of 12 men and a total population of 37 (1689), and 2 families and a total population of 7 (1700). Pouqueville in 1815 reported a population of 28 individuals. In 1829, the Expedition scientifique listed only 5 families at Hasan-Aga (sic), while Ramfos lists 42 individuals.8 OSMAN ACA The two "villages,or tchifliks"that Gell describes (Itinerary)would have been the most substantial remains visible from the low-lying coastal area of station 15. In the Narrative, he appearsto distinguish Hasan Aga, which he refers to as a "village,"from Osman Aga, which he describes as a "country-house."The name of this village was officially changed from Osmanaga to Koryfasio in May 1915,82one of the earliest in a series of renamings designed to purge the region of non-Greek (especially Turkish) village names. The name Osmanaga continues to be applied to the lagoon north of the Bay of Navarino, however, perhaps some indication of the importance in the region either of this individual or the village.83In 1716, the settlement was designated as a fiftlik and was called either Osman Aga or Buyuk ["Greater"]Pisaski, to distinguish it from nearby Ku,cuk["Lesser"] Pisaski.The Venetians employed a similar distinction, referringto the two villages as Pisaschipiccolo and grande, although Suman Ag'a is an alternative name for the latter.84 The relativeimportance of the two settlements seems not to have been one of size, since in 1716, Kui,iik Pisaski had thirteen sharecroppersas opposed to twelve at Buyuk. However, TT880 suggests a higher status for Osman Aga since it had a "big house" or saray ("palace")and a han, in addition to ten houses belonging to the f ftlik.85In the case of Ku,cukPisaski, only three "lower rooms" are mentioned as property of the fiftlik, but an annotation indicates that each of the sharecroppershad a house. Kiiyik Pisaski, therefore, seems more like a village than an estate. Moreover, in addition to its grand buildings, Buyuk Pisaski/Osman Aga also boasted the largest number of olives in the kaza (district) of Anavarin [Navarino]

81. TT 880, p. 80; Panayiotopoulos 1987, pp. 226,262; Pouqueville18261827, VI, p. 73, note 1; Puillon de Boblayeand Virlet 1833-1834, p. 85; Loukatos 1984, p. 219, note 1. 82. Politis 1915, p. 281. 83. Bory (1836, p. 140) rather puzzlingly notes that the name Osmanagais only correctlyappliedto the lagoon. In the Expedition scientifiquecensus, Osmanagaappears as Souman-aga(Puillon de Boblaye and Virlet 1833-1834, p. 85); in the census of 1815 it appearsas Osman-aga (Pouqueville1826-1827, VI, p. 73, note 1); and in 1830 as Sumanaga(Loukatos 1984, pp.211-212). 84. For the Venetiancensusesof 1689 and 1700 that relateto the "territorio" of Navarino,see Panayiotopoulos1987, pp.226 [1689], 262 [1700]. Note that Gell was wrong in supposingthat Osman Aga, the man, was his contemporary.Pouqueville (1826-1827, VI, p. 26) similarlynotes the "villagesou tchiftlicks qui portent les noms d'Osman et de Hassan, proprietairesauxquelstoute cette contree appartient." 85. TT 880, pp. 84-85. Compare Gell's referenceto a "country-house."

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

367

(1903 roots), the largest area of vineyards (baglari,"gardens";300 d5nuim, approximately27.5 ha), and 1500 do5num(approximately 137.5 ha) of arable land atTavarne and the nearbydeserted f ftlik of Lefku.86Many mulberry saplings (1500) are also recorded, presumably to nourish the silkworms that produced silk for the workshop (ipekhane),another feature unique in Anavarin to Osman Aga. Since Pouqueville's figures suggest that a situation similar to that described in 1716 still obtained at the time of Gell's trip, it is surprising that even in January Gell noticed no one working the land in the vicinity of either Osman Aga or Hasan Aga.87 Remains of the bridge with which this stage of Gell's itinerary ends are visible beneath the modern concrete bridge on the Selas River that is crossed by the modern asphalt road from Romanou to Tragana (Fig. 10). This bridge seems to have given its name to the mazra'arecorded in TT 880 that is explicitly called "the other Yufiri"to distinguish it from the Yufiri near Beli.88 Although the modern village of Romanou appearsnot to have been inhabited prior to the mid-19th century,the name Romanou was applied to the river,and the presence in TT880 of the toponym Rum Bag/Baglari ("Greek Garden[s]") must reflect this. FROM

THE

ROMANOU

RIVER

TO GARGALIANI

In his description of the third part of his journey, Gell gives far fewer local toponyms, although it is still possible to trace his route. It is clear that he did not follow either of the principal modern roads to Gargaliani, either the route that runs along the coast and approachesGargalianifrom Pigadia or the upperroute through Ambelofyto (formerlyAgorelitsa).89After crossing the bridge over the Romanou River (Fig. 10), Gell seems to have climbed to a point near the modern village of Tragana (what he calls "the top of the opposite bank"), and then to have descended into the upper reachesof the present Mati rema (which he variouslycalls Brisomero Nerro [Itinerary,p. 52] or Brusomavo [Narrative, p. 62]).90Because Gell makes no mention of the village of Mouzousta, he is likely to have followed a poor switchback trail that makes its way up a gorge on the precipitous southern flank of the hill of Profitis Ilias (Lefkis), and then crosses the area of Kalantina before finally arriving at the southern outskirts of the town of Gargaliani (Fig. 11).91 86. See above,p. 363 and note 75. 87. In 1815 Kiuck and Buiyuik Pisaskihad 22 and 29 individuals, respectively(Pouqueville1826-1827, VI, p. 73, note 1); and in 1829, 9 and 7 families (Puillon de Boblaye and Virlet 1833-1834, p. 85). TT880, pp. 80-81, also lists a f iftlik of Rustem Aga, some of whose arableland was located in the vicinity of Osman Aga fiftlik. 88. See above,note 49. 89. These routes areindicatedin the Atlas of the Expedition scientifique,pls. 111.3and 111.5(see Fig. 5). The track followed by Gell is not.

90. Bory (1836, p. 165) also in the mentions a "Brisomero-nero" courseof his journeyto Mouzousta; there were ancient remainsthere and the land was workedby farmersfrom Mouzousta (modern Lefki). Pouqueville too (1826-1827, VI, p. 26) refers to the "Vrysso-Nero."It seems clear from context that none of these authors refersto the sulphurousspringsat modernVromoneri.See Lyritzis 19821983, p. 123, concerningthe application of the toponym Vrysomylosto the streamfed from the springsat Mati, and to its vicinity.The village of

Traganawas founded after the Greek Revolution;it is not mentioned by the Expedition scientifique. 91. Pouqueville(1826-1827, VI, p. 26) pursuedpreciselythe same route in reverse,mentioning the chapel of Ayios Nikolaos on the outskirtsof Gargaliani, a semicircularvalley,a steep descent, and an aqueductfed by the spring of "Vrysso-Nero."The Expedition scientifique,on the other hand, traveled from Navarinoto Gargalianivia Mouzousta (Bory 1836, p. 167); the churchthere had been destroyed.

368

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Proftis

.S..

Ilas

II

__~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Gell's description closely fits this geography. The summit that he reaches, in the Itinerary"aftera steep and dangerous descent [sic],"92 must be the crest at the top of the switchback trail, a place that does, indeed, afford spectacularviews of the sea and the island of Proti. A distinct natural hollow between Profitis Ilias and the area of Mylolitha slopes west toward the sea, and behind it lies the semicircularvalley of Kalantina. A church of Ayios Nikolaos, now ruined, once lay at the edge of town.93 Had Gell taken the coastal, lowland route to Gargaliani, he would again have avoided signs of habitation and witnessed a similarly depopulated landscape.This is, in fact, precisely the impression conveyed by another travelermaking the journey along the coastal route a decade before Gell. A. L. Castellan traveled from Petrohori to Filiatra in 1799. He did not visit Gargaliani,but stayed on the coastal plain all the way.Thus, after crossing various rivers north of the Osmanaga Lagoon, he found himself on higher ground amidst an extensive olive forest: Ces oliviers, dont on ne parait pas recolter le fruit, etant livres entierement a la nature, sont devenus de grands arbres.Cet endroit se nomme les Cent-Villages,quoiqu'il n'y existe pas en ce moment une seule cabane. Au-dela du bois se trouve un terrain sterile, couvert de bruyeres,et peu apres le chemin se dirige vers une foret que le Zantiote nous a signalee comme la retraite des forbans de l'ile de Prodano.

Figure11. Switchbackroadleading fromthe Mati valleyto Kalantina

92. Cf. Gell 1823, p. 62: "We ascendedfrom hence by a steep and difficulttrackto a summit"(our emphasis). 93. Bory saw its ruins (1836, p. 169). It is clearfrom his description that the chapel lay nearthe hill of Hondrovouni,on the outskirtsof Gargalianiin the directionof Mouzousta. Its existence is attested alreadyin 1698 (Dokos 1976, p. 133).

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

369

There follows an account of making camp for the night and of an alleged attack by pirates. Castellan then continues: Le 29, a la pointe du jour, nous nous sommes hates de sortir de cette foret malencontreuse;ce que nous avons fait sans obstacle. Ayant jete les yeux en arrierenous nous sommes convaincus que toute cette c6te, en effet voisine de l'ile des Pirates, etait absolument deserte et inculte. Devant nous etait un pays plus riant, et a mesure que nous avancions nous avons retrouve avec plaisir des champs cultives et ensemences pour la seconde fois. Les collines etaient tapissees de vignes, et la fumee qui s'elevait catet la parmi des bouquets d'arbresindiquait des habitations. Ce spectacle ramena parmi nous la tranquillite, et nous laissames les craintes et les dangers derriere nous, dans le brouillard qui couvrait la foret.94 The party ultimately reached Filiatra in safety.

COASTAL SETTLEMENT IN THE SECOND OTTOMAN PERIOD AND THE ACCOUNTS OF TRAVELERS

94. Castellan 1808, pp. 91, 94. 95. On the devastationat this time and the placementof Greek and Egyptian forces,see Humphreysand Pecchio 1826. On battles for the fortressof Navarino,see Makriyiannis 1966. Makriyiannis(chapter3) describesthe large Greek force assembledat Hores (the areaof modern Hora); on the existenceof the name Hores before the name Hora officially replacedLigoudista,see also Bory 1836, p. 229, and the Atlas of the Expedition scientifique,pls. 111.3and "'I.5. 96. Gell's accountcan also be contrastedwith that of Perdicaris,who visited the areasoon after the Greek Revolution.Perdicaris(1845, p. 209) wrote:"Fromancient Pylos to the city of Arcadia,a distanceof about twentyfour miles, the country,though generallyin a state of neglect, is, in comparisonwith the desertedregions between Petalidiand Navarino,well cultivated,and prettywell inhabited."

Narratives like those of Gell paint a rather bleak picture of coastal settlement in the last century of Ottoman rule. Many scholars have concluded from such accounts that, prior to 1821, the Greeks had fled to the highlands, leaving the lowlands in waste. By the time the French expeditionary force arrivedin the Morea, such a picture is indeed accurate.Ibrahim Pasha and his troops had cut swathes of destruction through the Peloponnese. Much farmlandhad been intentionally destroyed and villages pillaged and burned.The areaof the Bay of Navarino was especially hard-hit.95During the sieges of Neokastro and of Sphakteria (1825), Ibrahim had camped at Yialova. Later he occupied Navarino itself (1825-1828). During the campaign of Papaflessas (1825), additional damage was inflicted on the district. But the lowlands around the Bay of Navarino do not appear to have been severely depopulated prior to the Greek Revolution. It is clear that the vision of a travelerlike Gell must have been restricted both physically and conceptually in order for him to have failed to comment more extensively on the considerable settlement and land use there, and that his descriptions must be counterbalanced by a review of newly available documentary and archaeological evidence. Although his route kept first to the coastal lowlands and then avoided the village of Mouzousta on the way to Gargaliani, it is inconceivable that Gell would have been unaware of the substantial human activities in the areas through which he traveled, had he cared to record them.96 It seems clear from archivalsources that the land surroundingthe Bay of Navarino was cultivated by Greek sharecroppersat the beginning of the Second Ottoman period and that the villages they occupied were not far

370

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

from the lands they farmed (Figs. 12, 14). Nor is it probable that by Gell's day the situation had changed to any appreciableextent. In fact, the settlement pattern in the district of Navarino, in terms of villages or fiftliks recorded, is quite stable- from the Venetian censuses of 1689 and 1700 (Fig. 13), through TT880 in 1716, through Pouqueville's"cadastreof 1815," to the Expedition scientifique's census of 1829. Only three populated villages included in Venetian censuses were deserted in 1716 (TT880: Other Papla [modern Kato Papoulia],KurdBey, and Elyas Aga [modern Lezaga]). Only Kurd Bey, now known as Yialova, had not been resettled by Pouqueville's1815 "cadastre,"although it was populated by 1829, the time of the census of the Expedition scientifique. On the basis of Pouqueville's cizye figures, the population of the district of Navarino appears to have almost doubled between 1716 and 1815 (cf. Figs. 14 and 15, and see Fig. 16 for population in 1829).97The impression of desertion in the Second Ottoman period conveyed by travelers'accounts like Gell's is thus illusory. Viewed in this light, Gell's narrativeseems less a straightforwarddescription of what he saw than a construction that reflects his own agenda in composing it.98It seems likely, for example, that Gell was influenced by earlier travelers'accounts, such as that of Castelian, who charted a similarly deserted landscape. Susan Sutton has demonstrated how formulaic travelers'accountscan become: at Nemea, desertionand isolation arethemes maintained consistently in narrativesof the 19th century in spite of the fact that documentaryand archaeologicalevidence demonstrates that there was dense habitation and land use.99Simopoulos has discussed Gell's general distaste for ordinary farmers,and his preference for the Greek Europeanized elite, the so-called xo-u4ocVTc6ce;ag.Gell may also have been influenced in part by the toposof desertion that had already been established by ancient authors. It is important to remember that, despite his frequent reference to the mundane (for which he was sometimes mocked he was a graduate of Cambridge University, a scholar enby reviewers),100 gaged in debates over the Homeric question, and a leader in the Society of Dilettanti.101Pausanias and Strabo served as his guides, as they did for other scholars of his day, and no doubt directed his vision away from regions and subjects that played no substantial role in their texts.102 Gell himself describes the land around Navarino as "a country where no one goes, because it affords no object of curiosity."103 97. Pouqueville1826-1827, VI, p. 73, note 1; Puillon de Boblayeand Virlet 1833-1834, p. 85. Pouqueville's (1820-1821, V, p. 15) figure for cizye billets of 447 is almost twice that for ispencelisted in TT 880: 269. 98. Other aspectsof his texts suggest that Gell at times consciously chose to omit significantinformation, perhapsin an attempt to defamiliarize his narrative.For example,although he was accompaniedby English servants and a western companion(1823, p. 366), they are rarelymentioned, and never named.

99. Sutton, in press. 100. Simopoulos 1985, p. 124. 101. Gell publishedsignificant studies of both Troy (1804) and Ithaka (1807). 102. Indeed, Gell himself published a geographicalcommentaryto Pausaniasand Strabo:Gell 1810. For Pausanias,there was not "muchto 'see' in the region,thanks to the yearslost to Spartandomination"(Alcock 1998a, p. 191). 103. Gefl 1823, p. 32. He searched for antiquitiesonly in the fortressof Old Navarino,but rejectsit as a likely

location for Nestor'sPylos (pp. 26-28). Pouqueville(1826-1827, VI, pp. 2627) also paints a pictureof scant cultivationand restrictedgrazing:"Le pays,parsemede groupesd'oliviers,ne presentea de grandesdistancesque peu de terrainscultives,entremelesde parcoursdans lesquelspassent une multitudede moutons et de chevres,car les grassesg6nisses ainsi que les taureauxqui faissaientl'opulencedu roi de Pylos, pasteurdes hommes, n'existentdepuis long-temps que dans lIliade et I'Odysseed'Homere."

WILLIAM

SIR

GELL

S ITINERARY

IN

THE

PYLIA

k ._ ...

......: . \:

\: 1'

.:... ..:.

. : !.

.. .

... .

.

.. ...b . .... . . . .. . ... !::: .:

:.

R l;,

.

8

s

g

S

'

.

'

R

.

P

ft~~~~~~~~~~~~~~~~~~~~~~ .. f. R':}} i-.....

.S

1.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .

..... ...... ... between Setilements the~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. TT880. were dashe dotted~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. .... ...... .... ....... ...di after .1716.. reassigned~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. (in~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~NM modern place-names Some

in

italics)

~~

.o refrece

~

~

~ ~

~

..

................

37I

372

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

J

*

9

13. Ppulton in 1700. ~~~~~~~~igure Place-names appear in original AfterPanayiotopoulos ~~~~~~~~~~~~~orthography. 1987, pp. 250,262. D. K. Harlan and

Ga

Families 00

~

ii

~

J.Robertson. ~~~~~~~~~~~~~~~~~~~R.

~

~

~

~

A

~~mume1

01.4 * 514 @15-24

p

o Ac d

odi Navarino Lazerefl

Figure 14. Population in ca. 1716. Place-names appear in original orthography.AfterTT880.D. K.Harlan R. J. Robertson.

Ga~~~~~~~~~~1i (1

~~~~~Gca7~rninl

Muzuutela

Agas

klad

agg

OPenhul

00~~~~~~~ 014 * 5-14 @125-34

3VW

o 0 mAusrn

oi

WILLIAM

SIR

in 1815. Figure15.Population Place-names appearinoriginal After Pouquevillel1826orthography. 1827, VI, p. 73, note 1. D. K. Harlan and R. J. Robertson.

GELL

S ITINERARY

(k(

11

9

y

PYLIA37

Woy

lGXt

g

H_

-

li

|

ll

*Lvxs_

l

>

0

V

)*

4

t

Y

Papo+,& hsJ

+ hl*,^8r

70/ DRhmyw c _Kmi

11i l

9

l

g

t}.

y

| | I113

C

%O ~~?and

|

f(\s \0%N\ <

Jg

l~~~~~

>\<\ Ctnt}<

,

|

R. J. and~~

Aj)

+

\\

B| |

Virle 183-13,p and~~~~~

THE

@

S

| |l |

16 Pouato Figure~~~~~

IN

*lbtem-opetro-Cho

| | |

I *>GowJ

| l l l PimCE c c

\~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

y^,*ocuaa

in 1829_ .

Lloputis_S>

85.9D. K.fqHarla

a^-0__z<^urgs

Robermtson

|-==

\(142\~0~nnl

I131

1-4.

Figurel6.Populationinl829.~~~

Place-namesappearinoriginal~~ OrthOgraphy.AfteruillondeBoblaye

Viltl3313,p and~~~~~~

8.D

.Haln~~~~

_ E

E>g\\3
14\7470' i

A

_

1524 fAV 25 -4/\9sC\SM P \\> 11111 fafdrgop lan

)>s>

p

lf i]s 5NDES)

M S. vW3

~ t\

m

;

374

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

Such conclusions are relevant to recent examinations by FrangakisSyrett and Wagstaff of the so-called question of "height zonation" mentioned earlier in this article (see above, p. 345). Many scholars have suggested that the lowlands of the Morea were widely deserted in the Second Ottoman period. More specifically,they have argued that the Greek peasant population fled from these areas in the face of excessive taxation and other oppression inflicted upon them by the Turkish administration, and forsook fertile lowlands for "barrenbut secure mountain heights."104Such a widespread desertion of lowlands, if accuratelyportrayed,would hardly provide strong support for the position of the many Ottoman historians who have argued of late that the incorporation of Turkish territories into expanding world systems in the 18th century led to an expansion of markets in the 19th century and increased commercialization of agriculturein the Ottoman Balkan provinces.105Frangakis-Syrett and Wagstaff have, however, questioned the supposition that lowlands were deserted, basing their argument on analysis of population figures and lists of settlement names included in published Venetian censuses, on early maps, and especially on data reported by the Expedition scientifique de Moree. But how reliable are the data employed by Frangakis-Syrett and Wagstaff for their purposes? Information recorded by the Expedition scientifique can only be used with great risk to sketch a typical picture of land use and settlement in the 18th century, as Frangakis-Syrett and Wagstaff note; the turmoil produced by Ibrahim Pasha'soccupation of the Peloponnese had, at least in part, created the settlement patterns that were witnessed and described by the French. In this instance, population variability,even radical change in the extent and density of human settlement, was hardly a long-term phenomenon, and landscape archaeologists cannot afford to ignore short-term events of this sort while privileging conjonctureor the longuedzuree.106 The results of our own investigations are thus a welcome opportunity to evaluate the position advanced by Frangakis-Syrett and Wagstaff. In our view, the pattern of settlement around the Bay of Navarino, as it existed in the Second Ottoman period, was to a large extent already established before the Venetian occupation of 1685-1715. For example, there are patterns in the distribution of settlement types that suggest strongly that direct Ottoman involvement in the rural economy diminished with distance from the fortress at New Navarino (Anavarin-i cedid). In particular,those settlements characterizedas "villages"(Turkish,karye)in TT 880 tend to have the highest populations, are also at the highest altitudes, and are farthest from New Navarino (Fig. 17). Settlements categorized as fiftliks, but with a non-Turkish village name, tend to fall in the middle range of distance from Anavarin-i cedid and are situated at lower elevations. In TT880 such settlements invariablyhave a Turkish personal name as an alternativeto the village name. These Turkish personal names never appear in Venetian documents, which might imply that the extensive establishment of estates (rftliks) in existing villages (signified by attaching the personal name of the Ottoman "owner"to an existing village name) happened not long before the Venetian reconquest of 1685. Finally, what seem to be the major Ottoman estates, bearing only the name of an Otto-

Figure 17 (oppositepage).Ottoman population (ispence)vs. altitude and distance from New Navarino (Anavarin-i cedid). J. Bennet

104. See McGrew 1976. 105. E.g., Kasaba1988. 106. The Expedition'sfiguresare based in part on reportssubmittedto Kapodistriasby the administratorof the district;see above,note 53.

SIR

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

375

so - - --------

so 40

uain(rwe

ppoiaedsac

107. One possible exceptionto this rule is Abdul KadirAga fiftlik (TT 880, p. 34; see Fig. 12), CadirAga in the Venetiancensuses (Panayiotopoulos 1987, pp. 226, 262), which may have been establishedspecificallyto exploit the areaaroundthe prosperousvillages of Ligoudistaand Kavallaria.This settlement becomes simplyTchifliki by the early 19th century(Pouqueville 1826-1827, VI, p. 73, note 1; Puillon de Boblaye and Virlet 1833-1834, p. 85; see Figs. 15, 16). 108. These are the mazra'aof Deli Ahmed f ftfik and the mazra'aof Arkadianuor Mufti fnftfik:TT880, p.91. 109. In the Second Ottoman period these three settlementswere under the jurisdictionof Arkadiye(modern Kyparissia).Ligoudista,along with Skarminga(modernMetamorfosis), Iklaina,and Pyla, is among those villageswith the longest attested history in the region, stretchingback into the 13th century(Bon 1969, pp. 430-433).

1m

Eliue(elx1)

4

man individual (Hasan Aga, Rustem Aga, Kurd Bey, Osman Aga, and Ali Hoca), dominate the lower reaches of the major valleys leading into the Bay of Navarino.These estates must have been established during the First Ottoman occupation, since their names do appearin Venetian census lists, and their locations seem to reflect a preference for positions close to the fortress.107 TT880 also indicates the existence of mazra'asin the immediate vicinity of Navarino itself, which were worked by the inhabitants of the suburb of the fort.108 There is some evidence to suggest that the Greek population did avoid the lowlands nearest the fort of Anavarin, and it is impossible to exclude the possibility that, in the earliercenturies of Ottoman control, non-Muslim populations had moved from these areasto settlements farther inland and therefore at higher elevations. In Venetian censuses, the largest concentration of population in the area (although just outside the borders of the kaza of Anavarin) had been in the Ligoudista/Kavallaria/KadirAga region (modern Hora), where a total of 157 families are listed in 1700.109 Siriol Davies notes that the Venetians repeatedly tried to encourage inhabitants to relocate to the town of Navarino, apparentlywithout success. The total number of households recorded in TT 880 is, in fact, substantially greater than in Venetian times (Ligoudista, 128; Kavallaria,73; and Kadir Aga, 4), offering further proof of the failure of Venetian policy. In general, it is also clear that the majority of the Greek population lived at elevations above the 100-m contour, the arbitrarythreshold employed by Frangakis-Syrett and Wagstaff. Nonetheless, in marked contrast to the Peloponnese as a whole, virtuallyall settlements in the district of Anavarin

376

BENNET,

DAVIS,

AND

ZARINEBAF-SHAHR

were located at elevations beneath the 200-m contour, and, even if lands at the higher elevations were not in all cases as fertile as the coastal plains, they were hardly barren.110 Such micro-regional patterns here and elsewhere in the Morea may well lie behind commonly held traditions that Orthodox populations took refuge from Turkish administrators in the mountains, more so than the macro-regionalpatternsthat Frangakis-Syrettand Wagstaff examine,which do not in the aggregate demonstrate that a disproportionate fraction of Greek settlements were located at higher elevations.Is the patternof settlement recordedin TT880 a reflection of state policy during the First Ottoman occupation of the Morea?The scanty evidence availablesuggests that at Navarino, at least, depopulation occurred at that time. A Frankish inventory of the propertyof Niccolo Acciaiuoli in 1354 recordsa large population in the village of Kremmydi, an area that appears to have been entirely depopulated by 1716. But it is impossible to be certain if this was a general pattern.Might the Ottoman pattern itself reiteratein part a settlement structureestablished alreadyin the Frankish Morea? No additional Frankish data can be brought to bear on these questions.1"'Nor is there yet availabledetailed information concerning patterns of settlement in the First Ottoman period. Conclusions must consequently be limited to the final phase of Ottoman rule, 1715-1821. The evidence presented in this paper appearsto offer support for the deduction of Frangakis-Syrett and Wagstaff that a downward movement of population into the lowlands had begun before the Greek Revolutionthat there was no flight from such areasin the Second Ottoman period.112 In this period, it seems clear that Greek settlement at Navarino shifted toward the harbor from the area of the older suburb near the gate of the citadel. Because of the problems with Pouqueville'sfigures, it is difficult to assess the Greek population of Navarino between 1716 and 1815, but it appears to have held approximately steady: 30 families in 1716 to 142 individuals in 1815.113 It also seems evident that this Greek settlement had begun to be integrated significantly into largerMediterranean economies before 1821. Such a trend was severely disrupted by the invasion of Ibrahim in 1824, but accelerated when the French occupation reestablished security in the Morea. By 1829, 62 families are listed in the census at Navarino, alreadya significant underestimate by 1830, according to the Expedition.114 It is not impossible that the establishment of new villages in the 18th century around the Bay of Navarino also represents an expansion of local economic systems. Two coastal settlements (Lefko and Petrohori) appear to have been resettled subsequent to 1716.115 At the same time, there appearsto have been a significant reduction in the population of the large settlement cluster in the area of modern Hora. There appearto have been only 85 individuals in 1815, and not until the time of the census of the Expedition do we see a return to the levels of the Venetian period, with a total of 153 families in Ligoudista, Kavellaria, and Tchifliki (= KadirAga).116 In conclusion, our results agree largelywith those of Frangakis-Syrett and Wagstaff. A more fine-grained approach using high-quality unpublished documentary information and archaeological data has, however,

110. TT880 includes information concerningthe productivityof various crops for a selection of settlements and mazra a.

111. See Longnon and Topping 1969, pp. 67-115; Gerstel 1998. 112. This is not to imply that the Greek communitydid not take measuresto ensuretheir personal securityin the Second Ottoman period. Facilitieswere built to protect property and produce,and settlementswere situatedin shelteredplaces. See Forbes, in press, an articlethat evaluatesthe "heightzonation"hypothesisfor anothermicro-regionof Greece, Methana. 113. See Pouqueville1826-1827, VI, p. 73, note 1, for 1815, almost certainlyindividuals(above,note 25). 114. Puillon de Boblayeand Virlet 1833-1834, p. 85: the populationhad risen to 300 families. Ramfos lists only 63 in 1830 (Loukatos 1984, p. 214). 115. See above,pp. 362-363. 116. Pouqueville1826-1827, VI, p. 73, note 1; Puillon de Boblaye and Virlet 1833-1834, p. 85.

SIR

117. E.g., Lee, in press,a study examiningthe recenthistory of the village of Margeli and its vicinity,in the far northeasternreachesof the PRAP study region.In Ottoman times this village lay in the administrativedistrict of Androusa.

WILLIAM

GELL

S ITINERARY

IN

THE

PYLIA

377

allowed a more nuanced settlement history to be developed for our geographically restricted and temporally focused case study; it has thus been possible to evaluate their conclusions in detail at a micro-regional level. Whatever depopulation of the lowlands existed prior to the Second Ottoman period, this trend was surely reversed in the 18th century. Nor can observations recorded in travelers'accounts that might be adduced in support of a picture of widespread abandonment of agriculturalpursuits in these areasbe accepted as uncontested facts when other categories of evidence contradict them. This seems particularlytrue of the two published works of Sir William Gell that have been examined here. Finally, we return to the archaeological implications of this research. It should be obvious to readersthat the meaning of information contained in Ottoman documentary sources is not always transparent.Not least, its interpretation requiresa substantial investment in time and energy in unraveling the toponymic history of an area such as the Pylia, a task that can only be accomplished by comparing side by side a great variety of source material, including new field research, firsthand accounts published by earlier travelers, maps, and administrative documents. Such exacting research is, however, an essential first step in planning any strategy for the study of a region if the archaeology of the recent past is to be exploited to the fullest possible extent. In our own case, we hope not only to have created a framework in which the results of PRAP can be interpreted, but also to have provided groundwork for future programs of researchinto the archaeology of the recent past.1"7

REFERENCES Aeschimann,W., andJ. Tucoo-Chala, eds.1984.Edgar Quinet,La Grece moderneet sesrapportsavecl'antiquite suivi dujournalde Voyage, Paris. Alcock, S. E. 1998a. "Liberationand Conquest:Hellenistic and Roman Messenia,"in Davis 1998, pp. 179198. .1998b. "Hasanaga:A Glimpse into the Ottoman Countryside,"in Davis 1998, pp. 262-266. Alcock, S. E., J. F. Cherry,andJ. Elsner, eds.In press.Pausanias.Traveland Memoryin RomanGreece,NewYork. Amaury-Duval,P. 1885. Souvenirs (1829-1830), Paris. Anderson, R. 1830. Observationsupon thePeloponnesus and GreekIslands Made in 1829, Boston. Andrews,K. 1953. Castlesof theMorea (GennadeionMonograph4), Princeton. Atlas = ExpeditionscientifiquedeMoree: Sectiondessciences physiquesV:Atlas, Paris 1835.

Baltas,C. A. 1997. THAo,: Nac3cap'vo, NdcuT0`apo, avcx-opo

Niuropo;. Touopiru-ox o&rlyo% xcc govroyj7 crop(c, Athens. Belle,H. 1881. Voyageen Grece:Trois ann'esen Grece,Paris. Bennet,J., and S. Voutsaki.1991. "A Synopsis and Analysis of Travelers' Accounts of Keos (to 1821),"in Cherryet al. 1991, pp. 365-382. Bessan,J. F. 1833. Souvenirs de l'expeditiondeMoreeen 1828, suivis d'unMe'moire historiquesurAthenes, avecleplan de cettevilleparj F Bessan,Lalognes. Bintliff,J. 1999. "The Ottoman Era in the Context of Long-Term Settlement History.A Case-Study:The ArchaeologicalSurveyof the Valley of the Muses, Boeotia, Greece," ArabHistoricalReviewfor Ottoman Studies 19-20, pp. 203-229. Blouet, A. 1831-1838. Expe'dition scientifiquedeMoree:Architecture, sculptures, inscriptions,et vuesdu

378

BENNET,

Peloponese,des Cycladeset de lAttique,

Paris. Bon, A. 1969. La More'efranque: Rechercheshistoriques,topographiques, et archeologiquessur laprincipaute d'AchaYe(1205-1430), Paris.

Bory de Saint Vincent, M. 1836. Expedition scientifique de Moree: Section des sciencesphysiques I: Relation, Paris. Castellan,A. L. 1808. Lettres sur la More'eet les iles de Cerigo, Hydra et Zante, Paris.

Cherry,J. F. 1994. "RegionalSurveyin the Aegean:The 'New Wave'(and After),"in Beyond the Site. Regional Studies in the,Aegean,Area,P. N. Kardulias,ed., Lanham,pp. 91-112. Cherry,J.F.,J. L. Davis, and E. Mantzourani.1991. "Introduction to the Archaeologyof Post-Roman Keos,"in Cherryet al. 1991, pp. 351-364. Cherryet al. 1991 = J. F. Cherry,J. L. Davis, and E. Mantzourani,eds., LandscapeArchaeologyas Long-Term History: Northern Keos in the Cycladic Islandsfrom Earliest Settlement until Modern Times, Los Angeles.

Davies, S. 1994. "Tithe-Collectionin the VenetianPeloponnese,16961710,"BSA 89, pp. 433-455. Davis,J. L. 1991. "Contributionsto a MediterraneanRuralArchaeology: Historical Case Studies from the Ottoman Cyclades,"JMA4, pp. 131-216. Davis,J. L., ed. 1998. Sandy Pylos:An ArchaeologicalHistory from Nestor to Navarino, Austin.

Davis,J. L., S. E. Alcock,J. Bennet, Y. G. Lolos, and C. W. Shelmerdine.1997. "The Pylos RegionalArchaeologicalProject, Part I: Overview and the Archaeological Survey,"Hesperia 66, pp. 391-494. Dokos, K. 1976. "'H 'v ITlXoovviclcp zxx-J:LxatarLxi z 9LoorLO xxt B' 'Evetoowg -civ opLoaov xpoc oc," ByzantinischneugriechischeJahrbucher21, pp. 43-

168. Dokos, K., and G. Panagopoulos.1994. -, cx-coMyLo To f3svsEvx6 x Bouv-'racs, Athens. Driault, E. 1930. L'expe'ditionde Crete et de More'e(1 823-1 82 8): Correspon-

AND

DAVIS,

ZARINEBAF-SHAHR

dance des consulsde France en Egypte et en Crete, Cairo. Duheaume, M. A. 1833. Souvenirs de la

Moree,Paris. Faroqhi,S. 1999.,Approaching Ottoman History:An Introduction to the Sources,New York.

Fokas,D. G. 1927. 'H vcoujcX'ocroo Nac3cpp'voo, 1827-1927, Athens. Forbes,H. 1997. "Turkishand Modern Methana,"in Mee and Forbes1997, pp. 101-117.

. In press."Securityand Settlement in the Medieval and Post-MedievalPeloponnese,Greece: 'Hard'History versusOral History," JMA 13. Frangakis,E., andJ. M. Wagstaff.1987. "SettlementPatternChange in the Morea (Peloponnisos)c. A.D. 17001830," Byzantine and Modern Greek

Studies11, pp. 163-192. Frangakis-Syrett,E., andJ. M. Wagstaff.1992. "The Height Zonation of Populationin the Morea c. 1830,"BSA 87, pp. 439446. Gell, W. 1804. The Topographyof Troy andIts Vicinity,Illustratedand Explained by Drawings and Descriptions, London. . 1807. The Geographyand Antiquities of Ithaca, London. . 1810. The Itinerary of Greece, with a Commentary on Pausanias and Strabo and an Account of the Monuments ofAntiquity at Present Existing in That Country; Compiled in the Years1801-06, London. . 1817. Itinerary of the Morea: Being a Description of the Routes of That Peninsula, London.

1819. TheItineraryof Greece, Containing100 Routesin Attica, Boeotia, Phocis, Locris, and Thessaly.

London. 1823. Narrative of a Journey in theMorea, London.

Georgacas,D. J., and W. A. McDonald. 1967. Place Names of SouthwestPeloponnesus, Minneapolis.

Gerstel, S. E. J. 1998. "The Estates of Niccolo Acciaiuoli,"in Davis 1998, pp. 229-233. Hahn, M. 1997. "ModernGreek, Turkish,and VenetianPeriods: The Pottery and the Finds,"in

The Greek-Swedish Excavations at the AyiaAikaterini Square, Kastelli, Khania, 1970-1987, E. Hallagerand

B. P. Hallager,eds., Stockholm, pp. 170-196. Humphreys,W. H., and G. Pecchio. 1826. A Picture of Greecein 1825:As Exhibited in the Personal Narratives ofJames Emerson, esq., Count Pecchio, and W H. Humphreys, esq., Comprising a DetailedAccount of the Events of the Late Campaign, and Sketchesof the Principal Military, Naval, and Political Chiefs, New York.

Inalclk,H. 1983. "Introductionto Ottoman Metrology,"Turcica 15, pp.311-348. . 1994. An Economic and Social History of the Ottoman Empire I: 1300-1600, Cambridge. Jameson,M. H., C. N. Runnels, and T. H. van Andel. 1994. "The Early Modern Period, 1537-1834," in A Greek Countryside: The Southern Argolidfrom Prehistory to the Present Day, M. H. Jameson,C. N. Runnels,

and T. H. van Andel, eds., pp. 126139. Kasaba,R. 1988. The Ottoman Empire and the World-Economy:The Nineteenth Century, Albany.

Kiel, M. 1987. "PopulationGrowth and Food Productionin 16th-Century Athens and Attica accordingto the Ottoman TahrirDefters,"in Proceedings of the Vlth CIEPO Symposium (VariaTurcica4), J. L.

Bacque-Grammontand E. van Donzel, eds., Istanbul,pp. 115-132. . 1990. "Remarkson the Administrationof the Poll Tax (Cizye) in the Ottoman Balkans and the Valueof Poll Tax Registers (Cizye Defterleri) for Demographic Research,"EtBalk 26, pp. 70-104. . 1997. "The Rise and Decline of TurkishBoeotia, 15th-19th Century (Remarkson the Settlement Pattern,Demography,and AgriculturalProductionaccording to Unpublished Ottoman-Turkish Census andTaxationRecords),"in Recent Developments in the History and,Archaeologyof Central Greece: Proceedings of the 6th International Boeotian Conference(BAR-IS 666),

J. L. Bintliff, ed., Oxford, pp. 315358.

SIR

Kiel, M., and F. Sauerwein.1994. OstLokrisin tiirkischer und neugriechischer Zeit, 1460-1981: Persistenzund Zasurenin der und Siedlungsentwicklung demographischen Geschichte MittelGriechenlands auf Grundosmanischtiirkischer und neugriechischer administrativerQuellen(Passauer Mittelmeerstudien6), Passau. Kontoyioryis,G. D. 1982. Kotvwv1x27 (TO(310XcD'7c.Oc EAA2v1MX xoLvd-rTEs T27sTOU9XOX9G(TW(g

Athens. Koukou,E. E. 1980. Oc xowvorixo' Osouol

u-cf KoxAat&

;xca

GELL

S ITINERARY

EZ&Xoc," Platon 34-35, pp. 106-139. Makriyiannis,Y. 1966. The Memoirs of GeneralMakriyannis, 1797-1804,

H. A. Lidderdale,trans.,London. Mangeart,J. 1830. Souvenirs de la More'e,recueillispendant le sejour des Franfais dans le Peloponese, Paris.

McDonald, W. A., and R. Hope Simpson. 1961. "Prehistoric Habitation in Southwestern Peloponnese,"4AJA 65, pp. 221-336. McGrew,W. W. 1976. "The Land Issue in the GreekWar of Independence,"in Hellenism and the First Greek WarofLiberation (18211830): Continuity and Change

-rv

roopxoxprcqc;, Athens. Koukoulis,T. 1997. "Catalogueof Churches,"in Mee and Forbes 1997, pp.211-256. Lair,J. 1902. La captivitMde Franfois Pouqueville en Moree (Publications diversede l'Institutde France23), Paris. . 1904. La captivitMde Franfois Pouqueville a Constantinople, 18001801 (9prairial, an VII-16 vent6se, an IX), Caen. Leake,W. M. 1830. Travels in the Morea I, London.

Lee, W. In press."PylosRegional ArchaeologicalProject,PartIV: Change and MaterialCulturein a Modern Greek Village in Messenia,"Hesperia 70. Locatelli, A. 1691. Racconto historico della VenetaGuerra in Levante diretta dal valore delprincipe Francesco Morosini contro l'Impero Ottomano,

Cologne. Lock, P., and G. D. R. Sanders.1996. The,Archaeologyof Medieval Greece

(Oxbow Monograph 59), Oxford. Longnon,J., and P Topping. 1969. Documents sur le re'gimedes terres dans la principaut de More'eau XIV' siecle,

Paris. Loukatos, S. D. 1984. "lIoXvtzLoypwopLxdKopco'v%5Mz0s%v)g, xcxL) Neox'co-poou, 1830,"in Hpax-oxru rob B' TowxobC Zuvsap`ouo

WILLIAM

MsUm7vtcxv

ZToo8u&v (KorcLuucva', 27-29 Noif3p('oo 1982), Athens,

pp. 209-231. Lyritzis,S. T. 1982-1983. "'O&pyxouog Au-C.Mzoovqxg -CNg oTotoqlog

(Institutefor BalkanStudies 6), N. P Diamandouros,J. P. Anton, J. A. Petropoulos,and P Topping, eds., Thessaloniki,pp. 111-129. Mee, C., and H. Forbes,eds. 1997. A Rough and Rocky Place: The Landscape and Settlement History of the Methana Peninsula, Greece,

Liverpool. Miller,J. P. 1828. The Conditions of Greecein 1827 and 1828, New York. Molin, A. 1900. "Relazionedel Nobil Homo ser Antonio Molin ritornato di ProvveditorEstraordinariodi Morea,"in S. P. Lambros, "'ExOicszLs -C5vBsvsc&v ITpovoycGiv-cg ITzlXo7ovv'cuoo," xcd EOvoAoytxc% 'EracpEs(cI' ;r 'Ettc(o,; 5, pp. 429-447. As,tc'ov -rs

'Iu-opix%r

Nixon, L., J. Moody, and S. Price.In press."SettlementPatternsin Mediaevaland Post-Mediaeval Sphakia:Issues from the Environmental,Archaeological,and Historical Evidence,"in Proceedings of the Mediaeval and Post-Mediaeval GreeceConference,Corfu, May 1998,

J. Bintliff and D. Tsoungarakis,eds., Corfu. Panayiotopoulos,V. 1987. 1A7Uv7oyoud XG(LOLXLU(IOcT27s TIS,OWOVVv,2U

13o,--18o c,; cOva,,Athens. Pellion,J.-P. 1855. La Greceet les Capodistriaspendant l'occupation franfaise de 1828 a 1834, Paris. Perdicaris,G. A. 1845. The Greeceof the GreeksII, New York. Politis, N. G. 1915. "To7uo)vov,uxa," Laografia 5, pp. 249-308.

Pouqueville,F. C. H. L. 1820-1821. Voyagedans la Grece,comprenantla

IN

THE

PYLIA

379

description ancienne et moderne de l'Epire, de l'Illyriegrecque, de la Macedoine Cisaxienne, . . . avec des considerationssur l'arche'ologie,la numismatique, les moeurs,les arts, l'industrie, et le commercedes habitants de cesprovinces, Paris. . 1826-1827. Voyagedans la Grece:Deuxieme edition revue, corrige'eet augmentee, Paris.

Puillon de Boblaye,M. M., and T. Virlet. 1833-1834. Expe'dition scientifique de More'e:Section des sciencesphysiques II: Geographie etgeologie, Paris.

Ridgway,D. 1996. "Gell, Sir William (1777-1836)," inAn Encyclopedia of the History of ClassicalArchaeology,

N. de Grummond,ed., Westport, pp. 483-484. Robinson,T. In press."Ottoman Recordsand ClassicalArchaeology," in Proceedings of the XVth International Congressof ClassicalArchaeology, R. Dokter and E. Mordmann,

eds., Amsterdam. Sauerwein,F. 1969. "Das Siedlungsbild der Peloponnesum dasJahr 1700," Erdkunde 23, pp. 237-244. Simopoulos, K. 1984. HIl, ('8cavoc ~ivo i vv EAcAoc roLl '21 5, Athens. .1985. SEvoi Tccw;

u-mv E2Atckk;, 1800-1810 3, pt. 1, 2nd ed., Athens. Sutton, S. B. In press."ATemple Worth Seeing:Pausanias,Travelers, and the NarrativeLandscapeat Nemea,"in Alcock et al., in press. Topping, P. 1972. "The Post-Classical Documents,"in The Minnesota Messenia Expedition. Reconstructing a BronzeAge Regional Environment,

W. A. McDonald and G. R. Rapp, eds., Minneapolis,pp. 64-80. . 1976. "Premodern Peloponnese:The Land and People underVenetianRule (1685-1715)," in Regional Variation in Modern Greeceand Cyprus:Toward a Perspective on the Ethnography of Greece,M. Dimen and E. Friedl,

eds., New York,pp. 92-108. Valta,E. 1989. L'EubJeealafin du XV, siecle:Economie etpopulation, les registresde l'anne'e1474, Athens.

Vroom,J. 1998. "EarlyModern Archaeologyin CentralGreece:

BENNET,

380

DAVIS,

The Contrastof Artefact-Rich and SherdlessSites,"JMA11, pp. 131-164. Wagstaff,M. 1982. "Post-Classical Exchange," inAn IslandPolity:The Archaeology ofExploitationin Melos, C. Renfew and M. Wagstaff,eds., Cambridge,pp. 236-243. . In press."Pausaniasand the Topographers:The Case of Colonel Leake,"in Alcock et al., in press. Woodward,A. M., and R. P. Austin. 1925-1926. "SomeNotebooks of

AND

Sir William Gell,"BSA 27, pp. 67-80. Wroth, W. 1921-1922. "Gell, Sir William (1777-1836)," Dictionary of NationalBiography7, pp. 994-996. Zangger,E., M. E. Timpson, S. B. Yazvenko,F. Kuhnke,andJ. Knauss. 1997. "The Pylos Regional ArchaeologicalProject,Part II: LandscapeEvolution and Site Preservation,"Hesperia 66, pp.549-641.

John Bennet UNIVERSITY

OXFORD

OF ARCHAEOLOGY

INSTITUTE

36

BEAUMONT

OXFORD

OXI

STREET

2PG

ENGLAND [email protected]

Jack L. Davis OF CINCINNATI

UNIVERSITY

OF CLASSICS

DEPARTMENT

OHIO

CINCINNATI,

4522I

[email protected]

Fariba Zarinebaf-Shahr OF CHICAGO

UNIVERSITY CENTER

5828

FOR

SOUTH

CHICAGO,

EASTERN

MIDDLE UNIVERSITY

ILLINOIS

[email protected]

60637

AVENUE

ZARINEBAF-SHAHR

STUDIES

Recently Published

6LORIA

S. M EERKER

The Sanctuary of Demeter and Kore: Terracotta Figurines of the Classical, Hellenistic, and Roman Periods

394 pages, 79 plates CorinthXVIII,iv. ISBN 0-87661-184-6. May 2000. Cloth $100.00

This volume publishes the terracotta figurines of the Classical, Hellenistic, and Roman periods from the Sanctuary of Demeter and Kore on Acrocorinth. The coroplastic finds from this site, numbering in total about 24,000 figurines and fragments, greatly enrich the known body of Corinthian figurines not only in number but also in the addition of many entirely new types and styles. Working far beyond the output of the Potters' Quarter workshops, the Corinthian coroplasts are revealed as inventive, often highly adept in technique, and attuned to stylistic developments in the plastic arts in general. Evidence of terracotta figurines and small bronze ones, especially in the 4th and early 3rd centuries B.C., through the use of shared models, provides a glimpse of the mostly lost bronze production of that period. The figurines are also important because they help to explicate the meaning and conduct of the cult of Demeter and Kore in Corinth. The figural coroplastic art provides additional data on the deities and heroes recognized at the sanctuary,the age and gender of the participants in the rituals, the offerings they brought, and the nature of their cultic activities. Beyond these data, the figurines are examined for what they may reveal through their imagery of the underlying ideas of the cult, how the deities were perceived,why they were approached,and how the cult functioned as a part of Corinthian society.

MARTHA HEATH WIENCKE

Just Published

The Architecture, Stratification, and Pottery of Lerna III

The Lerna III, or Early Helladic II, discoveries at the site of Lerna in the Argolid, from the excavations made during the 1950s, have provided the largest body of material yet available from that crucial period in the prehistory of southern Greece. The finds have been known until now chiefly from the preliminary reports published in Hesperiaby the director,J. L. Caskey. In the present volume the evidence for the stratification and architectural remains of the Lerna III phase is presented in full, with plans, sections, and photographs, including a detailed description, tentative reconstruction, and discussion of the well-known House of the Tiles. Fascicle 2 contains a complete catalogue of the context pottery with drawings, arranged chronologically according to newly established categories of shape and fabric. In following chapters the author discusses the history of the individual shapes, together with much comparative material, as well as painted patterns, tactile decoration, and potters'marks.This is the first detailed study of the development of EH II ceramics. An appended report by P. P. Betancourt and G. H. Myer presents the petrographic analysis of a select group of sherds. Small objects, including lithics, listed by location in the stratigraphic sections, will be fully presented in future monographs by E. C. Banks and C. Runnels.

In two volumes: fascicle 1: 332 pages, 108 figures,32 plans,29 sections; fascicle2: 496 pages, 103 figures, 24 plates,3 appendixes,37 tables LernaIV.

ISBN 0-87661-304-0. August 2000. Cloth $125.00

Forthcoming

JOSEPH W. SHAW, ALEYDIS VAN DE MOORTEL, PETER M. DAY, VASSILIS KILIlO6LOU

A LM IA Ceramic Kiln in South-Central Crete: Function and Pottery Production 175 pages, 66 figures,16 tables HesperiaSupplement30. ISBN 0-87661-530-2. December 2000. Paper$40.00 (est.)

An in-depth study of the Late Minoan IA cross-draft kiln found in excavation at Kommos, Crete. The kiln is of a type popular during the Neopalatial period, and its good state of preservation has allowed the authors to speculate about its original internal layout and use as well as about the roof that covered it. Much of the large quantity of obviously locally produced pottery found associated with the kiln is analyzed in detail, allowing for the first time the study of the shapes, decoration, and technical characteristicsof vases known to have been fired in a specific LM IA kiln. The book presents an integrated program of analytical techniques used to illustrate the range of firing temperatures, the compositional similarities and differences in the clay used, and aspects of the firing process and the upper kiln structure.Offered here is a valuable contribution to our knowledge of the technology and organization of ceramic production at the beginning of the Late Minoan period, which will form a basis for studies of pottery provenience and exchange.

MARTHA K. RISSER

Forthcoming

Corinthian Conventionalizing Pottery

Corinthian Conventionalizing pottery is a fine ware produced during the 6th, 5th, and 4th centuriesB.C. While Athenian workshopsproducedblackand red-figuredvases, their Corinthian counterpartswere decorating vases predominantly with black and red bands, patterns, and floral motifs. This book provides a full and comprehensive study of Corinthian Conventionalizing pottery found during the American School of Classical Studies Corinth Excavations. Through the examination of contextual information, shape development, and changes in the style of painting, a chronology of the vases is proposed. This is followed by a discussion of painters, workshops, and groups. Evidence for systematic export is also presented.

195 pages, 45 plates, 31 figures CorinthVII,v. ISBN 0-87661-075-0. December 2000. Cloth $60.00 (est.)

HESPERIA

SUPPLEMENTS

13 MarcusAurelius:Aspects of Civic and Cultural Policy in the East, by James H. Oliver (1970). $15.00 14

The Political Organization of-Attica, byJohn S. Traill (1975). $15.00

15

The Lettering of an Athenian Mason, by Stephen V. Tracy (1975). $15.00

16 A Sanctuary of Zeus on Mount Hymettos, by Merle K. Langdon (1976).

$15.00 17 Kallias of Sphettosand the Revolt ofAthens in 286

B. C.,

byT. Leslie ShearJr.(1978). $15.00 19 Studies in Attic Epigraphy, History, and TopographyPresented to Eugene Vanderpool(1982). $15.00 20 Studies in Athenian Architecture,Sculpture,and TopographyPresented to HomerA. Thompson(1982). $15.00 21 Excavations at Pylos in Elis, by John E. Coleman (1986). $25.00 22 Attic Grave Reliefs That Represent Womenin the Dress oflsis,

by ElizabethJ.Walters (1988). $40.00 23

Hellenistic Relief Moldsfrom the Athenian Agora, by ClaireveGrandjouan

(1989). $25.00 24 The Prepalatial Cemeteriesat Mochlos and Gournia and the House Tombsof BronzeAge Crete,byJeffrey S. Soles (1992). $35.00 25

Debrisfrom a Public Dining Place in the Athenian Agora,

by Susan I. Rotroff andJohn H. Oakley (1992). $35.00 26

The Sanctuary ofAthena Nike in Athens:Architectural Stages and Chronology, by Ira S. Mark (1993). $50.00

27

Proceedingsof the International Conferenceon GreekArchitectural Terracottasof the Classical and Hellenistic Periods, December12-15, 1991,

edited by Nancy A. Winter (1994). $120.00 28

Studies in Archaic Corinthian VasePainting, by D. A. Amyx and Patricia

Lawrence(1996). $65.00 29

TheAthenian Grain-Tax Law of374/3

$35.00

B.C.,

by Ronald S. Stroud (1998).

H THE OF

%SPE

JOURNAL

OF THE

CL A S S I C A L

AMERICAN

STUDIES

VOLUME 2000

69

AT

SCHOOL A THENS

Hesperiais publishedquarterlyby the AmericanSchool of ClassicalStudiesat Athens. Foundedin 1932 and devoted primarilyto the timely publicationof reportson School-sponsored andSchool-directed projects,Hesperia welcomessubmissionsfromall scholarsworkingin the fieldsof Greekarchaeology, art,epigraphy,history,and literature,from earliestprehistorictimes onward.Hesperiais a refereedjournal. Submissions:Manuscriptsand communicationsshould be addressedto Dr. TraceyCullen,Editor,Hesperia,AmericanSchoolof ClassicalStudiesat Athens, 6-8 CharltonStreet,Princeton,New Jersey08540; tel. 609-683-0800;fax 609924-0578;[email protected]. Manuscriptsandphotocopiesof illustrationsmust be submittedin triplicate;originalartworkand photographsshouldnot be sent unlesspriorarrangements aremadewith the Editor.A shortabstractsummarizing the majorconclusionsof the articleshould also be included.The style for manuscriptpreparation,notes, bibliography,and other informationon submissionscanbe foundin the GuidelinesforAuthorspublishedin Hesperia62, 1993, pp. i-xvi, on the School'swebsite(www.ascsa.org), or bywritingto ASCSA Publicationsat the aboveaddress.Abbreviationsusedarethoselistedin theAmerican JournalofArchaeology 95, 1991, pp. 4-16; for abbreviations of epigraphicand numismaticpublicationsnot foundin AJA, pleaseconsultourguidelines. The AmericanSchoolof ClassicalStudiesat Athenswill not knowinglyprintin Hesperiaor any of its other publicationsthe announcementor initial scholarly presentationof anyobjectacquiredafterDecember30, 1973,by anymeansother than throughan officiallysanctionedexcavationor survey,unlessthe objectwas partof a previouslyexistingcollectionorwas legallyexportedfromthe countryof origin.

The annualsubscription Subscriptions: price,payablein advancein dollars,is $60.00 for individuals,$70.00 for institutions,and $33.00 for studentswith valid I.D. Canadaandothercountriesadd$10.00postage.We acceptVISAandMasterCard. Publishedquarterly.Reprintsof Hesperia1-41, IndexVolume1 (Hesperia1-10 and Supplements1-6), and Supplements1-12 and Supplement18 shouldbe orderedfromSwetsandZeitlinger,b.v.,P.O.Box 810,2160 SZ Lisse,Netherlands. Singleissues(currentand backnumberswhen available)of Hesperia42 and following are availablefor $15.00 each plus postagefrom the AmericanSchool of ClassicalStudiesat Athens,6-8 CharltonStreet,Princeton,New Jersey085405232 U.S.A.OrderIndexVolume2 andSupplements13 andfollowingfromUniversityMuseumPublications,Universityof PennsylvaniaMuseumof Archaeol33rdandSpruceStreets,Philadelphia, ogy andAnthropology, 19104. Pennsylvania

Copyright (? 1999 The American School of Classical Studies at Athens Produced at Edwards Brothers, Inc., Ann Arbor, Michigan. Design by Ellen McKie. Periodicals postage paid at Princeton, NewJersey, and at additional mailing offices. Postmaster.Send address changes to Hesperia,American School of

ClassicalStudiesat Athens, 6-8 Charlton Street, Princeton, New Jersey 08540-5232 U.S.A.

ISSN 0018-098X ISBN 87661-500-0

VOLUME

69,

2000

PUBLICATIONS STAFF

C

N

[

T

AJOOTIAN

AILEEN

A RomanTableSupportat Ancient Corinth

487-507

EDITOR-IN-CHIEF

KerriCox EDITOR, Hesperia TraceyCullen INTERIM EDITOR, Hesperia M. B. Richardson ASSOCIATE

EDITOR

BENNET,

JOHN AND

FARIBA

JACK

L.

DAVIS,

ZARINEBAF-SHAHR

Pylos RegionalArchaeologicalProject,PartIII: SirWilliam Gell'sItineraryin the Pyliaand Regional Landscapesin the Moreain the SecondOttomanPeriod

343-380

Michael Fitzgerald EDITORIAL

ASSISTANT

SuzanneAbrams PRODUCTION

MANAGER

BOOKIDIS

NANCY

CorinthianTerracottaSculptureandthe Templeof Apollo

381-452

SarahGeorge Figueira CREATIVE

COORDINATOR

JordanPeled

D.

MICHAEL

DIXON

A New Latin and GreekInscriptionfromCorinth

335-342

PUBLICATIONS COMMITTEE EVELYN

Carol C. Mattusch (Chairman) George Mason University Darice Birge Loyola Universityof Chicago Kevin Clinton Cornell University Jack L. Davis Universityof Cincinnati JeniferNeils Case Western ReserveUniversity James P. Sickinger Florida State University KathleenW. Slane Universityof Missouri-Columbia Stephen V.Tracy (ex officio) The Ohio State University

B.

HARRISON

EumolposArrivesin Eleusis

267-291

ALAN W. JOHNSTON

BuildingZ at Kommos:An 8th-CenturyPotterySequence

DAVID

R. JORDAN

A PersonalLetterFoundin the AthenianAgora MARK

189-226

L.

91-103

LAWALL

Graffiti,Wine Selling,andthe Reuseof Amphorasin the AthenianAgora,ca. 430 to 400 B.C.

3-90

IAN

MCPHEE

Falaieff Bell-Kraters from Ancient Corinth DANIEL

453-486

J. PULLEN

The Prehistoric Remains of the Acropolis at Halieis: A Final Report

133-187

LEE ANN RICCARDI Uncanonical Imperial Portraits in the Eastern Roman Provinces:The Case of the Kanellopoulos Emperor LINDA

JONES

Roccos

Back-Mantle and Peplos: The Special Costume of Greek Maidens in 4th-Century Funeraryand Votive Reliefs STEPHEN

235-265

V. TRACY

Athenian Politicians and Inscriptions of the Years 307 to 302 NATALIA

105-132

227-233

VOGEIKOFF-BROGAN

Late Hellenistic Pottery in Athens: A New Deposit and FurtherThoughts on the

Associationof Potteryand SocietalChange JOHN

293-333

TRAILL

EpigraphicalIndex

509-510

H THE OF

E$PER

JOURNAL

OF THE

CL A S S I C A L

AMERICAN

STUDIES

SCHOOL

AT

ATHENS

VOLU ME

69:

NUMBER

OCTOBER-DECEMBER

~ |~~~~~~~~~~~~~~

American Schoolof Classical Studies at Athens 2000

4 2000

The AmericanSchoolof ClassicalStudiesat Athens is a researchand teachinginstitutiondedicatedto advancedstudyof the archaeology,art, history,philosophy,language,andliteratureof Greeceandthe Greek world. Establishedin 1881 by a consortiumof nine Americanuniversities, the Schoolnow servesgraduatestudentsand scholarsfrommore than 150 affiliatedcollegesanduniversities,actingas a basefor research and studyin Greece.The mainbuildingsof the Schooland its library arelocatedin Athens,with administrative andpublicationsofficesin Princeton,New Jersey.As partof its mission,the Schooldirectsongoing excavationsin the AthenianAgoraand at Corinthand sponsors all otherAmerican-ledexcavationsand surveyson Greeksoil. It is the officiallink betweenAmericanarchaeologistsand classicistsand the ArchaeologicalServiceof the GreekMinistryof Cultureand,as such,is dedicatedto the wise managementof culturalresourcesandto the disseminationof knowledgeof the classicalworld.Inquiriesabout membershipin the Schoolor participationin the SummerSessions shouldbe sent to the AmericanSchoolof ClassicalStudiesat Athens, 6-8 CharltonStreet,Princeton,New Jersey08540-5232. Hesperiais publishedquarterlyby the AmericanSchoolof Classical Studiesat Athens. Foundedin 1932 and devotedprimarilyto the timelypublicationof reportson School-sponsoredand School-directed projects,Hesperiawelcomessubmissionsfromall scholarsworkingin the fieldsof Greekarchaeology, art,epigraphy,history,andliterature,from earliestprehistorictimes onward.Hesperiais a refereedjournal.

OCTOBER-DECEMBER

2000

THE

JOURNAL

OF CLASSICAL

OF THE STUDIES

AMERICAN

SCHOOL

AT ATHENS

PUBLICATIONS

NANCY

STAFF

Corinthian Terracotta Sculpture and the Temple of Apollo

BOOKIDIS

381

EDITOR-IN-CHIEF

KerriCox INTERIM EDITOR, HesperiaF M. B. Richardson ASSOCIATE

IAN MCPHEE

Falaieff Bell-Kraters from Ancient Corinth

453

A Roman Table Support at Ancient Corinth

487

EDITOR

Michael Fitzgerald EDITORIAL ASSISTANT

SuzanneAbrams PRODUCTION

MANAGER

SarahGeorge Figueira CREATIVE COORDINATOR

JordanPeled

PUBLICATIONS COMMITTEE Carol C. Mattusch (Chairman) George Mason University Darice Birge Loyola Universityof Chicago Jack L. Davis Universityof Cincinnati JeniferNeils Case Western ReserveUniversity James P. Sickinger Florida State University KathleenW. Slane Universityof Missouri-ColumbiStephen V. Tracy (ex officio) The Ohio State University

Submissions: Manuscripts and communications should be addressed to Editor,Hesperia,AmericanSchool of ClassicalStudiesat Athens, 6-8 Charlton Street, Princeton,New Jersey08540; tel. 609-683-0800; fax 609-924-0578; e-mail [email protected] photocopies of illustrationsmust be submittedin triplicate;original artworkand photographsshould not be sent unless priorarrangementsaremadewith the Editor.A short abstractsummarizing the majorconclusionsof the articleshouldalso be included.Articles are submittedto a double-blindreviewprocessand authorsare requestedto prepare their manuscriptsaccordingly,without their name or affiliationappearing. The style for manuscriptpreparation,notes, bibliography,and other informationon submissionscanbe foundin the GuidelinesforAuthorspublished in Hesperia 62,1993, pp. i-xvi; on the School'swebsite (www.ascsa.org);or by writing to ASCSA Publicationsat the above address.

Copyright (C2000 The American School of ClassicalStudies at Athens

The AmericanSchool of ClassicalStudies at Athens will not knowinglyprint in Hesperia or any of its other publicationsthe announcementor initial scholarlypresentationof anyobjectacquiredafterDecember30, 1973, by anymeans other than through an officially sanctioned excavationor survey,unless the objectwas partof a previouslyexistingcollection or was legallyexportedfrom the countryof origin.

ClassicalStudies at Athens, 6-8 Charlton Street,Princeton, New Jersey08540-5232 U.S.A. ISSN 0018-098X ISBN 87661-500-0

Subscriptions:The annualsubscriptionprice,payablein advancein dollars,is $60.00 for individuals,$70.00 for institutions, and $33.00 for studentswith valid I.D. Canadaand other countriesadd $10.00 postage.We accept VISA and MasterCard.Published quarterly.Reprintsof Hesperia 1-41, Index Volume 1 (Hesperia1-10 and Supplements 1-6), and Supplements 1-12 and Supplement 18 should be orderedfrom Swets and Zeitlinger,b.v., P.O. Box 810, 2160 SZ Lisse, Netherlands. Single issues (currentand back numbers when available)of Hesperia 42 and following areavailablefor $15.00 each plus postagefromthe AmericanSchoolof ClassicalStudiesat Athens,6-8 Charlton Street,Princeton,NewJersey 08540-5232 U.S.A. OrderIndex Volume2 and Supplements 13 and following from UniversityMuseum Publications,Universityof PennsylvaniaMuseum of Archaeologyand Anthropology,33rd and SpruceStreets,Philadelphia,Pennsylvania19104 U.S.A.

Producedat EdwardsBrothers, Inc., Ann Arbor,Michigan. Design by Ellen McKie. Periodicalspostage paid at Princeton,NewJersey, and at additionalmailing offices. Postmaster:Send addresschanges to Hesperia, American School of

HESPERIA

69,

2000

Pages 38r-4S2

R I NIH

TE

RRACOTTA

$CU

LPTU T EM

T HE

A

To Henry S. Robinson

IAN

CO

PO

RE P LE

D

AN O

F

LLO

AB STRACT Excavations around the Archaic Temple of Apollo in Corinth in the 1960s and 1970s uncoveredforty-eightfragmentsoflarge-scaleterracottasphinxes.' Possiblyfrom acroteriaof the temple roof, the fragmentsdate from the 6th to at least the mid-5th century B.C. The publication of these pieces providesan opportunityto reviewthe corpusof sphinxes,together with their stylistic development, and to discusshow they were made.The latest and best-preserved statue shows that this tradition did not end in the Archaic period. A small appendixof fragmentsfrom earlierexcavationsof Corinth completes the picture of Corinth'sproductivityin terracottasphinxes. It has long been acknowledged that Corinth was a major center for the productionof terracottasculpture.Statues made in its workshops areknown from Olympia, Delphi, Thebes, Halai, Athens, Kerkyra,and possibly from Samos and from Sane-Ouranoupolis in the Khalkidike.2 In addition,

1. This articlewas to have formed one chapterin a volume on the excavations carriedout by Henry S. Robinson on Temple Hill in Corinth.To facilitate publication,the planned chapterswill now appearas a series of articlesin Hesperia. I would like to dedicatethis articleto Mr. Robinson in thanks for his giving me this materialto publish and for his assistancewith an earlierand far rougherdraft.But more than that, I would like to thank him on behalf of all of us working at Corinth for the interest he showed in Corinth when directorof the School and for his revitalizationof the excavationsafterthe war. I would also like to thank CharlesK. Williams II for discussingnumerous points with me; Nancy Winter for her architecturaladvice;Elizabeth G. Pemberton,Aliki Moustaka,and Peter

Danner,all of whom readearlierdrafts; and Guy Sanders,who translatedinto a computerizedversion the original drawingin Figure 1 by Donald Sanders. I am gratefulto Nikos Kaltsasfor permission to measurethe sphinxesin the National Museum.The photographsare the work of Ino Ioannidouand Lenio Bartziotou.Finally,I wish to thank the anonymousreadersfor their useful comments and, in particular,Molly Richardson for her perceptiveediting. 2. Bibliographicalcitationsfor most of this materialcan be found in Weinberg 1957, Billot 1977, and OlForsch XXII. The colossalsphinx from Samos, first publishedby Buschor (1957, pp. 34, Beil. 2), is now being restudiedby Aliki Moustaka.For the materialfrom Sane-Ouranoupolis,see preliminaryreports in VokotopoulouandTsigarida

1993, 1995, 1997, and especiallythe report by Tsigarida(1997, pp. 335-336), who tentativelysuggests that the statues might be Corinthian.The Kerkyrean sculpturesare presentedby Dontas (1997, pp. 87-88); among them, a head from Mon Repos, MR-818, which Dontas attributesto a Nike acroterion, may be an antefix,as suggestedby P G. Kalligas(1968, p. 307). Billot (1977, p. 387) states that except for the sphinxesfrom Athens and Kalapodhi,all terracottasphinxes were made in Corinth.To her exceptions I would add sphinx I from Olympia;possibly sphinx II (OlForsch XXII, pp. 104-115); and the sphinx from Colonna Hill, Aigina (Alt-Agina II, 4, pp. 13-15, figs. 3-4, pl. 9:a-c), if the inclusions areindeed volcanic, as stated in that publication.

382

NANCY

BOOKIDIS

terracottasphinxes from Kalydon,if not made locally by Corinthians, were probablymade under Corinthian inspiration.All of this exported material was apparently designed as architecturaldecoration for temples, treasuries, and stoas. With the exception of several pedimental groups from Corinth, and perhapsone from Kerkyra,4the remainingpieces areacroteria and take the form of animals, chiefly sphinxes, nikai, or combatant figures. In 1957 Saul S. Weinberg published his important work on the terracotta sculpture from Corinth, cataloguing forty-eight pieces that had derived from the excavations up to that time.5 That corpus has now been expanded by two groups of more recent discoveries.The first of these consists of a substantialbody of large, freestanding dedications of the late 7th to 4th centuries B.C. from the Sanctuaryof Demeter and Kore in Corinth.6 The second, and the subject of this article, is a smaller set of architectural sculpture that was excavated around the Temple of Apollo between the years 1968 and 1976 under the direction of Henry S. Robinson.7 Fortyeight fragments of terracottasculpturewere found on Temple Hill, thirtyeight of which have been numbered 1 to 15. The remaining ten pieces, considered too small or amorphous to be included in the catalogue, have been cited in footnotes where relevant. Only one set of fragments, Ila-c, was not found in the modern excavationsbut in late debrisjust to the east of Temple Hill. With three exceptions all belong to sphinxes. The exceptions are a wing of a possible Nike (12), a human foot (13), and a possible disc (15).' A fourth piece, a large strut (14), may or may not have belonged to a sphinx. All of the newly excavated fragments but one (13) were recovered from the east end of Temple Hill in the filling of the limestone quarrythat cut across the entire east side of the temenos (Fig. 1).10Filled in during the first third of the 1st century A.C., the quarry contained abundant pottery and small finds, roof tiles, and discardedblocks of the 7th-centuryTemple of Apollo that had been thrown in with the successivetipped fills of earth."1 3. According to Nicholls (1970), both the Royal Stoa and South Stoa I in the Athenian Agora were decorated with terracottaacroteria. 4. Weinberg 1957, nos. 8-12 and possibly no. 13, pp. 306-309, pls. 6567. Also perhapsWilliams and Russell 1981, no. 11 (SF-80-1), pp. 31-33, pl. 9; the head was stolen in 1990 and recoveredin 1999. The inventory numbersin parentheses,accompanied by the letters "SF"(sculpturefictile), referto a new inventorysystem reservedfor terracottasculpturein the Corinth collection.As for the Kerkyreanamazonomachy,Dontas (1997, p. 88) attributesthe group to acroteria,but it would be more suitable in the pediment of a small building. 5. Weinberg 1957. Objects publishedthereinwill be cited by his cataloguenumber,precededby "Weinberg."

6. These will be the subjectof a separatevolume in the series of Corinth XVIII. See also Bookidis 1988. 7. Preliminaryexcavationreports can be found in Robinson 1976a and 1976b. 8. Numbers in bold print referto the cataloguebelow. 9. An additionalfragment,SF-77-5, may preservedraperyarrangedin zigzag stackedfolds, but it is too small to providemuch informationand has thereforebeen omitted. With it can be associatedSF-88-1 from the areaeast of the Theater. 10. It is no longer clearwhere the quarriedstone was used. Still visible within the quarryare the partiallycut outlines of relativelysmall blocks. 11. Potterycontext lots 72-175, 74-89, 74-178, 74-181, 75-184, 76177, 76-178, 77-233. Robinson 1976a, p. 237; Robinson 1976b, p. 254.

TERRACOTTA

CORINTHIAN

S

max Figurel.Planofeastend,Temple~~

statue Hill,withfindspotsof

12. For a discussionof the Athena Trench,named for the sanctuaryof Athena Chalinitis once presumedto have stood there, see Shear 1925, pp. 388-391; Shear 1926, pp. 444-449; and Williams and Zervos 1984, p. 83. The areaof the Athena Trench excavationsfalls along the east side of Williams'TerracedBuilding, shown in his figures3 and 4. It is now clearthat the numerousfragmentsof terracotta sculpturefrom the Athena Trenchwere all part of a generalcleanupof the city in Roman times. Among those fragmentspublishedby Weinberg (1957), nos. 4,20,21, 32, and 36 come from the Athena Trench. 13. Milchhoefer (1879, p. 70, no. 5) mentions a sphinx from Corinth, its discoveryprecedingthe earliest excavationson Temple Hill by D6rpfeld in 1886. It is regrettablethat we do not know which fragmentthis was.

-

383

SCULPTURE

Q~~~~~T13

C

A

9

Most of the fragments of terracotta sculpture came from the northwest corner of the quarry,but a few were discovered farther to the southeast. Although pieces of the same sphinx, joining and nonjoining, were sometimes found in the same stratum,this was not always the case. The numerous pieces of sphinx 7 lay in different parts of the northwest corner as well as in different layers.Three fragments of wing 9 were scattered from the northwest corner to the east side of the quarry.In a few cases the fragments were found in Byzantine or modern disturbances of the quarryfill. Only 13 came to light further west, along the north edge of the temenos, roughly on line with the middle of the temple. Some associations are made here between the newly excavated material and fragments published by Weinberg. Of these earlier finds, 4a had been recovered from the southwest edge of the Hill, behind the west end of the Northwest Stoa, in Late Roman or Byzantine fill; 2a was discovered in the so-called Athena Trench northwest of Temple Hill and east of the Theater.12 Four more, 2c, 8b, 8c, and a joining piece of 7h, are of unknown proveniencebut could have derivedfrom earlierexcavationsaroundTemple Hill." In addition to the fifteen entries from Temple Hill presented in part I of the catalogue, eleven more pieces have been included as 16 to 21 in part II of the catalogue. Five of these, 16, 18a-c, and 19, are unpublished frag-

384

NANCY

BOOKIDIS

ments of sphinxes that derive from the early excavations of Corinth. Although their specific provenience is not known, some, if not all of these, could have come from the Sanctuary of Apollo. 18a-c may perhaps be from Temple E, while the four fragments of the sea monster 20a-d were found in the Athena Trench and its modern extension, known as the area east of the Theater. In view of the clear association of some finds from the theater areawith those from Temple Hill, I have included them here, although it is equally possible that they once belonged to the decoration of buildings elsewhere.With the publication of these examples, the corpus of sphinxes found at Corinth is complete, excepting any recent finds from the excavations in the Forum area. The sphinxes that are discussed here and in the catalogue that follows are extremely fragmentary,and one might ask why they should all be published. There are several reasons. Terracotta is consistently overlooked as a medium of sculpture.And yet for a city like Corinth, it was the medium in which much or most of its pre-Roman sculpture was made. It must be considered. Over the centuries, the multitudinous occupants of the site of Corinth have not been kind to the city's Greek art. With the exception of two stone sphinxes-a poros sphinx from the Forum and a second sphinx from the North Cemetery14-the Greek sculpturefrom the site is extremely fragmentary.But no matter how small the pieces, they give us valuable glimpses, if not whole views, of Corinthian workmanship.With regardto the sphinxes, in particular,the fragments published here also help to fill the enormous chronological gap that has separated the existing Archaic corpus from the Late Classical sphinx found at Thermon.

FUNCTION AND CHRONOLOGY OF THE FRAGMENTS FROM TEMPLE HILL The fragments from the excavations around the Temple of Apollo represent parts of no fewer than nine, and perhaps as many as eleven, statues of sphinxes, no two of which are alike."5This is a considerable number and raises the question of their function in the sanctuary-that is, whether they are architecturalor votive. Marie-Francoise Billot has discussed this issue with respect to the Louvre sphinx, which was probably found in a cemetery near the Temple of Apollo Ismenios."6Although the majority of terracottasphinxes that have been found were once mounted on roofs, two terracotta sphinxes might have been votive-namely, a small sphinx from the Colonna Hill on Aigina,'7 and a colossal sphinx from the Heraion on Samos."8 The Corinth fragments offer no assistance in gauging their functions, for they are too fragmentary.No evidence for votive offerings exists in situ on Temple Hill in the form of bases or rock cuttings for bases. At the same time, the 6th-century B.C. Temple of Apollo apparentlyhad acroteria,for an unpublished fragment of a terracottasima preservespart of the box for one, including a small portion of the floor on which the acroterion rested (Fig. 2).'9 Unfortunately, neither the height of the tympanum nor that of the temple as a whole is known, from which the probable height of its

14. Poros sphinx from the Forum: Wright 1977. Stone sphinx from the North Cemetery:ProtonotariouDeilaki 1973 andAlt-Agina II, 2, p. 55, figs. 64-65. 15. It is possible,of course,that fragmentsattributedto a single statue are actuallyfrom a largernumberof statueswhich were identical. 16. Billot 1977, pp. 419-420. Goldberg (1977, p. 172) suggests that one criterionfor distinguishing acroteriafrom votive sphinxesmay be the turn of the head. Sphinx acroteria, she argues,tend to turn their heads only slightly,an arrangementthat would suit our 7. The exceptionsshe notes are the sphinxesfrom the Temple of Aphaia on Aigina, which turn their heads 90 degrees.But the two terracotta sphinxesfrom Olympia,which areprobablyacroteria,also turn their heads 90 degrees.So do those from the Temple of Artemis on Parosand the Temple of Apollo on Aigina, according to P. Danner (1989, pp. 46-47 and note 184). See alsoAlt-Agina II, 2, p. 76. 17. Alt-Agina II, 4, pp. 13-15, figs. 3-4, pl. 9:a-c. 18. Buschor 1957, pp. 3-4, Beil. 2; OlForsch XXII, p. 108. Aliki Moustaka, who is working on this statue,informs me that it is probablya votive statue ratherthan an acroterion. 19. FS-23. P.H. 0.116, p.W. 0.155 m; single fragmentpreserving upperpart of torus and crowning astragalmoldings;it is decoratedwith a lotus-palmette chain in dark-on-light style.The top of the sima is cut down 0.018 m to hold the plinth of an acroterion.For the type see Roebuck 1990, p. 56, fig. 2 and pl. 6.

CORINTHIAN

TERRACOTTA

385

SCULPTURE

w top

Figure2. Box for an acroterion, Templeof Apollo, FS-23

20. Danner (1989, pp. 69-70) argues that from ca. 530 B.C., there is a relation between the height of the acroterionand the total height of the building on which it stood. Regrettably,there arevery few Archaic buildings for which both of

those dimensionsare known. 21. The only possible exception might be the fragmentaryleg 6. But becauseit is an unusualpiece, its scale cannot be clearlydetermined. 22. Weinberg 1957, no. 33 (SF-31-2), pp. 314-315, pls. 71-72. The statue has been newly mounted.A comparisonwith old and new photographswill show that the legs have been shortened,the body tipped forwardsomewhat. A considerableamount of the terracottasculpturepublishedby Wein-

acroteriacould be determined, if indeed there was a common ratio at this earlydate.20But since the temple was large,it must have had large acroteria. Of the sphinxes, only one, 7, is sufficiently well preserved to be reconstructed (Fig. 8). Its height is estimated to have been 0.94 m. As for the remaining fragments, one can say only that their sizes are not incompatible with that of 7.21At that size the sphinxes are considerablylarger than the well-preserved terracottasphinx published by Weinberg, which is about 0.63 m high (Fig. 33),22 and the poros sphinx from immediately south of the South Stoa, which is 0.77 m high.23The scale of the sphinxes published here, then, could argue for their association with the temple. In addition, the fired colors of the clay and paint used in the wing 9 and disc 15 are identical to those of a ridge palmette attributed to a Classical or Late Classical repairof the temple's roo.24 Given the statues'scale and the similarities among 9, 15, and the temple ridge palmette, it is possible that all of the statues once decorated the roof. berg was found in the region of the RomanTemple E. Fifteen to nineteen statueswere recoveredfrom a Late Archaicwell beside the temple and found joins with fragmentsfrom the immediate area.Initially also assigned to the well, Weinberg'ssphinx no. 33 was actuallyfound to the northeastof Temple E below the modern museum. For the correctionsee Hesperia 27, 1958, p. 79. Weinberg concludedthat all of this materialderivedfrom a coroplast'sworkshopin the area,based on the discoveryof kiln props and the assumedunfinished state of the amazonomachypediment no. 8. It is more likely that the pieces designatedas kiln props are simply internalsupportsfrom the statues;as for the unfinished-that

is, unpainted-state of the pediment, I would arguethat it was originally coveredwith white slip that has now been lost. It is unlikelythat a workshop with a kiln large enough to fire statues would have existed so close to the Temple of Apollo. But if this material once decoratedbuildings in the areaof Temple E, no evidence of those buildingshas been found. 23. Wright 1977. 24. For the ridge palmette see Robinson 1976a, p. 236, pl. 53:c (FR-lOlA). The clay of all three has fired a distinctiveyellow at surfaceand bright pink at the core;in addition, a light red color that does not appear elsewhereis used for their painted decoration.

386

NANCY

BOOKIDIS

Unless some of the pieces assigned to one sphinx are really part of a second, identical one, no two of the statues were alike.25Their number and stylistic variations may therefore mean that these statues, if architectural, were replaced with relative frequency. For those who have experienced a typical Corinthian winter with strong winds from the north, this should not be surprising. As will be clear from the cataloguebelow, the majorityof the terracotta fragments are small and difficult to date. If these statues are correctlyassociated with the temple, however,then the earliest pieces are no earlierthan the first decorated roof associated with that building. Since the 6th-century temple will be the subject of a separatestudy by Christopher Pfaff,26I will limit myself here to a few comments on its date, based on existing published works. Scholars have placed the construction of the temple variously in the first half of the 6th century,27570-560,28 ca. 550,29the early third quarter of the 6th century,30and 560-540 B.C.3"Of these dates, the early third quarterof the 6th century B.C. is perhaps the most objective, based as it is on sherds of both Late Corinthian I and II styles that were recoveredfrom the layer of poros working chips associated with the construction of the 6th-century temple. Whether one follows the traditional dates of 570 B.C. for the beginning of LCI, and 550 B.C. for LCII, or the more recent lowering of the beginning of LCI to ca. 560-555 B.C.,32it is difficult, based on the pottery,to defend a date earlierthan 550 for the building of the temple. Sphinx 2, virtuallythe only terracottastatue of the 6th century catalogued below for which some sort of independent stylistic date can be suggested, would suit a roof laid down early in the third quarter of the century.Not all of the fragmentsconsideredbelow,however,belong to this period. Some, such as 4,5, and 7, very clearly were made in the 5th century B.C.; wing 9 could even be later.Again, if all of these fragments are correctlyidentified as architecturalrather than votive sculpture,then clearly the decoration of the roof underwent more than one remodeling. 25. How similarthe terracotta acroteriawere on a building is difficult to assess,given the relativelysmall numberthat areknown from the same site. Danner (1989, pp. 64-65) states that in the Archaicperiod the acroteria of both facadesof a building represented the same subjectbut that the individualstatuescould differin size and modeling. He would, for example, put all four sphinxesfrom Kalydon on Temple A, placing the larger sphinxeson the east facade,the smaller ones on the west. Others would divide them betweenTemples A and B. Differencesin executioncan be seen among them. On the other hand, the cornerNikai acroteriafrom ancient Sane-Ouranoupolisin the Khalkidike, now mounted in the museum at

Polygiros,arevirtuallyidentical,even to the point of both having their right hand raised;variationsare confined to the painted decoration. 26. For a preliminarydiscussionof the temple, see Pfaff, forthcoming,in CorinthXX. The attributionto Apollo is discussedin Bookidis, forthcoming, in CorinthXX. 27. CorinthI, i, p. 124. 28. Robinson 1976a, p. 218. See also Roebuck (1990, p. 55), who favorsa date around560 B.C. 29. Williams 1984, pp. 69, 71. 30. Weinberg 1939. 31. Winter 1993, p. 24, 550-540 B.C., but possiblybegun in 560 B.C. 32. CompareAmyx 1988, pp. 397434, esp. p. 428, and Tiverios 19851986.

CORINTHIAN

TERRACOTTA

SCULPTURE

387

TECHNIQUE All of the statuespresentedbelow were madelocally.The claysused are fairlyhomogeneousin theirfiredcolorsand in the kinds and sizes of inclusionsaddedas temper.The fired colors of the clays are pale, and fit what one traditionallythinksof as Corinthian,thatis, buffto off-whiteto pale yellow-green;33 rarelyarethey as darkas pale brown.34Firingis seldom consistentthroughoutthe thicknessof a piece, and shadesof light pinkto darkerpinkandof light grayarecommonfor the interiorsurfaces and cores.35Sphinx7, which firedpale buff for almostall of its surface, showspatchesof tan on its posteriorthat mayreflectthe statue'sposition in the kiln and an uneven exposureto heat.There are no fragmentsin darkerreddish-tanclays such as appearamong the freestandingstatues from the Sanctuaryof Demeter and Korein Corinth.36If those darker claysarealso Corinthian,then the absenceof sphinxesin such claysmay be of interestandcouldreflecta consciouspreferencefor certainclaysfor certainkindsof sculptureor the availabilityof specificclaybedsto specific workshops.37 One furtherobservationmay be made on the clay.In at least two examples,7 and 9, individual,unblendedpellets of white and red clays couldbe discernedwithin the fabric.Superficiallyresemblingharderinclusions, these can be distinguishedby their very smooth texture and streakedoutline.Their presenceindicatesthat two differentclays were mixed.38 Why claysweremixedis not entirelyclear:perhapsfor purposes of firedcolor,addeddurability,or plasticity.The practice,however,may invalidate,or atleastcomplicate,conclusionsdrawnfromchemicalormineralanalysesof the clays. In allof the statues,two consistenciesof clayhavebeenused(cf. breaks visiblein Figs.3, 5, 6).The basicoutlinesof the figuresaremodeledin clay to which temperhas been added.This temperis most commonlymudstone,39the colorof whichis affectedby firingconditions.Where the clay has firedyellowto buff,the mudstoneis black;wherethe claysarepinkto reddish,the mudstoneis red.In 10 and 17, a few redparticlesoccuralong with the black.The inclusionsvaryin size fromless than0.001 to 0.0020.003 m andareonlyrarelylarger.In shapethey areangularto subangular, spherical,platy,andtabular,andsometimestheiredgescanbe quitesharp,

33. Translatedinto the colors of the Munsell charts 1975, these rangefrom 5Y 7-8/2-3 (four statues)to 2.5Y 78/2-5 (seven statues). 34. Or, 10YR 7-7.5/3-4 (three statues). 35. These range from 5Y 7/2 to 7.5YR 7/5 and to 5YR 6/6. 36. Bookidis and Fisher 1972, pl. 65:a-b, from 5YR 5/6 to 7.5YR 6/6. 37. A sphinx from Olympia is modeled in clay that is redderin color, or 5YR 7/6. See OlympiaIII, pp. 40-

41, pl. VIII:4; Kunze 1941, pp. 120121, fig. 103; OlForsch XXII, group II, pp. 106-112. Most recently,Aliki Moustaka (OlForschXXII, p. 102) has attributedthis sphinx to the northeast Peloponnese.It does not resemblethe existing corpusof Corinthiansphinxes, however. 38. See Whitbread 1995, pp. 298299. 39. This identificationwas verified by the late George Viele of the Universityof Missouri.A petrologic

analysisof a fragmentof terracotta sculpturefrom the Sanctuaryof Demeter and Koreproducedthe following results:mudstone 76%, monocrystallinequartz11%,polycrystalline quartz5%,limestone 3%, feldspar1%,unidentified 1%.See resultslisted byWhitbread (1995, p. 298, fig. 5:9), who carriedout this analysis.The firing color of the inclusions is also affectedby the presenceof iron oxide in the mudstone.

388

NANCY

BOOKIDIS

producing hackly fractures.40In general, the inclusions arewell mixed, but exceptions occur.As Weinberg observed, the inclusions used in the sphinx wings 2c (see section showing break in Fig. 6) and 11 are "stratified"by size of particle, for a central layer of clay mixed with coarser particles is sandwiched between layers with fine inclusions.41This method was not widely used, and it offered no apparent advantage. A second layer of the same clay,but without added temper, masks the coarse core (Figs. 3, 6, 33). The thickness of this layervaries from statue to statue and from part to part.While averaging0.001-0.003 m in thickness, it occasionally reaches a thickness of 0.017 m or more. Generally speaking, it is the medium in which the finer modeling was executed. In places the layer is so thin that it looks as if it had been smeared or brushed on like a slip.42 Weinberg has discussed the way in which statues were modeled, insofar as his material permitted. An ever-growing corpus of sculpture allows for additional observations to be made, most of which will be discusseacin conjunction with the more abundant material from the Sanctuary of Demeter and Kore. Here it will suffice to concentrate on the making of sphinxes on the Greek mainland, as exemplified primarily by the statues from Corinth.43 Today clay sculpture is generally modeled solidly over a wire armature. By contrast, ancient statues were hollow and lacked such a unified armature,for the figure was built in parts that were then joined.44Where necessary,sticks and props were used for individual parts such as arms and legs (Figs. 4,22). Such supportswere generallyremoved before firing, however, to provide ventilation channels for the circulation of air during firing and to prevent the drying clay from cracking. Ancient statues were modeled freehand, in molds, or by a combination of both methods. In her recent study of the Louvre sphinx from Thebes, Marie-Francoise Billot counted five molds for the head alone as well as at least one mold for each wing.45While Billot's proposed number of molds may be excessive,46the 40. I would like to thank Guy Sandersfor assistingme in standardizing the descriptionsof the clay and the inclusions.He has done considerable experimentationwith the firing of local clays,which will, it is hoped, be the subjectof a futurestudy.See Sanders 1999, pp. 477-478 for useful chartsand for the terminologyused here in the descriptionof fabrics. 41. Weinberg 1957, p. 294, no. 34B and C, pl. 75. 42. It is this fine outer layerof varyingthicknessthat is absentin Etruscanterracottas.See Lulof 1991, pp.130-131. 43. For recent importantdiscussions of the technical aspectsof modeling, in conjunctionwith the architectural terracottasfrom Satricumin Latium, see Lulof 1991, pp. 115-136 and Lulof 1996, pp. 11-24,175-182. As Lulof

herself observes,however,the techniques used there aredifferentfrom those used on the Greek mainland. One small but importantdetail is the multiplicityof vent holes used in Etruscanstatuesthat do not appearon Greek mainlandpieces. 44. Accordingto Lulof (1996, p. 20), lower arms and hands were sometimes not only modeled separately but also fired separatelyandjoined thereafter.I know of no such practice among the mainlandpieces. Separate firing alwaysraisedthe possibilityof differentcolors and differentrates of shrinkage. 45. Billot 1977, p. 392. 46. I have not had the opportunity to corroborateBillot's observations.I would have thought, however,that the frontalmassesof hair could have been formed in the same mold as the face.

CORINTHIAN

47. Weinberg (1957, p. 298) cites the layeringof differenttexturesof clay in the statueWeinberg34 as evidence of the use of a mold. 48. SF-64-12, for which see Stroud 1968, pl. 95:e; the pigtail has flaked awayin the same way. 49. MF-8778, to be publishedby Gloria Merkerwith the other finds from the Tile Works.Merkersuggests, however,that the mold may have been intended for a bronzewing ratherthan a terracottaone (pers.comm.). Billot (1977, p. 392) states that the Louvre sphinxwing was made in a mold. 50. Weinberg 1957, pp. 315-316, pl. 72; the piece should be turned so that the scales are horizontal.

TERRACOTTA

SCULPTURE

389

finds from the Sanctuaryof Demeter and Kore show that faces were definitely made in molds. For bodies the evidence is less clear. Various criteria can be used to determine whether a mold has been used. One such criterion is the mechanical or dull appearancethat the use of a mold is said to give to a surface. But in large statues, subsequent retouching can effectively modify that impression. Deep fingerprints on the interior, particularlybehind noses and chins where clay has been pushed into a cavity,are a second criterion.A third might be lengthyjoints such as appearbetween the front and back of moldmade figurines. Less clear as criteriafor molds aretwo other features,namely,a smooth interior surface combined with a consistent thickness of the statue wall, and the use of fine surface clay over coarse core clay for modeled details, which is characteristicprimarily of Archaic sculpture. It has been argued that only the layering of fine clay in a mold can have produced such a relatively thin surface (as seen in the break shown in Fig. 6).4 There may be some evidence to suggest that it is not always so. The fine surface clay of the sea creature20c has flaked away to reveal a very thin fine undercoat that covers the coarse core to act as a binder. A similar arrangement is observable on the back of an Archaic draped statue from the Sanctuaryof Demeter and Kore.48Occasionally, too, fingerprints can be seen on the surface of the coarse core, where the fine clay is missing. These details suggest that in those pieces the layering progressed from core to surface. A mold for a sphinx wing of the late 5th or 4th century B.C. was found in the Tile Works of Corinth.49Although smaller in scale and stylistically more elaboratethan the sphinxes from Temple Hill, the mold nevertheless suggests that a mold could have been used for the wings of the sphinxes here. A mold offers several advantages.It facilitates the even reproduction of the feathers while maintaining a fairly consistent thickness (or thinness) for the wing as a whole. It must also help in the process of consolidating the clay over such a large area.A sign that some of the wings from Corinth were made in molds is the absence of cutting marks at the junctures between successive feathers executed in relief. But making a wing with a mold does not requirethat the coroplast simply rolled out a slab of clay and pressed it into the mold, cutting off the excess. The broken surfaces on the Temple Hill wings suggest that the clay was laid down in a variety of ways. Some wings, such as 2c (Fig. 6) and 17 (Fig. 29), appearto be formed of a well-compacted horizontal layer or layers of coarse clay beneath the fine surface layer.Wing 2d is modeled in alternating layers of fine and coarse clays (Fig. 5). Other wings, such as lb, 7g, and 7h (cf Figs. 16, 17), show evidence of having been built up by means of wads or snakes of clay. See below in conjunction with 7h. Molds may have been used for more than wings. As Billot has observed, the wings of a sphinx give the statue its maximum width. Viewed from the front, the shoulders and humeri align with and appear to be of about the same thickness as the wings (Fig. 33). Therefore, the wings, shoulders, and humeri could have been made in one piece. The shoulders are sharply offset from the chest, the juncture often reinforced on the exterior with a strip of clay made into a fillet (see Fig. 33). While the interior of the sphinx Weinberg 33 does not show anyjoint in the clay at this juncture, the fragmentarybreast,Weinberg 37, does.50

390

NANCY

BOOKIDIS

Further evidence that Weinberg 37 was made in a mold is the consistent thickness of the wall in section and the smoothness of the interior surface,mentioned above. Despite the heavy reinforcement of the interior of Weinberg 33, it is still possible to see an irregularjoint running down the length of the belly and extending roughly from the forelegs toward the hindquarters(Fig. 33). Presumably,the two sides of the statue were united here and reinforced. The body thickness confirms this presumption, being thinner through the side (0.019 m), thicker through the underside (0.032 m), where clay was added by hand to reinforce the joint and to fill out the form. The interior surface of the back is rougher, an indication of modeling by hand. This would suggest the possibility that in the 6th century, at least, each wing and shoulder was built up as a unit in one or more molds; the breast was either made in another mold or was built by hand; while the belly and back were surely modeled by hand. At the same time, it is useful to observe the front of Weinberg 33, where the overlying fine clay has broken away (Fig. 33). Where this has occurred,the exposed surface of the coarse core is quite rough. It seems unlikely that it would have been so rough had it been pressed into a mold lined with fine clay, for surely its surface would have been smoothed by the act of pressing. As for the legs, those of the Classical sphinxes such as 7 and 8 were probablymodeled by hand (Figs. 20, 21, 22). Their forms arefairly simple, and their surfaces not well finished. The Archaic legs such as la (Figs. 3, 4), which are quite different, may also have been modeled free-hand. They are far more complex in shape, having a variety of ribs and grooves; there are few traces of tooling, and the surfacefinish is superb.If these had been made in a two-part mold, however, there would have been two vertical seams, and there are no traces of such joints on these legs. For support the legs were built up around round sticks or rods (Figs. 4, 22), which continued up through the shoulders to the exterior surface,as in sphinxes D and G from Kalydon.5' It is not clear whether the upper ends were subsequently masked. Presumably,these sticks were removed after the clay had become firm but before final drying and firing, to prevent cracking as the clay contracted. While it is possible-and I emphasize "possible"-that molds were used extensively in the 6th century,they were probablyused far less in the 5th century. Our best 5th-century example, 7, dated to the middle of the century,shows little evidence of the use of molds. Unfortunately, the head is missing,but among the remaininganatomicalparts-namely, neck, shoulder,breast,belly,back, hindquarters,and legs-the treatment of both interior and exterior surfaces and the varying wall thicknesses suggest the use of modeling by hand. In contrast to the interior surfaces of Weinberg 33, which seem to be fairly uniformly smooth, these parts of 7 vary considerably in the treatment of their surfaces. Around the neck and shoulders (Figs. 10, 12) large wads of clay,which were added to reinforce the juncture between the two, have been left rough. Further down, the inside of the breast shows deep diagonal furrows,where the modeler's fingers consolidated wads of clay while the clay was still pliable (Fig. 12). The interior surfaceof the back 7j is extremely uneven, the wall thick, probablybecause support was needed to counter the weight of the wings. Within the belly,

51. Dyggve 1948, pl. XXI:D, G.

CORINTHIAN

52. Weinberg 1957, p. 314. The edges of the holes, on both Weinberg's statue and 7, are chipped,and there is no preservedevidence of metal in them apartfrom the single hole on the right wing of Weinberg 33. In one case, at least, on 7g, the hole was cut through when the clay was fairlyfirm, for shallow rings on the sides of the channel have been left from the tool that was used. 53. Such partitionsoccurin the terracottawarriorfrom Olympia (Kunze 1941, p. 126, fig. 104) and in severalstatuesfrom the Sanctuaryof Demeter and Korein Corinth.

TERRACOTTA

SCULPTURE

39I

7e, the surface has been smoothed, and the clay is well consolidated. The wall is thinner at the top, to lighten the weight, and thicker at the bottom, where the greaterthrust must have fallen. The broken surfacesof the right humerus, 7c, show a series of cracks along vertical joints where clay was added in layers (Fig. 13, upper right), but this need not mean that a mold was used. The outer layer of fine clay is basically paper-thin, but it thickens where necessary to fill out irregularitiesin the underlying coarse core. This suggests that the fine clay was applied to a preexisting core. As for the wings of 7, the broken surfaces show that in some places the clay fabricwas well consolidated. Elsewhere, such as at the break on 7h (Fig. 17), snakes of clay seem to have been pressed together for the core of the wing, then faced top and bottom with a thin layer of coarse clay, succeeded by a paper-thin layer of fine clay on the surface. This fine clay surface coat is quite different from that on the Archaic wing 2c. The latter (Fig. 6) forms a thick crust on top of the coarse core, while the fine surface on the wings of 7 is so thin that irregularitiesare visible through it (Fig. 16). The outer edges of the wing fragments 7g and 7h are quite carelessly trimmed, and cutting marks are visible along parts of the feathers, in particular along the lowest primary feathers on 7e (Fig. 15). In general, more tooling is apparenton the surfaces of 7 than is customary in Archaic statues. These details of surface treatments on the various anatomical parts of 7 suggest that the statue was built up largely,perhaps entirely,by hand. Sphinx 7 is especially tantalizing because of the numerous preserved strut holes, described in the catalogue below (see the holes labeled H1-H6 in Figs. 11, 14, 15, 16, 18). Because the exact relation among the fragments of 7 cannot always be determined, the system and purpose of the struts remain unclear.In some cases, at least, sticks seem to have been used to strengthen the joint between two parts, such as the juncture between the left shoulder and the base of the left wing. Other preserved holes on 7 may have facilitated ventilation during firing. Although no evidence of external vent holes is preserved,one may have existed in the crown of the head. It is not entirely clear what purpose was served by the two holes that pierced the wing on 7g (Fig. 16). Weinberg identified similar holes on statue no. 33 as parts of lead mends.52Indeed, lead is preserved in one but it simply ends against the surface of the wing and doesn't appearto extend across a break.Perhaps the holes helped to reduce the pressureof the wind against the statue in its position high up on the roof. A feature that appearsin the two sea creatures,20 and 21, as well as in severalArchaic and Classical standing figures from Corinth and elsewhere, is a vertical interior partition, or internal wall.53In the case of 20, and also of 21 (Fig. 32), the partition runs down the long axis of the statue from top to bottom, giving support to the dorsal fins. No such interior support appears in the sphinxes, presumably because the greatest pressure was exerted by the head and wings. All of the statues were painted to some degree. Hair, facial features, scales, wings, and, in some cases, claws, were regularlypainted, while the hindquarterswere left plain. The fragments document the use of a simple three-color system of matt red, black, and the natural clay color, or, as in the instance of 7, a four-color system that included two shades of red.

392

NANCY

BOOKIDIS

Contrary to the view of J. V. Noble, the black was not simply the result of conditions of reducing in the kiln, but was without doubt made with manganese.54These colors were applied before firing and were fired with the statue, but whether in one or two firings is not known. Marked differences between the Archaic and Classical statues are apparent in their finish. On the former a thicker layer of fine clay is used to mask the coarse core, and surfaces are carefully smoothed and burnished. The leg la exemplifies the finest Corinthian workmanship in this respect; tool marks are virtually eradicated, and the surface shines like polished metal. Moreover, the Archaic statues are more carefully painted, and, in the case of wing lb, faint guide lines were incised before paint was applied. The Classical sphinx 7, by contrast, is thinly masked with fine clay; its surface preservesvery little polish, while fragments such as 16 have none. It is possible that these differences may be of some use in distinguishing Archaic sculptures from Classical. Finally, with regard to size, it is likely that most terracotta sphinxes were under 1.00 m in height. The complications of firing a figure much larger than that would have been considerable. Exceptions exist here too, however, most notably the colossal sphinx from Samos now under study by Aliki Moustaka, and possibly 17 below.

CHRONOLOGY

OF TERRACOTTA

SPHINXES

In recent years severalgeneral works on terracottasphinxes have appeared. Among these are Marilyn Goldberg's synopsis of her dissertation on Archaic acroteria (1982); Peter Danner's studies of acroteria on the Greek mainland (1989), in Western Greece (1997), and in Etruria (1993); and Aliki Moustaka's publication of the terracotta sculptures from Olympia (OlForschXXII, 1993). In addition, Marie-Franyoise Billot's lengthy publication (1977) of the Corinthian sphinx from Thebes, now in the Louvre, provides a wealth of information on the subject. It therefore seems to be a fitting time to review the subject in general with the Corinthian material in mind and to examine certain aspects of chronology. Both Billot and Goldberg providereaderswith lists of known terracotta sphinxes together with the relevant bibliography. Only a few additions need to be made to these lists. A paw found in the early excavations of the Aphaia Sanctuary was attributed by Furtwangler to the Propylon.55The headless body of a small sphinx from the Colonna Hill excavations on Aigina has now been published.56To the wing and the knee from Kalapodhi can be added a face.57A similar face was found at Rhamnous in the area 54. See Bothmer and Noble 1961, pp. 20-22. Tests to determinewhat pigments were used as coloring agents were not performedon the Temple Hill pieces, but manganesewas identified on one black-paintedfragmentfrom the Sanctuaryof Demeter and Kore, SF-69-21. 55. Furtwangler1906, p. 385,

no. 123, fig. 318. Also publishedby Van Buren (1926, p. 175, no. 21). 56. Alt-Agina II, 4, pp. 13-15, figs. 3-4, pl. 9:a-c. 57. For the wing and knee, see Felsch and Kienast1975, p. 21, fig. 30. The face appearsin Felsch, Kienast, and Schuler1980, p. 114, fig. 103. See also Hiibner 1990.

CORINTHIAN

TERRACOTTA

SCULPTURE

393

around the temples.58Its similarity to the Kalapodhi face makes likely its identification as a sphinx. From the Heraion at Samos come fragments of a colossal terracottasphinx, mentioned above.59 Despite this growing corpus, the number of terracotta sphinxes that can be dated on grounds other than style is relatively small. The reasons are several.In most cases the statues are extremely fragmentary,like those from Temple Hill. Virtually none come from a datable context, and only a few can be assigned to specific buildings as acroteria,thereby gaining some chronological reinforcement from the date of the architecture.In the case of Athens, of Delphi, and, to a considerable extent, of Corinth, the fragments derive from late secondary fills. According to the evidence preservedthus far,the earliest sphinx acroteria have not been found in Greece but in Etruria, and date to ca. 600 B.C.60 On the Greek mainland, the earliest terracotta sphinx acroterion is apparently one from Olympia that was tentatively associated by Aliki Moustaka with the so-called "friiheblattstab"roof and dated on stylistic grounds to ca. 580 B.C.61 That roof has now been attributedto the Treasury of Syracuse.62More important for our purpose is the terminusante quem provided by the discovery of one fragment of that roof in a context dating no later than 560 B.C. It is tempting to see the Corinthians as the transporters of the type from Etruria to Syracuse and back to Greece.63But although Aliki Moustaka attributes the Olympia sphinx to a Corinthian or northeast Peloponnesian workshop, the roofing system is strictly Sicilian. Apart from the Olympia sphinx, and the items catalogued below that are attributed to the Temple of Apollo at Corinth, perhaps the only other terracottasphinxes that can be assigned with some security to the roofs of specific mainland buildings are those from Kalapodhi64and Kalydon.65 The two temples at Kalapodhi were initially assigned to ca. 580 B.C., based on the discoveryof Middle Corinthian sherds in the construction fill. More recently,that construction date has been lowered to the second quarterof the 6th century because of the discovery of Late Corinthian sherds in the terracing.66The presence of Late Corinthian sherds brings those temples closer in time to the Temple of Apollo at Corinth, whether the higher or lower chronology for Late Corinthian is followed. 58. For the most recent illustration see Petrakos1987, pp. 300-301, pl. 55:a.Yet anothersphinx face from Syracusehas much in common with the faces from Rhamnousand Kalapodhi;see Danner 1997, A73, pp. 34-35, pl. 9:2. 59. Buschor 1957 and OlForsch XXII, p. 108. A head from Dyrrachium (Durres),Albania,may also belong to a sphinx,but its resemblanceto two antefixesfrom Apollonia throws this identificationinto question. See Eggebrechtet al. 1988, no. 109, p. 240, Durres AM 4861; Homann-Wedeking 1942, col. 370, fig. 44. The date of ca. 580-570 B.C. for the head from

Dyrrachium,given by Eggebrechtand by Holtzmann (1991, p. 149, note 95), seems too high. I would agreewith Moustaka (OlForschXXII, p. 109), who placesit in the third quarterof the 6th centuryB.C. 60. Danner 1993, pp. 99-100, fig. 14, the ridge acroterionfrom Murlo. 61. OlForsch XXII, group I, pp. 104108, 159. 62. OlForsch XXIV, room 37, pp. 83-87, especiallyp. 86, where Heiden questionswhether a roof with a continuous sima would have had corner acroteria. 63. See Williams 1980.

64. The sphinx from Kalapodhiis preliminarilypresentedin Felsch and Kienast 1975, p. 21, fig. 30; Felsch, Kienast,and Schuler1980, p. 114, fig. 103. 65. Danner (1989, pp. 20-22) disassociatesa numberof stone sphinxesfrom temples to which they had previouslybeen assigned. 66. See KalapodiI, pp. xvi, 232-235; Hiibner 1990, p. 168. Hiibner suggests that the sphinx may have been reused from an earlierbuilding, dated to the 7th centuryB.C., but since that roof was of reeds,it is difficult to imagine how a terracottasphinx could have been mounted on it.

394

NANCY

BOOKIDIS

From Kalydon there are remnants of three or four terracottasphinxes in at least two different sizes.67These have been assigned to one or to two temples.68Relying on the evidence of the well-known sphinx head H from Kalydon, now in the National Museum at Athens, scholars have regularly dated these sphinxes on stylistic grounds without regardfor their architectural context. The torso G can be securely associated with the so-called lion-sima roof, the profile of which is closer to the sima on the Temple of Apollo at Corinth than to the cavetto simas of the 580s, the date commonly assigned to the head. Statues G and H show stylistic similarities to each other, at least in comparison of the same anatomical portions. The face of H also has certain elements in common with the sphinxes from Spata and from the Kerameikos,the latter dated ca. 550 B.C. by Harrison,69 and with the stone sphinx found in the North Cemetery of Corinth.70 While the profiles are different, the overall structureof the faces, the integration of the features, and the modeling of the ears seem similar.I would therefore place all of the above examples near the middle of the 6th century. Such a date would also agree with Nancy Winter's placement of the lion-sima roof at ca. 550-540 B.C.71

STYLISTIC DEVELOPMENT

OF SPHINXES

Lacking the assistance of architecture,one must turn to style as a criterion for determining the chronology of terracotta sphinxes. In so doing, one must also considerthe more abundantand better-preservedstone examples, while keeping in mind the differences between the two media. One example may serve to illustrate this point. A consistent convention in terracotta sphinxes of the 6th century is the sharp definition of the chest from both shoulder and humerus, with the chest recessed from the shoulders and the joint generally emphasized by a raised band (Fig. 33). This sharp definition does not appear in stone.72It may begin to disappearin clay in the late 6th century, as exemplified by the sphinx from Halai,3 and is certainly gone by the later 5th century, as is illustrated by sphinx 7 (Fig. 67. Sphinx G (Dyggve 1948, pl. XXII) is the smallest;the torso D-E is larger.According to Dyggve (1948, p. 177), the breastF is about the size of D, althoughits preserveddimensions are smaller;but accordingto Rhomaios (1951, p. 40) it is smallerthan D. While Dyggve states that the head H is the size of D, Rhomaios notes that its neck is 0.02-0.03 m largerthan that of D. At present,H is mounted on F, which seems too small for it. Two more fragments-hair and ear Al and breast Bl-are problematic,both having been assignedto a sphinx or a kore.In supportof the latter attribution,made by Dyggve, are the extensivepainted bordersaroundthe neck, which would be unique for a sphinx. Recently,JoachimHeiden (1987,

pp. 66-67) has reassessedthe fragments. He divides them into two sizes, G and F being smaller,D-E larger. According to him, the head H fits onto the body F "breakfor break." 68. In favorof a single temple are Rhomaios (1951, p. 45), Stucchi (19521954, p. 36), and Danner (1989, p. 48). In favorof two buildings are Dyggve (1948, pp. 222-224,235-236) and Winter (1993, pp. 128-130). 69. AgoraXI, p. 12. For the two sphinxessee Richter 1961, no. 12, p. 16, figs. 40-41, and no. 16, p. 17, figs. 5457. 70. Protonotariou-Deilaki1973 and Alt-Agina 11,2, p. 55, figs. 64-65. 71. Winter 1993, pp. 125-130. Winter places the Kalydonroofs into a separatenorthwest Greek system.

72. Only a few stone sphinxesdisplay a featureresemblingthe articulating band:an earlyone from the Metropolitan Museum, Richter 1961, no. 2, p. 10, figs. 8-9; that from the Ny Carlsberg Glyptothek, Richter 1961, no. 3, pp. 10-11, figs. 10-15; the Kerameikos sphinx, Richter 1961, no. 11, pp. 15-16, figs. 34-39; the laterlimestone sphinx in the MetropolitanMuseum, Richter 1961, no. 14, pp. 16-17, figs. 46-49. But in all cases, these bands or incised lines are merelydecorative,because there is no distinction in modeling between chest and shoulder.Neither of the stone sphinxesfrom Corinth preservesthis feature. 73. Goldman 1940, no. 5, p. 443, fig. 115, torso with incipient breasts,if this is indeed a sphinx.

CORINTHIAN

74. Richter 1961, nos. 1-3, pp. 1011, figs. 1-15. 75. Richter 1961, no. 11, pp. 15-16, figs. 34-39. 76. Comparethe photographin Richter 1961, fig. 51, with Wright 1977, pl. 57. In the latterphotograph the legs have been lengthened slightly, and there is a correspondingrise in the angle of the body to the ground. 77. Richter 1961, no. 37, pp. 27-29, figs. 96-103 and no. 38, p. 29, figs. 110-114. 78. Raepsaet1982;Trianti 1975. 79. OlForsch XXII, J12-16, pp. 108109, 113-114, pls. 89-92. 80. Goldman 1940, p. 443, fig. 115. 81. See Richter 1961, no. 12, p. 16, figs. 40-41, Athens NM 28; its head from the top of its polos to the chin is 0.205 m high; the breastis 0.198 m high and 0.202 m wide. 82. For the Piraeussphinx see Richter 1961, no. 19, p. 18, figs. 64-65, Athens NM 76; its head height is 0.25, the chest proper0.196, chest width 0.24 m; regrettably,I do not have the height of the chest to the elbow. 83. Richter 1961, no. 16, p. 17, figs. 54-57, Athens NM 2891; head, and breastto the elbow,are 0.19 m high, while the breastalone is 0.156 m high.

TERRACOTTA

SCULPTURE

395

10), but previously it was the rule. Undoubtedly, the reason behind this band is a technical one, as has been described above. One cannot, however, ignore the evidence of stone, particularlysince stone statues aremore abundant and better preserved than clay ones. Although statues of animals are extremely difficult to assign to a date, certain basic changes can be observed to have occurredin the course of the 6th century.These changes can be seen, in both terracottaand stone sphinxes, in the profile of the shoulders and chest and in the angle of the body and wings. A quick review of the stone sphinxes depicted in Richter's Archaic GravestonesofAttica shows the following changes in features.The earliest Attic sphinxes have large and very round chests, the maximum projection tending to be low. Legs aremassive and strictlyvertical,the body lies nearly parallel to the ground, and the wings stretch out horizontally before curving upward.74By the time of the Kerameikos sphinx in the middle of the 6th century,the profile of the chest has flattened somewhat, although the general outline is still large.75At the same time, the position of the body has become more upright-here because the sphinx is crouching with its hindquartersoff the ground-and the wings rise at a slightly sharperangle before turning up. This steeper angle is apparent in the stone Corinthian sphinxes from the North Cemetery and the Forum area,76but the chest is still round, the legs massive. Change occurring in the third quarterof the 6th century is exemplified by the Metropolitan sphinx, which is associated with the boy-and-girl stele, as well as by the Boston sphinx.77In both cases the chest is round but the curvature is more compressed, the maximum projection higher; the legs now angle back somewhat from the vertical, and they are more slender and somewhat longer, especially between the carpals and the paw. In addition, the angle of the wing is steeper. By the late 6th century, the wings are steeper yet, as evidenced by the marble sphinx from Colonna Hill, Aigina.78The chest has flattened out, becoming deeper once again and making way for the next stage, as incipient breasts begin to break the otherwise smooth contours of the chest. How do terracotta sphinxes fit into this progression?The chests of Kalydon statues G and D are relatively boxy, their profiles describing a flattened curve that is in keeping with the first half of the 6th century. This profile is quite different from the tightly curved chest of Weinberg 33, whose figure fits better within the third quarterof the 6th century than within the last quarterof the century, as Weinberg had dated it. A profile similar to that of Weinberg 33 can be found in sphinx II from Olympia.79 Halai torso no. 5 from the late 6th century also has a boxy chest with the That of sphinx 7 below has a relatively straight suggestion of breastsA80 profile with a slight compound curve that again suggests incipient breasts. Another aspect to be investigated as an indicator of the chronology of the sphinxes is that of relative proportions. A very rough proportion that appears to apply to 6th-century sphinxes is an approximate equivalence between the height of the head without the neck, the height of the chest from shoulders to the top of the leg, and the length of the leg from top to metacarpal joint. These proportions hold for the Corinth stone sphinx from the North Cemetery, and also for the marble sphinx from Spata,81if one includes the full height of her polos. In addition, this canon works for the sphinx from Piraeus,82from the Themistoklean wall,83 and also for

396

NANCY

BOOKIDIS

Thermon sphinx D,84 if one measures the chest to the elbow joint at the base of the covert feathers. These proportions do not apply to the 5thcentury marble sphinx from Aigina: the height of the chest here equals the combined height of head and neck, hence the apparent enormity of its chest.85It is unfortunate that the head and complete legs are not preserved on 7 from Corinth to provide a point of comparison. Marilyn Goldberg suggested several technical criteriafor sorting Archaic from Classical sphinxes.86In the former,the holes for rods in the legs are large, the claws are rendered only in paint, and the inside surfaces of the wings areunpainted. In the latter,the holes are narrow,the claws plastic, and the insides of the wings are painted. These distinctions are maintained in 7 and in Weinberg 33, where the holes left by the rods are, respectively,0.006 and 0.015 m thick, but they are not so clear in the case of 8. If correctly assembled, statue 8 combines a large hole (Diam. 0.02 m) with plastic claws. Moreover, the inner faces of the wings of the Kalydon sphinxes, which Goldberg was unable to see, are painted. An interesting feature of many Corinthian sphinxes is the treatment of the wing. The Corinthian sphinx wing is upright, its outline making a fan-shaped curve (Fig. 33). In contrast to many Attic sphinxes, whose wings have a smooth lower edge,87the Corinthian feathered wing has a scalloped outline, starting from the body and continuing to the tip. This form is apparentin the poros sphinx from Corinth, a small bronze sphinx from Perachora,88Weinberg 33, and statue 7. It can be found as well on a limestone sphinx in the Metropolitan Museum that is said to have come from Attica, but which is stylisticallycloser to Corinthian examples.89That this feature appears on the locally made terracotta sphinx from Colonna Hill, Aigina, may be a sign of Corinthian influence.

CONCLUSIONS A few years ago, in attempting to characterizeArchaic Corinthian sculpture from the very fragmentary material that has survived, I concluded that it could best be described as conservative.90This conclusion also applies to the sphinxes considered here. Despite the poor condition of many of these fragments, it is possible to see that naturalismwas often sacrificed for pattern. Sphinx 7, made in the third quarterof the 5th century, looks back to Archaic models, despite its more developed anatomy. Whether this conservatism is an indication of stylistic preference or of conscious adaptation of the statue to its Archaic setting is difficult to assess without more material of this date. Our uncertainty notwithstanding, this conservatism led to the creation of very satisfying sculpture in which a pleasing balance was struck between the body and the large, fan-shaped wings, between form and decoration.This balance is apparentin the poros sphinx from Corinth, which differs dramaticallyfrom contemporary Attic statues.91It is perhaps best expressed in the small terracottasphinx, Weinberg 33 (Fig. 33), which seems to hang in the air.Together with technical excellence, this is what Corinth could offer to the Greek world. That these qualities were appreciated is shown by the predominance of Corinthian terracotta sculpture in cities outside of Corinth.

84. Dyggve 1948, pl. XXII. 85. Alt-Agina 11,2, no. 52, pp. 80, 118-120, pls. 35, 38-40. 86. Goldberg 1977, p. viii. 87. See, for example,the Kerameikossphinx,Richter 1961, no. 11, pp. 15-16, figs. 34-39. 88. Payne et al. 1940, p. 135, pl. 43:1-2; Alt-4gina 11,2, p.58, figs. 66-67. 89. Richter 1961, no. 14, pp. 1617, figs. 46-49. 90. Bookidis 1995. 91. Wright 1977; Richter 1961, figs. 50-53.

CORINTHIAN

TERRACOTTA

397

SCULPTURE

CATALOGUE I.

FINDS

FROM

THE

EXCAVATIONS

ON TEMPLE

HILL

The catalogue of the fragments is organized as follows.92Fragments at-

tributedto the same statuehavebeen groupedtogetherunderone catalogue entry.Those fragmentsthat are less securelyassociatedhavebeen assigned consecutive catalogue numbers.93Dimensions, given in meters, are primarily intended as an indication of the size of a given piece. The size of a sphinx cannot be restoredfrom the proportions established by the width of its breast or the length of a feather. An attempt has been made to approach something like a consistent nomenclature for the various parts of a sphinx. Given the fact that the sphinx was a mythological beast and that the author is not a zoologist, the results are undoubtedly far from correct.Terms used here include: shoulders; humeri for the upper part of the forelegs that are, in a sense, "encased" in the body; radius and ulna for front and back bones of the forelegs; carpalsfor the joint at the base of the radius and ulna; metacarpalsfor the section between the carpals and the toes; dew-claw for the vestigial fifth toe; thigh for the fleshy upper part of the leg; and hock for the joint between upper and lower leg bones. In addition, scapularsdenote the short feathers on the body, coverts the feathers at the base of the wing, primaries the long flight feathers.94Where possible, the catalogued entries are reproduced at 1:2 in order to make clear their relative sizes; reproductions of a few details are included at a smaller scale. 1

92. Since the mudstoneinclusions which were addedto the clay have been describedin some detail above,we will not repeatthe descriptionshere. In addition,the valuesof the Munsell soil color charthave been translatedinto words commonly employedfor Corinthianclay,since the Munsell words areless graphicfor one who hasn'tthe book in hand. 93. When the year of discoveryis known, its final two digits are cited afterthe identifyingletters, and are followed by a serialnumber,e.g., SF-76-1. Where the year of discovery is not known, a simple serialnumberis used. In addition,a few fragmentsof 7, too small to inventory,have been given numberswithin their storagecontext, e.g., lot 72-240-8. 94. See Ryder1969, fig. 1, skeleton of a dog (mammal).

Left front leg, wing of sphinx

a. SF-74-la. P.H. 0.138,W. at top 0.043, D. at top 0.068, W. at bottom 0.032, D. at bottom 0.053, Diam. hole 0.006. Three joining fragments, upper part of left foreleg, broken at top and bottom. b. SF-74-lb. P.H. 0.058, W. feather 0.034, p.Th. 0.043. Wing, preserving front face and lower edge of primary feathers, broken at back.

Figs. 3-4 EarlyRomanquarryfill with Turkishintrusions,quarrytrench4, lot 74-178. Clay:fine surfacelayer,Th. 0.002-0.008, firedgray-buff2.5Y 7.5/2; core:few smallblackinclusions (chiefly0.002), rarefine white inclusions,firedto a pinkergray-buff lOYR7/2. Paint:red-brown,now faded;black.

la derives from the left foreleg above the carpals.The anatomical rendering of the leg is highly stylized. In addition to the broad horizontal curve of the radius and the constricted curve of the ulna, a series of rounded and sharp ridges along either side seem to have no basis in feline anatomy. On either side, a rounded protrusion runs the length of the leg and gives it a diamond-shaped plan in horizontal section. On the outside of the leg (Fig. 3) a narrow rib sets this off from the ulna; on the inside of the leg (Fig. 3) two sharp ridges separate the protrusion from the ulna. Nearly solid, the leg was built around a square rod 0.006 m thick (Fig. 4). A highly lustrous layer of very fine clay 0.002-0.008 m thick masks the coarse core. lb preserves a small portion of a rather large and extremely thick wing, namely, parts of two stepped primary feathers and the bottom

398

NANCY

inside

BOOKIDIS

outside

Left leg la

Wing lb

edge of the wing (Fig. 3). This edge is not wholly straight but flares slightly toward the left break, suggesting that the fragment may have lain near the body. If so, then the wing belonged to the left side. Surrounded by a reserved stripe, the centers of the feathers are painted alternately matt black and red-brown, beginning at the bottom with black. Along the outer edge of each feather is a stripe in the same color as the center. Only the outer edge of the wing is not painted. Faint guide lines, incised before firing, delineate the various painted stripes.The one complete feather is 0.033 m wide. Wads of coarse clay form the core of the wing, while the feathers are executed entirely in fine clay, the thickness of which varies from 0.003-0.006 m. A thick layer of fine clay masks the outer edge of the wing.

Figure3. Sphinx1. Scale1:2

These two fragments have been associated on the basis of the distinctive color of their clay. They are considerably larger than corresponding fragments of Weinberg 33. Little can be said about the wing, apart from the careful execution of the primary feathers, but parallelsfor the leg can be found within the Corinth collection. Two sets of legs published by Weinberg as 33b95and 39 (see Fig. 4)916 are virtually identical to la in modeling and execution.917All three sets of legs exhibit the same stylized anatomy and highly lustrous finish, representing Corinthian work at its very best. In addition, Weinberg 33b-c and la are very close in composition of clay and fired color. That they are not from the same statue is shown by the difference in execution of the ridges that run down the inside of each leg. In addition, the fine clay surface of Weinberg 33b-c is far thinner than that of la and does not completely cover the coarse core. Whereas la is built up around a square rod, a round rod was used in the modeling of Weinberg 33b. Of the three, la is the largest, and Weinberg 39 the smallest, though the differences are not very great. More important is the fact that there were three such sphinxes at Corinth.

95. Weinberg 1957, pp. 314-315, pl. 72, SF-31-2b-c. 96. Weinberg 1957, p. 316, pl. 72, SF-32-12. 97. The light tan surfaceof Weinberg 39a-b lies roughlybetween lOYR 7.5/4 and 7.5YR 7.5/4 on the Munsell chart.

TERRACOTTA

CORINTHIAN

399

SCULPTURE

0

0 la

0 Weinberg 39

Weinberg 33b

0

8c Figure 4. Rods used in modeling sphinx legs (in cross-section)

98. The field notebook recordsthe discoveryof the torso,wing fragments, and front right paw in the areato the northeastof Temple E, roughlybeneath the presententrancehall to the Museum, but no mention is made of legs. On the other hand, the notebook accountof the well nearTemple E, which producedmuch of the sculpture publishedby Weinberg,mentions at least two sphinx legs. One of these is Weinberg 39; it is tempting to identify Weinberg 33b, c as fragmentsof the second. Indeed, in fired color,and size and color of inclusions they closely resemblethe chest of a sphinx from that same well, Weinberg37 (1957, pp. 315-316, pl. 72, SF-32-11). This second chest is largerin scale than Weinberg33a, its preservedwidth being 0.089 in contrastto 0.055 m for the latter. 99. For the date of Weinberg 33, see above,p. 395.

7m 5 cm

Weinberg associated the legs 33b and 33c with the torso and wings, 33a, on the assumption that they were found together, but they were not.98 Moreover, the legs are slightly grayer (5Y 7.5/2) in fired color than the body (near 2.5Y 6/4), and are larger in scale; furthermore, the stumps of the legs preserved on 33a do not suggest a lower leg with the contours of 33b or 33c. If the legs are disassociated from the body, they are then free to move in date,99but how much or, indeed, whether at all cannot be determined. Regrettably,we know very little about the evolution of sphinx legs because relatively few have survived either in stone or clay. The legs of sphinx D from Kalydon (Dyggve 1948, pp. 176-177, pl. XXII; Rhomaios 1951, pp. 40-41, figs. 23-24) and the earliest sphinx at Olympia, sphinx I (OlForschXXII, J8, J9, p. 112, pls. 87:b, 88:c, 92:a), both sphinxes dating to the first half of the 6th century,are more simply modeled than ours are. A better parallel is the later sphinx II from Olympia (OlForsch XXII,J13, pp. 113-114, pls. 91:c-d, 92:d), assigned to the third quarter of the 6th century B.C. Although not as stylized as ours, it is leaner and bonier than its predecessor at Olympia, has multiple ridges running down the inside, and is similar to 1 in horizontal section. If we are correct in associating our sphinx with the Temple of Apollo, then the Olympia parallel corroborates a date in the mid- to third quarter of the century. Mid- to third quarter 6th century B.C.?

400

2

NANCY

Head, right and left wings of sphinx

a. SF-26-1. P.H. 0.073, Th. 0.026. Right half of forehead. From the Athena Trench, in Roman fill north of Building 5. b. SF-74-3. Max. p.dim. 0.073, Th. 0.026-0.028. Hair over the forehead, broken on all sides. Medieval fill overlying quarry, basilica trench 1, lot 74-149. c. SF-6. Max. p.dim. 0.25-0.26, Th. 0.045 (lower tip)-0.064 (top right). Right wing, from the scapulars to the base of the primary feathers. Exact provenience within Corinth unknown, but found prior to 1915. d. SF-74-2. P.H. 0.063, Th. 0.030-0.032. Tip of right wing. Early Roman quarryfill with Byzantine intrusions, quarrytrench 1, lot 74-88.

BOOKIDIS

Figs. 5-6

Clay:fine surfacelayer,Th. 0.004-0.017; core:commonfine blackinclusions(< 0.001). On thick wings,centrallayerwith coarser temper(to 0.002) betweenlayersof fine temper;firedtan 10YR7/4 throughout,exceptfor light pink core of b whereverythick,and 10YR7/5 surfaceof skin (a, b). Paint:black near1OR2.8/1; red-brownscapulars 1OR4/4, primaries2.5YR 4/4. Published. a, c: CorinthIV,i, p. 113, Ml, andfig. 46:a;Koch 1915, no. 6, p. 82, fig. 38;Weinberg1957, nos. 4, 34a, pp. 305, 315, pls. 64, 72.

These four fragments have been grouped as part of one sphinx because of their fired color, painted decoration, and distinctive use of very fine temper.Together they preserve a very small portion of the head and parts of one wing of a large-scale sphinx. Two of these fragments (b, d) belong to the recent excavations on Temple Hill, and two (a, c) were catalogued by Weinberg. As Weinberg has described it, a preserves the forehead, crowned by four wavy locks, just to proper right of the center of the head (Fig. 5). The locks are rendered as parallel strands that are round in section. They frame a forehead that is relatively high and triangular.A broad brownpainted stripe demarcates the eyebrow.Below, the fragment breaks at the line of the eyeball. As the break indicates, the eyeball curved in more deeply at the inner corner,coming closer to the surface at the outer corner.If we assume that the center of the head's height fell through the center of the eyes, then we can estimate that its full height was roughly 0.17-0.18 m, or nearly twice that of the head of the Louvre sphinx (Billot 1977, p. 405). Fragment b preserves seven rows of parallel,wavy strands of hair (Fig. 5). A tiny patch of skin is visible in one indentation, indicating that this piece, too, lined the forehead. Although the exact relation of b to a is uncertain, the greater number of locks of hair on b place it closer to the left ear. On both fragments, the skin is reserved and lustrous, while the hair is painted matt purplish black. The head is hollow. Somewhat perplexing for the relative placement of these two fragments is the fact that the interior of a is irregular,with clearly superimposed wads, while that of b is far flatter.This flatness may be a further indication of the position of b on the side of the forehead.The hair is worked entirely in fine clay,which varies in thickness from 0.004 to 0.01 m. Fragment c preserves part of the right wing from the scapularsup to the base of the primary feathers (Fig. 6). Behind the scapularsis a

CORINTHIAN

TERRACOTTA

SCULPTURE

40I

Head 2a Head 2b

Rightwing 2d

2d,lowerbreak(1:1)

Figure5. Sphinx2. Scale 1:2except as noted

narrow portion of the shoulder,which is plain except for a single drop of uarudacruaro(Da.001)nolgethr.A ut.4m black paint from the hair.The scapularfeathers consist of vertical rows of scales, alternately dark red-brown and black, each outlined by a reserved band and a black stripe. A plastic border divides the scapulars from the coverts.This border consists of a red band that is decorated with a row of conical depressions and is framed by deep, unpainted grooves and two black ribs. These depressions, left the reserved color of the clay,were never filled with any substance. A broad V-shaped groove and a narrowplastic black band articulate the outer edge of the coverts. At present the surface of the coverts is unpainted and lustrous but it was once covered by a matt cream-colored paint that appears to overlie a thin wash of white slip. If this is indeed white slip, then we have the earliest attested use of this substance as a base for overpainting. Modeled like the coverts, with deep groove and raised framing band, the primary feathers are cut back slightly from them. Moreover, while the coverts are all on the same plane, the primaries project in low relief, one above the other. Their centers are painted alternatelylighter red-brown and matt black within the usual borders of reserved and black stripes, beginning with black at the lower break.The wing has probablybroken near its bottom edge, for the fine clay used for the surface turns down at the end, much as on lb and 2d. Viewed from the back, c breaks at left, at the wing's j'unctureto the bodyi-(Fig.c 6). Just- above this, isza niproetingr mass o-fca that waszbuil;t

402

NANCY

BOOKIDIS

deep, the hole left by the rod cuts midway through the thickness of the wing; its sides are smooth and its floor consists of a plug of clay (Fig. 6). The remainder of the back of c is smoothed, but toward its lower tip can be seen a bit of lustrous slip, better preserved on the back and outside of d. The feathers of c are modeled in fine clay and are 0.004-0.012 m thick. Since no sharp cut marks are visible in the grooves that define the various feathers, the wing was probably made in a mold, with coarse clay then pressed in behind it. Fragment 2d (Figs. 5, 6) derives from the tip of the wing and preserves the end of one red primary feather as well as the black outline of the next one. The outer edge of the wing is flat, the transition to the flat back rounded. Both the edge and back are unpainted but somewhat lustrous. When Weinberg published c (Fig. 6; Weinberg 1957, no. 34b, pl. 75), he associated with it two more fragments of a left wing (Weinberg 1957, no. 34b-c, p. 315, pl. 72), on which the primary feathers were rendered only in paint. All three fragments show the same use of temper: a layer of clay with very fine temper on either side of a central layer with coarser temper. Weinberg attributed c to the main, right side of the sphinx, the other two to the secondary left side. Weinberg is undoubtedly correct in associating c with the right side, for the absence of hair on the preserved shoulder suggests that the head was turned to proper right. But a comparison of the two fragments of the left wing shows that their painted feathers taper in opposite directions. Thus, unless the flight feathers tapered toward both their bases and their tips, we must assume that these fragments preserveboth a right and a left wing. They differ slightly from the molded wing in fired colors and are somewhat thinner.We have thus disassociated them from this statue, despite the similar arrangementof the temper. Having now associatedyet a third fragment with them, we have included them below as lla, c. For the dating of this statue, there is little, apart from the hairstyle, that can usefully be compared to other works. Characteristicof the hair is the steeply descending wave that creates a high, triangularforehead. The wave is typical of Corinthian protomes on pottery from the Middle Corinthian phase and on, and in particular,Middle Corinthian III (see Amyx 1943 for a survey). A relatively close parallel can be found in the sphinx from Thebes, now in the Louvre, as Weinberg had observed. But there are also differences. Fewer waves frame the forehead of 2a because they are larger and looser in format. Moreover, they do not impinge as closely on the eyebrow as do the waves of the Louvre head. Weinberg had placed 2a after the Louvre sphinx, presumablyon the assumption that the stylistic trend moved from tighter to looser waves, but the opposite seems more likely. Nor is the high triangularforehead a feature of Late Archaic style. At the lower end of the chronological spectrum are pieces such as the head of a terracottaNike from Delphi, dated to the early 5th century B.C. by Ducat (FdD II, pp. 235-236, Nike1,1, pl. 87). Here the waves are very tightly crimped, standing out in slight relief from the head. A central

CORINTHIAN

TERRACOTTA

SCULPTURE

403

part is no longer used, and the hair instead forms a continuous band circling the forehead. At the opposite end of the chronological spectrum are pieces like the Wix head in Copenhagen, most recently attachedto a sphinx body and dated ca. 570-560 (Holtzmann 1991, esp. p. 142). A marble head of a kore from Rhodes, now in Copenhagen (Richter 1968, no. 77, p. 53, figs. 244-247), is of about the same date. On both of these heads the descending waves are pulled out in a more horizontal direction to create a rounder forehead. Marie-Francoise Billot has proposed that the date of the Louvre sphinx be lowered to ca. 525 B.C. (Billot 1977, pp. 406-418).1?o Our head should then fall somewhere between it and the earlierpieces. A date early in the third quarterof the 6th century is suggested for 2a. Weinberg, however, attributed the right wing to the first half of the 6th century on the plausible grounds that such careful work and stylization of the feathers with the relief bands and dotted border should be earlier rather than later. Dotted borders are paralleled elsewhere in clay. There are three fragments from Delphi, two of which are only painted (FdD II, no. 37, p. 256, pl. 93; no. 39, p. 257, pl. 93; no. 44, pp. 259-260, pl. 94). A dotted border appears on Kalydon sphinx D, with the difference that the holes pierce the left wing as if something had been pinned there. Yet another example is the small terracotta sphinx from the Sanctuary of Apollo on Aigina (Alt-Agina 11, 4, pp. 13-15, figs. 3-4, pl. 9:a). While in general avoiding dates for the more fragmentary material from Delphi, Ducat supportedWeinberg'stenets, putting reliefwork before simple painting (FdD II, p. 263). Both Kalydon D and the sphinx from Aigina are dated ca. 560 or 560-550 B.C. Yet, if the Corinth wing is correctly associated with the head, then its date must be lowered. Early third quarter 6th century B.C. 3

Right foreleg of sphinx SF-76-5. P.H. 0.063, p.W. 0.042,Th. wall 0.012-0.018. Nearly halfthe circumference of the rightleg with dew-claw,brokentop and bottom.EarlyRomanquarryfill, quarrytrench12, lot 1976-180.

100. Although I accept the general parallelsBillot drawsto Acropolis 679 (Peplos kore),I do not agreethat the profilesof the two heads are close. Characteristicof the two is the way in which the cheeks are defined from the lower half of the face by a slight concavitythat extends from the nose to nearlythe ears.But the renderingof the sphinx'seye, the shape of her forehead, and the painted lotus chain on her stephaneseem to be more conservative elements.I would thereforeplace the Louvrehead before the Acropolis head, in the 530s.

Fig. 7 Clay:fine surfacelayer,Th. 0.002-0.007, firedyellow-green5Y 7.5/3; core:commonfine to medium blackinclusions(0.001-0.003), fired gray-green,near5Y 7/2.5. Paint: black.

A small fragment derives from the inside and back of a right foreleg. At the lower break is preserved the dew-claw, which is modeled in relief and painted matt black; a single black line is painted above the claw. Although the leg is basically cylindrical, the back is lightly facetted to create a faint ridge down the axis. The leg is hollow and was built up around a rod 0.012 m in diameter. Over the coarse core is a thick layer of fine clay that varies in thickness from 0.002-0.007 m and is lustrous. It is this clay from which the claw is modeled. This is a perplexing piece, for its scale cannot be easily estimated. Its circumference appears small and suggests that the leg was considerably smaller than would be suitable for 7. At the same time, both the carpals and the bone that normally projects at the back of the carpals are missing. If Weinberg 33b is used as a parallel, then this bone should appearjust

NANCY

404

BOOKIDIS

cm

_5

lef bra

2d and 2c

_25

iue6

phn

.Rgtwnc

Section

Section

through shoulder

through strut

TERRACOTTA

CORINTHIAN

SCULPTURE

p - ~_

Stephane4b

Stephane4a

Right foreleg3

Rightwing tip 5

Figure7. Sphinxes3-6.

Scale

1:2

Right hock6

405

406

NANCY

BOOKIDIS

above and behind the dew-claw, while the carpals should be just above it. On Weinberg 33b the carpalslie 0.02 m above the dew-claw, but on 3 the leg continues 0.04 m above the dew-claw, with only the faintest flare to indicate a possible thickening of the leg. The leg may therefore either have been larger than it appears or have been very simply modeled. A date in the 6th century is suggested by the surface treatment and the use of black paint on the dew-claw. 6th century B.C. 4

Stephane, forehead of sphinx

a. SF-33-1. PH. 0.067, H. stephane 0.03, Th. through forehead 0.018. Stephane and hair over forehead; upper half of stephane badly chipped. Found in 1933 in Late Roman or Byzantine fill along the southwest edge of Temple Hill. b. SF-75-5. PH. 0.045, p.W. 0.064. Front and top of stephane, broken all around;upper half of stephane broken. Early Roman fill,

Fig. 7 quarrytrench 8, lot 75-184. Clay: fine surface layer,Th. 0.002, fired pale yellow 2.5Y 8/3; core: frequent fine to small (to 0.002) well-mixed black inclusions, fired to a gray-white, or slightly greener than 5Y 7.5/1.5. Paint: brown-black 7.5YR 4/2; light red hair, near 5YR 6/5. Published:a:Weinberg 1957, no. 42, p. 316, pl. 72.

Both fragments derive from a female head wearing a stephane. In addition, a preserves a bit of the forehead and two rows of wavy locks. Although too little remains of the head for a reconstruction of its height, it appearsto be similar in scale to 2a. Weinberg assigned a to the left side of the face above the left cheek, but it better fits the right side. In that way the locks of hair are pulled nearly horizontally from the slightly indented central part (largely missing) before bending into two deeply folded waves. The forehead then becomes low; the framing hair descends in continuous crimped locks at a very slight angle toward the missing right ear.A deep channel divides the two rows, and the surface of each is lightly scored to simulate fine strands.The hair is covered with a pale red paint. A low stephane with straight flaring profile sits on top of the head. At its top inner break, a very faint lip may mark the start of the projecting crown of the head such as occurs on the Louvre sphinx. The stephane is decorated as follows. A lotus-palmette chain, of which only the bottom half is preserved, and two bands below it are painted in a matt brown-black on a reserved, slipped ground. The palmette is formed of seven petals, the tips of which are round and slightly drooping. The top of the stephane is very thinly slipped and unpainted. The relative positions of the two pieces are uncertain. 4b, which shows a slight asymmetry in horizontal section not apparent in a, could have lain to the right or left of a. Furthermore,the painted bands and the lines of floral chain are much thicker on b than on a. One cannot reject the possibility therefore that the two fragments derive from very similar but separate statues. As fragment a better reveals, the head is hollow, its preserved wall thickness varying from 0.018 m through the forehead, 0.040 m through the stephane, to 0.017 m at the top back break. A clearjoint is visible between the wads that form the forehead and hair and those of the

CORINTHIAN

TERRACOTTA

SCULPTURE

407

stephane and top of the head. This suggests that the two sections were made separatelyand were joined. Strips of fine clay were used for the hair. Given the very deep and irregularchannels that separate the two waves, the hair was probably added by hand and scored. For this procedure, the clay must have been soft, because the locks look as if they had been squeezed out of a serratedpastry tube. A fine skin-like slip masks the stephane. It is possible that the head broke and was repairedin antiquity, for a small, horizontal hole 0.005 m in diameter pierces the stephane of b at the right break, on line with the upper painted band. Weinberg placed a in the second quarterof the 5th century because of the form of the palmettes which decorate the stephane. A similar lotuspalmette chain appears on the stephane of statue 7, with the difference that the palmette of statue 4 has seven rather than eleven petals. Parallels are discussed in conjunction with 7, and a date around the middle of the 5th century is suggested for that statue's modeling. Such a date also suits better the hairstyle of 4. Parallel rows of crimped wavy locks are relatively common on sculptures and terracottasof the first half of the 5th century. See, for example, the terracotta half-figures published by Poulsen (1937, pp. 50-52) or the Attic figurine head found in Ithaca and now in the British Museum (Higgins 1954, no. 680, p. 181, pl. 89). But if we compare our piece with a typical example of the Severe style-the head of Athena on the Olympia metope that depicts the gardenof the Hesperides (Ashmole and Yalouris 1967, pls. 188, 191)-there is a difference. On the latter, the evenly sized waves are crimped but rigidly parallel. On Corinth sphinx 4, the strands are of uneven thickness and relief height, and the waves make large loops that virtually bend back on themselves. A good parallel in a small-scale, Attic figurine, now in the Louvre, has been dated variously to 450 or 450-440 B.C. (Mollard-Besques 1954, nos. C10-11, p. 83, pl. 56; Charbonneaux 1943, pl. 42). The identification of this piece as a sphinx is largely based on the similarities of its clay color and of its composition to the wing, 5, which follows, and, less certainly,to the hock, 6.

Ca. 450 B.C. 5

Right wing tip of sphinx

SF-76-4. Max. p.dim. 0.200, Th. 0.030 (tip)-0.038 (top right). Parts of four feathers, including parts of three tips, surface chipped. Upper edge somewhat weathered. Early Roman fill, quarrytrench 10, lot 1976-177. Clay: thin fine surface layer,

Fig. 7 Th. 0.001 m or less, firedpaleyellow, nearly2.5Y 7/4; core:commonand well-mixed,fine to medium,black mudstoneinclusions(0.001-0.003), firedgray-green5Y 7/2. Paint:black lOYR3/1; red-brownlOR 4/4.

Parts of three primary feathers remain from a large-scale right wing. The surface of the wing is flat, with successive feathers defined by a broad groove that has a rounded section. From groove to groove, the feathers are roughly 0.044-0.046 m wide. They curl up to the rounded tips, which are cut into a scalloped edge. The feathers are painted alternately matt black and dark red-brown and each is outlined by a very

408

NANCY

BOOKIDIS

narrow reserved stripe. Black paint is applied to the grooves. The unpainted outer face and back of the wing are either covered by a very thin slip or simply wiped, so that although the inclusions still show, these surfaces have a faint luster. Although no certain association is possible, the wing closely resembles the stephane, 4, in fabric and color. In section the wing is essentially a flat slab of well-compacted clay that is thicker toward its base, thinner toward the tip, and rounded at the juncture of edge to back. In front, the coarse core is masked by a thin layer of fine clay no thicker than 0.001 m. This simply provides a good base for the painted decoration, and it is the layer through which the defining grooves are cut. A very close parallel can be found in sphinx no. 46 from Delphi (FdD II, no. 46, p. 261, pl. 94), with the difference that those feathers are simply divided by a black painted line ratherthan by a groove. The Delphi wing is also apparently thinner than ours. Difficult to date in its own right, our wing, if correctlyassociated with 4, would fall roughly in the middle of the 5th century. Ca. 450 B.C. 6

Right hock of sphinx

SF-77-4. P.L.0.112,Th. 0.062. Hock, brokenall around.Early Romanfill, quarrytrench12, lot 1977-233.

Fig. 7 Clay:fine surfacelayer,Th. 0.001-0.003, firedyellow-green5Y 7.5/3; core:abundantfine to medium well-mixedblackinclusions(0.001 to 0.003), firedgray-green5Y 7/3.

Preserved is the right hock, with part of the adhering thigh, of a crouching sphinx. A sphinx from Halai, which has all of its thigh, upper leg, and part of the rump (Goldman 1940, no. 1, p. 443, fig. 106), better illustrates the exact position of our piece. The Corinth fragment comes at the joint between the upper part of the leg, which is pressed against the thigh, and the lower part, which bends downward free of the body. In so far as the leg is preserved, the bones are well articulated.The outer surface of the leg is lustrous while the existing portion of the thigh is not. The leg is solid but the body is hollow, the body wall averaging 0.03 m in thickness. Because the fine clay covers the surface incompletely, inclusions show through in places. Paring marks are apparent, and in general the quality of workmanship is not as good as that of 1 and 2. In fired color and clay composition, the hock is similar,but not identical, to 4 and 5, for the surface of the leg has fired to a greener shade. While this difference in firing can be explained by placement of the pieces at opposite ends of the same statue, it is also possible that the fragments are from different statues. Too little remains to permit a dating of this piece independent of 4. At the same time, the careless finishing of the surface is more in keeping with the 5th century than with the 4th. 5th century B.C.?

CORINTHIAN

7

TERRACOTTA

SCULPTURE

Sphinx

a. SF-72-3a. P.H. 0.079, p.W. 0.153, est. Diam. 0.21. Stephane, broken at either end, in back, and below; top heavily weathered. Early Roman fill, west quarry trench, lot 72-175. b. SF-72-3b. P.H. 0.325, p.H. chest to sternal notch 0.20, max. p.L. ca. 0.222, est. W. neck 0.10, est. W. mid-chest 0.20-0.22, Th. wall 0.035-0.052. Six joining fragments, preserving two-thirds of circumference of neck with hair; most of chest, left shoulder, and humerus; and upper part of left side. Paint worn. Partly restored in plaster. Early Roman fill in west quarrytrench and quarrytrench 1, lot 72-175; and Early Roman fill with Byzantine intrusions, lot 74-88. c. SF-72-3d. P.H. 0.212, p.W. 0.092,Th. 0.046-0.065. Three joining fragments from outside of right humerus and right leg, surface of leg badly chipped. Early Roman fill in west quarrytrench, lot 72-175; and Byzantine disturbance, lot 72-192. d. SF-72-3e. P.H. 0.108, p.W. 0.138, Th. 0.030-0.045. Two joining fragments, left side, part of scapularsand three covert feathers; paint nearly gone. Finding as a. e. SF-72-3f. P.H. 0.102, p.W. 0.14, Th. 0.037-0.050. Two joining fragments, left side and belly, base of wing with parts of four coverts; partly restored in plaster. Finding as a. f. SF-72-3j. Max. p.dim. 0.10, Th. 0.027. Single fragment, upper edge of left wing, parts of three coverts, base of one primary feather, broken at back and at three edges. Finding as a. g. SF-72-3g. P.H. 0.252, p.W. 0.16, Th. 0.018-0.052. Three joining fragments, left wing, parts of seven primary feathers, including three tips. Finding as a. h. SF-72-3i + SF-5. P.H. 0.22, p.W. 0.23, Th. 0.026-0.04. Three joining fragments, upper part of left wing, five primary feathers, including three tips; top of wing heavily weathered. Finding as a; early excavations, provenience unknown. i. SF-72-3k. Max. p.dim. 0.106,

409

Figs. 8-21 Th. 0.023-0.033. Right wing, three primary feathers with part of one tip. Early Roman fill, quarrytrench 1, lot 74-89. j. SF-72-3p. Max. p.dim. 0.153, Th. 0.025-0.050. Back, broken all around. Finding as a. k. SF-72-3u. Max. p.dim. 0.091, Th. 0.018-0.035. Flank, broken all around; surface worn. Finding as a. 1. SF-72-3r. Max. p.dim. 0.12, Th. 0.022-0.034. Two joining fragments, hind end(?), broken all around. Early Roman fill, west quarry trench, lot 72-175; 19th-century intrusion, no lot. m. SF-72-31. P.L. 0.143, W. top 0.074, W. bottom 0.064. Upper part of left foreleg, broken below at carpals. Finding as a. n. SF-72-3m. P.L. 0.150, W. bottom (side to side) 0.047. Three joining fragments, right foreleg from above and below carpals, preserving dew-claw and back bone of leg. Finding as a. o. SF-72-3q. P.L. 0.145, p.W. ca. 0.11,Th. 0.028-0.05. Two joining fragments, left haunch. Finding as a. p. SF-72-3n. P.L. 0.103, p.H. 0.037, W. across toes 0.067, W. leg behind paw 0.048. Hind paw and bit of leg. Finding as a. q. SF-72-3s. P.H. 0.059, p.W. 0.124; p.L. 0.138. Plinth, preserving bottom, back edge, and bottom of sphinx. Early Roman fill, quarry trench 1, lot 74-88. r. SF-72-3t. P.L. 0.08, Diam. 0.046. Strut for wing support, broken at either end. Finding as a. Clay: fine surface layer varying in thickness from less than 0.0010.004; core: common fine to medium black mudstone inclusions, rare white, chiefly 0.001-0.002, to 0.004 maximum, and at least one piece of unmixed red clay. Clay fired buff throughout, ca. lOYR 7.5/3; where wall thicker, fired to grades of pink at core; on 1, o, q, surface mottled to a darkertan, 5Y 8/2.5-SY 7.6/2.5. Paint: light red 2.5YR 3/6; dark red lOR 4/4; black. Published.Robinson 1976a, p. 236, pl. 54:a, b. For part of 7h, see CorinthIV, i, p. 113, M5 (SF-5).

4IO

Is

i I1i

I "I

,--Z-

Illa

t

\1

CORINTHIAN

Figure 8 (opposite). Sphinx 7. Restored drawing. R. Doxan

TERRACOTTA

SCULPTURE

4II

Sphinx 7 is the both the latest sphinx from the area of Temple Hill and the best preserved or, perhaps one should say,represented by the greatest number of fragments. Eighteen pieces are catalogued herein; another ten were considered too small or too unimportant for full description but have been cited where useful. Despite the number of fragments, however, reconstruction of the sphinx has been extremely problematic, and the drawing by Roxanna Doxan, shown in Figure 8, should be regarded as a suggestion rather than as a definitive restoration. All of the pieces were found in that part of the quarrythat lies immediately northeast of the Temple. While they were not all found together in one spot, they occurred in successive tips of the same Early Roman filling. Only joining fragments of b, c, and 1were recovered from later disturbances of that filling, while part of 7h derives from early excavations. Half again as large as Weinberg 33 (H. 0.627 m), our sphinx is estimated to have been roughly 0.94 m high. She crouches with her head turned very slightly to proper left, the torsion apparent in the muscles of her neck. Whether this torsion was carried down into the chest or shoulders cannot now be determined. The sphinx wears a stephane, a, that is decorated with a floral chain described below (Fig. 9). The top of the stephane projects freely from the crown of the head and is scalloped to reflect the painted ornament. Beneath the stephane is a smooth surface (partly reserved and partly painted brown) which follows the curvatureof the head. It does not resemble hair and is therefore either a band along the bottom of the stephane or a separate fillet. The shoulder-length hair is preserved only from the height of the neck and down on b (Figs. 10-12). In Figure 11, it can be seen to viewer's right of the neck. It is arrangedin a modified version of the traditional Archaic etagenperlicke. In place of the usual stylized waves that frame the face are long loose curling locks, modeled in high relief, that are preserved only on the proper left side. The locks continue onto the neck in paint. In back the hair abruptlychanges to smooth horizontal waves. As these curve around to the right side of the head, the otherwise smooth, wavy surface is broken by parallelvertical incisions to indicate individual strands of hair.Whether there were loose curls framing the right side of the face is no longer clear. The neck, b, is well modeled. Somewhat irrationally,the prominent sternal notch is pulled to proper left of center as a reflection of the head's position. A shallow concavity, the axis of which falls just to proper right of the sternal notch, marks the center of the chest. The subtleties of the chest and shoulders are best appreciatedwhen compared to Weinberg 33. The chest of the latter sphinx (Fig. 33) is narrow and without modeling; a ridge separatesthe chest from either shoulder. In profile the chest and the front of the humeri make a continuous convex curve, and the overall impression is bird-like. Our Temple Hill sphinx has a broad, rather square chest, with parallelsbetter drawn to felines. In addition to the concavity that runs down the center of the chest, a further concavity,in place of the sharp ridges of Weinberg 33, marks the

412

NANCY

BOOKIDIS

Figure9. Sphinx7. Stephane7a. Scale 1:2 (color photo not to scale)

transition to either humerus. In profile, the shoulder and humerus form a compound curve. At a later stage in the development of the sphinx, the upper part of the chest will develop into breasts, as on the late sphinx from Thermon (Koch 1915, p. 67, figs. 28-29). On the sides, b and c, the body flattens unnaturallyto make the transition to the wings, d. It is unfortunate that this part of the sphinx's body is poorly preserved for, as a result, certain questions about the form of the wings and their attachment to the body must remain problematic. The wings begin just behind the humeri and are recessed slightly from that surface. 7d (Fig. 14) preserves the offset as well as part of the covert feathers of the left wing, while c (Fig. 13), from the right side, breaks off just short of the offset. As is shown by g (Fig. 16), h (Figs. 17-18), and i (Fig. 19)-together with lot 72-240-8, an uncatalogued fragment of the right wing tip-both wings are modeled in exactly the same way. On both fragments c and d, covert and primary feathers are worked in relief. On e (Fig. 15), the coverts appear to begin just behind the foreleg. Although the lowest feathers are not preserved, the outline of a tip appearson e along the left break,where it overlaps the second and third primary feathers. 7e probably lay close to d. Because the length of the coverts is unknown, the length of the wings cannot be determined with certainty.It is clear, however, that the upper edge of the wing did not form a continuous line from shoulder to tip, as is customary and clearly preserved on Weinberg 33, but was articulated at the juncture of the coverts and primaryfeathers. Evidence for this is provided by f (Figs. 15, 18), which preserves the tips of the uppermost three covert feathers and the base of a primary one. The tips of the first, and half of the second

TERRACOTTA

CORINTHIAN

~

I

I

~

~

~

~_

~

SCULPTURE

4I3

-

-

II~

~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~ ~~~ ~~~ ~~~ ~~~ ~~~~~~~~~~~~~~~~~~~

4I4

NANCY

BOOKIDIS

_~~~~~~~~~~~~~~~~~~~~~~~~

t

I

Figure 11. Sphinx 7. Neck, chest, and left side 7b, left proffle. Photo scale 1:2

TERRACOTTA

CORINTHIAN

isucran fasion thi

h

rmryfahr

roneiscetn

415

SCULPTURE

tpoti ele clopdotie

_ntecnetoa

hywdnfo

Figure12. Sphinx7. Neck, chest, and left side 7b, backview.Scale1:2 covert, are finished, and therefore curled out from the wing, while a primary feather continues upward from the rest of the second and third coverts. In Figure 8 an attempt has been made to render the relative positions of the wing fragments f, g, and h based on the diminishing widths of the feathers and on their alternating colors, which are described below. Further assistance comes from the supposition that the top of the wing was probably at least as high as the top of the head. On the back of g,abroken line along its bottom edge undoubtedly represents the point of attachment of the wing to the side of the body. Above this point there were eleven primary feathers, the uppermost of which is missing but can be restored from f. But the exact distance between the wing g and belly e

NA N CY

4I6

I

A

I

1111

I ~ I 11Il= _~~~~~~~~~

B OO KI DI S

CORINTHIAN

_

Figure14. Sphinx7. Left wing 7d. Scale1:2

TERRACOTTA

SCULPTURE

4I7

~~~~~~~~~~~~H

narrow bases toward the tips. In section, the wings are thicker toward the body and thinner toward all edges, and the back face is rather carelessly finished. One or more struts served to prop apart the wings. One such strut, r (Fig. 21), is a solid cylinder of clay, roughly finished and painted like the back of the wing. At mid-length, a narrow hole 0.01 m in diameter cuts from the surface obliquely into the strut to pierce one end. On the back face of the wing h (Fig. 17) is a roughly circularbreak and a central hole that cuts into the thickness of the wing. Although the strut r does not join h, it may have belonged close to it. Of the body of the sphinx beyond the wings, only odd fragments survive.These are the lower back j (Fig. 19), the left rib cage e (Fig. 15), and the flank k (Fig. 19). From these one can see that the ribs are rendered as soft ripples, and the flank is tautly concave as if the body were being stretched. A fourth body fragment, 1 (Fig. 19), is problematic. It consists of a gently rounded portion of anatomy, the surface of which breaks in a roughly straight line along one preserved edge. To one side of this edge is a large broken surface with curved outline that is pierced obliquely by a hole. Around the hole the surface is fired tan. Wall thickness and firing color suggest that the piece should belong to the hindquarters of the sphinx. It is tempting to attach a tail to the curved break and run it up the back along the straight broken line, but the back then does not seem to be sufficiently rounded. Just below the primary feathers on the belly, e, is a projecting broken surface that may have to do with the attachment of the left foreleg. Parts of both forelegs, m (Figs. 4, 20) and n (Fig. 20), are preserved down to the carpals.Two other fragments, not included here (SF-72-3o and lot 72-240-1), may be from the upper part of the foreleg. Although the legs are well modeled, they show some lingering of the Archaic in that each seems to consist of four bones rather than two: front, back, and two sides. The dew-claw is rendered as a flat and slightly undercut flap, as is the projecting bone of the carpals. On the outside of the right leg, at the upper break, is a small conical depression, presumably some sort of decorative element like the conical depressions on 2c. The haunch, o, is problematic. As oriented in Figure 21, it is the thigh of a bent hind leg,

Wing andbelly7e _

Figure15. Sphinx7. Scale1:2

_

CORINTHIAN

_~~~~~~~~~~~~~~~~~~

TERRACOTTA

SCULPTURE

419

A

Figure 16. Sphinx 7. Leftwing7g. Scale 1:2 except as noted

the line of which is restored from the lower break. But turned upside down, the piece could also represent the lower leg pressed against the missing thigh. Since such a lower leg seems too large for the forelegs, we have preferredthe first interpretation.The one surviving paw, p (Fig. 21), a hind one, has long, well-separated claws with fat pads and projecta ifr"nd;I, smoted on, th bottom.xi Onv

etca

dei

peevd,psil

NANCY

420

.,~~~~~~~~~~~~~~' _

.

.......

BOOKIDIS

......

... ... ... ...

.... .....

blaccay ofteltsi_loatnae,

left break(1-1)

ihfaigtp

n

ih

the back one, for a thickness of 0.025 m. Worked in one with it may be the bottom of the sphinx, represented by a sloping finished surface that slants up from the plinth toward the back and breaks off 0.08 m from the back edge. This surface displays the same pinkish-tan discoloration of the surface as do several fragments from the hindquarters.If correctly understood, this piece indicates that the sphinx was sitting on its hind end, supported by its front legs. The sphinx is painted as follows, in the typical three-color system of matt black, red-brown in varying shades, and reserved clay. On the reserved face of the stephane is a lotus-palmette chain composed of an scrll The petals ar ln

and slender the ipsfulan

dopn.

The

Figure 17. Sphinx 7. Left wing 7h. Scale1:3exceptas noted

CORINTHIAN

TERRACOTTA

SCULPTURE

42I

Figure18. Sphinx7. Left wings 7f, 7h, restored. red bud. To the left of the first lotus is a broad black stripe above a curling tendril where palmette petals should be. Possibly, therefore, the floral chain was confined to the front half of the stephane or else the painting here was simply careless.Above, the scalloped edge of the stephane is painted light red while the badly weathered inner face appearsto be reserved.The sphinx's hair is a brown-black with wispy curls painted onto the neck. Around the base of the neck is a decorative collar, consisting of a row of large black dots, framed above and below by a broad black stripe. Similar borders, running vertically,separate the chest from shoulder and humerus, and the shoulders from the wing. The chest, shoulders, and humeri are covered with long-stemmed, scaly feathers, arrangedin rows of light red, alternating with black within a reserved outline. Over the chest, only the red feathers are further ornamented with reserved diamond centers. Whereas the scales of the chest point downward, those on the shoulder are turned sideways or tilt slightly downward, depending on the angle of the chest. All of the scales appear to have been painted free-hand. No two are identical in size, and they grow or contract according to the anatomy.Above the forelegs, the scales are interrupted by another border of dots and bands that makes an arc above the outside of the leg and curves down sharply inside it. Both covert and primary feathers are painted with alternating red or black centers, surroundedby a reserved stripe, and by a black stripe along only the outer edge of each feather.The black coverts overlap red primaries and vice versa.Two shades of red are used, a darkerone for the primary feathers and backs of the wings, a lighter shade for the coverts and scales. In addition, the scalloped outer face of the wing is painted black or red in agreement with each feather.The body parts are left in the natural clay color and are virtually without luster. A very faint polish

422

NANCY

BOOKIDIS

Back7j Rightwing 7i

Flank7k

Hind end(?)71

can be seen on a few parts such as the forelegs, but in no way does this correspond to the high luster of the Archaic sphinxes. We have spoken earlier of the interior of this statue. It exhibits considerable variety,which probably is the result of modeling by hand. Regrettably,the very fragmentary state of the statue does not allow us to reconstruct the way in which the various parts went together. The thickest and most irregularsections are those around the base of the neck and chest. In the neck, large wads of clay are pressed one against another. At the base of the neck and juncture with the body, the wads are left rough rather than smoothed, perhaps because the sculptor could not get his hand in to consolidate the clay. By contrast, the inner surface of the breast shows deep vertical furrows that were probably made by the fingers. The surface of these furrows is finely striated. It is not clear whether this is simply the result of stroking coarse clay and pulling up inclusions or whether something textured like a rough cloth or sponge was used. Whereas the chest and shoulders are quite thick, the belly and hindquarters are thinner. Although finger strokes are also visible within the belly, the body wall is more consistent in thickness, the inner surface smoother. These variations in wall thickness from one part of the statue to another, as well as the irregularfinish of the interior, are characteristic of post-Archaic work.

Figure19. Sphinx7. Scale1:2

CORINTHIAN

TERRACOTTA

SCULPTURE

423

Left foreleg 7m

shudr ,t bek thog

h ako h ef h ufaeil rn.utpsil

ieo

h It_ed(i.1:H) o i ak

hr

Right foreleg7n

Figure20. Sphinx7. Scale1:2

A number of fragments preserve impressions of rods or sticks used during modeling to hold the various segments together. Because the exact position of any one fragment in relation to another is uncertain, the system behind the rods cannot be restored, and each hole is delogol1. A11.

I

)

s

424

NANCY

BOOKIDIS

......~. ....I|

|

_

Left haunch7o

Plinth~~~~~~~~ Paq7 Figure.21.

thyd cotas,a

Sphix

o apa oconiewthaysrt ithhlei teseodprmryfate

ntebckfc.B o ntonypire

appear to be two separate drillings with slightly different angles. A single small hole (Diam. 0.005-0.006 m) runs up either leg-c and n; m (Fig. 4)-presumably from the plinth or paw to the top of the leg, and ends below the body cavity.In this hole the impression of a stick is clearly preserved. The left wing preserves a number of holes, described here beginning at the base and working upward. One cuts obliquely through the surface of the top covert on d near its base (Diam. 0.008-0.009 m; Fig. 14: H2). A second hole enters the wing, e, horizontally through the fourth primary feather from the bottom, to a depth of 0.019 m but does not go through it (Fig. 15: H43).The bottom of this impression is smoothly concave. Two more holes pierce g through the third and sixth primary feathers

(i1:4H_ahu

e

pr

t

w

completely,

7.

Sca

CORINTHIAN

TERRACOTTA

SCULPTURE

425

the wing but continues into the strut in back (Figs. 17, 18: H6). The outer surface,however, is broken here and it is not clear whether the opening was masked or exposed. Finally, a hole (Diam. 0.011 m) passes obliquely through the thickness of the anatomical fragment 1,visible in Figure 19 at lower left. Around this hole, the surface is broken as if something had once been pinned here. These holes do not seem to form an armature,in the modern sense of the word, around which the sphinx was built. Rather, sticks or rods were either used to pin two parts together during modeling and drying or, perhaps, to prop the sphinx from the outside, as on e-otherwise, it is difficult to understand why holes such as that on e did not pass entirely through the body wall. It is interesting that there are no ventilation holes which communicate with the hollow body chamber and the exterior,the only possible exception being the incompletely preserved hole in d. As for the holes on the left wing that are unrelated to struts, these can be paralleled on Weinberg 33. We suggested in the introduction that these could have been for mends with lead, although the absence of abrasion on the surface surrounding them makes this questionable. Alternatively, they might have served some sort of aerodynamic purpose to reduce the thrust of wind. Sphinx 7 is unquestionably post-Archaic in style. The plastic modeling of the hair, neck, and chest is ample testimony of a later date. At the same time, the very fragmentarystate of preservationof the statue and the lack of later comparandain terracottamake close dating of the piece somewhat difficult. As noted above, the structureof the chest differs from that of 6th-century animals, and the primaryfeathers of these wings areblunter than theirs. But perhaps the most useful criterion is the palmette-lotus ornament on the stephane. Within the realm of architecturalterracottas, some comparison can be drawn between painted ornaments on simas and our stephane. The critical elements are the form of the lotus, the use of an eleven-petal palmette, the slight droop in the petals of the palmette, and, to a lesser degree, the direction of the spirals beneath the palmette. In looking for parallels, it soon becomes clear that a lotus of our type, with two flaring petals and a single bud, drops out of the repertoryof architecturalterracottas by about the middle of the 5th century.A sima from Corinth, FS-6 (Roebuck 1991, p. 40, pl. 9:C), dated by Mary C. Roebuck as being from the second quarterof the 5th century to ca. 450 B.C., has a similarlotus, though with more exaggerated petals, and an eleven-petal palmette. But the form of the palmette differs from ours. It makes a semi-circle in outline and has shorter and fuller petals. A second sima from Corinth, FS-876 (Roebuck 1991, p. 43, pl. 10:D), is decorated with petals that droop more like those on the stephane, but again the palmette is fuller,the central petal short, the heart triangular,and the lotus more complex in form. Dated to ca. 430 B.C. by Roebuck, this piece probably gives a rough terminus ante quem for our statue. Parallels in vase painting are less satisfactory,perhaps because the floral ornamentation is affected by the space and contours which define it. Of specialinterest,however,is a bronze couch appliquein Berlin (Jacobsthal

426

NANCY

BOOKIDIS

1927, p. 149, pl. 117:B), which consists of four antithetic palmettes. Each palmette has eleven petals, its central leaf is elongated, and the general outline of the palmette is slightly higher than it is wide. In addition, the heart is rounded and sits above two scrolls that curl in toward the center. Dated ca. 460 B.C. byJacobsthal, it provides a close parallelfor our piece. If any differences are to be cited, it is perhaps the greater droop of our petals, which may point to a slightly later date for our sphinx. Although Pernice (1904, p. 30) suggested that the applique had come from Boiotia, one cannot help but wonder if it had not been made in Corinth. Based on these parallels, our sphinx should probably be dated around the middle of the 5th century or just slightly earlier. At that date, there are few pieces with which it can be compared. Perhaps the most prominent example is the well-known marble sphinx from the Sanctuary of Apollo on Aigina (Alt-Agina II, 2, no. 52, pp. 80, 118-120, pls. 35, 38-40). Both sphinxes turn their heads slightly to one side but there the similarities stop. The hairstyle of the Aiginetan sphinx is elaborate and unusual, while that of the Corinthian statue is a kind of modernvariationon an Archaic theme. In profile,the chest of the Aiginetan sphinx is heavy and blockish, while the Corinthian one shows the beginnings of breasts.In contrastto the massivenessof the chest of the Aiginetan sphinx, the lower part of its torso is slender and bony, especially in the haunches. The torso of the Corinthian sphinx is less exaggerated.WalterKarydicalls the Aiginetan sphinx the first to abandonthe traditionalsickleshaped wing, whereas the wings of the Corinthian sphinx are absolutely conventional and follow local traditions established in the 6th century. It is possible that these differences are owing to the materials in which the animals were modeled. The differences may also be the result of the differing functions of the two sphinxes. As a freestanding votive, the marble sphinx could have been carved in the style of the period in which it was dedicated. But an acroterion that was designed for a pre-existing Archaic temple may have had to keep more of the flavor of the original structure. Ca. 460-450 B.C. 8

Legs, paw of sphinx

a. SF-75-3. P.H 0.069, W. top 0.066, Th. wall 0.022-0.033. Half or quarter of the circumference of lower leg, broken top and bottom. Early Roman filling with Byzantine intrusions, quarrytrench 6, lot 75182. b. SF-lb. P.H. 0.115; p.W. 0.071 (top)-0.057 (bottom). Lower part of left foreleg, broken at top, bottom, side. Early excavations, exact provenience unknown. c. SF-la (Weinberg 38). P.H. 0.118, D. top (front to back) 0.075, Diam. rod 0.02. Lower part of left foreleg, preserving carpals and part of metacarpals.Finding as b.

Figs. 4, 22 d. SF-75-4. P.H. 0.042, p.L. frontto back0.068, p.Th. 0.024. Paw, brokenall around,preservingtwo toes. Findingas b but with Byzantine intrusions,lot 75-181. Clay:fine surfacelayer,Th. 0.002-0.005, firedyellow,near2.5Y 8/5 to 2.5Y 7/5 (d);core:abundant fine to mediumblackinclusions(to 0.003), firedpink 7.5YR 7/5. Published:c: Weinberg1957, no. 38, p. 316, pls. 72, 75.

CORINTHIAN

TERRACOTTA

SCULPTURE

Lowerleg 8a

427

Left foreleg8b

Paw 8d

Left foreleg8c Figure22. Sphinx8. Scale 1:2 8a (Fig. 22) preserves either half the circumference of a leg from radius to ulna, or, as the scale of c suggests, the front quarterfrom the inside of a large, right leg. The fragment tapers slightly from top to bottom. In horizontal section, the inside is flattened but makes a wide curve around the radius,which is offset toward the front by a shallow concavity.A much deeper concave channel offsets the radius toward the back of the leg. Presumably,the ulna began just behind this. The surface is well polished. A large hole 0.018 m in diameter runs down the axis of the leg. Not included by Weinberg, although an early find, b (Fig. 22) preserves the back quarterof the inside of the left leg immediately above the carpals c (Figs. 4, 22). That b and c go together is suggested by two technical details. On both, inclusions show through to the surface on the back of the leg, and a roughened surface,which runs down one side of the central hole made by the supporting rod in b, can be aligned with a similar surface in c.

NANCY

428

BOOKIDIS

Already published by Weinberg, c preserves the left leg at the carpals and top of the metacarpals.As with a, the bones of the leg are separatedby concave channels, one on either side of the radius. At the back of the carpals is the large projecting bone, partly broken and painted black. On the inside of the leg toward the front is the dew-claw. This is undercut to stand out from the leg, and the claw proper is cut back from the surrounding flesh. Below the joint, the ridges of the metacarpals are marked and curve outward rather sharply.A central hole 0.02 m in diameter remains from the structuralrod around which the leg was built (Fig. 4). 8d (Fig. 22) derives from a very large front paw that is larger in scale than the hind paw of 7. Whether it is a left or right paw is uncertain. Preserved are one of the middle toes and part of an outer one, set well back from the first. In profile the toes are quite high, the pads fat and round. The claw is outlined by fairly deep incisions but is not cut free of the pad. A shallow groove delineates the skin above the claw. Another groove sets off the small claw from the larger one. Paring marks are apparenton the slightly lustrous surface. The four fragments are distinguished by their very large scale and by their color.They have been fired at a high temperature,which has turned the core clay a bright pink and the fine surfaceslip a deep yellow, especially on b (2.5Y 7/5). The high temperaturehas also made the fragments somewhat brittle. If we compare these legs with 1 and with its Weinberg parallels, we find here none of the schematic renderings of anatomy of the earlier pieces. The elements are simplified, and there is more reliance on rounded forms than on sharp ridges. In this respect, 8a-d more closely resemble 7, while in scale they appear to be slightly larger than the fragments of that sphinx. The legs may be tentatively associated with the two wing fragments catalogued below as 9a-b. Classical. 9

Right wing of sphinx

a. SF-72-14a + c. Max. p.W. 0.17, Th. 0.022-0.028. Two joining fragments from tip of wing, preserving two feathers. Early Roman fill with Byzantine intrusions, quarry trench 6, lot 75-182; and Early Roman fill, west quarrytrench, lot 72-175. b. SF-72-14b. PH. 0.10, p.W. 0.142, Th. 0.023-0.025. Two feathers, including one tip, and edge of third. Early Roman fill with

Fig. 23 moderndisturbance,quarrytrench2, lot 74-87. Clay:fine surfacelayer,Th. 0.001-0.005, firedyellow,or slightly moreyellowthan 5Y 8/3, to buff lOYR7.5/4 on back;core:moderate amountof well-mixedred andblack fine to smallinclusions(to 0.002), with small,unmixedpelletsof red andwhite clay,firedpink,ca. 7.5YR 6.5/6. Paint:red-brown,ca. 5YR 5.5/4.

Two nonjoining fragments of a large right wing preserve the uppermost two primary feathers and three more primary feathers from the edge. The feathers overlap in steps that are not neatly and sharply cut but somewhat carelesslybeveled. The uppermost primary feather

CORINTHIAN

TERRACOTTA

I

SCULPTURE

Il

_

|

_ _I

_

l

l

I_

I | I_

l

_

_

429

l ll l

l

11

..

|

1-

i

_ w_

_

_

_

_w Rightng

_i _ _l

_

_ - _M l _S _ _ _l

_

_

_

_

_l

_-I

9a

|I i |_

|l I l

__

l l __. l |

| _.

|_

I I b,

i

1|1,

-

Rightng

Figure23. Sphinx9. Scale

1:2

9b

430

NANCY

BOOKIDIS

curls in and is bent out somewhat. A small projection of the coarse core at the base of this feather probably marks the start of the topmost covert feather, much as on 7. Complete feathers measure respectively 0.0630.066 (b) and 0.067 m wide (a). The outer contour of the wing is not sharply cut but somewhat irregularlyfinished. The back is quite flat, with the result that the wing is fairly uniform in thickness, increasing slightly toward the outer edge. Painted uniformly light red-brown, the centers of the feathers are outlined by a broad reserved band between matt black stripes. A very dilute red-brown paint is applied to the edge of the wing, while the back face is both unslipped and undecorated. In section, the coarse core of b, 0.02-0.023 m thick, is clearly distinguishable from the fine clay in which the feathers are modeled. This fine layer tapers in thickness from 0.005 m at the outer edge of each feather to a thin skin at the inner edge. Irregularitiesare more marked on a. For example, the uppermost feather is thicker (to 0.028 m) than the others. Furthermore,a curious feature is visible on the side break of fragment a. A thin line of fine slip, lying 0.003 m below the surface, appearsto outline a thinner and shorter wing that ends 0.038 m from the present wing tip. The adjoining feather,which has broken just beyond this point, has also broken along a joining surface, as if the original wing had been thickened and lengthened. Whether the main part of the wing was made in a mold or handmade is unclear. These two fragments differ stylistically and in quality of workmanship from all of the other wings. The feathers are much wider and yet, given the scale,the wing is extremelythin. This thinness is especiallystriking when we compare the piece with 16, which has feathers of similar width but over twice the thickness. Also noticeable is the rather poor quality of workmanship. The surface of a is rough, the slip unevenly applied or smoothed; inclusions show through to its surface.The tip of its uppermost feather turns outward toward the viewer, as if the piece had warped in firing. Moreover, the black stripes are quite carelesslypainted. The quality of finish and drawing would suggest a date in the 5th century,at the earliest, rather than in the 6th. In his preliminary work on the temple's roof tiles, Henry Robinson identified a fragment of the roof that might reflect a Hellenistic repair.101 Our wing is identical to that fragment in both clay color and surface treatment and may well belong to the same remodeling. Whether this took place as late as the Hellenistic period is not clear,but it was surely no earlier than the 5th century. 5th century B.C. or later. 10 Anatomical part of sphinx? SF-72-15. Max.p.dim.0.021, Th. 0.071-0.082. Anatomicalpart, brokenon all edges,preservingouter andinnersurfaces.EarlyRomanfill, west quarrytrench,lot 72-175. Clay:fine surfacelayer,Th. 0.008, firedlight yellow2.5Y 8/3;

Fig. 24 core:well-mixedcommonfine inclusionswith raresmallto medium ones (0.001-0.003), all of which are blackwhereclayis buff,redwhere clayfiredpink,firedsalmonpink 5YR 6/6 or slightlyredder.

101. See above,note 24, and 15, below.

CORINTHIAN

TERRACOTTA

SCULPTURE

43I

of Figure~~~~~~~~ 24. Anatmica ~par ~~

o Fping(?) 10. 1.Sntmcale Scale pa:2 sphinx(?) 1:2so

Anatomical part of a very large animal having very little modeling far as preserved.The fragment is relativelyflat, with a slight compound curve along the long axis, from concave at the thinner end to slightly convex toward the thicker end as if from the side of an animal. The surface is plain but lustrous, with inclusions coming through to the surface in a few places. The interior surface is roughly smoothed. There are traces of what may have been two holes for supports. One runs vertically to the surface but breaks off 0.02 m below it, while a second, at the opposite end of the piece, lies horizontally 0.05 m below the surface. The great scale of this piece is attested by its wall thickness, greater than that of any other terracottafound at Corinth. Because of its size and lack of features, its identification must remain uncertain. Its scale is close to that of the horse published by Weinberg (1957, no. 7, SF-39-3, pp. 305-306, pls. 65, 75), although it is clearly not a part of that composition. Late 6th to 5th century B.C. 11 Wings of a sphinx a. SF-7a.Max.p.dim.0.15, Th. 0.035-0.041. Left wing,brokenon all edges.Foundbefore1929. b. SF-28-3. P.L.0.115,Th. 0.030. Left wing, preservingupper edge.Fromthe AthenaTrench southeastof the Theater. c. SF-7b.Max.p.dim.0.092,Th. 0.026-0.031. Rightwing, preserving outeredge.Foundin 1925 in late levelsoverthe hemicyclejust northeastof TempleHill andwest of the LechaionRoad.

Fig. 25 Clay:fine surfacelayer,Th. 0.001-0.003, firedpaleyellow2.5Y 7/4; core:abundantfine to small blackinclusions,firedslightlygray. Paint:darkred-brownlOR3/3; black 5YR 3/1. a, c: CorinthIV, i, Published: pp. 113-114, M2, M9, fig. 46:b; Koch 1915, p. 83, fig. 38;Weinberg 1957, no. 34b-c, p. 315, pl. 72.

432

NANCY

BOOKIDIS

k.A

backs(1:3)

Left wings

outer~~~~ or loeed,

flloe

by on_e

n

n

lckpiayfahr

Rightwing tic Figure 25. Sphinx 11. Scale 1:2 except as noted

Three fragments preserve parts of two wings of a sphinx. Ila (Fig.

25) preserves three primary feathers of the left wing from a point near their base. Rendered in paint only, these consist of alternating red-brown and black centers, flanked by reserved stripes and outlined in a thin black stripe; those feathers which are preserved here are, from top to bottom, black, red-brown, and black. The feathers average 0.04-0.042 m wide. I b (Fig. 25) gives the upper edge of the same wing, as well as two primary feathers, black (upper feather) and red-brown (lower). The feathers on b are the wider of those on the two fragments, and b stood roughly 0.03-0.04 m to the right of a.

CORINTHIAN

TERRACOTTA

SCULPTURE

433

Since these feathers are equal in width to those on a, fragment c must also have lain near the base of the wing, though it is somewhat thinner than a. The better-preserved left wing appears to have tapered in thickness from bottom to top. Both top and bottom edges of the wings are cut quite sharply and left unpainted, while the back surfaces are slipped and left plain. The wings consist of flat slabs of clay.As with 2c, the clay is layered. Beneath the fine surface is a thick layer with abundant fine inclusions (< 0.001 m), then a layer with small inclusions (to 0.002 m). The fine clay that forms the surface reaches a thickness of 0.002-0.003 m on the outer face, but measures less than 0.001 m on the back face. As we stated in our discussion of statue 2, fragments a and c were originally associated by Weinberg with the wing published here as 2c. But becausethe featherson lla and llc taperin opposite directions,they should belong to similar points on opposite wings. We have therefore separated them from 2c and have added a third piece, 1lb, which was recently found with context pottery from the Athena Trench. Given the find-place of c, the sphinx undoubtedly once stood on Temple Hill, and for that reason has been included in Part I. If we are correctin assigning both wings to the same sphinx, as similar firing and painting suggest, then both wings were rendered in the same way.With so little preserved, however, it is not possible to assign a date to this group. 12 Wing of a Nike(?) SF-74-4. P.L. 0.074; p.W. 0.065; Th. 0.022-0.024. Wing, preserving both faces, broken on all edges. Early Roman fill, quarrytrench 4, lot 74178.

Fig. 26 Clay:fine surfacelayer,Th. 0.001-0.003, firedbuff2.5Y 7.5/2; core:frequentfine blackinclusions (0.001 or less),firedpalepinkish buff,or slightlypinkerthan lOYR 7/3.

A fragmentarywing, possibly the left, is modeled on both faces. On the front face (Fig. 26) is part of one feather, below which is the edge of a second feather that is set back from the first. Above, the fragment has broken along the joint with the next feather, giving a width of ca. 0.05 for the extant one. At the upper right, the fragment has broken along a V-shaped joint. On the back (Fig. 26) are parts of two overlapping feathers, the lines of which do not precisely correspondwith those on the front face. The surface of the lower feather thickens slightly to the right, where the surface breaks along a regularcurving line, presumably at the juncture with another projecting surface such as the coverts. The single preserved front feather is painted according to the usual convention: red-brown center, now largely peeled, offset by narrow reserved bands and a red-brown band over the step. The back face is modeled but not painted, and inclusions appear on the surface.The wing is executed largely in coarse clay that is masked on both faces by a thin layer of fine slip.

NANCY

434

BOOKIDIS

Wing of a Nike(?)12

Humanleftfoot13 ..

l##lS#..................... ......a'. 'u

| I|| | | | | |

I I111 l

1

- | | | | | || | ,.',.'1.''',,....".'':...:.:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..

11 _1

ll

I I II_E

_"...'.,,.

'.t,'t,?.,1'

On the principle that sphinx wings are modeled on the outer face only,102our piece should not belong to a sphinx but to a Nike. It certainly finds no parallelsamong the other fragments published herein. It is generally thinner than those, smaller in scale, and with a more pronounced curve to the feathers. If we compare it with the Nike published by Weinberg as no. 28 (1957, SF-34-1, p. 313, pl. 70), however,there are severalnoticeable differences. Although the two pieces are similar in thickness, the feathers of Weinberg 28 areworked in relief only on the front face; in back they are painted. Moreover, the feathers of Weinberg 28 are approximatelyhalf the width of those on our piece, suggesting that our Nike may have been the substantially larger of the two. Despite these differences, no other interpretation for our wing comes to mind. Late 6th century B.C. or later? 13 Human left foot SF-69-8. P.H. 0.022, p.L. 0.051, W. 0.058. Left foot from ball to toe tips, broken at attachment to plinth. North edge of Temple Hill, roughly north of center of temple, in 13thcentury fill, no lot.

Fig. 26 Clay:fine thin surfacelayer,fired buff lOYR8/3; core:raresmallblack mudstoneinclusions,firedpink 7.5YR 7/5.

Figure 26. Miscellaneous pieces. Scale 1:2

102. Goldberg 1977, pp. 174,239. At Halai, wing no. 4 of sphinx I (Goldman 1940, figs. 112-113) appears to presentpainted featherson both faces. But that fragmentis also unusual in preservingoutlined scales that continue down the body,beyond the wing. For similarscales see Weinberg 1957, no. 36 (SF-27-1), p. 315, pl. 72, now 20c below; this is no longer maintainedto belong to a sphinx,but to a sea creature.

CORINTHIAN

TERRACOTTA

SCULPTURE

435

A bare left foot of a human, roughly half life-size, which originally projected from beneath the hem of a long garment, the scar from which runs across the top of the foot at the ball. The toes rest flat on the missing plinth but are skewed slightly to proper left. The longest toe is the second toe, which extends well beyond the big toe; from it the other toes diminish in size to the little one, which curls in against the fourth toe. In profile the toes are relatively flat; they are long, first and second toes are separated,and the nails are square in outline, cut straight across the tips. Somewhat hastily modeled, the surface of the foot preserves numerous paring marks, and it is not polished. There are traces of black paint on the preserved edge of the plinth that surroundedthe foot. The foot is solid. From the underside it appears to have originally been shorter,its toes closer together, and the first two toes the same length. Additional clay was apparentlyadded to change the form. The underside of the foot is rough where it separated from its plinth. A slight projecting ridge of clay below the ball may mark the edge of a ventilation hole. Too little remains to help determine whether the fragment belonged to an acroterion or a freestanding figure. That it doesn't belong with the wing, 12, is shown by the considerable difference in fired colors of the two clays. A foot of a Nike from Olympia (OlForschXXII, F6, p. 87, pl. 60:f, g), dated to the beginning of the 5th century B.C., provides a useful comparison. The first two toes of the Olympia foot are equal in length, and all five, including the little toe, are straight. In profile the toes slant downward. A similar foot belongs to one of the striding figures on the Amazonomachy pediment from Corinth (Weinberg 1957, pp. 306-307, no. 8, SF-32-3, pl. 65), also dated to the early 5th century.The differing lengths of the first two toes of our foot, the curling little toe, the flatter profile, and the hasty finish of the surface argue for a date no earlier than the second quarterof

the 5th centuryB.C. 5th century B.C. 14

Strut?

SF-72-16. P.L.0.17, W. 0.055, p.H. 0.048-0.070. Strut,brokenat eitherend. EarlyRomanfill,west quarrytrench,lot 72-175.

Fig. 26 Clay:thin fine layeron one side; surfaceand core:commonfine black inclusions,raresmall(chiefly0.002 but ranging from < 0.001 to 0.003), fired yellow throughout, 5Y 8/2-8/3.

A massive bar, modeled solidly of clay; approximatelyrectangularin section, its upper surface slightly rounded. At one end, the upper surface rises at an angle toward the broken end; at the other end, both sides and the upper surface begin to flare.The bottom is smoothly flat but in no way finished. One long side is carefully masked with a fine slip which extends partly onto the top. The rest of the top, and the back faces, are both irregularand without a surface slip. This piece has been included for purely technical reasons.Tentatively identified as a large strut, it differs from the lighter cylindrical struts, such

436

NANCY

BOOKIDIS

as 7r and 18c, which generally prop apart sphinx wings. While it too may have been a prop for wings, the finish of the bottom suggests that the piece originally rested on something equally flat, such as a plinth. The one smoothed face was clearly what was to be seen. Where this strut might have been used is unclear; perhaps, in some way, it braced the legs of a sphinx. That sphinx legs were not always completely free is shown by Weinberg 39 (Fig. 4; Weinberg 1957, SF-32-12, p. 316, pl. 72), which preserves a thin partition down its length. At the same time, an alternate interpretation cannot be overlooked, namely, that the fragment derives from the back leg of a large, seated sphinx or lion like one from Olympia (Moustaka 1984, especially pl. 31:2; OlForschXXII,Lll, p. 124, pl. 1O1:a). In accordancewith that parallel, the angled end (Fig. 26, at right) might mark the beginning of the "elbow,"while the other end might be widening to form a paw.There is, however, no evidence of anatomical modeling of the bones of the legs such as appears on the Olympia example. Moreover, the lack of slip over the entire leg becomes problematic. For these reasons the identification of the fragment as a strut seems preferable.The piece does not precisely resemble any of the other fragments from Temple Hill. Whereas its surface color is closest to 3 and 4, that of its core is yellow rather than gray-green. With so little preserved, a date cannot be assigned to this piece. 15 Disc(?) a. SF-74-5a. P.H. 0.060, p.W. 0.028, max. Th. 0.037. Broken on all edges, preserving part of two faces and intermediate strut. Early Roman fill, quarrytrench 4, lot 74-181. b. SF-74-5b. P.H. 0.080, p.W. 0.067, Th. 0.015. Broken on all edges, preserving decorated face. Modern surface find on Temple Hill, 1972.

Fig. 27 Clay:fine thin surfaceslip, Th. < 0.001, firedbuff lOYR8/4; core:frequentfine to small(to 0.002) red andblackinclusions,fired pink 6.5YR 7/6. Paint:red-brown 5YR 6.5/4.

These two nonjoining fragments are problematic and do not fit in with the other pieces described thus far. 15a consists of two flat and parallel surfaces, one decorated, the other plain, which arejoined by a strut 0.013 m thick; an airspace of ca. 0.01 m is left between them. The surface of the plain face was smoothed while the clay was moist; inclusions are apparent.More of the decorated surface is preserved on b. This consists of a roughly flat surface,the center of which is framed by a broad groove that, on a, flattens into an angular offset. The groove on b appears to form a curving line, and as such would outline a central tondo, but too little is preserved for certainty.The interior surfaces of both pieces are irregularand clearly were not seen. Although carelessly drawn, the decorated face on 15b is painted with what could be the face of an owl or a Gorgon, in so far as it is preserved. Parts of two large round eyes, closely set, are drawn in a thick black outline on the buff ground and are fringed along the inner circumference with black lashes. The small, solid black pupil is also fringed

CORINTHIAN

TERRACOTTA

437

SCULPTURE

V 15a

Figure27. Disc(?).Scale1:2 with black lashes. A dull light reddish brown is used for the face. Above and between the eyes is a triangle, drawn in short horizontal strokes of black, which might represent the ruff of feathers that surrounds the eye of an owl. At the same time, it must be admitted that the drawing of the lashes makes more sense if the fragment is turned 90 degrees in either direction so that one eye lies above the other, for the lashes within the better preserved eye then fall more naturally.'03The shallow groove that frames the face is painted black. Above this are the tips of possibly vertical stripes, in red and reserved colors, outlined in black. Continued onto fragment a are both the stripes and the framing band, below which is a much smaller triangle outlined in black. The strut that separatesthe two surfaces of 15a lies beneath the stripes at a right angle to them and therefore outside the central tondo. No trace of such a strut exists on b. This could mean that the strut was not continuous but only occurred at intervals, or that the two fragments derive from two distinct but similar objects.

103. My thanks to RoxannaDoxan for this observation.

Both the function and the subject matter of the fragments are unclear. Clay color, inclusions, and painted colors arevirtually identical to those of wing 9, and yet it is unlikely that they could have derived from the same statue. The walls are relativelythin, indeed extremely so, and yet the total thickness of a is greater than that of 9. The back face of a is clearly a finished surface, much like the backs of sphinx wings; it is impossible, therefore, that, like an aegis, the piece could have decorated the front of a human statue. It is also unlikely that it was part of a shield. The shield carried by the terracotta warrior at Olympia (OlForschXXII, B20-21, pp. 38-40, pls. 25-30) is, in fact, modeled like a bronze one. As such it is convex on the exterior, not flat like our piece. In addition, it is a single

438

NANCY

BOOKIDIS

thickness, not double like the Corinth disc. We have cited the similarities that exist among 15, 9, and a ridge palmette from the temple in terms of the fired colors of both clay and paint (above, note 24), which are unparalleled among the other terracottas.These parallels suggest an architectural function for 15. For want of a better identification, we have called it a disc because of its circulardesign. But we have avoided the term disc acroterion, since this form of architecturaldecoration is alien to the Corinthian tradition. The scale of our piece is also too small for that function. With regardto the image, the eyes could belong to either an owl or a Gorgon. That it is a Gorgon seems unlikely for the following reasons.The eyes are entirely circular,whereas Gorgons have large circularpupils within a human-shaped eye; the fringing of both the pupil and the outline of the eye is not human but seems more suited to an animal. Finally, the deep V or triangle created by the shaggy eyebrows recalls the feathered ruff that surroundsan owl's eyes. Compare it, for example, with the head of an owl on an Early Corinthian aryballos,the name piece of the Painter of Delos 330 (Amyx 1988, p. 89, pl. 38:c). By contrast,the outline of the nose would generally be drawn between the eyes of a Gorgon. The date of 15 is unknown. Stylization of the representationsuggests an Archaic date; at the same time, the carelessness of the workmanship is unusual for that period. In keeping with 9, it should date no earlier than the 5th century B.C. 5th century B.C. or later. II. PIECES

FOUND

UNKNOWN

PROVENIENCE

OUTSIDE

16 Shoulder of a sphinx SF-18. P.H. 0.120, p.W. 0.055, Th. 0.027-0.045. Shoulder,broken on all edges.Provenienceunknown. Clay:fine surfacelayer,Th. 0.005, firedtan lOYR7/5; core:

OF TEMPLE

HILL

OR OF

Fig. 28 abundantfine blackand rareredand white inclusions(to 0.001), firedpale yellow2.5Y 7/2. Paint:purplishbrownlOR4/3; black5YR 3/1.

Part of the shoulder of a sphinx that is close in scale to Weinberg 33, broken at or near the juncture to the chest (Fig. 28). Although the piece may derive from either shoulder, the left seems more likely. So placed, the fragment preserves the front of the shoulder as it breaks before the side. A slight contraction of the curve of the shoulder is apparent at the lower break, presumably as it tapers to the start of the leg. On the interior (Fig. 28) can be seen a spear-shaped impression of two flattened surfaces, at acute angles to each other, that may have been made by sticks used for support during modeling. Scaly feathers, painted on the shoulder, are arrangedin rows that rise slightly toward the upper side. Carefully drawn but not quite uniform in size, the scales have short stems, are arrangedin vertical rows of purplish red, then black, and have reserved and heavy black outlines. The fragment very much resembles Weinberg 33 and may well date from about the same time, in the third quarterof the 6th century B.C. Third quarterof 6th century B.C.?

TERRACOTTA

CORINTHIAN

439

SCULPTURE

_ _ _ _ _ _K -

L

_ _

,.

_F _d

|t L

.

,Ei

.

>

l

Figure28. Sphinxshoulder16.

_

Scale 1:2

17 Right wing of a sphinx SF-24. P.L. 0.183, p.H. 0.185, Th. 0.055 (top)-0.073 (bottom). Front and back faces of wing, broken on all edges. Early excavations, provenience unknown, found by

1915.

Fig. 29 clayfiredyellow,redwhereclayfired pink,firedpalepink.Paint:redbrown5YR 5/4 to 2.5YR 4/2 to lOR 3/3. Published: CorinthIV,i, p. 113, M7; Koch 1915, p. 82, no. 1.

Clay: fine surface layer,Th. 0.005-0.01, fired pale yellow, ca. 2.5Y 7/3; core: abundant fine to small inclusions (to 0.002), black where

A small portion of an enormous right wing preserves parts of three primary feathers and the tip of one covert feather.The primaries are stepped up in relief, the single complete feather having a width of 0.075 m. Each center is painted brown, shading to red-brown where more dilute, and each is set off by a broad reserved stripe with a redbrown stripe along the outer edge. Unlike the modeled primary feathers, the single covert feather is simply painted in a brown outline with reserved center. A measure of the piece's large scale is shown by the fact that the feathers show only a very slight curve. In terms of technique, the feathers are modeled entirely in fine clay,varying in thickness from 0.005-0.01 m, while the flat back is finished with a paper-thin slip that is mottled from firing but not painted. In execution, the wing is not equal to 2c, but it represents the largest terracotta sphinx that has survived in Corinth. From the thickness of the fragment and its weight (2.25 kg) we can conclude that the making of the whole figure was a technological accomplishment of the highest order. This appearsto be the piece published without illustration in Corinth IV, i, p. 113, as M7. It does not belong with M5, however, as suggested therein, since the latter piece has now been joined to 7h and is quite different both in scale and treatment. Nor is Koch's observation (1915, p. 82, note 6) correct, that it derived from a large circular acroterion, for the

NANCY

440

6th~~ cetr

BOOKIDIS

B.C.?

Figure29. Sphinxrightwing 17. Scale 1:2

feathers of the sphinx wing are quite clear. Its date is uncertain, but to cliiay, a 6h-ctur

de

is l

CORINTHIAN

TERRACOTTA

SCULPTURE

18 Leg, tail, strut of sphinx a. SF-28b. P.H. 0.097, p.W. 0.050, Th. wall 0.014-0.045. Back half of foreleg, surface chipped. b. SF-28a. P.H. 0.105, Diam. tail 0.024-0.032. Tail, broken at attachment to rump and before tip. c. SF-28c. L. 0.124; Diam. 0.033-0.036. Strut, complete, broken at attachment to wings.

44I

Fig. 30 Early excavations, provenience unknown. Clay: thin fine surface layer, Th. < 0.001, fired unevenly at surface, buff to tan, 2.5Y 8/3 to 7.5YR 7.5/8; core: frequent fine black inclusions (to ca. 0.001), fired pale buff, ca. lOYR 8/3.

Three nonjoining fragments are grouped on the basis of clay color and mottled firing. 18a (Fig. 30) preserves the back half of the foreleg of a sphinx that may have been the size of 7. It differs somewhat from other legs in that it is not rounded or oval in back but virtually triangular, the sides thinning to a very narrow,flattened ulna. The ulna ends just above the carpals in a pronounced bump, the tip of which is broken. Of further interest is the fact that the leg may have been modeled freehand, as is suggested by the generous central cavity, rather than built up around a central rod. The tail, b, which tapers in thickness from rump end to tip, makes a complete loop, then crosses to the left of the thicker end before breaking off. It is broken along a joint, which preserves the impression of a rounded surface,presumably the back of the sphinx. Finally, the strut, c, which originally must have propped apart the wings, is round in section like 7r. It is pierced from end to end by a narrow hole 0.01 m in diameter and flares at either end where it was attached to either wing. The strut is more carefully finished than 7r, was not painted, and exhibits the same mottled firing that characterizes 18a and b. All of the fragments are finished with a very thin layer of fine clay that is unpolished and that does not completely cover the underlying inclusions.

104. The mold may be that describedin CorinthXV, i, p. 112, as coming fromTemple E.

The pieces described here have been catalogued from a box of fragments from early excavations that lack specific provenience. Among these were a torso that maybelong to the Amazonomachy publishedby Weinberg, and a mold for an ornamental disc,104both of which may have come from the well by Temple E. Therefore it is possible, but by no means certain, that this sphinx also came from that area. Comparison with the better preserved marble sphinx in the Metropolitan Museum in New York (Richter 1961, no. 37, pp. 27-29, figs. 97, 100) clarifies the exact position of our tail. It formed a figure eight, turning up from the rump against the hindquartersbefore making a loop and ending against the flank. The crossing of the tip apparentlyoccurs on the side toward which the sphinx is looking. Therefore, our sphinx must have faced left. The thickness of the tail and the size of the loop suggest a sphinx larger than Weinberg 33. The surface finish, which closely resembles that of sphinx 7, argues for a date in the 5th century B.C. It may be of interest to note here a difference between at least two of the Kalydon sphinxes and those from Corinth in terms of technique. The

NANCY

442

BOOKIDIS

Tail 18b

Foreleg18a

Figure30. Sphinx18.Scale1:2

Strut18c wings of sphinxes D and G from Kalydon are reinforced at their base in the following manner: a solid shelf-like wall projects from the upper back for an estimated 0.05-0.06 m, running from wing to wing to give them support and to reinforce their juncture with the body. The "wall"is preserved on sphinx D, and can be restored on G on the evidence of the broken surface on the back. This system does not appearon Weinberg 33 nor on statue 7, above.The circularstruts used in Corinth are not attested at Kalydon; however, the wings of the Kalydon sphinxes may not be preserved to a sufficient height to display this feature. It would be interesting to know whether the difference is chronological or regional.

5th centuryB.C. 19 Scaly body, sphinx(?) SF-19. P.H.0.104, p.W. 0.106, Th. 0.024-0.034. Body,brokenall around.Provenienceunknownbut apparentlyfoundbefore1915. Clay:thin fine surfacelayer,Th. 0.002-0.005, firedyellow-green5Y

Fig. 31 7/3; core: frequent fine to small angular red and rarewhite inclusions, fired bright pink 5YR 6.5/6. Paint: red-brown 2.5YR 4/4 where thickest; black.

An anatomical fragment, with a pronounced convex curve in vertical section, thickens at the bottom right as the interior surface of the body turns inward. A portion of smooth body surface runs along the lower third of the fragment. As the fragment is oriented in Figure 31, this

CORINTHIAN

TERRACOTTA

443

SCULPTURE

Scalybody,sphinx(?)19

20a

20c

Sea creature20

Figure31. 19 and20. Scale1:2

20d

smooth surface is broken only at the left edge by a single recessed scale. The body is otherwise unpainted and lightly polished. From this body, a surface projects in low relief (ca. 8 mm); its lower edge is scalloped to form at least two sharp points or lobes. This surface is decorated with scales running horizontally.They are painted in vertical rows of black, alternating with rows of red-brown, within a reserved outline, and are of a uniform size. The statue is hollow, its interior surface smooth.

444

NANCY

BOOKIDIS

What animal this represents is not clear.Although the scales suggest a sphinx, parallels are lacking for a sphinx with lobed wing or for one with scales extending so far along the side. Nor is it clear that the orientation shown in Figure 31 is correct.That we have chosen this position is owing to the vertical curvature as contrasted with the flatness of the piece in horizontal section. If the piece is turned 180 degrees so that the smooth body falls at the top rather than the bottom, its interpretation becomes even more perplexing, for the vertical curvatureis too steep for a shoulder. As for a date, the color scheme of the scales and the lightly polished surface treatment of the body suggest that it was made in the 6th century B.C., but probably no earlier than the second half, given its relatively complex form. Second half 6th century B.C. 20 Sea creature a. SF-87-lb. Max.p.dim.0.114, Th. 0.022-0.029. Part of body and dorsal spine. East of the Theater, east of Building 3, in surface fill. b. SF-87-la. Max. p.dim. 0.147, Th. 0.02-0.042. Two joining fragments, part of body and fin. East of the Theater, east of Building 1, in surface fill. c. SF-27-1 (Weinberg 36). P.H. 0.092, p.W. 0.099, Th. 0.03-0.032. Body, broken on all edges. Athena Trench, east of Building 3, in surface fill.

Fig. 31 d. SF-26-4. P.W.0.061,Th. 0.013-0.018. Finding as c. Clay: fine surface layer,Th. 0.001-0.002, fired yellow 2.5Y 7.5/ 4.5; core: frequent fine inclusions, chiefly black and a few red, and rare small inclusions, fired pale pink lOYR 6.5/4. Paint: black 7.5YR 3/1. Published:c: Weinberg 1957, no. 36, p. 315, pl. 72.

Four tantalizing fragments belonging to a sizable sea creatureof some sort, preserving only a small part. The animal to which they belong has a body that is round and covered with scales, in so far as the body is preserved.These scales are simply outlined in a broad black line against the yellow clay of the surface,but, given that they should point away from the head, they help to establish the orientation of the fragments. Described individually,the fragments are as follows. 20a probably belongs to the upper right side of the creature,for it preserves what may be the dorsal spine. This consists of a black-painted sloping surface that runs the length of the piece. Beneath this are the scales, of which are preserved part of one row (W. 0.035 m) and the tips of a second. A black cross-line cutting through the bottommost scale may be the remnant of some further decoration. In horizontal section the fragment is slightly concave. 20b preserves the largest scales, of which parts of three rows are preserved. Slightly concave along the long axis of the scales, the body is partly covered by a large fin, of which two pointed lobes remain. The fin stands out in relief from the body, and plastic ribs extend to the tip of each lobe; the upper lobe presses down against the body while the lower one projects more freely.The surface of the fin is reserved,but a black

CORINTHIAN

TERRACOTTA

SCULPTURE

445

stripe outlines its edge. The scales of b diminish in size both from presumed top to bottom and from right to left. Since the largest scales (0.04 m wide) lie just beneath the fin, and since the remaining scales decline in width from 0.04 to 0.036 m, the fin should lie near the base of the fish's head. Because of the way in which the lower lobe opens out from the body, we assume that the fin is a ventral one. Two vertical rows of scales and the tip of a third are preserved on c. These diminish in size both from top to bottom and from right to left (W. 0.031-0.024 m). Slightly convex both horizontally and vertically, the piece in vertical section does not make an even curve but is fuller toward the bottom and flatter toward the top. Finally, d, the smallest, preserves one row of scales and the base of a second. Within the uppermost scale is a second black stroke, the purpose of which is unclear.This piece too is very slightly concave along the axis of the scales. The statue is hollow. Its wall is not of a consistent thickness: c is the thickest, a is close in thickness to c, and the bottoms of b and d are the thinnest. A vertical partition 0.013 m thick runs the length of the body beneath the dorsal spine, 20a, presumably to give the spine support. The body is worked chiefly in coarse clay, including the core of the fin and spine. On c, the fine surface clay has broken away,exposing a second, thin fine skin directly over the coarse, as if the outer layer had been added to a figure that had alreadybeen blocked out. 105. Weinberg 1957, no. 36, p. 315, pl. 72; Goldman 1940, p. 443, nos. 4-5, figs. 112-115. There is much about this statue that is unusual,such as the scales that continue over the lower body and the squaretorso with incipient breasts. One wonders if it could have been a siren ratherthan a sphinx. 106. A useful chartwith fish terminologycan be found in McPhee andTrendall1987, p. 171, fig. 2. 107. A possible exceptionis the terracottaMelian relief,depicting a Triton holding a youth, where a fin of a differentform is visible along the bottom of his body.See Jacobsthal 1931, no. 3, p. 17, pl. 3. For an Etruscan ivory from Corneto, now in the Louvre, on which is a fish-tailed monsterwith numerousfins, see Shepard1940, p. 32, pl. VI:45. A useful surveyof sea monstersin vase-paintingis in Ahlberg-Cornell 1984. 108. For the Agrigento pyxis C948 now in Wiirzburg,see Boehringer 1929, p. 84, figs. 9:a-c, especiallyc, and McPhee andTrendall1987, p. 60, pl. 12:A. A drawingof the vase from Selinos appearsin Kekulevon Stradonitz1884, pl. 57 and p. 83.

One of these fragments, c, was known to Weinberg, who assigned it to a sphinx on the analogy of the unusual sphinx from Halai.105Clearly that interpretation cannot be correct, for when we put these few pieces together, we have a sea creaturewhose body is round, not quite cylindrical, and roughly 0.16-0.17 m in diameter.The few existing fragments suggest that the body was flexed but to what extent cannot be determined. Presumably,the body tapered from one end to the other, as the diminishing scales indicate. It apparentlyhad a dorsal spine that may have supported one or more dorsal fins. In addition, it had a large web-like fin, which we tentatively place behind the head, on the belly.106 It may be easier to say what 20 certainly isn't. It isn't a dolphin, for a dolphin's skin is not scaly, as the acroterion of a dolphin from Olympia makes clear (OlForschXXII, O1, pp. 140-145, pls. 110-111). It also differs from the possible dolphin published by Weinberg (1957, no. 32, p. 314, pl. 70). It probably is not part of a fish-tailed monster like Typhon, for he is generally shown without fins like this.107 But the fragments could have belonged to some other kind of sea creature,such as a ketos or a seahorse. That sea creatureswere modeled in clay is shown by two fragments from Kalydon (Dyggve 1948, pp. 185-188, fig. 197), although neither is quite like ours. A Boiotian red-figure pyxis in Wiirzburg and a relief vase from Selinos provideglimpses of what such an animalmight have been,108namely, a creaturewith a horse-like head, serpentine body, numerous dorsal fins, and fewer ventral ones. Even better parallels for the flipper-like fin, here placed on the belly, can be found in representationsof a ketos in Scythian art. Among the examples are a gold applique of the late 4th century B.C.

446

NANCY

BOOKIDIS

from Kul' Oba (Michel 1995, p. 87, fig. 26) and a Hellenistic gold plate from Alexandropol (Minns 1913, pp. 154-155, fig. 42).109 While the Corinthian piece is not as late as these examples, it may well depict the same kind of animal. The function of this piece is by no means clear. If it was as big as its circumference suggests, it was probably too large for an acroterion.110It might, however, have filled the corner of a small pediment. Scales of this sort appearon the sphinx from Halai, cited above and dated to the end of the 6th century B.C. They also decorate the breast of a sphinx from Olympia (OlForschXXII,J18, pp. 109-110,114-115, pl. 93:b). The carefulplastic rendering of the fin may suggest a date in the 5th century B.C. End of 6th to 5th century B.C. 21 Sea creature(?) SF-25. P.L. ca. 0.16, p.W. 0.12, Th. 0.015-0.05. Back of an animal, with part of dorsal fin, tips of which are broken away,surface chipped. Provenience unknown.

Fig. 32 Clay:fine surfacelayer, Th. < 0.001, firedpaleyellow,ca. 2.5Y 8/3; core:frequentfine to small blackandredinclusions,at leastone largeredinclusion(0.007),firedpale pink7.5YR 7/4.

A single fragment may be further evidence of the use of sea creatures as architecturaldecoration at Corinth. Preserved is the rounded back of an animal, from which projects a heavy fin (Fig. 32). Both ends of the fin are preserved, as well as the stumps of three projecting lobes. On each of these is painted a single spine in black. At the base of the fin are horizontal rows of parallelzigzags in relief that extend both along and down the body; at least four rows are preserved on the right side and two on the left. These zigzags are quite angular near the spine but graduallyflatten out into the surface of the back. They are painted a matt black against a reserved ground. Where the relief zigzags flatten out, the solid black paint breaks into wispy streaksthat curl, much like the ends of locks of hair. If the piece is viewed from the front-in cross-section (Fig. 32)the following is apparent.The fin is a slab of clay, 0.016 m thick, which, at its base, curves out on either side to form a body of the same thickness. Onto this structureat the base of the fin has been added a mass of clay as much as 0.038 m thick, in which is modeled the zigzag pattern. If one looks at the animal from its right side and, more particularly,at the broken surface that runs along the present base of the right side, the addition of this clay to the fish body becomes even more apparent.Here the wall tapers in thickness from 0.047 m, at right, to 0.017 m at left. In the thicker right half, two distinct layers can be distinguished. The underlying body consists of a layer of coarse clay, covered by a thin surface of fine clay, the whole 0.018 m thick. Over this is a second layer of coarse clay, finished with fine clay that forms the skin of the fish. At the right edge, the lower layer of fine clay preserves a small patch of black paint. If this lower fine surface represents the body of the fish, then the rest may be thought of as an added element. Modeled like 20, the fish is hollow but has a vertical partition 0.01 m thick that supports the dorsal fin.

109. See also Boardman1987 for useful representationsof the ketos. 110. I regretthat Eva Rystedt's dissertationon the earlycut-out acroteriafrom Murlo and Acquarossa was not availableto me here in Athens in orderto comparethe Corinth piece with a possible marineacroterionfrom there.

CORINTHIAN

TERRACOTTA

..447

SCULPTURE

':~~~~~~~~~~~~~~~I:

21. Scale 1:2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I E Figure32. Seacreature(?) Once again, that we are dealing~~~ wIt an auaianmlsesiky hte t ane pne.Wa sntceri atta stepaeca oo n htaloh mottled.z some way blackis something appliedto it, or whether the body is in Zig The rliefzigzas an wisp pait frrtherobscre th idetifiction

from~~~~ th doslfnwt ~ ~~ th ihi.h the~~~~~~~~~~ bod ofS

NANCY

448

BOOKIDIS

.......... ... ...... .............

...

............

..........

N,

`2

.....

..........

...

':f

........... :ffW

=

:=

;...... : ...... :; :

..........

......

........

:l,fz::f:::::fn.:f;;:,' ................ .......................... ... ..........

...........

.............

... .....

. .

. ..... ......

... .... .... .................

.. .... ... . ..........

::

. .. ... ..... ::f: :n

... ... . ... ..

. .....

.. ................

2

.............. ...

... .....

..

. ........................ .... .................

. .....

..

. .....

... .... .... ll..::

..................

... ... .. ... ... .... ... . .. ....

.. ... .............. .

..

..... ...............

... . .. . ... .... .. ..

.. .....

. ..........

... ......

........

...

... ..

........

.........

... .

...........

........

... .....

.... .....

.....

.....

... .. .... ......... .... .........

f

.....

M :f':tff:".

....

. ......

..... .... ..... ...

......... -n ffs

ff

...

.....

............

..

............ ..... ..... ..... .... :':: : ::: f- f :Nf. : f' ............. . . .. .....IN:................... f.. ; %X;;:

..........

................. ..........

............ ..............

.....

.

.. .......

...... ....

.. ... . ...

: :;M .':'.:::';:: ... ..... ......... .........

.... ............................. ..............................

.... ...

..................

....

..

............... :";:::;I:; . ..........

.......................... ............

... ..

. .......

.... .............. ..... ........ .... ................. .... . .....................................

.

.. .....

..................

.............

t=msmMl

... . ....

.....

............. ...

...

...

.: . ..... .... .....

..

.. . ....

.... . .

................ : ;.:.,-:!;:::::,:.;:!:::::.:. ..... ..... 11................:.

.

... ....

....

......

..............

........

...

........................

.... .. ......

...

...

......

..

..

..

.......... ............................

... ...

....................

....

......

...

...

....

............. ..........

...

Ut

View of interior through

. . ...........

neck

CORINTHIAN

TERRACOTTA

449

SCULPTURE

zags are a common convention for hair and the wispy finish to the black bands is also quite reminiscent of hair.But a hairy fish is unlikely.Nor is it easy to imagine how a human figure, riding a fish, could have its hair extended along the fish'sback. Zigzags could also be a convention for waves and it is possible that what is representedis a fish rising from or swimming in the sea. A fish with zigzags on its back occurs in the pebble mosaic floor that decorated the pronaos of the Temple of Zeus at Olympia (L'Orange and Nordhagen 1966, pl. V). Although upright waves form the upper border of the fish panel, possibly the zigzags on the fish are a further reference to water.The wispy lines of dilute paint on our statue might then represent water as the waves thin against the fish's body. But if this is the case, it seems strange that the water should have been added as a thick mass of clay to the upper front half of the fish, where it should be thinnest. The dolphin acroterion from Olympia reflects the more conventional way of depicting a fish in water, leaping above a sprayof water, there represented by curling sprays. The interpretationof this piece must remain problematic,as is its date. It should be no earlier than the late 6th century and possibly as late as the 5th century B.C. Like 20, it could also have decorated the fa9ade of a small building. Late 6th-5th century B.C. CONCORDANCE

INV. No. SF-la SF-lb SF-5 SF-6 SF-7a SF-7b SF-18 SF-19 SF-24 SF-25 SF-26-1 SF-26-4 SF-27-1 SF-28a SF-28b SF-28c SF-28-3 SF-33-1 SF-69-8 SF-72-3a SF-72-3b SF-72-3d SF-72-3e SF-72-3f SF-72-3g SF-72-3i SF-72-3j SF-72-3k SF-72-31

CAT. No. 8c 8b 7h 2c lla llc 16 19 17 21 2a 20d 20c 18b 18a 18c llb 4a 13 7a 7b 7c 7d 7e 7g 7h 7f 7i 7m

INV. No. SF-72-3m SF-72-3n SF-72-3p SF-72-3q SF-72-3r SF-72-3s SF-72-3t SF-72-3u SF-72-14a SF-72-14b SF-72-14c SF-72-15 SF-72-16 SF-74-la SF-74-lb SF-74-2 SF-74-3 SF-74-4 SF-74-5a SF-74-5b SF-75-3 SF-75-4 SF-75-5 SF-76-4 SF-76-5 SF-77-4 SF-87-la SF-87-lb

CAT. 7n 7p 7j 7o 71 7q 7r 7k 9a 9b 9a 10 14 la lb 2d 2b 12 15a 15b 8a 8d 4b 5 3 6 20b 20a

No.

450

NANCY

BOOKIDIS

REFERENCES Agora XI = E. B. Harrison,Archaic and Archaistic Sculpture, Princeton 1965. Ahlberg-Cornell,G. 1984. Herakles and the Sea-Monster in Attic Black-Figure Vase-Painting, Stockholm. Alt-Agina 11,2 = E. Walter-Karydi,Die aginetische Bildhauerschule,Mainz

1987. Alt-Agina 11,4 = K. Hoffelner and M. Kerschner,Die Sphinxsaule: Votivtriger,Altare, Steingerate. Perirrhanterien und Becken, Mainz

1996. Amyx, D. A. 1943. "CorinthianVases in the Hearst Collection at San Simeon,"CPCA 1, pp. 207-240. . 1988. Corinthian Vase-Painting oftheArchaic Period, Berkeley. Ashmole, B., and N. Yalouris.1967. Olympia. The Sculpturesof the Temple of Zeus, London.

Billot, M.-E 1977. "Recherchessur le Sphinx du Louvre CA 637,"BCH 101, pp.383-421. Boardman,J. 1987. "'VeryLike a Whale': ClassicalSea Monsters,"in Monsters and Demons in theAncient and Medieval Worlds:Papers Presented in Honor of Edith Porada,

A. E. Farkas,P. 0. Harper,and E. B. Harrison,eds., Mainz, pp. 7384. Boehringer,E. 1929. Die Miinzen von Syrakus,Berlin. Bookidis, N. 1988. "Classicismin Clay," in T4paxrtxa' IOV XII flcz&Ovoo Eovo6pl'oo KAcacxcr ApXaWo2oylc'csIII, 1983, Athens, pp. 18-

21. . 1995. "ArchaicCorinthian Sculpture:A Summary,"in Corinto e l'Occidente (AttiTaranto 34), pp.231-256. . Forthcoming."The Sanctuaries of Corinth,"in Corinth XX. Bookidis, N., andJ. E. Fisher.1972. "The Sanctuaryof Demeter and Kore on Acrocorinth.Preliminary Report IV: 1969-1970," Hesperia 41,pp.283-331. Bothmer,D. von, andJ. V. Noble. 1961. An Inquiry into the Forgery of the Etruscan Terracotta Warriorsin the Metropolitan Museum ofArt (Papers

no. 11), New York.

Buschor,E. 1957. "Altsamischer Bauschmuck,"AM,pp. 1-34. Charbonneaux,J. 1943. La sculpture grecqueclassiqueI, Paris. CorinthI, i = H. N. Fowlerand R. Stillwell,Introduction,Topography,Architecture, Mass. Cambridge, 1932. CorinthIV, i = I. Thallon-Hill and L. S. King,DecoratedArchitectural Terracottas,Cambridge,Mass. 1929. CorinthXV, i = A. N. Stillwell, ThePotters'Quarter,Princeton1948. Danner,P.1989.GriechischeAkrotere der undklassischen archaischen Zeit (RdA Supplement5), Rome. . 1993. "Die Dekoration auf First und Giebelschragenin der archaischenBaukunstMittelin DeliciaeFictiles. italiens," Proceedings of theFirstInternational on CentralItalicArchitecConference turalTerracottas, Rome,10-12 Dec. 1990, E. Rystedt,C. Wikander,and O. Wikander,eds., Stockholm, pp. 93-107. .1997. WestgriechischeAkrotere, Mainz. Dontas, G. S. 1997. "`xE%t4;, xoat nPRo-rasstgEL 1lnPopx1tiati,olty yt(a T-v yXtdYCTL)nxT-) at) GTovU; aPxPCovL X 1pWttloUt;

X\c(JLxob;

KEspxtPc; Tot) xeo6voD;,"

in -Ew=xivos Icodvvoo K. Lx7rx6V7p7zlT'Oo, B. Petrakos,ed., Athens, pp. 53-170. Dyggve, E. 1948. Das Laphrion, der

Tempelbezirkvon Kalydon,

Copenhagen. Eggebrecht,A., et al., eds. 1988. Albanien: Schiitze aus dem Land der Skipetaren, Mainz. FdD II = C. Le Roy andJ. Ducat, Fouilles de Delphes, II Topographieet architecture.Les terres cuites architecturales.La sculpturedJcorative en terre cuite, Paris 1967.

Felsch, R. C. S., and H. S. Kienast. 1975. "Ein Heiligtum in Phokis," AAA 8, pp. 1-24. Felsch, R. C. S., H. S. Kienast,and H. Schuler.1980. "Apollonund Artemis oder Artemis und Apollon?"AA95, pp. 38-115. Furtwangler,A. 1906. Aegina, Munich.

CORINTHIAN

Goldberg,M. Y. 1977. "Typesand Distributionof Archaic Greek Acroteria"(diss. Bryn Mawr College). Goldberg,M. Y. 1982. "ArchaicGreek Acroteria,"AJA 86, pp. 193-217. Goldman, H. 1940. "The Acropolis of Halae,"Hesperia9, pp. 381-514. Heiden, J. 1987. Korinthische Dachziegel,Frankfurt. Higgins, R. 1954. Catalogueof the Terracottas in theDepartmentof Greekand RomanAntiquities,British Museum,London. Holtzmann, B. 1991. "Une sphinge archaiquede Thasos,"BCH 115, pp. 125-165. Homann-Wedeking,E. 1942. "ArchaologischeGrabungenund Funde in Italien,Albanien und Libyen,"iA457, cols. 277-389. Hiibner,G. 1990. "Die Dachterrakottender archaischen Tempel von Kalapodi,"Hesperia 59, pp. 167-174. Jacobsthal,P. 1927. Ornamente griechischer Vasen,Berlin. . 1931.Die MelischenReliefs, Berlin. KalapodiI = R. C. S. Felsch, ed., Kalapodi I, Mainz 1996. Kalligas,P. G. 1968. "Kepxopoc," ArchDelt 23, B2, Xpovixoc, pp. 302322. Kekulevon Stradonitz,R. 1884. Die Terrakottenvon Sicilien, Berlin. Koch, H. 1915. "Studienzu den CampanischenDachterrakotten," RM30, pp. 1-115. Kunze, E. 1941. "Terrakottaplastik," in ITTBerichtfiberdieAusgrabungen in Olympia, E. Kunze and H. Schleif, eds., pp. 119-132. L'Orange,H. P., and P.J. Nordhagen. 1966. Mosaics, London. Lulof, P. S. 1991. "Monumental TerracottaStatuesfrom Satricum" (diss. Universityof Amsterdam). . 1996. The Ridge-Pole Statues from the LateArchaic Temple at Satricum (Scrinium XI, Satricum V),

Amsterdam. McPhee, I., and A. D. Trendall.1987. GreekRed-Figured Fish-Plates

(AntK-BH 14), Basel. Michel, S. 1995. DerFisch in der skythischenKunst, Frankfurt.

TERRACOTTA

SCULPTURE

Milchhoefer,A. 1879. "Sphinx,"AM 4, pp. 45-78. Minns, E. H. 1913. Scythiansand Greeks,Cambridge. Mollard-Besques,S. 1954. Mus6e nationaldu Louvre.Catalogue raisonnedesfigurineset reliefsen terre-cuitegrecs,etrusques, et romains I, Paris. Moustaka,A. 1984. "Fruihklassische Lowenplastikaus Olympia,"AM99, pp. 177-183. MunsellSoil ColorCharts,Baltimore 1975. Nicholls, R. 1970. "Architectural TerracottaSculpturefrom the Athenian Agora,"Hesperia 39, pp. 115-138. OlForschXXII = A. Moustaka, Olympische XXII. Forschungen aus Tonin Olympia, Grossplastik Berlin 1993. OlForsch XXIV = J. Heiden,Olympische Forschungen XXIV. Die Tonddacher von Olympia,Berlin1995. OlympiaIII = G. Treu,Olympia.Die Ergebnissedervon demdeutschen Reichveranstalteten AusgrabungIII. Die Bildwerkevon Olympiain Stein undThon,Berlin 1897. Payne,H., et al. 1940. Perachora,the Sanctuariesof HeraAkraiaand LimenaiaI, Oxford. Pernice,E. 1904. "Erwerbungender Antikensammlungenin Deutschland,"AA4 19, pp. 17-46. Petrakos,B. C. 1987. "To Ns,u.60ov tob Potpvobvto;,"in Otitor' Iei FEcWpyiovE. MoAcov&vII, Athens, pp. 295-326.

Pfaff, C. Forthcoming."Archaic CorinthianArchitecture,ca. 600 to 480 B.C.," in Corinth XX. Poulsen,V. H. 1937. "Der strengen Stil,"ActaArch 8, pp. 1-148. Protonotariou-Deilaki,E. 1973. "'H AAA 6, KopLvOoo," ZqpLy un5; pp. 181-188. Raepsaet,G. 1982. "Reliefsfuneraires attiquesde l'archaYsme tardif,"in Rayonnementgrec.-Hommages a CharlesDelvoye, L. Hadermann

Misguich and G. Raepsaet,eds., Brussels,pp. 137-142. Rhomaios, K. 1951. Kipopqot -u; KocAot3)vog,Athens. Richter,G. M. A. 1961. TheArchaic

45I

Gravestones ofAttica,New York. .1968. Korai,London. Robinson, H. S. 1976a. "Excavationsat Corinth:Temple Hill, 1968-1972," Hesperia45, pp. 203-239. . 1976b. "TempleHill, Corinth," in NeueForschungen in griechischen U. Jantzen,ed., Heiligtuimern, Ttbingen, pp. 239-260. Roebuck,M. C. 1990. "Archaic ArchitecturalTerracottasfrom Corinth,"Hesperia59, pp. 47-63. . 1991. "Architectural

Terracottasfrom Classicaland Hellenistic Corinth,"Hesperia Supplement27, pp. 39-52. Ryder,M. L. 1969. Animal Bonesin Archaeology, Oxford. Sanders,G. D. R. 1999. "ALate Roman Bath at Corinth:Excavations in the PanayiaField, 19951996,"Hesperia68, pp. 441-480. Shear,T. L. 1925. "Excavationsat Corinth in 1925,"4AJA 29, pp. 381397. . 1926. "Excavationsin the Theatre District of Corinth in 1926,"AJA30, pp. 444-463. Shepard,K. 1940. TheFish- Tailed Monsterin GreekandEtruscanArt, New York. Stroud,R. S. 1968. "The Sanctuaryof Demeter and Kore on Acrocorinth, PreliminaryReport II: 1964-1965," Hesperia37, pp. 299-330. Stucchi,W. 1952-1954. "Nota introduttivasulle correzioniottiche nell'artegrecafino a Mirone," ASAtene14-16, pp. 23-73. Tiverios, M. A. 1985-1986. "ArchaischeKeramikaus Sindos," MocxEdo&vxoi 25, pp. 70-85. Trianti,I. 1975. "Hocpxayxjgp6yy0c EOv.M. 77,"AAA 8, pp. 227-241. Tsigarida,E.-B. 1997. "Avocgxccqpxj ipwovoc Grrqv76PLO)XY15;

xcWxocto0;

1990ZovNs-ODpxvCo67roX; 1996,"in To ocpycouoXoyxo' spyo xoxt(ppcx- 10A, ar-jvMocxscovwoc 1996 [1997], Thessaloniki, pp. 333-346. Van Buren, E. 1926. GreekFictile Revetmentsin theArchaicPeriod, London. Vokotopoulou,J., and E.-B. Tsigarida. 1993. "AvocgxocqpLxj 6peovtca-oc Nioc P6o&cXocXx&iax';," in

452

NANCY

To poxpAcxio2oytxo uz' ?pyo Macxe6o&v'xxxt Opc,xv 4, 1990

[1993],Thessaloniki, pp. 455-468. . 1995. "Avocxaocxx

6peovoc

txocNxocPo6c XocXxt&x';,"in To CpXctoAoytx0 EPYO67Z7 Macxxveovcxxaxt Opacxv76, 1992

-

[1995], Thessaloniki,pp. 467-474. 1997. "Avmaxotp9txn ipwovoc uxoc NxocP6o&cXocXx&iax';," in To cpxpcsoAoytx0 ,oyO 617 Macxsdowv'Cxaxt pocxv77, 1993

[1997], Thessaloniki,pp. 445-454. Weinberg, S. S. 1939. "On the Date of the Temple of Apollo at Corinth," Hesperia8, pp. 191-199. . 1957. "TerracottaSculptureat Corinth,"Hesperia26, pp. 289-319. Whitbread, I. K. 1995. GreekTransport Amphorae.APetrologicaland Archaeological Study(BSA, Fitch

BOOKIDIS

LaboratoryOccasionalPaper4), Athens. Williams, C. K., II. 1980. "Demaratus and Early CorinthianRoofs,"in Zz,(i,7. T6poio ei Mv'ruv

NixoAc'ooKovro?ovrov,; Athens. . 1984. "DoricArchitectureand Early Capitalsin Corinth,"AM 99, pp. 67-75. Williams, C. K., II, and P. Russell. 1981. "Corinth:Excavationsof 1980,"Hesperia 50, pp. 1-44. Williams, C. K., II, and 0. H. Zervos. 1984. "Corinth,1983:The Route to Sikyon,"Hesperia 53, pp. 83-122. Winter, N. A. 1993. GreekArchitectural thePrehistoricto the Terracottasfrom End of theArchaicPeriod,Oxford. Wright,J. C. 1977. "APoros Sphinx from Corinth,"Hesperia 46, pp.245-254.

NancyBookidis CORINTH

EXCAVATIONS

AMERICAN

SCHOOL

54 ODOS

SOUIDIAS

ATHENS GR io6 GREECE

76

OF CLASSICAL

STUDIES

AT ATHENS

HESPERIA

69,2000

Pages453-486

FALAI

[FF

FROi

CO

13[LL-

RAT[RS

ANCIENT

RI

NTH

ABSTRACT A rareshape in Greek pottery of the 4th century B.C. is the bell-kraterof socalled Falaieff type. Red-figure and black-glaze examples made in Athens have long been known, but this article presents fragments from the American excavations in ancient Corinth that represent a contemporary local coarsewareversion. Catalogues of the Athenian and Corinthian pieces are followed by a discussion of the chronology,based particularlyupon deposits at Corinth, and by an examinationof the relationshipbetween the two series. Antecedents of the Falaieff kraterin Etruscan Bucchero and possible functions for the shape are suggested in the final section. One of the rarershapes produced by Greek potters in the 4th century B.C. was the bell-krater of so-called Falaieff type.1This unusual form of krater was named by J. D. Beazley after the Russian family that once owned the two Attic red-figure examples that are today in the Louvre.2Characteristic of the shape are the deep bowl, the tall lip, and especially the inner collar pierced with holes. When Stella Drougou discussed the Attic fineware kratersof Falaieff shape in her important article of 1979,3 she was able to cite seven examples. In the present article I add some new fragments of Attic red-figureFalaieffs,particularlyfrom Corinth; show that a coarseware version of the shape was produced in Corinth during the 4th century; and discuss the chronology, origins, and function of the shape. I begin with the Attic fineware kratersof this type. To the seven examples known to Drougou, I add a fragment from ancient Istros, perhaps another from the Athenian Agora, and a further four or five examples that have been found in the American excavations at Corinth.4 1. I would like to expressmy gratitude once again to Charles K. Williams II, formerDirector of the Corinth Excavations,who kindly gave me permissionto publish the fragmentary vases presentedin this article.I would also like to thank most warmlyNancy Bookidis, Assistant Director,for her kind and patient help in Ancient Corinth. T. Leslie ShearJr.and Mary Moore readilygave me permissionsome years ago to illustratethe unpublishedFalaieff kraterfrom the Athenian Agora;andJan Jordanexpeditedmy study of the vase.

I am also most gratefulto Martine Denoyelle in the Louvre,and to Ulf Johnsson and Marten Snickarein the National Museum, Stockholm, for supplyingphotographsof, and information about,vases in their care;and to John McK. Camp and Elizabeth R. Gebhardfor allowing me access to materialin the Athenian Agora and at Isthmia, respectively.Elizabeth G. Pembertonand the late A. D. Trendall read earlydraftsof this article,and I have benefited from their expertise, particularlyfrom Dr. Pemberton's

extensiveknowledge of the Classical pottery of Corinth.The photographsof vases in Corinth are the contributionof L. Bartzioti and I. Ioannidou.Claude Perucichand Daniela Lentini of Campus Graphicsat La Trobe University preparedthe profilesfrom my drawings. 2. See ARV2 1469-1470; Talcott and Philippaki 1956, p. 66. 3. Drougou 1979, pp. 268, 270. 4. Assuming that A14 comes from a Falaieffkrater,the fragmentsfrom Corinth must representfour or five vases.

454

IAN

MCPHEE

ATTIC FALAIEFF KRATERS5 Al St. Petersburg,HermitageMus., Ch.1903.55 (inv.14105),from Chersonesos. Fig. 1 H. 0.445;Diam. rim 0.357;H. lip 0.09; Diam. centralopening0.11 (Belov),0.13 m (Drougou). ARV2 1469, note 1; Belov 1945, pls. I-II; Drougou1979, p.268,no. 1,pp.271-273. Side A: Dionysosseated,with two maenadsand a satyr.Side B: two youthspursuingtwo women (Dioskouroiand the daughtersof Leukippos?).Lip:youths,women,and Eros. A2 Paris,Louvre,CA 229 (G 529), ex FalaieffCoil. Fig. 1 H. 0.435;Diam. rim 0.37; Diam. centralopening0.185 m.

ARV2 1470, no. 162 (Group G); BeazleyAddenda2, p. 380; Drougou

1979, pp.268,274,276, no.2; LIMCVIII.2,1997, pls. 344-345, Arimaspoi 40. Side A: Dionysos seated, with maenad and two satyrs. Side B: grypomachy. Lip: grypomachy. A3 Paris, Louvre, CA 228 (G 530), ex Falaieff Coil. Fig. 1 H. 0.43; Diam. rim 0.38; Diam. central opening 0.175 m. ARV2 1469, no. 161 (Group G); BeazleyAddenda2, p. 380; Drougou 1979, pp. 270,275, no. 3; Boardman 1989, fig. 414; LIMC VIII.2, 1997, pls. 343-344, Arimaspoi 39.

Side A: Dionysosseated,with maenad,satyr,and Eros.Side B: grypomachy.Lip:grypomachy. A4 Stockholm,Nat. Mus., G III A V:2 (D 19310),ex Hamilton Coll. Fig. 1 H. 0.43;Diam. rim0.43; Diam. centralopening0.18-0.19 m. ARV2 1470, no. 163 (Group G);

p. 494; Beazley Paralipomena, Addenda2, p. 380; Drougou1979, pp.270,277,279, no.4. Side A: Dionysosreclining,with Ariadne(or a maenad)seatedand playinga harp;a satyr;and seated Eros.Side B: grypomachy.Lip: grypomachy.

WU ...,.

Al

A2

. .!:...

..

A4 Figure1. Attic Falaieffkraters(AlA4).

-

-A3

5. In the cataloguethat follows, the term "lip"refersto the whole element that rises from the bowl of the vase; "rim"refersto the edge of the lip. Grid coordinatesof the findspots referto the plan of the centralareaof Corinth, ca. 400 B.C., given in Williams 1980, p. 112, fig. 2, and in Pemberton1997, p. 91, fig. 17. C and CP numbersare Corinth inventorynumbers.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

455

Both2:5

.

I

1:2

PN-P244a

PN-P244b

_ A_

A6

1:2

Figure 2. Attic Falaieffkraters: from Athens (A5-A6). Scale as indicated

A5 Athens,AgoraMus., PN-P 244a-b, fromAthens (Pnyx). Fig. 2 Two fragments.Diam. rim 0.44; H. lip at least0.105 m. TalcottandPhilippaki1956, pp. 65-66, pl. 9, no. 318; Drougou 1979,p.270,no.5. Lip:battleof pygmiesand cranes.

A6 Athens, Agora Mus., P 23747ab, from Athens (Agora). Fig. 2 Two fragments. Diam. rim 0.38; H. lip 0.14; W. collar 0.06; Diam. central opening 0.147 m. Talcott and Philippaki 1956, p. 66; Drougou 1979, p. 270, no. 6. Black except for a reserved handle-zone decorated with a black ivy-vine.

456

IAN

MCPHEE

__~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

t .. ... . . . .................. .........

q

-

0

5

10014

Figure 3. Attic Falaieff krater, from Athens (A7).

Figure 4. Attic Falaieff krater, from Istros (A).

A7 Athens,Third Ephoreia,from Athens. Fig. 3

A8 Bucharest,Instituteof Archaeology,V 8732, fromIstros. Fig. 4

Fragmentary. Diam. rim0.395; Diam. centralopening0.18 m. Drougou1979, pp.266-267, 269,270, no. 7. Blackexceptfor a reserved handle-zonedecoratedwith a black

Singlefragment.Est. Diam. rim 0.40 m. Alexandrescu1978, p. 78, no. 460, pl. 54. Lip:seatedwoman(maenad?). Recognizedby Alexandrescuas part of a Falaieffkrater.

ivy-vine.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

457

A9, fragmentc A9 A9, fragmenta

AIO Figure5. Attic Falaieffkraters,from Corinth(A9-A10). Scale 1:2

A9 C-1980-21a-c, fromCorinth. Fig. 5 Threefragments.Fragmenta (fourjoining sherds):profileof lip almostcompletedownto junction with bowl,wherethereis preserved half of one of the holes pierced throughthe innercollar;outeredge of rim moldedwith two blackridges and two reservedgrooves;flat top of rim andinnersurfaceof lip glazed black;undersideof collarreserved. PH. 0.10;Th. 0.009-0.012 m. Fragmentb (singlesherd):small sectionof rim.PW. 0.036 m. Fragmentc (singlesherd):upperpart of lip. PW. 0.022 m. Findspot: ForumNortheast,grid38:A-B, fill overstarting-lineof earlierracecourse. Dionysiacscene.Fragmenta preservesthe top of a smalltreeat

the lefthandbreak;at right,a satyr (left leg, rightbuttock,tail,left hand, crownof head)creepingto right,his rightleg forwardandraised,his left armthrownback.Fragmentc preserves,in profileto right,the face of a secondsatyr(bearded)and the uppercontourof his right(?)arm outstretchedandbent up at the elbow.Tracesof preliminarysketch on both figures.Reliefcontourfor the branchesof the tree,the thighs and the handof the one satyr,the armof the other.White scumbled with diluteglaze for the firstsatyr's wreath.See alsothe commentsunder A10, All, andA12.

A10 CP-3293, fromCorinth. Fig. 5 Singlefragmentfromjunctionof lip, shoulder,and innercollar,preservingabouthalf of a hole pierced throughthe collar.Narrowridge, with reservedgroovebelow,at base of lip on outside.Innersurfaceof lip anduppersurfaceof collarglazed brownish-black. Undersurfaceof collarand shoulderreserved.Max. dim. 0.061;Th. lip 0.011-0.012; Th. collar(innerbreak)0.006 m. Exactfindspotuncertain:perhaps one of the wells in the JulianBasilica excavatedin 1915. On the lip, the raisedleft foot, partof the rightfoot, and the hem of the garmentof a figuremovingto left. On the shoulder,a patternof verticaltongues(not, I think,eggs as on A4). Possiblyfromthe samevase as A9.

458

IAN

I

I

% %

I I

.

MCPHEE

All

I I

Figure6. Attic Falaieffkraters,from I

Corinth(All-A12). Scale1:1

A12 Fig. 6

A12 CP-1692, from Corinth. Fig. 6

Single fragment from junction of lip with bowl and inner collar, preserving trace of pierced hole on inside and black ridge between two reserved grooves above lower break on outside. Black on inside of lip; streaks of black on inside of bowl. P.H. 0.031; p.W. 0.046; Th. (upper break) 0.007 m. Findspot: Southeast Building, mixed fill. At the right, the left foot of a figure standing to left. At the lefthand break, two toes of the right foot of a second figure standing to right, and, above, the hem of a garment worn by this figure. Preliminary sketch on the foot. Relief contour for the foot, toe, and drapery. Though both C-46-9 (All) and C-1980-21 (A9) were found along the eastern side of the Forum of Corinth, the different thickness of the lips shows that they cannot belong to the same vase.

Single fragment from lower part of lip at junction with bowl; ridge on outside at lower break;broken on inside where inner collar joined base of lip. Brownish-black glaze on inside of lip. P.H. 0.056; p.W. 0.051; Th. lip 0.011-0.016 m. Exact findspot unknown. All that remains of the picture is the right leg (foot to knee) of a male(?) figure, standing to right; and at the upper lefthand break, relief lines defining an object (treebranch?). Preliminary sketch-lines on the leg. The thickness at the base of the lip, and the different relationship of collar to bowl, indicate that this fragment does not come from the same vase as A9.

All C-46-9, from Corinth.

FALAIEFF

BELL-KRATERS

FROM

Figure7. PossibleAttic Falaieff hraters (A13-A14). Scale 1:1

ANCIENT

CORINTH

459

A14

Possiblyalso: A13 Athens,AgoraMus., P 32276, fromAthens (Agora). Fig. 7 Unpublished.Lip with redfiguredecoration:chariot-horses gallopingto left.6 A14 C-28-75, fromCorinth. Fig. 7

6. John McK. Camp, Director of the Agora Excavations,has kindly allowed me to illustratethis recent find.

Singlefragmentpreservingpart of rim andlip. P.H. 0.065;p.W. 0.148;est. Diam. rim0.42-0.44; Th. lip 0.008-0.012 m. Upperbodyof a maenad,facing to right,beatinga tympanon.To left and right,the headsof two thyrsoi. McPhee 1976, p. 384, no. 7, pl. 85 (with profile).In the original publicationthis fragmentwas saidto come froma bell-kraterof special

type,which is correctinsofaras it goes, but the moldedrimhas the sameformas that of A9, which suggeststhat the piece maycome froma Falaieffkrater,and the estimateddiameterwouldalso be appropriate. However,the fabric andglaze indicatethat the fragment musthavecome froma different vasethanA9-A12. The date,"ca. 410-400,"given in the original publicationis a little early:the fragmentshouldbe placedabout 400-380 B.C.

460

IAN

MCPHEE

All the Attic kratersof Falaieff type seem to have been large vases, the height of the better preserved examples (Al-A4) ranging from 43 to 44.5 cm, the diameter of the rim in the case of Al-A8 and A14 ranging from 35.7 to 44 cm. They are characterizedby a solid ring-foot in one or two degrees; a deep bowl; two handles placed high on the bowl; and a tall, flaring lip with a rim that is either rounded and articulatedby one or more grooves on the outside, or that is flat on top and with a rounded or ridged profile. A narrow,horizontal ridge or fillet marks the junction of lip and bowl; and on the inside, where the lip and bowl join, there is a broad collar, which at the junction with the lip is pierced by four to eight holes (four in the case of Al, five of A2-A4 and A6, eight of A7).7 Some distinctions in shape, however, may be established within the series. The two kraters in the Louvre (A2, A3) and that in Stockholm (A4) all display a foot of inverted echinus shape, a bowl that is low and broad, handles set horizontally below the join of bowl and lip, and a lip that is particularlytall and flaring.Whether the three vases are the work of a single potter, as Frel maintains,8I cannot confirm, but they seem to have been painted by a single hand, and are probablyproducts of a single workshop. In contrast, the krater in St. Petersburg (Al) clearly represents a different form: the foot consists of two elements, the bowl is taller and more ovoid, the handles are set diagonally, and the lip is shorter.Among the remaining pieces, the fragmentary black krater A7 may perhaps be associatedwith the Louvre and Stockholm Falaieffs,9while the Agora fragment A6, to judge from the profile of the lip and the collar, is perhaps contemporary,though not necessarilyby the same potter; and A13 has the same type of rim as A4. Not enough remains of the red-figure sherds A5, A8, A9-A12, and A14 to associate them on the basis of shape with the more complete vases, although A9 and A14 share a similar type of rim. In all cases these kraters are fineware. Two (A6 and A7), both fragmentary,seem to have been painted black except on the bowl between the handles where a reservedband is ornamented with a horizontal vine of ivy leaves and berries; and the style of the ivy is similar in both, so far as one can tell.10The remaining kraters all employ the red-figure technique, on bowl and lip where both elements of the vase are preserved.Three, A2A4, may be attributed to a single hand, the Griffin Painter, the principal artist within Beazley's Group G.1"St. Petersburg Ch.1903.55 (Al) and the Pnyx fragment (A5) were painted by two different hands, unrelated to Group G.12The style of A8 is unconnected with any of these pieces. Alexandrescu places this fragment near the Black Thyrsus Painter, but though the style does seem to relategenerally to the Telos Group (ofwhich the Black Thyrsus Painter is a member), I do not myself believe that it is possible to make a closer attribution from what remains.13None of the fragments from Corinth can be definitely assigned to a particular artist, though the style of A14 seems to me to recall most closely the Erbach Painter.'4

The absolute chronology of these Attic vases depends less upon archaeological context than upon stylistic analysis of the red-figure decoration. The contextual evidence is slight. The precise findspots of the kraters in Paris and Stockholm (A2-A4) are unknown, and the contexts of A7

7. In the case of A6 only four holes arepreserved,but their even spacing allows for the restorationof a fifth; in the case of A4, there aretracesof four holes, but againprobablyfive originally. 8. Frel 1964, p. 124. 9. The bowl of A7 may be too tall in the restoredprofile publishedin Drougou 1979, p. 267, fig. 2 (here Fig. 3). 10. A similardecorativescheme was employed for the contemporaryblackglazed deep bowls, such as Athens, Agora Mus. P 26062 and P 8606, AgoraXII, pp. 56-57, nos. 78 and 79, p. 242, pl. 4. 11.iARV21469-1470, nos. 161-163. Talcott and Philippaki(1956, pp. 6566), as well as Drougou (1979, p. 275), remarkon the unity of the style of these three.The most recent discussionsof Group G will be found in Margos 1980 (but I do not alwaysaccepther stylistic divisions),Isler-Kerenyi1982, and Robertson 1992, p. 274. 12. So alsoTalcott and Philippaki 1956, pp. 65-66; and see Beazley, ARV2 1469. 13. Alexandrescu1978, p. 78. Telos Group:ARV2 pp. 1425-1434. Black Thyrsus Painter:ARV2 1431-1434. 14. ErbachPainter:ARV2 pp. 14181419.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

46I

and A8 are also unhelpful. The kraterin St. Petersburg (Al) was found in 1903 in a tomb at Chersonesos, but the tomb seems to have contained multiple cremations and the contents have not yet, to my knowledge, been fully published, so that it is not possible to establish a close dating. Although the fill which yielded fragment A5 and many other red-figure sherds, brought in during a rebuilding (Period III) of the Pnyx, included Hellenistic and Roman material,the bulk of the pottery favorsa date around 340-335 B.C. for the rebuilding.We cannot, however, argue that this provides a terminus ante quem of ca. 330 for A5, since the chronology of the deposit depends in part upon the fragments of red-figure-a clear example of circularreasoning.1" None of the Attic red-figurefragmentsfrom Corinth (A9-Al2, A14) provides a useful stratigraphiccontext.16 In an effort to obtain a more precise date for the Attic red-figure Falaieffs, we must turn to the subjective evidence of style. In her recent study, Stella Drougou dates all the examples then known (A2-A7) to ca. 370-360 B.C., except for the St. Petersburg krater (Al), which she places about 390. This dating basically follows that suggested by Karl Schefold for the three Falaieffs of Group G (A2-A4).'7 However, the chronology of red-figure pottery in the 4th century has not been determined with great precision,18 and I think that a date of 370-360 for A2-A4 may be too high by about a decade. On the other hand, the St. Petersburgkrater(Al) does seem to belong in the first quarterof the 4th century.19And the fragments A9-A12 and A14, from Corinth, may be assigned to the same period.20If this subjective stylistic dating were to be correct,it would suggest that the kraterof Falaieff shape was first produced in the Athenian potters' quarter at some time during the first quarter of the 4th century, earlier perhaps rather than later, and that its popularitywas relativelybrief, for there is no clear evidence at present that it was still being manufactured during the third quarterof the century,and indeed none of the Attic examples may be later than 350. These Attic fineware kraters,to judge from the findspots, were apparently potted not only for the home market (A5-A7, A13), but also, and perhaps mainly, for export: to Corinth (A9-Al2, A14), to Chersonesos and Istros on the Black Sea (Al, A8), and, probably,to Campania (A215. See Talcott and Philippaki, pp. 5-6, for the chronologysuggestedby the red-figurepottery.A date for Pnyx Period III in the third quarterof the 4th centurywas originallyproposedin Thompson and Scranton1943, pp. 293301. In AgoraXIV, pp. 49-50, it is suggestedthat the reorganizationof the Pnyx may have been begun under Lykourgosbut have been "brokenoff'"in the late 4th century.The pottery has now been restudiedby Susan Rotroff, who supportsa date of ca. 340-335: Rotroff 1996; Rotroff and Camp 1996, pp. 275-278; AgoraXXIX, pp. 20-23. 16. A9, to be sure,came from the fill

(lot 1980-57) dumped over the starting-lineof the first racetrack,but that fill contains materialdating as late as the first half of the 3rd centuryB.C.: see Williams 1981, p. 11 with note 12. 17. Drougou 1979, pp. 268, 270; Schefold (1934, pp. 17 and 78) places the three Falaieffsof Group G about 370-360. 18. The problemsin the chronology of Attic red-figurein the 4th century have been remarkedupon by Robertson (1981, p. 67), and have been highlighted most recentlyby Panos Valavanis' study (1991) of nine Panathenaic amphoraefrom Eretriaand their

stylistic relationswith contemporary red-figure. 19. One of the anonymousreaders has suggested that there is a stylistic connection between the St. Petersburg vase and an epinetronin Rhodes (ARV2 1503, no. 6; Chalki Group), which comes from a gravedatableby the associatedpottery.I comparedthe two vases when I first began to work on this article,but could not see any close relationshipin style.This is still my opinion. 20. A8, however,probablybelongs, as Alexandrescusuggests,to the second quarterof the 4th century.

462

IAN

MCPHEE

A4).21The St. Petersburg vase (Al) was unearthed, as we have noted, at Chersonesos in a tomb, where it was employed as an urn for the cremated bones of the dead, the interior opening covered with a single black-glazed bowl.22Doubtless, A2-A4 also came from tombs. But other pieces were found in settlement contexts, so a funerary use was clearly a secondary function for these vases. Before, however,we proceed to a considerationof the origins and function of the Falaieff krater,we must first examine some evidence recovered by the American excavations in Corinth indicating that the local potters were producing a coarsewareversion of the shape. At present I am aware of sixteen fragments that come from such Corinthian Falaieff kraters,23 representing, at a minimum, thirteen different vases.24

CORINTHIAN Cl C-71-524.

FALAIEFF KRATERS Fig. 8

P.H. 0.204; Diam. rim 0.305 (inner); 0.34 (outer); W. inner collar 0.085; Diam. interior opening 0.115; Th. (lower break) 0.005-0.007 m. Hard, coarse fabric;light red (5YR 6/8), but in places fired light gray (7.5YR 6/2) in core; tempered with many small to large grits (0.5-6 mm, averaging 0.5-2 mm), mainly red, gray, and yellow, but also some white; voids. All surfaces smoothed; outer surface of lip and bowl fired a creamish color (7.5YR 8/4-6). Findspot: Forum Southwest, grid 55:M-L, northeast of pottery deposit near walls 7 and 8 (drain 1971-1). Two sections of the lip, almost half of the interior collar, and small

21. The Stockholm krater,A4, was once part of the first collection formed by Sir William Hamilton, and so was probablyfound in Campania. The two vases today in the Louvre,A2 and A3, were said by their former Russianownersto have been given to their family at the beginning of the 19th centuryby the King of Naples (Pottier 1922, p. 287). Pottier doubted the Campanianproveniencebecause the style and subjectmatterof the picturesremindedhim of vases from S. Russia.Schefold (1934, p. 17) and Drougou (1979, p. 268, note 12) have

partsof the bodyremain,together with one handleandtwo knobs. Lip,projectingelementof molded rim,bowl,handle,knobs,andinterior collarall formedseparately. Wall curvesin sharplybelowthe handle. Pointof supposedseparationof lip frombowl articulatedby a narrow horizontalstripof clay.Preserved handlehas circulardiskswhereit joins the wall.Both handlesmight havebeen flankedby smallknobs, althoughonly two remain(fromthe sameside of the vase).Lip flaresout slightly,with a jog in outerprofile, and ends in a modeledrim.Horizontal collarattachedto insideof vessel at a pointbetweenjog and claystripon outersurface,about 2 cm abovethe latter.Collaris piercedwith threerowsof holes,

with an incisedguidelinefor each row,andhas a raisedrim at its inner edge.Thirteenholespreservedof firstrow (thatnearestthe lip of the vessel);nine,of second;and six,of third.Originally,the firstrowmay havehad as manyas twenty-nineor thirtyholes;the secondrow,about twenty;the third,some fourteenor fifteen.Eachhole is about4 mm in diameter.

followed Pottier in postulatinga Russianproveniencefor the Louvre vases, and Drougou (loc. cit.) suggests a Russianproveniencealso for the Stockholm krater.There are,however, no substantialgroundsfor doubting a Campanianorigin for all three vases. 22. Belov 1945, p. 142. 23. A seventeenthfragmenthas come to light in the excavationsat Isthmia:IP 8479. It has the same coarsefabricas the pieces from Corinth, and had at least two rows of holes. The piece was first identified by Elizabeth Pemberton.

24. Lot numbersin the following catalogueare numbersgiven to fragmentsnot inventoriedbut kept in the pottery lots stored in Ancient Corinth. I ought to say that, although I have looked through all the inventoried pottery and many storedlots containingpottery of the second half of the 5th and the 4th century, especiallythose excavatedsince the mid-1960s, I have not examinedevery preservedlot, and there may be fragmentsof other CorinthianFalaieffs awaitingdiscovery.

FALAIEFF

......

.. . .. ..

....

....

BELL-KRATERS

FROM

ANCIENT

463

CORINTH

.

...

.

......

.

.

.

.

...

..

. ..

..

. .

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.

...

1.2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~z 1

-;--

iR.

R

..

-.

-

.

4

..

_

.

'.'

':S

.

.'S

.

..

' _ , '

.

.. ...

.

...

...

...

..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Ti

.

..

.....

:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

.................................................. ........

,.

,

|

;

..

tE

r

i

s

iW,

*S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ g

~

~ ~

~

~

~

~ ~

1

IAN

464

MCPHEE

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ,,,'.,,.,,,,3git.C |.,l.,"j=3. .S . ...

..

. ..

l~~~~~~~~~~~~~~~~~~~~~~~~~~~..... ...''.,..,,,'.,,,

C-71-641~

tl,

_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~U

~ ~ ~ ~ ~~~~~~~~~ ....

.. ..

...

~*

......

^'

.

_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... :.~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....... > :-... ..... ..:. '..:' ,. ''

...:,.

...

.'3S'

.

. | wN

.:._

,,

... .... ...... .:........................................................ . ,:.....

. _ .X ...................................................... ..

_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........... S "'|q ........................................... ..

EM ,':S"~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ..

M

Coitha

:~~~~~~~~~~~~~~~~~~~C) ...,,....4 . r,s

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~ecp

C2 C-71-523.

Fig. 9

PH. 0.244; Diam. rim0.345 (inner);0.375 (outer);Th. (lower break)0.003-0.007 m. Hard,coarse fabric;grayto light gray(2.5YR5/0, 7.5YR 6/0) in core;temperedwith smallto mediumgrits (< 2 mm) and some largegrits (< 1 cm), mainlygray andyellow,but some red andwhite; voids.Innerand outersurfaces smoothedand firedorange(2.5YR6/ 8). Findspot:ForumSouthwest,grid 54:K,potterydepositeastof Building II (drain1971-1). About a thirdof the lip, and a sectionof the bodyremain.Bodyand interiorcollarapparentlyformed together,lip madeseparatelyand joined,withjunctionconcealedby a stripof clayhorizontallygrooved. Lowerwall of bowl curvesin strongly andis remarkably thin at lowerbreak.

No traceof handles.Lip flaresout fromwall and ends in an overhanging rim modeledin two elements.Only a narrowsectionof interiorcollar remains,but enoughto showthat it was piercedby at leasttwo rowsof holes:tracesremainof sevenfrom outerrow (i.e., that atjunctionof collarandlip), threefrominner. Preservedholes of outerrowaverage about5 mm in diameter.It is possible to estimatethat originallytheremust havebeen abouttwenty-threeor twenty-fourholes in this row. A fragment(mendedfromtwo sherds)fromthe samefindspot, C-71-641 (Fig. 9), which preserves abouta thirdof a ring-foot(est. Diam. 0.10 m), has the samefabricas C-71-523 and maycome fromthe base of this vase (or one like it).25

rm

rigfo a efi noted

Faaif figmn

dptasC

krater C-71-641 .Scl

:

25. This fragmentis unlikelyto be the foot of a coarsewarelekane, for the feet of these lekanaiare somewhat differentin form, and in any case the fabricis alwaysfiner.I cannot, however, rule out the possibilitythat it could be the ring-foot of a "blisterware" amphoralike C-1971-185 (Williams 1972, p. 156, no. 19, pl. 24).

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

465

CORINTH

Both1:4

1.2

C3

1:2

C4 2:5

Figure10. CorinthianFalaieff kraters(C3-C4). Scale as indicated C3 C-69-352.

Fig. 10

P.W. 0.185; Th. (lower break)

0.005;Th. lip 0.010-0.015;Th. collar (inner break) 0.008 m. Hard, coarse fabric;light red (2.5YR 6/8) in core, shading to 5YR 7/6 and 7.5YR 7/6 near surfaces;tempered with small to large grits (usually < 3 mm, some larger), mainly red, but some white and gray, also some sparkling inclusions; large voids. Smoothed surfaces orange to light brown, 5YR 6/6-8, where not rubbed; lime incrustation on surfaces of collar. Findspot: Sacred Spring Central, trenches 1-4, dumped fills above floor 2 and drain. Single fragment preserving complete profile of lip, part of inner collar, and beginning of bowl. Tall,

flaringlip with modeledrim.Collar piercedwith at leastthreerowsof unevenholes (5 of outerrow preserved,perhaps4 of secondrow, and2 of inner).Holes piercedfrom top, eachabout6-8 mm in diameter. Collarandbowl formedin one piece, lip formedseparatelyand attached. C4 Lot 72-92-2a, b.

Fig. 10

Fragmenta: p.W. 0.205;Th. (lowerbreak)0.007-0.008; est. Diam. (baseof band)0.43; L. of handle 0.15, W. 0.032 m. Fragmentb:p.W. 0.053;Th. 0.008-0.010 m. Hard, coarsefabric;gray(2.5YR5/0) in core,shadingto light red (1OR6/8) nearsurfaces;creamishpalered (5YR 7/3-4) on outersurface;interiorand probablyexteriorsurfaceperhaps

givena streakygrayishwash,and smoothed;smallto medium(< 3 mm) grits,mainlygray,red,and white,with someyellow.Findspot: ForumSouthwest,grid61:D, southwestcorner,pit in wall 3 in area of drain. Two nonjoiningfragments. Fragmenta (foursherds)preservesa sectionof bowlwith most of one handle(centralridge)and,at the upperbreak,partof a horizontal band(convexprofile).Fragmentb (singlesherd-not illustrated)comes fromupperwall of bowl,with partof samehorizontalbandtowardupper break.C4 probablybelongsto the samevaseas C5, and perhapsto the samevaseas C6.

466

IAN

MCPHEE

I

1:1

I

C5

1:2

C6 Figure11. CorinthianFalaieff kraters(C5-C6).Scaleas indicated

C5 Lot 72-87-1.

Fig. 11

PH. 0.064;p.W. 0.075;Th. 0.007-0.008 m. Hard,coarsefabric; light red (1OR6/6-8) in core,but alsogray(2.5YR5/0) sandwiched betweentwo light redlayers;pale red outersurfacewith some streaksof a possiblegraywash;light redon inner surface,againwith tracesof possible graywash;smoothedsurfaces;gray, red,and a few white grits,mainly smallto medium(< 3 mm). Findspot: ForumSouthwest,grid61:D, fill for foundationtrenchof northstylobate of South Stoa. Singlefragmentfromupper bowl,preservinga horizontalband

(convexprofile)at upperbreak,and at lowerlefthandbreak,partof a knob.This fragmentprobably belongsto the samevaseas C4: both werefoundin grid 61:D, both share the distinctivewashon the interior, andboth havethe sametype of band. See also C6. C6 Lot 72-92-3.

Fig. 11

P.H. 0.071;p.W. 0.178;Diam. rim 0.40 (inner),0.43 (outer);Th. rim 0.015 m. Hard,coarsefabric; gray(2.5YR5/0) in corebetweentwo light red (1OR6/8) layers;surfaces creamishpalered(5YR 7/3-4), smoothed;tracesof graywash

outside;red,gray,and somewhite grits,mainlysmallto medium, averaging0.5-3 mm, a few larger. Findspot:ForumSouthwest,grid 61:D, southwestcorner,pit in wall 3 in areaof drain. Singlefragment(mendedfrom two sherds)of lip, brokenatjunction with innercollar.The rim is flat on top, andoffset on the outersurface by a horizontalgroove.Fabricand surfacetreatment,as well as findspot, suggestthat C6 maybe the lip of the samekrateras C4 and C5.

FALAIEFF

BELL-KRATERS

C7, fragment a

FROM

ANCIENT

CORINTH

467

C7, fragment c

C8, fragment b C8, fragment a

Figure 12. Corinthian Falaieff kraters (C7-C8); ring-foot fragment, lot 7079-277, from same findspot as C8. Scale 1:2 C7 Lot 7079-275a-c.

Fig. 12

Fragment a: max. dim. 0.102; Th. wall 0.010-0.012; Th. flange 0.009-0.010 m. Fragment b. max. dim. 0.064; Th. 0.009-0.011 m. Fragment c:Th. 0.026 m. Hard, coarse fabric, light gray core sandwiched between reddish-brown (5YR 6/4) layers;many small to large grits (< 4 mm, some larger), mainly red but some gray and white, and voids; surfaces smoothed and fired a very pale brown (1OYR 7/4-6). Findspot: Forum Southwest, grid 55:M, drain between Buildings I and II (drain 1971-1).

r

Lot 7079-277 Fragment a comes from the upper wall of the bowl, and preserves at the top the beginning of the lip and the inner collar; b is a wall fragment; c is a handle, round in section and slightly upturned. C8 Lot 7079-276a, b.

Fig. 12

Fragment a: max. dim. 0.10; Th. wall 0.006-0.007; Th. handle 0.033 m. Fragment b. p.H. 0.068; p.W. 0.075; Th. 0.006-0.007 m. Hard, coarse fabric, light red (2.5YR 6/6-8) in core, more orange (5YR 7/6-8) on inner surface;outer surface cream (7.5YR 8/4); small to large grits (< 3

mm), mainlyred andgray,some white;voids.Findspot:Forum Southwest,grid54:K,drainbetween BuildingsI and II (drain1971-1). Fragmenta preservesa section of the wall,with the stumpof a handle,roundin sectionand upturned;b (twojoiningsherds)is also a fragmentof wall,with three horizontalridgesat the top andhalf of a knobat the bottom.Possiblypart of the baseof the samepot aretwo joining sherds,lot 7079-277 (Fig. 12), that preserveabout40% of a ring-foot(Diam.0.13 m) and a sectionof the lowerwall.

468

IAN

MCPHEE

1:2

C9

,1'o

13. CorinthianFalaieff

C10;Figure

kraters(C9-C1O). Scaleasindicated C9 Lot72-92-6.

Fig. 13

P.H. 0.060; p.W. 0.122; Th. collar at break 0.008; Th. lip (upper break) 0.011 m. Hard, coarse fabric; light red (1OR 6/8-2.5YR 6/8) in core; mainly red and gray grits but some yellow and white, small to medium, averaging 0.5-3 mm, with a few larger;surfaces smoothed; outer surface and underside of the collar, creamish pale red. Findspot: Forum Southwest, grid 61:D, southwest corner, pit in wall 3 in area of drain. Single fragment from junction of bowl, lip, and inner collar.Junction masked on outside of vase by three horizontal bands, convex in section. At one end of collar, about 1 cm in from junction of collar and lip, is preserved half of a pierced hole. The profile of the inner collar continues that of the bowl.

C10 Lot 72-92-4.

Fig. 13

P.H. 0.064; p.W. 0.051; Th. rim 0.015 m. Hard, coarse fabric;light red (1OR 6/6) in core, but shading to light gray sandwiched between light red layers;gray,red, and some white grits, small to medium, averaging 0.5-3 mm, with a few larger;surfaces smoothed: inner, pale red (5YR 7/34), outer and top of rim, creamish (1OYR 8/3). Findspot: Forum Southwest, grid 61:D, southwest corner, pit in wall 3 in area of drain. Single fragment of lip, broken below at the junction with inner collar. Rim flat on top, and offset on outer surface by a horizontal groove. Not, apparently,from same vase as C6: fabric, and form of rim, are both slightly different. It is possible that lot 72-92-5 (not illustrated), a fragment from the bowl of a krater, with a similar fabric, may come from the same vase.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

469

C"l

C12 Figure14. CorinthianFalaieff kraters(C11-C12). Scale1:2 Cll

of the 26. I owethe identification of the bronzeto the conservator Miss Stella CorinthExcavations, Bouzaki.Dr.NancyBookidisandI havebothexaminedthebronze,butwe An areuncertainof its explanation. likethatto A7, would ancientrepair, normallyhavebeendonein lead.There is no signthatanyof the otherholes wereplugged,noris thereanytraceof on the surfaceof bronzestainanywhere the collar.

Lot 1976-100-1.

Fig. 14

PH. 0.117; p.W. 0.11; p.W. inner collar 0.034; Th. lip (upper break) 0.007; Th. collar (inner break) 0.009 m. Hard, coarse fabric;mainly light red (1OR 5-6/8) in core, but gray (5YR 6/1) where collar, lip and bowl join; pale red (1OR 6/8) for inside of lip and bowl, and for upper surface of collar; cream to very pale brown (7.5YR 7/4) for outer surface. Traces of possible gray wash on inside of lip and upper surface of collar. Small to medium grits (< 2 mm), a few larger (< 5 mm); mainly yellow, gray,and red. Findspot: Forum Southwest, grids 73:A-ZZ, fill in Basin Room below Roman Cellar Building. Single fragment with parts of lip, collar, and bowl. Where collar joins wall, five holes from outer row remain, spaced somewhat irregularly, each about 4-5 mm in diameter.Two holes side by side seem to indicate two different attempts to pierce a hole, presumablybecause the first was felt to be too close to its neighbor. Another hole has a section of bronze preserved around half the

circumference.26Traces also of two holes of a second row. Outside of bowl articulated by a horizontal band in relief. Collar made separately,bowl and lip perhaps in one piece. C12 Lot 7084-1.

Fig. 14

P.H. 0.077; p.W. 0.092; Th. (lower break) 0.009; Th. collar (inner break) 0.008 m. Hard, coarse fabric; gray (5YR 5-6/1) in center of core, but orange to light red (1OR 6/8) toward surfaces;outer surface of bowl and lip fired creamish (7.5YR 8/4); small to medium grits (< 3 mm) and some larger (< 8 mm); mainly white and yellow but also gray and red; some voids; all surfaces smoothed. Find-spot: Forum Southeast, well 1971-1. Single fragment from junction of bowl, lip, and collar.Toward lower break on the outside, three narrow horizontal ridges. Two preserved holes, one at either end of the preserved section of the collar,where it joins the wall, about 6-7 mm in diameter and about 7 cm apart. Lip seems to have been made separately.

470

IAN

MCPHEE

C13

1:1

Figure15. CorinthianFalaieff kraters(C13-C14). Scale1:2,except asindicated

C14

C13 Lot 1979-64-1.

Fig. 15

P.H. 0.086;p.W. 0.122;Th. (lowerbreak)0.012 m. Hard,coarse fabric;light red (1OR6/8-2.5YR 6/8) in core,shadingto purplishred (2.5YR6/4), but creamishpalered (5YR 8-7/4) on the surface;white, red,andgraygrits,mainlysmallto medium(< 2 mm),with a few larger. Surfacessmoothed.Findspot:Forum Southwest,grid 63:C, fill overfoundationtrenchII, in northernpartof BuildingIV. Singlefragmentbrokenon all sides,fromjunctionof lip, bowl,and collar.At lowerbreak,on outside,a narrow,horizontalridge,probably uppermostof three.Little of inner collarremains,but whereit joined the wall thereis a traceof a pierced hole. Lip probablymadeseparately, innercollarandbowl in one piece.

C14 Lot 1978-40-1.

Fig. 15

P.H. 0.065;p.W. 0.091;Th. collar(innerbreak)0.013;Th. lip (upperbreak)0.013 m. Hard,coarse fabric;gray(1OR6/1) in core,shading to light red (1OR6/6-8) near surfaces;grayandredbut somewhite andyellowangulargrits,mainlysmall to medium(< 2 mm),with some larger.Surfacessmoothed,and fireda creamishpaleredto light orange. Verysimilarin fabricto C4 and C5. Findspot:ForumSouthwest,grid 61:B,fill of foundationtrenchfor northwall of Shop 31 of South Stoa. Singlefragment,brokenon all sides,fromjunctionof lip, bowl,and innercollar.At lowerbreak,on outside,a horizontalridgeand beginningof a secondbelow. Remainingsectionof innercollar preservestwo holes fromouterrow, spacedabout4 cm apartand 1 cm in fromjunctionof lip and collar,and one hole froma secondrow.Holes areabout7-8 mm in diameter.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

47I

C15

Figure16. CorinthianFalaieff kraters(C15-C16). Scale1:2,except

C16

as indicated

C15 C-64-12.

Fig. 16

Max.dim. 0.105;Diam. inner rim of collar0.11;Th. collar0.07 m. Moderatelyhard,coarsefabric;core light redto paleorange(5YR 7/8), but smoothedsurfacescloserto 7.5YR 7/6; numeroussmallto mediumgrits(< 2 mm), somelarger, mainlyredbut alsowhite andgray. Surfacesratherabradedbut outer mayoriginallyhavereceiveda thin creamishwash.Findspot:surfacefind on line of west citywallbetweenthe PhliasianGate andAcrocorinth,June 1964. Partof collar.Rim is roundedon inside,beveledon outeredge.Four holes remainof innerrow,spaced 1.5-2.0 cm apart.Preservedof a secondrowaretwo adjacentholes, spaced1.2 cm apart.There may originallyhavebeen a thirdrowat junctionof collarandlip of vase. Holes areabout5 mm in diameter.

C16 Lot 7079-341.

Fig. 16

Max.dim. 0.085;Th. collar 0.007-0.011;W. collarat least0.053 m. Hard,coarsefabric,light redin core(2.5YR5/8-6/6), but more orange(5YR 7/6) underneath; creamish(7.5YR8/4) on upper surfaceand in holes;frequentsmall to verylargegrits,mainlyred and gray,someyellow.Findspot:Forum Southwest,grids53:J,53-54:K, drain betweenBuildingsI andII (drain 1971-1). Partof collarof a Falaieffkrater, brokenon all sides.Therewerethree rowsof holes:only a singlehole remainsof the rowat the junctionof collarandlip;two holes (spaced5.3 cm apart)arepreservedof the second row;and a singlehole of the third row.Eachhole is about7 mm in diameter.

472

IAN

MCPHEE

In contrast to the Attic Falaieffs, the Corinthian examples are all coarseware. No fragments of local fineware versions from Corinth have been identified. The fabric of fragments Cl to C16 is essentially the same, and the variations in the color of the fired clay areprobablydue to differing conditions in the kiln. The basic fabric is a red clay, tempered mainly with red and gray grits (mudstone and tuffite), but also with white and yellow inclusions; it fires light red to orange, or gray,or gray in the center of the core and light red near the surfaces. It is hard and impermeable. The surfaces have been smoothed, and often have a creamish orange color: I do not know whether this is due to the addition of a very thin slip of fine yellow clay,or is the result of a self-slip, perhaps caused by the addition of potash to the matrix.27In some cases there seem to be traces of a thin wash that has fired a dark gray.28The same fabric was employed in the manufacture of Corinthian A transport amphorae during the 4th century,2' and essentially the same fabric, often termed "blister-ware,"was used for certain types of kitchenware and lamps at Corinth during the Classical period.30 Chemical and petrological analyses support the identification of this fabric as Corinthian.3"There can be no doubt, therefore, that these fragmentary Falaieff kraterswere all made in Corinth, and not imported. Like their Attic counterparts,the Corinthian Falaieffswere largevases. None is well enough preserved to give a total height, but C2 is over 24 cm high without the lower body and base, while the outer diameter of the rim can be estimated at about 37.5 cm. We can also determine that the rim of Cl was about 34 cm in diameter,and that of C6 even greater,about 43 cm. It will be apparentthat these fragmentaryvases display different conceptions of the same basic shape. The variation is especially apparent in the five examples which preserve the profile of the lip. In the case of Cl a tall, straight lip is divided into two unequal parts by a horizontal ridge, and the rim is not only flat on top but the lower element has a straight outer edge. C2 and C3 also have a tall lip, but it is more flaring and not articulated on the outer surface,and the rim is formed differently,with a rounded lower element. C6 and C10 have the same form of lip: short, with slightly convex profile outside, and a squared rim separated off by a horizontal groove. The join of body and lip is masked on C2 by a broad band divided into three ridges by horizontal grooves, and the form seems to have been similar on C4, C5, C8, C9, C12, C13, and C14, though all three ridges are preserved only on C8, C9, and C12; on Cl the three ridges have been reduced to a single one, and on Cll to a simple flat band. Though the ridges on C2 and C9 were clearly practical as well as decorative, intended to mask the junction of body, lip, and collar, the ridge or ridges on most fragments where they are preserved are situated well below the junction, so that their function can only have been decorative.The inner collar on C2 was formed in one piece with the bowl, continuing its curve, and the lip was added separately.This seems also to be the case with C3, C9, and apparentlyfor C12-C14. The horizontal inner collar of Cl was, however, made separately,and this seems also to be true for Cll. The full width of this collar is preserved only on Cl: it is pierced with three rows of holes (perhaps originally twenty-nine or thirty in the outer row, about twenty in

27. For the possible addition of potash to the clay body of Corinthian A transportamphoraeto producean orange surface,see Vandiverand Koehler 1986, pp. 180, 208-209. 28. A similariron-richwash is used on CorinthianA transportamphoraeof the second half of the 4th centuryB.C.: see Vandiverand Koehler 1986, p. 180. 29. For the fabricof CorinthianA transportamphorae,see now Whitbread 1995, pp. 255-293, but also Vandiverand Koehler 1986. 30. Good parallelsfor the fabricare providedby the contemporaryblisterware table amphoraeC-71-185, Williams 1972, p. 156, no. 19, pl. 24; and C-1990-63, Williams 1991, p. 35, no. 37, pl. 12 (the captions of nos. 37 and 38 have been reversed);as well as by the oinochoe C-71-188, Williams 1972, p. 157, no. 23, pl. 25. Blister-ware is describedby Pemberton(1970, p. 300) and Edwards(in CorinthVII, iii, pp. 144-146). For comments on the fabricof Corinthiankitchenware,see CorinthVII, iii, pp. 117-119; and Corinth XVIII, i, p. 69. 31. The earlierscientific studies of Corinthianclays are summarizedin Jones 1986, pp. 174-189, 713, to which may now be added the important discussionin Whitbread 1995, pp. 255346.

FALAIEFF

32. Similar,then, to the handles on the Athenian finewareFalaieffA7 (Drougou 1979, pp. 266-267). 33. Terminologyvaries:in the archaeologicalliterature,they are also referredto as "nipples,""bosses,"or "warts." 34. See, for example,the remarksof R. S. Young in CorinthXIII, p. 41. 35. Coarsehydriai:CorinthXIII, p. 21, no. 14-1, pl. 6; p. 22, no. 15-1; p. 23, no. 16-9, pl. 6; p. 25, no. 17-5; p. 27, no. 18-8, pl. 8; p. 28, no. 20-3, pl. 6; CorinthXV, iii, p. 20, no. 30, pl. 3; Williams 1976, p. 102, no. 5, pl. 18. Fineware:CorinthXIII, pp. 24-25, no. 17-1, pl. 7. 36. CorinthVII, ii, p. 129, An 159, pl. 62; p. 155, An 293, pl. 78. Remarks in CorinthXV, iii,p. 345 under no. 2132. 37. There are at times knobs beside the handles of convexpyxidesdecoratedin ConventionalizingStyle, but these arevestiges of reflexhandles:see the 4th-century exampleillustratedin Williams 1972, pl. 27, no. 46.

BELL-KRATERS

FROM

ANCIENT

CORINTH

473

the second, and fourteen or fifteen in the inner), and has a raised rim. Such a rim also exists on C15, though the form is different. Both C3 and C16 also preserve evidence of at least three rows of somewhat irregularly spaced holes; C2 had at least two rows of holes (about twenty-three or twenty-four holes in the outer row); and Cll, C14, and C15 also had at least two rows. Normally the holes are 4-6 mm in diameter, although in the case of C16 the diameter is about 7 mm; and they are always pierced from the top. Cl, C4, C7, and C8 preserve handles, and although they were all positioned horizontally, the forms are varied: C4 is broad, straplike, with a central ridge;32Cl is round, curves up strongly in the middle, and has distinct disks where it meets the wall of the vase; C7 is round in section like Cl but does not curve up and lacks the handle-plates; while C8 was larger,like C4, but did not have a central ridge. In the case of Cl, C5, and C8 the handles seem to have been flanked by knobs,33although one cannot say whether they were a constant feature of these Corinthian kraters.The bowl seems to have been deep, and in some cases quite thickwalled, tapering in toward the base. At present it is not possible to point with certainty to any base from a Corinthian Falaieff, but if the fragments C-71-641 (under C2, Fig. 9) and lot 7079-277 (under C8, Fig. 12) do belong to kratersof this type, they would indicate that the vase was supported by a low ring-foot 10-13 cm in diameter. The principal morphological differences between the Corinthian and Attic series lie in the form of the rim; the nature of the inner collar; the variation in the manner of masking the junction of bowl, lip, and inner collar;the form of the handles;and the addition of knobs beside the handles on at least some of the Corinthian versions. Of the five Corinthian Falaieffs that preserve the profile of the lip, C2 and C3 seem to be closer in general shape to the Attic vases. Cl is quite different, not only in the form of the rim, but in the nature and position of the collar,the type of handle, and the presence of at least two knobs. And the low, convex lip with squaredrim of both C6 and C10 is even further removed. Given the small number of vessels, it is not possible at present to prove any definite typological (or chronological) development. One cannot, of course, assume that the simpler form of lip on C6 and C10 is earlier than the more complex forms on Cl-C3, and all five vases may in fact be contemporary products of different Corinthian potters. Two odd elements of the shape require further comment: the presence of knobs on at least Cl, C5, and C8, and the attachments of the handle on Cl. The use of knobs has a respectable ancestry in Greek pottery on both small and large shapes.34In Corinth the earliest examples after the Bronze Age occur on coarse hydriai of the Middle Geometric II and Late Geometric periods,as well as on the occasionaldecoratedoinochoe of the same time.35The practice continued into the 7th century on a few coarsewarepots,36but is not found, to my knowledge, in Protocorinthian or Corinthian fineware of the Archaic period, nor are there any instances in Corinthian pottery of the 5th century.37In the 4th century, however, knobs areemployed frequentlyon Corinthian finewarestemless bell-kraters decorated both in red-figure and black-glaze (with ivy, laurel, or a necklace design in the handle-zone rendered in thin slip or in added white).

474

IAN

MCPHEE

Perhaps the inspiration for the knobs on Cl, C5, and C8 came from this contemporary type of Corinthian krater.38Indeed, it is quite possible, though at present unprovable, that the potters who manufactured the stemless kratersmay also have produced the coarsewareFalaieffkraters.9 The preservedhandle on Cl is remarkablein that its ends do not pass directly into the wall of the vase, but terminate in disks set against the wall. This is not the normal manner ofjoining a handle to the body of a ceramic krater.It looks to be an imitation of the soldered handle-attachments of a metal vase.40Indeed, the whole appearanceof Cl, including the form of the rim and the horizontal, rather than slightly curved, collar, seems to reflect a metal prototype. These fragmentaryCorinthian examples of the Falaieff kratercannot be closely dated on the basis of shape. One can only say that the form of the lip on Cl-C3, and the form of the preservedhandles, particularlythat of Cl, suggest a date broadly in the 4th century.Nor do they display any decoration that might be chronologically significant. Fortunately,the archaeological contexts of some of the pieces provide important evidence. Most significant for the inception of the Corinthian Falaieff krater are the contexts of the five fragments C4, C6, C9, C10, and C14, all of which came from the Forum Southwest, under the west end of the South Stoa. C14 was found in the fill of the foundation trench for the north wall of Shop 31.41The pottery from this particulardeposit, mainly black-glazed drinking-vessels and plainware and coarsewarevessels, does not appearto include anything later than the early 4th century B.C. C4, C6, C9, and C10 came to light under the terraceof the South Stoa, north of the foundations for Ionic columns 32 and 33, and below the originalfloor of Building IV along its east wall.42The associatedpottery,which again consists mainly of black-glazed drinkingvessels as well as plainwareand coarsekitchenware and a little red-figure, seems to extend later than that found with C14, but does not include anything certainly datable after the first quarter of the 4th century.On the other hand, other large deposits of the first quarterof the 4th century, such as well 1937-1 and drain 1937-1,43 have not produced any fragments of Falaieff kraters,nor have any been identified from deposits that cease in the last half of the 5th century, such as wells 193410, 1939-1, and 1936-10.44From this evidence we may infer that the shape was introduced at Corinth during the first quarterof the 4th century,but that it was not very common. Let us now consider the contexts of the other fragments of Corinthian Falaieffs. C5 came from fill used in the foundation trench for the north stylobate of the South Stoa, in the same general area (Building IV) as C4, C6, C9, and C10; and C13 was also found in a deposit on the northern side of Building IV.45In both cases the latest pottery seems to run down approximately to the middle of the 4th century. This accords with the suggestion of the excavatorthat Building IV was "abandonedabout 350 or slightly thereafter."46Five fragments, Cl, C2, C7, C8, and C16, formed part of the large deposit associated with drain 1971-1.47This drain,which ran approximatelynorth-south between the east wall of Building II and the west wall of Building I, in the southwest corner of the later Roman Forum, went out of use at the time of the destruction of Building I, when

38. For the knobs on stemless bellkraters,see McPhee 1997, pp. 133-134. Another possible sourceof inspiration for these knobs is suggestedin note 68 below. 39. It is at least probablethat the CorinthianFalaieffswere made in the same workshopsas the CorinthianA transportamphoraeand blister-ware amphorae. 40. For the handle-attachmentsof bronze hydriai,cf. Diehl 1964, pls. 811, 18. Comparealso the handleattachmentsof a bronze bowl from Derveni:Themelis andTouratsoglou 1997, p. 31, A51, pl. 32. 41. Williams 1979, p. 127. 42. Williams 1973, pp. 17-18, nos. 17-19. 43. For these, see CorinthVII, iv, pp. 18-21, Deposits 4-5; and McPhee 1997, p. 124. 44. Well 1934-10: Pease 1937; CorinthVII, iv, pp. 17-18, Deposit 3; CorinthVII, iii, p. 201, Deposit 10. Well 1939-1: CorinthVII, iii, p. 200, Deposit 8. Well 1936-10: CorinthVII, iii, p. 201, Deposit 11. 45. For the excavationfrom which C13 came, see Williams 1980, pp. 111115, with p. 113, fig. 3. 46. Williams 1979, pp. 105, 130. 47. Plans of the areaarein Williams 1972, p. 150, fig. 3, and p. 166, fig. 5. A selection of the pottery from the drainis presentedon pp. 154-163, pls. 24-27. See also McPhee 1997, p. 125. This large and importantdeposit is currentlybeing studied by Dr. Elizabeth Pembertonand myself.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

475

it was filled with a great quantity of broken pottery and some other objects (terracottas,broken roof-tiles, coins). The pottery consists mainly of blackglazed vessels (skyphoi,bowls of varioustypes, saucersand plates),plainware (two-handled jugs, bowls), coarse kitchenware (cooking-pots, casseroles), transport amphorae (mainly Corinthian A, A', and B), some lamps, and a few scraps (mainly kraters)of Attic and Corinthian red-figure. This pottery seems to be generally datable to the second and third quartersof the 4th century, the latest pieces belonging to the last quarter and reaching perhaps as late as 320-310.48 C12 came from the fill in well 1971-1, located toward the east end of the South Stoa and partly cut into by the building's stylobate. The pottery from this fill seems again to be relatively uniform, the latest pieces dating to about 320-310. Well 1971-1 and drain 1971-1 therefore contained contemporary deposits, and both were filled during the same construction program-the levelling of the areain preparation for the building of the South Stoa. Thus, Cl, C2, C5, C7, C8, C12, and C13 may all be dated before ca. 300. At present there is no contextual evidence that kratersof Falaieff type were being produced in Corinth after ca. 300.49The shape enjoyed a period of popularity from the first to the last quarterof the 4th century,but particularlyduring the second and third quarters.It is possible that it outlived its Attic fineware counterpart. This Corinthian version of the Falaieff krater does not seem ever to have been produced in large numbers, and it is remarkablethat, with the exception of C3, Cll, and C15,50 all these fragmentarykratersfrom Corinth were found in the southwest areaof the later Roman Forum, in association with a group of buildings of the Classical period.5'We have alreadyestablished that Cl, C2, C7, C8, and C16 were part of the large deposit of pottery from drain 1971-1, between Buildings I and II. Now, that deposit consisted mainly of shapes connected either with the preparationand serving of food or with the storage, mixing, and drinking of wine. This essentially homogeneous fill of pottery may have been taken from somewhere in the immediate area of the drain, possibly from Buildings I and II and from the badly preserved structureto the east of Building 1.52 The excavator conjecturedthat the deposit did not represent domestic debris from an ordinary house but refuse from a building or buildings with some civic and/or cultic function that involved dining.53C4-C6, C9, ClO, C13, and

48. For this dating, see Williams 1972, p. 153; Williams 1976, pp. 115116; and Williams 1977, pp. 51-52. The destructionand rebuildingat this time was, as Williams (1976, pp. 115116) points out, not localized but widespreadin Corinth.Williams (1976, p. 116, note 20) even postulatestwo destructions,one ca. 330-320, the other ca. 310-300. Most recently,Williams (1995, pp. 44-45) has suggestedthat the South Stoa, the successorof

Buildings I-IV, was constructedwith "Siciliancampaignloot sent back to Corinth after341 B.C." 49. I have not found any fragments in the deposits associatedwith the use of the South Stoa. 50. C3 comes from a large dumped fill (lot 5777) in the areaof the Sacred Spring;the pottery is mixed, including pieces of the first quarterof the 3rd century,but most pieces belong to the 5th and 4th centuries.The most recent

discussionof the deposits of the Sacred Spring is in Steiner 1992. The contexts of Cll and C15 arenot chronologically significant. 51. Even Cll is probablyconnected throughits context with the Classical buildings that predatethe South Stoa. 52. Williams 1972, pp. 164-165. See also McPhee 1997, p. 127. 53. Williams 1972, pp. 164-165, 172-173.

476

IAN

MCPHEE

C14 were all found in the general area of Building IV.54This building, which seems to have gone out of use about 350, earlier than Buildings IIII, was probably not an ordinaryhouse, but a construction of more elaborate form, perhaps again with a public function that included dining (suggested by the pottery) and cultic activity involving small terracottaaltars.55 Given this evidence, it may be proposed that the krater of Falaieff shape was introduced at Corinth in connection with a specific civic or ritual activity involving dining.56It is remarkablethat the Corinthian stemless bell-krater,which was decorated in red-figure as well as black-glaze, was produced contemporaneouslywith the Falaieff krater,and that the majority again come from deposits in the Forum Southwest.57Were both shapes, the one fineware, the other coarseware,used for public dining in connection perhaps with the cults in the area?

ORIGINS As we have seen, the evidence of archaeological context, of shape, and of style suggests that the kraterof Falaieff type began to be manufacturedin clay in both Athens and Corinth early in the 4th century.In neither center,however, is there any local ceramic precursorfrom which the form may be derived. However, if we look further afield, we find in the Bucchero pottery of Etruria a shape that is remarkablysimilar and that must be included in any consideration of the origins of the Greek Falaieff krater. The basic discussionsof these Etruscanvases areto be found in Camporeale 1971 and Brommer 1980. Camporealegives a list of five examples,to which Brommer added the vase in Essen, and Hayes that in Toronto.58 El

Florence, Mus. Arch., 2859, from Orvieto. Fig. 17

H. 0.21 m. Camporeale 1971, p. 258, no. 1, pl. LXXII. E2 Rome, Villa Giulia. H. 0.21 m. Camporeale 1971, p. 258, no. 2; Mingazzini 1930, p. 56, no. 209, pl. VI:8-9. E3 Essen, Museum Folkwang, A 112, said to have been found at Chiusi. Fig. 17 H. 0.231 m. Brommer 1980, pl. 115; Froning

1982, pp. 110-112, no. 43; EAA Suppl. 11/1 (1994), p. 766, fig. 851. E4 Toronto, Royal Ontario Mus., Fig. 17 918.3.165A.

H. 0.183-0.188, with handles

0.25 m. Hayes 1985, pp. 92-93, 188. E5 Rome, Villa Giulia, Gorga Coll. 85. H. 0.19 m. Camporeale 1971, p. 258, no. 5. E6 Rome, Villa Giulia, TR 31-TR 43, from Trevignano. H. 0.215 m. Camporeale 1971, p. 258, no. 3; Arte e civ/tc, p. 52, no. 93. E7 Orvieto, Museo dell'Opera del Duomo, 570, from Orvieto. Fig. 17 H. 0.19 m. Camporeale 1971, pp. 258-259, no. 4, fig. 1; Brommer 1980, p. 336, fig.2.

54. For the excavationof Building IV, see Williams 1973, pp. 17-18; Williams 1979, pp. 125-136; Williams 1980, pp. 111-116. 55. The altarsare discussedin Williams 1979, pp. 136-140. 56. Certainlythere were a number of cults establishedin this part of the Classicalcity: for hero cults, see most recentlyWilliams 1978, but also Broneer 1942, and for the hero shrine nearBuilding I, Williams 1972, pp. 149-151; for Dionysos in the area of the Forum,see Williams 1975, pp.28-29. 57. For a discussionof the function of the stemless bell-kratersee McPhee 1997, pp.126-127. 58. Hayes 1985, pp. 92-93, 188.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

477

~~~~~~~~~~~~wgs

E1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

*_

_.E_ _f

E3

478

IAN

MCPHEE

Although there are variations in shape within this group, all seven vases have certain common elements. They are relatively small, ranging from 18 to 23 cm in height (excluding the handles). They all have a low foot, a spreading bowl, a tall lip, and an internal collar pierced with two groups of holes. There is clearly a general similarityin shape to the Falaieff krater.59 There are, of course, some differences between the two series:the lip in the Buccherovases is waved to form two "spouts,"or is at least pinched in two places; the collar is pierced with many holes in two zones adjacent to the spouts; five of the vases (El-E5) have the handles positioned vertically from the shoulder to the lip; and some at least (El, E4?, E6, E7) were providedwith a lid. But the main differenceis one of size:were the Etruscan vases used for mixing or for drinking or for both functions? Certainly,they must have servedas strainersin a mannersimilarto the CorinthianFalaieffs. The Bucchero vases are not precisely dated, but on the basis of their decoration and of the tomb context of E6 they have been placed between the end of the 7th and the middle of the 6th century.60This leaves a gap of at least a century and a half before the appearance of the earliest Greek kraters of Falaieff type.61It may be, in this case, that we have to do with two independent creations, perhaps with somewhat different functions; nevertheless, the relationship in shape between the Etruscan vases and the Greek Falaieffs is so striking that it is hard to see the two forms as completely independent, particularlyin view of the highly specialized shape. Of course, it may be that the late 6th- and 5th-century Greek ceramic imitations of these Bucchero vases have not yet been recognized, even though this seems rather unlikely.62On the other hand, it is possible that the development from the Bucchero shape to the Falaieff krateroccurred first in metal versions, even if none has been preserved. In this regardwe have alreadynoted that the shape of one Corinthian Falaieff (Cl) strongly implies a metal prototype.63 At present it is also not possible to be sure whether the Falaieff shape was produced first in Athens or Corinth.64However, the evidence may be thought to favor Athens. All the fragments of Attic Falaieffs found at 59. The comparisongiven in Brommer1980, p. 336, figs. 1 and 2, is, however,quite misleadingsince Brommerdoes not indicate that the scales are entirelydifferent. 60. This is the date given to E3 in Froning 1982, p. 111; Camporeale (1971, p. 261) suggests the end of the 7th and the first decadesof the 6th century;Hayes 1985, p. 93, early6th centuryfor E4. Typologically,El and E2 may be earlier,and E6 and E7, which arecloser to Greek Falaieffs, may be later.It is not possible to be certainwhere these vases were made, but Orvieto may have been one center, as Camporeale(1971, p. 260) suggests. 61. Geneva MF 254, a small Etruscanred-figurecolumn-krater,

which has a piercedmouth or inner collarlike the Athenian Falaieffs,may be evidencefor the influence of the Greek Falaieffsratherthan for the continuityinto the 4th centuryof a version of the Etruscanshape:see EVPE p. 301; Schauenburg1980, p. 50, pl. 29; Brommer1980, p. 339. 62. There is, however,one other curiouspiece of evidencethat may be relevant.Corinth lot 1978-100-1 (Fig. 18) is a fragment(H. 0.05 1, W. 0.16, est. Diam. rim 0.25 m) from the upper wall of a vessel with a flanged rim (as though to take a lid), tracesof the stumps of a handle, and the beginning of an inner collarpiercedwith holes (some ten remain)about 4 mm in diameterand 7-10 mm apart.The

exact shape is unclear,but may have been similarto the Roman vessels mentioned in note 71. The outer surfaceis undecoratedexcept for a horizontalband.The fabricis fine but slightly micaceous,hard,pinkish (5YR 6-7/4) in the core but light brown (5YR 7/6) on the surface.It is not Corinthian,but may be Attic. The archaeologicalcontext, although disturbed,suggests a date in the mid5th centuryor earlier. 63. For the relationshipbetween metal and ceramicvases, see Vickers 1986. 64. Brommer(1980, p. 337), being unawareof the existence of Corinthian Falaieffs,assumedthat the shapewas taken up at Athens.

FALAIEFF

I

BELL-KRATERS

FROM

ANCIENT

CORINTH

479

a

I

I

I

Figure18. Corinthlot 1978-100-1, Attic(?).See note 62. Scale:profile 1:1, photos 1:2

65. There is no evidencethat the shapewas imitated in the local redfigure at Corinth. 66. Corinthiancommercewith the Western Mediterraneanis discussedin Munn 1983. Fragmentsof Etruscan Bucchero,mainly kantharoi,reached Corinth in some quantityduringthe first half of the 6th century,just at the time when the Buccherokraterswere being produced:see MacIntosh 1974; furtherexampleshave been found in excavationssince 1973.

Corinth seem to belong to the first quarter of the 4th century, whereas most of the Corinthian versions probably belong to the second and third quarters.If the shape developed first in Athens, red-figure examples (and the metal version?) very soon reached Corinth, where the shape was imitated, but only in coarseware and perhaps metal.65On the other hand, Corinth had developed a brisk commerce with Greek and Etruscan centers in Italy in the Archaic and Classical periods.66The city also had an important tradition of metalwork and utilitarian pottery, and during the 5th and early 4th century her potters continued to experiment with new or adapted shapes.67It is possible, then, that a ceramic version of the shape was first developed in coarsewareat Corinth, inspiredby models in metal.68 The shape would soon have been taken up in Athens where the potters began to produce fineware examples, decorated in red-figure, for export to Corinth and elsewhere: it would not have been the first such venture.69 Because of its possible derivation from the Etruscan Bucchero shape, Brommer attempted to identify the kraterof Falaieff type with the xpatilp -vtopp-ovxo;("Etruscankrater")that is listed in inscriptions from Delos.70 It may be objected that these inscriptions belong to the 3rd century, after the Falaieff kraterhad ceased to be produced in pottery,but metal versions

67. For Corinthianmetalworkand clay imitations of metal, see Pemberton 1981. Local versionsof Athenian white-ground lekythoi and of shapes decoratedin red-figureor stamped black-glazearefound after440-430: see Steiner 1992, pp. 391-399 on white-groundlekythoi;Pemberton 1997 on stampedblack-glaze. 68. It may be noted that one of the Buccherokraters,E5, has additions ("appendici")beside the handles (Camporeale1971, p. 261, note 9); such

additionsmay have providedthe inspirationfor the knobs on the CorinthianFalaieffs. 69. A much earlierinstance is the Nikosthenic neck-amphora:see Rasmussen1979, pp. 74-75, 168. See also the remarksin Miller 1993, pp. 110-112. Comparealso the Attic red-figureversionsof the Apulian trozella:Jentoft-Nilsen 1988. 70. Brommer1980, pp. 337339.

480

IAN

MCPHEE

of the shape may well have outlasted the ceramic. Still, this ingenious suggestion seems unlikely, for the Delian inscriptions do not give a clear description of the xpcxp so that an identification with the upp-vcx6, Falaieff krateris entirely conjectural.

FUNCTION The exact function of the krater of Falaieff type has not yet been determined with complete satisfaction, and the following discussion will not resolve this situation.7"The impermeability of the Etruscan, Corinthian, and Athenian vases indicates that they were intended to hold liquid. That the liquid was wine is a reasonable assumption, based not only upon the shape of these vases but also upon the Dionysiac iconography of the Attic Falaieffs.72As we have seen, the Bucchero vessels are small, have two "spouts"at right angles to the handles, and have two groups of holes in the collar adjacent to the spouts: whether or not they served for mixing wine and/or for drinking,they must certainlyhave functioned asvessels for straining and pouring.73Since the central opening was covered in some cases with a lid, Camporeale believed that the Bucchero vases held a heated, perhaps aromatic, liquid.74 On the other hand, the Greek Falaieff krater,whether fineware or coarseware,was a large vessel, not intended for pouring, but presumablyin some way connected with the general preparation of wine for drinking. Any more specific interpretation must take into account the tall lip and the peculiar inner collar pierced with holes. We can, I think, dismiss any suggestion that a kraterof this type served as a wine-cooler (psykter).75 In her study of the shape, Stella Drougou tentatively offered two different suggestions for the function of the Falaieff krater.76First, she suggested that the kratermay have been used for the preparationof a form of mulled wine. We know that the Greeks mixed wine with boiling water, honey, salt, and different spices to produce various kinds of mulled drink.77 Drougou quotes in particular the following fragment of the comic poet

71. A much earlier,Middle Minoan vessel from Knossos,with a piercedinteriorcollar,may have had a function similarto that of the Falaieff kratersdiscussedin the presentarticle: Pendleburyand Pendlebury1932, pp. 66-67, and PM IV, pp. 72-73, fig. 45. This may also be the case with a type of Roman cylindricalvessel of the lst-2nd centuriesA.C., examples of which have been unearthedin Corinth:Slane 1986, p. 287, no. 57, pl. 64, and p. 310. These vessels have been thought of as incense-vases,winecoolers,and pots for scaldingmilk or for heating wine mixed with water

and spices:Nuber 1969-1970, where earlierliteratureis cited. See also the vase mentioned in note 62 above. 72. Drougou (1979, pp. 277-278) discussesthe iconographyof A1-A7, and notes the connection of the grypomachywith Dionysos. In this respect,I may add that the god is sometimes shown riding a griffin on Attic red-figurevases of the 4th century:ARV2 1453, nos. 810; 1694, no. 9bis;perhaps1523, no. 6. 73. Camporeale(1971, p. 260) and Brommer(1980, p. 337) saw that they

must have been used for pouring. Froning (1982, p. 111) disputes this interpretationon the grounds that the rim is merelypinched, not waved,but the position of the holes and the use of verticalhandles supporta pouring function. 74. Camporeale1971, p. 260. 75. Drougou (1979, p. 278) presents a sensible argumentagainstthis function. 76. Drougou 1979, pp. 280282. 77. For additives,see Pricket 1980, pp. 54-56. A more general discussionis in Younger1966, pp. 130-133.

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

481

Alexis, who was active at Athens in the middle and second half of the 4th century B.C.78 JDXL8p?q 8? xpTxeI'p 00tqx?sLo;

?V ?a

eCrG-Xe XeXOO VSXt-Xp0 7rOCXOXLYeVOD

7rXY, 1P1 ?,x&

b xsV6V

V: OV

-cpr4oc;,7roNax;?XBXtpov, POxkA f6cCav a-uCijoc;, 1

auV6ix

Xc piTEoL;

XLGCYoL

X?(o?ot;

On the basis of these lines, Drougou wondered whether the Falaieff krater might have contained "foaming or bubbling"wine, which, if it overflowed the brim of the containing inner collar, would have been retained by the high lip, flowing back into the bowl through the holes. A corollaryof this might be the identification of the Falaieffkraterwith the Therikleian krater mentioned by Alexis.79 Drougou's second suggestion was that the shape was used in the purifying of wine. She considered this less likely than her first conjecture,but the two are not mutually exclusive and perhaps the Falaieff krater performed both functions. Much of ancient wine must have been bad, containing sediment and many impurities, and Corinthian wine was not always much prized.80In addition, as we have noted, the wine might be spiced with aromatic herbs or mixed with water heated and spiced. The process of straining the wine (or wine and water) was essential, so that it is not surprising to see a strainer (qOc6o;,q,O6,vtov), along with a ladle and a jug, sometimes representedin scenes of symposia on Attic red-figure vases of the Classical period.8' The wine might have been strained as it was poured from the jug into the drinker'scup, but it might also have been strained at an earlierstage, when decanted from the storage vessel into the 78. Athenaeus 11.472a (Edmonds 1959, no. 119): "Therestands shining a thericleianmixing-bowl right in our midst, filled with white nectarof ancient vintage,all a-foam;I had taken it empty and polished it up, making it bright;I set it firmlyon its base, and wreathedit with berry-ladensprigsof ivy which I had plaited together." (trans.C. Gulick, Loeb) 79. The latest discussionof Therikles andTherikleianvases seems to be Gill 1986, pp. 9, 19-23, which is useful for bibliography,but which shows more zeal than prudencein the discussion.Our sources(e.g., Athenaeus 11.470f) make it perfectly clear that Therikles,a Corinthian,was a potter,not a metalworker.We are told that he was active (whetherat Corinth or Athens cannot be determinedwith certainty)duringthe lifetime of

Aristophanes,so probablyabout the last quarterof the 5th or first quarterof the 4th century.The earlyreferences,in the 4th centurycomic poets (Athenaeus 11.467d, 469b, 470e472d), imply that his most renowned "invention"was a particularshape of drinking-vesselthat was producedin clay,wood, and metal.The clay exampleswere decoratedin a particular manner,possiblylustrousblackglaze, sometimeswith gilded clay ornament: on this aspect,see Shefton 1971, p. 110. The most revealingnew evidenceis providedby the inscribedblack cupkantharoifrom Kafizin in Cyprus publishedby Mitford (1980, pp. 29-32, nos. 40-42; the readingof OpLxXetov on the fragmentarylagynos no. 46 on pp. 35-36 is unlikely).The Kafizin cupkantharoiarereallyelongatedversions of the originalstandardcup-kantharos

with molded rim that begins about 400-380 B.C.:AgoraXII, p. 118, nos. 648-670. Curiously,the earliest examplesquoted in AgoraXMI, especiallyno. 649, which have the distinctivemolded rim, come from Corinth. For a differentidentification of the cup-kantharos,see Pfrommer 1986, pp.3-24. The "Therikleian"kraterdescribed by Alexis seems to have been metal. 80. At least in Athens. Alexis could say (Athenaeus 1.30f; Edmonds 1959, no. 290): otvo; evLtx6iocp,v: roeyocpKopt'v8c

lL. POcGOCvT[Iv6,; "Therewas importedwine on hand; for the Corinthianstuff is torture." (trans.C. Gulick, Loeb) 81. See Hill 1942, pp. 44-45.

482

IAN

MCPHEE

krater.Drougou therefore suggested that the inner collar served to support a strainer,the lqOt6; ExP?-c8p?o 82 Such a strainer could have been of terracotta-a simple bowl or jug of suitable dimensions, pierced with holes83-or of metal, possibly just a flat perforated sheet.84On occasion, however, a more ornate metal strainer might have been employed, as Drougou has suggested. One of the principal Greek types of strainer in the 4th century B.C., made in silverand bronze, has a broadrim, two handles set in the same plane as the rim and ending in duck- or swan-heads, and either a shallow bowl perforated with tiny holes or a molded funnel covered with a pierced metal disk.85This form of metal strainermight fit over the central opening of a Falaieff krater,resting upon the raised rim of the projecting collar,86but it must be admitted that the arrangement seems rather awkward.In any case, whether the strainerwas a ceramic pot or a simple metal sheet, any wine that spilled onto the coliar during decanting would have flowed through the holes into the bowl. One other piece of circumstantialevidence in support of this proposal may be cited. Metal strainer-lids were frequently fitted to metal volutekratersfrom the second half of the 6th to the second half of the 4th century, as examples from Vix, Trebenischte, and Derveni and in the Ortiz Collection illustrate.87If bronze or silver versions of the Falaieff krater were ever manufactured,they may have been suppliedwith similarstrainerlids supported by a narrow inner flange. In the terracotta Falaieff kraters, however, a broad inner collarwas more practical,providing both the opening necessaryfor ladling the wine, and the support necessary for a separate strainer.

But we must also take into consideration a curious difference in form between the inner collar of the Athenian and Corinthian versions of the 82. PoUux10.108. See also Amyx 1958, pp. 261-264. On strainersin general,Hill 1942; Crosby 1943; Strong 1966, pp. 92-93,217; Mertens 1976, pp. 71, 73-74, 78-80; Oliver 1977, pp.44-45,47. 83. We have alreadynoted that the St. PetersburgFalaieff,Al, was reportedlyfound with a bowl covering the centralopening, though it is not stated that the bowl was pierced. Sparkesand Talcott (AgoraXII, p. 106) remarkthat in the Classicalperiod a pottery strainerwas normallymade by piercingthe bottom of a suitable shape, such as a bowl.There are examplesof such piercedbowls and jugs at Corinth. 84. It is worth recallingat this point the piece of bronze that remainsin one of the holes of Cll, although its correct explanationis unclear. 85. The strainerwith perforated bowl and duck- or swan-headhandles is type 6 in Hil 1942, p. 54, table 6;

for the variantwith funnel instead of bowl, see Oliver et al. 1987, pp. 192193. Here are a few examples,in silver and bronze:Thessaloniki,from Vergina, Andronicos 1984, p. 148, fig. 108; Thessaloniki,from Vergina,Andronicos 1984, p. 211, fig. 178;Thessaloniki,5145, from Potideia,Michaud 1970, p. 1069, fig. 392; Thessaloniki, from Derveni A, Themelis and Touratsoglou1997, p. 37, A14, pl. 43; Thessaloniki,from Derveni B, Themelis andTouratsoglou1997, p. 69, B4, pls. 11 and 72; Thessaloniki,from Derveni D, Themelis andTouratsoglou 1997, p. 104, Dll, pls. 21 and 115; Thessaloniki,M 452, from Nikesiani, Lazarides,Rhomiopoulou,and Touratsoglou1992, pp. 22-23, pl. 7; Thessaloniki,A 2581, from Nikesiani, Lazarides,Rhomiopoulou,and Touratsoglou1992, p. 42, pl. 26; Baltimore,WaltersArt Gallery,57.910, fromThessaly(?),Oliver 1977, pp. 4445; Istanbul,1415, from Kastamonu,

Erdal 1989, p. 334, figs. 4-5; Sofia, from Duvanlij,Filow 1934, p. 176, fig. 195; Boston, Museum of Fine Arts, 24.874, from Meroe, Smith 1960, p. 186, fig. 124; New York,Fleischman Coll., True and Hamma 1994, pp. 7778, no. 31d. 86. The diameterof the central opening of A6 is 0.147 m; the inner diameterof the vase at the level of the collar,about 0.27 m. In the case of A4 the centralopening was larger,about 0.18 m in diameter.The bowl of these metal strainersseems usuallyto have a diameterof about 0.09 to 0.12 m, and the total length acrossthe handles seems to be about 0.18 to 0.22 m. 87. Joffroy1954, pl. 16 (Vix) and pl. 19 (Trebenischte);Gioure 1978, pl. 5 (Derveni);Ortiz 1994, no. 149. A similarstrainer-lidfor a volute-krater has come to light at Olympia:Gauer 1991, pp. 8 and 255, M27, pl. 83 (I am most gratefulto Elizabeth Pemberton for this reference).

FALAIEFF

BELL-KRATERS

FROM

ANCIENT

CORINTH

483

shape: in the former there is only a single row of four to eight holes, in the latter the norm was probably three rows of closely-spaced holes. In the Attic red-figure Falaieffs, the collar cannot itself have been intended as a strainer: these vases were showpieces, intended to impress the guests at fashionable symposia, where costly metal strainers might have been important accessories.The coarseware Corinthian vases were more humble but more practicalpots, in which the inner collar must itself have served as a strainer,whatever was placed over the central opening. Given the size of the holes in the Corinthian Falaieffs, we might conjecture that a heated, spiced wine was poured into the krater,so that the holes helped to catch the added herbs. In this situation it might have been convenient to have a lid for the central opening, as in the case of the earlier Etruscan Bucchero vases. In fact, experiment with Cl shows that some at least of the ceramic lids made for Corinthian A transport amphorae could also have been employed in a secondary function to cover the central opening of the Falaieff krater.88

CONCLUSION

88. For these lids see Koehler 1986, pp. 55-56, with figs. 5-6.

Fineware kraters of Falaieff type were made in Athens during the 4th century B.C. This has long been known, but in this article I have shown that Athenian examples of the shape were exported to the important commercial city of Corinth. Moreover, there is now evidence that Corinthian potters occasionally produced a coarsewareversion of the shape. Whether the Corinthian or Attic Falaieffs are the earlier cannot yet be determined. Nor are the origins of the Falaieff krater certain, though a very similar, if smaller,vessel was made in Etruria in the later 7th and 6th centuries, and the connection may have been provided by examples in metal that have not survived.The Attic kraters,and presumablythe Corinthian, were used in the preparationand consumption of wine at the symposion, though the specific function of the shape remains conjectural. Despite these uncertainties, we have been able to add a new shape to the 4th-century ceramicrepertoryat Corinth, a shape that once again shows the Corinthian potters experimenting with a plain, utilitarian form. And another connection has been forged between the potters in Corinth and Athens in the Classical period.

REFERENCES AgoraXII = B. A. Sparkesand L. Talcott, Blackand Plain Potteryof the 6th, 5th, and 4th CenturiesB.C., Princeton 1970. AgoraXIV = H. A. Thompson and R. E. Wycherley,TheAgoraof Athens,Princeton 1972. AgoraXXIX = S. I. Rotroff,Hellenistic Pottery:.Athenian and Imported TableWareand Related Wheelmade Material,Princeton 1997.

Alexandrescu,P. 1978. Histria IV: La ceramique d'epoquearchaiqueet classique(VIIe-IVe s.), Paris. Amyx, D. A. 1958. "The Attic Stelai, Part III: Vasesand Other Containers,"Hesperia 27, pp. 163-307. Andronicos,M. 1984. Vergina:The RoyalTombs,Athens. Arte e civilta = Arte e civilta degli Etruschi,Torino 1967. Belov, G. D. 1945. "Krasnofigurnye

IAN

484

Krateriz Chersonesa,"Trudy GosydarstvennogoErmitazha 1,

pp. 141-145. Boardman,J. 1989. Athenian Red Figure Vases:The Classical Period, London. Brommer,F. 1980. "Krater Tyrrhenikos,"RM 87, pp. 335-339. Broneer,0. 1942. "Hero Cults in the CorinthianAgora,"Hesperia 44, pp. 1-29. Camporeale,G. 1971. "Vasoda filtro di Bucchero,"ArchC/23, pp. 258-261. Corinth VII, ii = D. A. Amyx and P. Lawrence,Archaic Corinthian Pottery and theAnaploga Well,

Princeton 1975. Corinth VII, iii = G. R. Edwards, Corinthian Hellenistic Pottery,

Princeton 1975. Corinth VII, iv = S. Herbert, The RedFigure Pottery, Princeton 1977. Corinth XIII = C. W. Blegen, H. Palmer,and R. S. Young, The North Cemetery, Princeton 1964. Corinth XV, iii = A. N. Stillwell and J. L. Benson, The Potters' Quarter. The Pottery, Princeton 1984. Corinth XVIII, i = E. G. Pemberton, The Sanctuary of Demeter and Kore: The GreekPottery, Princeton 1989.

Crosby,M. 1943. "ASilverLadle and Strainer,"4AJA 47, pp. 208-216. Diehl, E. 1964. Die Hydria, Mainz. Drougou, S. 1979. "Ein neuer Krater aus Athen: Die Gruppe'Falaieff,'" AA, pp.265-282.

Edmonds,J. M. 1959. The Fragments of Attic Comedy II, Leiden. Erdal,B. 1989. "Miizelerimizde sergilenenta?lnlrarkeolojikve ethnografikkiiltiirvarliklarinizin bozulma sebeplerive koruma (konservasyon)metotlarl," TuirkArkDerg28, pp. 323-335. Filow, B. 1934. Die Grabhiigelnekropole bei Duvanlij in Siidbulgarien, Sofia. Frel,J. 1964. "Un cratereattique retrouve,"Eirene 3, pp. 123-125. Froning,H. 1982. Museum Folkwang. Katalog der griechischenund italischen Vasen,Essen. Gauer,W. 1991. Die Bronzegefasse von Olympia I (OlForschXX), Berlin.

Gill, D. 1986. "ClassicalGreek Fictile Imitationsof PreciousMetal Vases," in Pots and Pans. A Colloquium on PreciousMetals and Ceramics,

M. Vickers,ed., Oxford,pp. 9-30.

MCPHEE

Gioure, E. 1978. 'OKp,oxpa,; -rov Apf3Pv(oV, Athens. Hayes,J.W. 1985.Etruscanand Italic Potteryin theRoyalOntarioMuseum: A Catalogue,Toronto. Hill, D. K. 1942. "Wine Ladles and Strainersfrom Ancient Times," JWalt5, pp. 40-55. Isler-Kerenyi,C. 1982. "Il trionfo di Dioniso,"NumAntCl11, pp. 139155. Jentoft-Nilsen, M. 1988. "TwoAttic Vasesof the Unique Shape,"in Proceedings of the Symposiumon AncientGreekand RelatedPottery, Copenhagen, August31-September4, 1987, J. Christiansenand T. Melander,eds., Copenhagen, pp.278-283. Joffroy,R. 1954. Le Tresorde Vix (MonPiot48.1), Paris. Jones, R. E. 1986. Greekand Cypriot Pottery:AReview of Scientific Studies,Athens. Koehler,C. 1986. "Handlingof Greek ContainerAmphoras,"in Recherches surlesamphores grecques(BCH Supplement13), J. Empereurand Y. Garlan,eds., Paris,pp. 49-67. Lazarides,D., K. Rhomiopoulou,and G. Touratsoglou.1992. 'O T6jif3of -r Ntx,oatcvv,; Athens. MacIntosh,J. 1974. "Excavationsat Corinth, 1973, Appendix I: EtruscanBuccheroPotteryImports in Corinth,"IHesperia43, pp. 34-45. Margos, R. 1980. "Unepelike attique a figuresrouges du IV siecle avant J.-C.,"in Etudesd'archbologie grecque prsenMesle 16.1111978a Violette Brussels. Verhoogen, McPhee, I. 1976. "AtticRed Figure of the Late 5th and 4th Centuries from Corinth,"Hesperia 45, pp.380-396. . 1997. "StemlessBell-Kraters from Ancient Corinth,"Hesperia 66, pp. 99-145. Mertens,J. 1976. "AHellenistic Find in New York,"MMAJ 11, pp. 71-84. Michaud,J. P. 1970. "Chroniquedes fouilles en 1968 et 1969,"BCH 94, pp. 883-1164. Miller, M. C. 1993. "Adoptionand Adaptationof Achaemenid MetalwareFormsin Attic BlackGloss Ware of the Fifth Century," AMIran 26, pp. 109-146.

FALAIEFF

BELL-KRATERS

FROM

Mingazzini, P. 1930. Vasidella collezione Castellani, Rome. Mitford, T. B. 1980. The Nymphaeum of Kafizin: The Inscribed Pottery (Kadmos SupplementII), Berlin.

Munn, M. L. 1983. "CorinthianTrade with the West in the Classical Period"(diss. Bryn Mawr College). Nuber,H. U. 1969-1970. "Ein Siebgefassaus Heddernheim,"Rei Cretariae Romanae Fautorum Acta

XI-XII, pp. 70-75. Oliver,A. 1977. Silverfor the Gods, Toledo. Oliver,A., et al. 1987. Antiquitiesfrom the Collection of Christos G. Bastis,

New York. Ortiz, G. 1994. In Pursuit of the Absolute:Art of theAncient World from the George Ortiz Collection,

London. Pease,M. Z. 1937. "AWell of the Late Fifth Centuryat Corinth,"Hesperia 6,pp.257-316. Pemberton,E. G. 1970. "The Vrysoula ClassicalDeposit from Ancient Corinth,"Hesperia 39, pp. 265-307. . 1981. "The Attributionof CorinthianBronzes,"Hesperia 50, pp. 101-111.

. 1997. "CorinthianBlackGlazed Potterywith Incised and StampedDecoration,"Hesperia 66, pp. 49-97. Pendlebury,H. W., and J. D. S. Pendlebury.1932. "Two ProtopalatialHouses at Knossos," BSA 30 (1928-1930) [1932], pp. 53-73. Pfrommer,M. 1986. Studien zu alexandrinischer undgrossgriechischer ToreutikfriihhellenistischerZeit,

Berlin. Pottier,E. 1922. Vasesantiques du Louvre III, Paris. Pricket,J. 1980. "AScientific and TechnologicalStudy of Topics Associatedwith the Grapein Greek and Roman Antiquity"(diss. Universityof Kentucky,Lexington). Rasmussen,T. 1979. BuccheroPottery from Southern Etruria, Cambridge. Robertson,M. 1981. "The Attic BlackFigure and Red-Figure Pottery,"in V. Karageorghiset al., Excavations at Kition IV: The Non-Cypriote Pottery, Nicosia.

ANCIENT

CORINTH

485

. 1992. TheArt of Vase-Painting in ClassicalAthens, Cambridge. Rotroff,S. I. 1996. "PnyxIII: Pottery and Stratigraphy," in The Pnyx in the History ofAthens. Proceedings of an International Colloquium Organised by the Finnish Institute atAthens,

B. Forsenand G. Stanton, eds., Helsinki, pp. 35-40. Rotroff,S. I., andJ. McK. Camp. 1996. "The Date of the Third Period of the Pnyx,"Hesperia 65, pp. 263-294. Schauenburg,K. 1980. "Skyllaoder Tritonin?Zu einer Gruppe canosinischerAskoi,"RM 87, pp. 29-56.

Schefold, K. 1934. Untersuchungenzu den KertscherVasen,Berlin. Shefton, B. B. 1971. "PersianGold and Attic Black-Glaze:Achaemenid Influenceson Attic Potteryof the 5th and 4th CenturiesB.C., AnnArch 21, pp. 109-111. Slane, K. W. 1986. "TwoDeposits from the Early Roman CellarBuilding, Corinth,"Hesperia 55, pp. 271-381. Smith, W. S. 1960. Ancient Egypt as Represented in the Museum of Fine Arts, Boston, Boston.

Steiner,A. 1992. "Potteryand Cult in Corinth: Oil and Water at the SacredSpring,"Hesperia 61, pp.385-408. Strong,D. E. 1966. Greekand Roman Gold and Silver Plate, London. Talcott, L., and B. Philippaki.1956. Small Objectsfrom the Pnyx. II, Part I, Figured Pottery (Hesperia

Supplement 10), Princeton. Themelis, P., and G. Touratsoglou. 1997. Ot Tcyot rozo As,pv('ov, Athens. Thompson, H. A., and R. L. Scranton. 1943. "Stoasand City Walls on the Pnyx,"Hesperia 12, pp. 269-383. True,M., and K. Hamma, eds. 1994. A PassionforAntiquities:.AncientArt from the Collection of Barbara and Lawrence Fleischman, Malibu. Valavanis, P. 1991. HocvocOEvocxxot( ocyqopEl' oc7ro -rrp EpE'-roloc. cqy,BopRJ xrr-7vocTEpx .

ocyyEtoypocgoc zroo4ov 7r.X oc.,

Athens. Vandiver,P. B., and C. G. Koehler. 1986. "Structure,Processing, Properties,and Style of Corinthian

486

IAN

TransportAmphoras,"in Ceramics and Style, and StyleII: Technology W. D. Kingeryand E. Lense, eds., Columbus,pp. 173-215. Vickers,M. 1986. "Silver,Copper,and Ceramicsin Ancient Athens,"in Pots & Pans.A Colloquiumon PreciousMetalsand Ceramics, M. Vickers,ed., Oxford,pp. 137151. Whitbread, I. K. 1995. Greek Transport Amphorae:APetrologicaland Study,Athens. Archaeological Williams, C. K., II. 1972. "Corinth, 1971: ForumArea,"Hesperia 41, pp. 143-174. . 1973. "Corinth,1972:The ForumArea,"Hesperia 42, pp. 1-32. . 1975. "Corinth,1974: Forum Southwest,"Hesperia 44, pp. 1-29. .1976. "Corinth,1975: Forum Southwest,"Hesperia 45, pp. 99137.

Ian McPhee LA TROBE DEPARTMENT

Bundoora, Australia

UNIVERSITY OF ART

Victoria

HISTORY

3083

[email protected]

MCPHEE

. 1977. "Corinth,1976: Forum Southwest,"Hesperia 46, pp. 40-81. .1978. "Pre-RomanCults in the Area of the Forumof Ancient Corinth"(diss. Universityof Pennsylvania). .1979. "Corinth,1978: Forum Southwest,"Hesperia 48, pp. 105144. . 1980. "CorinthExcavations, 1979,"Hesperia 49, pp. 107-134. .1981. "Corinth:Excavationsof 1980,"Hesperia 50, pp. 1-33. 11991. "Corinth,1990: SoutheastCornerof Temenos E," Hesperia60, pp. 1-40. 1995. "Archaicand Classical in Corintoe l'Occidente Corinth," (AttiTaranto34), pp. 31-45. Younger,W. 1966. Gods,Men, and Wine,London.

HESPERIA

69,

2000

Pages 487-So7

In memory of

A

ROMAN $ U

P PORT

CO

RI

TABLE

AT

ANC{I

EN

T

NTH

ABSTRACT

CharlesM. Edwards(1953-i992) A Roman sculpturedtable support,discoveredin Frankishdestructionlevels at ancient Corinth, representsa rare iconographicaltreatment of a popular theme, Hermes carryingthe baby Dionysos. Dating to the second half of the 2nd century A.C., the trapezophoros is adorned with a muscular,torsional Hermes supportingDionysos perchedon his outstretchedleft arm.The closest parallelsfor this composition are the large three-dimensional groups of satyrsand babiesproducedat Aphrodisias.The Hermes and Dionysos group mayhaveremindedits Frankishdiscoverersof St. Christopher,patronoftravelers and transporterof the infant Christ. Recent excavations at Corinth in the area southeast of Temple E have revealed a Byzantine complex with enclosed court (Unit 1), and a funerary chapel in Unit 2. An unusual Roman sculpture of Hermes holding the baby Dionysos (Fig. 1), discovered in Unit 2 during the 1993 season, represents a hitherto unknown iconographical treatment of this popular and enduring theme, and may also illustrate how this motif, and the work itself, took on new meaning in Frankish Greece.' The sculpturewas discovered during the removal of destruction debris that ran east to west in the tile-paved Room C of Unit 2, which included a series of rooms surrounding the church at the northern end of the Byzantine monastery altered in the Frankish period (Fig. 2:a-b).2 According to C. K. Williams II, Unit 2 1. S-1993-2. I thank C. K. Williams II for his generouspermission to study and publish this sculpture, and for readingdraftsof the paper. N. Bookidis helped in manyways, discussingthe sculpturewith me and readingthe manuscript.In additionto the originalphotographsshe prepared, Dr. Bookidis suppliedadditionalimages on short notice;I owe her a special debt of thanks.Thanks also to Luke Walker, GraphicDesign student at the Universityof Mississippi,who producedFigure 7.

B. S. Ridgwayalso read a draftof this essay,makingvaluablesuggestions.The anonymousrefereesmade many helpful recommendationsand corrections. Remaining mistakesare my own responsibility. 2. Corinth NB 863, pp. 18,24 (June 22-23, 1993). This room is called C in the 12th-centuryByzantinephase (see Williams and Zervos 1995, pp. 11-15); for Room C in its Frankishform (Room 3) see Williams et al. 1998, pp. 237-239; Williams and Zervos 1993, pp. 9-12, and 1994, pp. 24-26.

AI LEEN AJOOTIAN

488

_

_ _

__

_S

_ _

_

| _

,

:

_ _

_

_

__

_

_ _1

_

_

_

_

_

_

; I

i

.

-

-

S r

,

i

1.

Figure _

j

l

K

'

S-1993-2^

i

Corinth __

Hermes and Dionysos

table support

A ROMAN

Figure 2a. Corinth, temenos of Temple E: Frankish remains of the end of the 13th century. From 11-

1 r7

lxr Williams and Zervos 1996, fig. 1 o

TABLE

OF 1965 ~~~~~EXCAVATION

SUPPORT

--

ANCIENT

AT

489

CORINTH

_--

--~~1

Q:V

UNIT3

UNIT_4

0

Figure 2b. Corinth, Unit 2: actualstate plan, composite remains of the late 12th and 13th centuries. From

VI

Williams and Zervos 1996, fig.2

7

6

H

g

\1|

G

(5.j

&

8

9

nt5e

r < 0\tt

rti ,~~~

q

i*

?>

igse

H

2

anUNIT 0

3. Williams et al. 1998, p. 237; Williams, Barnes,and Snyder 1997, p. 21; Witliams and Zervos 1995, pp. 11-12. 4. Williams et al. 1998, p.239.

20M.

5 SM. '' -3g 9

was probably constructed in the first third of the 12th century.3Williams first coupled this destruction stratum in Room C with the Catalan invasion of 1312, but more recently has connected it with the earthquakeof ca. 1300.4 Room C was repairedearly in the 13th century,and went out of use shortly after.During this brief last phase, the north wall of Room C had a door at its center that opened into a courtyard at its north side.

?

490

AILEEN

AJOOTIAN

The excavatorhypothesizes that the fragment was immured in an external wall of Room C until it collapsed into a stratum of architecturaldebris.5 Sculpted of fine-grained white marble, the male, nude except for a chlamys fastened over his right shoulder, supports a baby on his left arm. The fragment has a preserved height of 0.53 m. The height of the adult's head is 0.13 m. Multiplying this dimension by 7, a standardratio of head to body height, we can hypothesize that he stood ca. ninety centimeters tall. Some patches of the highly polished surface are preserved on the left side of the neck and shoulder and in the valleys between folds on the chlamys. The right side of the sculptureis considerablyweathered, suggesting that it may have been exposed for some time to the elements. Its surface considerably damaged, his short-brimmed, close-fitting hat, preserving stumps of wings on both sides of the crown, identifies this personage as the god Hermes.6 His hair, where it emerges beneath the hat, was arranged in short curls, executed with a profusion of drill holes, their finished surfaces now largely broken away. Locks of hair frame his face. These strands can be seen most clearly on the left side of the head where it was protected by the baby's body. Much of Hermes' face from the nose down is broken away. His brow, however, marked with a deep furrow,survives, along with the left eye and the inner half of the right eye. A deep drill line defines the contour of the upper right lid and most of the deeply set left eye. The tearduct is carefully articulated.Pupils and irises are not incised; these details may have been added in paint. The appearance,in photographs, of deliberately cut detail in the center of the left eye is the result of damage. Hermes now lacks his raised right arm from the shoulder, the left from mid-upper arm, and both legs from mid-thigh. The infant is missing his head and neck, and a portion of his torso on the left side, including the arm, part of the left leg to the thigh, and most of the right leg to midthigh. Hermes raised his right arm clear of his head, perhaps above it, holding an attribute toward which the infant reached. The god held his bent left arm away from his body. He may have grasped the baby'sleft leg with his left hand or raised his left forearm toward his charge; there are parallels for both poses, as we shall see. Or he may have brandished a caduceus. The baby sits on Hermes' left shoulder, close to the god's head, dangling both legs down over Hermes' chest: scars remain on the chlamys where they rested. The child stretches to the right, resting his right hand on the front of Hermes' head. The god, turning to his own left side, looks up at him in a close, intimate pose. Hermes, given an athlete'sbody, active and torsional, twists his lower body to his right side. He probably stepped forward on his left leg. His neck and torso are strikingly muscular.The neck preserves the remains of a bulging tendon on the left side, along with horizontal ridges of flesh that intensify the dramatic upturned position of the head. The muscularity of the epigastric region is apparent,despite the broken and abraded surface. Some details of digitation survive, especially on the left side of the torso. The iliac crests are prominently expressed. This sculpture of Hermes as a kourotrophos, probably shouldering the infant Dionyos, was not freestanding but adorned a support. Part of a

5. C. K. Williams II (pers.comm.). 6. LIMC V, 1990, pp. 319-321, s.v. Hermes(G. Siebert).

A

ROMAN

TABLE

SUPPORT

AT

ANCIENT

CORINTH

49I

:~~~~~~~~~~~~~~~~~~ ... ........ ........~~~~~~~~~~~~~~~~. .....: :....... . .

: :.: : ,., ., ...

.........

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

i _I 1|I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ..... .....

S. .5V

Figure 3. Corinth S-437, satyr table support

7. Figure M, Temple of Zeus at

Olympia,470-456 B.C. (Ashmoleand Yalouris1967, figs. 33-36). 8. S-1934 A, B: Sturgeon 1975, pp.292-293, no.3. 9. These table supportswere being producedin Attic workshopsby the

middleof the 1stcenturyA.C. and through the third quarterof the 3rd century.Also on trapezophorasee Moss 1988 and Cohon 1984. 10. Stephanidou-Tiberiou1993, p.261.

pillar, measuring about ten centimeters on each side, carved in one piece with the figure, survives on the back. The sides of the pier were worked with a rasp, and the back was finished with the claw chisel. Now broken off at top and bottom, the support extended above and below the figure, which was carved completely free of this pillar from the middle of the shoulder blades and downward.The plasticallyrenderedchlamys, its edges and bottom now broken away in back, was carved in one piece with the support, but billowed out on both sides, providing a dramatic background for the figure. This garment fell over the back of Hermes' left shoulder, where draperyfolded over on itself; it was carved more carefullyhere than on the right side, but the figure was probablymeant to be viewed from the front. While his eyes are not drilled, Hermes' common snail curl coiffure is characterized by deep drill holes. The use of the drill to articulate short, tightly curled locks is attested by the second quarter of the 5th century B.C., so this feature is not diagnostic for the dating of this sculpture.7In Roman Imperial portraits of women, the combination of drilled hair and blank eyes is found in Augustan through Trajanicworks. It may be risky, however, to use large Imperial portraits as comparanda for small mythological compositions, especially in the case of this piece, unusual in form and function. At Corinth, two nonjoining fragments from the head of a large, freestandingRoman Hermes with the short curly hairstylehave been given an Antonine date.8These pieces preserve not only drill holes, but also comma-like drilled grooves defining individual locks of hair.The coloristic quality of the hair,the defined tearduct, active pose, dramatic musculature,and polished surface may place the Hermes and Dionysos group in the Hadrianic period, and perhaps even later in the second half of the 2nd century. The pillar served a structuralfunction for the sculptureitself, stabilizing the load supported on Hermes' neck. Square trapezophoros supports with roughly consistent dimensions from top to bottom are less common than struts which taper toward the top.9 According to StephanidouTiberiou, the thicker,squarepillarwas more stable.'0Parallelsfor the square support can be found among Attic figural trapezophora, including two at Corinth. The upper torso of a satyr (Fig. 3) excavated at the site preserves

AILEEN

492

AJOOTIAN

:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ..g:.... |E _ il ...... g

I

.

._

_

_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I_.: '. ::: l ............ I .,.s ........... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~z:

....:......*

part of a similar square strut with the same measurement of 0.10 m on a side." The figure, however, has a preserved height from neck to waist of aroundthirteen centimeters,versusca. twenty centimetersfor the Hermes.'2 Another fragment of a possible support at Corinth (Fig. 4) preserves most of the head of a bearded male, probably Herakles, with short, drilled curls. His head, like that of Hermes, measures 0.13 m and was carved in a single piece with the top of a substantialstrut,similarlyproportioned,whose capital bears a round cutting for the tenon of the table leaf, or whatever element it supported.'3This work corresponds in the dimensions of both figure and support to the Hermes and Dionysos group, but the statue, a herm, is centered frontally on the pillar, and probably adhered to it from the head down. In Athens at the National Museum, a frontal Pan wrapped in himation decorates a similar squared shaft, measuring 0.10 m per side (Fig. 5).34 Figures partially or wholly released from their pillars are uncommon. Moss compiled a small group that includes a satyr, a young male resting against a herm, a female of unknown identity, and images of Erotes carrying birds.'5Other types can be added to this assemblage. Figures of Dionysos alone or with satyrs and Pans customarily are carved nearly in the round. Some "Hanging Marsyas" figures adorning table supports were carved almost free of the tree trunk support that was also an essential narrativeingredientof the scene.16 It may be that the larger,freestandingprototypes for these figural supports encouraged a three-dimensional quality in their presentation. More complex group compositions include struts supporting individual sculptures in which the pillar itself more clearly functions as the support for the table leaf. One parallelfor the nearlythree-

-

Figure 4 (left). Corinth S-2329, Herakles table support Figure 5 (center,right). Athens, National Museum NM 251, Pan tablesupport.Photograph by the author

11. S-437: Stephanidou-Tiberiou 1993, no. 71, and see Appendix, below. 12. Another fragmentary trapezophorosin Athens has a comparablesupport,but the figure,a Ganymede,is smaller(NM 5730; Stephanidou-Tiberiou1993, p. 281, no. 132). 13. S-2329: unpublished. 14. Athens, NM 251: StephanidouTiberiou 1993, p. 261, no. 80, pl. 42. 15. Moss 1988, p. 16, note 8. 16. Weis 1992, nos. 12, 13, 18,20, 45,46,56,57.

A ROMAN

E...:

SUPPORT

AT

ANCIENT

'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Y

>

t ' '

TABLE

.

w

_l~

. _l

Figure 6 (left). Rome, Vatican 2445,

Ganymedeand eagletablesupport. After Ridgway1997, pl. 58

Figure7 (right).Reconstructionof CorinthS-1993-2, Hermesand Dionysos tablesupport

17. Vatican2445: Ridgway 1997, p. 247, pl. 58; Moss 1988, p. 390, no. A5. I thank B. Ridgway for pointing out this parallel.In addition to the Vaticanexample,a more compressedAttic version also presents the youth in a pose similarto that of Hermes, looking up at the eagle (Stephanidou-Tiberiou1993, no. 136). More commonly on trapezophora, Ganymede leans languidlyagainst the supportwith head lowered (Stephanidou-Tiberiou1993, pp. 125131). 18. Trapezophoramay measure as high as 1.275 m or even taller (Stephanidou-Tiberiou1993, p. 42, no. 100).

_

~

CORINTH

493

Yvvupw

~~~~1.

dimensionalCorinthHermesis the torsionalfigureof Ganymedeon an Antoninetrapezophorosin Rome.The youth,carvedvirtuallyfreeof the strut,reachesup towardZeus,who is in the guiseof an eagle,spreadingits wings abovehim (Fig. 6).'7 While the Ganymedetrapezophorospreservesits plinth,we do not knowthe shapeandheightof thebasethatoriginallysupportedthe Corinth work,nor can we be certainhow much fartherthe pillarextendedabove the figures'heads.The top of the support,with its capital,had to clear Hermes'raisedrightarm,held freeof the head.It mayhaverisenas much as anotherfifteento twentycentimeters.If one restoresHermes'height at aroundninety centimeters,then the completework,withoutbase,could havemeasuredone meterandten to twentycentimetersin height.18 These dimensionsconformto thoseof the well-preservedPansupportin Athens (NM 251), with its well-preservedcapitaland base,carvedin one piece with the figureand pillar(Fig. 5). Pan in this supportmeasures0.91 m, andthe totalheightof the supportis 1.12 m. Forthe HermesandDionysos supportat Corinthwe can tentativelyrestorea capitaland a base to the pillar,all carvedfromone piece of marble.The well-preservedtop of the fragmentarypossibleHerakles(Fig. 4), mentionedabove,consistsof an extensionof the pillar,finishedin frontwith a simple cavettomolding. Perhapsthe Hermes and Dionysos supporthad a similarcrowningelement (Fig. 7). What this figuralpillarmay have supportedis uncertain.A marble tableleaf is the most likelyburden,but anotherclassof figuredsupports, for balustrades,shouldbe mentionedbrieflyhere.Hermswith male and femaleheadsadornedthe stagefrontof the Odeion of Agrippain its first

494

AILEEN

AJOOTIAN

phase.'9The herm shafts, 0.92 m high, measure 0.19 x 0.16 m. A Roman example recovered from Omega House in Athens, on the north slope of the Areopagus, takes the form of a bearded Silenos or Dionysos (?) herm wearing a full-length, long-sleeved garment.20Camp has reconstructed the room, which contains a horseshoe-shaped swimming pool, with its west wall extending up ca. 1.50 m, surmounted by a balustrade.People on the same story as the balustrade could have looked down into the pool below. The figure, ca. 1.09 m high with its tapered shaft, supported the railing,which kept viewers from falling into the pool. He inclines his head, as if also peering down into the pool. It is just possible that the fragmentary Herakles pillar and fragmentary satyr support mentioned above, together with the Hermes and Dionysos group discussed here, supported a railing or balustrade.In any case, there are no surviving parallels for such an installation. If it adorned a trapezophoros,the Hermes and Dionysos composition is unusual in that respect. Cohon, Moss, and Stephanidou-Tiberiou provide no comparandain the surviving assemblage of Roman trapezophora for the theme of Hermes holding the infant wine god. In fact, Hermes rarely adorned table supports, and Dionysos only infrequently appeared on them depicted as a child, although the adult god occurred in various poses, alone or accompanied by members of his thiasos.2' There are numerous examples with the drunken adult god supported by an adolescent satyr.2 The composition at Corinth appears to be rare in the corpus of furniture supports and among sculpted images of Hermes and the infant Dionysos. While the motif of Hermes Kourotrophosis familiar in Roman sculpture,parallelsfor the Corinth work, with Hermes in a torsional pose looking up at the baby seated on his shoulder, are rare, as a survey of this theme in vase painting and sculpture reveals.23In addition, sculptural comparanda for the baby's pose, with both legs resting against Hermes' chest, are scarce.24 Relatively few Classical images of Hermes carryingDionysos survive. While there are no depictions of Hermes bearing Dionysos on the reverses of Classical coins, the adult god appears with Arkas, eponymous hero of Arcadia, on reverses of silver tetradrachms minted by the city of Pheneus, in northeastern Arcadia. These tetradrachms, minted between 370 and 300 B.C., show Hermes carrying Arkas, who is identified by inscription.The child perches on Hermes' outstretched left forearm.Hermes grasps a caduceus in his right hand as he runs to his right, looking back at the child, who raises his right hand to the god's head.25 19. Thompson 1950, pp. 64-68, pis. 46, 48-49. 20. Camp 1989, p. 52, pl. 9:5, p1. 10:10.

21. For trapezophorawith Dionysos and a young satyrsee StephanidouTiberiou 1993, pp. 86-96. A Flavian trapezophorosfrom Pompeii takes the form of a "Lysippan"Silenos holding a baby.The silen'sbody grows out of acanthus leaves, and the supportitself

terminatesin a feline foot: Vermeule 1981, fig. 11. 22. Stephanidou-Tiberiou1993, pp. 96-98. On this motif in Roman sculpturesee Pochmarski1990. 23. On the myth see Huskinson 1996, pp. 30-31; Motte 1996; Beaumont 1998, pp. 71-76; Beaumont 1995, pp. 341-344; LIMC V, 1990, pp. 285387, s.v.Hermes(G. Siebert);Baratte 1989; Hadzisteliou-Price1978, pp. 70-

72; Loeb 1979, pp. 28-56; Zanker 1965, pp. 45-55; Greifenhagen1931; Heydemann 1885. 24. For the motif on a late 5th-early 4th-century B.C. Attic hydria,see CVA, Berkeley1 [USA 18], pp. 48-49, pls. xlvii-xlix. 25. Gardner[1883] 1981, p. 193, no. 13, pl. xxxvi:7.P.

A ROMAN

TABLE

SUPPORT

AT

ANCIENT

CORINTH

495

In ancient sculpture, at least two Classical traditions of Hermes Dionysophoros can be identified: 1. Narrativesin reliefsillustratingthe birth of Dionysos from Zeus' thigh, or Hermes deliveringthe newbornto the Nymphs. Hermes, often wearingtunic and chlamys,in an activepose, moves toward the Nymphs.26One of the earliestsurvivingsculpturedimages of Hermes Dionysophorosappearson a late-4th-century B.C. votive relief in Athens.27The relief's cave borderframesa large cast of divinities,with Hermes, dressedfor travel,in the center.He steps up on a rock,passing the bundled infant to a waiting Nymph. 2. Large,three-dimensionalsculpturesof Hermes holding the baby Dionysos. No Classicalexamplesof this kind survive,but numerous Roman works, discussedbelow, arethought to reflect earlier traditions.Greek kourotrophoiin the round appearto have become popularby the 4th centuryB.C. The Elder Kephisodotos'statue of Eirene and Ploutos had been set up in the Athenian Agora probablyby 369 B.C. One of the earliestlarge,three-dimensional images of Hermes and Dionysos may have been the bronze also by This lost work Kephisodotos,possiblythe father of Praxiteles.28 suggeststo some critics that there may have been family interestin the theme.29 Pausanias (5.17.7) is the only ancient authority who mentions the Hermes and Dionysos group at Olympia. He attributed it to Praxiteles, the famous 4th-century Athenian sculptor,but the evidence for this association continues to be disputed.30Basically frontal, Hermes extends his body laterally in space. Resting his weight on the right leg, he leans to his

26. The decorationof the lost Throne of Apollo at Amyclae (ca. mid6th centuryB.c.) by the Ionian sculptor Bathyklesmay have included one of the earliestsculpturedrepresentationsof Hermes holding the baby in an isolated vignette (Faustoferri1991). According to Pausanias(3.18.11-12), the mythological scenes on the Throne included Ino and Semele with babyDionysos, and Hermes carryingthe infant to Olympos, presumablyafter Semele's death, to be gestated in Zeus' thigh. But see Pipili 1991. A fragmentaryplaster cast in Munich, said to have been taken from the neck of a metal rhyton tentatively dated ca. 5th/early 4th century B.C., may reflectthe earliestsurviving sculpturedevidencefor the delivery theme, albeit in miniaturerelief (Froning 1981, p. 54, pl. 6; Richter 1960). 27. Giintner 1994, pp. 113-114;

Edwards1985, pp. 419-437, no. 14. Dedicated by Neoptolemos from the Attic deme Melite, the reliefwas found reusedin Omega House on the south slope of the Areopagus. 28. Pliny,NH 34.87. 29. A Roman type has been connected with the Hermes by Kephisodotos I (Rizzo 1932, pp. 7-10; Blanco 1957, pp. 40-41, no. 39E) but see AgoraXI, pp. 163-165. For a similar image on Roman coins of Pautaliain Thrace (Caracalla)and Anchialos on the Danube, see Head [1877] 1979, pp. 232, 235. Pausanias(3.11.11) reports on a Hermes Agoraios Dionysophoros at Sparta,but does not providedescription, attribution,or date. 30. Ridgway1997, pp. 261-263; Corso 1996; Ajootian 1996, pp. 103110;Todisco 1993, pp. 65-79; Stewart 1990, pp. 77-81; Ridgway1990, pp. 90, 93.

496

AILEEN

AJOOTIAN

left, holding the baby on his left arm,with the help of a tree trunk support. While this work has established-for modern critics-the standardsculptured image of Hermes and Dionysos, its source and date are problematic, and in fact there are few Roman copies.31 Generally in Roman marble sculpture, a frontal Hermes looks down at the baby he supports, usually on the left arm, although sometimes the pose is reversed.32The god is often partly draped in a chlamys, as in the much restored example in the Boboli Gardens, where he holds the baby away from his body on his own right side.33There are at least two large torsos, one from Pergamon, the other now in the Bardo Museum (Tunis), where the child was held in front, close to the chest.34A Hermes from the Hadrianic Baths at Leptis Magna is a variant of the Jason/Sandalbinder type. He rests his right leg on top of a large turtle. The infant, of which only fragments survive, sat on the upraised knee.35In another unusual Roman composition, a seated Hermes holds the baby in his lap.36 The composition at Corinth does not resemble canonical large-scale frontal Roman sculptures of the pair. Closer parallels for this sculpture exist in the more torsional groups in the round with young satyrs carrying infants. One of the earliest examples, a terracottafigurine (2nd-lst centuries B.C.) from the necropolis at Myrina, stands in a torsional pose that might be a dance step (Fig. 8).37Turning his head to his right, he looks up at the child sitting on his right arm. There are also convincing parallels in large-scale, three-dimensional marble.Two examples in different scales have been recoveredfrom a sculptor'sworkshop at Aphrodisias. A third, related, piece, signed by the sculptor Flavius Zeno, was one of several sculptures discovered together on the Palatine in Rome.38In all three, the satyrsupportsthe baby,who sits astride his left shoulder, and steadies him by grasping the child's left hand in his

31. SeveralLate Hellenistic works correspondingto the Hermes in pose and musculaturehave been recovered, so it is possible that a stock body type was reworkedfor differentcompositions (Ridgway1984, p. 85). There are a few works that might be associated with the group at Olympia:a fragmentary infant in Corfu (see ES 1325) correspondsin pose and the arrangement of draperyto the Olympia Dionysos; the Roman torso of a satyr with tracesof a child supportedon his arm,in Athens (NM 4800), resembles the pose of the Hermes at Olympia. This satyrand babygroup may in fact be importantto our understandingof the famous piece since a Pompeianwall painting of a satyrholding Dionysos is the usual sourcefor the restorationof the lost right arm and attributeof the Olympia Hermes (Schefold 1957, pp. 124-125).

32. A bronze figurineof Hermes in the Louvrehas been linked with large marbleversions (Waldstein 1882). 33. Zanker 1967, p. 40, no. 40. 34. Ridgway1984, p. 49, note 39. 35. Ridgway1997, p. 307; Bianchi Bandinelli 1966, pp. 99-100, fig. 155. For the Sandalbindertype see Moreno 1995; Ridgway1990, pp. 81-82; Ridgway1984, pp. 88, 94; Inan 1975, pp. 92-95, pl. 42:1-2; Ridgway1964. If, as Ridgwayhas suggested,the type is actuallya Late Hellenistic creation, predatedby a varietyof similarimages in relief at least as earlyas the Parthenonfrieze, then its adaptation for an image of Hermes Kourotrophos supportsthe view that types of the god in this role arelate developments. 36. Ridgway1997, pp. 260-261; Gualandi 1976, pp. 141-142, no. 110, fig. 147. For a similargroup with Pan supportingan infant on his lap:from

y1M

Figure 8. Paris, Louvre, Myr 185, Satyr and baby Dionysos, terracotta group. After Bieber 1955, fig. 570

Nimes, Maison Carree,Naples MusNaz 155747: Marquardt1995, pp. 78-79. 37. Louvre,Myr 185: MollardBesques 1963, I, p. 80, pl. 96:b. A traditionof satyrfamilylife in Classical Greek vase painting precedessculptured groups of satyrsand children.The earliestexampleis the well known image of a satyrfathersupportinga baby on his shoulders(Flying Angel Painter'skrater in Baltimore,WaltersArt Gallery 98.882; Beazley 1918, p. 59, no. 4, fig. 37). The piggybackmotif also occurred in Imperialsculpture,for examplein the alimentapanel,Arch of Trajanin Benevento,A.D. 114-118. 38. On the satyrsat Aphrodisiassee Smith 1998, pp. 255-259; Smith 1996, pp. 60-63; Rockwell 1991, p. 130; Erim 1974. On the Palatinestatue,now in Copenhagen, see Moltesen 1990, pp. 139,145-146.

A ROMAN

TABLE

SUPPORT

AT

ANCIENT

CORINTH

497

own; both hands are now broken away in the work from Aphrodisias illustrated here (Fig. 9). Poised on tiptoe, the satyr steps forward on his right foot. He holds a lagobolon in his right hand, resting it against his right shoulder.39 R. R. R. Smith has recently reevaluatedthe problematic dating of the workshop finds. Epigraphical evidence supports a date for Flavius Zeno within the second quarter of the 4th century A.C.,although it is possible that some items in the workshop assemblagecould be earlier,finished pieces reworked by fledgling sculptors.40The Aphrodisias groups have been connected with a lost Hellenistic original of the 3rd to 2nd century B.C., but apart from these three examples, no other large-scale replicas of this type are known.4' Another vanished Hellenistic work is thought to have inspired at least five Roman copies of a different Satyr and Dionysos group, including a much restoredversion in the Villa Albani (Fig. 10).42 The child, a toddler, rides piggyback on the satyr'sshoulders. Dionysos leans over to his right and the satyr looks up toward him, grasping the child's right ankle in his right hand. He raises his left hand toward the boy. The satyr rests his weight on the right side with the left foot drawn well back. These sculptures express the close, playful relationship between child and adult. But while both the Aphrodisias groups (cf. Fig. 9) and this type resemble the Corinth piece, they do not correspond to it in all respects, since the child at Corinth appearsyounger and does not straddle Hermes' upper arm or shoulders. These three-dimensional groups have counterparts in relief on Roman Dionysiac sarcophagi. Matz recognized at least six types, occurring first on sarcophagi dated to just before the mid-2nd century A.C.43 In all these scenes, the satyrs are members of crowded, complex thiasoi. Some types closely resemble the large sculptures in the round discussed here. The three-dimensional Villa Albani Satyr with Dionysos as a toddler (Fig. 10), for example, corresponds to Matz's KindertragerA (Type 75, Fig. 11). This piggyback group is the rarest of the relief types, occurring

39. While generallysimilarin pose and in their sinewy,muscularbodies and unkempthair,some variations occur among these three sculptures, aside from differencesin scale. In both works found at Aphrodisiasthe baby sits on a goatskindrapedover the satyr'sleft arm,but this pelt is lacking in the Copenhagenexample.In the smallerof the two replicasat Aphrodisiasthe infant leans back;the other one sits straightup. 40. Smith 1998, pp. 258-259. Erim suggestedthat the unsigned groups found at Aphrodisiasmight have been sculptedby a follower of Zeno (Rouech6and Erim 1982). On stylistic grounds,other worksbearing

this artist'ssignaturehad been consideredHadrianicor Antonine (Erim 1974, p. 772). Erim, while favoringa 4th-century date for the artist,proposedan alternatedating for pieces bearinghis signature.He suggestedthat they might be earlier and suppliedto a sculpturesrevworked 4th-century clientele.According to Rockwell (1991, p. 130), both satyr groupsfound at the Aphrodisian workshopmay have been finished pieces undergoingrepairs,so they might have been in the workshopfor a long time. 41. Erim 1974, pp. 773-774. 42. Villa Albani 148: Bieber 1955, p. 139; Helbig4II, p. 247, no. 384. On

the type see Kell 1988, pp. 40-43; Schneider1990, pp. 288-298, no. 232, pls. 202-203. For a list of replicassee Gercke 1969; also Minto 1913. The best preservedreplicais a statuette,the only figurepreservingHermes'legs. The baby holds grapesin his raised right hand, his human earsconfirming his identity as Dionysos ratherthan a baby satyr(Andren 1964, pl. 25). For other baby/satyrgroupssee Paribeni 1959, pp. 118-119, no. 335, pl. 156; also the fragmentarysatyrand baby sculpturein Baltimore:Reeder 1988, p.27. 43. Matz 1968, nos. TH 75-TH 80, pp. 40-50.

A ROMAN

THi75

TH75 Figure 11. After Matz 1968

44. Matz 1968, pp. 48-49. 45. Matz 1968, p. 49, no. 76. 46. Matz 1968, pp. 255-256, no. 115. 47. Matz 1968, pp. 48-50.

TH L76

TH 76

TABLE

-

TH 76

SUPPORT

-

AT

ANCIENT

!

.

TH 76

499

CORINTH

7

TH 77

just a few times.44Matz sees the terracotta dancing Satyr and Baby in the Louvre (Fig. 8), along with a small group of Roman wall paintings and the Villa Albani satyr type, as ancestors of this motif in Dionysiac funerary art. The Aphrodisias groups (cf Fig. 9) resemble Matz's KindertragerC (Type 77, Fig. 11). Some seven instances of this motif survive on sarcophagi, some with the pose reversed. None of the sarcophagus reliefs is dated earlierthan the early Severan period, but they may still be considerably earlierthan the Aphrodisias statues.The Corinth sculpture,with few comparanda in the round, is close to KindertragerB (Type 76, Fig. 11), with pose reversed, because of the baby's frontal position, with both legs over the satyr'schest.45The earliest occurrenceof the relief image is on an early Antonine sarcophagus in Florence.46Most of the other examples date from later in the 2nd century,possible contemporaries of the Corinth sculpture. Matz' analysis of satyr types on Roman sarcophagi reveals that poses remain the same, even when other burdens are substituted for babiesthyrsoi, askoi, amphoras, sacrificial animals. He proposes that in all the cases considered here, the small relief versions on sarcophagiwere inspired by earlier sculptures in the round.47Some important changes in meaning do occur,however, in the shift from three to two dimensions, if in fact this is the direction of influence. In the round, satyrs carrythe baby Dionysos, depicted as a human child. In relief, they carry infants of their own race, distinguished by pointed ears, and sometimes tails. The identity of both personages in the Corinth group varies from those in the relief versions. A repertoire of poses and personas apparentlywas manipulated to suit various genres and themes in Roman sculpture. As we have seen, the Corinthian group of Hermes and Dionysos is unusuallycompressed,with the babyvery close to Hermes and raised above him. This arrangement may have been important for the legibility of the composition, as well as for its stability.The substantialpier behind Hermes supports his neck and the child resting against him. This pose, suggesting an intimate physical and emotional relationbetween the two figures, found more commonly in satyr/baby groups, hieremay also serve a structural purpose. From the Classical period on, the image of Hermes carryingDionysos represented one stage of a venerable sacred narrativesequence that docu-

500

AILEEN

AJOOTIAN

mented the wine god's prematuredeliveryfrom the womb of dying Semele, his rebirth from Zeus' thigh, his subsequent foster homes and parents.48It is likely that whatever its original setting, the Corinth sculpture expressed to its ancient creator and viewers a precise message appropriatein a specific context at the ancient city. Several other Roman sculpturesof Hermes have been recovered from the site; various types and poses can be identified.49While no other sculptured images of Hermes Kourotrophos have been recognized at Corinth, Trajanicbronze reversetypes of Hermes with Dionysos were minted for the city.50Running to his left, Hermes supports a baby on his outstretched left arm, which he holds so far away from his body that the infant does not touch his chest. The more usual relationship of adult to infant appearson Corinthian reverses,with Hermes' head higher than that of his charge.The adult wears petasos and chlamys and holds a caduccus in his right hand. As Sturgeon and Fuchs have shown, several Roman sculptures of Hermes Dionysophoros, in relief and in the round, come from theaters, an appropriate setting for such images.5"From the Theater of Dionysos in Athens, four surviving Hadrianic or early Antonine panels, which may originally have decorated the scacnae frons, include one with a frontal, chlamys-draped Hermes holding Dionysos on his left side. This pair and a seated Zeus are flanked by nude males carryingshields and identified by modern critics as the Korybantes.52Hermes in this Athens relief resembles Roman three-dimensional sculpturesof the god and not the more heavily draped, active figure of earlier narrative reliefs. Sturgeon connects this imagery with scenes on Dionysiac sarcophagi,altars,and other reliefs, suggesting that these scenes might have influenced the theater program.53 This theme apparentlywas more popular in the decoration of eastern theaters;at least four stage fronts, at Perge (ca. A.D. 120), Side, Sabratha,and Hierapolis, included birth of Dionysos cycles with images of Hermes and baby Dionysos.54At Corinth's theater,this motif does not appear.Hermes may have been a participant in the Gigantomachy frieze adorning the scaenae frons, but he does not survive as a kourotrophos.55 Sculpturesof Hermes and Dionysos also adorned Roman baths, occasionally the frigidarium. We have already considered the unusual Jason/ 48. See Hutchinson 1991, p. 224; also pp. 494-495 and note 23 above. 49. S-686, S-705, S-718, S-1934, S-3585. 50. Imhoof-Blumer and Gardner 1964, p. 21, pl. E:LXXXVIII. 51. Sturgeon 1977. On Hermes Dionysophorosin theaters:Fuchs 1987, pp. 37, 133, 191. These works include a small-scale(H. 0.975 m) sculpturein the round from Minturno (Naples, Fuchs 1987, p. 37, no. ElI1). From the Theater of Dionysos in Athens see the large Silenos in hairysuit carryingDionysos,who holds a mask (NM 257, Karouzou1968, p. 98; LIMC III, 1986, p. 480, no. 687, s.v.Dionysos[C. Gasparri]).The toddler Dionysos wears

long-sleeved costume and leggings, tunic, mantle, and boots. "Silenos"could be an actorin costume. Gasparri connects this Roman work with Pliny's reference(NH 36.29) to a statue in the PorticusOctaviae.For a similargroup of Hairy Silenos and Dionysos, see Marcade1970, p. 189, no. A4143, pl. 23. 52. Sturgeon 1977, pp. 34-35. 53. Sturgeon 1977, p. 36. The other panels reusedin the so-called base of Phaedruspresent subsequentphases of Dionysos' career,and Sturgeon (1977, pp. 51-52) proposedthat other pieces, now missing, included other episodes from his childhood. 54. In all these scenes, Hermes is much more active than he is in the

Athens frieze. Hermes Dionysophoros at Perge (Fuchs 1987, p. 133; Bernardi Ferrero1966-1974, III, p. 152); at Side (BernardiFerrero1966-1974, II, p. 141; Fuchs 1987, p. 133); at Sabratha (Fuchs 1987, p. 133; Caputo 1959, pl. 38, fig. 66); at Hierapolis (Fuchs 1987, p. 133; BernardiFerrero19661974,1, p. 59, fig. 97). 55. CorinthIX, ii, pp. 14,22, nos. G 7-1, G 7-2. The torso of a large, three-dimensionalstatue of Hermes with the remainsof a caduceusadhering to his left armwas recoverednear the theater.The fragmentarypurse mentioned by Sturgeon(CorinthIX, ii, p. 22, no. G 7-1) might belong with this figure.

A ROMAN

TABLE

SUPPORT

AT

ANCIENT

CORINTH

50I

Sandalbindertype from the bath at Leptis Magna.56Possibly these sculptures in public baths alluded to a subsequent episode in Dionysos' childhood, when he is bathed by the Nymphs, the sculpturesevoking this mythic lustral episode in an actualbath complex.57There are several Roman baths at ancient Corinth. A probable site for our sculpture may be the area tentatively identified as a bath in its Roman phase, close to the Frankish complex in which the work was found. This bath lies on the western side of the excavated Frankish remains south of Temple E, some fifty meters southwest of where the sculpture was discovered. Preliminary investigations of this areareveal an installation whose first phase was possibly Augustan; it apparentlysurvived into the 5th or 6th century A.C. A complete 1lth- or 12th-century rebuilding disturbed most of the earlierwalls.58A fragment of sculpture, possibly a nonjoining portion from one of Hermes' limbs, was found in this area.59This might be a section from an upper arm or thigh and it has a preserved length of 8 cm and a diameter of 9 cm. It is important to keep in mind that our sculptureprobably adorned a table support, with Attic parallels in terms of pillar and figure size, even though there are few surviving thematic comparanda. Evidence for trapezophora from Delos, Pompeii, and other sites, in addition to literary testimonia, demonstrates that Roman stone tables with figural supports served a variety of functions, secular and religious.60In temples, shrines, and household lararia,such tables were used as altars, offering tables, and supports for portable altars,patera, or lamps.6' In domestic contexts they served as stands for statues of lares. Examples from Pompeii and Herculaneum are thought to have displayed bronze figurines in the atria or peristyles of villas.62While monopodia, tables with a single central support, were not used for dining, wall paintings depict them laden with glassware and vessels and other banqueting supplies.63 The western end of the Roman Forum at Corinth, in the general area where this sculpture was found, is dominated by sacred buildings and precincts.The Hermes and Dionysos support could have been part of the sacred furniture in the precinct of Temple E, over which the ByzantineFrankishcomplex had been built. Much sculpturehas been recoveredfrom this area,including at least one Roman table support.64Or, the piece might belong to the unusualbuilding with centralhemicycle.This building, which terminates the west end of the Central Shops, was originally identified as a Temple of Dionysos. In his revised interpretation of Pausanias' route through the Roman Forum, Williams has identified this complex as the 56. Also from baths,primarilyin North Africa:Manderscheid1982, no. 116H (Agnano, statuette,Trajanic/ Hadrianic);no. 272 (Shahat,Hadrianic/ EarlyAntonine); no. 304 (Leptis Magna, Hadrianic).At Corinth, the fragmentaryhead of a large Polykleitan Hermes was found in Room 1, south pool of the Lechaion Road Bath (CorinthXVII, p. 41; Sturgeon 1975, pp. 290-292, no. 2). In addition,Biers (in CorinthXVII, p. 48) reportsfragments of a ram,possibly an attribute

from a statue of Hermes, from Rooms 3 and 5. 57. For bathing scenes on Dionysiac sarcophagisee Matz 1968. 58. The building is securely identified as a bath in the Byzantine period;much less can be said of the earlierremains,although segments of flooring built of waterproofherringbone tiles suggest the presenceof an earlybath here, accordingto C. K. Williams II (pers.comm.); and see Williams and Zervos 1995, p. 11;

Williams, Barnes,and Snyder1997, pp. 37-40. 59. S-1994-3. 60. Cohon 1984, pp. 6-7; Moss 1988, pp. 241-292; StephanidouTiberiou 1993, pp. 58-73. 61. Moss 1988, pp. 279-280. 62. Moss 1988, pp. 279-280. 63. Moss 1988, p. 274. 64. For trapezophorafound at Corinth, see Appendix,below.

502

AILEEN

AJOOTIAN

Hermes sanctuarymentioned by the traveler.65 The foundations, probably from an altar,ca. 2 meters to the north and on axis with the entrance of this structure, support its identification as a temple. The central room, with its multicolored marble flooring, appearsto have had severalbases, or perhaps tables, along the edge of the interior.66 As we have seen, its secondary context, possibly a Frankish wall in Room C, one of a suite of rooms west of the church, has been connected with a destruction of ca. A.D. 1300. And as observed above, Hermes' body is noticeably more weathered on his right side than on the left. If the statue had been embedded in the exterior face of a wall (perhaps the predestruction north wall of Room C) with the right side projecting, such a placement would account for this uneven weathering.Just possibly, as spoIla immured in the outer wall of a room associatedwith the funerarychapel of a Frankish religious complex, the statue had a new life not simply as building material, but as sculpturewhose pagan iconography was reinterpreted with a Christian message.67A secondary, Christian, function for the Hermes and Dionysos trapezophoros will be explored briefly here. The statue's intimate juxtaposition of grown-up and child may have evoked, for Frankish Christians at Corinth, images from Christ's life depicted in contemporary church art. In the Byzantine world, by the 8th century, some episodes from the pagan cycle of Dionysos' birth became models for the childhood of Christ. The pose of the reclining Virgin Mary echoed that of recumbent, dying Semele,68and the bathing of the Christ Child, at least once employed as a metaphor for his later baptism, drew on earlier scenes of Dionysos bathed by his nymph nurses.69The question of how pagan sculpturewas used, perceived, and interpreted by both Byzantine and Western Christians has received considerable recent attention. Whether pagan statues, inscriptions, or even architecturalelements embedded in Christian churches had specific religious meanings for patrons, designers, and congregations continues to be debated.70Mango has suggested that building blocks, as well as sculptures,may have been perceived as having apotropaic power, and served specific functions when built into a church facade. Saradi-Mendelovici has recently examined the earlierambivalent Byzantine view of Classical and Roman sculpture, in which this sculpture was perceived both to be a haven for demons and to have artistic worth.71 Hermes Dionysophoros in the Dionysos birth cycle may have inspired the later figure of St. Christopher,who unwittingly carriedthe infant Christ across a river.The 8th-century Passionof Christopherplaces him in Syria; his martyrdommay have occurredca. A.D. 250.72 A mid-5th-century Greek dedicatory inscription from a church in Bithynia provides the earliest evidence for the saint's cult.73Jacobus de Voragine, in The GoldenLegend, recounts the best-known features of his story.74A giant first named Reprobus,he felt his destiny was to serve the greatest king on earth, whom he eventually determined to be Christ. Christopher carriedpeople across a river at a point where there was no bridge. One of his burdens was Christ, disguised as a mortal child. In the West, St. Christopher, one of the Fourteen Holy Helpers, was a medieval patron of travelers;anyone who glimpsed his image would be safe from death that day.75Large murals of the saint bearing the baby Christ commonly decorated the north walls of English and northern

65. Williams and Zervos 1990, pp. 351-356; Williams 1989. 66. CorinthI, iii, pp. 89-90. 67. See Engemann 1998, esp. pp. 173-177 for a discussionof the possible Byzantinereceptionof a Roman Dionysos and Satyrtable supportfound in the oven of a pottery kiln dating to ca. the 9th centuryA.C. at Abu Mina in Egypt. 68.Juhel 1991; Kitzenger 1963, pp. 100-105; Weitzmann 1960, pp. 5253. 69. Kitzenger 1963; Nordhagen 1961. 70. On Byzantinereceptionof Greek and Roman sculptureand other artworkssee Engemann 1998; Alchermes 1994; Saradi-Mendelovici 1990; Brenk 1987; Mango 1963. 71. Saradi-Mendelovici1990, p. 47. 72. Kaftal 1965, p. 24; Drake and Drake 1916, p. 26; Jameson 1894,11, pp. 439-450; ReallexikonfiirAntikeund ChristentumII, cols. 1241-1250, s.v. Christophoros (A. Hermann). 73. Benker 1975, p. 8. 74. de Voragine1948, pp. 377-382. 75. Whaite 1929, pp. 8-9.

TABLE

A ROMAN

SUPPORT

AT

ANCIENT

CORINTH

503

4~~~~~~~~~~~~~4

Zan~~ipolo), jVenie Giovnni Bellin Figure 12. St. Chri-topher. Detai

p.

1 9 6.>

.0 ,

Detail Figure12.St.Christopher.

!

r0'iin:

b

e

;':

. ................ s.N

.

?

S,>,.."'.... s w\', X > et t t'S. g' ................................................... ' 'S,,'.,. aSSw;

'

SantiGiovanniePaolo(SanU|N4; 1999, (1430-1516). AfterTemnpestini

76. Farmer1987, pp. 868-887; Whaite 1929, pp. 8-9. 77. Benker 1975, pp. 46-47. The cynocephalicChristopherof the Easternchurchappears,infrequently,as kourotrophos,beginning in the 14th century(Loeschcke 1965, pp. 54-55, pl. 85:3). 78. On Greek monasteriesas pilgrims'shelterssee Ciggaar 1996, chapter1.

. ia

oS,!,L

European churches, opposite the entrance, and could be viewed from outside by passersby.76These wall paintings first appear in the 13th century, although the earliest images of St. Christopher carrying the infant Christ date to the beginning of the 12th.77Medieval and later images of St. Christopher with the infant Christ often echo the intimate pose of Hermes and Dionysos at Corinth, where the adult, supporting the baby on his left side, looks up at him, their heads close together (Fig. 12). Just possibly, the medieval discoverers of the trapezophoros saw a connection between the pagan group and the Christian travelers'saint. They may have considered their complex at Corinth, possibly a pilgrims' refuge, an appropriate setting for this unusual Roman sculpture.78For its 13th-century Frankish viewers, the pagan composition of Hermes holding baby Dionysos at ancient Corinth may have resonated with meaning that gave it a second life in the Christian world.

504

AILEEN

AJOOTIAN

APPEN DIX ROM/\AN TRAPEZOPHORA AT ANCIENT CORINTH Excavations at Corinth have produced many Roman trapezophora.They are listed below by theme, with their provenience, reference to StephanidouTiberiou 1993, and Corinth Excavation notebook (NB) entry,when available. Aphrodite S-3745 Attis S-1962 Dionysos S-731 S-1462 S-1527 S-1600 S-2470 S-2532 S-71-32 Eros S-2830 Ganymede S-56 S-2729

Lechaion Road; Stephanidou-Tiberiou 1993, no. 164 North of School; Stephanidou-Tiberiou 1993, p. 58, note 65

Asklepieion; Stephanidou-Tiberiou 1993, p. 157 Temple E (NB 127, pp. 96-97) Agora Southeast (NB 135, p. 148); Stephanidou-Tiberiou 1993, no. 39 Agora North Central (NB 171, p. 69); StephanidouTiberiou 1993, no. 11 St. John's (NB 169, p. 200) Gymnasium (NB 536, p. 92); Stephanidou-Tiberiou 1993, no. 40 Stephanidou-Tiberiou 1993, no. 102 Stephanidou-Tiberiou 1993, no. 127 Near Peirene (NB 8, p. 4); Stephanidou-Tiberiou 1993, no. 124

S-2903 Herakles St. John's (NB 169, p. 360); Figure 4 S-2329 Hermes and Dionysos (NB 863, pp. 18,24); Figure 1 S-1993-2 Satyr S-437 Northwest Shops(?) (NB 15, p. 51); Figure 3; StephanidouTiberiou 1993, no. 71 S-2742 Satyr and Dionysos S-762 Anaploga; Stephanidou-Tiberiou 1993, no. 69 Theater T-108 (bases) S-460 S-1029 S-1376 S- 1726 Agora Southeast (NB 142, p. 18) S-75-6 Forum Southwest T-917

A ROMAN

TABLE

SUPPORT

AT

ANCIENT

CORINTH

505

REFERENCES Agora XI = E. B. Harrison,Archaic and Archaistic Sculpture,Princeton 1965.

in Ajootian, A. 1996. "Praxiteles," Personal Styles in GreekSculpture,

0. PalagiaandJ.J. Pollitt, eds., Cambridge,pp. 91-129. Alchermes,J. 1994. "Spoliain Roman Cities of the Later Empire: LegislativeRationalesand Architectural Reuse,"DOP 48, pp. 167-178. Andren, A. 1964. Antik Skuiptur i Svensk Samlinger, Stockholm. Ashmole, B., and N. Yalouris.1967. Olympia: The Sculpturesof the Temple of Zeus, London.

Baratte,F. 1989. "Lanaissancede Dionysos sur un couverclede sarcophagede Musee d'Amiens," RA, pp. 143-148. Beaumont,L. 1995. "Mythological Childhood:A Male Preserve?"BSA 90, pp.339-361. . 1998. "BornOld or Never Young?Femininity,Childhood, and the Goddesses of Ancient Greece," in The Sacredand the Feminine in Ancient Greece,S. Blundell and M. Williamson, eds. London, pp. 7195. Beazley,J. D. 1918. Attic Red-Figure Vasesin American Museums,

Cambridge. Patron Benker,G. 1975. Christophorus: der Schifer, Fuhrleute und Kraftfahrer,

Munich. BernardiFerrero,D. de. 1966-1974. Teatri classici in Asia Minore, 4 vols., Rome. Bieber,M. 1955. The Sculpture of the HellenisticAge, New York. Blanco,A. 1957. MuseodelPrado, Catalogo de la Escultura, Madrid. Brenk,B. 1987. "Spoliafrom Constantine to Charlemagne:Aesthetics versusIdeology,"DOP 41, pp. 103109. Camp,J. McK. 1989. "The Philosophical Schools of Roman Athens,"in The GreekRenaissance in the Roman Empire, S. Walkerand A. Cameron,

eds., London, pp. 50-55. Caputo, G. 1959. Il Teatro di Sabratha, Rome. Ciggaar,K. N. 1996. WesternTravellers to Constantinople, Leiden. Cohon, R. 1984. "Greekand Roman Table Supportswith Decorative

Reliefs"(diss. New YorkUniversity). CorinthI, iii = R. L. Scranton, Monumentsin theLowerAgoraand North

of theArchaicTemple,Princeton 1951. CorinthIX, ii = M. Sturgeon,Sculpture. TheReliefsfromthe Theater,Princeton 1977. CorinthXVII = J. Biers,TheGreatBath on theLechaionRoad,Princeton 1986. Corso,A. 1996. "The Hermes of Praxiteles,"NumAntCl25, pp. 131153. de Voragine,J. 1948. The GoldenLegend G. Ryanand ofJacobusde Voragine, H. Ripperger,trans.,New York. Drake, M., and W. Drake. 1916. Saints and TheirEmblems,London. Edwards,C. M. 1985. "GreekVotive Reliefs to Pan and the Nymphs" (diss. New YorkUniversity). Engemann,J. 1998. "EinTischfuss mit Dionysos-Satyr-Darstellungaus Abu Mina/Agypten,"JAC41, pp.169-177. Erim, K. T. 1974. "The Satyrand Young Dionysos Group from Aphrodisias,"in MelangesManselII, Ankara, pp. 767-775. Farmer,D. H. 1987. The Oxford Dictionaryof Saints,2nd ed., Oxford. Faustoferri,A. 1991. "The Throne of Apollo at Amyclae:Its Significance and Chronology,"in Sculpture from ArcadiaandLaconia:Proceedingsof an InternationalConference Held at theAmericanSchoolof ClassicalStudiesin Athens,0. Palagia and W. Coulson, eds., Oxford, pp. 159-166. Froning,H. 1981. MarmorSchmuckreliefs mitgriechischen Mythenim 1.Jh. v. Chr., Mainz. Fuchs,M. 1987. Untersuchungen zurAusstattungromischer Theater in Italien undder Westprovinzen desImperiumRomanum, Mainz. Gardner,P. [1883] 1981. British Museum,Catalogueof GreekCoins, repr.Bologne. Gercke,P. 1969. "Satyrmit dem Dio-

5o6

AILEEN

nysos Knabenauf dem Schulter,"

in OpusNobile:Festschriftzum 60. Geburtstagvon UyJantzen, P. Zazoff, ed., Wiesbaden,pp. 4850. Greifenhagen,A. 1931. "Kindesheitmythos des Dionysos,"RM 46, pp. 27-43. Gualandi,G. 1976. "Sculturedi Rodi," ASAtene 38, pp. 7-259. Guntner,G. 1994.Gwttervereine und Gatterversammlungen aufattischen zur Weibreliefs: Untersuchungen TypologieundBedeutung,Wiirzburg. Hadzisteliou-Price,Th. 1978. Kourotrophos, Leiden. Head, B. V. [1877] 1979. British MuseumCatalogue,GreekCoinsof Thrace,repr.London. Hermann,A. 1954. Reallexikonfzir Antikeund ChristentumII, cols. 1241-1250 (Christophoros). Heydemann,H. 1885. "Dionysos Gebuirtund Kindheit,"HallWPr 10, pp. 3-58. Huskinson,J. 1996. RomanChildren's Sarcophagi,Oxford. Hutchinson, V.J. 1991. "The Cult of Dionysos/Bacchusin the GraecoRomanWorld,"JRA4, pp. 222230. Imhoof-Blumer,F.W., and P. Gardner. 1964.AncientCoinsIllustrating LostMasterpieces of GreekArt, A. N. Oikonomides,ed., rev.ed., Chicago. Inan,J. 1975. RomanSculpturein Side, Ankara. Jameson,1894.SacredandLegendary Art, 2 vols., London. Juhel,V. 1991. "Le bain de l'EnfantJesus,"CA 39, pp. 111-132. Kaftal,G. 1965. Iconography of the Saintsin Centraland SouthItalian Schoolsof Painting,Florence. Karouzou,S. 1968. NationalArchaeologicalMuseumCollectionof Athens. Sculpture, Kell,K. 1988.Formuntersuchungen zu spat- undnachhellenistischen Gruppen,Saarbriicken. Kitzenger,E. 1963. "The Hellenistic Heritage in ByzantineArt,"DOP 17, pp. 97-115. Loeb, E. H. 1979. Die GeburtderGotter in dergriechischen Kunstder klassischen Zeit,Jerusalem. Loeschcke,W. 1965. "Neue Studien

AJOOTIAN

zur Darstellungdes tierkopfigen Christophoros,"in Beitrage zur Kunst des christlichenOstens Erste Studien-Sammlung, Recklinghausen,

pp. 37-80. Manderscheid,H. 1982. Die Skulpturen-ausstattung der kaiserzeitlichen Thermenanlagen, Berlin.

Mango, C. 1963. "AntiqueStatuary and the Byzantine Beholder,"DOP 17, pp. 55-75. Marcade,J. 1970. Au Musje de Delos, Paris. Marquardt,N. 1995. Pan in der hellenistischen und kaiserzeitlichen

Plastik,Bonn. Matz, F. 1968. Die dionysischen SarkophageI, Berlin. Minto, A. 1913. "Satirocon Bacco Fanciullo,"Ausonia8, pp. 90-103. Mollard-Besques,S. 1963. Catalogue raisonne desfigurines et reliefsgrecs et romains, 2 vols., Paris.

Moltesen, M. 1990. "The Aphrodisian Sculpturesin the Ny Carlsberg Glyptothek,"in Aphrodisias Papers I (JRA Supplement 1), C. Rouech6 and K. T. Erim, eds., Ann Arbor, pp. 133-146. Moreno, P. 1995. "Ermeteche si slaccia il sandalo,"in Lisippo lArte ela Fortuna, Florence,pp. 230-23 1. Moss, C. F. 1988. "RomanMarble Tables"(diss. PrincetonUniversity). Motte, A. 1996. "Les enfancesdivines dans le mythe Grec,"EtCl 64, pp. 109-125. Nordhagen,P.J. 1961. "The Origin of the Washing of the Child in the Nativity Scene,"ByzZeit 54, pp. 333-337. Paribeni,E. 1959. Catalogo delle Sculture di Cirene, Rome. Pipili, M. 1991. "Hermesand the Child Dionysos:What Did PausaniasSee on the AmyklaiThrone?,"in Stips Votiva:Papers Presented to C. M. Stibbe, M. Gnade, ed., Amsterdam,

pp. 143-147. Pochmarski,E. 1990. Dionysische Gruppen, Vienna. Reeder,E. 1988. HellenisticArt in the WaltersArt Gallery, Baltimore. Richter,G. M. C. 1960. "AnAncient PlasterCast in Munich,"in Theoria: Festschriftffir W-H. Schuchbardt,

F. Eckstein, ed., Baden-Baden, pp. 179-183.

Ridgway,B. S. 1964. "The Date of the So-Called LysippanJason,"AJA68, pp.113-128. . 1984.Roman Copiesof Greek Sculpture: TheProblemof the Originals,AnnArbor. . 1990.HellenisticSculptureI, Madison. . 1997.Fourth-CenturyStylesin GreekSculpture,Madison. Rizzo, G. 1932. Prassitele, Milan. Rockwell,P. 1991. "UnfinishedStatuaryAssociated with a Sculptor's in AphrodisiasPapersII Workshop," (JRA Supplement2), R. R. R. Smith and K. T. Erim, eds., Ann Arbor,pp. 127-143. Rouech6,C., and K. Erim. 1982. "Sculptorsfrom Aphrodisias: Some New Inscriptions,"PBSR 50, pp. 102-115. Saradi-Mendelovici,H. 1990. "Christian Attitudes towardPaganMonuments in Late Antiquity and Their Legacy in Later Byzantine Centuries,"DOP 44, pp. 47-61. Schefold,K. 1957. Die WandePompejis, Berlin. Schneider,R. M. 1990. "Satyrtragt Dionysosknaben,"VillaAlbaniIII, Berlin. Smith, R. R. R. 1996. "Archaeological Researchat Aphrodisias,19891992,"in AphrodisiasPapersIII (JRA Supplement20), C. Rouech6 and R. R. R. Smith, eds., Ann Arbor,pp. 11-72. . 1998. "HellenisticSculpture under the Roman Empire:Fishermen and Satyrsat Aphrodisias," in RegionalSchoolsin Hellenistic Sculpture: Proceedings of an International Conference Held at the AmericanSchoolof ClassicalStudiesat Athens,March15-17, 1996, 0. Palagia,ed., Oxford, pp. 253260. Stephanidou-Tiberiou,Th. 1993. mePlastikeDiakosmese, Trapezophora Athens. Stewart,A. 1990.GreekSculpture:An Exploration,New Haven. Sturgeon,M. 1975. "ANew Group of Sculpturesfrom Ancient Corinth," Hesperia44, pp. 280-301. . 1977. "The Reliefs on the Theater of Dionysos in Athens," AJA 81, pp. 31-53.

A ROMAN

Tempestini,A. 1999. GiovanniBellini, New York. Thompson, H. A. 1950. "The Odeion in the Athenian Agora,"Hesperia 19, pp. 31-141. Todisco,L. 1993.SculturaGrecadelIV Secolo,Milan. Vermeule,C. C. 1981. "Benchand Table Supports:Roman Egypt and Beyond," in Studiesin AncientEgypt, theAegean,and theSudan.Studiesin Honorof Dows Dunham,W. K. Simpson and W. M. Davis, eds., Boston, pp. 180-192. Waldstein, C. 1882. "Hermeswith the Infant Dionysos,"JHS3, pp. 107110. Weis,A. 1992.TheHangingMarsyas andIts Copies,Rome. Weitzmann, K. 1960. "The Survivalof Mythological Representationsin Early Christianand ByzantineArt andTheir Impact on Christianity," DOP 14, pp. 43-68. in Whaite, H. C. 1929. St. Christopher EnglishMedievalPainting, London. Williams, C. K., II. 1989. "AReevaluation of Temple E and the West End of the Forumof Corinth,"in

AileenAjootian DEPARTMENT UNIVERSITY UNIVERSITY,

OF CLASSICS OF MISSISSIPPI

MIss.

[email protected]

38677-I848

TABLE

SUPPORT

AT

The GreekRenaissance in the Roman Empire, S. Walkerand A. Cameron,

eds., London, pp. 156-168. Williams, C. K., II, and 0. H. Zervos. 1990. "Excavationsat Corinth, 1989:The Temenos of Temple E," Hesperia 59, pp. 325-369.

. 1993. "FrankishCorinth: 1992,"Hesperia 62, pp. 1-52. . 1994. "FrankishCorinth: 1993,"Hesperia 63, pp. 1-56. . 1995. "FrankishCorinth: 1994,"Hesperia 64, pp. 1-60. . 1996. "FrankishCorinth: 1995," Hesperia 65, pp. 1-56.

Williams, C. K., II, E. Barnes,and L. M. Snyder.1997. "Frankish Corinth:1996,"Hesperia 66, pp. 747. Williams, C. K., II, E. Barnes, L. M. Snyder,and 0. H. Zervos. 1998. "FrankishCorinth:1997," Hesperia 67, pp. 223-281. Zanker,P. 1965. Wandelder Hermesgestalt in der attischen Vasenmalerei,Bonn. . 1967. Klassizistische Statuen,

Mainz.

ANCIENT

CORINTH

507

69,2000

GREEK EPIGRAPHICAL INDEX

PERSONS 'Ayocxo6fo(u)Rol,fabricant of Rhodian amphora 146-108 a., 328 (1612) 'Ayoc[Oolo]X][],fabricant of Knidian amphora 97-88 a., 328 (92_3); 'Ayoc[Oolox;k],328 (1023) 'Ayayc[oXLs], duovir on Knidian amphora 97-88 a., 328 (111); 'Ayan6ol[?k], 328 (1212) 'Av8Dpoaxo,owner of oinochoe and lekane, graffitifin. V a. 60 (89), 86 (= AgoraXXI F 131 a); 'A[v81]p(xos, 60 (90) (= AgoraXXI F 132); [....]axos, 59 (88) (= AgoraXXI F 131 com); 'Av8po[axol], 62 (92) (=Agora XXI F 97) duovir on Knidian amphora 97-88 a., 327 (42); ['A]paT63o,PovXos, ['A]pvro6[foDkovl], 327 (32) 'Apvrv[0]lloc[o], eponym on Knidian amphora 97-88 a., 328 (91-2); 'Apv-rv[0]loa[o]s, 328 (101 2) eponym on Knidian amphora 97-88 a., 328 (131); 'Ap&[aToxp&ok-r], 'A[poa[roxpo]6c[Y], 328 (141) ALo,v6 os, eponym on Knidian amphora 108-98 a., 327 (72); A[Lo]v6moo,328 (82) fabricant of Knidian amphora 97-88 a., 328 (142); LAoxoDpL&8ocs, 328 (132) [LAoxoDpLo]8oxs, 'Ep6o'qocv-os,duovir on Knidian amphora 97-88 a., 328 (112); 328 (122-3) 'Epto6lqxovros, fabricant of Knidian amphora 97-88 a., 327 (52) `Ep6ooqxocv-o, "'Eptcov,eponym on Knidian amphora 97-88 a., 327 (22) Ev`Poukos,duovir on Knidian amphora 97-88 a., 327 (62) EDv0?z[oo,duovir on Knidian amphora 97-88 a., 327 (13) 'Jciacov,duovir on Knidian amphora 97-88 a., 328 (151)

duovir on Knidian amphora 108-98 a., 327 (74); FlI]TcTo6a-vpovoo, 328 (84) 'J[TcTc]crvpcTo5, eponym on Knidian amphora 108-98 a., 327 (71); KoAL8s K[oYrpoc-ro, 328 (8k) Kocpovz c4, duovir on Knidian amphora 97-88 a., 327 (6k) K?A(---)?,owner? of amphorasfin. V a., 58 (84), 59 (85)

5IO

AaX7- duovir on Knidian amphora 97-88 a., 327 (12) Aco,v, owner? of amphora, graffitofin. V a., 51 (68) Ai(an, author of lead letter init. IV a., 951 [M]?socvToc[s], duovir on Knidian amphora 97-88 a., 327

(43);

[M?so?vTocs],327 (33)

[M6aTs], duovir on Knidian amphora 97-88 a., 328 (151)

H

owner?of amphorafin.V a., 57 (83) ocxv[---]?, ?voxXNi, addressee of lead letter init. IV a., (951)

HILLaEvo;, eponym on Knidian amphora 97-88 a., 327 (51)

[YIo;D6vLxos], fabricant of Knidian amphora 97-88 a., 327 (23) duovir on Knidian amphora 108-98 a., 327 (75); HoX6cDXPoo0, HI[o?6oclx]po[og],328 (85)

'Po8q, owner of platefin. II - init. I a., 298 (7)

ETHNIC Kv(Los: Kv6&ov,328 (132, 142); Kv6&[ov], 327 (53, corrected); Kv[61&[ov],328 (93); [Kv6&ov],328 (103, corrected);Kv68ov], 327 (73, corrected); [Kv61&ov,327 (21, corrected);K[v68Lov,328 (83,

corrected)

DEITY n;koDTov:MI;o6I[-v&vos], 3366-7, corrected

INSCRIPTIONS

STUDIED OR EMEND ED

AgoraInventoryNumber IL 1702 P P P P P P P P

2067 18620 23821 23835 23968 23991 26387 26389

95-101 51-52 62 60 60 58-59 57 59 59

(68) (92) (89) (90) (84) (83) (85) (88)

CorinthInventoryNumber 97-1

336-337

CORRTGENDA: Natalia Vogeikoff-Brogan, "Late Hellenistic Pottery":p. 328, nos. 9-11, 13, 15, 16, for 1956-NAK-273-278, read 1956-NAA\-273-278; p. 328, no. 16, for'Ayxof3o,olcluster??Bouread'Ayxof3o,o(6)Icluster? kou.

HESPERIA

SUPPLEMENTS

13 MarcusAurelius:Aspectsof Civicand CulturalPolicyin theEast, by James H. Oliver (1970). $15.00 14 ThePoliticalOrganizationofAttica,byJohnS.Traill(1975). $15.00 15 TheLetteringof an AthenianMason,by Stephen V. Tracy (1975). $15.00

16 A Sanctuaryof Zeuson MountHymettos,byMerleK. Langdon(1976). $15.00 17 Kalliasof Sphettosand theRevoltofAthensin 286 B.C., byT. Leslie ShearJr. (1978). $15.00 Presentedto Eugene 19 Studiesin AtticEpigraphy,History,and Topography (1982). $15.00 Vanderpool Presentedto Sculpture,and Topography 20 Studiesin AthenianArchitecture, HomerA.Thompson(1982).$15.00 21 Excavationsat Pylosin Elis, by John E. Coleman (1986). $25.00

22 Attic GraveReliefsThatRepresentWomenin theDressofIsis, by ElizabethJ. Walters (1988). $40.00

23 HellenisticReliefMoldsfrom theAthenianAgora,by ClaireveGrandjouan (1989). $25.00 at Mochlosand Gourniaand theHouseTombsof 24 ThePrepalatialCemeteries BronzeAge Crete,byJeffrey S. Soles (1992). $35.00

25 Debrisfrom a PublicDining Placein theAthenianAgora, by Susan I. Rotroff andJohn H. Oakley (1992). $35.00 Stagesand Chronol26 TheSanctuaryofAthenaNike in Athens:Architectural ogy, by Ira S. Mark (1993). $50.00

on GreekArchitectural 27 Proceedingsof theInternationalConference Terracottas of the ClassicalandHellenisticPeriods,December12-15, 1991, edited by Nancy A. Winter (1994). $120.00

28 Studiesin ArchaicCorinthianVasePainting,by D. A. AmyxandPatricia Lawrence(1996). $65.00

29 TheAthenianGrain-TaxLaw of374/3 $35.00

B.C., by RonaldS. Stroud(1998).