Proceedings of the Sixth International Congress of Genetics, Volume I, Transactions and General Addresses, Ithaca, New York, 1932

PROCEEDINGS OF T H E

SIXTH INTERNATIONAL CONGRESS OF GENETICS Ithaca, New York, 1932 VOLUME I

TRANSACTIONS AND GENERAL ADDRESSES Edited by DONALD F. J O N E S

Published at Menasha, Wisconsin by the BROOKLYN BOTANIC GARDEN Brooklyn, New York, U.S.A.

Composite Reprint of Complete Proceedings Volumes 1 and 2 BY T H E GENETICS SOCIETY OF AMERICA Genetics Business Office, P. 0. Drawer U, University Station Austin, Texas 78712

Compoud. Printed and Bound by we-?.-

George Bmta Publishing Company

Mmarha, Wnuonnln SECOND P~JNTINC

.

1968

UNIVERSITY OF TEXAS PRINTING DIVISION AUSTIN. TFXAS

FOREWORD

Volume one contains the general information about the congress, the opening addresses, and the papers presented at the morning sessions. Volume two contains the condensed papers presented in the sectional sessions and the descriptions of the exhibits. Volume two was printed in advance of the meeting and distributed to members as they enrolled. All members who did not attend received a copy after the congress. The present volume is sent to all members. Additional copies of both volumes may be obtained from the BROOKLYN BOTANICGARDEN,1000 Washington Ave., Brooklyn, New York, U.S.A. O n account of financial limitations, it was not possible to print the sectional papers in full. Not all papers were received in time for publication in volume two. All that have been received are added to this volume. The accompanying index includes items in both volumes. I t is printed in a separable section to be put at the end if both volumes are bound together. The congress is indebted to the BROOKLYN BOTANIC GARDEN 'for undertaking the publication of these proEXceedings and to the CONNECTICUT AGRICULTURAL PERIMENT STATION for the facilities avaiIabIe in the editorial office. The editor wishes to acknowledge with gratitude the help of the other members of the publication committee consisting of Doctors C. E. ALLEN, E. M. EAST,and H. D. KING.Much credit is due Mrs. WILLIAM D. MILLERfor the preparation of copy and correction of proofs for both volumes. New Haven, Connecticut, December 6,1932

DONALD I?. JONES

TABLE OF CONTENTS PAGE

......................................... 1 Preparation for the sixth congress ................................................ 4 Organization of the sixth congress ............................................... 7

History of preceding genetics congresses

Permanent international committee Committee on arrangements

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

8

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

Organization committee for the Sixth International Congress of Genetics

.........

8 9

......................................................................... 9 Executive council and committees .............................................. 9 Committee of one hundred ................................................... I 1

Officers

...................................................... 15 Official delegates from foreign governments ...................................... 20 Official delegates from institutions and societies .................................... 21 Institutional members ............................................................ 22 Contributing members ............................................................ 22 Members of the congress ......................................................... 23 Program ........................................................................ 50 Exhibits ......................................................................... 68 Excursions and entertainments ................................................... 74 Group conferences ............................................................... 75 Transportation. tours. and lectures ................................................ 76 Treasurer's report ............................................................... 79 Opening address. EDMUND B. WILSON.............................................. 81 Address of welcome. A . R . MANN................................................ 84 ................................................. 86 Response. RICHARD GOLDSCHMIDT Address of the president. THOMAS HUNTMORGAN. The rise of genetics .............. 87 Minutes of plenary meetings

General addresses given in morning sessions (for addresses given in afternoon sessions and descriptions of exhibits. see volume two and also appendix. page 367 of this volume) .......................................................................104 BLAKESLEE. ALBERTF..The species problem in Datura CREW.F. A . E.. Inheritance of educability

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

I04

.....................................121

.......................................... EMERSON. R . A.. The present status of maize genetics .......................... FEDERLEY. HARRY. The conjugation of the chromosomes ....................... DAVENPORT. C. B.. Mendelism in man

FISHER. R . A.. The evolutionary modification of genetic phenomena GOLDSCHMIDT. R.. Genetik der geographischen Variation

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

HALDANE. J . B . S., Can evolution be explained in terms of known genetical facts? MOHR.OTTOL.. On the potency of mutant genes and wild-type allelomorphs .... MULLER.H. J.. Further studies on the nature and causes of gene mutations ......

...................... STADLER. L . J.. On the genetic nature of induced mutations in plants ............

SAX.KARL.The cytological mechanism for crossing over

STERN.CURT.Neuere Ergebnisse iiber die Genetik und Zytologie des Crossing over STURTEVANT. A . H., The use of mosaics in the study of the developmental effects of genes ................................................................... TIMOI&EFF.RESSOVSKY. N . W.. Mutations of the gene in different directions ......

.

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

VAVILOV. N I.. The process of evolution in cultivated plants WINCE.0.. 'The nature of sex chromosomes

WRIGHT. SEWALL. The roles of mutation. inbreeding. crossbreeding. and selection

................................................................ pendix ........................................................................ lex of volumes one and two ................................................... in evolution

ILLUSTRATIONS President of congress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Frontispiece Willard Straight Hall

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Group picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Facing page

48

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

69

Growing-plant exhibits Maize chromosome map

.......................................................... 71

vii

H I S T O R Y O F PRECEDING GENETICS CONGRESSES The increasing interest in the production of new forms of plants and the growing appreciation of a need for fuller knowledge of the process of heredity induced the council of the Royal Horticultural Society of England in 1899 to hold an "International Conference on Hybridisation and on the Cross-Breeding of Varieties." The meeting was held July 11 and 12. Special invitations were sent to 125 well-known hybridists and botanists, and all others interested were invited. There is no statement as to the number that attended, although we note that covers for 130 were laid a t the banquet. Six addresses were scheduled for the first d a y and eight for the second. These details and the papers presented and sent to the conference may be found in volume twenty-four of the Journal of the Royal Horticztltural Society for 1900, published by the society in London. I n an address entitled "Hybridisation and Cross-Breeding as a Method of BATESON made Scientific Investigation," presented at this meeting, WILLIAM the following statement: "Of all the methods which are open to us for investigating the facts of Natural History, there is perhaps none which is more likely t o bring forth results of first-rate importance. . . . I t is perhaps simpler to follow the beaten track of classification or of comparative anatomy, or to make for the hundredth time collections of the plants and animals belonging to certain orders, or to compete in the production or cultivation of familiar forms of animals o r plants. But all these pursuits demand great skill and unflagging attention. Any one of them may well take a man's whole life. If the work which is now being put into these occupations were devoted to the careful carrying out and recording of experiments of the kind we are contemplating, the result, it is not, I think, too much to say, would in some five-and-twenty years make a revolution in our ideas of species, inheritance, variation, and the other phenomena which go to make up the science of Natural History. W e should, I believe, see a new Natural History created." In 1902, an "International Conference on Plant Breeding and Hybridization" was held in New York City under the auspices of the Horticultural Society of New York. Three days, September 30 to October 2, were given to a program dealing mostly with the practical application of hybridization to plant breeding, Seventy-five members were enrolled. Thirty addresses were presented and thirteen papers were read by title. These were published as volume one of the Memoirs of the Horticultural Society of New York by the society in New York City. Four .years later the Royal Horticultural Society again called an "Inter-

2

PROCEEDINGS O F T H E S I X T H

national Conference oh Hybridisation and Plant Breeding," to be held in London, England, July 30 to August 3, 1906. About 300 were specially invited. There is no record of the number that attended. The president of the congress was WILLIAMBATESON.I n his inaugural address he said: "Like other new crafts, we have been compelled to adopt a terminology, which, if somewhat deterrent to the novice, is so necessary a tool to the craftsman that it must be endured. But though these attributes of scientific activity are in evidence, the science itself is still nameless, and we can only describe our pursuit by cumbrous and often misleading periphrasis. T o meet this difficulty I suggest for the consideration of this congress the term Genetics, which sufticiently indicates that our labours are devoted to the elucidation of the phenomena of heredity and variation: in other words, to the physiology of Descent, with implied bearing on the theoretical problems of the evolutionist and the systematist, and application to the practical problems of breeders, whether of animals or plants. After more or less undirected wanderings we have thus a definite aim in view." Unlike the two previous conferences the program included animal as well as plant topics for presentation and discussion. About forty papers were listed in the program and forty-six were published in the report of the conference, together with descriptions of twenty-two exhibits, the latter dealing entirely with plants. A l t h o ~ g hin the preliminary announcement the meeting was called a "Conference on Hybridisation and Plant Breeding," the supplementary volume of the Joz~rnalof the Royal Horticult~ralSociety that contains the addresses and other information was given the title of "Report of the Third I n t e ~ n a tional Conference on Genetics." In this way the word, genetics, first proposed during the meeting, was put to immediate use. The fourth genetics congress, held in Paris, France, in 1911, enrolled 234 members. The actual attendance was not stated. Fifty-eight papers were included in the report, published in 1913 by MASSONET CIE., Bditeurs, Libraires de 1'AcadCmie de Medecine, Paris. A feature of this congress was a number of excursions to various places of interest to geneticists, including, among others, L'INSTITUTPASTEUR,the ET CIE., L'~?COLE experimental plots of the MAISONVILMORIN-ANDRIEUX VBTBRINAIRED'ALFORT,and the MUSEUMD'HISTOIRENATURELLEDE PARIS. At many of these places, plant and animal material of genetic interest was exhibited. After the fourth congress, it was generally understood that the succeeding international genetic gathering should be scheduled for 1916. Circumstances ordained otherwise, and it was not until sixteen years later, in 1927,

INTERNATIONAL CONGRESS O F GENETICS

3

that the Fifth International Congress of Genetics met in Berlin, Germany. The rapid growth of the science of genetics was shown by the increase in membership and in the diversity and number of topics discussed. The enrollment reached 966, there were 903 members present, and 35 countries were represented. Excursions were made to many institutions and places of genetic interest. The proceedings were issued in two volumes, totaling 1646 pages and 14 plates, as a supplement to the Zeitschrift fur Induktive Abstammungs-und Vererbungslehre, published by GEBRUDER BORNTRAEGER, Leipzig, in 1928. One hundred and forty-eight papers were included in these proceedings.

4

PROCEEDINGS O F THE SIXTH

P R E P A R A T I O N F O R THE S I X T H CONGRESS C. B. Davenport

A t the Fifth International Congress of Genetics, held in ~ e r . l i nin 1927, it was voted, a t a business meeting held September 17, to hold the next congress in the United States in case the United States presented an invitation. When a permanent international committee to look after the next congress was appointed, this committee suggested that Ithaca would be a desirable meeting place. About December 12, 1927, a letter was written to a dozen geneticists in the United States asking for suggestions as to the best meeting place of the congress in 1932. A number of places were suggested, with a concentration on Ithaca and New Haven. A t the Nashville meeting of the American Association for the Advancement of Science, held in December, 1927, a committee to make arrangements for the place of the next congress was elected. Suggestions having been made to the committee that suitable places for the meeting wduld be at Ithaca or UNIVERSITY and YALEUNINew Haven, inquiries were made of CORNELL VERSITY and cordial and generous invitations were received from each of these universities to hold there the congress of 1932. In March, 1928, a three-page mimeographed letter, signed by the committee appointed a t Nashville, was sent to the geneticists of the United States, setting forth the advantages that each of the two places offered and asking a vote of preference. Of 153 persons who returned ballots 106 voted in favor of Ithaca and 47 favored New Haven. Thus by vote of the geneticists of the country Ithaca was decided upon as the place of meeting. This decision was approved by the international committee. On April 30, 1929, there was held at Washington a meeting of the orH U N TMORGAN was unanimously elected presganizing committee. THOMAS R. A. EMERSON chairman of the ident of the sixth genetics congress and C. C. LITTLEwas unanimously nominated as secretary local con~n~ittee; general. C. B. DAVENPORT, as chairman of the organizing committee, was authorized to get the full vote of the organizing committee as to vice presidents. As a result of letters sent out t o members of the organizing committee a list of preferences for vice presidents from each country was drawn up. The organizing committee meanwhile drew up a list of leading geneticists to constitute a general committee of the congress. This group was called the committee of one hundred. At a meeting a t Woods Hole, August 19, 1929, C. C. LITTLEwas elected secretary general, and preliminary lists of persons were drawn up for the following committees: finance, transportation, exhibits, program and publi-

INTERNATIONAL CONGRESS O F GENETICS

5

cation. The chairmen of these committees and that of the local committee together with the secretary general and treasurer constituted the executive council. The functions of the organizing committee were turned over to this executive council in March, 1930. The financing of the sixth congress was, from the outset, recognized as a matter of prime importance. With the progress of the ever-increasing depression its work became still more difficult and complicated. R. C. COOK, editor of the Journal of Heredity, was selected as treasurer of the congress. The finance committee consisted of CHARLESMCALPINPYLEas chairman with eleven additional members. In March, 1931, owing to the resignation was elected chairman of the finance comof C. M. PYLE,C. B..DAVENPORT mittee. The finance committee drew up a plan for raising the necessary funds: ( 1 ) by listing all geneticists and soliciting a ten dollar membership fee from each; (2) by correspondence with each member of the committee of one hundred to induce him to become responsible for a certain quota of contributions; ( 3 ) by preparing a list for solicitation of practical breeders; (4) by drawing up a list of possible donors with suggestions for handling each case; ( 5 ) by formation of subcommittees to solicit contributions from industries related to genetics. Some 30 subcommittees, called contact committees, were thus drawn up and these had the double function of interesting the industries and the breeders in the coming congress and soliciting contributions from them for the success of the congress. Especial mention must be made of the following contributions: from the OF NEW Y ORK,through the CARNEGIEINSTITUCARNEGIECORPORATION TION OF WASHINGTON, $5,000; CARNEGIE INSTITUTION OF WASHINGTON, indirectly through the Department of Genetics, about $1,000; from the CARNEGIE ENDOWMENT FOR INTERNATIONAL PEACE,through L. C. D U N N , $1,352.31 ; and from COLUMBIA UNIVERSITY the sum of $1,500. The following is a summary of the receipts of other amounts: E . P. PRENTICE, $300; and sustaining and institutional member$750; E . C. MACDOWELL, JAMES, ships or patrons of $100 each from the following: MRS. WORTHAM JAMES F. PORTER,HARVARD UNIVERSITY, COLUMBIA UNIVERSITY, JOHNS HOPKINS UNIVERSITY,BUCKNELLUNIVERSITY,CORNELLUNIVERSITY, DARTMOUTH COLLEGE,BROWNUNIVERSITY,CARLETONCOLLEGE( 1931) , UNIVERSITYOF MISSOURI,TEXAS AGRICULTURAL EXPERIMENT STATION (1931), SMITHSONIAN INSTITUTION; also the GENERALELECTRIC COMPANY ; AMERICAN FRUIT GROWERS'ASSOCIATION ; HAWAIIANPINEAPPLE CANNERS;INSTITUTE OF FOREST GENETICS;RUSSELL-MILLERMILLING COMPANY, Minnesota; GENERALMILLS, INCORPORATED, Minnesota; HAWAIIAN SUGARPLANTERS' ASSOCIATION, Honolulu; ARMOURAND COM-

6

PROCEEDINGS O F THE SIXTH

PANY,Chicago; PILLSBURY FLOUR MILLS COMPANY, Minnesota; MINNESOTA CROPIMPROVEMENT ASSOCIATION ; AMERICAN GUERNSEY CATTLE CLUB;GALLATINVALLEYSEED COMPANY, Montana; TRI-STATESOFT WHEATIMPROVEMENT ASSOCIATION, Ohio. Contributions of $50 or less were received from the following institutions and individuals in addition to their membership fee: ALBERSBROTHERS MILLINGCOMPANY, California, WESLEYAN UNIVERSITY, J. BELLING,C. B. BRIDGES, J. T. BUCHHOLZ, LEONJ. COLEand students, J. L. COLLINS,H. E. CRAMPTON, C. H. DANFORTH, C. B. DAVENPORT, B. M. DAVIS,D. FAIRCHILD, H. D. GOODALE, J. W. GOWEN,M. F. GUYER,H. K. HAYES,J. H. KELLOGG, H.'D. KING,D. E. LANCEFIELD, F. R. LILLIE,M. T. MACKLIN, H. J. MULLER,C. H. MYERS,R. PEARL,H. D. PLOUGH,K. SAX,A. F. SHULL,C. R. STOCKARD, F. B. SUMNER,E. N. WENTWORTH, 0 . E. WHITE, P. W. WHITING,D. D. WHITNEYand S. WRIGHT.

INTERNATIONAL CONGRESS O F GENETICS

7

ORGANIZATION O F T H E S I X T H CONGRESS C. C. Little

The organization of the Sixth International Congress of Genetics was carried out by the executive council. Due to the unique financial situation. during the years 1930-1932 inclusive, certain complications were encountered which it is hoped will not occur in future congresses. The original plans for the congress were distinctly de luxe. They included a budget of approximately $100,000, of which expenses of foreign delegates and the exhibits were to consume a large part. As economic conditions became worse rather than better simplification of the program and downward revision of the budget naturally followed. This was aided very considerably by the coijperation and understanding of the heads of the various committees of the council. On the program committee, E. M. EASTdeveloped a comprehensive but simple plan which cut travel expenses of invited participants t o a minimum consistent with adequate representation of the various foreign countries. On who succeeded E . C. MACDOWELL as the exhibits committee, M. DEMEREC, chairman, showed remarkable skill and industry in arranging an exhibit that proved to be in the minds of many the chief feature of the congress. This was done at less than one-sixth the cost of the original estimate of exhibits; R. A. EMERSON and other members of the Department of Plant Breeding a t CORNELLUNIVERSITY helped greatly. On the publication committee, D. F. JONES with the coijperation of C. S. GAGERand the BROOKLYN BOTANIC GARDEN made a most economical and advantageous arrangement for the publication of the Proceedings. The local committee, under the chairmanship of R. A. EMERSON, carried out the detailed organization of the congress most effectively and economically. Treasurer R. C. COOKused much ingenuity in saving expense to the congress in many ways. L. C. D U N N ,in charge of transportation, performed a difficult and constantly changing duty in the ortaking over ganization of steamship and railway facilities. C. B. DAVENPORT, the chairmanship of the finance committee in place of C. M. PYLE,organized a large number of subcommittees which adequately covered the various a p plied phases of genetics. A n account of this type should, I believe, be brief, making no effort to cover the myriad details which must of necessity differ in some degree in each succeeding congress. In general the chief weakness of the kind of organization that prepared the Sixth International Congress of Genetics is its geographical distribution. The need for constant conference is so great that it is the writer's belief that the council of future congresses should be a body the membership of which

PROCEEDINGS O F THE SIXTH

8

is chosen entirely from the locality or localities at which the congress will be held. Various devices can be employed to enlist a wide interest under the direction of the council. The detailed problems are, however, so definitely local that the working body should be closely knit. Advice can be obtained as desired from many people at more or less distant points. The experience of the Sixth International Congress of Genetics has shown that the great preponderance of successful experimental geneticists are or can become successful organizers. No outstanding executive or organizer as such appears to be needed. A reasonable amount of high grade clerical help is an essential. With it the organization of a congress becomes readily the part-time activity of several scientists rather than the full-time work of a few. Organization of a congress, in my opinion, should start not less than two full years in advance but need not be begun sooner. In closing I wish to thank especially Miss MARGARET CAMERON, who for eighteen months acted as full-time secretary to the office of the Secretary General, and Miss MARYE. RUSSELL,who gave considerable time to the work before Miss CAMERON'S appointment. PERMANENT INTERNATIONAL COMMITTEE APPOINTED BY THE FIFTH INTERNATIONAL CONGRESS O F GENETICS

TSCHERMAK-SEY-Japan: S. IKENO Netherlands: J. P. LOTSY Belgium: V. LATHOUWERS Norway: K. BONNEVIE Russia: N. KOLTZOFF Denmark: W. JOHANNSEN France: L. BLARINGHEM Sweden : H. NILSSON-EHLE Switzerland : A. ERNST Germany: E. BAUR Great Britain: R. C. PUNNETT United States of America: T. H. MORGAN, C. B. DAVENPORT Italy: P. ENRIQUES Austria: E.

VON

SENEGG

COMMITTEE O N ARRANGEMENTS

This committee was appointed by the Joint Genetics Sections of the Botanical Society of America and the American Society of Zoologists to arrange for the place of holding the Sixth International Congress of Genetics. Chairman C. B. DAVENPORT, E. B. BABCOCK W. E. CASTLE L. J. COLE

R. A. EMERSON D. F. JONES T. H. MORGAN G. H. SHULL

INTERNATIONAL CONGRESS O F GENETICS

9

ORGANIZATION C O M M I T T E E FOR THE SIXTH INTERNATIONAL CONGRESS

O F GENETICS

This committee was appointed by C. B. DAVENPORT and T . H. MORGAN, representatives from the United States on the permanent international committee, to select the president, vice presidents and the executive council. Chairman C. B. DAVENPORT, H. S. JENNINGS L. J. COLE T . H. MORGAN E. M. EAST G. H. SHULL R. A. EMERSON OFFICERS

President T . H. MORGAN, CALIFORNIA INSTITUTE OF TECHNOLOGY, Pasadena, California, U.S.A. Vice Presidents E. VON TSCHERMAK-SEYSENEGG, Y. TANAKA, Japan H. DE VRIES,Netherlands Austria RE, L. COCKAYNE, New Zealand V. G R ~ ~ G O IBelgium A. H. R. BULLER,Canada 0. L. MOHR,Norway E . MALINOWSKI, Poland 0. WINGE,Denmark H. FEDERLEY, Finland N. VAVILOV, Russia L. C U ~ N O France T, A. ZULUETA,Spain H. NILSSON-EHLE,Sweden C. CORRENS, Germany J. B. S. HALDANE, Great Britain A. ERNST,Switzerland P. ENRIQUES,Italy EXECUTIVE C O U N C I L

C. C. LITTLE,chairman and general secretary of the congress R. C. COOK,treasurer C. B. DAVENPORT, chairman finance committee M. DEMEREC, chairman exhibits committee L. C. D U N N chairman , of the transportation committee and secretary of the council E. M. EAST,chairman program committee R. A. EMERSON, chairman local committee D. F . JONES,chairman publication committee F I N A N C E COMMITTEE

C. B. DAVENPORT, Chairman L. C. D U N N

D. FAIRCHILD J. W . GOWEN

10

PROCEEDINGS O F THE SIXTH

C. R. STOCKARD E. N. WENTWORTH F. A. WOODS

H. H. LAUGHLIN F. OSBORN E. P. PRENTICE F. D. RICHEY

EXHIBITS COMMITTEE

M. DEMEREC,Chairman OAKESAMES E. B. BABCOCK R. E. CLELAND C. H. DANFORTH L. C. DUNN DAVIDFAIRCHILD H. H. LAUGHLIN

FRANK E. LUTZ E. C. MACDOWELL L. F. RANDOLPH MARCUS M. RHOADES F. D. RICHEY SOPHIASATINA SEWALLWRIGHT

TRANSPORTATION COMMITTEE

C. H. MYERS L. C. DUNN,Chairman M. DEMEREC E. W. SINNOTT WALTERLANDAUER Foreign Advisory Members A. S. SEREBROVSKY, MOSCOW J. KRIZENECKY, Brno T. H. SHEN,Nanking H. NACHTSHEIM, Berlin T, TANAKA, Mi~azaki M. PEASE,Cambridge I. F. PHIPPS, Adelaide PROGRAM COMMITTEE

E. M. EAST,Chairman W. E. CASTLE R. E. CLELAND H. D. KING H. H. LAUGHLIN

G. H. SHULL A. H. STURTEVANT E. N. WENTWORTH SEWALLWRIGHT LOCAL COMMITTEE

General chairman: R. A. EMERSON Bulletins and daily news sheet: A. C. FRASER Decorations: E. A. WHITE Entertainment of women and children: Mrs. R. A. EMERSON, Mrs. C. H. MYERS Excursions: R. G. WIGGANS Geneva Session : RICHARD WELLINGTON Housing, dormitories, camp sites: R. A. EMERSON Information service and guides: J. R. LIVERMORE

INTERNATIONAL CONGRESS O F GENETICS

11

Lecture room and exhibit assignments: R. A. EMERSON, M. DEMEREC Lecture room facilities: W. D. SWOPE Music: P. J. WEAVER Photographs: M. GORDON Picnic: L. H. MCDANIELS Press service: BRISTOW ADAMS,L. C. BOOCHEVER Reception: R. A. EMERSON Signs: H. S. PERRY Sports: R. G. WIGGANS Stenographic service: F. FEEHAN Transportation, train and automobile: C. H. MYERS PUBLICATION COMMITTEE

D. F. JONES,Chairman C. E . ALLEN

E. M. EAST H. D. KING

COMMITTEE OF ONE HUNDRED

C. E. ALLEN,UNIVERSITY OF WISCONSIN, Madison, Wisconsin OF KENTUCKY, Lexington, Kentucky W . S. ANDERSON, UNIVERSITY E . B. BABCOCK, UNIVERSITYOF CALIFORNIA, Berkeley, California A. M. BANTA,BROWNUNIVERSITY, Providence, Rhode Island OF MICHIGAN, Ann Arbor, Michigan H. H . BARTLETT, UNIVERSITY J O H N BELLING,UNIVERSITY OF CALIFORNIA, Berkeley, California A. F. BLAKESLEE, CARNEGIE INSTITUTION OF WASHINGTON, Cold Spring Harbor, New York C. G. BOWERS, Maine, New York Pasadena, CaliC. B. BRIDGES,CALIFORNIA INSTITUTEOF TECHNOLOGY, fornia OF WISCONSIN, Madison, Wisconsin R. A. BRINK,UNIVERSITY UNIVERSITY OF ILLINOIS,Urbana, Illinois J. T. BUCHHOLZ, E . ELEANOR CAROTHERS, UNIVERSITYOF PENNSYLVANIA, Philadelphia, Pennsylvania W. E. CASTLE,Bussey Institution, HARVARD UNIVERSITY,Forest Hills, Boston, Massachusetts OF CALIFORNIA, Berkeley, California R. E. CLAUSEN,UNIVERSITY R. E. CLELAND,GOUCHERCOLLEGE,Baltimore, Maryland L. J. COLE,UNIVERSITY OF WISCONSIN, Madison, Wisconsin J. L. COLLINS,UNIVERSITY OF HAWAII,Honolulu, Hawaii

12

PROCEEDINGS O F THE S I X T H

E. G. CONKLIN,PRINCETON UNIVERSITY, Princeton, New Jersey H. E. CRAMPTON, Barnard College, COLUMBIA UNIVERSITY, New York, New York C. H. DANFORTH, STANFORD UNIVERSITY, Stanford University, California C. B. DAVENPORT, CARNEGIEINSTITUTION OF WASHINGTON, Cold Spring Harbor, Long Island, New York B. M. DAVIS,UNIVERSITY OF MICHIGAN, Ann Arbor, Michigan INSTITUTION OF WASHINGTON, Cold Spring HarM. DEMEREC,CARNEGIE bor, Long Island, New York L. C. DUNN,COLUMBIA UNIVERSITY, New York, New York E. M. EAST,Bussey Institution, HARVARD UNIVERSITY, Forest Hills, Boston, Massachusetts R. A. EMERSON, CORNELL UNIVERSITY, Ithaca, New York W. H. EYSTER,BUCKNELL UNIVERSITY, Lewisburg, Pennsylvania DAVIDFAIRCHILD, UNITEDSTATESDEPARTMENT OF AGRICULTURE, Washington, District of Columbia A. C. FRASER,CORNELLUNIVERSITY,Ithaca, New York H. B. FROST,CITRUSEXPERIMENT STATION, Riverside, California W. H. GATES,LOUISIANA STATEUNIVERSITY,Baton Rouge, Louisiana J. M. GEROULD, DARTMOUTH COLLEGE,Hanover, New Hampshire H. D. GOODALE, 257 W. Main Street, Williamstown, Massachusetts H. B. GOODRICH, WESLEYAN UNIVERSITY, Middletown, Connecticut OF CALIFORNIA, Berkeley, California T. H. GOODSPEED, UNIVERSITY J. W. GOWEN,T H E ROCKEFELLER INSTITUTE,Princeton, New Jersey OF WISCONSIN, Madison, Wisconsin M. F. GUYER,UNIVERSITY E. B. HANSON, WASHINGTON UNIVERSITY, St. Louis, Missouri H. K. HAYES,UNIVERSITY OF MINNESOTA, University Farm, St. Paul, Minnesota S. J. HOLMES,UNIVERSITY OF CALIFORNIA, Berkeley, California H. R. HUNT,MICHIGAN STATECOLLEGE, East Lansing, Michigan Montreal, Canada C. L. HUSKINS,MCGILLUNIVERSITY, H. L. IBSEN,KANSASSTATECOLLEGE,Manhattan, Kansas JOHNS HOPKINS UNIVERSITY, Baltimore, Maryland H. S. JENNINGS, D. F. JONES,CONNECTICUT AGRICULTURAL EXPERIMENT STATION,New Haven, Connecticut VERNONKELLOGG, NATIONALRESEARCHCOUNCIL,Washington, District of Columbia HELEND. KING,WISTARINSTITUTE,Philadelphia, Pennsylvania D. E. LANCEFIELD, COLUMBIA UNIVERSITY, New York, New York

INTERNATIONAL CONGRESS O F GENETICS

13

WALTER LANDAUER, STORRSAGRICULTURAL EXPERIMENTSTATION, Storrs, Connecticut H. H. LAUGHLIN, CARNEGIE INSTITUTION OF WASHINGTON, Cold Spring Harbor, New York J. W. LESLEY,CITRUSEXPERIMENT STATION,Riverside, California F. R. LILLIE,UNIVERSITY OF CHICAGO, Chicago, Illinois E. W. LINDSTROM, IOWASTATECOLLEGE, Ames, Iowa C. C. LITTLE,ROSCOEB. JACKSON MEMORIAL LABORATORY, Bar Harbor, Maine H. H. LOVE,CORNELL UNIVERSITY, Ithaca, New York C. E. MCCLUNG,UNIVERSITY OF PENNSYLVANIA, Philadelphia, Pennsylvania E. C. MACDOWELL, CARNEGIE INSTITUTION OF WASHINGTON, Cold Spring Harbor, New York MADGEMACKLIN,UNIVERSITY OF WESTERNONTARIO, London, Ontario, Canada P. C. MANGELSDORF, TEXASAGRICULTURAL EXPERIMENT STATION, College Station, Texas J. W. MAVOR, UNIONCOLLEGE, Schenectady, New York C. W. METZ,CARNEGIE INSTITUTION OF WASHINGTON, Cold Spring Harbor, New York A. R. MIDDLETON, UNIVERSITY OF LOUISVILLE, Louisville, Kentucky T. H. MORGAN, CALIFORNIA INSTITUTE OF TECHNOLOGY, Pasadena, California H. J. MULLER,UNIVERSITY OF TEXAS, Austin, Texas OF AGRICULTURE, Ithaca, New C. H. MYERS,NEW YORKSTATECOLLEGE York K. K. NABOURS, KANSAS STATECOLLEGE, Manhattan, Kansas I-I. H. NEWMAN, UNIVERSITY OF CHICAGO, Chicago, Illinois J. H. PARKER, KANSAS STATECOLLEGE, Manhattan, Kansas RAYMOND PEARL,JOHNSHOPKINSUNIVERSITY, Baltimore, Maryland J. C. PHILLIPS,Wenham, Massachusetts H. H. PLOUGH, AMHERST COLLEGE, Amherst, Massachusetts P. POPENOE,T H E HUMANBETTERMENT FOUNDATION, Pasadena, California L. F. RANDOLPH, CORNELL UNIVERSITY, Ithaca, New York F. D. RICHEY,UNITEDSTATESDEPARTMENT OF AGRICULTURE, Washington, District of Columbia 0. RIDDLE,CARNEGIE INSTITUTION OF WASHINGTON, Cold Spring Harbor, New York

i4

PROCEEDINGS OF THE SIXTH

ELMERROBERTS, Agricultural College, UNIVERSITY OF ILLINOIS,Urbana, Illinois UNIVERSITY, Forest Hills, Boston, KARLSAX,Bussey Institution, HARVARD Massachusetts J. H. SCHAFFNER, OHIO STATEUNIVERSITY, Columbus, Ohio A. F. SHULL,UNIVERSITY OF MICHIGAN, Ann Arbor, Michigan G. H. SIIULL,PRINCETON UNIVERSITY, Princeton, New Jersey E. W. SINNOTT,Barnard College, COLUMBIA UNIVERSITY, New York, New York L. H. SNYDER, OHIOSTATEUNIVERSITY, Columbus, Ohio L. J. STADLER, UNIVERSITY OF MISSOURI, Columbia, Missouri C. R. STOCKARD, CORNELL UNIVERSITY MEDICAL COLLEGE, New York, New York A. B. STOUT,NEWYORKBOTANICAL GARDEN, New York, New York A. H. STURTEVANT, CALIFORNIA INSTITUTE O F TECHNOLOGY, Pasadena, California F. B. SUMNER,SCRIPPSINSTITUTE OF OCEANOGRAPHY, LaJolla, California W. T. SWINGLE,UNITEDSTATESDEPARTMENT OF AGRICULTURE, Washington, District of Columbia W. P. THOMPSON, UNIVERSITY OF SASKATCHEWAN, Saskatoon, Canada R. B. THOMSON, UNIVERSITY OF TORONTO, Toronto, Ontario, Canada A. WEINSTEIN,JOHNSHOPKINSUNIVERSITY, Baltimore, Maryland RICHARD WELLINGTON, NEWYORKAGRICULTURAL EXPERIMENT STATION, Geneva, New York E. N. WENTWORTH, ARMOUR AND COMPANY, Chicago, Illinois 0 . E. WHITE,UNIVERSITY OF VIRGINIA, Charlottesville, Virginia P. W. WHITING,UNIVERSITY OF PITTSBURGH, Pittsburgh, Pennsylvania D. D. WHITNEY,UNIVERSITY OF NEBRASKA, Lincoln, Nebraska R. G. WIGGANS,NEWYORKSTATECOLLEGE OF AGRICULTURE, Ithaca, New York F. A. WOODS,Brookline, Massachusetts SEWALLWRIGHT,UNIVERSITY OF CHICAGO, Chicago, Illinois C. ZELENY,UNIVERSITY OF ILLINOIS, Urbana, Illinois

INTERNATIONAL CONGRESS O F GENETICS

15

M I N U T E S O F T H E F I R S T P L E N A R Y SESSION O F T H E S I X T H I N T E R N A T I O N A L CONGRESS O F GENETICS H E L D I N W I L L A R D S T R A I G H T HALL, AUGUST 24, 1932 The meeting was called to order at 8:30 P.M. by the chairman of the council, C. C. LITTLE. 1. O n motion from the floor it was voted that the next international congress of genetics shall be held in 1937. 2. On motion from the floor it was voted that the chair appoint a committee to nominate members of the permanent international committee and to report to the plenary session of August 30. The chair appointed the following committee: T . H . MORGAN, chairman, KRISTINEBONNEVIE,F. A. E . CREW,G. P. FRETS,ALESSANDRO GHIGI, HANS NACHTSHEIM. 3. The chairman announced that the council had considered selection of the place of the next congress and recommended to the plenary session that because of economic uncertainty and other conditions, authority to select

16

PROCEEDINGS O F T H E S I X T H

the place of the next congress be delegated to the permanent international committee. A motion from the floor to this effect was passed. 4. The chairman announced that several suggestions had been received that the congress appoint a standing committee to consider matters of genetical nomenclature. There was no discussion and no action was taken. 5. The chairman requested official representatives of foreign governments and organizations who had not already deposited their credentials to do so at the registration desk before Monday, August 29. 6. On motion from the floor it was voted that the chair appoint a committee to prepare the greetings of the congress to be sent to those of our absent colleagues whom we wish to remember at this time. The chair apGINI, pointed the following committee: G. H. SHULL,chairman, CORRADO and 0. WINGE. 7. Voted: that the chair appoint a committee to prepare resolutions and to present at the next plenary session of the congress, on Tuesday evening, August 30, any business which might properly come before the session. chairman, R. R. GATES,C. L. HUSKINS,CURT Committee: H. FEDERLEY, ~ N. VAVILOV,ANTONIODE STERN,A. F. SHULL, R E N VANDENDRIES, ZULUETA. 8. There being no further business brought before the meeting it was adjourned at 8:45 to meet again at 8:00 P.M. August 30.

INTERNATIONAL CONGRESS OF GENETICS

17

MINUTES O F T H E SECOND PLENARY SESSION O F T H E S I X T H INTERNATIONAL CONGRESS O F GENETICS H E L D IN BAKER AUDITORIUM, CORNELL UNIVERSITY, AUGUST 30, 1932 The meeting was called to order at 8:30 P.M. by the chairman of the council, C. C. LITTLE.The minutes of the plenary session of August 24 were approved. 1. The chairman called the roll of accredited delegates of governments and societies, and the delegates rose to receive the greetings of the congress. 2. Report of the nominating committee: The nominating committee, consisting of T. H. MORGAN, chairman, KRISTINEBONNEVIE, F: A. E. CREW, GHIGI and HANSNACHTSHEIM, presented the G. P. FRETS,ALESSANDRO following nominations for membership in the permanent international committee: Austria Belgium Denmark Finland France Germany Great Britain Italy Japan Netherlands Norway Sweden Switzerland Union of Socialistic Soviet Republics United States of America On motion from the floor the report of the nominating committee was accepted, and a unanimous vote was cast electing the above named members of the permanent international committee. The nominating committee thought best to designate a temporary chairman who might initiate correspondence within the committee since it appears unlikely that the committee will be able to meet. OTTOMOHRwas asked to serve in this capacity. The chief functions of the committee are: to represent the international congress until the council of the next congress is formed and to designate the country in which the next congress is to be

18

PROCEEDINGS O F T H E SIXTH

held. I t was understood that the committee is to reach its decision concerning the place of the next congress as representative of the congress as a whole without special consideration for the countries represented by its chairman or members. When it has reached this decision it will transfer its functions to a committee to be formed in the country in which the congress is to be held. 3. Report of the committee on greetings to absent colleagues: The report GINI and of tLe committee, consisting of G. H. SHULL,chairman, CORRADO OWINDWINGE,was presented by the chairman. The committee recommended that greetings be sent in the name of the congress to:

T o the vice presidents of the congress:

It was voted unanimously that the report of the greetings committee be accepted and that the greetings be sent. 4. Report of the resolutions committee: .The resolutions committee, consisting of HARRYFEDERLEY, chairman, R. R. GATES,C. L. HUSKINS,CURT ~ VANDENDRIES, N. VAVILOV, ANTONIO DE STERN,A. F. SHULL,R E N DE ZULUETA,presented the following motions: 5. "Resolved that a vote of thanks be extended to the officials of CORNELI UNIVERSITYfor the facilities and hospitality received." The motion was voted unanimously. 6. "Resolved that the thanks of the congress be extended to the members of the organizing committee and executive council of the congress for the tremendous amount of work and time devoted to the congress and especially to the genetical gardens and the setting up of exhibits which have been such a successful feature of the congress." The motion was voted unanimously. 7. "Resolved that a vote of thanks be extended to the local committee

I N T E R N A T I O N A L CONGRESS O F GENETICS

19

for the many ways in which they have added to the pleasure of the delegates by the arrangements made and the splendid hospitality received." The motion was voted unanimously. 8. "Resolved that in all experimental work involving investigations of a pathological kind, for example the production of cancer, or of a physiological or psychological nature with animals or plants, the importance of using genetically pure strains be emphasized." After discussion by A. F. SHULL, E. C. MACDOWELL, R. A. EMERSON, and H. K. HAYESopposing the resolusupporting it, a motion to tion and by Mme. DOBROVOLSKAIA-ZAVADSKAIA table the resolution was carried. 9. "Resolved that the congress approves the suggestion of TINETAMMES that the problem of standardizing genetical symbolism and nomenclature be reconsidered and that the genetics societies of all countries concerned be asked to appoint committees which shall cooperate and prepare recommendations tcr be published two years before and to be discussed at the next international congress of genetics." The motion was carried. There being no new business to come before the meeting, it was adjourned sine die a t 9:30 P.M.

a0

PROCEEDINGS O F T H E S I X T H

OFFICIAL DELEGATES FROM FOREIGN GOVERNMENTS Belgium: Monsieur le Professeur L. FRATEUR Monsieur le Professeur R. VANDENDRIES Chile : Sefior Don MANUEL ELGUETO Y GUERRIN Denmark: Professor Doctor 0. WINGE Finland : Professor Doctor HARRYFEDERLEY France : Monsieur Ie Professeur A. VANDEL Great Britain: GATES Professor R. RUGGLES Professor F. A. E. CREW Italy: Professor Professor Professor Professor

GHIGI Doctor ALESSANDRO Doctor CESAREARTOM Doctor FABIO FRASSETTO Doctor CORRADO GINI

Norway: Professor Doctor OTTOLOUISMOHR Spain : Seiior Don ANTONIO DE ZULUETA Y ESCOLANO

INTERNATIONAL CONGRESS O F GENETICS

21

OFFICIAL DELEGATES FROM I N S T I T U T I O N S A N D SOCIETIES Argentina : University of Buenos Aires: Seiior S. HOROVITZ Brazil : Agronomical Institute of the State of Campinas: Seiior CARLOSARNALDO KRUG British Empire : Canada-University of Laval: Monsieur 1'AbbC MAURICEPROULX University of Montreal: Monsieur le Professeur HENRIPRAT West Indies-The Empire Cotton Growing Corporation: Doctor S. C. HARLAND France : Ministry of Public Instruction: Monsieur le Professeur HENRI PRAT Germany : German Genetics Society: Professor Doctor R. GOLDSCHMIDT Kaiser Wilhelm Society for Advancement of Science: Professor Doctor R. GOLDSCHMIDT Italy : Central Institute of Statistics of Italy: Professor Doctor CORRADO GINI Committee for the Study of Population Problems: Professor Doctor CORRADO GINI Italian Society of Genetics and Eugenics: Professor Doctor CORRADO GINI, Professor Doctor CESAREARTOMand Professor Doctor ALESSANDRO GHIGI GHIGI, Ministry of National Education: Professor Doctor ALESSANDRO GINI and Professor Doctor C E ~ A R E Professor Doctor COI~RADO ARTOM Netherlands : Dutch Genetical Society: Miss Doctor J. A. LELIVELD Norway : University of Oslo: Professor Doctor KRISTINEBONNEVIE Poland : Free University of Poland: Doctor M. S K A L I ~ S K A United States of America : Department of Agriculture and Commerce, ruerto Rico: Doctor ARTUROROQUE National Institute of Social Sciences: Doctor ALBERTF. BLAKESLEE University of Florida: Doctor B. A. BOURNE

22

PROCEEDINGS O F T H E S I X T H

I N S T I T U T I O N A L MEMBERS American Fruit Growers' Association American Guernsey Cattle Club A m o u r and Company Association of Hawaiian Pineapple Canners Brown University Bucknell University California Institute of Technology Carleton College Carnegie Corporation of New York Carnegie Endowment for International Peace Carnegie Institution of Washington Columbia University Cornell University Cornell University M-edical College Dartmouth College Gallatin Valley Seed Company General Electric Company General Mills, Incorporated Goucher College Harvard University Hawaiian Sugar Planters' Association Institute of Forest Genetics Johns Hopkins University Long Island University Minnesota Crop Improvement Association New York College of Agriculture at Cornell University Pillsbury Flour Mills Company Russell-Miller Milling Company Smithsonian Institution, United States National Museum Texas Agricultural Experiment Station Tri-State Soft Wheat Improvement Association University of Chicago University of Missouri CONTRIBUTING MEMBERS JAMES,$100 Mrs. WORTHAM E. C. MACDOWELL, $300

JAMESF. PORTER,$100 E. PARMALEE PRENTICE,$750

INTERNATIONAL CONGRESS O F GENETICS

23

MEMBERS O F T H E CONGRESS The total paid membership in the congress is 856. Of this number 33 are institutional members. The registration at Ithaca totals 562. The enrollment from countries other than the United States is 163, of which 103 registered at Ithaca. Members marked with an asterisk (*) did not register at Ithaca. ENROLLMENT BY COUNTRIES

Algeria, 1 Argentina, 3 Australia, 1 Belgium, 1 Brazil, 1 British West Indies, 1 Canada, 49 Chile, 1 China, 11 Cuba, 1 Czechoslovakia, 1 Denmark, 2 Finland, 1 France, 9 Germany, 11 Great Britain, 15 India, 1 Italy, 8

Japan, 6 Lithuania, 1 Morocco, 1 Netherlands, 6 Netherlands East Indies, 1 Norway, 6 Poland, 5 Portugal, 2 Roumania, 3 South Africa, 1 Spain, 3 Sweden, 1 Switzerland, 2 Tunis, 1 Union Socialistic Soviet Republics, 5 United States, 693 Uruguay, 1 Total, 856

MEMBERS O F THE SIXTH INTERNATIONAL CONGRESS O F GENETICS

*AAMODT,0. S., Dept. of Field Crops, Univ. Alberta, Edmonton, Alta., Canada. *AASE, HANNAH,College Court, Pullman, Wash. *ACHESON, JR., W . W., 5029 Marewood PI., Pittsburgh, Pa. *ADOLPH,E. F., Physiological Lab., Univ. Rochester, Rochester, N. Y. * A ~ o o s SOLOMON, , 209 South St., Boston, Mass. R., Dept. Genetics, Univ. Wisconsin, Madison, Wis. ALBRECHT,HERBERT ALDERMAN, W . H., University Farm, St. Paul, Minn. *ALDRICH,JR., EDWARD K., 155 Brown St., Providence, R. I. *ALLEN,C. E., Biology Bldg., Univ. Wisconsin, Madison, Wis. *ALLEN,EZRA,4367 142 St., Flushing, N. Y.

24

PROCEEDINGS OF THE SIXTH

AMERICAN GENETIC ASSOCIATION, 306 Victor Bldg., Washington, D. C. AMERICAN FRUIT GROWERS' ASSOCIATION. AMERICAN GUERNSEY CATTLECLUB,Peterboro, N. H. *ANDERSON, Waller-Franklin Seed Co., Guadalupe, Calif. ANDERSON, ARTHUR, College Agric., Lincoln, Nebr. ANDERSON, EDGAR, Arnold Arboretum, Jamaica Plain, Boston, Mass. *ANDERSON, W. S., Univ. Kentucky, Lexington, Ky. *ARCINIEGA, DE, A., El Jefe Del Ser., Pecuario, A1 Apartado de Correos Num. 53, Bilbao, Spain. ARMOUR AND COMPANY, Chicago, Ill. ARMSTRONG, J. M., Dept. Botany, McGill Univ., Montreal, Quebec, Canada. ARNASON, THOMAS J., 26 N. Lake St., Madison, Wis. ARONESCU, ALICE,600 W. 161 St., New York, N. Y. ARTOM,CESARE, Regis Universiti di Pavia, Pavia, Italy. ASDELL,SYDNEY ARTHUR, Lab. Anim.Nutr., State College Agric., Cornell Univ., Ithaca, N. Y. *ASMUNDSON, V. S., Univ. California, Berkeley, Calif. OF HAWAIIAN PINEAPPLE CANNERS. ASSOCIATION *ATKINSON, ALFRED,Montana State College, Bozeman, Mont. *AVERY,PRISCILLA, 29 Westall Ave., Oakland, Calif. *BAASHUS-JESSEN, J., Parkwien 8 111, Oslo, Norway. E. B., Univ. California, Berkeley, Calif. BABCOCK, *BAEHR,VENCESLAS DE,Univ. Warsaw, Warsaw, Poland. BAGG,HALSEY, J., Memorial Hosp., 2 W. 106 St., New York, N. Y. *BAILEY,L. H., Ithaca, N. Y. BAMFORD, RONALD, College Ave., College Park, Md. BANCROFT, WILLIAMA., Garland, Me. BANGSON, JOHNS., Dept. Biology, Berea College, Berea, Ky. *BANKER, HOWARD J., 14 Myrtle Ave., Huntington, N. Y. BANTA,A. M., Dept. Zoology, Brown Univ., Providence, R. I. BARON, A. L., 212 E. 53 St., Brooklyn, N. Y. BARRONS, KEITHC., 1455 Grantham Ave., St. Paul, Minn. *BARROWS, EDWARD FLETCHER, 220 N. East St., Monmouth, Ore. BARROWS, FLORENCE L., R.F.D.2, Box 85, Stafford Springs, Conn. *BARTLETT, H. H., Univ. Michigan, Ann Arbor, Mich. *BARTSCH, PAUL,Smithsonian Inst., U.S. National Museum, Washington, D.C. *BAUMGARTNER, W. J., 1209 Ohio St., Lawrence, Kans.

INTERNATIONAL CONGRESS O F GENETICS

BEADLE,G. W., California Inst. Tech., Pasadena, Calif. BEAUMONT, J. H., University Park, Md. *BECKWITH, CORAJ., Vassar College, Poughkeepsie, N. Y. BEERS,CATHERINE V., BOX60,411 W. 116 St., New York, N. Y. BELFIELD,'EILEEN, 1 Pentley Park, Welwyn Garden City, Herts., England. BELFIELD,SYDNEY, Mrs., 1 Pentley Park, Welwyn Garden City, Herts., England. *BELLING, JOHN,Univ. California, Berkeley, Calif. *BENNETT, H. W. N., 913 Elm St., Manchester, N. H. *BERGER, CHARLES A., Woodstock College, Woodstock, Md. BERGNER, A. DOROTHY, Carnegie Inst., Cold Spring Harbor, N. Y. BESLEY,HELEN,457 Park Ave., Rochester, N. Y. BETTEN,CORNELIUS, 3 The Circle, Ithaca, N. Y. BIDDLE,RUSSELLL., College City New York, 139 St., New York, N. Y. BIRD,J. S., 304 W. 21st St., Hays, Kans. BITTNER,J O H N J., Jackson Memorial Lab., Bar Harbor, Me. *BIXBY,WILLARDG., 32 Grand Ave., Baldwin, Nassau Co., N. Y. *BLAIR,J. C., 125 New Agric. Bldg., Urbana, Ill. BLAKESLEE, ALBERT F., Carnegie Inst., Cold Spring Harbor, N. Y. *BLARINGHEM, L., kcole Normale Supkrieure, 45 Rue d'Ulm, Paris, France. *BOERGER, ALBERTO, La. Estanzuela, Dpto. Colonia, Rep. 0 . del Uruguay. *BOEUF,FELICIEN,Service Botanique de Tunisie, L'Ariana, near Tunis, Tunis. *BONNETT, R. K., Washburn-Wilson Seed Co., Moscow, Idaho. BONNEVIE, KRISTINE,Univ. Oslo, Oslo, Norway. *BOOYE, B. T., 101 California St., San Francisco, Calif. BORODIN, D. N., Woods Hole, Mass. BORST,H. L., Dept. Farm Crops, Ohio State Univ., Columbus, Ohio. BOSTIAN, C. H., North Carolina State College, Raleigh, N. C. BOURNE, B. A., Florida Agric. Expt. Sta., Bella Glade, Fla. BOWERS, CLEMENT GRAY,Maine, Broome Co., N. Y. BOWSTEAD, J. E., Univ. Alberta, Edmonton, Alta., Canada. BOYDEN, ALAN,Stelton, N. J. *BRANCH, HAZELE., Univ. Wichita, Wichita, Kans. BRANDT, A. E., 2917 Oakland, Ames, Iowa. BREGGER, JOHNT., Bangor, Mich. BREHME,KATHERINE S., 545 W. End Ave., New York, N. Y. BREWBAKER, H. E., 900 Elizabeth St., Fort Collins, Colo. BRIDGES, C. B., California Inst. Tech., Pasadena, Calif.

26

PROCEEDINGS O F T H E SIXTH

BRIESE,ESTELLA,1112 Second St. N. W., Rochester, Minn. BRIGGS,FRED N., Univ. California, Davis, Calif. BRINK,R. A., Dept. Genetics, Univ. Wisconsin, Madison, Wis. BRINSMADE, JR., J. C., N. G. P. Field Sta., Mandan, N. Dak. BRITTINGHAM, WILLIAMH., 1509 Ward Terrace, Portsmouth, Va. *BROWN,HARRYB., University Station, Baton Rouge, La. BROWNUNIVERSITY, Providence, R. I. BROWNELL, WALTERD., Little Compton, R. I. *BROZEK,A., Plant Physiological Inst., 433 Benatecka St., Praha 11, Czechoslovakia. BRUNSON, A. M., Kansas State College, Manhattan, Kans. BRYAN,A. A., Iowa State College, Ames, Iowa. BRYAN,R., P.O. BOX402, Hilo, Hawaii, T. H. *BRYAN, W. E., Univ. Arizona, Tucson, Ariz. BUCHHOLZ, JOHNT., Univ. Illinois, Urbana, 111. BUCKNELL UNIVERSITY, Lewisburg, Pa. BURHOE,SUMNER O., Dept. Zoology, Univ. Maryland, College Park, Md. BURKHOLDER, W. H., Cornell Univ., Ithaca, N. Y. BURLINGAME, LEONAS L., Room 426, Stanford University, Calif. BURNHAM, CHARLESR., Dept. Field Crops, Univ. Missouri, Columbia, Mo. *BURNS,ROBERTH., Dept. Wool, Univ. Wyoming, Laramie, Wyo. *BURR,CHARLES W., 1918 Spruce, Philadelphia, Pa. BUSSELL,FRANK P., Cornell Univ., Ithaca, N. Y. BUTLER, A. N. LOWE,11 Fairmont Ave., Ottawa, Ont., Canada. *BUTLER,NICHOLASMURRAY, 405 W. 117 St., New York, N. Y. BYLMER, H. T., Ambon, Moluccas, Netherlands East Indies. *CALDWELL, OTIS W., 433 W. 123 St., New York, N. Y. CALIFORNIA INSTITUTE OF TECHNOLOGY, Pasadena, Calif. *CAMPBELL, M. H., 32 Wilson St., Burlington, Vt. *CAPPER,ARTHUR, Topeka, Kans. CARLETON COLLEGE, Northfield, Minn. CARNEGIE CORPORATION OF NEWYORK. CARNEGIE ENDOWMENT FOR INTERNATIONAL PEACE. CARNEGIE INSTITUTION OF WASHINGTON, 16th and M Sts. N.W., Washington, D. C. CAROTHERS, E. ELEANOR, Univ. Pennsylvania, Philadelphia, Pa. CARTER,GEORGE S., Clinton, Conn. CARTLEDGE, J. LINCOLN, 203 Biology Hall, Univ. Pittsburgh, Pittsburgh, Pa.

INTERXATIONAL CONGRESS O F GENETICS

CASTLE,W. E., Bussey Inst., Forest Hills, Boston, Mass. *CATTELL,J. MCKEEN,Garrison, N. Y. *CATTELL,J. MCKEEN,Mrs., Garrison, N. Y. *CAULLERY, MAURICE, Boulevard Raspail 105, Paris 6, France. *CHALMERS,R. ELIZABETH,Dept. Anatomy, Univ. Pittsburgh, Pittsburgh, Pa. CHANG,H . T., Dept. Biology, Univ. Soochow, Soochow, China. *CHANG,TEHJEN, Agric. Expt. Sta., Yenching Univ., Peiping, China. CHAPMAN, ARTHURB., 1427 University Ave., Madison, Wis. CHARLES, DONALD R., Dept. Zoology, Columbia Univ., New York, N. Y. CHENEY,RALPH HOLT, Long Island Univ., 300 Pearl St., Brooklyn, N. Y. CHESLEY,PAUL,BOX8, Schermerhorn Hall, Columbia Univ., New York, N. Y. CHILD,GEORGE P., New York Univ., 100 Washington Sq. E., New York, N. Y. CHOU,C. Y., National Central Univ., Nanking, China. CHRISTIAN, C. STUART,2248 Carter Ave., St. Paul, Minn. CHROBOCZEK, EMIL, Skierniewice, Poland. CHUNG,C. H., Dept. Botany, Ohio State Univ., Columbus, Ohlo. CLARK,ARTHURB., 205 Church St., New Haven, Conn. H., Bussey Inst., Forest Hills, Boston, Mass. CLARK,FRANK CLARKE,ALFREDE., Dept. Genetics, Univ. California, Berkeley, Calif. CLAUSEN, J., Carnegie Inst., Stanford University, Calif. CLAUSEN, ROYE., 205 Hilgard Hall, Berkeley, Calif. CLELAND, R. E., Goucher College, Baltimore, Md. *CLOUDMAN, ARTHURM., BOX558, Jackson Memorial Lab., Bar Harbor, Me. *COE, WESLEYR., Yale Univ., New Haven, Conn. COFFMAN, FRANKLIN A., 114 N. Va. Ave., Lyon Village, Clarendon, Va. COLBY,ARTHURS., Univ. Illinois, Urbana, Ill. COLE,L. J., Dept. Genetics, Univ. Wisconsin, Madison, Wis. COLIN,E. C., 5527 Kimbark Ave., Chicago, 111. *COLLINS,G. N., Lanham, Md. COLLINS,J. L., A. H. P. C., Univ. Hawaii, Honolulu, T. H . *COLTON,HAROLDS., BOX307, Flagstaff, Ariz. COLUMBIA UNIVERSITY, New York, N. Y. *CONKLIN,E. G., Dept. Biology, Princeton University, Princeton, N. J. *COOK,0. F., Lanham, Md. COOK,ROBERT C., Victor Bldg., Washington, D. C.

28

PROCEEDINGS O F THE SIXTH

COOPER,DELMERC., Dept. Genetics, Univ. Wisconsin, Madison, Wis. CORNELL UNIVERSITY, Ithaca, N. Y. CORNELL UNIVERSITY MEDICALCOLLEGE, New York, N. Y. *COSGROVE, E. B., Minnesota Valley Canning Co., Le Sueur, Minn. *COVILLE,FREDERICK V., Dept. Agric., Washington, D. C. *COY,HAZEL,R.F.D.l, Walbridge, Ohio. CRAIG,EDNA,92 Montgomery St., Newburgh, N. Y. CRAIG,W. T., Cornell Univ., Ithaca, N. Y. CRAMPTON, H. E., Barnard College, Columbia Univ., New York, N. Y. *CRAWFORD, J. L., Thaison Bldg., Laredo, Tex. CRAWFORD, MARYM., 333 E. 57 St., New York, N. Y. CREIGHTON, HARRIET B., 408 Dryden Road, Ithaca, N. Y. CREW,F. A. E., Inst. Animal Genetics, Kings Bldg., W. Mains Rd., Edinburgh, Scotland. *CREW,WILLIAMH., Briarcliff Farms, Pine Plains, N. Y. CROFTS,JOHNT., 6424 Kenwood Ave., Chicago, Ill. *CROZIER, W. J., Dept. Physiology, Harvard Univ., Cambridge, Mass. *CRUMBLEY, JAMESB., Dept. Genetics, Iowa State College, Ames, Iowa. CURRENCE, T. M., University Farm, St. Paul, Minn. CURTIS,LAWRENCE C., Connecticut Agric. Expt. Sta., New Haven, Conn. *CURTIS,MAYNIER., 1145 Amsterdam Ave., New York, N. Y. CUTLER,G. H., Purdue Univ., W. Lafayette, Ind. *DAHLBERG, HENRYW., Great Western Sugar Co., Denver, Colo. *DALE,ERNESTE., Dept. Biology, Union College, Schenectady, N. Y. *DANFORTH, C. H., Dept. Anatomy, Harvard Medical School, Boston, Mass. DARLINGTON, C. D., John Innes Horticultural Inst., London, S. W. I. G., England. DARROW, GEORGE M., Glenn Dale, Md. DARTMOUTH COLLEGE, Hanover, N. H. C. B., Station for Experimental Evolution, Cold Spring HarDAVENPORT, bor, N. Y. DAVENPORT, GERTRUDE C., Cold Spring Harbor, N. Y. DAVID,PAUL R., Storrs Agric. Expt. Sta., Storrs, Conn. DAVIS,BARBARA, 1001 St. Paul St., Baltimore, Md. DAVIS, BRADLEYMOORE,Botanical Lab., Univ. Michigan, Ann Arbor, Mich. DAVIS,DONALDW., 349 W. Scotland St., Williamsburg, Va. DAVIS,MARYELEANOR, 1922 Mt. Royal Terrace, Baltimore, Md. DAWSON,WALKERM., Dept. Animal Husb., Univ. Illinois, Urbana, Ill.

INTERNATIONAL CONGRESS O F GENETICS

29

DEAKIN,ALAN,Central Expt. Farm, Ottawa, Ont., Canada. *DEDERER,PAULINE H., Connecticut College, New London, Conn. DEMEREC,M., Station for Experimental Evolution, Cold Spring Harbor,

N. Y. DEMEREC, MARY,Cold Spring Harbor, N. Y. DENTON,CAROLINE A., 2019 St. Paul St., Baltimore, Md. DERMEN,HAIG,Bussey Inst., Jamaica Plain, Mass. *DICK, G. A., 39 St., and Woodland Ave., Philadelphia, Pa. *DICKERSON, L. M., South Greenwood, Lebanon, Tenn. DOBROVOLSKAIA-ZAVADSKAIA, N. A., Inst. de Radium, 26 Rue d'Ulm, Paris, France. THEODOSIUS G., California Inst. Tech., Pasadena, Calif. DOBZHANSKY, DODGE,BERNARD O., New York Botanical Garden, Bronx Park, N. Y. *DOLLEY,JR., WILLIAML., Univ. Buffalo, Buffalo, N. Y. DOMM,L. V., Whitman Lab., 5700 Ingleside Ave., Chicago, Ill. *DONALDSON, HENRY,Wistar Inst., Philadelphia, Pa. *DOUGHTY, L. R., 54 White St., Derby, England. DOVE,W . J., 142 Park St., Orono, Me. D u B o ~ sA , N N EM., Carnegie Inst., Johns Hopkins Medical School, Baltimore, Md. D u B o ~ sETHEL, , 1449 Broadway, Hewlett, N. Y. DUCHEMIN, WARREN J., Prince of Wales College, Charlottetown, Prince Edward Island, Canada. J., 9 Rue de CondC, Paris 6, France. DUFRENOY, DUNN,L. C., Dept. Zoology, Columbia Univ., New York, N.. Y. DUNNING,WILHELMINA F., 1145 Amsterdam Ave., New York, N. Y. DURHAM, GEORGE B., Dept. Hort., Rhode Island State College, Kingston, R. I. EARL,R. O., Queen's Univ., Kingston, Ont., Canada. EAST,E. M., Bussey Inst., Forest Hills, Boston, Mass. EATON,ORSONN., 4 Carroll, Hyattsville, Md. EINSELE,WILHELM,24 Rheinstrasse, Karlsruhe i.B., Germany. ELGUETO, MANUEL,Casilla 53F, Santiago, Chile. ELLIS, ZENASH., Fair Haven, Vt. EMERSON, R. A., Cornell Univ., Ithaca, N. Y. EMERSON, STERLING, California Inst. Tech., Pasadena, Calif. EMSWELLER, S. L., University Farm, Davis, Calif. ERLANSON, EILEENW., Univ. Michigan, Ann Arbor, Mich. *ERNST,A. Botanisches Inst., Ziirich, Switzerland. ESTABROOK, ARTHURH., 1 Bank St., New York, N. Y.

30

PROCEEDINGS OF THE SIXTH

*EVERS,ROBERT A., 1018 Bluemont Ave., Manhattan, Kans. EWING,E . C., Scott, Miss. EYSTER,W. H., Bucknell Univ., Lewisburg, Pa. *FAIRCHILD, DAVIDG., U . S. Dept. Agric., Washington, D. C. FEDERLEY, HARRY,Univ. Helsingfors, Helsingsfors, Finland. FEKETE,ELIZABETH, Roscoe B. Jackson Memorial Lab., Bar Harbor, Me. FELDMAN, HORACE, Lab. Vertebrate Genetics, Univ. Michigan, Ann Arbor, Mich. FELL, HONORB., Strangeways Research Lab., Cambridge, England. FENG,C. C., National Central Univ., Nanking, China. FENG,C. F., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. FERGUSON, MARGARET C., 46 Dover Rd., Wellesley, Mass. *FERNANDEZ, MIGUEL,Univ. Casilla, Correo F2, Cordoba, Argentina. FISHER,R. A., Rothamsted Expt. Sta., Harpenden, England. FISK, EMMA L., Dept. Botany, Univ. Wisconsin, Madison, Wis. *FITCH,C. L., Iowa State Vegetable Growers Assoc., Ames, Iowa. *FLORY,JR., WALTERS., BOX 176, Bridgewater, Va. FORBES,WILLIAMT . M., Roberts Hall, Cornell Univ., Ithaca, N. Y . FORE,R. E., 508 W. Elm, Urbana, Ill. *FOSDICK,ELINORWI-IITNEY,Smith College, Northampton, Mass. FOWLDS,MATTHEW,Brookings, S. Dak. FRASER,ALLANC., 119 The Parkway, Ithaca, N. Y. FRASSETTO, FABIO,R. Univ. di Bologna, Bologna, Italy. FRENCH,A. P., Massachusetts State College, Amherst, Mass. FRETS,G. P., HOSP.for Mental Diseases, Maasoord-Poortugaal near Rotterdam, Netherlands. FRETS,G. P., Mrs., Hosp. f o r Mental Diseases, Maasoord-Poortugaal near Rotterdam, Netherlands. *FRIESNER, RAYC., Butler Univ., Indianapolis, Ind. FRISCH,JOHN A., Loyola College, Baltimore, Md. FROST,H. B., Citrus Expt. Sta., Riverside, Calif. FULLMER, E. L., 30 Fifth Ave., Berea, Ohio. *GAGER,C. STUART,Brooklyn Botanic Garden, 1000 Washington Ave., Brooklyn, N. Y. GAINES,E . F., Route 1, Pullman, Wash. *GAINES,W. L., Dept. Dairy Husb., Univ. Illinois, Urbana, Ili. GAISER,L. O., McMaster Univ., Hamilton, Ont., Canada. GALLATIN VALLEYSEEDCOMPANY, Bozeman, Mont. GARBER,R. J., West Virginia Univ., Morgantown, W . Va. GARLAND, SARAHMAY,606 W. 122 St., New York, N. Y.

INTERNATIONAL CONGRESS O F GENETICS

31

GATES,R. RUGGLES, Univ. London, King's College, Strand W.C.2, London, England. GATES,WILLIAMH., Univ. Sta., Baton Rouge, La. H., Rockefeller Inst. for Medical Research, Princeton, GAY,ELIZABETH

N. J. GENERAL ELECTRICCOMPANY, Schenectady, N. Y. GENERAL MILLS, INCORPORATED, Minneapolis, Minn. GENETICS LAB.,Dept. Zoology, 806 Schermerhorn Hall Extens., Columbia Univ., New York, N. Y. *GEROULD, J. H., Dartmouth College, Hanover, N. H. GERSHOY, ALEXANDER, 307 Colchester Ave., Burlington, Vt. GHIGI,ALESSANDRO, Rettore del'Universit8, Bologna, Italy. GIBSON,ROYE., 937 Phoenix St., S. Haven, Mich. GILMORE,KATHRYN, Univ. Pittsburgh, Pittsburgh, Pa. Societa Italiana di Genetica e Eugenica, 10 Via Delle GINI, CORRADO, Terme, Rome, Italy. GLASS,H . BENTLEY,Anatomical Inst., The University, Oslo, Norway. GLASS,LEROYC., Dept. Zoology, Univ. Idaho, Moscow, Idaho. GLOYER,W. O., New York Agric. Expt. Sta., Geneva, N . Y. GOLDSCHMIDT, R., Kaiser Wilhelm-Institut fiir Biologie, Berlin-Dahlem, Germany. *GOLDSMITH, WILLIAMW., Municipal Univ. Wichita, Wichita, Kans. *GONDIN,IDELINO,Rua Ponta Delagada, 49, Lisboa, Portugal. * G o o c ~ MARJORIE, , 615 N. Wolfe St., Baltimore, Md. GOOD,HOWARD, Fable Bros., Westport, Conn. GOODALE, H . D., 257 W . Main St., Williamstown, Mass. H. B., Wesleyan Univ., Middletown, Conn. GOODRICH, GOODRICH, JULIA IRENE,Carnegie Inst., Cold Spring Harbor, N. Y. *GOODSPEED, T. H., Univ. California, Berkeley, Calif. C. D., Dept. Genetics, Univ. Wisconsin, Madison, Wis. GORDON, McGraw Hall, Cornell Univ., Ithaca, N. Y. GORDON, MYRON, GOT^, S E I T A R 660 ~ , Nishi Sugamo, Tokyo, Japan. GOUCHER COLLEGE,Baltimore, Md. GOWEN,J. W., Rockefeller Inst., Princeton, N. J. *GRAHAM, W. R., Ontario Agric. College, Guelph, Ont., Canada. GREEN,C. V., Roscoe B. Jackson Memorial Lab., Bar Harbor, Me. P. W., Univ. California, Davis, Calif. GREGORY, GRIFFE,F., 35 Park St., Orono, Me. GROSSMAN, EDGARF., 546 113 St., New York, N. Y. GUMBEL,E. J., 39 Beethovenstr., Heidelberg, Germany.

32

PROCEEDINGS OF THE SIXTH

GUTHRIE,MARYJANE,Lefevre Hall, Columbia, Mo. GUYER,M. F., Univ. Wisconsin, Madison, Wis. HAINES,GEORGE, Office of Experiment Stations, Washington, D. C. HALDANE, J. B. S., Roebuck House, Ferry Lane, Cambridge, England. HALL,G. O., 513 Dryden Rd., Ithaca, N. Y. HAMILTON, L. H., Macdonald College, Quebec, Canada. HAMMOND, JOHN,School of Agric., Univ. Cambridge, Cambridge, England. *HAMMOND, WARNERS., 2523 13 St. N.W., Washington, D. C. HANNA,W. F., Dominion Rust Research Lab., Winnipeg, Man., Canada. HARLAND, S. C., Cotton Research Sta., Trinidad, British West Indies. HARNLY, MORRISH., 65 Morton St., New York, N. Y. HARRIS,REGINALD G., Biological Lab., Cold Spring Harbor, N. Y. *HARRISCO., INC.,JOSEPH,Coldwater, N. Y. *HARRISON, ROSS G., Osborn Zoological Lab., Yale Univ., New Haven, Conn. *HARTMANN, CHARLES J., BOX32, Saratoga Springs, N. Y. HARVARD UNIVERSITY, Cambridge, Mass. *HASEGAWA, NOBUMI,Utsunomiya Agric. College, Utsunomiya KotoNorin-Gakko, Japan. *HASKELL, HARRYG., 9048 duPont Bldg., Wilmington, Del. *HASKELL,HENRYS., 405 W. 117 St., New York, N. Y. HATFIELD,WILSONC., State Dept. Agric., Dover, Del. ASSOCIATION, Honolulu, T . H. HAWAIIANSUGARPLANTERS' HAWRYLUK, THOMAS,3300 Bailey Ave., Bronx, N. Y. HAYDEN,MARGARET A., Wellesley College, Wellesley, Mass. HAYES,H . K., University Farm, St. Paul, Minn. A., Massachusetts State College, Amherst, Mass. HAYS,FRANK HEARNE,MARIE,Dept. Botany, McGill Univ., Montreal, Quebec, Canada. HEH, C. M., Univ. Nanking, Nanking, China. HEIMBURGER, CARLCONSTANTINE, 115 Eddy St., Ithaca, N. Y. HEINICKE,ARTHURJOHN,Dept. Pomology, Cornell Univ., Ithaca, N. Y. HEIZER,E. E., Ohio State Univ. College of Agric., Columbus, Ohio. *HELD,C. E., 39 Hawthorne Ave., Akron, Ohio. HELLMAN, MILO, 57 W . 57th St., New York, N. Y. *HENDERSON, W. W., Dept. Entomology and Zoology, Utah State Agric. College, Logan, Utah. HENNING,WILLIAML., 203 Agricultural Bldg., State College, Pa. *HERMINIO, GIORDANO, Fitotecnial-M. de Agric., Paseo Colon 97, Buenos Aires, Argentina.

INTERNATIONAL CONGRESS O F GENETICS

33

HERRIOTT,FRANK W., Mrs., 425 Riverside Dr., New York, N. Y. HERSH,A. H., Dept. Biology, Western Reserve Univ., Cleveland, Ohio. *HESS,WALTERN., Hamilton College, Clinton, N. Y. HETZER,HERBERTO., Iowa State College, Ames, Iowa. HEYN,HANSH., Blandy Expt. Farm, Univ. Virginia, Boyce, Va. HILL, E . LILLIAN,345 Dryden Rd., Ithaca, N. Y. HILL, HENRYE., E . Killingly, Conn. HILL,J. BEN,Dept. Botany, Pennsylvania State College, State College, Pa. *HILLEBRECHT, HERBERT E., 5108 Wentworth Ave., Chicago, Ill. HINMAN, R. B., Dept. Animal Husb., Cornell Univ., Ithaca, N. Y. HODSON,CORAB. S., Mrs., 406 Fulham Rd., London, S. W. 6, England. T. L., Kansas Flour Mills Corp., Kansas City, Mo. *HOFFMAN, *HOFFMAN, W . H., Cerro 593 Labovatono Finlay, Havana, Cuba. HOLMES,S. J., Univ. California, Berkeley, Calif. *HOOTON, ERNEST A., Peabody Museum, Cambridge, Mass. HOOVER, MAXMANLEY,Univ. West Virginia, Morgantown, W . Va. *HORLACHER, W . R., Agric. and Mech. College of Texas, College Station, Tex. HOROVITZ, SALOMON, Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. HOUGHTALING, HELEN,Coytesville, N. J. HOUK,W . G., 4 0 Judson St., Canton, N. Y. HOWLETT,F. S., Ohio Agric. Expt. Sta., Wooster, Ohio. *HOYT,WILLIAMDANA,Washington and Lee Univ., Lexington, Va. *HUBER,G. CARL,1330 Hill St., Ann Arbor, Mich. HUBERT,KURT,Julius Kuhns St. 31, Halle a/dJS, Germany. ALFREDF., Washington Square College, Washington Sq. E., HUETTNER, New York, N. Y. HULL,FRED H., Agric. Expt. Sta., Gainesville, Fla. HUME,EDWARD P., 606 W . 122 St., New York, N. Y. HUMPHREY, ESTELLA,1 8 Shailer St., Brookline, Mass. HUMPHREY, HARRYB., Cabin John, Md. HUMPHREY, LLEWELLYN M., Dept. Botany, Iowa State College, Ames, Iowa. HUNT,BARBARA, 11 Leighton Rd., Wellesley, Mass. R., 501 Sunset Lane, E . Lansing, Mich. HUNT,HARRISON *HUNT,MILDREDB., 327 S. Highland Ave., Los Angeles, Calif. *HUNTER,JAMES A., T'unghsien, Peiping, China. *HUNTINGTON, ELLSWORTH, Yale Univ., New Haven, Conn. HURST,C. C., 4 Brookside, Cambridge, England. HUSKINS,C. L., Dept. Genetics, McGill Univ., Montreal, Quebec, Canada.

34

PROCEEDINGS O F THE SIXTH

HUSTED,LADLEY,The Blandy Expt. Farm, Univ. Virginia, Boyce, Va. *HUTCHISON, C. B., College Agric., Univ. California, Berkeley, Calif. HUTT,F. B., Poultry Div., Univ. Farm, St. Paul, Minn. *HYMAN,HARRIET,35 16 Ave., Columbus, Ohio. IBSEN,H . L., Kansas State College, Manhattan, Kans. *ILLICK,J. T., Univ. Nanking, Nanking, China. IMMER,F. R., University Farm, St. Paul, Minn. INSTITUTE OF FOREST GENETICS,Placerville, Calif. IRWIN,M. R., Dept. Genetics, Univ. Wisconsin, Madison, Wis. JAAP,ROBERTGEORGE, 813 Grant St., Madison, Wis. JACKSON, V. W., 737 McMillan, Winnipeg, Man., Canada. Mrs., St. James, Mo. *JAMES,WORTHAM, JEFFERS,KATHARINE R., Radnor Hall, Bryn Mawr, Pa. *JEFFREY,EDWARD C., 47 Lakeview Ave., Cambridge, Mass. JENKINS,MERLET., 1120 Harding Ave., Ames, Iowa. *JENNINGS, H . S., Johns Hopkins Univ., Baltimore, Md. JOHNSHOPKINSUNIVERSITY, Baltimore, Md. JOHNSON, R. E., Dept. Botany, McGill Univ., Montreal, Quebec, Canada. *JONES,BRUCEP., 315 Plant Science Bldg., Ithaca, N. Y. JONES,D. F., BOX1106, New Haven, Conn. JONES,HENRYA., University Farm, Davis, Calif. *JONES,JENKINSW., Cereal Crops and Diseases, Bureau Plant Industry, U. S. Dept. Agric., Washington, D. C. JUCCI, CARLO,R. Universiti degli Studi, Scuola di Farmacia, Sassari, Italy. JUHN,MARY,Whitman Lab., 5700 Ingleside Ave., Chicago, Ill. *JULL,MORLEY A., Bureau Animal Industry, U. S. Dept. Agric., Washington, D. C. *KADAM,BARBURAO S., Rice Breeding Sta., Karjat (Kolaba), Bombay, India. *KAGAWA, F., Utsunomiya Agric. College, Utsunomiya Koto-Norin-Gakko, Japan. KAISER,SAMUEL, 960 43 St., Brooklyn, N. Y. KALISS,NATHAN, 2158 E. 17 St., Brooklyn, N. Y. KAMENOFF, RALPHJ., 2023 Valentine Ave., Bronx, N. Y. *KANDER, IRVING J., New Paltz, N. Y. KAUFMANN, BERWIND P., BOX2042, University P. O., Ala. *KEARNEY, THOMAS H., Cosmos Club, Washington, D. C. KEELER,CLYDEE., Bussey Inst., Forest Hills, Boston, Mass. DAVID,College Agric., Univ. Nebraska, Lincoln, Nebr. KEIM, FRANKLIN

INTERNATIONAL CONGRESS O F GENETICS

KELLEY,TRUMAN L., Stanford University, California. *KELLOGG, JOHNHARVEY, Battle Creek Sanitarium, Battle Creek, Mich. KELLOGG, VERNON, National Research Council, Washington, D. C. KELLY,JAMESP., Dept. Botany, Pennsylvania State College, State College, Pa. KEMP,TAGE,Univ. Inst. for Gen. Path., Juliane Mariesrej 22, Copenhagen, Denmark. KENDALL, JAMES,Sterlington, N. Y. KENT,0 . D., Libertyville, Ill. KERNS,K. R., Beymester 20, Berlin, Steglitz, Germany. KEY, WILHELMINE E., Fernwold, Somers, Conn. *KEZER,ALVIN,Colorado Agric. College, Fort Collins, Colo. *KHADILKER, T. R., 506 Dryden Rd., Ithaca, N. Y. *KIHARA, H., Dept. Agric., Kyoto Imperial Univ., Kyoto, Japan. KING,HELEND., Wistar Inst., 36 and Woodland Ave., Philadelphia, Pa. KING,ROBERT L., Dept. Zoology, Univ. Iowa, Iowa City, Iowa. KING,S. P., 129 Linden Ave., Ithaca, N. Y. *KIRBY,VIOLETM., Canyon Ranch, R.F.D. 2, Redlands, Calif. KIRK,L. E., Dominion Experimental Farm, Ottawa, Ont., Canada. KNOWLES, EMERSON G., McMaster Univ., Hamilton, Ont., Canada. KNOX,C. W., U. S. Dept. Agric. Expt. Farm, Beltsville, Md. KNUDSON, LEWIS,Plant Physiology, College Agric., Ithaca, N. Y. *KOFOID,CHARLESA., Life Science Bldg., Univ. California, Berkeley, Calif. *KOMAI, TAKU,Zoological Inst., Kyoto Imperial Univ., Kyoto, Japan. *KOONCE, DWIGHT,Fort Lewis School Agric., Hesperus, Colo. KOPF,KENNETH, Dept. Genetics, Iowa State College, Ames, Iowa. KORNHAUSER, S. I., Biological Lab., Cold Spring Harbor, N. Y. KRANTZ, FRED A., University Farm, St. Paul, Minn. *KRAUSS, F. G., Univ. Hawaii, Honolulu, T. H. *KRUEGER, WILLIAMF., BOX102, Sca. A, Toledo, Ohio. A., 210 College Ave., Ithaca, N. Y. KRUG,CARLOS *KULESHOV, N., Inst. Applied Botany, 44 Herzen St., Leningrad, Union of Socialistic Soviet Republics. LAXMAN GOPAL,Genetics Div., University Farm, St. Paul, KULKARNI, Minn. KWAN,CHIACHI, Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. LAANES, THEOPHIL, Carnegie Inst., Cold Spring Harbor, N. Y. *LAGOMARSINO, EARL,University Farm, Davis, Calif. LAMARTINE-YATES, Rose, Barley near Royston, Herts., England.

36

PROCEEDINGS O F THE S I X T H

LAMB,C. A., Dept. Agronomy, Ohio Agric. Expt. Sta., Wooster, Ohio. LAMBERT, W. V., Dept. Genetics, Iowa State College, Ames, Iowa. LAMPE,LOIS, Dept. Botany, Ohio State Univ., Columbus, Ohio. LANCEFIELD, D. E., Columbia Univ., New York, N. Y. LANDAUER, WALTER, Agric. Expt. Sta., Storrs, Conn. LANDONE, BROWN, Stickle Pond Rd., Newton, N. J. LANGE, MATHILDE M., Wheaton College, Norton, Mass. *LATHE,F. E., National Research Council, Ottawa, Ont., Canada. *LATIMER, HOMER B., Dept. Anatomy, Univ. Kansas, Lawrence, Kans. LATTES,LEONE,Modena, Italy. LAUGHLIN, H. H., Eugenics Record Office, Cold Spring Harbor, N. Y. LAUMAN, G. N., 504 Thurston Ave., Ithaca, N. Y. LEBEDEFF, GABRIEL A., Carnegie Inst., Cold Spring Harbor, N. Y. LELIVELD, J. A., Botanical Institute, Amsterdam, Netherlands. LENDERKING, RUTH,718 N. Fulton Ave., Baltimore, Md. LEONARD, WARREN H., Colorado Agric. College, Fort Collins, Colo. *LESLEY,J. W., Citrus Expt. Sta., Riverside, Calif. LEWIS,R. D., Dept. Farm Crops, Ohio State Univ., Columbus, Ohio. *LEWIS,RUNDALL M., Slaterville Rd., Ithaca, N. Y. L'HERITIER,PHILIPPE,45, Rue d'Ulm, Paris, France. LIBRARIAN, Univ. Nanking, Nanking, China. *LILLIE,F. R., Univ. Chicago, Chicago, Ill. *LINCOLN, EDWARD A., 4 Oak Knoll, Arlington, Mass. LINDEGREN, CARLC., West Pennsylvania Hospital, Pittsburgh, Pa. LINDSAY,RUTHH., Dept. Botany, Wellesley College, Wellesley, Mass. *LINDSEY, A. W., BOX782, Granville, Ohio. LINDSTROM, E. W., Iowa State College, Ames, Iowa. *LINFIELD,F. B., Expt. Sta., Bozeman, Mont. LITTLE,C. C., Roscoe B. Jackson Memorial Lab., Bar Harbor, Me. LIVERMORE, J. R., Cornell Univ., Ithaca, N. Y. LODS,GMILEA., Macdonald College, Ste. Anne de Bellevue, Quebec, Canada. LONGISLAND UNIVERSITY, Brooklyn, N. Y. LONGLEY, W. H., Goucher College, Baltimore, Md. LORZ,ALBERT,Blandy Experimental Farm, Univ. Virginia, Boyce, Va. *LOTBINI~RE, DE,A. JOLY,Pointe Platon Co., Lotbinii.re, Quebec, Canada. *LOVE,H. H., Dept. Plant Breeding, Cornelf Univ., Ithaca, N. Y. LUSH,JAYL., Agric. Hall, Iowa State College, Ames, Iowa. LYNCH,CLARAJ., Rockefeller Inst., Ave. A and 66 St., New York, N.Y. MA, PAULC., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y.

INTERNATIONAL CONGRESS O F GENETICS

37

MACARTHUR, JOHN W., Dept. Biology, Univ. Toronto, Toronto, Ont., Canada. *MACAULAY, T . B., 109 Westmount Boulevard, Westmount, Quebec, Canada. *MACAULAY, T. B., Sun Life Assurance Co., Montreal, Quebec, Canada. MCCLINTOCK, BARBARA, Dept. Biology, California Inst. Tech., Pasadena, Calif. MCCLUNG, C. E., Dept. Zoology, Univ. Pennsylvania, Philadelphia, Pa. MCCONKEY, O., Ontario Agric. College, Guelph, Ont., Canada. *MCCRACKEN, ELIZABETH, International House, Berkeley, Calif. *MCCRADY, JR., EDWARD, Wistar Institute, Philadelphia, Pa. MACDANIELS, L. H., Dept. Pomology, Cornell Univ., Ithaca, N. Y. MACDOWELL, CHARLOTTE G., Cold Spring Harbor, N. Y. MACDOWELL, E. C., Station for Experimental Evolution, Cold Spring Harbor, N. Y. MCEWEN,ROBERTS., 208 Forest, Oberlin, Ohio. MCFADDEN, E. S., Redfield, S. Dak. *McGEE, ANITANEWCOMB, BOX363, Southern Pines, N. C. MCGIBBON, W. H., Macdonald College, Quebec, Canada. * M c G o u ~JR., , RALPHC., 13 S. Prospect St., Amherst, Mass. MCGREGOR, W . GRANT,Central Experimental Farm, Ottawa, Ont., Canada. *MACKIE,W. W., 124 Hilgard Hall, Univ. California, Berkeley, Calif. MACKLIN,MADGE,Univ. of W . Ontario, London, Ont., Canada. MACOUN, W. T., Central Experimental Farm, Ottawa, Ont., Canada. MCPHEE,HUGHC., Bureau Animal Industry, U. S. Dept. Agric., Washington, D. C. MACRAE,NORMAN A., Central Experimental Farm, Ottawa, Ont., Canada. MCROSTIE,G. P., 495 Strodbrook Ave., Winnipeg, Man., Canada MAGRUDER, ROY,908 B St. S.W., Washington, D. C. MAHONEY, CHARLES H., Dept. Hort., Michigan State College, E. Lansing, Mich. *MAINS,E. B., Dept. Botany, Univ. Michigan, Ann Arbor, Mich. *MAIRE,R E N ~Professeur , i l'UniversitC, Algeria, Africa. *MANGELSDORF, A. J., H.S.P.A. Expt. Sta., Honolulu, T.H. MANGELSDORF, P. C., Texas Agric. Expt. Sta., College Station, Tex. MARGOLIS, OTTOS., New York Univ., 100 Washington Sq. E., New York, N.Y. MARSHAK, ALFRED,Bussey Inst., Forest Hills, Boston, Mass.

38

PROCEEDINGS OF THE SIXTH

MARSHALL,RUTH,Rockford College, Rockford, Ill. MARTIN,J. HOLMES,Dept. Poultry, Univ. Kentucky, Lexington, Ky. *MARTIN,JOHNH., Bureau Plant Industry, U. S. Dept. Agric., Washington, D. C. MAVOR,J. W., Union College, Schenectady, N. Y. MAW,A. J. G., Macdonald College, Quebec, Canada. MAW,W. A., Macdonald College, Quebec, Canada. MERRELL,WILLIAMDAYTON, Univ. Rochester, Rochester, N. Y. METZ,CHARLES W., Johns Hopkins Univ., Baltimore, Md. J., 308 The Parkway, Ithaca, N. Y. METZGER, HERBERT *MIDDLETON, A. R., Univ. Louisville, Louisville, Ky. *MIEGE, EMILE,Dir. Sta. Sel. Sciences, Ave. de Tfmara No. 6F, Rabat, Morocco, Africa. Dept. Agric. and Stock, Brisbane, Q., Australia. MILES,L. GORDON, *MILLER,HELENM., Osborn Zoological Lab., Yale Univ., New Haven, Conn. MILLER,J U L I A N C., Dept. Hort., Louisiana State Univ., Baton Rouge, La. *MILLER,KENNETHR., Court House, Pulaski, N. Y. *MILLS,H . STRYCKER, BOX358, Bristol, Pa. MILLS,W . D., 417 Utica St., Ithaca, N. Y. *MILNOR,GEORGE S., Farmers' National Grain Corp., Alton, Ill. MINNESOTA CROPIMPROVEMENT ASSOCIATION. MINNS,LUAA., 217 Mitchell St., Ithaca, N. Y. N , ALFRED,Vinderen Laboratorium, Kristiana, Norway. M J ~ ~ EJON MOHR,OTTOLOUIS, Anatomical Inst., The Royal Frederiks Univ., Oslo, Norway. MONTALENTI, GIUSEPPE,29F, Via Cola di Rienzi, Rome, Italy. MOODY,JULIAE., Hallowell House, Wellesley, Mass. MOORE,ARTHURRUSSELL,Univ. Oregon, Eugene, Ore. *MOORE,BARRINGTON, 191 Ninth Ave., New York, N. Y. *MOORE,J. PERCY,Zoological Lab., Univ. Pennsylvania, Philadelphia, Pa. MORGAN, T. H., California Inst. Tech., Pasadena, Calif. T. H., Mrs., California Inst. Tech., Pasadena, Calif. MORGAN, MORGAN, W. P., 41 15 Otterbein Ave., Indianapolis, Ind. F. B., Dept. Animal Husb., Cornell Univ., Ithaca, N. Y. MORRISON, *MORRISON,GORDON, Ferry-Morse Seed Co., Rochester, Mich. MOYER,RAYMOND T., W . Columbia Ave., Lansdale, Pa. MULLER,H. J., Univ. Texas, Austin, Tex. MUMM,WALTERJ., 110 Old Agric. Bldg., Urbana, Ill. *MUNERATI, O., R. Stazione Sperimental di Bietecultura, Rovigo, Italy.

INTERNATIONAL CONGRESS O F GENETICS

39

*MUNOZ, ROBERTO, Dept. Plant Pathology, Cornell Univ., Ithaca, N. Y. MURRAY, JOSEPHMERRITT,Carnegie Inst., Cold Spring Harbor, N. Y. MURRAY, JOSEPHM., 18 Hancock St., Bar Harbor, Me. *MUTTKOWSKI, R. A., Univ. Detroit, 6 Mile Rd. at Livernois, Detroit, Mich. MYERS,C. E., 304 W. Fairmount Ave., State College, Pa. MYERS,C. H., Cornell Univ., Ithaca, N. Y. NABOURS, R. K., Kansas State College, Manhattan, Kans. NACHTSHEIM, HANS,Inst. f. Vererbungsforschung, Berlin-Dahlem, Germany. *NAHM,LAURAJ., BOX146, Flat River, Mo. NEAL,NORMAN P., 803 State St., Madison, Wis. NEBEL,B. R., Agric. Expt. Sta., Geneva, N. Y. NEEDHAM, JAMESG., 6 Needham PI., Ithaca, N. Y. NEELY,WINSTON,Dept. Agron., Cornell Univ., Ithaca, N. Y. *NEETHLING, J. P., Univ. of Stellenbosch, Stellenbosch, C.P., S. Africa. *NEWBOLD, JR., ARTHURE., 15 and Walnut St., Philadelphia, Pa. NEWHALL, A. G., 208 Stewart Ave., Ithaca, N. Y. NEWMAN, H. H., 5712 Dorchester Ave., Chicago, Ill. NEWMAN, L. H., Ottawa, Ont., Canada. *NEWTON,ROBERT,Dept. Field Crops and Plant Biochem., Univ. Alberta, Edmonton, Alta., Canada. NEW YORKCOLLEGE OF AGRICULTURE AT CORNELL UNIVERSITY, Ithaca, N. Y. *NICHOLAS, J. S., Osborn Zoological Lab., Yale Univ., New Haven, Conn. NICHOLS,M. LOUISE,The Dreycott, Haverford, Pa. NOLLA,J. A. B., Cornell Univ., Ithaca, N. Y. *ODLAND, T. E., Kingston, R. I. *O'KELLY,J. FRED,Agric. and Mech. College, College, Miss. P., Washington Univ., St. Louis, Mo. OLIVER,CLARENCE *OSBORN, FREDRICK, 52 Broadway, New York, N. Y. OTTLEY,ALICEM., 46 Dover Rd., Wellesley, Mass. *OWEN,F. V., 1810 S. Main St., Salt Lake City, Utah. *PAGE,DAVID,Thomas Page Milling Co., N. Topeka, Kans. PAINTER,REGINALDH., Dept. Entomology, Kansas State College, Manhattan, Kans. *PAINTER,T. S., Univ. Texas, Austin, Tex. *PALMER,E. F., Vineland Station, Ont., Canada. PARK,J. B., 145 E. Webber Rd., Columbus, Ohio. PARKER, JOHNH., Kansas State College, Manhattan, Kans.

PROCEEDINGS O F T H E SIXTH

40

*PARODI,LORENZO R., Thames 1225, Buenos Aires, Argentina. PASSMORE, SARAF., Mendenhall, Pa. PATTERSON, J. T., 1903 Cliff St., Austin, Tex. *PAYNE,FERNANDUS, 620 Ballantine Rd., Bloomington, Ind. Inst. for Biol. Res., Johns Hopkins Univ., Baltimore, *PEARL,RAYMOND,

Md. PERRY,H. S., 219 Kelvin PI., Ithaca, N. Y. *PETERSON, R. F., 2125 Como Ave. W., St. Paul, Minn. *PETRY,E. J., 1730 B Ave. N.E., Cedar Rapids, Iowa. *PETZKE,ERNESTA., Hixton, Wis. PHELPS,LILLIANA., 714 E. 33 St., Kansas City, Mo. PHILLIPS,E. F., Cornell Univ., Ithaca, N. Y. *PHILLIPS,JOHNC., 77 Mt. Vernon St., Boston, Mass. *PIESCU,A., Institut Expirimental pour la Culture et la Fermentation du Tabac, Bucarest-Baneasa, Roumania. PILLSBURY FLOUR MILLSCOMPANY, Minneapolis, Minn. PINCUS,G., Inst. Biology, Harvard Univ., Cambridge, Mass. *PIZA,S. DE TOLEDO, Escola Agric. Superior, "Luiz de Queiroz," Piracicaba f. de S. Paulo, Brazil. PLATT,EMILIEL., 52 Hinckley PI., Brooklyn, N. Y. PLATT,LOIS I., Carnegie Inst., Cold Spring Harbor, N. Y. PLOUGH,H. H., Amherst College, Amherst, Mass. PLUNKETT,C. R., New York Univ., Washington Sq., New York, N. Y. *PONTIUS,BYRONE., 124 Russell St., W . Lafayette, Ind. *POPE,ALVINE., New Jersey School for the Deaf, Trenton, N. J. POPENOE,PAUL,2495 N. Marengo Ave., Altadena, Calif. *PORTER,JAMES F., 105 W. Adams St., Chicago, Ill. PORTER, R. H., Botany Hall, Iowa State College, Ames, Iowa. *POST,R. H., Quogue, N. Y. POTTER,JAMES S., Cold Spring Harbor, N. Y. POWERS,LEROY,University Farm, St. Paul, Minn. POWSHER,LOUIS,224 Sullivan St., New York, N. Y. PRAT,HENRI,1265 Rue Saint Denis, Montreal, Quebec, Canada. *F'RENTICE,E. PARMALEE, 5 W . 53 St., New York, N. Y. PROULX, MAURICE, Ste. Anne de la PocatiPre, Quebec, Canada. *PRZYBOROWSKI, JOSEPH,Lobzowska 24, Krakow, Poland. *PUTNAM, EBEN,Wellesley Farms, Mass. QUISENBERRY, J. H., Room 103, Animal Genetics Lab., Urbana, Ill. RAFFEL,DANIEL,40 Wall St., New Haven, Conn. RAHN,OTTO,107 Maple Ave., Ithaca, N. Y.

INTERNATIONAL CONGRESS O F GENETICS

*RALEIGH, S. M., 2089 Carter Ave., St. Paul, Minn. *RAMALEY, FRANCIS, Univ. Colorado, Boulder, Colo. RAMELLA, RAUL,124 Catherine St., Ithaca, N. Y. L. F., Cornell Univ., Ithaca, N. Y. RANDOLPH, RATNER, BRET,515 W. End Ave., New York, N. Y. RAYMOND, L. C., Macdonald College, Quebec, Canada. REDDICK, DONALD, College Agric., Ithaca, N. Y. *REED,ERNEST,844 Sumner Ave., Syracuse, N. Y. REED,GEORGE M., Brooklyn Botanic Garden, Brooklyn, N. Y. REED,H. D., 214 Wait Ave., Ithaca, N. Y. REED,SHELDON C , Alpha Tau Omega House, Hanover, N. H. REEVES,R. G., BOX280 F. E., College Station, Tex. RHOADES, MARCUS M., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. RICHARDS, JR., A. GLENN,Dept. Entomology, Cornell Univ.;Ithaca, N. Y. RICHARDS, MILDRED HOGE,434 Chautauqua, Norman, Okla. *RICHARDSON, MARGARET, 2551 Rose Walk, Berkeley, Calif. RICHEY,FREDERICK D., Bureau Plant Industry, U. S. Dept. Agric., Washington, D. C. RIDDLE,OSCAR,Station for Experimental Evolution, Carnegie Inst., Cold Spring Harbor, N. Y. RIEMAN,G. H., 205 Church St., New Haven, Conn. *RIFE, D. CECIL,62 W. 10 Ave., Columbus, Ohio. RILEY,HERBERT PARKES, Graduate College, Princeton, N. J. ROBB,R. CUMMING, 447 S. Beech St., Syracuse, N. Y. *ROBERTS, EDITHA., Vassar College, Poughkeepsie, N. Y. ROBERTS, ELMER,College Agric., Urbana, Ill. ROBERTSON, D. W., Colorado Agric. College, Fort Collins, Colo. *ROBERTSON, EGBERT, 1257 Continental Illinois Bank Bldg., Chicago, Ill. ROBERTSON, W. R. B., Univ. Iowa, Iowa City, Iowa. ROBINSON, T. R., Bureau Plant Industry, U. S. Dept. Agric., Washington, D. C. ROEMER, THEODOR E. M., Ludwig Wuchererstr 2, Halle-Saale, Germany. ROMELL,L. G., 115 Eddy St., Ithaca, N. Y. ROQUE,ARTURO, Insular Expt. Sta., Rio Piedras, Puerto Rico. ROSENBAUM, LOUISE,Dept. Zoology, Univ. Pennsylvania, Philadelphia, Pa. R o s ~ f i s ~BOLESLAW, r, Lwhw, Delugosza, Poland. ROUSSEAU, JACQUES, Univ. Montreal, Montreal, Quebec, Canada. Ru, S. K., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. * R u n , CARLOS, Madrid, Spain.

42

PROCEEDINGS OF THE SIXTH

RUSSELL-MILLER MILLINGCOMPANY, Minneapolis, Minn. RUTTLE,M. L., Agric. Expt. Sta., Geneva, N. Y. SAENKO, S. M., Academy of Agric., .Moscow, Union of Socialistic Soviet Republics. *SALAMAN, REDCLIFFEN., Homestall, Barley, Royston Herts., England. SALISBURY, GLENNW., 527 E. Buffalo St., Ithaca, N. Y. SALMON, S. C., Div. Cereal Crops and Diseases, U. S. Dept. of Agric., Washington, D. C. SANDERS, J., Heemraadssingel 240, Rotterdam, Netherlands. *SANDERS, JOSEPH, 2612 Tilden St., Washington, D. C. SANDO, W. J., 3021 S. Dakota Ave. N.E., Washington, D. C. SAPEHIN,A., Odessa, Union of Socialistic Soviet Republics. SATINA,SOPHIA,Carnegie Inst., Cold Spring Harbor, N. Y. *SAULESCU, NICOLAI,Capita Postala, 219, Sta. de Ameliorarea Ilantelor, CIuj, Roumania. SAUNDERS, A. P., Hamilton College, Clinton, N. Y. SAWIN,PAULB., Biol. Lab., Brown Univ., Providence, R. I. *SAWYER, M. LOUISE,Dept. Botany, Wellesley College, Wellesley, Mass. SAX,KARL,Arnold Arboretum, Jamaica Plain, Boston, Mass. SCHAFFNER, J. H., Dept. Botany, Ohio State Univ., Columbus, Ohio. SCHMEISKE, WILLIAMF., Kirkwood, N. Y. SCHMID, A., Universitatsstrasse 2, Ziirich, Switzerland. SCHNEIDER, BURCHH., 555 Old Agricultural Hall, Univ. Illinois, Urbana, , Ill. *SCHOLES,JOHN C., Extens. Service, Manitoba Dept. Agric., Winnipeg, Man., Canada. SCHOTT,R. G., 315 Lynn Ave., Ames, Iowa. *SCHRADER, FRANZ,Dept. Zoology, Columbia Univ., New York, N. Y. *SCHULTZ, EDITHC., 676 Riverside Dr., New York, N. Y. SCHULTZ, JACK,California Inst. Tech., Pasadena, Calif. *SCHWEIT~ER, MORTOND., Dept. Zoology, Columbia Univ., New York, N. Y. SCOTT,ALLANC., Dept. Zoology, Columbia Univ., New York, N. Y. SCOTT,CARL,Plant Science Bldg., Cornell Univ., Ithaca, N. Y. SCOTT,GEORGE C., Dept. Biology, College City New York, 138 and Convent Ave., New York, N. Y. SCOTT,JOHN W., 1409 Garfield St., Laramie, Wyo. SCOTT,MARTHAHUGHES,5708 Maryland Ave., Chicago, Ill. SENN,HAROLD A., McMaster Univ., Hamilton, Ont., Canada. SENN,P. H., College Agric., Univ. Florida, Gainesville, Fla.

INTERNATIONAL CONGRESS O F GENETICS

*SETCHELL, WILLIAMALBERT,2441 Haste St., Berkeley, Calif. SHARP,LESTERW., 107 Irving PI., Ithaca, N. Y. SHAW,RICHARD N., Garland, Me. SHEN,S. T., Yenching Univ., Peiping, China. SHOEMAKER, D. N., 6800 Eastern Ave., Takoma Park, D. C. SHOWALTER, HIRAMM., Blandy Expt. Farm, Univ. Virginia, Boyce, Va. SHRIGLEY, EDWARD WHITE, Dept. Genetics, Iowa State College, Ames, Iowa. SHULL,A. FRANKLIN, 431 Highland Rd., Ann Arbor, Mich. SHULL,G. H., 60 Jefferson Rd., Princeton, N. J. *SIMPSON,F. B., Cuba, N. Y. SIMPSON, JENNIEL. S., Mrs., Hunter College, New York, N. Y. SINGLETON, WILLARD RALPH,Connecticut Agric. Expt. Sta., New Haven, Conn. SINNOTT,E. W., Barnard College, Columbia Univ., New York, N. Y. *SIVICKI~, P. B., Pasto Deze 130, Kaunas, Lithuania. SKALIASKA, Mrs. M., Free Univ., Warsaw, Poland. SLATE,G. L., Agric. Expt. Sta., Geneva, N. Y. *SMITH,FRANCIS L., 21 19 Addison St., Berkeley, Calif. SMITH,HELENBERENICE, BOX 1085, Johns Hopkins Univ., Baltimore, Md. SMITH,LUTHER,1209 Paquin, Columbia, Mo. SMITH,0 . D., Conway, Ark. SMITH,S. G., Dept. Botany, McGill Univ., Montreal, Quebec, Canada. SMITH,STUART N., Dept. Genetics, Iowa State College, Ames, Iowa. SMITH,T. L., Dept. Zoology, Columbia Univ., New York, N.Y. SMITH,WILLIAMK., Univ. Wisconsin, Madison, Wis. *SMITH,W. W., Purdue Univ., West Lafayette, Ind. SMITHSONIAN INSTITUTION, UNITEDSTATESNATIONAL MUSEUM, Washington, D. C. SNELL,GEORGE D., Dept. Biology, Brown Univ., Providence, R. I. *SNOW,SYDNEY B., 5700 Woodlawn Ave., Chicago, 111. SNYDER, L. H., Ohio State Univ., Columbus, Ohio. *SOUZA DA CAMARA, DE,ANTONIO, Instito Superior de Agronomia, Lisboa, Portugal. SPEIDEN,NORMAN R., BOX265, Mohegan Lake, N. Y. SPENCER, WARREN P., 701 N. Bever St., Wooster, Ohio. SPIER,JANE,Dept. Botany, McGill Univ., Montreal, Quebec, Canada. SPRAGUE, GEORGE F., 353 Hamilton, Clarendon, Va. STADLER, L. J., Univ. Missouri, Columbia, Mo.

44

PROCEEDINGS O F THE SIXTH

*STAKMAN, E. C., University Farm, St. Paul, Minn. STANTON, T. R., 116 Jackson Ave., Hyattsville, Md. *STAPLES-BROWNE, R., Butler's Court, Alvescot, Oxfordshire, England. 450 E. 64 St., New York, N. Y. STARK,MARYB., STAUFFER, JAMES,325 Dryden Rd., Ithaca, N. Y. Hamilton, N. Y. STEBBINS, JR., G. LEDYARD, STEELE,D. G., Connecticut Agric. College, Storrs, Conn. STEGGERDA, MORRIS,Carnegie Inst., Cold Spring Harbor, N. Y. *STEIL,W. N., 748 N. 23 St., Milwaukee, Wis. STEIN,KATHRYN F., Mount Holyoke College, South Hadley, Mass. STENE,A. E., BOX25, Kingston, R. I. *STEPHENS,F. E., State Capitol Bldg., R.308, Salt Lake City, Utah. STERN,CURT,Kaiser Wilhelm-Tnst. f. Biologie, Berlin-Dahlem, Germany. STEVENSON, F. J., 908 B St., Washington, D. C. STOCKARD, C. R., Cornell Medical College, First Ave. and 28 St., New York, N. Y. STONE,R. E., Guelph, Ont., Canada. STOUT,A. B., New York Botanical Garden, New York, N. Y. STRANDSKOV, HERLUFH., Dept. Zoology, Univ. Chicago, Chicago, Ill. STRINGFIELD, G. H., Expt. Sta., Wooster, Ohio. LEONELL C., Jackson Memorial Lab., Bar Harbor, Me. STRONG, *STRUBLE, FRED,Clear Lake Cannery, Upper Lake, Calif. *STUBBE, HANS,Kaiser Wilhelm-Institut, Miincheberg-Mark, Germany. STUCK,FLORENCE, 531 W. 122 St., Apt.C35, New York, N. Y. STURTEVANT, A. H., California Inst. Tech., Pasadena, Calif. R., Macdonald College, Ste. Anne de Bellevue, Quebec, Can. SUMMERBY, *SUMNER, F. B., Scripps Inst. Oceanography, La Jolla, Calif. SURINA, A. A., Elkland, Pa. SUTHERLAND, J. R. G., The Maples, Limehouse, Ont., Canada. SWANSON, A. F., Fort Hays Expt. Sta., Hays, Kans. *SWINGLE,CHARLES F., 2901 Legation St., Washington, D. C. *SWINGLE, WALTERT., U. S. Dept. Agric., Washington, D. C. SWOPE,W. D., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y. *TANAKA, YOSHIMARO, Kyushu Imperial Univ., Fukuoka, Japan. *TANNER, J. R., Norris Grain Co., 940 Bd. of Trade Bldg., Kansas City, Mo. TAYLOR, LEWISW., Div. Poultry, Univ. California, Berkeley, Calif. *TAYLOR, RAYMOND S., Newtown, Pa. *TAYLOR, W. R., Univ. Michigan, Ann Arbor, Mich. *TEODOREANU, N. I., Dieria Palas, Constantza, Roumania.

INTERNATIONAL CONGRESS O F GENETICS

*TERMAN, LEWISM., 761 Dolores St., Stanford University, Calif. TEXASAGRICULTURAL EXPERIMENT STATION, College Station, Tex. THOMAS,L. C., 1003 S. Busey St., Urbana, Ill. THOMPSON, H. C., College Agric., Ithaca, N. Y. THOMPSON, ROSSC., 1650 Harvard St. N.W., Washington, D. C. *THOMPSON, W . P., Dept. Biology, Univ. Saskatchewan, Saskatoon, Sask., Canada. *THOMSON, R. B., 586 Spadman Ave., Toronto, Ont., Canada. TIMOF~EFF-RESSOVSKY, H. A., Mrs., Kaiser Wilhelm-Inst. f. Hirnforschung, Berlin-Buch, Germany. TIMOF~EFF-RESSOVSKY, N. W., Kaiser Wilhelm-Inst. f. Hirnforschung, Berlin-Buch, Germany. *TIMOLAT, JAMES G., 59 Fourth Ave., New York, N. Y. *TINKLE,WILLIAMJ., 2104 E. 12 Ave., Huntington, W . Va. *TJEBBES,KLAS, Hilleshog Inst. for Sugarbeet Research, Box 82, Landskrona, Sweden. TORVIK,MAGNHILD M., Dept. Zoology, Univ. Pittsburgh, Pittsburgh, Pa. *TRANKOWSKY, DANIELA., Wolodarskaya, 22 fI.4, Moscow 4, Union of Socialistic Soviet Republics. *TREADWELL, AARONL., Vassar College, Poughkeepsie, N. Y. TRI-STATESOFT WHEATIMPROVEMENT ASSOCIATION, 2221 Front St., Toledo, Ohio. *TROWBRIDGE, E . A., Agric. Bldg., Columbia, Mo. *TUFF, P., Landbrukshiskolen, Aas, Norway. TULLY,EVAM., 9406 34 Rd., Jackson Heights, New York, N. Y. TURK,KENNETHL., Mount Vernon, Mo. UNIVERSITY OF CHICAGO, Chicago, Ill. UNIVERSITY OF MISSOURI, Columbia, Mo. OF STATE,Division of International ConUNITEDSTATESDEPARTMENT ferences, Washington, D. C. UPP, CHARLES W., University Station, Baton Rouge, La. VANDEL,A., FacultP des Sciences, AllCe Saint Michel, Toulouse, France. VANDEL,A., Mrs., FacultC des Sciences, AllCe Saint Michel, Toulouse, France. R E N ~Rixensart, , Near Brussels, Belgium. VANDENDRIES, *VANHEEL,J. P. DUDOK,"Die Rietkraag," Naarden, Netherlands. VAN LONE,ELDYNE,508 S. Baldwin St., Madison, Wis. F. A., American Univ., Washington, D. C. VARRELMAN, VAVILOV,N., Inst. Applied Botany, Leningrad, Union of Socialistic Soviet Republics.

46

PROCEEDINGS O F T H E SIXTH

*VEATCH,COLLINS, Ozark, Mo. VICARI,E. M., Cornell Univ. Medical College, New York, N. Y. VIJ JAKICH,L. S., 301 Bryant Ave., Ithaca, N. Y. VILMORIN, DE,ROGER, Vilmorin-Andrieux et Cie., Paris, France. *VOITELLIER, CHARLES, 89 Rue Erlanger, Paris, (16") France. *WAARDENBURG, P:J., Velperweg 22, Arnhem, Netherlands. WACHTER, W. L., Lafayette College, Easton, Pa. *WADE,B. L., Washburn-Wilson Seed Co., Moscow, Idaho. WAGGENER, ROYA., 210 E. First St., Northfield, Minn. *WAITZINGER, L. A., Lingnan Univ., Canton, China. WALDRON, L. R., Dept. Agronomy, Agric. Expt. Sta., State College Station, Fargo, N. Dak. WALKER, NOLANA., 719 Hitt, Columbia, Mo. *WALLACE, H. A., Des Moines, Iowa. WALLER, A. E., Dept. Botany, Ohio State Univ., Columbus, Ohio. *WALTER, HERBERT EUGENE, Brown Univ., Providence, R. I. WANG,S., Univ. Nanking, Nanking, China. WARBRITTON, VIRGENE,Lefevre Hall, Columbia, Mo. *WARKENTIN, C. B., Midland Flour Milling Co., Kansas City, Mo. WARREN, D. C., Kansas State College, Manhattan, Kans. WARREN,HERBERTS., 1405 Greywall Lane, Overbrook Hills, Philadelphia, Pa. WARREN, PAULA., Tufts College, Mass. WARWICK, BRUCEL., Agric. Expt. Sta., College Station, Tex. WATERS,NELSONF., Dept. Poultry, Iowa State College, Ames, Iowa. WEINSTEIN,ALEXANDER, Dept. Zoology, Johns Hopkins Univ., Baltimore, Md. WEISMANN, MAXWELL NAPIER,5716 11 Ave., Brooklyn, N. Y. WEISS,PAULA., Marine Biological Lab., Woods Hole, Mass. WELLINGTON, RICHARD, New York Expt. Sta., Geneva, N. Y. *WENRICH,D. H., Zoological Lab., Univ. Pennsylvania, Philadelphia, Pa. WENSTRUP, EDWARD J., St. Vincent College, Latrobe, Pa. WENTWORTH, E. N., Armour and Co., Chicago, 111. WENTZ,JOHNB., 1023 Brookridge Ave., Ames, Iowa. WHETZEL, H. H., Cornell Univ., Ithaca, N. Y. WHITAKER, THOMAS W., Bussey Inst., Forest Hills, Boston, Mass. WHITE,E. GRACE,Wilson College, Chambersburg, Pa. WHITE,ORLAND E., Biological Bldg., University, Va. WHITING,ANNAR., Pennsylvania College for Women, Pittsburgh, Pa. WHITING,P. W., Univ. Pittsburgh, Pittsburgh, Pa.

INTERNATIONAL CONGRESS O F GENETICS

WHITNEY,D. D., Univ. Nebraska, Lincoln, Nebr. WHITNEY,LEONF., Oakwood Rd., Orange, Conn. WHITTAKER, ELIZABETH L., 205 College Ave., Elmira, N. Y. *WICKS,STANTON D., 604 Syracuse Savings Bank Bldg., Syracuse, N. Y. *WIEBE,G. A., Davis, Calif. WIEGAND, KARLM., 109 E. Upland Rd., Ithaca, N. Y. *WIEMAN,H. L., Univ. Cincinnati, Cincinnati, Ohio. WIGGANS, R. G., New York State College Agric., Ithaca, N. Y. WILCOX,A. N., University Farm, St. Paul, Minn. *WILLEY,CHARLESH., New York Univ., University Heights, Bronx, N. Y. *WILLIAMS,ELMERF., Williams and Wilkins Co., Baltimore, Md. WILLIAMS,LEONARD F., 2958 Belrose Ave., Pittsburgh, Pa. WILLMAN, JOHNP., New York State College Agric., Ithaca, N. Y. WILSON,B. H., Dominion Experimental Farm, Indian Head, Sask., Canada. WINGE,o., Genetic Lab., The Royal Vet. and Agric. College, Rolighedsvej 23, Copenhagen V., Denmark. WINTERS,L. M., University Farm, St. Paul, Minn. WITSCHI,EMIL,Dept. Zoology, Univ. Iowa, Iowa City, Iowa. WOOD,THELMA R., Dept. Biology, Brown Univ., Providence, R. I. *WOOD,JR., WILLIAMP., 11 S. 14 St., Richmond, Va. WOODWORTH, C. M., 511 Pennsylvania Ave., Urbana, Ill. WOOLLEY, G. W., 1717 University Ave., Madison, Wis. WORZELLA, WALLACE W., 356 Northwestern Ave., La Fayette, Ind. WRIGHT,SEWALL, 5762 Harper Ave., Chicago, Ill. YAMPOLSKY, CECIL,230 Franklin Ave., Grantwood, N. J. YANG,YUNKUEI,Dept. Farm Crops, Ohio State Uuiv., Columbus, Ohio. *YARNELL, RAY,Capper's Farmer, Topeka, Kans. YARNELL, S. H., 4301 W. 26 St., Bryan, Tex. *YEAGER, A. F., State College Station, Fargo, N. Dak. YERKES,ROBERTM., Yale Univ., 333 Cedar St., New Haven, Conn. *ZEIMET,AGNESL., Dept. Genetics, Univ. Wisconsin, Madison, Wis. ZELENY,C., Univ. Illinois, Urbana, Ill. ZULUETA, ANTONIO DE, Museo Nacional Ciencias Natural, Madrid, Spain.

G R O U P P H O T O G R A P H O F M E M B E R S O F T H E S I X T H I N T E R N A T I O N A L G E N E T I C S CONGRESS A T I T H A C A ARRANGED BY N U M B E R S

F. A. E. Crew

F. B. Hutt Katherine S. Brchme R G aap Ehwa'rd Wenotrup Sara F. &assmore Elorence L. Barrows Helen Besley Helen Houghtaling Solomon Horowitz G. L. Slate W. H. Alderman John T. Bregger D a v ~ dH. Thompson A. P. French Glen Sallsbury E. E. Heizer Kenneth L. Turk -Stuart N. Smith Jack Schultz L. J. Stadler A. C. Fraser T. H. Mornan R. A. Emerson F. P. Bussell C. C. Hurst

130 131 132 133 134 135 I36 137 138 139 140 141 142 143 144 145 146 147 148

C. K. Parris F. D. Richey

R. A. Fisher Alexander Weinstein Daniel Raffel Mrs. A. Vandel H. R. Hunt K u r t Hubert A. E. Brandt P. W. Gregory G. L. Stebbins, Jr. A. P. Saunders A. B. Stout C. G. Bowers J. T. Buchholz G. W. Woolley H. 0.Hetzer M. T. Macklin N. I. Vavilov

~

&.

G. B. Durham Reginald H. Painter B. Nebel Lillian Hollingshead Hill George S. Carter R. J. Garber Max M. Hoover A. B. Nolla W. Jackson ohn T Crofts H. i o n g l e A. Franklin {hull T. Patterson E. Lancefield F. A. Varrelman Herbert S. Warren Virgene Warbritton G. Pincus C. B. Davenport R. S. McEwen E. C. MacDowell E. M. East Fred N. Briggs L. Collins Eleanor Carothers P. W. Whiting Roger de Vilmorin J. W. MacArthur R. K. Nahours John J. Bittner Wm. H. Gates Mrs. Wm. H. Gates Ruth Marshall 0. WinEe A. M. Banta S. C. Harland 0. E. White R. A. Brink D. N. Shoemaker Paul A. Warren A. F. Swanson E. P. Hume 0.D. Smith T. Hawryluk G. A. Lebedeff C. W. Metz Ralph E. Cleland Karl Sax B. 0.Dodge B. L. Warwick Mrs. Adeline Van L.one E. E. Van Lone E. Bowstead G. Steele H. R. Albrecht D. C. Cooper Ruth H. Lindsay P. H. S c n n

;::b.

204 205 206

h.

;:;i.

209 210 211 212 213 214 215 216 217

H. Gerould 2;: k. k. Timofheff-Ressov: F. R. Immer

t.R. Powers

z:2,";'"

h. . f. b

E. W. Llndstrom Anastasia J. Romanoff N. Dobrovolskaia-Zavadskaia Mrs. F. W. Herriott George Haines Th. Dobzhansky L. C. Dunn Lua A. Minns R. J. ~ a m e i l o f f Marcus M. Rhoades Barbara McClintock Virginia H. Rhoades ~a;riet B.,Creighton Luther Smith L. H. Newman G. H. Cutler W. W. Worzella w: R . B. ~ o b e r t s o n F. A. Hays MIS. F. A. Hays B. Rosiriski Zenas H. Ellis Myron Gordon J. B. S. Haldane L. R. Waldron E. F. Gaines Margaret Gaines Mrs. J. Rheinheimer J. Rheinheimer George H. Shull G. P. Frets N. R. Speiden C. F. Feng E. S. McFadden

195 196 197 198 199 200 201

169 170 171 172 173 174

ane Spier arle Hearne S. Bangson P. Kelly C. D. Gordon Arthur B. Chapman ohn H. Quisenberry alker M. Dawson William H. Eyster W. T. Macoun R. Summerby H. Nachtsheim Harry Federley Kristine Bonnevie Mrs. Hansen C. H. Mahoney Mrs. Helen D. Hill 1. Ben Hill f . Sanders E. J. Gumhel Rene Vandendries A. Zulueta Jac ues Rousseau L. Cole E. N. Wentworth F. B. Morrlson C. J. Lynch Wilhelmina F. Dunnin A Schmid E. W. Sinnott A. F. Blakeslee R. Ruggles Gates H. H. Strandskov Edward C. Colin M. N. Wcismann Nathan Kaliss A. C. Scott Haig Dermen

4.

193 194

220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246

;:;b.

249 250 251 252

C. Strong

H. D. King L. F. Whitnev

enoe.

Paul Pop

Mrs. W. E. Castle Mrs. F. S. Tulloss Beatrice Johnson-Little 331 H. A. Senn Robert Cook 332 Eva M. Tully 333 S. H. Emerson 334 Ruth E. Lenderking 335 Emerson G. Knowles 336 Phili pe L'Herrt~er L. R. Watson 337 charyes Zeleny J. R. Livermore 338 M. R. Irwin R. B. Hinman 340 J. 339 G. I. D.Kendall Snell W. Neely Arturo Roque 341 H. Bentley Glass Carlos A. Krug 342 E. M. Perry Taee Kemo 343 M. M. Perry J. 'Clausen' 344 Mrs. B. 0.Dodge L. C. Thomas 345 Hally J. Sax W . F. Hanna 346 Wm. H . Brittingham L. E. Kirk 547 S. K. Ru 0. McConkey 348 Otto S. Marzolis 349 A. L. BaronG. P. McRostic T. R. Wood 350 George P. Child Brown 351 L. H. Hamilton leanor Davis 352 M. H. Harnly 353 R. 0. Earl Donald W. Davis 354 E. Grace White Martha H. Scott 355 M. A. Hayden Donald F. Jones 356 L. W. Taylor M. Demerec R. G. Wiggans 357 N. A. MacRae 358 T. J. Arnason A. M. Brunson 359 W. H. McGibbon S. G. Smith 360 Howard B. Frost H. D. Goodale H. E. Warfel 361 Floyd Ingersoll Mrs. B. P. Kaufmann 362 W. V. Lambert 363 C. R. Burnham Ernest C. Diver 364 F. A. Coffman A. E. Clarke 365 A. 1. G. Maw Alan Deakin 366 F. B. Meacham A. H. Hersh 367 1. B. Cotner T. W. Whitaker 368 C. H. Bostian LadlPy Husted 369 R. A. Derick Wilbur M. Luce 370 A. G. Whiteside Merle T. Jenkins 371 1. R. G. Sutherland W. R. Singleton 372 'Edward W. Sbrigley T. L. Smith 373 H. J. Fitz atrick H. M. Showalter 374 W. G. ~ c e r e g o r S. H. Yarnell 375 J. M. Armstrong H. L. Ibsen 376 G . H. Stringfield Albert Lorz 377 W. K . w t h Alfred Marshak 378 P. B. Sawln S. 0. Burhoe 379 F. H. Clark P. C. Mangelsdorf Edgar Anderson Kenneth Kopf W. J. Duchemin W. E. Castle 385 Elliot C. C. Little 386 T. M. Carrence N. F. Waters 387 A. N. Wilcox C. V. Green 388 G. H. Rieman Horace Feldman 389 Herbert P. Riley J. M. Murray J. B. Park

2,' &

G R O U P P H O T O G R A P H O F M E M B E R S O F THE S I X T H I N T E R N A T I O N A L C O N G R E S S O F G E N E T I C S A T I T H A C A A R R A N G E D A L P H A B E T I C A L L Y Albrecht 249 ~ l d e r m a h ,13 Anderson 315 ~ r m s t r o n h ,375 Arnason, -358 Babcock. 91 Bangson; 159 Banta 229 Baron: 349 Barrows. 8 Beers, 118 Belfield, Mrs., 83 Bennion, 78 Besley 9 Bittne;, 224 Blakeslee 187 ~ o n n e v i e :170 Bostian, 368 Bowers 143 BowsteAd, 247 Boyden, 326 Brandt. 138

Burnham 363 Bussell. 56 Caroth;E;219 Carter 199 castle' W. E. 318 castle' MIS. W. E., 2 ~ h a p d a n ,162 Child 350 ~ h r i s i i a n 99 Chrohocz;k, 101 Chung, 106 Clark E. B. 381 lark' F. H.: 379 Clark; 299 ~ l a u s e h 279 ~ l e l a n d . '241 Coffman, 364 Cole. 180

Cooper 250 ~ o t n e r 367 ' ~ r a w f d r d 72 ~reighton),43 Crew 1 croft;, 204 Currence. 386 Cutler, 46 - Darlington, 68 Davenport, 213 Davis Mrs. Barbara, ~ a v i s :D. W., 288

Davis. M. E., 287 Dawson, 164 Deakin, 300 Demerec, 291 Derick. 369 ~ e r m e n ,194 Diver 298 ~ o b r d v o l s k a i a - ~ a v a daia, Dobzhansky, 35 Dodge B. 0. 243 ~ o d g e 'Mrs. 'B. O., 44 Duchehin, 317 Dunn, 36 Dunning, 184 Durham. 19; Earl 353 ~ a s t :216 Einsele, 113 Elliot 385 Ellis '52 ~ m e ; s o n , R. A. 25 Emerson, S. H:, 333 Erlanson, 330 Estabrook. 88 Evster. 165 F;derley, 169 Feldman, 322 Feng, C. F., 63 Fisher, 132 Fitzpatrick. 373 Forbes, 152 Fowlds 65 ~raser,'23 French, 16 Frets. 61 Frost' 360 ~ a i n g s E. F. 56 ~ a i n e s 'M. 5? ~ a r b e ; 20b Gates h. R., 188 ~ a t e s :W. H. 225 Gates, Mrs. H., Gay 29 ~ e r A u l d ,153 Ghigi, 96 Gibson, 84 Gini 94 - ~~~. glass,.^. B. 341 Glass L. C. '81 ~olds)chmidt;97 Good, 258 Goodale 295 ~ o o d r i c h 126 Goidon, k. D., 161 Gordon. M.. 53 Gowen,' 28 Green. 321 GregoLy i39 ~ r o s s m a n 149 Gumbel, i76 Haines 34 ~ a l d a d e .54 Hall, 120 Hami!ton 351 Hanna, 2'81

G.

Hansen, 171 Harland, 230 Harnly, 352 Hart, 259 Hawryluk, 238 Hayden, 355 Hays, F. A. 49 Hays, Mrs. 'F. A., 50 Hearne, 158 Heizer, 18 Herriott, 33 Hersh, 301 Hetzer, 146 Hill, Mrs. H. D., 173 Hill J. B. 174 ill; L. H:, 198 Hinman, 274 Holmes, 79 Hoover, 201 Horowitz, 11 Houghtaling, 10 Howlett 76 ~ u h e r t , 137 ' Hume, 236 Hunt, 136 Hurst, 27 Huskins, 80 Husted, 303 Hutt, 2 Ibsen, 310 Immer, 155 Ingersoll, 361 Irwin, 338 Jaap, 4 ackson, 203 enkins 305 ohnsod-little, 266 ones, 290 Kaliss, 192 Kamenoff, 39 Kaufmann B. P. 270 ~ a u f m a n n :Mrs. B. P., Kelly 160 ~ e m i 278 , Kendall 339 King, 2k0 Kirk, 282 Knowles 335 ~opf3 ,16' Kornhauser 151 Krantz, 384' K r u g 277 ~ u l k i r n i ,98 Kwan, 129 Lambert. 362 Lampe, 89 Lancefield 208 Lehedeff, i 3 9 Lenderking, 334 Leonard, 110 Lesley, 257 L'Heritier, 336 Lindegren, C. C., 75 Lindegren, Mrs. G., 7 Lindsay, 251

Reed, 380 Rheinheimer, 59 Rheinheimer, k r s . J., Rhoades, M. M., 40 Rhoades V. H. 42 ~ i c h a r d sA. 8; Richey, Richards:131 M.' H., 67 Rieman 388 Ritzma; 262 R O ~ 383 . Roberts. 269 Robertsbn,-E. W. 108 Robertson, W. R.' B., Romanoff. 31 Roque, 276 Rosiriski. 51 Roussea", 179 Ru, 347 Saenko, 82 Salisbury 17 Sanders, '175 Sando. 109 Saund'eis, 141 Sawin, 378 Sax, H. J., 345 Sax, K. 242 schaffngr J. H. 103 ~ c h i f f n e r :Mrs. 'J. H.: Schrn~d 185 ~ c h n e i d e r ,271 Schultz 21 Scott, A. C. 193 Scott, M. $. 289 Senn, H. A,: 331 ~ h o e kP. Senn a kH e r. b33 252

f

~howalter,'308 Shrigley 372 Shull A: F. 206 G. H.; 60 Simpson 7 singletoh 306 Sinnott i86 Slate, i 2 Smith, L., 44 Smith, 0. D., 237 S m ~ t hS. G. 294 smith: S. N.', 20 Smith, T. L., 307 Smith, W. K., 377 Snell, 340 Snyder, L. H., 92 Snyder, Mrs. L. H., 9 Speiden, 62 Soencer. 77 spier, i57 Sprague, 112 Stadler, 22 Stanton, 111 Stark, 119

hull:

Parris,' 130 Passmore 6 pattersod, 207 Perry, E. M., 342 Perry, M. M., 343 Phelps, 38 Pincus, 212 Plough, 327 Plunkett. 329 ~ o p e n o e ; 263 Powers, 156

Stehbins 140 Steele i 4 8 stern: 66 Stockard, 87 Stout. 142

Sturtevant. 328 Summerhv: 167 ~ u t h & l a n x , 371 Swanson, 235 ~homa's.356 280 Taylor Thomosbn. 15 T i m o f b e f f - ~ e s s o v s kH ~ ~ i m o f ~ e f f - ~ e s s o v N.*; sk~: Tulloss 265 TulIy, i 3 2 Turk. 19 Upp 1 2 5 vanhe1 A. 95 Vandel' M;S. A. Vandeidries, 17: Van Lone Mrs. A Van E. E., Varrelman 209 Vavilov, 148 Vicari 117 ~ i l m o ; i n de, 221 ~ a l d r o n . ' 55 Waller. 325 Warbritton 211 Warfel, 296 Warren, D. C., 124 Warren H. S. 210 warren: P. A,: 234 Warwick. 244 Waters 320 ats son) 272 weinstgin 133 weisman; 191 Wenstrup ' 5 ~ e n t w o r t k 181 Whitaker 502 White, E! G., 354 White 0. E. 231 ~ h i t e i i d e .3 j 0 Whiting 220 ~ h i t n e ; D. D. 150 ~ h i t n e y :L. F.,' 261 Wiggins, 292 Wilcox 387 Winge ' 228 Wood ' 285 wooll;y 145 ~ o r z e l l ; 47 Yang, 105' Yarnell 309 zeleny,'337 Zulueta. 178

on;

MEMBERS O F T H E SIXTH INTERNATIONAL CONGRESS O F GENETICS AT ITHACA 24-31 August 1932

PROCEEDINGS O F T H E S I X T H

PROGRAM E. M. East

In making up the program for the Sixth International Congress of Genetics, a list of nearly 3000 biologists interested in genetics was compiled. From this list five different geneticists independently selected the names of about 1200 active workers to whom invitations were sent to present papers before the congress. The preponderance of Americans on the morning programs was due to the inability of invited Europeans to attend. SUMMARY O F PROGRAM

Wednesday, August 24 Registration. Inspection of exhibits. Demonstration papers. Tour of the campus, the experimental plots, and the city. Opening plenary meeting of the congress. Informal reception. Thursday, August 25 General papers. Inspection of exhibits. Demonstration papers. Tour of the campus, the experimental plots, and the city. Trip to the plant-breeding experimental plots. General evening program. Organ music by Frederick S. Andrews, Assistant University Organist. Address of welcome by Provost A. R. Mann. Response by Doctor Richard Goldschmidt. Address by the President, Doctor Thomas Hunt Morgan, on "The Rise of Genetics." Friday, August 26 General papers. Inspection of exhibits.

INTERNATIONAL CONGRESS OF GENETICS

Demonstration papers. Picnic at Taughannock Falls State Park. Group conferences. Saturday, August 27 General papers. Sectional papers. General Genetics I. Cytology I. Animal Genetics I. Human Genetics. Methods and Technique. Genetics and Phytopathology. Excursion to Watkins Glen State Park. Inspection of exhibits. Trip to Enfield Glen State Park. Group conferences. Sunday, August 28 ~ x c u r s i o nto Niagara Falls. Organ recital in Sage Chapel by David Hugh Jones, Organist Westminster Choir School, Princeton. Monday, August 29 General papers. Sectional papers. Cytology 11. Animal Genetics 11. Plant Genetics I. Chromosome structure and crossing over. Genetics of species hybrids. Excursion to Watkins Glen State Park. Inspection of exhibits. Demonstration papers. Trip to Enfield Glen State Park. Trip to the sheep and swine barns. Trip to the experimental plots of the Department of Floriculture and Ornamental Horticulture. Group conferences. Tuesday, August 30 General papers. Sectional papers.

52

PROCEEDINGS OF THE SIXTH

General Genetics 11. Cytology 111. Plant Genetics 11. Drosophila. Problems relating to sex and fertility. Genetics and Pathology. Inspection of exhibits. Trip to the plant-breeding experimental plots. Trip to the experimental plots of the Department of Vegetable Crops. Final plenary meeting of the congress. Wednesday, August 3 1 [Program at the New York State Agricultural Experiment Station, Geneva, New York] Papers on fruit and vegetable breeding. Inspection of exhibits, and tours.

General Papers

Morning Sessions Thursday, August 25 T. H. Morgan, Chairman; C. B. Bridges, Vice Chairman and Secretary.

1. Mendelism in man. C. B. Davenport, Carnegie Institution of Washington, Cold Spring Harbor. 2. Inheritance of educability. F. A. E. Crew, Institute of Animal Genetics, Edinburgh. 3. The use of mosaics in the study of the developmental effects of genes. A. H. Sturtevant, California Institute of Technology, Pasadena.

INTERNATIONAL CONGRESS O F GENETICS

53

4. The present status of maize genetics. R. A. Emerson, Cornell University, Ithaca. Friday, August 26 B. Davenport, Chairman; A. F. Shull, Vice Chairman and Secretary. Special subject: Mutations. 1. On the potency of mutant genes and wild-type allelomorphs. 0. L. Mohr, Anatomical Institute, The University, Oslo. 2. Mutations of the gene in different directions. N. TimofCeff-Ressovsky, Kaiser Wilhelm-Institut fiir Hirnforschung, Berlin-Buch. 3. The genetic nature of induced mutations in plants. L. J. Stadler, United States Department of Agriculture, University of Missouri, Columbia. 4. Further studies on the nature and causes of gene mutations. H. J. Muller, University of Texas, Austin. Saturday, August 27 R. R. Gates, Chairman; L. W . Sharp, Vice Chairman and Secretary. Special subject : The interrelations of cytology and genetics. 1. The interrelations of the genotype and the karyotype and their bearing upon some genetic problems. M. S. Navaschin, Timiriazeff Institute, Moscow. 2. The cytological basis for crossing over. Karl Sax, Arnold Arboretum, Harvard University, Jamaica Plain. 3. Neuere Ergebnisse uber die Genetik und Zytologie des Crossing Over. C. Stern, Kaiser Wilhelm-Institut fur Biologie, BerlinDahlem. 4. The nature of sex chromosomes. 0. Winge, Royal Veterinary and Agricultural College, Copenhagen. Monday, August 29 K. Bonnevie, Chairman; L. C. Dunn, Vice Chairman and Secretary. Special subjects: Genetics of species hybrids (first three addresses). Contributions of genetics to the theory of organic evolution (last address). 1. The species problem in Datura. A. F. Blakesfee, Carnegie Institution of Washington, Cold Spring Harbor. 2. Konjugation der artfremden Chromosomen. Harry Federley, The University, Helsingfors.

54

PROCEEDINGS OF THE SIXTH

3. Species hybridization as a means of form genesis. G. D. Karpetchenko, Botanical Institute, Leningrad. 4. Genetik der geographischen Variation. R. Goldschmidt, Kaiser Wilhelm-Institut fiir Biologie, Berlin-Dahlem. Tuesday, August 30 R. Goldschmidt, Chairman; D. F. Jones, Vice Chairman and Secretary. Special subject: Contributions of genetics to the theory of organic evolution. 1. The process of evolution in cultivated plants. N. Vavilov, Institute for Applied Botany, Leningrad. 2. The evolutionary modification of genetic phenomena. R. A. Fisher, Rothamsted Experimental Station, Harpenden. 3. Can evolution be explained in terms of at present known genetical causes? J. B. S. Haldane, John Innes Horticultural Institution, Merton. 4. The rbles of mutation, crossbreeding, inbreeding, and selection in evolution. S. Wright, University of Chicago, Chicago.

Sectional Papers General Genetics I, Saturday, August 27 Julian Huxley, Chairman ; E. C. MacDowell, Vice Chairman and Secretary. 1. The importance of the genus Partula for the problem of heredity and environment in nature. H . E. Crampton, Barnard College, New York. 2. Progress of genetics and selection in animal breeding, Union of Socialist Soviet Republics. A. S. Serebrovsky, Timiriazeff Institute, Moscow. 3. Dependence of the size of xenoplastically induced organs upon the size of the host. 0. E. SchottC, Yale University, New Haven. 4. Genetics of evolution. C. C. Hurst, Trinity College, Cambridge. 5. The nature of the genes in relation to mutation and evolution. A. L. Hagedoorn, Soesterberg. 6. Investigations on the problem of causality of Mendelian results. V. RGzitka, Charles University, Prague.

INTERNATIONAL CONGRESS O F GENETICS

55

7. Concerning two modes of evolution in the horse. R. C. Robb, Syracuse University, Syracuse. 8. ( a ) General scheme concerning the mechanism of organic evolution from the standpoint of modern genetics. (b) Step allelomorphism and the theory of centers of the structure of the gene. N. P. Dubinin, Timiriazeff Institute, Moscow. 9. Anlage und Lebensraum in Tierzucht und Eugenik. H. Krzemer, Tierzuchtinstitut, Giessen. 10. Breeding habits of the Louisiana deer. W . H. Gates, Louisiana State University, Baton Rouge. 11. The genetics of modified endocrine secretion and associated form patterns among dog breeds. C. R. Stockard, Cornell University Medical College, New York. Cytology I, Saturday, August 27 C. E. Allen, Chairman; R. E. Cleland, Vice Chairman and Secretary. 1. Variability of the karyotype. G. A. Lewitsky, Institute of Applied Botany, Leningrad. 2. Genetic and cytological correlation of chromosomal aberrations of DrosoQhila melanogaster. C. P. Oliver and E. W . Van Atta, Washington University, St. Louis. 3. Genetic analysis of synapsis and maturation in eggs of Habrobracon. P. W. Whiting and Kathryn A. Gilmore, University of Pittsburgh, Pittsburgh. 4. Change in dominance of genes lying in duplicating fragments of chromosomes. T . G. Dobzhansky and A. H. Sturtevant, California Institute of Technology, Pasadena. 5. Structural changes in the chromosomes of maize. R. A. Brink and D. C. Cooper, University of Wisconsin, Madison. 6. Cytological observations in Zea inays on the intimate association of non-homologous parts of chromosomes in the mid-prophase of meiosis and its relation to diakinesis configurations. Barbara McClintock, California Institute of Technology, Pasadena. 7. The association of non-homologous parts in a chromosomal interchange in maize. C. R. Burnham, California Institute of Technology, Pasadena. 8. Chromosome unbalance and the asynaptic condition as induced in Nicotiana sytvestris by X-radiation. T. H. Goodspeed, University of California, Berkeley. 9. Male biparentalism in Habrobracon. Anna R. Whiting and Magnhild M. Torvik, Pennsylvania College for Women, Pittsburgh.

56

PROCEEDINGS O F THE SIXTH

10. Conservation of a morphological individuality of the chromosomes a t the resting nucleus. S. de Toledo Piza, Jr., Escola Agricola Superior, Piracicaba. 11. Alle ed Autopoliploidisme negli studii di Genetica. C. Artom, R. UniversitL di Pavia. Animal Genetics I, Saturday, August 27 Carlo Jucci, Chairman; H . Nachtsheim, Vice Chairman and Secretary.

1. The effects of inbreeding and crossbreeding on swine. H. C. McPhee, Bureau of Animal Industry, Washington. 2. The amount and kind of inbreeding which has occurred in the development of breeds of livestock. J. L. Lush, Iowa State College, Ames. 3. The nature of growth factors in domestic breeds of cattle. P. W. Gregory, University of California, Davis. 4. Die genetischen Beziehungen zwischen Korperfarbe und Augenfarbe bei Saugern. H. Nachtsheim, Institut fiir Vererbungsforschung, Berlin-Dahlem. 5 . Modifying factors in guinea pigs. H. L. Ibsen, Kansas State Agricultural College, Manhattan. 6. Mutations in a strain of captive gray Norway rats. Helen Dean King, Wistar Institute, Philadelphia. 7. "Leaden," a recent color mutation in the house mouse. J. M. Murray, Jackson Memorial Laboratory, Bar Harbor. 8. The inheritance of cataract and allied eye defects in the house mouse. A mutation involving eye lesions without the aid of Xrays or any other artificial means. Leone11 C. Strong, Jackson Memorial Laboratory, Bar Harbor. 9. Genetic studies in hare-lip. A. M: Cloudman, Jackson Memorial Laboratory, Bar Harbor. 10. The Aexed-tailed mouse. H. R. Hunt, Michigan Agricultural College, East Lansing. 11. The genetics of the ear of the house mouse, Mus musculus. H. W. Feldman, University of Michigan, Ann Arbor. Human Genetics, Saturday, August 27 S. J. Holmes, Chairman; G. P. Frets, Vice Chairman and Secretary.

1. Mental and physical differences in identical twins. H . H . Newman, University of Chicago, Chicago. 2. Variabilitatsanalyse des menschlichen Korpers (nach Forschungen

INTERNATIONAL CONGRESS O F GENETICS

3. 4.

5. 6.

7. 8. 9.

10.

11.

12.

57

an 800 Zwillingspaaren) . 0. von Verschuer, Kaiser Wilhelm-Institut fur Anthropologie, Berlin-Dahlem. Probleme der multiplen Allelie beim Menschen. Giinther Just, Zoologisches Institut, Greifswald. Uber Geschlechtseinfliisse bei autosomal bedingten Augenmerkmalen des Menschen und uber die Frage, ob es erblichveranlagte einseitige und bilateral-asymmetrische Augenmerkmale gibt. P. J. Waardenburg, Arnhem. Differential sex mortality and its genetic basis. S. J. Holmes, University of California, Berkeley. Chinese-Hawaiian crosses. H. L. Shapiro, American Museum of Natural History, New York. The application of statistics to the problem of inheritance of cancer. Madge Macklin, university of Western Ontario, London. Family investigations on the heredity of eye color in man. G. P. Frets, Maasoord-Poortugaal near Rotterdam. Uber idiodispositionelle entzundliche Erkrankungen der Nebenhohlen der Nase und des Ohres. Hans Griineberg, Die Universitat, Bonn. auf Messungen von Erbgang der grossen Begabung,-gestutzt iiber 1000 Familien in zwei oder drei Generationen. J. A. Mj$en, Vinderen Laboratorium, Oslo. Does the environment cause genetic change in man ? B. Rosiriski, Institut Antropologiczno-Etnologiczny, Lw6w. A study of twins. J. Sanders, Rotterdam.

Methods and Technique, Saturday, August 27 J. Clausen, Chairman; C. G. Bowers, Vice Chairman and Secretary. 1. Calculating linkage intensities from F, data. F. R. Immer, University of Minnesota, St. Paul. 2. On biological life tables. E . J. Gumbel, Die Universitat, Heidelberg. 3. The relative growth function in its application to the individual and to the group. A. H. Hersh, Western Reserve University, Cleveland. 4. Experimental methods in taxonomy. J. W . Gregor, Scottish Plant Breeding Station, Corstorphine. 5. Character recombination as a genetic tool. Edgar Anderson, Arnold Arboretum, Jamaica Plain. 6. Principles for a joint attack on evolutionary problems. J. Clausen, Stanford University, Palo Alto. 7. Storage, shipment and artificial germination of Rhododendron pollen. C. G. Bowers, Maine, N. Y.

58

PROCEEDINGS O F THE SIXTH

8. The technique of securing and hatching sexual eggs for use in studying biparental inheritance in Cladocera. Thelma R. Wood and A. M. Banta, Brown University, Providence. 9. Genetical engineering. H. D. Goodale, Hopedale Farm, Williamstown. Genetics and Phytopathology, Saturday, August 27 RenC Vandendries, Chairman; H. K. Hayes, Vice Chairman and Secretary. 1. Problems in the genetics of phytopathogenic fungi. E. C. Stakman, University of Minnesota, St. Paul. 2. Die Bedeutung der genetischen Analyse fiir die theoretische Resistenzforschung. W. H. Fuchs, Die Universitiit, Halle (Saale). 3. Breeding crop plants resistant to ins6cts. J. H. Parker, Kansas Agricultural Experiment Station, Manhattan. 4. Vererbungsstudien an anthraknoseresistenten Bohnen. F. Schreiber, Quedlinburg (Harz) . 5. The manner of inheritance of smut reaction in maize. M. M. Hoover and R. J. Garber, University of West Virginia, Morgantown. 6. Reaction of a wheat cross to three physiologic forms of bunt. E. F. Gaines, State College of Washington, Pullman. 7. The inheritance of resistance to bunt (Tilletia tritici) in wheat hybrids. F. N. Briggs, University of California, Davis. 8. Inheritance of resistance to loose and covered smuts in hybrids between certain susceptible oat varieties and Black Mesdag. G. M. Reed, Brooklyn Botanic Garden, Brooklyn. 9. Uber die Vererbung der Resistenz des Weizens gegen Ustilago tritici. T. Roemer, Die Universitiit, Halle (Saale). 10. The genetics of stem rust resistance in wheat. H. K. Hayes, University of Minnesota, St. Paul. 11. An apparently inseparable association of one type of rust resistance with a peculiar susceptibility to heat injury in wheat. E. S. McFadden, South Dakota Agricultural Experiment Station, Redfield. Cytology 11, Monday, August 29 G. A. Lewitsky, Chairman; C. W. Metz, Vice Chairman and Secretary. 1. The general bearings of recent research in Oenothera. R. R. Gates, King's College, London. 2. Cytological studies on the diploid offspring of a haploid Oenothera franciscana. J. A. Leliveld, Botanical Institute, Amsterdam. 3. The genetics and cytology of triploids and tetraploids from Oeno-

INTERNATIONAL CONGRESS OF GENETICS

59

thera franciscana. B. M . Davis, University of Michigan, Ann Arbor. 4. The fulfillment of predictions as to chromosome configuration in hybrids of Oenothera, and its significance. R. E. Cleland, Goucher College, Baltimore. 5. Cytological and genetical features of monosomic derivatives in Nicotiana Tabaczcm. R. E. Clausen, University of California, Berkeley. 6. Polyploidy in Sphaerocarpos. C. E. Allen, University of Wisconsin, Madison. 7. Chromosomes and phylogeny in Crepis. E. B. Babcock, University of California, Berkeley. 8. Comparative cytogenetic studies of tetraploid tomatoes from different origins. E. W. Lindstrom and E. W. Humphrey, Iowa State College, Ames. 9. Morphological and cytological characteristics of triploid pineapples. J. L. Collins, University of Hawaii, Honolulu. 10. Chromosome relations in somatic and meiotic divisions in violet species-hybrids. Alexander Gershoy, University of Vermont, Burlington. Animal Genetics 11, Monday, August 29 John Hammond, Chairman; P. W. Whiting, Vice Chairman and Secretary. 1. Melanic pigmentation of the mammary glands of black breeds and a red breed of pigs. Alan Deakin, Central Experimental Farm, Ottawa. 2. Genetics of silkworms. Carlo Jucci, R. Universiti di Sassari, Sassari. 3. An analysis of Mendelian phenotypes in the goldfish. H. B. Goodrich and Rowena Nichols, Wesleyan University, Middletown. 4. The inheritance of rate of growth in Daphnia longispinu. A. M . Banta and Thelma R. Wood, Brown University, Providence. 5. Genetic studies on selective segregation of chromosomes in Sciara coprophila Lint. Helen Berenice Smith, Carnegie Institution of Washington, Baltimore. 6. A case of non-disjunction in the domestic fowl. F. A. E. Crew, Institute of Animal Genetics, Edinburgh. 7. An inhibitor of gold color in chickens. L. W. Taylor, University of California, Berkeley. 8. Autosomal characters independently inherited in the domestic fowl. D. C. Warren, Kansas State Agricultural College, Manhattan.

60

PROCEEDINGS O F THE SIXTH

9. Inbreeding in White Leghorn fowls. N. F. Waters, Iowa State College, Ames. 10. Crossing, production, and exhibition of Rhode Island Reds. Frank A. Hays, Massachusetts Agricultural Experiment Station, Amherst.

Plant Genetics I, Monday, August 29 S. C. Harland, Chairman; 0. E. White, Vice Chairman and Secretary.

1. La possibiliti de transfirer par croisement plusieurs caractcres recessifs dans un mtme type de betterave. Ottavio Munerati, Stazione di Bieticoltura di Rovigo, Rovigo. 2. Linkage and the criteria of independence of genes in Oenothera. G. H. Shull, Princeton University, Princeton. 3. Recapitulation of seedling characters by nucellar buds developing in the embryo sac of Citrus. W. T. Swingle, Bureau of Plant Industry, Washington. 4. Fruit characteristics of autotetraploids in Citrus. H. B. Frost, Citrus Experiment Station, Riverside. 5. Variability and heredity in Beta vulgaris L. V . F . Savitzky, Ukrainisches Forschungsinstitut fur die Zuckerindustrie, Kiew. 6. Chromosomal aberrations as a result of transgenation. E. J. Khareeko-Savitzkaya, Ukrainisches Forschungsinstitut fur die Zuckerindustrie, Kiew. 7. The genetic basis of dimensional traits in Cucurbita fruits. E. W. Sinnott, Columbia University, New York. 8. Genetic interrelationships of some foliage, pod, and cotyledon factors in Pisum. 0.E. White, University of Virginia, University. 9. The inheritance of some plant colors in Brassica oleracea, var. capitata. Roy Magruder and C. H. Myers, Bureau of Plant Industry, Washington, and Cornell University, Ithaca.

INTERNATIONAL CONGRESS O F GENETICS

61

10. Pine and walnut breeding for timber production, Lloyd Austin, Institute of Forest Genetics, Placerville. 11. The occurrence and use of haploid plants in the tomato, with especial reference to the variety Marglobe. Gordon Morrison, Oakview Seed Breeding Station, Ferry-Morse Seed Company, Detroit. 12. Genetic association between qualitative and quantitative characters in the tomato. T . M. Currence, University of Minnesota, St. Paul. 13. Mutations in sweet potatoes. J. C. Miller. Chromosome Structure and Crossing Over, Monday, August 29

C. D. Darlington, Chairman ;j. Schultz, Vice Chairman and Secretary. 1. The general meiosis problem in the light of new facts, and its signification for the chromosome theory of heredity. M. V. Tschernoyarov, Botanical Garden, Kiew. 2. Observations bearing on the mechanism of meiosis and crossing over. C. L. Huskins, McGilI University, Montreal. 3. Meiosis as a genetic character. J. W . Gowen, Rockefeller Institute for Medical Research, Princeton. 4. Chromosome structure in Drosophila. B. P. Kaufmann, University of Alabama, University. 5. A cytological map of the X chromosome of Drosophila melanogaster. T. S. Painter and H. J. Muller, University of Texas, Austin. 6. A theoretical and experimental analysis of crossing.over. A. Weinstein, Johns Hopkins University, Baltimore. 7. Genetic behavior of a closed X chromosome of Drosophila melanogmtef. Lilian V. Morgan, California Institute of Technology, Pasadena. 8. Studies on the mechanism of crossing over in Drosophila melanogaster. I. Experiments with attached-X chromosomes. S. H. Emerson and G. W . Beadle, California Institute of Technology, Pasadena. 9. Studies on the mechanism of crossing over in Drosophila melanogaster. 11. Experiments with certain translocations. G. W . Beadle and S. H. Emerson, California Institute of Technology, Pasadena. 10. Regional differences in crossing over as a function of the chromosome structure. C. A. Offermann and H. J. Muller, University of Texas, Austin.

62

PROCEEDINGS O F T H E SIXTH

Genetics of Species Hybrids, Monday, August 29 H. Federley, Chairman; R. E. Clausen, Vice Chairman and Secretary. 1. Genetics of interspecific crosses. A. SapEhin, Odessa. 2. Hybrid emergence in grouse locust color patterns. R. K. Nabours, Kansas State College, Manhattan. 3. Inheritance of weight in a mouse interspecific cross. C. V. Green, Jackson Memorial Laboratory, Bar Harbor. 4. Variation in fertility of dove hybrids in successive generations. L. J. Cole, University of Wisconsin, Madison. 5. On heritable individual differences in the biochemical composition of the red blood cells in dove hybrids. M. R. Irwin, University of Wisconsin, Madison. 6. Heredity in guinea fowls. Alessandro Ghigi, The University, Bologna. 7. The genetical study of natural populations of Helix nemoralis and Helix hortensis. C. Diver, London. 8. Studies on hybridization of fish species in nature. Carl L. Hubbs and Laura C. Hubbs, University of Michigan, Ann Arbor. 9. Genetic and cytological studies in hybrids of Zea X Tripsacum. P. C. Mangelsdorf and R. G. Reeves, Texas Agricultural Experiment Station, College Station. 10. A study of interspecific hybrids of Vicia. Irene Sveshnikova, Timiriazeff Academy, Moscow. 11. Crepis nicaeinsis X Crepis setosa and some of the derivatives. S. L. Emsweller, University of California, Davis. 12. Species hybrids in Paonia. A. P. Saunders and G. L. Stebbins, Hamilton College, Clinton, and Colgate University, Hamilton. 13. Remote ancestral characters appearing in first-generation hybrids of Citrus and Poncirus. T. R. Robinson, Bureau of Plant Industry, Washington. General Genetics 11, Tuesday, August 30 G. H. Shull, Chairman ;J. T . Buchholz, Vice Chairman and Secretary. 1. Inheritance of thyroid-size and thyroid-structure in six crosses of purebred dogs. E. M. Vicari, Cornell University Medical College, New York. 2. The inheritance of mental aptitudes in dogs. L. F. Whitney, New Haven. 3. Genetic aspects of a socially important primate behavior pattern. R. M. Yerkes, Yale University, New Haven.

INTERNATIONAL CONGRESS O F GENETICS

63

4. Certain principles of physiological genetics. W. J. Crozier and G. Pincus, Harvard University, Cambridge. 5. Temperature modifications of pigmentation in different races of Epilachm chrysomelina. HCl6ne Timofieff-Ressovsky, Kaiser Wilhelm-Institut fiir Hirnforschung, Berlin-Buch. 6. Temperature as a tool of research in phenogenetics: methods and results. C. R. Plunkett, New York University, New York. 7. Congenital protein hypersensitiveness. Bret Ratner, New York University Medical College, New York. 8. The occurrence of gene mutations in Paramecium a.urelia. Daniel Raffel, Yale University, New Haven. Cytology 111, Tuesday, August 30 B. M. Davis, Chairman; Barbara McClintock, Vice Chairman and Secretary. 1. The morphology of the pollen grains of Petunia in relation to hybridity, polyploidy, and sterility. Margaret C. Ferguson, Wellesley College, Wellesley. 2. Incompatibility in cherries, plums, and apples. C. D. Darlington, M. B. Crane, and W. J. C. Lawrence, John Innes HorticuItural Institution, Merton. 3. The relationship of chromosomal irregularities in megasporogenesis to the fertility and fruitfulness of varieties of Mdus malus. F. S . Howlett, Ohio AgriculturaI Experiment Station, Wooster. 4. Cytological mechanism of segregation in the progeny of an allotetraploid Aquilegia. M. Skaliriska, Free University of Poland, Varsovie. 5. Multiple association of chromosomes and an instance of fragments in Rosa. Eileen W. Erlanson, University of Michigan, Ann Arbor. 6. Variation and chromosome behavior in Fragaria. S. H. Yarnell, Texas Agricultural Experiment Station, College Station. 7. Chromosome elimination during cleavage in the eggs of Sciara coprophila. Anne M . DuBois, Carnegie Institution of Washington, Baltimore. 8. Male sterility of Nicothna rusbica. M. Christoff, University of Sofia, Sofia. 9. Cytogenetics of a Nicotiana and a Triticum triple hybrid. D. Kostoff, University of Sofia, Sofia. 10. Cytological aberrations in Triticunz d g a r e . LeRoy Powers, University of Minnesota, St. Paul.

64

PROCEEDINGS O F THE SIXTH

11. Some cytological and genetical studies in the genus Melilotus. A. E. Clarke, University of California, Berkeley. Plant Genetics 11, Tuesday, August 30

M. Skaliriska, Chairman; R. A. Brink, Vice Chairman and Secretary. 1. The transmission of genes affecting pollen-tube growth in Datura. J. T. Buchholz and A. F. Blakeslee, University of Illinois, Urbana, and Carnegie Institution of Washington, Cold Spring Harbor. 2. The production of mutations in American upland cotton (Gossypium hirsuturn) by radiation. W. R. Horlacher and D. T. Killough, Texas Agricultural College, College Station. 3. ( a ) Uber die Methodik der okologischen Klassifizierung des Ausgangs-Materiels bei ziichterischen Arbeiten. (b) Das Vorriicken des Weizens nach Norden. V. Pissarev, Institute of Applied Botany, Leningrad. 4 The interaction of specific genes determining sex in dicecious maize. D. F. Jones, Connecticut Agricultural Experiment Station, New Haven. 5. Complete elimination of certain classes of gametes in Zea mays. W. R. Singleton, Connecticut Agricultural Experiment Station, New Haven. 6. Variability of sweet-corn hybrids as affected by genetic constitution. J. B. Park, Arthur Anderson, and M. T. Myers, Ohio State University, Columbus. 7. Inheritance in barley. D. W. Robertson, Colorado Agricultural College, Fort Collins. 8. Prevalence and origin of fatuoids in Fulghum oats. F. A. Coffman and J. W. Taylor, Bureau of Plant Industry, Washington. 9. Turkestan autogamous rye (Secale turkestanicum Bensin). B. M. Bensin, New York Botanical Garden, New York. 10. Genus Beta L. in the light of the new data of cytology and anatomy. V. Zossimovit, Genetics Laboratory, Ukrainian Research Institute for Sugar Industry, Kiew. 11. Mosaic segregation and chromosome behavior in Petunia. E. Malinowski, Institute of Genetics and Plant Breeding, College of Agriculture, Skierniewice. 12. Mosaic segregation in Phaseolus vulgaris. H. Bankowska, Institute of Genetics and Plant Breeding, College of Agriculture, Skierniewice.

INTERNATIONAL CONGRESS O F GENETICS

65

Drosophila, Tuesday, August 30 A. H. Sturtevant, Chairman; S. Zarapkin, Vice Chairman and Secretary. 1. The mechanism of mosaic formation in Drosophila. J. T . Patterson, University of Texas, Austin. 2. Specific suppressors in Drosophila melanogarter. C. B. Bridges, California Institute of Technology, Pasadena. 3. The developmental system affected by the genes for eye color in Drosophila melanogaster. J. Schultz, California Institute of Technology, Pasadena. 4. A study of dominant mosaic eye-color mutants in Drosophila w lanogaster. H. B. Glass, University of Texas, Austin. 5. The effect of long-continued subjection to constant temperatures in darkness upon inbred bar-eyed Drosophila. Charles Zeleny, University of Illinois, Urbana. 6. New evidence of the production of mutations by high temperature, with a critique of the concept of "directed mutations." H. H. Plough and P. T:Ives, Amherst College, Amherst. 7. The temperature-effective period for the lengthening of the vestigial wings of Drosophila melanogaster. M. H . Harnly, New York University, New York. 8. The development and minute structure of certain hereditary tumors in Drosophila. Mary B. Stark, New York Homeopathic Medical College, New York. 9. Changes in the instability of miniature-3 gene of Drosophila virilis during ontogeny. M . Demerec, Carnegie Institution of Washington, Cold Spring Harbor. 10. The analysis of body-stature in Drosophila funebris. S. Zarapkin, Kaiser Wilhelm-Institut fiir Hirnforschung, Berlin-Buch. 11. Sex-linked inheritance in Drosophila hydei. W. P. Spencer, Wooster. Problems Relating to Sex and Fertility, Tuesday, August 30 Hans Nachtsheim, Chairman; Franz Schrader, Vice Chairman and Secretary. 1. Inheritance of sex in oysters. W. R. Coe, Yale University, New Haven. 2. Sex and intersex in pigeons. Oscar Riddle, Carnegie Institution of Washington, Cold Spring Harbor. 3. Genetics of sexual dimorphism in plumage. C. H. Danforth, Leland Stanford Jr. University, Stanford University.

66

PROCEEDINGS O F THE S I X T H

The inheritance of fertility in the rabbit. John Hamrnond, School of Agriculture, Cambridge. The effect of X-rays on the fertility of the male house mouse. G. D. Snell, University of Texas, Austin. Genotypische und phanotypische Geschlechtsbestimmung bei Zahnkarpfen. Curt Kosswig, Die Universitit, Miinster i. W. Intermediate aphids and the time-of-determination theory. A. F. Shull, University of Michigan, Ann Arbor. Production of diploid binuclear oidia, diploid binuclear chlamydospores, and haploid mononuclear oidia, on the same diploid strain of Pholiota aurivella Batsch. Ren6 Vandendries, Rixensart. The production of vestigial and sterile sex-organs through sex-reversa1 and neutral sexual states. J. H. Schaffner, Ohio State University, Columbus. Studies of self- and cross-incompatibility in the petunia "Rosy Morn." A. B. Stout, New York Botanical Garden, New York. Selbstfertilitiit bei diozischen Pflanzen mit besonderer Beriicksichtigung von Versuchen an A n t e m r i a dioica. Gerta von Ubisch, Die Universitat, Heidelberg. Genetics and Pathology, Tuesday, August 30 Crew, Chairman; L. C. Strong, Vice Chairman and Secretary. Inheritance of resistance to disease in animals. Elnier Roberts, University of Illinois, Urbana. A lethal factor in sheep. G. K. Constantinescu, The University, Bucharest. Eight new mutations in the domestic fowl. F. B. Hutt, University of Minnesota, St. Paul. Genetic selection for resistance to fowl typhoid in the chicken. W. V. Lambert, Iowa Agricultural Experiment Station, Ames. Concerning the existence of genes with a specific effect upon one germ layer. Walter Landauer, Storrs Agricultural Experiment Station, Storrs. The genetic basis of resistance to paratyphoid in mice. R. G. Schott, Rockefeller Institute for Medical Research, Princeton. Hereditary anomalies in mice descending from stock raised (1921) by Little and Bagg. Kristine Bonnevie, Zoological Laboratory, Oslo. The genetics of spontaneous cancer in mice. I. Cross between dilute browns and yellows. C. C. Little and B. W. McPheters, Jackson Memorial Laboratory, Bar Harbor.

INTERNATIONAI, CONGRESS OF GENETICS

67

9. Mouse leukemia. E. C. MacDoweII, Carnegie Institution of Washington, Cold Spring Harbor. 10. Further studies on the inheritance of tumor susceptibility in mice. Clara J. Lynch, Rockefeller Institute for Medical Research, New York. 11. Heredity of cancer susceptibility in mice. N. DobrovolskaiaZavadskaia, Institute for Radium Research, Paris. 12. Genetic studies on the transplantation of tumors. IV. Linkage in tumor 19308A. J. J. Bittner, Jackson Memorial Laboratory, Bar Harbor. Wednesday, August 31 Section of Fruit and Vegetable Breeding, at the New York Agricultural Experiment Station, Geneva, New York W. T. Macoun, Chairman; R. Wellington, Vice Chairman and Secretary. 1. Address of welcome. U. P. Hedrick. 2. Recent progress in the raising of blight-immune potatoes. R. N. Salarnan, Barley, Herts. 3. Observations on the genetics of the potato: (a) Effects of inbreeding. (b) Interacting factors affecting tuber color. F. A. Krantz, University of Minnesota, St. Paul. 4. A survey of bud mutations among deciduous fruit varieties. J. T. Bregger, Washington State College, Pullman. 5. The importance of the parental genotype in the breeding of fruits. A. N. Wilcox, University of Minnesota, St. Paul. 6. Somatic segregation of an environmental character (hard shell) in pure lines of beans. W. 0. Gloyer, New York Agricultural Experiment Station, Geneva. 7. The morphological expression of diceciousness in the grape. M. J . Dorsey, University of Illinois, Urbana. 8. The value of the European grape (Yitis zkifera) in breeding grapes for New York State. R. Wellington, New York Agricultural Experiment Station, Geneva. 9. The Northern Spy apple : a parent in breeding new varieties. W . T. Macoun, Central Experimental Farm, Ottawa. 10. Metaxenia and xenia in apples. B. R. Nebel, New York Agricultural Experiment Station, Geneva. 11. Metaxenia in the date palm, and its genetic implications. R. W. Nixon, Bureau of Plant Industry, Washington. 12. Raspberry and strawberry breeding at the New York Agricultural Experiment Station. G. L. Slate, New York Agricultural Experiment Station, Geneva.

68

PROCEEDINGS O F THE SIXTH

EXHIBITS M.Demerec

The readiness with which geneticists took part in exhibits indicates that exhibits will not remain a special feature of the sixth congress but that they may appear again at future congresses as a prominent part of future program. Our experience in organizing exhibits may then be of help to the organizers of congresses. With this in mind there is here given a short review of our experiences. The organization committee of the Sixth International Congress of Genetics, in carrying out the proposal to make exhibits a special feature of the congress, appointed a committee on exhibits, the chairman of the committee being a member of the executive council. On February 20, 1931, a meeting of the exhibits committee was held with CLELAND,DEMEREC, LAUGHLIN,RANDOLPH, RICHEY,SATINAand WRIGHTattending. A t this meeting the scope of exhibits was decided upon and the organization plan outlined. It was agreed that an attempt be made to bring together an extensive collection of characters and types of organisms used in genetic investigations, to show methods employed in genetic research, and to present the results obtained. It was decided that exhibits be arranged under organisms and that the organization be decentralized, the responsibility for each section being given to one or more persons interested in the material which this particular section was to represent. The work on organizing exhibits began immediately after the meeting of the exhibits committee. Requests were sent to a selected group of geneticists asking them to take the responsibili'ty of organizing the different sections of exhibits. All those who accepted were requested to prepare as soon as feasible a tentative outline of exhibits planned for their section. These outlines were circulated among the members of the exhibits committee and the members of the council for suggestions and criticism. I n October, 1931, the exhibits number of the Genetics Congress Quarterly was published giving a list of exhibits organized up to that time and asking geneticists to suggest any improvements of or additions to the list. Several suggestions were received. During the summer of 1931 several tests were made in the garden to determine the best planting time for several plants which were to be grown for exhibits. Also extensive tests were made to find out most suitable grass composition for the paths between the exhibit plots. I n October information about exhibits was sent to all those in charge of exhibits giving details as to garden and laboratory facilities, micro-

INTERNATIONAL CONGRESS O F G E N E T I C S

69

scopic equipment, and the arrangement for publishing the descriptions. Instructions were also given regarding the exhibit signs and the shipment of exhibit packages. Requests were made that seed for live plant exhibits, titles for large signs and an estimate of the number of microscopes needed be sent as soon as feasible. April 20 was set as the last date for the submission of exhibits manuscripts and statements regarding the wall space and the table space requirements. When it became known that about 200 pages of the second volume of the Proceedings would be available for the descriptions of exhibits it became essential to organize the material so as to conform with the space allotted. The space was therefore apportioned between different sections according

to the number of individuals taking part in exhibits of the respective sections. In March, 1932, letters were sent to those in charge of sections notifying them about the space in the Proceedings allowed for descriptions of their exhibits. The work on setting up garden exhibits began in February, 1931, when the Oenothera seed was planted. From that time until the end of the congress meetings, garden exhibits required continuous attention, thought and labor. This was freely contributed by R. A. EMERSON and M. M. RHOADES ; the latter especially deserves much credit for the success of the garden exhibits since he was in charge and spent much time and effort on this feature. The chairman of the exhibits committee spent July and August at Ithaca. During that time estimates of the space requirements received from persons in charge of sections were worked over, and assignments of exhibits to rooms were made; copies of the plan of each exhibit room were prepared and sent to all persons organizing the exhibits; main signs for indoor ex-

70

PROCEEDINGS O F THE SIXTH

hibits and all signs for garden exhibits were made; demonstration periods for exhibits were arranged; a classified and a chronological list of exhibits was prepared for the program. T o facilitate the installation of exhibits, a graduate student was assigned to every room to help the exhibitors. Tools and a supply of thumb tacks, string, wire, hooks, etc., were available in each room. Exhibits were open during the whole session of the congress. I n addition each exhibit was demonstrated, usually twice, during a period stated in the program by either the exhibitor o r a person familiar with the material. In spite of the large number of microscopes on hand not enough instruments were available to go around. The instruments, therefore, had to be transferred daily t o the rooms where the demonstration of exhibits was held that particular day. The responsibility for the efficient handling of the microscope situation rests with L. F. RANDOLPHwho had charge of the optical equipment. Upon application, the commissioner of customs recognized the congress as an edtlcational institution which gave us a right to import for exhibit purposes, under bond, dutiable articles. As a special courtesy toward the members of the congress, customs officials of the Port of New York took special precautions in examining exhibit packages brought as personal baggage. These arrangements facilitated greatly the handling of foreign exhibits. Exhibit packages which arrived early were opened and, whenever feasible, the exhibits were put up. The majority of exhibits were taken down and packed by the exhibitors. Extensive exhibits on fruit genetics, breeding and cytology were organized and were shown a t the NEW YORKAGRICULTURAL EXby R. WELLINGTON PERIMENT STATION a t Geneva in conjunction with the meetings of the fruit and vegetable breeding section of the congress. One of the main advantages of exhibits is the broadening influence they have on the program of a scientific congress by making it possible to include material which it is not feasible to handle by platform papers. This is especially true where summaries of extensive problems are shown including material already published. Exhibits, in such a case, give a greater degree of freedom and make the presentation more complete. They may prove an efficient means of relieving the program of short papers which are usually so numerous as to make it impossible for any one person to hear everything he is interested in. Another important function of exhibits is the opportunity they afford to the members of the congress to become familiar with the

INTERNATIONAL CONGRESS O F GENETICS

71

material used in research. This is of special importance where live plant exhibits and cytological exhibits can be arranged. Exhibits also stimulate informal discussion which tends to bring out problems which otherwise might pass unnoticed. During the congress exhibit rooms served as convenient meeting places for congenial gatherings using the material shown in exhibits as a basis for discussion. This was especially evident on Sunday when exhibit rooms and the garden were well filled by groups of interested spectators. Finally, the exhibits offer the opportunity for personal contacts as does no other function of the congress. Even a most

retiring person will easily find an opportunity to approach a person demonstrating his exhibits, in order to ask a cluestion 01- to start a discussion. It is not easy to find at one place facilities necessary for extensive exhibits, and this fact may prove to be a serious obstacle which may prevent the exhibits from beconling a pernlanent part of the program. The exhibits at CORNELL UNIVERSITY, for example, occupied more than two acres (about one hectare) of garden space, 38 laboratories on the average 14 X 14 meters, and three corridors. Exhibitors requested 161 microscopes with either fluorite or apochronlatic lenses, 135 oil inlmersion microscopes, 239 low power n~icroscopesand 36 Greenoughs. Even the excellent equipment available at CORNELL UNIVERSITY, consisting of 6 0 microscopes with either apochronlatic or fluorite lenses, 100 microscopes with oil immersions. 186

72

PROCEEDINGS OF THE SIXTH

low power microscopes and 77 Greenoughs was not sufficient to satisfy this demand. It was necessary to borrow lights and high power oculars from manufacturers of optical instruments and to arrange for the high power microscopes to be available to the exhibitors during demonstration periods only. I n connection with the use of microscopes a difficulty with the electric current was encountered. In the Drosophila and Sciara exhibits, for example, 98 microscopes were used at once in two adjacent rooms. The microscope lights required more current than the wiring equipment could carry so that it became necessary to bring in three additional circuits to supply the necessary energy. Similar arrangement was required for three other rooms. The main credit for the success of the exhibits is due to the splendid cooperation of the geneticists who contributed time and effort and did not hesitate a t the expense necessary to make the exhibits an outstanding feature. Well over ninety percent of them gave a ready response to the request for exhibits. Over 400 took part in the undertaking. A few, however, failed even to answer repeated correspondence. Fortunately for the organizers, the percentage of the latter was smaller than could have been expected in an average population, and the inconvenience they caused was not significant enough to affect the results. Organizers of sections also deserve a large share of the credit for the success. They took a heavy load of responsibility off the shoulders of the exhibits committee and relieved the chairman of a great deal of correspondence. With the present organization more than 1200 letters and more than 500 circular letters were sent by the chairman. This correspondence would have been much heavier if the work had not been decentralized. for his helpful attitude and conCredit is also due to R. A. EMERSON tinuous interest in the exhibits, to MARCUSM. RHOADESfor the garden for the efficient handling of the microscopes, exhibits, to L. F. RANDOLPH to the members of the local committee for cooperation, and to graduate R. M. HAFF,S. HOROWITZ, W . G. HOUK,0 . A. KRUG, students F. CORREA, P. MA, W. R. MILLS, L. G. MILES, J. A. B. NOLLA,G. W. SALISBURY, S. T. SHEN,K. L. TURKand L. E. WOLF for the efficient assistance they gave the exhibitors. which unhesiT o the biological departments of CORNELLUNIVERSITY, tantly gave use of their facilities and optical instruments for exhibits purposes and to the administrative officers of the university, who helped whenever an occasion arose, appreciation of the exhibits committee is extended. OF WASHINGTON helped make the exhibits The CARNEGIE INSTITUTION a success by allowing the chairman to spend much of his time on exhibits

INTERNATIONAL CONGRESS O F GENETICS

73

work and by appropriating funds from which the office expenses and the secretarial expenses of the exhibits committee were defrayed. Miss MIRIAM KORTRIGHT and Miss EUNICEWHITEdid creditable work as part-time secretaries. Acknowledgment is here made of loans by the BAUSCHA N D LOMBOPTICAL COMPANY of the automatic balopticon, six euscopes with lights, thirty 15X oculars, and eight microscopes and eight lamps for the Geneva exhibits; by the SPENCERLENS COMPANY of seventeen pairs of oculars and sixty lamps; by the EASTMAN KODAKCOMPANY of fifty Wratten filters; and by the BAUSCHA N D LOMBOPTICALCOMPANY, E . LEITZ, I N CORPORATED, the SPENCERLENS COMPANY, and CARL ZEISS, of optical equipment with recent improvements.

71

PROCEEDINGS OF THE SIXTH

EXCURSIONS A N D E N T E R T A I N M E N T S Entertainillents provided for the members and guests of the Sixth International Congress of Genetics were all informal. On Wednesday evening, August 24, an informal reception was given in Willard Straight Hall by the executive council. On Friday afternoon, August 26, from 4 to 8:30 a picnic a t Taughannock Falls State Park, in charge of L. H. MCDANIELS, was attended by about 800 members and guests of the congress. The afternoon sports, baseball, horse-shoe pitching, and swimming, as well as the picnic supper and later the community singing, were typical of large-scale American picnics. A special feature arranged by E. A. BATES, primarily for the entertainment of foreign delegates, was ceremonial songs and dances by a group of American Indians from the Onondaga Reservation near Syracuse, New York. The music of the week was in charge of PAULJ. WEAVER.On Sunday afternoon, August 28, an organ recital was given in Sage Chapel by DAVID HUGHJONES,organist of the WESTMINSTER CHOIRSCHOOLof Princeton, New Jersey. The university chimes were played three times each day. The program consisted largely of the music of the several countries represented at the congress. Nun~erousexcursions to points of interest were a prominent feature of the congress. These were organized by R. G. WIGGANSwith transportation arranged for by C. H . MYERS.Each afternoon automobile trips were taken to either Enfield Glen State Park, Taughannock Falls State Park, Watkins Glen State Park, o r to points of interest in the city of Ithaca, through the campus of the university, and to the agricultural college farm and experimental fields and gardens. On Sunday, August 28, an all-day excursion was made by railway train to Niagara Falls. Special entertainment was provided for the women guests by a committee under the chairmanship of Mrs. R. A. EMERSON,and Mrs. C. H. MYERS.Afternoon teas were served in the library of Willard Straight Hall. Tours were made of the university campus with visits to some of the buildings of general interest; some of the more interesting private gardens of Ithaca were visited and excursions were made to near-by glens and state parks. Luncheons and card parties were also given. Entertainment of boys and of girls from seven to fourteen years of age was provided on certain days by the Ithaca Boy Scout Council and by the Ithaca Girl Scout Co~tncil.Children below the age of seven years were cared for by a committee of young women at the nursery school of the College of Home Economics.

INTERNATIONAL CONGRESS O F GENETICS

75

G R O U P CONFERENCES R. A. Eltaerson

The program of the Sixth International Congress of Genetics was so arranged that the evenings were left open for impromptu discussions by individuals who chanced to be together and for informal conferences by small groups with some common interest. The latter were arranged from time to time during the week of the congress and announced in the Daily News Bulletin and were held in the "student-activity" rooms in Willard Straight Hall. Seven such conferences were held during the week at which the following topics were discussed: Gene problems Unstable genes Poultry genetics Maize genetics

Mouse genetics Human genetics Sire valuation

In addition to these conferences, luncheons or dinners were held by the OF MINNEformer students and members of the staff of the UNIVERSITY SOTA,by a similar group known as the "Linkage Group" of the UNIVERSITY OF WISCONSIN,and by the "Synapsis" Club of CORNELL UNIVERSITY. the latter celebrating its twenty-fifth anniversary at this time.

76

PROCEEDINGS OF THE SIXTH

TRANSPORTATION, T O U R S A N D L E C T U R E S L.C. Dunn

The transportation committee was charged with making arrangements for travel to Ithaca by delegates from the United States and from abroad, with the entertainment of foreign delegates in New York and with arrangements for tours in the United States after the congress. Its chief efforts were expended in making those arrangements which would enable the largest number of visitors from abroad to come to the congress and to this end it obtained favorable rates and accommodations for group travel from Europe, arranged for the entertainment o.f delegates on their arrival in New York, and undertook to obtain lecture engagements for visiting geneticists which might help to defray the cost of travel to the congress. Entertainment of foreign members of the genetics and the eugenics congresses while they were in New York City preceding the genetics congress was in charge of the New York entertainment committee which had the following membership: A. F. BLAKESLEE,L. C. DUNN,W . M. FAUNCE, E . W . SINNOTT, C. S. GAGER,F. E. LUTZ,E . D. MERRILL,F. SCHRADER, A. B. STOUT,and C. R. STOCKARD. Entertainment in New York City was provided through the generosity of the CARNEGIE ENDOWMENT FOR INTERNATIONAL PEACE. COLUMBIA UNIVERSITYarranged to have its facilities held available for members of the congress and delegates were lodged in its dormitories and provided with breakfasts at the Faculty Club. O n Sunday, August 20, an excursion was arranged to Cold Spring Harbor where the foreign delegates were guests of the Department of INSTITUTION OF WASHINGTON, and inspected the Genetics of the CARNEGIE EVOLUTION and the Eulaboratories of the STATIONFOR EXPERIMENTAL GENICS RECORD OFFICE. On Monday, August 22, delegates had the choice of two all day excursions: one taking them to the NEWYORKBOTANICAL GARDEN and to the BOYCET H O M P SINSTITUTE O~ FOR PLANT RESEARCHat Yonkers, where of the Boyce they were entertained at lunch by Doctor and Mrs. CROCKER Thompson Institute; and the other to Doctor STOCKARD'S dog farm near OF PHYSICIANS AND SURGEONS and the Peekskill by way o f the COLLEGE PRESBYTERIAN HOSPITAL MEDICALCENTER.This group was entertained at luncheon by Doctor STOCKARD. Delegates were luncheon guests of the AMERICANMUSEUMOF NATURAL HISTORYon both Monday and Tuesday and opportunity was given t o inspect the museum under the guidance of the staff. Student guides from COLUMBIAUNIVERSITYunder the supervision of

INTERNATIONAL CONGRESS O F GENETICS

77

D. E . LANCEFIELD and F. SCHRADER stood ready on request to conduct the delegates to places of special interest in the city. O n Tuesday there was arranged a sight-seeing tour of the city which included a trip to the top af the Empire State building. Tuesday evening there was a dinner a t the Waldorf-Astoria Hotel given by the CARNEGIE ENDOWMENT FOR INTERNATIONAL PEACE; E . W . SINNOTTwas in charge of dinner arrangements and F. T. MCLEANwas responsible for the table decorations which consisted of genetic material including fruits of hybrid EXPERIMENT STApeaches presented by the NEW JERSEYAGRICULTURAL TION. A. F. BLAKESLEE presided and introduced the following speakers: THOMASDARLINGTON who represented New York City; E. W . SINNOTT who read a communication from E. B. WILSON who was unable to be present on account of illness; the president of the genetics congress, T. H . MORGAN; the president of the eugenics congress, C. B. DAVENPORT, and J, B. S. HALDANE,KRISTINEBONNEVIE,R. GOLDSCHMIDT, and N. I. VAVILOV. Transportation in Ithaca was in charge of C. H . MYERSwho, with the help of the local committee, arranged also an excursion to Niagara Falls. After the congress a group of delegates from England, Finland, Norway, France, Italy, Belgium, Poland, Denmark and Germany went on a week's excursion through New England, which had been organized by W. LANDAUER. They were entertained a t Williamstown, Massachusetts, by PRENTICEof Mount Hope F a r m ; a t AMHERSTCOLLEGE Mr. PARMALEE EXand the MASSACHUSETTS STATECOLLEGE,at STORRSAGRICULTURAL PERIMENT STATIONA N D CONNECTICUT AGRICULTURAL COLLEGE,WESLEYAN UNIVERSITY, CONNECTICUT AGRICULTURAL EXPERIMENT STATION, YALEUNIVERSITY, WELLESLEYCOLLEGE,and HARVARD UNIVERSITY. The last day was spent a t the MARINEBIOLOGICAL LABORATORY at Woods Hole, Massachusetts, and several of the delegates remained at Woods Hole as guests of the laboratory. Plans for other excursions had to be abandoned because of the small number of applicants. The committee, with the cooperation of the Institute of International Education and of American colleges, universities and research institutions, was able to make a number of arrangements for lectures by delegates to the congress during the fall and to enable them to travel in the United States and Canada. As a result ten visitors from abroad lectured at twentyfive institutions. The committee gratefully acknowledges the receipt of $1500 from CoLUMBIA UNIVERSITY which was expended for secretarial assistance between December 1, 1929 and September 1, 1932; and of $1352.31 from

78

PROCEEDINGS OF THE SIXTH

the CARNEGIE ENDOWMENT FOR INTERNATIONAL PEACE which Was expended for entertainment of foreign delegates in New York as follows: Lodging and meals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .$ Transportation and comn~unication . . . . . . . . . . . . . . . . . Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

890.85 143.31 193.75 63.50 60.90

Detailed vouchers for all expenditures are on file in the office of the Bursar of COLUMBIA UNIVERSITY. and later Mrs. G. W. LITTLE served as secretary to Mrs. C. H. HECHT the con~n~ittee, an arrangement made ~ossibleby the generosity of COLUMGI.\ UNIVERSITY.

INl'ERNATIONAL CONGRESS O F GENETICS

79

TREASURER'S R E P O R T R. C. Cook

I t has proved difficult to allocate all expenditures accurately to all the various committees whose combined activities made possible the congress. The principal reason for this has been the cooperative handling of the work. Thus, much of the printing, stationery, badges, etc., were ordered through the treasurer's office, as he happened to have contacts with printers and manufacturers that made this desirable. Similarly the secretary general's office assisted in handling the correspondence of the program committee, and during July and August, 1932, assisted the local committee and exhibits committee at Ithaca. The accounts of the exhibits and local committees were handled jointly at Ithaca and during August and September were paid from a fund set up at CORNELL UNIVERSITY. TO allocate the expenditures of each of these conln~ittees accurately would hardly enlighten us as to how they performed such marvels on so little, so the attempt has not been made. The treasurership of the congress has necessarily involved more than bookkeeping, as membership cards and transportation certificates were also issued at the treasurer's office. A considerable correspondence was necessary in connection with the ptlblishing of the bulletin, acknowledging contributions, answering inquiries, etc. The treasurer's office also maintained a stencil file of congress members and published the quarterly bulletin of the congress, issued during 1931 and 1932. This and other literature sent to all congress members, and to selected lists of geneticists and workers in allied fields, was handled by the treasurer. This involved considerable labor in checking lists (in cooperation with the secretary general's office), but the support obtained more than justified the effort and outlay. The facilities of the AMERICAN GENETICASSOCIATION (office equipment, stencil cutting ancl printing machines, multigraph and mimeograph, etc.) were placed at the disposal of the congress. Thus it was possible to take care of this phase of congress organization with no expense for equipment or office rent, and with only a small outlay for supplies and for part time clerical assistance. Without this equipment, the handling of the accounts and other activities mentioned above could have been carried on only at a much greater expense. The accounts for 1930 and 1931 have been audited by committees appointed by the council. The report of these auditing committees is to be found in the minutes of the council meetings. I t will not be possible to render a final accounting of congress funds

80

PROCEEDINGS OF T H E SIXTH

until the first volume of the proceedings, now in preparation, has been printed and distributed. There are also a few items due the congress which have not been collected. As soon as possible after the proceedings are published, the treasurer will submit the accounts of the congress to the council for a final audit. The grants of $1352.31 and of $1500 expended by the transportation committee and the grant of about a thousand dollars expended by the CARNEGIE INSTITUTION OF WASHINGTON were disbursed directly by COLUMBIA UNIVERSITY and by the CARNEGIE INSTITUTION. They were in addition to the sums disbursed by the treasurer. Thus the actual total budget of the congress was by these amounts in excess of $17,583.58. I wish also to acknowledge my indebtedness to Mrs. F. S. TULLOSS,whose competent handling of many matters alone made it possible to undertake much of the congress work handled through this office. RECEIPTS Personal memberships . . . . . . . . . . . . . . . . . . . . . . . . . . .$ 7,582.38 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7,007.85 Institutional memberships . . . . . . . . . . . . . . . . . . . . . . . 2,850.00 143.35 Interest and miscellaneous . . . . . . . . . . . . . . . . . . . . . . .

. Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .$17,583.58 DISBURSEMENTS Secretarial and clerical assistance . . . . . . . . . . . . . . . . . $ 4,395.24 Printing: Volume 2 . . . . . . . . . . . . . . . . . . . . . . . . . . $2,378.08 . Estimated cost of volume 1 . . . . . . . . . . . . 2,550.00 4,928.08 Travel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2,445.58 Correspondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 15.62 Exhibits and local expenses . . . . . . . . . . . . . . . . . . . . . . 2,681.90 Entertainment of foreign delegates at Ithaca . . . . . . . . 523.08 394.29 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Balance December 15, 1932 . . . . . . . . . . . . . .$3,949.79 Less estimated cost of volume 1 . . . . . . . . . 2,550.00 1,399.79

. Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .$17,583.58

INTERNATIONAL CONGRESS O F GENETICS

OPENING ADDRESS' Edmund B. Wilson, Columbia University, N e w York, N e w York

Members of the genetics and eugenics congresses: We had hoped and expected that the members of the congresses would be greeted tonight by COLUMBIA UNIVERSITY, in the person of its president, Doctor NICHOLAS MURRAY BUTLER,but to our acute regret he found it impossible to be present. In his unavoidable absence I have been honored by the request to speak for him. Unfortunately I am prevented by illness from being here in person and must send my message in written form. I cannot in any degree fill Doctor BUTLER'Splace but I am most happy to convey to you a warm greeting and welcome from him and from the university of which he is the head. Every geneticist should find himself at home at COLUMBIA UNIVERSITY which has for many years been a centre of research in this field under the leadership of the president of the genetics congress, Professor T. H. MORGAN. A prominent feature of genetic research at COLUMBIA UNIVERSITY has been the study of the phenomena of crossing over and their cytological interpretation. The discovery of these phenomena was one of commanding importance in the history of genetics, although we must admit that we do not yet fully comprehend their underlying mechanism. But while we have been trying to puzzle this out we have discovered another kind of crossing over, one that is no less important for the advance of our science and may in the end prove to be even more significant for the advance of our civilization. The presence here to-night of our friends from Europe bears witness to this, for they have crossed the sea in order to enter into synaptic union with us, to perform a friendly exchange of genes and, let us hope, to induce rejuvenescence in both. Crossing over of this type is a phenomenon of good augury, not alone for the advancement of science but also for the promotion of good will and better understanding .between the nations of the world. Our welcome to those of our friends who have come from older countries beyond the seas is therefore one of especial warmth in these troubled times. W e Americans are sometimes thought of as a rich and powerful creditor nation which seems to stand coldly aside with selfish indifference to the severe trials through which the older nations of Europe are passing. It must be said that a few of our public men have sometimes seemed to give such an impression, but these men do not represent the attitude of our most

'Address at the dinner tendered to the delegates to the genetics and eugenics congresses at the Waldorf-Astoria Hotel in New York City on August 23, 1932.

82

PROCEEDINGS O F THE S I X T H

intelligent and right-thinking people. And I am proud to say that in my opinion no man in the United States works more constantly, more intelligently and more effectively for the cause of international reconstruction, peace and good will than the man whose greeting to you I bear, the president of COLUMBIA UNIVERSITY. For my own small part, I am one of the multitude who are neither rich nor powerful, and certainly my knowledge of international finance is so little that only a microscope of high power could make it visible; but let me assure you that we men of science, above all perhaps we of the older generation, can never forget nor sufficiently recognize the enormous debt that we owe to our brothers beyond the seas. T o speak of this debt is to utter a comn~onplace.But for me, a veteran of the old guard, it is far from a commonplace. Do not grudge me the pleasure of saying that I can never forget the uniform and helpful kindness and good will which during my LYanderjahre in Europe I received from men of science, old and young alike. Some of these men I knew only slightly. Others, like ANTONDOHRN BOVERI,came to be numbered among my dearest friends. Some or THEODOR were famous leaders in their day, such as HUXLEY,HAECKEL,LEUCKART others were still obscure students. But I cannot remember and VIRCI~OW; one instance of rudeness or harshness on their part and from many I drew encouragement and inspiration. Let me turn for a few moments to the field of genetics. I do this with some diffidence ; for, as you well know, I am not and never was a geneticist and can be accepted as such only by courtesy. I t may therefore surprise you if for a moment I boast of a certain achievement in that field, one for which I have never received any credit. More than forty years ago I discovered a new and strongly dominant Mendelian character which I was able to recognize as such long before the resuscitation of MENDEL'Swork in 1900. That character is well known to you all, for it is at this moment H. MORGAN.At the honored president of this genetics congress, THOMAS that time he, like myself, was squandering his talents on embryology, a subject for which he had a passion from which he has never recovered. I t is in fact an open secret that even now he sometimes escapes from the austere heights where Drosophila has its home in order to indulge in the illicit pleasures of the egg and its development. But even a t that early time I saw great possibilities for genetic experiment with this character. The opportunity for such experiment did not come until the year 1904 when I was able to effect a cross between T. H. MORGANand the Department of Zoology a t COLUMBIA UNIVERSITY of which I happened to be a t that time the executive officer.

INTERNATIONAL CONGRESS O F GENETICS

83

The experiment succeeded beyond my wildest expectations. Already in the F, generation appeared a group of strong dominants, such as A. H. STURTEVANT, CALVINBRIDGES,H. J. MULLERand others. After this, I had only to stand, so to speak, on the side lines, a mere observer of the successive stages of the grand game which they played. I watched with sympathetic interest the successive discoveries in that little laboratory by which the stately edifice of the Drosophila genetics was reared. It has been suggested that this should culminate in a Zeitschrift fiir Drosophilalehre und Morganhunde. This periodical has not yet actually appeared, but I hope to live to see it realized in my time. When it takes on actual form, I shall claim it as my greatest achievement. All good wishes for the success of the congress and greetings to you all.

84

PROCEEDINGS O F THE SIXTH

ADDRESS OF WELCOME A. R. Mann, Cornell Uniz*ersity,Zthaca, New Yorkl

Mr. Chairman, Ladies and Gentlemen : CORNELL UNIVERSITY is sensible of the honor bestowed upon her by the selection of this university as the meeting place for the Sixth International Congress of Genetics. The presence here of the NEW YORKSTATECOLLEGE OF AGRICULTURE, with its achievements in genetics and its productive work in many related fields of plant and animal science, has made it posUNIVERSITY to serve as host to this congress. sible for CORNELL By reason of our physical remoteness from many of the world's great centers of learning, the participation of the American states in international gatherings of men of science and learning hds never been in proportion to our interest in such gatherings or our desire to share in their benefits. When, therefore, such meetings are held on this continent, drawing to us the leaders from many lands, our anticipation of them assumes large proportions. This is the fourth international congress dealing with scientific problems of biology and agriculture held at Ithaca within a period of seven years, bringing to this campus distinguished figures from all the world. I t has been an extraordinary privilege. In any field of learning, and especially in the experimental sciences, contact among the workers in the field is a recognized aid toward reliable progress and a great incentive to superior achievement; and it is conducive to that humility of spirit which appropriately envelops careful scientific study. Correspondence and the interchange of publications between individuals-never sufficiently well done-must always be the main dependence ;but the association is vastly enriched when personal acquaintance has entered and confidence is established for the free exchange of ideas. This is, indeed, one of the most valuable products of such international gatherings. This congress is primarily representative of the colleges and universities of the world. The character of your program emphasizes the significance of the universities as places of research and as contributors to the world's store of knowledge and to man's understanding of the universe. It is a revelation if not a consternation to the laymen whither your studies lead' you into the hidden mysteries of life. The geneticist is beginning to pry open some of the deepest concerns of life, age-old in their interest and speculation, of major importance, yet long baffling understanding. The rays of light already admitted stir the imagination not only as to impending Provost of CORNELL UNIVERSITY.

INTERNATIONAL CONGRESS O F GENETICS

85

knowledge of life but also as to the ultimate utility for human, plant, and animal betterment. When one contemplates such work he sees not only its importance in the enlargement of human understanding; he is equally impressed with the enrichment it brings to the university as a place for the higher training of students for productive careers both in learning and in the manifold occupations and professions in the world of affairs. It is well that even grave physical distress and economic crises shall not check or thwart the progress of discovery and interchange of ideas of such significance for human development. The field of exploration called genetics is one of the newest and most difficult outreaches of scientific effort, even though mankind has always speculated about the processes of reproduction and heredity. Set apart as a specialized field of inquiry less than a third of a century ago, clothed with obscurity as to method of attack, it has, within the present generation and with the aid of closely related fields of science, established its place, marked its course, and drawn into itself some of the ablest and keenest minds in the universities of the world. The membership of this congress and the content of your program are expressive of the high worth and recognition of your field of science. That this congress will give clearer insight and greater meaning and impetus to your efforts there can exist no question. Honored by the presence at this time of so great a host of fellow workers from our sister states in America, and doubly honored by the great number of distinguished leaders in genetics and in related fields of biological science from other nations, whose presence makes this congress particularly places every facility of this institution a t notable, CORNELLUNIVERSITY your command. We wish to serve you in whatever ways will make your brief sojourn here most comfortzable and profitable. W e trust that you will avail yourselves of our readiness to respond to your interests and desires. On behalf of my colleagues and of the administrative officers of the university, I cordially greet you and bid you hearty welcome.

S6

PROCEEDINGS O F THE SIXTH

RESPONSE Richard Goldsclzri~zdt, Kaiser Willzelm-lfzstitttt fur Biologic, Berlirz-Dnhlem, Germany

May I be permitted to thank you, Mr. PROVOST, in the name of the members of this meeting for the friendly and gracious words of welco~newhich we enjoyed hearing now. There is not a single one among us who has not looked forward to this meeting five long years. Finding ourselves now on this beautiful campus which is towering above the ordinary abode of men like an island of peace and beauty, our mind most naturally turns to the Homeric Island of Ithaca and the adventures of the hero Odysseus passing through all possible dangers, his heart set to his return to Ithaca. Certainly the modern Odysseus, the geneticist, had to undergo no less hardships on his way to Ithaca. There was pessimism, disguised as the nymph Calypso, who tried to keep the wanderer lulled into lassitude in the glittering cave of inactivity, from where he would never again have found his way home to Ithaca. There were Scylla and Charybdis, one urging to cancel the ineeting, the other to postpone it. But Odysseus sailed clear of both towards his goal. And there was the Cyclops Polyphemus in the garb of econoillic depression, trying to smash up everything and everybody. Though he succeeded in decimating the number of Odysseus' crew, the hero finally escaped his fangs. Safely he has landed now in Ithaca to take possession of his own. Here his spouse Penelope, that is, the local geneticist, is expecting him, Penelope who lived all these years bent over her spindle as it fits a cytologist and geneticist. And like the hero of old the geneticist Odysseus has strung his bow and arrows over his shoulder, the bow of constructive imagination and the arrows of analytical experiment, and as Homer says, they are clattering on his shoulders as he marches along. He is ready to shoot his arrows at the treacherous suitors of Penelope, at ignorance of every description. And we are confident that there will be good shooting these clays. But here the comparison with old Ithaca fails at one decisive point. When classic Odysseus returned home, he had to hide first in the hut of the divine swineherd Eumaeus, and he entered the palace disguised as a blind beggar. Modern Ithaca however has admitted us at once into its palatial grounds. If nevertheless some of us are feeling rather like blind beggars, it is because the splendor of this seat of learning has blinded us and because the greatness of your hospitality makes us feel like beggars. Let me then repeat the assurance of our heartfelt gratitude. W e shall try in our work of these days to prove worthy of the beautiful setting of this classic island of Ithaca.

INTERNATIONAL CONGRESS O F GENETICS

87

T H E R I S E OF GENETICS1 7'. H. Morgan, California Institrcte of Teclznology, Pasadeita, California

The new developments in science that occur from time to time can generally be traced either to the invention of a new method or to the discovery of a new fact that has far-reaching consequences, or to the elaboration o f a new theoretical principle that suggests new lines of investigation. In the latter case, it is the prerogative of science, in comparison with the speculative procedure of philosophy and metaphysics, to cherish those theories that can be given an experimental verification and to disregard the rest, not because they are wrong, but because they are useless. In the case of genetics the situation was in some respects different from any of these procedures; for it began with the discovery of a discovery that had been made 35 years before. W e can date the beginning of genetics, then, from the resurrection of MENDEL'Spaper in 1900. Its rehabilitation was not, however, due to a literary find, but to a need resulting from similar experiments by DE VRIES, CORRENS and TSCHERMAK that unveiled a series of phenonlena identical with the facts of MENDEL'Searlier work. The significant fact is that ;,hen the time was ripe to appreciate its fundamental significance, MENDEL'Sforgotten paper was discovered with the amazing result that hundreds of biologists, as the program of this present congress bears witness, had the direction of their scientific careers entirely redirected, or begun along new lines. The discoveries that rapidly followed, showing that the same laws applied widely to the other plants and to animals also, brought about realization that a great step forward in biology had been made. But before we consider the rise of genetics after the year 1900, it is proper on this occasion to pay tribute to the earlier work in hybridizing that furnished the background of procedure to which MENDELhimself probably owed a considerable debt. Let us pause for a moment and recall a bit of history, for it would be unfair to forget or to underrate everything prior to the first year of the present century. If to-day we express surprise that MENDEL'Spaper remained unnoticed so long, let us recall that this is not an unfamiliar experience in biological science. Between the experimental proof of sex in plants by CAMERARIUS (1694) and the prize essay of LINNAEUS on the sex of plants (1760) sixtysix years elapsed. A t about this time the scientific study of hybridizing may also be said 'Address of the president of the Sixth International Congress of Genetics at CORNELL UNIVERSITY, Ithaca, New York, August 25, 1932.

88

PROCEEDINGS O F THE SIXTH

to have been begun by LINNAEUS and his students, and especially by KOHLREUTER in several memorable papers (1760-1766). (1793) observations Then, thirty-three years elapsed before SPRENGEL'S on the natural cross-pollination of plants by insects, which made clear that such fertilization is of widespread occurrence in flowering plants. More interesting, perhaps, to modern geneticists are the pioneer experiments on peas that, in a very real sense, were the precursors of MENDEL'S work. I t is not as generally known as it should be that some of the facts on which MENDEL'Sresults with garden peas rested had been recorded by KNIGHT,42 years before several earlier experimenters. In 1823 THOMAS MENDEL,described a cross between a pea with a gray seed coat and one with a white that gave seeds which were uniformly gray coated. These seeds when grown produced in the next year both gray and white seeds. JOHN Goss had in 1822 also reported experiments with garden peas and found that the first generation of offspring had seeds like the paternal race. From these in the next generation he obtained peas of two kinds, one like those of the original grandpaternal race, the other like those of the grandmaternal. Separating these he found that the blue peas produced in F, only blues, and the white peas both blues and whites. Here is an example of what to-day we call dominance and recessiveness, as well as segregation in F,. I n the same year (1822) ALEXANDER SETON reported similar results. Nearly fifty years later THOMASLAXTON(1866-1872), working with peas, recorded numerous facts similar to those first spoken of, and in addition he mentioned cases in which two pairs of contrasted characters were present. Assortment between the pairs was found-which result is familiar to students to-day and which MENDEL'Swork established. Amongst the earlier hybridists the name of NAUDIN(1861-1864) is most often referred to as a forerunner of MENDEL,and it is sometimes stated that he anticipated MENDEL'Sdiscoveries. His principal prize paper appeared in 1863, two years prior to MENDEL'Spaper before the Briinn Society, and was followed by two others in 1864 and 1865. NAUDINlaid emphasis on the identity of individuals of the first generation hybrids, including reciprocal crosses. H e insisted on the intermediate character of the F1 hybrid, with the important reservation that the intermediate forms do not stand always equally distant from the two parents. W e now know that, taken character by character, sometimes an intermediate condition, sometimes complete dominance, may be found. But whichever condition holds for a particular character, the phenomenon of segregation in the germ cells of the F, hybrid remains unaffected. NAUDINstated explicitly that in the second and later generations there

I N T E R N A T I O N A L CONGRESS O F GENETICS

89

is a mixture of forms, including some which are like the original parents and others that approach these in various degrees. Then follows his most important deduction, namely, that the second generation results find their explanation in the disjunction of the two specific essences, derived from the parents, in the ovules and in the pollen of the hybrid. Here we have a highly significant contribution, for, not only did NAUDINsee clearly that the results are explicable on the principle of disjunction (or, as we say now, segregation), but that this, taking place both in the egg and in the pollen, gives the kinds of characters that appear. S o important historically is this fact that there should be included his specific statement showing that he had a perfectly clear idea as to how disjunction accounts for the diversity in the second generation. If, he says, an F, pollen grain bearing the characters of the male parent meets an egg of the same kind, a plant that is a reversion to the paternal species will result; similarly for the maternal species. But if a pollen grain of one kind meets an egg of the other kind, a true crossfertilization takes place like that of the first generation, and an intermediate form will result. I t will be agreed, I think, on all hands that this was a brilliant interpretation of results based on first-hand experience. I t falls short of MENDEL'Swork in two or three important aspects: (1) The failure to put the hypothesis to a test by backcrossing; (2) the failure to see what the numerical results should be on the basis of disjunction of the elements in the hybrid. His use of the words "disordered variation" in the F, and later generations brings out the essential difference between NAUDINand MENDEL. I t is the orderly result of disjunction or segregation that is the important feature of MENDEL'Swork; and finally, the clearness with which MENDELstated and proved the interrelation between character-pairs in inheritance, when more than one pair is involved, places his work distinctly above everything that had gone before. Nevertheless, the genial abbot's work was not entirely heaven born, but had a background of one hundred years of substantial progress that made it possible for his genius to develop to its full measure. If, in this brief review, I have neglected to bring in the names of a number of well-known selectionists whose work has been in the main in the field of agriculture, it is not because I do not realize the importance of their work or the great difficulties they overcame, but because, for the moment, we are interested especially in the development of our theoretical knowledge of genetics. So far I have spoken only of plants. What part, may be asked, has the study of animals played in the pre-Mendelian history of genetics, that is, down to 1865 ?

90

PKOCEEDINGS O F THE SIXTH

The question of sex in plants that took botanists a hundred years to decipher was not so difficult for zoologists. If we may accept the traditional story, it was not unknown in the Garden of Eden. ARISTOTLEhad a good deal to say about it. The credit of finding a sex-determining mechanism can properly be claimed by zoologists, but this happened only in the opening years of the present century. Hybridizing was also familiar to zoologists, but in pre-Mendelian times occupied only a relatively small part of their interest. What was known has been recorded by DARWINin his Ani~nnlsand plants under domestication. This scattered and loose information was incorporated after 1859 in the discussions of the theory of evolution. The chief contribution of zoologists to present-day genetics was along different lines. In the latter half of the last century there was great activity in the field of cellular morphology. The important facts concerning chromosome division and the extraordinary changes that take place a t the time of maturation of the germ cells and at fertilization were first made out by zoologists. The names of KOLLIKER,FLEMMING, FOL, V A N BENEDEN, HERTWIGand BOVERIare landmarks in the history of cytology, CorrespondDE BARYand ingly for plants the names of HOFMEISTEP,STRASBURGER, GUIGNARD run a parallel course. WEISMANN'Stheoretical contributions have also played an important historical r61e. The continuity of the gernz-plasnz served to counteract the and his successors, whose views if all-too-prevalent influence of LAMARCK correct would undermine all that MENDEL'Sprinciples have taught us. WEISMANN'S speculations on the origin of new variations by recombination of elements in the chromosomes, while not to-day acceptable as stated by him, nevertheless focused attention on an important subject. His discussion of the interpretation of the maturation divisions played, I believe, a leading r61e in directing attention to a subject that was destined very soon to have great importance for genetics. Thus at the end of the last century some extraordinary advances had been made in unraveling the changes that take place in the maturation of the germ cells. These advances led to the recognition of a mechanism that was to place the theoretical elements of MENDEL'Shypothesis on a firm foundation of fact. But this, however, was not apparent until 1903. GENETICS AT THE B E G I N N I N G O F T H E C E N T U R Y

W e come now to the fateful year 1900, when three lines of fundamental significance for genetics were ready to be brought together. I refer, of

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

91

course, to the mutation theory of DE VRIES,to the rediscovery of MENDEL'S paper, and to the application of the discoveries in cytology to the new theories. The intimate connection between the mutation theory, as first propounded, and the origin of the characters that follow MENDEL'Slaws was not immediately evident, since DE VRIES laid emphasis on the many character changes that result from each progressive mutational step. In fact, a t about this time DE VRIESrecognized three types of mutational changes: Progressive changes-changes that introduce something new, leading to the sudden appearance of a new elementary species; retrogressive changes, the result of something lost or becoming latent; and degressive changes, in which old characters are revived. This nomenclature, in so far as it is purely descriptive and based on characters rather than changes in the germ-plasm, covers broadly many of the facts with which we are familiar to-day. But in the light of the work of the last 30 years, especially when applied to genes, this description can no longer be accepted as fundamental; for now we have information that gives a more consistent picture of the changes produced by the genes. For example, the evidence from hybridizing elementary species, on which DE VRIES based in part his distinctions, has to-day a different interpretation. We: no longer hold that a progressive change introduces an entirely new, unpaired element into the germ track, for the unpaired chromosome ip cases of heteroploidy can surely not be regarded as the usual step for progressive evolution. Again, the permanence of certain hybrid combinations, whenever such exceptional cases arise, are not now regarded as due t o the introduction from each parent of a new unpaired element, but can be interpreted in different ways in different cases. I t was the emphasis that DE VRIES laid on mutational changes in the germinal material as sharply discontinuous, irrespective of the effect on the character, that has had important and far-reaching conseqtlences for genetic work and theory. The groundwork for discontinuous phenotypic variation had in 1894 been laid by BATESON'S contribution on discontinuous variation. While we recognize that some of the examples BATESON collected are not inherited but are phenotypic (which confused the picture), nevertheless his insistence on the importance of discontinuity prepared the way for the acceptance of the more fundamental distinction that DE VRIESmade later. But I wish to emphasize that the revolution in our ideas that took place a t this time was not so much due to the insistence on discontinuity of somatic

92

PROCEEDINGS O F THE SIXTH

structures, but discontinuity in the hereditary elements. A n example will serve to illustrate the difference. When a gene changes, its effects on new characters, taken individually, are generally very different. Some of them may be sharply marked off from the original character. The character showing the greatest effect is the one generally picked out for genetic work. But a t the same time there are changes in other organs that are less conspicuoussome of the characters are so little affected or so variable that, taken by themselves, they would give a picture of continuity rather than of discontinuity. They would often pass unnoticed were not attention drawn to them by the discovery of the major change. For the theory of evolution some of these inconspicuous changes may be more significant than the more obvious discontinuous change. In fact, if evolutionary advances are more often through invisible physiological mutational changes rather than morphological ones, we can better understand the paradoxical situation in which taxonomists find themselves, to wit, that the sharp structural differences, that are used for diagnostic separation of species, relate to characters that seem often to be unimportant for the wellbeing of the individual. The new point of view is a complete reversion of much of the thinking in which the evolutionary theory indulged in the past. As I have said, the rapid expansion of genetics after 1900 has been intimately connected with the applications of the chromosome theory to the experimental work in genetics. The integrity of the chromosomes and their continuity from one cell generation to the next, the constancy in number of the chromosomes in each species and the absence of mixing of the materials of the conjugating chromosomes at the time of meiosis have furnished the basis on which genetics rests. I think we can not overemphasize the significance of this relation between the theoretical side of genetics and the factual side as observed in the known behavior of the material basis of heredity. T o put the matter bluntly, the recognition that there is a mechanism to which genetic theory must conform, if it is to be productive, serves to keep us on the right track and acts as a check to irresponsible speculation, however attractive it may seem in print. Some one may reply that it is not always an advantage to keep one's nose to the grindstone. Granted! but realizing how often ingenious speculation in the complex biological world has led nowhere and how often the real advances in biology as well as in chemistry, physics and astronomy have kept within the bounds of mechanistic interpretation, we geneticists should rejoice, even with our noses on the grindstone (which means both eyes on the oculars), that we have at command an additional means of testing whatever original ideas pop into our heads.

INTERNATIONAL CONGRESS O F GENETICS EXPANSION SINCE

93

1900

I come now to the expansion of the Mendelian theory that has taken place in the last 30 years. If I refrain from giving the names of the numerous contributors to this advance, it is because many of the discoverers are before me in person; or, if not, will get reports of the congress. Future congresses will probably be better able to evaluate individually the merits of those who have made the significant contributions in this generation. I t must have been evident to many geneticists after 1903 that if the chromosomes are the bearers of MENDEL'Selements, there would be only as many independently inherited characters as there are chromosomes, if the then current idea of the integrity of the chromosomes were true. This would place limitations on MENDEL'Ssecond law-the law of independent assortment. In fact, the genetic evidence can now be said to have firmly established that owing to interchange between linkage groups there are more characters inherited than there are chromosomes. Thus linkage turned out to have its limitations, and it was these very limitations that made it possible to determine the location of the genes in the chromosomes. I refer, of course, to crossing over. Since localization of the genes is to-day the basis of much of the quantitative work in genetics, I may be allowed to elaborate the theory. The outstanding genetic fact is that these interchanges take place only between homologous chromosomes-that is, between members of the same pair. The second important genetic fact is that when the interchange takes place, large blocks of the chromosomes are exchanged. This can be proved only in cases where more than two loci are involved, and best when a considerable number of well-spaced genes have been located. Until recently the evidence that large blocks of genes are involved in crossing over was known only genetically. N o certain cytological proof was known. To-day, however, the proof has been found. Without doubt this cytological evidence will be presented and discussed at this congress. I t has also been determined on genetic evidence that more than one interchange may take place between a pair of chromosomes ; this can be checked only in cases where there are enough intermediate loci between two pairs to serve as markers. A moment ago I said that crossing over has furnished the basis for the theory of localization. May I give an illustration, in the hope of removing a criticism of the localization technique that is based, I believe, on a misunderstanding? It has been said, for example, that the changes made from time to time in the genetic map of the Drosophila chromosomes discredit the method

91

PROCEEDINGS OF THE SIXTH

by which the localization is determined. I t might as well be said that the method by which the atomic weights in chemistry were gradually improved discredits the procedure of the chemist. Two illustrations will serve our present purpose. Let us suppose a new mutant character is found and its chromosome group--that is, its linkage group--determined by familiar methods. W e may ~roceed,then, to find its relation to two known loci in that chromosome. If these are far apart, the crossover data will give only its approximate position. Having found this, it may turn out that the locus lies near another gene in that region, but whether above or below may be uncertain. W e next proceed to find its more exact location with respect to this third gene, using either of the other two genes as a second point. In this way the new gene is more accurately placed with respect to the third locus. Further work will then be necessary if there are other genes in this region. The second illustration concerns a distinction between crossing over data given in the actual experiment and their conversion into map distance. For very small values, say 5 points, the two are the same because double crossing over is not present. But in longer distances the crossing over data may depart widely from the map figures because double crossing over makes the figures too low. In Drosophila the sex chromoson~eis 70 units of map distance, but for long distances the crossing over data are found to give not over 50 units. In this case. the map distance has been built up piece by piece through the summation of crossing over data of loci so near each other that double crossing over is eliminated. In other animals and plants, where few loci have as yet been found, the incomplete data are generally put down as map distance. This may be f a r from the real map distance, and since the actual amount of double crossing over in such less-worked-out forms is unknown, and since crossing over is different in different species, the loci must be regarded as only tentative. There is another factor to be taken into account. The theory of localization was based in a general way on the assumption that crossing over in one region of a chron~osomehas the same frequency as in other regions. The Drosophila workers have long known that this is not exact, and in fact they had invented methods to show that crossing over is different in different regions. The crowding of the genes in some regions of the genetic map and their scarcity in other regions has been shown to be due to the different frequencies of crossing over per unit of absolute distance in the cytological chron~osome.This seems to be a fundamental relation for all chromosonles. In the X chromosome of Drosophila, which appears to be a special case, most

INTERNATIONAL CONGRESS O F GENETICS

95

of the genes are crowded at the two ends of the chron~oson~e with a middle region of undetermined length having few or no genes, in the sense that the Y chromosome is empty of genes. These facts do not invalidate the purpose for which the maps were invented, since the relative position of the genes remains the same. I t is their position relative to each other, allowing very precise prediction of the topographical relation of a new gene to all other known genes, as determined by corrected crossover data, that is important. This brings us to one of the most recent fields of modern genetics-the study of the redistribution of the linkage group by translocation. Treatment with X-rays has been found to be a prolific source for material of this kind, but it should not be forgotten that translocation had been discovered and utilized for genetic interpretation several years before X-rays were used. Even to-day, with much evidence before us, the way in which X-rays bring about this result puzzles us. In a crude way we might picture the electron shooting holes in the chromosomes, thus breaking them apart. But when the relative sizes of the electron and the chromosome are considered, it is difficult to see how such a disruption would result from a single shot. Even more surprising is the fact that the broken end of a piece may reunite with the end of some other chron~osomeand, acquiring thereby an attachment fiber, form a new linkage group. Of course it does not follow that such a reunion occurs whenever a chromosome is broken. I t is only those cases where reunion does occur that are recovered and studied by geneticists. When no such union is brought about the piece, lacking an attachment point, will be lost, and the zygote without it will probably die. As I have said, the astonishing fact remains that the broken end becomes a t times attached to the end of another chromosome. Without the objective evidence of this union that we have to-day, it might have been supposed that the broken-off fragment would rather have made, or retained, a side-to-side union with a corresponding part of its homologous chromosome. Haowever, since the conditions of the cell that permit conjugation of like chron~oson~es occur only once in the life cycle, such a union is not, then, to be expected if the breaking has occurred after that event. If it had occurred earlier in the germ track the piece would no doubt have been lost before meiosis came on. Here, then, we have a field inviting speculation. Let us hope that it will not long remain there, but that evidence concerning these puzzling relations may soon be forthcoming. In this connection I need hardly recall to mind that, on the current theory of crossing over, the linear order of the genes is broken a t the same level in two of the strands, and a new lengthwise reunion of the broken ends takes place. Whether this breaking and reunion is a comparable process to that

96

PROCEEDINGS O F T H E S I X T H

seen in translocation we do not know, and it would be unprofitable at present even to make a guess. POLY PLOIDY

In even a passing review of present-day genetics, the numerous problems connected with the increase in number of the chromosomes, or polyploidy in technical language, can not be ignored. But how can one hope even to summarize the work that is pouring in with the arrival of every new number of the genetic journals? The importance of polyploidy for the evolution doctrine is perhaps clear, but needs cautious handling in the light of the past history of phylogenetic interpretation of the facts of comparative anatomy. I hope that that history, a t least, will not be repeated when the story of k the exchange genetics comes to be written, for, in the light of recent w ~ r on of limbs between non-homologous chromosomes, and on translocations, the comparison of chromosome numbers without this knowledge may be very misleading. The determination of the linkages of the genes is the only safe basis for such comparisons. A t present I can do no more than briefly indicate some of the obvious and salient points. In many families of plants, and also in a more limited number of animals, chromosome groups are present that are multiples of a basal number, usually of the haploid number of the lowest member of the group. These are frequently double or triple, or quadruple groups of a basal number, generally assumed to be the haploid number. A good many of our cultivated plants are also known to show multiples of a real or postulated basal number of chromosomes. It is natural to assume that, in many cases, this has come about by the actual doubling of the whole chromosome group rather than by the breaking of the chromosomes, that would also lead to doubling their number. I t is more consistent to assume that doubling is the method by which the number of chromosomes is increased, because of the evidence from the sizes of the chromosomes, from their method of conjugation and from the relation of chromosomes t o the attachment fiber. There are several known ways in which we can bring about a doubling of the number of chromosomes in a cell. The usual way is to suppress the cytoplasmic division of the cell at the time when the chromosomes divide. When this is done the chromosomes do not reunite, but the descendants of that cell will forever possess twice the original number of chromosomes. Theoretically, the process might go on forever, unless there are upper limits of a physiological nature preventing an indefinite increase. Doubling of diploids gives tetraploids. These crossed to diploids give triploids. Double tetraploids (or octoploids) crossed to tetraploids give hexaploids, and so on.

INTERNATIONAL CONGRESS OF GENETICS

97

This work furnishes an opportunity for the solution of certain genetic problems of theoretical interest, for, without this knowledge, some of the known genetic ratios would have been difficult to interpret. With this knowledge they are found to conform to recognizable general principles. I t is perhaps ungracious to point out that the mere study of chromosome numbers in different species may in itself become mere hack work. I t looks as though it may become as popular for academic work as section-cutting of embryos was at an earlier period. It is more generous, perhaps, to regard the work on chromosome counts as pioneering, and therefore preliminary work in the search for new materials, some of which will certainly be of value for deeper-lying genetic problems. This is especially evident in the study of hybrids whose parents, whether cultivated or wild types, have different numbers of chromosomes. The erratic behavior of the chromosomes, often seen in the maturation of the germ cells of such hybrids, clearly explains the exceptional and often abnormal results that follow. Without this information we might be tempted to indulge in much profitless and arbitrary speculation. Not only are we familiar with cases where a multiplication of the same group of chromosomes is brought about within the species, but there are a few cases where an increase has been brought about by crossing distinct species with different numbers of chromosomes, and chromosomes that do not mate a t meiosis. These situations are full of interest for students of genetics, presenting a wide range of new possibilities. Of great importance for the genetic interpretation of polyploidy in terms of chromosomes is the identification of chromosomes that carry specific genes. Only a few years ago this was known in only one animal, but the number of cases is steadily increasing. Until information of this kind becomes more general there will be, as at present, a good deal of guessing as to the interrelation of chromosome groups having different numbers of chromosomes. INFLUENCE OF THE GENES ON THE CYTOPLASM

If another branch of zoology that was actively cultivated a t the end of the last century had realized its ambitions, it might have been possible today to bridge the gap between gene and character, but despite its highsounding name of Entwicklungsrnechanik nothing that was really quantitative or mechanistic was forthcoming. Instead, philosophical platitudes were invoked rather than experimentally determined factors. Then, too, experimental embryology ran for a while after false gods that landed it finally in a maze of metaphysical subtleties. It is unfortunate, therefore, that from

98

PROCEEDINGS O F THE SIXTH

this source we can not add, to the three contributory lines of research which led to the rise of genetics, a fourth and greatly needed contribution to bridge an unfortunate gap. I say this with much regret, for, during that time and even now, I have not lost interest in this fascinating field of embryological experimentation. It is true that a great deal of factual evidence came to light, and it is true that many misleading ideas were set aside, but the upshot was negative so far as the formulation of any of the factors of development, whether mechanistic or otherwise, are concerned. This may be because the work was pioneer and largely qualitative. Perhaps my disappointment a t the outcome of the work has led me to an overstatement of its failures. Something did emerge that the future may show to be of fundamental importance for genetics. I mean the experimental demonstration that the immediate factors in the differentiation of the embryo are, at the time of their activity, already in the cytoplasm of the cell. Second only in interest was the discovery that, within certain limitations, the already determined specificity may be reversed, or rather, shall I say, the initial steps already taken are reversible by factors extraneous to the individual cells. These statements call for further elaboration, because they are unconsciously in the background of much of our thinking about genetic problems, and should if possible be more sharply formulated. That the form of cleavage of the egg is determined by the kind of chromosomes it contained before the egg reached maturity has been sufficiently proved; and since the foundations of all later differentiation are laid down a t this time, the den~onstrationis of first-rate importance for genetics, because it shows that we are not obliged to suppose the genes or chromosomes are functioning only a t the moment of the visible appearance of characters. This is demonstrated by introducing into the egg foreign sperm of a species having another type of development. Although the chromosomes from the sperm are present from the first cleavage onward, they produce at first no effect on the cleavage; only after a time do they succeed in bringing about changes in the embryo. This evidence is, as I have said, important for our genetic analysis, for it serves as a warning that the time relations between gene and cytoplasm may have a relation different from that of an immediate dynamic change in the cytoplasm. The preparation for the effect may have taken place long before the actual event. The second inference is no less significant. I need not labor the point a t this late date that the characters of the inilividual are the product both of its genetic make-up and its environment. The earlier, premature idea, that for each character there is a specific gene-the so-called unit-character-was

INTERNATIONAL CONGRESS O F GENETICS

99

never a cardinal doctrine of genetics, although some of the earlier popularizers of the new theory were certainly guilty of giving this impression. The opposite extreme statement, namely, that every character is the product of all the genes, may also have its limitations, but is undoubtedly more nearly in accord with our conception of the relation of genes and characters. A more accurate statement would be that the gene acts as a differential, turning the balance in a given direction, affecting certain characters more conspicuously than others. But let us not forget that the environment may also act as a differential, intensifying or diminishing, as the case may be, the action of the genes. The best illustration of this double relation is seen in the determination of sex. When an unpaired chromosome is present, in one or in the other sex, its genes determine, as a rule, whether a male or a female develops from each egg. Under environmental conditions which, as we say, are normal, the differential acts almost perfectly; but under other unusual conditions and in a few special cases its power may be partially overcome and even a reversal may take place. These unusual environmental conditions may be external agents, such as temperature or light. They may also be internal factors, such as hormones. Even "age" itself may bring about a reversal of sex in certain types. These statements are commonplaces to-day. The only differences of opinion concern the emphasis that one theorist places on the environment, and another on the genic composition. In passing, a word may be said about the genes as sex factors or differentials. All through the 32 years of the present century there have been attempts to isolate (in a genetic sense) the sex-determining factors. A t first, when the chromosome mechanism was discovered, the idea prevailed that one X, let us say, made a male, a i ~ dtwo X's a female. The sex chromosome itself was then taken as the differential. Very soon after this the idea that the sex chromosome was the carrier of a gene for sex suggested itself, and a search was started t o locate such a gene or genes in this chromosome. More recent work on translocations has shown the probable futility of such an interpretation. The tendency at present is rather to look upon all the genes, or at least on many of them, as sex determining in exactly the same sense, as all or many of the genes have an effect on the development of each character. It may well be, however, that certain genes in the sex chromosome (as in other chromosomes) are more influential than others in turning the balance one way or the other, but even so, it does not at the present moment-in the light of recent evidence-seem probable that a single gene for sex determination is to be found in the X chromosome any more than, in the

100

PROCEEDINGS O F THE SIXTH

contrary sense, there is a single gene for sex in any special autosome. Here again, some one or a few genes may be more influential than others, but this is also true to varying degrees of the gene for any other character. The theory of balance between the intracellular products of the genes is the most direct contribution to physiology that modern genetics has made. I t is an idea familiar to classical physiology as applied to organ systems, but a distinctly new contribution to cellular physiology. I t may be a long time before these intracellular genic substances are isolated and purified (since there may be many steps between the actual primary substances and the endproduct of such substances in the cell plasm) ; nevertheless as a point of view the presence of genic materials rather than a dynamic action of the nucleus is supported by some analytical evidence. Already there is afoot in several quarters, and by methods partly genetic, partly physiological, partly embryological, partly physical and chemical, a decided effort to approach this problem. I f we could obtain these substances in pure condition we might then be in a position to speak more confidently of a quantitative study of gene activities in the sense that chemistry is quantitative. Meanwhile there are other more practical methods by which we may construct provisional hypotheses as to the nature of the intracellular substances that are the products of the genes, namely, through a study of triploids, trisomic types, fragments of chromosomes and by analysis of crosses between different species. This statement does not, of course, exclude the possibility of the discovery of entirely new methods of approach. Let us not forget that the idea of balance, as seen in the character, is really an old and familiar one to geneticists. For example, the intermediate character of the F1 hybrid was generally interpreted as due to a conflict between the old and the new gene. Again, the familiar statement that characters are often affected by modifying genic action that enhances or diminishes the effect of the primary gene, is another example of the intracellular balance of the activity of the genes. What has been said so far relates to the action of the gene on the cytoplasm of its own cell-its intracellular action. Those of us working with insects or plants are apt to think of genetic problems in this way, and are inclined to consider mainly the effects that do not reach beyond the cells in which they are produced. But in other groups, especially birds and mammals, the effects of the genes are not always so limited. W e are on the threshold of work concerned with the isolation of the so-called sex-hormones, the endproducts of the thyroid gland, the pituitary, the thymus, and the substances

INTERNATIONAL CONGRESS O F GENETICS

101

isolated from the suprarenal bodies. All these substances produce their effects outside of the cells that manufacture them. In themselves they are far removed from the primary action of the genes. In this connection certain work carried out by experimental embryologists should not be overlooked, beginning with the early experiments of LEWIS ill 1904 and culminating in the more recent work of SPEMANN. Here it appears as the result of grafting experiments that certain organs of the body develop in response to the vicinity of other organs, as when, for example, the lens of the eye of the frog is shown to be a response to the presence beneath the skin of the optic lobe. Similar and more far-reaching effects have been recently found for other organs of the embryo. The simplest interpretation, perhaps, is the setting free of a hormone by an embryonic organ or group of cells that calls forth a response in neighboring regions. This and other evidence goes to show that gene activity may produce results outside of the cells in which the first steps are initiated. The problem a t present is one of immediate importance in the study of gynandromorphs, 'mosaics and intersexes. EVOLUTION

Sooner or later every geneticist is asked what bearing this work has on the theory of evolution. In the early years of the century when genetics was new, some of us tried to sidestep the question, partly on the grounds that genetics was not ready to discuss the bearing of the new work on evolution, but mainly because it seemed unfortunate to compromise the precise results of the new procedure with those of the evolution doctrine which, because it dealt with a historical problem, was largely speculative. After 32 years of activity, caution may still be the wiser course to pursue; yet, on the other hand, we are now prepared, I think, to make a more definite commitment. I t is, of course, obvious that only those characteristics that are inherited can take part in the process of evolution. The only characters that we know to be inherited are those that arise first as mutants, that is, discontinuously, or, as we say, by a change in a gene. Here genetics has made a very important contribution to evolution, especially when it is recalled that it has brought to the subject an exact scientific method of procedure. If we compare our present status in this respect with the discussions of the old school of evolutionists concerning variability, there can be no question but that genetics has contributed valuable information. In the second place, the objection has been not infrequently made that geneticists are dealing only with aberrant or abnormal characters-hence their results, however accurate, can have nothing to do with the kind of

102

PROCEEI3INGS O F THE SIXTH

progressive changes that have made evolution of new types possible. Such objections have come largely froin those who ignore what geneticists have done and are doing. The same objections have also come from those whose minds are closed to new evidence, or who can not distinguish between the value of tested and verifiable theories and vague views or juvenile impressions with a teleological background or bias. Without elaborating, I wish to point out briefly that there is to-day abundant evidence showing that certain differences, distinguishing the characteristics of one wild-type or variety from others, follow the same laws of heredity as do the so-called aberrant types studied by geneticists. Even this evidence may not satisfy the members of the old school because, they may still say, all the characters that follow MENDEL'Slaws, even those found in wild species, are still not the kind that have contributed to evolution. They may claim that evolutionary characters are in a class by themselves, and not amenable to Mendelian laws. If they take this attitude, we can only reply that here we part company, since c x cathccEra statements are not arguments, and an appeal to mysticism is outside of science. There remains still the question of the causal origin of mutations. Here also some progress has been made, but the subject is admittedly by no means or1 the same footing as is our knowledge of the laws of inheritance. I t behooves us, then, to be careful, for our progress in this respect has been slow and to some extent erratic. I mean by this that we have not yet found a method of producing specific results-that is, a method by which particular genes can be changed in a particular way. Even here, however, something has been done. In the work with X-rays and heat the same mutants appear that are already known, and that have come up without treatment. In addition, new mutants appear, as they do also without treatment. If it can be shown on a large scale that the same ratio for known mutations holds for X-ray and for spontaneous mutations, we may have found an opening for the further study of the causes of certain types of mutation. I have been challenged recently to state on this occasion what seemed to be the most important problems for genetics in the immediate future. I have decided to try, although I realize only too well that my own selection may only serve to show to future generations how blind we are (or I have been, at least) to the significant events of our own time. First, then, the physical and physiological processes involved in the growth of genes and their duplication (or as we say their "division") are obviously phenomena on which the whole process of reproduction rests. The ability

INTERNATIONAL CONGRESS O F GENETICS

103

of the new genes to retain the property of duplication is the background of all genetic theory. Whether the solution will come from a frontal attack by cytologists, geneticists and chemists, or by flank movements, is difficult to predict. Second: An interpretation in physical terms of the changes that take place during and after the conjugation of the chromosomes. This includes several separate but interdependent phenomena-the elongation of the threads, their union in pairs, crossing over, and the separation of the four strands. Here is a problem on the biological level, as we say, whose solution may be anticipated only by a combined attack of geneticists and cytologists. Third: The relation of genes to characters. This is the explicit realization of the implicit power of the genes, and includes the physiological action of the gene on the rest of the cell. This is the gap in our knowledge to which I have referred already a t some length. I may say the Fourth: The nature of the mutation process-perhaps chemico-physical changes involved when a gene changes to a new one. Emergent evolution, if you like, but as a scientific problem, not one of metaphysics. Fifth: The application of genetics to horticulture and to animal husbandry, especia1,ly in two essential respects; more intensive work on the physiological, rather than the morphological, aspects of inheritance ; and the incorporation of genes from wild varieties and species into strains of domesticated types. Should you ask me how these discoveries are to be made, I should become vague and resort to generalities. I should then say: By industry, trusting to luck for new openings. By the intelligent use of working hypotheses (by this I mean a readiness to reject any such hypotheses unless critical evidence can be found for their support). By a search for favorable material, which is often more important than plodding along the well-trodden path hoping that something a little different may be found. And lastly, by not holding genetics congresses too often.

PROCEEDINGS OF THE SIXTH

T H E S P E C I E S PROBLEM I N D A T U R A Albert F . Blakeslee, Carnegie Institution o f Washington, Cold Spring Harbor, New York

The discussion of our topic will be largely an explanation of the Datura exhibit. The exhibit, however, has an advantage in that the exhibitors who carried on the work are able to explain their own findings in person. Our present summary is based on the results of a number of collaborators. A. D. BERGNER, with S . SATINA, is in charge of the work in cytology and J. L. CARTLEDGE of the work in ~ o l l e nabortion; J. T. BUCHHOLZ is leader in studies on internal in the studies on pollen-tube behavior, E. W. SINNOTT has charge of the care of the plants and the tabuanatomy and A. G. AVERY early demonstrated that chromolation of the segregating cultures. BELLING sonles can be matched up like dominoes and worked out the chromosomal basis of the "B" races. It is easy to play with dominoes after we have been taught the rules of the game. The species problem in Datura has had an evolution. A t first we attempted to analyze the differences between species by means of hybridization, selfing, and isolation of types through continued inbreeding, without knowing much about the genetic elements with which we were dealing. Our present program attempts to relate differences between species in terms of established standards. That so f a r our work has been more with our standards than with different species may not be a. disadvantage to our ultimate objective. In triangulation, the surveyor finds it practical to first know accurately his base lines. T o the geneticist pure living reagents of known composition are as iillportant as are pure chemical reagents to the chemist. The investigations in Datura may be logically classified under four main headings, although these divisions do not represent the actual order o f research nor of our discussion. We have studied: ( 1 ) differences within a standard line or biotype, our line 1; ( 2 ) intraspecific differences within a single standard species, Datura stramonium; ( 3 ) interspecific differences between 7 of the species within the genus Datura; (4) synthesized or laboratory new "species." DIFFERENCES WITHIN T H E STANDARD L I N E

Our standard line 1 of D. stramoniunz has been inbred by selfing for 19 generations and has been once passed through a haploid. Each season we grow a considerable number of plants of this standard race and scrutinize them closely for possible morphological variations. In addition the pollen of several hundred line 1 plants is examined microscopically each year for ab-

INTERNATIONAL CONGRESS O F GENETICS

105

norn~alitiesthat might not show in external appearance. Very few variants have been discovered during our experience with this line that could be attributed to gene mutations. Chromosomal mutations, however, have been not infrequent. Out of 13,345 line 1 plants grown in our breeding plots from 2n parents, 22 or 0.17 percent have been identified as 2n+ 1 chromosomal abnormalities. Since there are in Datura 12 different kinds of chromosomes each with two different ends, there are 24 different kinds of ends. In line 1 normal plants the ends of the chromosomes have been given consecutive numbers beginning with the odd number. The largest chromosome, therefore, is labeled [ l . 21 and the smallest [23 . 241. They may be represented diagrammatically as dominoes with numbered ends. In assigning consecutive numbers to the two ends of the line 1 chromosomes we do not wish to imply that this race is the most primitive from the evolutionary standpoint. The numbering merely emphasizes again that we are interpreting Daturas in terms of this single standard. As shown in table 1, the 2n+ 1 types have been classified as primary types with the extra chromosome unmodified and like the other two in the set, as secondaries with the extra chromosome composed of one half doubled, as tertiaries with the extra chromosome composed of parts of two normal chromosomes, and as fragment types with only a part of a chromosome extra. These types have been figured and discussed elsewhere (BLAKESLEE 1927, 1929, 1931, BLAKESLEE and BELLING1924) and one of the primaries with its two secondaries is shown in figure 3, so we need not dwell on their peculiarities. Capsules are convenient- parts with which to illustrate their diversity in external appearance, but probably all parts of the plant are more or less strongly affected by the presence of a particular extra chromosome. SINNOTT,for example, has found marked differences in their internal anatomy, CARTLEDGE in pollen grains, and BUCHHOLZ in the behavior of their pollen tubes as can be seen in their respective exhibits. Table 1 shows types with extra chromosomes, all of which are in line 1 and have appeared spontaneously more than once with the exception of the [2 91 which has been found but a single time. Since they all tertiary 2n belong to the same highly purified standard line, we can assume with a fair degree of assurance that the peculiarities of these types are due entirely to the extra chromosome. W e have in these types, therefore, a means of learning something about the assemblage of factors in normal whole chromosomes and in major parts of chromosomes, at least of such factors as induce recognizable effects when present in an extra dose. This is especially true of

+

-

PROCEEDINGS OF T H E SIXTH

106

+

Primary, secondary, and tertiary Zn 1 tyfies ilt standard line I . (Extra chromosomesare shown with numbered ends) CEROMOBOUE 81EE CLAW

COND DART CaROYOBOUEB

PRIMART CEROMOSOYBB

BECONDARI CEROMOSOXBB

TEBTIART CEROMOBOUES

OXNEB LOCAmD IN PARTICULAR CaROYO80YB

Bz in . 4 ; Qs and pa, in 3 . 4 , half not determined

fw,

to in .16; $1, in 15.16, half not yet determined; rj, in

15.16or7.8

The ends l o , 8O, lo0, 12O, 16O, 20°, 21' and 24' are characterized by terminal humps.

primaries which may be secured by segregation from 3n parents where no break has been necessary in their formation as is the case for secondaries or tertiaries. These 12 chromosomes of line 1, together with parts of some of them, are our genetic standards for the genus Datura. They are not as precise standards as we might wish since they are obviously blocks of many

INTERNATIONAL CONGRESS OF GENETICS

107

interacting units, but they have the advantage we believe of representing more surely the normal chromosomal constitution than standards made up of allelomorphs inferred from single gene mutations. We have in our standard race a rough genetic scale or balance with which we can weigh any chromosome that can be added to this race in terms of the unbalance which it exerts over the standard 2n chromosomes. How individual chromosomes of other species may be isolated and weighed as extras in our standard line 1 balance will be discussed later. The tertiary forms that have arisen spontaneously in line 1 show that chromosomes may be formed with different end arrangements from those of our standard line 1 chromosomes which are used as the ultimate testers for ends of chromosomes. The standard chromosomes are separable into different size classes and in the majority a terminal hump distinguishes one end from the other. A humped chromosome is capable of identifying the end of a tertiary chromosome to which it is attached at reduction division. The double half chromosomes in secondary 2n+ 1 types are of especial value as chromosomal testers since both ends are identical. It will be noted from table 1 that we have secondaries for all the primaries except for the smallest two chromosomes, [21 221 and [23 241. The [21 . 221 chromosome and the [23 241 chromosome are marked by a terminal hump. For both these two chromosomes other testers for their ends are also available in good tertiaries not listed in the table since they did not arise spontaneously within line 1. A consideration of other line 1 types will be more profitable after a discussion of intraspecific differences.

.

.

INTRASPECIFIC DIFFERENCES

It was early seen that in D. stramonium there were cryptic races in nature in respect to the breeding behavior. Thus only the 2n [17 181 primary gave trisomic ratios when heterozygous for "A" whites whereas both the primaries 2n + [ l 21 and 2n + [17 181 threw abnormal ratios when heterozygous for "B" whites. By means of ratios from a more convenient type, "Nubbin," a considerable number of our white and purple races from nature were classified into A's and B's. BELLING(BELLING AND BLAKESLEE 1926) showed that the basis for this classification was segmental interchange in the origin of the "B" race such that instead of the [l 21 and the 117 181 chromosomes of line 1, which was an "A" race, we had in the "B" race the chromosomes [l 181 and [2 171. These two chromosomes we had already had as extras in 2114- 1 tertiary types which appeared to be related to both the [ I . 21 and the [17 . 181 primary 2x14- 1 forms.

+

.

.

.

PI
108

Instead of using breeding tests for the identification of "A" and "B" races, we studied the configurations in the F, generations between line 1 and various races from nature. As might have been expected "B" races induced a large circle of 4 when combined with line 1 chromosomes having the formula 1 - 22.17

I

I

1 . 18 - 18 . 17

1929) has BERGNER with the help of SATINAand others (BLAKESLEE studied about 550 races from nature by means of the configurations which they induce in F, generations with our standard line 1. As seen by the table, only 5 cryptic chromosomal types have surely been discovered in nature in D. stramonium. Three of them differ from line 1 by interchange of segments of non-homologous chromosomes and one apparently by interchange of humps. A rather wide range of countries is represented and it seems unlikely therefore that a further search would greatly increase this number. Types such as these which are homozygous for modified chromosomes are called prime types. Each of the prime types from nature has been repeatedly backcrossed to line 1, chiefly through compensating types, until each resembles line 1 in appearance and presun~ablyhas all line 1 chromosomes except the two modified by segmental interchange. Presumably they now have practically the same assortment of genes since they are alike in appearance, but the grouping of the genes is different due to the interchanged chromosomes. Since the chromosome peculiarities are not obvious they have been called "cryptic types." Modified chromosomes from each prime type have been added to line 1 to form 2n+ 1 tertiary forms. The 2n + [l . 181 and the 2n 12 . 171 types, f o r example, shown in the exhibit and on the screen (BLAKESLEE 1927, plate 10) are to be classed as tertiary forms in terms of line 1. They are primary 2nf 1 types so far as the "B" race is concerned and have both been secured in the offspring of "B" triploids. The tertiary forms shown in table 1 have arisen spontaneously in our cultures. Since they appear to have been formed through segmental interchange they raise the question whether segmental interchange leading to homozygous prime types is not frequent in our material of D. stramonium. This seems not to be the case for several reasons. The spontaneous occurrence of these tertiaries is relatively rare. One, 2n [13 181, arose twice from a 2n + [17 . 171 parent heterozygous for a articular gene which may have been responsible in some way for its appearance. Of the remaining two, the 2n + [4 61 occurred twice and the 2n [2 . 91 occurred once. These three

+

+

+

-

INTERNATIONAL CONGRESS O F GENETICS

109

tertiaries are not related to the tertiaries in the relatively few prime types in nature. Furthermore a test by BERGNER of 22 sublines of line 1 derived from continued selfing of the various 2n+ 1 types failed to disclose any which induced abnormal configurations in F1 generations with a single line 1 plant. W e can conclude that neither in our cultures nor in nature has segmental interchange leading to prime types been a frequent phenomenon in L l . strantoniu~~z. Fragmentations, translocations and segmental interchanges as well as gene mutations have been common in this species, however, following radiation treatment. Together with the 5 prime types from nature and the results of radiation we now have isolated in the homozygous condition a total of over 75 prime types. Twelve of these are shown in the accompanying table (table 2 ) which is part of a larger table that is published elsewhere (BERGNER, SATINA,and BLAKESLEE 1933). The ends of the modified chromosomes in about 40 prime types have been identified by 'attachments with tester chromoson~esat meiosis. In our collection we have at least one homozygous prime type tester for the ends of each of the 12 chron~osomesexcept the [5 . 61 chromosome and for this chromosome we have a prime type in the heterozygous condition which we are attempting to render homozygous. Possibly among those with an 1 chromosome involved, which have not yet had their ends determined, we may already have isolated a homozygous prime type for this last chromosome. (Since the paper was read the prime type involving the [5 . 61 chromosome has been rendered homozygous.) This collection of prime types is being found of value as a source of chromosomal testers for various purposes and as material for the synthesis of compensating types and pure-breeding morphological types or synthetic L species." It would take us too far to recount all the uses of prime types. I t should be mentioned, however, that prime type testers have replaced 2n-t 1 forms in the identification of the ends of chromosomes from other prime types and from other species, by means of chromosome attachments. Prime types contain tertiary chromosomes and therefore are a source of compensating types (BLAKESLEE 1931) which we are beginning to use in interspecific studies. "Nubbin" (BLAKESLEE 1927), which was secured from radiation treatment in 1921, may be used as an illustration of a compensating type. Its formula showing chromosome attachments may be written as follows: 6

r

T

r

f

f ~b gamete

[9'10]-[10.9]-[9'1]-[1'2]-[2'5]-[5'6]-[6'5]

I

1

5

1 In gamete

PROCEEDINGS OF THE SIXTH

110

TABLE2 Prime types (races homozygous for modified chromosomes).

DIManATIon

Ootmamnon ~nPI WITH BT*mBD

p~mm POLLEN ABOBTIOR

mo~oWMAL

0I.m CIA-0

LIND

1

CEBOY*

YODIIIED CEBOYMMMZ0

WYE0 INVOLVED

oBIaIn

LInEl

la Fl

PT 1

12 Bivalents

OK

PT 2

04

OK

L= 1 . 2 ma17.18

1.18 2.17

Nature

PT 3

04

125

M=11.12 S=21.22

11.21 12.22

Nature

PT 4

04

OK

I= 3.4

3.21

Nature

S-21.22

4.22

PT 5 PT 6

Chain 4 Chain 4

Nature

..

L= 1.2 M=ll.12

OK

2.11.12

L= 1.2 M=11.12

OK

e

l

2.11.12

Radium treatment Radium treatment

PT 7

-4

f 50

M- 9.10 m=19.20

9 . lPO 19. 2010

Nature

PT 8

-4

150

M = 9.10 M=11.12

9.101' ll.lZ1"

Radium treatment

OK-25

ma19.20 s-23.24

20.19.23 .24

X-ray treatment

L= 1.2 Ms13.14

1.13 2.14

X-ray treatment

Mzll.12 M-13.14 ma17.18

11.13 12.17 14.18

Radium treatment

M-13.14 ~023.24

13.23 14.24

Radium treatment

PT 9

Chain 4

PT 10

04

OK

PT 11

06

125

PT 12

04

OK

.

There is only a single [ I . 21 chromosome but the italicized parts 1 and 2 "compensate" to form the equivalent of a [ I . 21 chromosome leaving the 9 and 5 portions shown in bold-faced type as excess material to affect the appearance of the plant. Each I n gamete, indicated by the arrows pointing down in the formula, contains the single [1 21 chromosome. The gamete with the chromosomes [I 91 and [2 51, indicated by the arrows

.

.

-

INTERNATIONAL CONGRESS O F GENETICS

111

pointing up in the formula, will lack the intact [ l 21 chromosome but have its equivalent. The excess . 9 and . 5 material will prevent this latter gamete from going through the pollen. If now the intact [ l 21 chromosome contains a particular gene, all the In female gametes, as well as all the effective pollen, will have this gene. A particular [1 21 chromosome, therefore, may be retained and a highly heterozygous race purified and rendered homozygous for the original [l 21 chromosome by continued male backcrossing of the heterozygous compensating type to a line 1 type that compensates for the same chromosome. Similarly, if a prime type such as a "B" race which involves the [ l 21 chromosome is crossed onto our compensating type Nubbin, a11 the I n gametes will be B's and the 2n offspring from selfing in consequence will all be B's as is shown by the following formula :

.

.

.

f

7

t

[9' lo]-[lo. 91-[9 .I]-[l .181-[18.17]-[I

t 7.2J-[Z.Sj-[S. 61-[6

1 7 Nb gamete '

51

1 1 1 5 1In gamete The method of continued backcrossing to appropriate line 1 compensating types has been used in retaining the interchanged chromosomes in cryptic types in nature while at the same time replacing the other chromosomes by those of line 1. The method is also being used in isolation of prime types and the retention of particular chromosomes from other species. The method outlined is fully successful only when crossing over does not occur. The compensating type, however, does not of itself prevent crossing over and hence a single chromosome passed through a series of compensating types will tend ultimately to resemble the homologous chromosome of line 1 so far as the genes which cross over are concerned. The method is more effective in the isolation and purification of prime types than of single genes since the end arrangement of chromosomes in the prime type is not changed, so far as is known, in passing through a compensating type. However, crossing over between the "B" race and the chromosomes of Nubbin is known to occur, and continued use of a compensating type should tend to render the genic content of the cryptic type similar to that of line 1 without, however, changing the order of the loci in the interchanged chromosomes. Compensating types have now been secured for 7 out of the 12 chromosomes. INTERSPECIFIC DIFFERENCES

Datura is not a large genus. Its species are shown in table 3. The chromoor by BERGNER somes of all the species listed have been studied by BELLING with the exception of some of the species of tree Daturas. All examined have

112

PKOCEEDIKGS OF THE SIXTH

12 pairs of chromosomes which show size ciasses somewhat similar t o those in line 1 of D. stramonium The evolution within the genus, therefore, has not been by changes in chromosonle number. The chronlosomes of the first 7 have been brought directly or indirectly into contact with those of line 1, D. stramo~zizlnz,with results which show that there are only 24 different kinds of ends in these 7 species. Evolution apparently has been accompanied Ily changes in the arrangement of the ends rather than in the ends themselves as attachment organs. The nature and origin of these terniinal organs of attachment are a major problem of chron~osoineevolution. TABLE 3 Species of Datura zcnder czrltivation. I. Stramonium group ( 1 ) D. stramonium (species used as standard) ( 2 ) D. ferox ( 3 ) D. quercifolia IT. Meteloides group (4) D. leichardtii (5) D. meteloides (6) D. innoxia ( 7 ) D. pruinosa ( 8 ) D. discolor (9) D. mete1 (several horticultural varieties) 111. (10) D. ceratocaula I\'. Urugmansia group (tree Daturas).

As ;r pi-eliininary to an analysis of these chroinosomal changes, there has been established for each species a tester race which has been used in the same way in which line 1 has been used as a standard for D. stratnotziunz in terms of which its cryptic chromosomal races are interpreted. In no other species has the chromosomal analysis been carried so far as in D. strantotziunz. In all species in which more than two races have been tested, however, cryptic types have been discovered. They are now known to exist in 11. stramoniztnz, D. quercifolia, D. Leiclzurdtii, D. nteteloidcs, D. innoxia and D. nzetel. Considerable progress has been tnade in the chromosomal analysis of the 3 species in the 1). stranzowizti~zgroup ( D . stranzonizcnz, D. ferox and 11. quercifolia). This species triangle is illustrated in the Datura exhibit and in a previous publication (BERGNERand BLAKE~LEE 1932) and need not be discussed in detail here. The chromosomes of D. strartzoniurtz that have taken part in the chronlosomal evolution of this group are given below the label. The numbered dominoes between the 3 species in figure 1 repre-

INTERNATIONAL CONGRESS O F G E N E T I C S

1

DATURA FEROX

I

113

DATURA QULRCIFOLIA

FEROX X STRAMONIUM

STRAMONIUM X QUERCIFOLIA 7 I

17

I

DATURA STRAHONlUH

FIGURE 1.-Diagram to show chromosomal differences between the three species Datura stramonium, D. ferox and D. quercifolia shown dt the corners of the triangle. The chromosomes which are not the same in all three species are represented by models. Those along the sides of the triangle represent the configurations which they form in hybrids between the species.

PROCEEDINGS O F THE S I X T H

114

sent the chromosomes and the configurations which they form in hybrids between these species. What the chromosome ends are in these configurations has been determined by crossing to appropriate prime type testers and noting the attachments of the chromosomes at metaphase in reduction division. In the 2 hybrids with line 1, D. stramonium, we recognize the "B" circle. Both D. ferox and D. quercifolia are "B"s (or P T 2). They are both also prime type 3, the cryptic prime type from Peru. If races from Peru, which are at the same time "B" and the prime type 3, were used as the D. stramonium tester, there would be only 1 circle of 4 in F, with D. quercifolia and a single configuration of 6 with D.ferox. These remaining configurations would then represent interchanges which have not yet been found within D. stramonium. So far as investigated in this and also in other groups, species hybrids always show configurations due apparently to segmental interchange. There does not seem, however, to be a close connection between the degree of relationship of the species and the number of attached chromosomes in their hybrids. Thus D. ferox and D. quercifolia are obviously more closely related to each other than either to D. stramonium. It depends upon what tester race is used in D. strantoniunz whether D. quercifolia gives the same number of attached configurations with D. ferox as with D. stramonium or fewer. The modified chromosomes of D. ferox and D. quercifolia have been freed from the other chromosomes by continued backcrossing to line 1, D. stramonium. The different prime types have been isolated or are being isolated by selfing and identifying the prime types by crossing to testers or by the use of compensating types. Thus through the use of the compensating type Nubbin we have isolated the B races of both ferox and of qwrcifolia and thus have now the B races of these 3 species to compare in otherwise line 1 standards. As a further test of these "B" chromosomes we shall isolate the interchanged B chromosomes in 2n [ l 181 and 2n [2 . 171 types and compare the morphology of the plants which result. It should be stated that in the process of purification by backcrossing to a standard, the end arrangements may be altered by crossing over. Thus from backcrossing D. ferox, two chromosomes ( [ I 9 20]"' and [7 203") are formed which are unknown in this species. Other species have been studied by means of their chromosomal configurations in hybrids, but their analysis has not proceeded so far a s with prelimthe species triangle just discussed. Figure 2 represents BERGNER'S inary findings in hybrids with 5 species. D. leichardtii acts as a bridging

+

+

INT,ERNATIONAL CONGRESS O F GENETICS

115

CONFIGURATIONS IN SPECIES HYBRIDS

0 DATURA

INNOXIA

-+ (3

\

+

@

+ @ +

5 BlVALENTS

+

DATURA METELOIDES

0

0J

FIGURE 2,-Diagram of hybrids between 5 species of Datura. Chromosomal configurations in F, generations, consisting of circles of 4, 6, 8 and of pairs, are shown on lines which connect the species.

species between our standard species D. stranzonium and the meteloides group with which D.stramonium will not cross directly. D. leichardtii differs from the line 1 D. stranzonium tester by only two interchanges. If a "B" race had been used as a tester they would appear to differ by only a single interchange. Specifically, however, they are very dissimilar. We are in the process of getting leichardtii chromosomes into otherwise line 1 plants

116

PROCEEDINGS O F THE S I X T H

in order to weigh the interchanged chron~osomesas extras and to learn what we can about the genic content of the leichardtii chromosomes in terms of those of the D. stramoniz~mstandard. W e are also getting line 1 testers into leichnrdtii to use in analyzing the chromosomes of the other species. The interchanged chromosomes of the other species we are freeing from their accon~panyingchrmosomes by continued backcrossing to D. leiclzardtii. Difficulties are involved in crossability. BUCHHOLZ (BUCHHOLZ and BLAKESLEE 1927) has found that pollen-tube growth may be a factor in preventing certain hybrid combinations. Thus the pollen of nletcloides grows

+ . +

FIGURE 3.-In center, plant of the 2n [ l 21 primary "Rolled." At right its 2n secondal-y "Polycarpic" and a t left its 2n [ 2 ' 21 secondary Sugarloaf."

+ [l

11

well on D. stranzoniuwz whereas the pollen of D. stranzoniunz bursts in the style of D. meteloides. Our survey of species has shown that segmental interchange has accompanied the formation of biotypes within a single species and also the differentiation of the larger groups called species. Since it has been shown that prime types homozygous for interchanged chromosomes do not necessarily differ from normals, it is not clear what if any role segmental interchange has played in bringing about the differences which characterize distinct species. I t has been somewhat difficult to conceive of species being evolved by the gradual accun~ulationof gene mutations. A method of adding whole

INTERNATIONAL CONGRESS O F GENETICS

117

blocks of genes at a single stroke, such as would be afforded by simple translocations, would be illore satisfying. It must be confessed, however, that our studies of Datura species have as yet given no evidence that such methods have been used in nature. SYNTHESIZED NEW "SPECIES"

Artificial new "species" can be synthesized by the addition of extra chromosomal material as is shown in mpre detail in another publication (BLAKESLEE, BERGNER and AVERY1933). The photograph (figure 3 ) rep-

FIGURE 4.-Two synthesized new "species," morphologically similar but chromosomally different. Each has two extra doses of the 2 half chromosome and breeds true. Capsules of these plants a r e shown in figure 5.

-

+

resents a 2n [ l 21 plant between its two secondaries, 2n + [ l . 11 and [2 . 21. The latter type has 2 extra doses of the a2 half chromosome. 2n This extra [2 . 21 chromosome BUCHHOLZ (BUCHHOLZ and BLAKESLEE 1932) has shown can be transmitted through the pollen. In consequence, it has been possible to build up types homozygous for this excess - 2 material, as is shown by figures 4 and 5. So far we have four types that are morphologically similar but chromosomally different. The first is the 2n [2 21 secondary called "sugarloaf" which is shown in figures 3 and 5 . I t has 25 chromosomes and since the extra chromo-

+

+

-

PROCEEDINGS O F THE S I X T H

118

FIGURE 5.-Capsule of a normal 2n plant from line 1 in comparison with those of four types which are different in chromosomal constitution but alike in having two extra doses of the 2 half chromosome. The 2n+ [2 21 is the secondary "Sugarloaf" with 25 chromosomes and can not breed true. T h e other three have similar sugarloaf-shaped capsules and are true breeding.

.

-

INTERNATIONAL CONGRESS O F GENETICS

119

some is carried by only part of the gametes, this type can not breed true. The second type (figures 4 and 5 ) lacks the normal [ I 1 . 121 chromosomes but is homozygous for a [ I 1 . 121 chromosome to which is attached a - 2 fragment. I t has two extra doses of the . 2 and breeds true. The third type lacks any [I . 21 chromosome. I t is homozygous for the [2 21 chromosome and also for the . 1 fragment. The - 1 and the - 2 portions compensate to form the equivalent of the missing [I . 21 chromosome leaving the other - 2 part in excess. Since the plant is homozygous it will accordingly have 2 extra doses of .2 and breed true. The fourth was synthesized by combining three prime types, PT 9, P T 10 and P T 12, which are listed in table 2. I t has two extra doses of the - 2 half and breeds true. I t was obtained too late this summer for a planting in the garden but capsules from offspring grown in the greenhouse, one of which is shown in figure 5 , show the sugarloaf shape due to the extra .2 material. The capsules of the four types with double doses of the - 2 half shown in figure 5 may appear slightly different, but as great differences in appearance could be secured in capsules taken from a single plant of any of the four types. At the time when the photograph was taken the choice of capsules was limited. The secondary 2n [2 . 21 plant in figure 3 and the two types in figure 4 are indistinguishable in gross appearance. The last type shows certain characters found in the 2n + 123 . 241 primary that suggest that it contains excess material of -23 or of .24 in addition to that of the - 2 half. It has been seen that we have produced two morphologically similar types that bieed true, the one with 24 and the other with 26 chromosomes. If we had in our collection of prime types the proper translocations we should he able to secure a pure-breeding plant with 22 chromosomes. Thus if we had the half translocated to the 113 . 141 chromosome, for example, we should be able to obtain a plant with the following formula:

+

a

1

Such a plant should be normal in appearance if no excess - 1 or a 2 material were present but would differ from normals if excess material were left over after the compensation. The homozygous types that have been obtained differ from random gene mutations in that the characters that they show have been definitely planned for. They offer a new method of bringing about variations in economic

120

PROCEEIIINGS O F THE S I X T H

fori~lsthat are propagated by seed. In producing such pure-breeding races we may perhaps be said to have exercised a measure of control of speciation, if such synthetic forms are not eliminated from the category of species by definition, because their origin is known. Whether new species with the same or different chromosome numbers have been produced by such methods in nature is not known, but the ease of their formation artificially when the excess material can he transmitted through the pollen suggests that the same process may have been utilized in natural evolution. In this as in many other processes devised by man nature may be found to have had the priority. L I T E R A T U R E CITED

BELLIXC, JOHN,and BLAKESLEE, A. F., 1926 On the attachment of non-homologous chromosomes at the reduction division in certain 25-chromosome Daturas. Proc. Nat. Acad. Sci. Washington 12: 7-11. EERGNER, A. DOROTHY, and BLAKESLEE, A. F., 1932 Cytology of the ferox-qzhercifoliastrantoiiiui~t triangle in Datura. Proc. Nat. Acad. Sci. Washington 18: 151-159. EERGNER, A. DOROTHY, S . ~ T I ~S., ' A ,and BL.ZI
+

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

121

I N H E R I T A N C E O F EDUCABILITY A FIRST REPORT O N A N ATTEMPT TO E X A M I N E PROFESSOR MCDOUGALL'S C O N C L U S I O N S R E L A T I N G TO IIlS E X P E R I M E N T FOR TIIE T E S T I N G O F

T H E HYPOTHESIS O F L A M A R C K

I;. A. E. Crezw, I+zstitute of Anir?ial Genetics, Edi~rburgk,.Ycotland

It will be well known to all here assembled that in 1927 there appeared the first report of WILLIAMMCDOUGALL'S Expcri~gzerzt.for the testing of the hypotlzesis of Ln~~zarclz and that his second report was published in 1930. These reports have attracted considerable attention, as was to be expected, since they deal with a sul~jectthat always arouses controversy and provides ample opportunity for the clashing of opinions and prejudices. It has to be agreed, however, that of all the experiments OII which the neo-Lamarckian case is based these of MCDOUGALL stand in a class by theillselves for the reasons that he used as his experimental material an animal stock which can be regarded illore or less as pure line and that he took very considerable precautions to avoid selection. You will remember that his experiment took the form of dropping rats 3 to 4 weeks old in the iniddle blind compartn~entof a water tank out of which there were two ways which led to platforms, one of these (alternately on the right and left) being illuminated and so wired that a rat stepping out of the water on to it received an electric shock. The rat had the choice of going either to the dim platforln and thus out of the tank without receiving a shock or to the lighted platforn~and getting a shock. Each rat was dropped in the tank 6 tiines daily until it learned to avoid the illuminated and live platform. The rat had learned when 12 times in succession it had taken the dim, safe platform. As the experiment proceeded the technique employed underwent several modifications, but after 12 generations of rats had been trained the procedure was finally standardized. In 9 generations thereafter the average number of errors fell from 80 to 25 and the number made by the best from 42 to 3. In the case of another related group the number of errors fell from about 170 in the first generation to 114 in the fifth. Finally, the worst performers were selected during 2 generations, I;ut in spite of this adverse selection the time needed for training fell. T o obviate the possibility of tradition, mothers of a sligl~tlytrained stock were mated first with slow-learning and subsequently with quick-learning males. The first inating gave an average number of 164 errors, the second of 62. MCDOUGALL'S tentative conclusions are as follows: "Twenty-three generations of rats have been trained in the tank to the

1 22

PROCEEDINGS OF THE SIXTH

performance of a specific task. The rats of the successive generations have displayed increasing facility in mastering this task. Whereas rats of the control stocks make on the average 165 errors (and receive the same number of shocks) before learning to avoid the shock, rats of the twenty-third generation of trained stock make on the average only 25 errors, the latter having acquired a greatly increased facility in mastering the task, the increase being measured by the difference between 165 and 25 shocks required for learning." "The average degree of facility shown by any group of rats is in the main a function of their genetic constitution." "In the light of our present knowledge there would seem to be only two ways in which such change of constitution as is shown by the rats of the trained stock can be brought about; first, by steady selection of such variations and mutations as may occur in the direction of such change; secondly, by transmission of modifications acquired by the rats in the course of training." "It seems to me very improbable that the first process, selection, can have played any appreciable part in producing the change of constitution and still more improbable that selection can have been the main or the sole process." "It begins to look to me as though Lamarckian transmission were a real process in nature and I submit for criticism the proposition, if continuance of the experiment, combining training with strongly adverse selection should result in steadily increasing facility, the reality of Lamarckian transmission will have been demonstrated." Now it so happened that I was sufficiently misguided as to consent to resecond report, and while doing this I became intrigued, view MCDOUGALL'S puzzled, worried, impressed, and shortly nothing would satisfy me until I had built a tank, collected some rats and had chosen my career-that of bathroom attendant to a rat colony-a pleasant enough billet even though the hours are long and the vacations few. The not unreasonable conclusions is arriving as a result of careful experimentation to which MCDOUGALL possess such importance that they cannot be washed away by a flow of words. His attitude, save perhaps when he seems to assume that biologists generally must ne'cessarily be hostile to his views, commands respect and makes it most necessary to examine with the greatest fairness and care the strong prima facie case that he has made out. H e himself, as I myself know from many conversations with him, would be the last t o deny that it is reasonable t o suspend judgment for a while until certain legitimate criticisms have been answered. records have never I t is the case, for instance, that since MCDOUGALL'S

INTERNATIONAL CONGRESS OF GENETICS

123

becn published in full it has been quite impossible for anyone to glean from them whether or not there is ground for thinking that capacity for learning is inherited in some particular fashion. I t is doubtful indeed that the plan could adequately examine this of experimentation adopted by MCDOUGALL matter. If capacity for learning were inherited, then of course any process of change would have been greatly accelerated if selection had unwittingly taken place. Secondly, as noted by SONNEBORN, the intensity of the shock found that rats varied considerably and was not measured. MCDOUGALL subjected to light shocks took nearly three times as long to learn as when the shocks were relatively heavy. I t follows therefore that a steady and progressive increase in the shock intensity could account for the results further points out that if the method of choosing obtained. SONNEBORN two rats "at random" from a litter was to take the first two available in the cage this could tend to the selection of rats of a peculiar psychological disposition. Then again it is found that in one of his experiments, broken found that the times needed for off after 4 generations, MCDOUGALL training actually increased generation after generation. Various other criticisms can be made, but debate alone can not satisfy: actual experimentation is demanded. Commonly, it is impossible accurately to repeat the experiment of another owing to differences in the genetic constitutions of different animal stocks and to differences in environments, but at least it is possible to devise a n experiment to yield results that can profitably be compared with those of a previous experiment conducted elsewhere by another investigator. This I have tried to do. Perhaps I am ill advised to communicate to a meeting of this kind a story that must necessarily be so incomplete. I have arrived a t n o conclusions other than that the inherent difficulties of such experimentation are varied and profound. Perhaps caution should have determined that I should have waited for 10 or 20 years when I might have exhausted the experimental approach to the problems that are involved and then announced that the riddle was solved or beyond solution by me. But I have gone far enough to realize very clearly that I need the advice of many colleagues if I am to continue the work hopefully and really intelligently. You will understand me when I say that there can never be the same joy in repeating someone else's work, however important this may be, as in prosecuting one's own. There is no real lasting satisfaction in proving that someone else is right or wrong. Yet this motive has been the mainspring of much endeavor. In planning this work I decided that a s far as possible I would imitate MCDOUGALL, but that I would so arrange matters that the data that emerged could be expected to illuminate the qction of any genetic factors that might

121

PROCEEDINGS O F THE S I X T H

1)e opei-ating. I l~uilta tank that as far as I was able to judge was a fair copy MCDOUGALL has visited me and of the latest edition of MCDOUGALL'S. agrees that save in one respect lny tank is to all intents and purposes similar to his own. The difference lies in the fact that the intensity of the light on the illuminated side is greater and more widely spread than in his tank. His light is situated beyond the platform; mine ( 5 candle power) is in the roof of the lateral passage and illuminates both the platform and the whole of the surface of the water on that side. My electric contraption differs from his. He uses a secondary coil. My supply is taken from the main which is 230 volts A.C. The current passes through a neon lamp (0.5 watt) to be reduced to 0.002 amp. In addition, an anode resistance, which usually is cut out, of 500,000 ohms is inserted in order that I can, when occasion requires it, temper the shock to the weakling. The rat with part of his body in the water and part on the platform closes the circuit. When he is entirely on the platform and moving on along the tunnel, the platform tips and the circuit is broken. Though I have no exact knowledge of the quantity of electricity that passes through the rat, 1 am allowed to think that it is fairly constant. I t should be remarked, however, that special manipulation of the switch is required in order to accommodate, on the one hand, the rat that rushes the platform and, on the other, the rat that is so feeble that it takes a relatively long time to crawl out of the water. I recognize that this system is far from satisfactory since by it I cannot hope to attain absolute uniforlnity. However, it works, and moreover I cannot afford to get an equipment such as is regarded by KNIGHT DUNLAPas the most desirable. He found that to obtain exactly the same current in each application and the minimum current necessary to evoke the response desired, it was necessary to en~ployhigher voltage and higher external resistance than were available to me. H e standardizes his shock a t 0.00015 amp., and this has no deleterious effect upon the rat. The shock used by me, like that used by MCDOUGALL, is such as will tetanize a rat which grips the platform with its teeth and will cause paralysis of the hind limbs, bladder and rectum in such as hold on to the platform for more than 3 seconds. This refers to the rats as a group; it is not necessarily true of any given individual, for rat differs from rat in the most remarkable fashion: a shock that merely tickles one rat will tetanize another, and danger attends every trial in the case of the rats during the first week of their training; for at this time they are younger and smaller. Even though one standardized the shock, one could not hope t o standardize the rats, so far as I can see. Therefore in the table that I shall show, the number of shocks is not a n exact measure of the total amount of current that has passed through the

INTERNATIONAL CONGRESS OF GENETICS

125

rat, since the time during which the circuit is closed by the rat varies from rat to rat and fro111 trial t o trial. Wistar derivatives, taken The rats I have used are, like MCDOUGALL'S, from the Institutional rat colony in which 1,000 breeders are regularly maintained and recorded. I began by taking the rats for training at 4 weeks old but I was forced to relinquish this practice for the reason that at this age my rats, unlike MCDOUGALL'S apparently and possibly because of poor coat development, could not endure 6 immersions in water a t 60' to 62'. After 2 or 3 trials the majority became water-logged and exhausted and were then not in a fit condition to make any considered choice, and moreover quite slight shocks (that is, of minimum duration) either killed or crippled. I decided to begin training at 7 weeks. On the day before training commenced, the individuals of the litter, previously weaned and thereafter kept away from their parents, were taken one by one from the cage between 10 and 12 midday and earmarked with serial numbers, No. 1 male being the first male to be taken, No. 1 female the first female to come to hand. MCDOUGALLtested and discusses the relative performances of generations. From the litters provided by one generation he took some individuals from each to be tested in their turn and to become the parents of the next generation. The performances of offspring and parent respectively are not given. The plan adopted by myself was to take twelve pairs of pedigreed rats from the Institutional rat colony and to train these, to take the pair with the lowest scores, the pair with the next lowest scores and so on, and thereafter to train every individual produced by these pairs, to take from among the individuals of first litters the pair with the best records, or with the worst records, and to continue this practice generation after generation. My methods will become clearer as the story unfolds. The original twelve pairs of rats were of exactly the same age and were taken straight from the rat colony when 3 weeks old. They were from 4 different litters out of parents closely related to each other. Their training commenced when they were 7 weeks old. They registered the following scores (one male and one female were paralyzed and discarded) : Pair 88 99

1 19 69

2 44 75

3 58 78

4 5 6 67 72 73 81 82 86 A n average o f 77.8

7 8 73 76 93 97 errors.

9 1 0 1 1 78 84 95errors 101 104 107 errors

Their training being completed, certain pairs were placed in separate cages to become the progenitors of the following lines : Line A $ 1 and ? 1

Line B $ 2 and O 2

Line C

8 3 and ? 3

Line G $ 7 and ? 7

Liqle I Line K $ 10 and 9 10 $ 11 and 9 11

126

PROCEEDINGS O F THE SIXTH

Line J has not produced a single litter. Line K has produced 1. The rest when compared with the general colony stock not exposed to the experimental procedures involved have produced relatively few litters. But there are in the colony plentiful instances of low reproductivity, and I cannot be certain that the training is responsible for the infertility that characterized this first generation of the experiment. On the other hand I am not yet sure that it will not be shown, as the experiment proceeds, that a considerable number of shocks and a low reproductive rate are associated. This remains an impression and nothing more for the present, but it is certainly the case that as yet my task is being complicated by the fact that on the whole my slowest learners are the poorest reproducers. If this proves to be a verifiable fact it will throw considerable light upon the question as to how it is that one generation can give better scores than its predecessor. As time passed litters appeared, to be trained in their turn, and soon it became very obvious that the scores of the individuals of this generation were to be infinitely better than those of their parents. When 39 individuals had been trained the best among them had a score of 0 and the worst one of 80, while the average of the whole was 23.8. This drop from an average score of 78 to one of 24 in one step was most disturbing and demanded prompt explanation. Naturally I jumped to the conclusion that it was my own capacity for learning and not that of the rats that had been weighed and not found wanting. So I obtained, over a limited period of time, a total of 100 more colony rats, 50 males and 50 females, and put them through the tank. The results I obtained were startling. The best performer, as was the case in the I?, made not a single error. Six times a day for 30 days in succession this rat was dropped into the tank and on each of the 180 occasions it emerged from the dim tunnel. It invariably avoided the light and therefore the shock. Here then was my first really serious complication. Was I dealing with a stock which, in spite of a hundred generations of inbreeding, included within it individuals differing from the majority in that they were photophobic? Now it so happened that I had decided to put all the rats 6 times through the tank with the lights working but with the shock current cut out before the real training began. I did this in order to accustom the rats to the water, the lighted channel alternately right and left, the platform and the tunnel and the handling. Referring to this rat's "pre-shock" record I found that it had been twice to the light a t the beginning. The F1 rat with a score of 0 had been to the light 4 times. In the case o f these rats it seemed to be true that, though they went to the light to begin with, when once they had become accustomed to the water and to the intricacies of the tank they thereafter always chose the dim side.

INTERNATIONAL CONGRESS O F GENETICS

127

The worst record scored by an individual of this 100, known to me as the Repeat P, (RP,) was 81 and the average of the whole was 18.09, a figure very near that of the so-called F1group. (These symbols, P and F,

have no genetical significance of course but they came readily to my tongue.) I show you a distribution of errors and individuals in RP1. It is extraordinary and surely significant o f something though o f what I confess I do not know. The great difference between the average scores of my original P1 and

128

PROCEEDINGS 01; THE SIXTH

of the Repeat PI demanded an explanation. I was quite willing to think that'improvements in technique were possibly responsible, since unquestionably many small improven~entshad become incorporated as time passed. But I was not, and still am not, sure that this is the case, particularly in view of the fact that recently I have been getting scores even higher than those of the original PI. There is this to be said. When I got the original PI individuals from the colony, rats of the right age were not plentiful and there was strong coinpetition for such as there were. I certainly received rats that were not wanted for other purposes by my colleagues. But I only borrowed the Repeat P, and this at a tiine when rats were plentiful. I then got the best, according to the judgment of those whose first charge is to select ainong young stock for future breeders. This may seem to be a matter of no importance, but I am not inclined to disregard it for on the whole it seems that such rats as are physically the inost vigorous, that is, such as would be retained as breeders, put up the better scores. ox0

I now possessed two individuals each with a score of 0. Fortunately one was a male, the other a female. I mated these t o found a separate line, for it seemed to me that if I had photophobia in my stock I must denlonstrate the relation of this to rapidity of learning. There was no reason for not assuining that two factors, photophobia and speed of learning, were involved in the experiment and that a photophobic rat might still be a stupid rat. If, perchance, a photophobic rat is a rat of relatively poor constitution ther, I might find myself first selecting for photophobia and then later against it. Furthermore, a con~plicationmight arise if photophobia proved to be genetically simple and speed of learning polygenic and greatly affected by environinental agencies. Again, since MCDOUGALL'S light is less intense than i~~inE, it might well be that in his case photophobics are not identified. These two rats have so far produced 2 litters which have been trained. The scores of their offspring were as follows: The best was 1, the worst 71, and the average 13.7. Froin the first litter I took the two best and the two worst males and females and mated them, best to best and worst to worst. S o far only the latter couple have given ille a litter which has been

128

PROCEEDINGS 01; T H E SlXTH

of the Repeat P, demanded an explanation. I was quite willing to think that'improvements in technique were possibly responsible, since unquestionably many small improvements had become incorporated as time passed. But I was not, and still am not, sure that this is the case, particularly in view of the fact that recently I have been getting scores even higher than those of the original P,. There is this to be said. When I got the original P, individuals from the colony, rats of the right age were not plentiful and there was strong competition for s ~ ~ cash there were. I certainly received rats that were not wanted for other purposes by my colleagues. But I only borrowed the Repeat P, and this at a time when rats were plentiful. I then got the best, according to the judgment of those whose first charge is to select aillong young stock for future breeders. This may seem to be a matter of no importance, but I am not inclined to disregard it for on the whole it seems that such rats as are physically the most vigorous, that is, such as would be retained as breeders, put up the better scores.

I now possessed two individuals each with a score of 0. Fortunately one was a male, the other a female. I mated these to found a separate line, for it seemed to me that if I had photophobia in my stock I must deinonstrate the relation of this to rapidity of learning. There was no reason for not assuming that two factors, photophobia and speed of learning, were involved in the experiment and that a photophobic rat might still be a stupid rat. If, perchance, a photophobic rat is a rat of relatively poor constitution then I might find myself first selecting for photophobia and then later against it. Furthermore, a conlplication might arise if photophobia proved to be genetically simple and speed of learning polygenic and greatly affected by enlight is less intense than vironmental agencies. Again, since MCDOWGALL'~ mine, it might well be that in his case photophobics are not identified. These two rats have so far produced 2 litters which have been trained. The scores of their offspring were as follows: The best was 1, the worst 71, and the average 13.7. From the first litter I took the two best and the two worst males and females and mated them, best to best and worst to worst. S o far only the latter couple have given me a litter which has been

INTERNATIONAL COIUGKESS OF G E N E T I C S

129

trained. The best record among these generations was 0, the worst 118, and the average 48.1. The male with 0 never went to the light, not even during his pre-shock career. I shall mate him to his grandmother. I t is too early for me to discuss this particular aspect'of the problem further. I have examined the pre-shock career of my trained rats in order to see whether or not the number of times a rat went to the light and to the right or left had any relation to its subsequent record. The following table gives the relevant figures: To A

To Light

No. o f Anima!s

No. Expected

Average N o . of Errors

Times

35 40 51 59 72 53 40

5.47 32.81 82.03 109.38 82.03 32.81 5.47

34.66 19.88 27.14 23.41 19.82 18.85 22.25

0 1 2 3 4 5 6

No: o f

No.

Avernqr No. of Errors

5.47 32.81 82.03 109 38 82.03 32.81 5.47

20.71 21.67 21.08 26.65 20.49 19.55 24.00

A n ~ n ~ a l s Erpectrd

7 30 63 140

76 31 3

I t is clear that the way a rat habitually goes during the pre-shock period is not determined by chance. Many of the rats exhibit the tendency to be strongly right- or left-handed. The slight asymmetry of distribution shows that there are rather more right-handed than left-handed rats. I t would seem, perhaps, that the strongly left-handed rats learn somewhat less quickly, but, of course, greater numbers are required before this point can be settled. The figures for the number of times the rats go to the light during the pre-shock period have to be related to those of right- and left-handedness. For the first of the six trials the light is at the right-hand and thereafter is alternated from side to side. The two classifications are not independent as is shown in the following table: A A A A A A A

rat going to the right 0 times will go to the light 3 times rat going to the right 1 time will go to the light 2 or 4 times rat going to the right 2 times will go to the light 1, 3 or 5 times rat going to the right 3 times will go to the light 0, 2, 4 o r 6 times rat going to the right 4 times will go to the light 1, 3 or 5 times rat going to the right 5 times will go to the light 2 or 4 times rat going to the right 6 times will go to the light 3 times

The fact that there is a large excess of totally right- or left-handed rats necessarily means that there must be an excess of rats that go to the light 3 times. This is the case. The completely left-handed rats make lower scores: this is reflected in the higher scores made by rats that went 3 times to the light. I t will be noted, however, that the dependence of the light classification on the direction classification does not affect the symmetry of the

130

PROCEEDINGS O F T H E S I X T H

former, and, actually, the light distribution is symmetrical. There is no reason to suppose from the figures given that the rats tend to avoid the light or vice versa. This makes the suggestion of photophobia very unlikely. It is clear that right- and left-handedness. are much more important in relation to the score made by a rat than is reaction to light. Now let me turn to such results as I have so far obtained and consider them as a whole. The following table presents them in a concise form: Generation

PI PI Fl F, Fa

F4 RP1

No: o Animafs

22 10 39 77 76 15 100

Coeficient Variationo f

Mean

25.3 34.5 78.7 76.7 291.6 68.0 72.4

77.8 71.1 23.9 20.4 27.9 54.9 18.1

I confess I do not know what to make of these figures. The large coefficient of variation in F, and F, is due to the appallingly bad scores of one litter in each, and even in these some individuals have exceptionally good scores. Apart from the F, the variability of all the generations is slight. If I explain the drop from PI to F1 as a reflection of refinements in technique how am I to explain the worsening seen in Fz- F4?It cannot be technique when some individuals in a litter give me scores of 4 while others give 140. I have tried to think that a seasonal influence is operating, but I cannot satisfy myself that this is so. My numbers are not yet sufficient for genetical treatment, and all that I can hope to do for the present is to explore such as I have in order to gain ideas that later can guide experimentation. As I have said, the PIanimals with the higher scores have given the fewest offspring, and so it is that the parents as a group are more rapid learners than are the offspring as a group. The mean of 63 parents (that is, of the parents of each of all the litters) is 13.1 ; that of 168 offspring is 26.9. There is no help here. I then divided the rats according to their scores into 6 classes as follows. The classification is an arbitrary one, being based on my own impressions that there were significant differences between rats falling into the different groups and that the divisions were roughly equal. Score

q-5 6-10 11-16 17-30 31-50 51+

Class

= = =

1 2

=

3 4

= =

5 6

INTERNATIONAL CONGRESS O F GENETICS

131

I considered the performances of offspring in relation t o those of the parents : Matings

Classes Of Offspring

2

1

3

4

5

6

I t is seen that I am getting more high scorers out of low by low than low scorers out of high by high. The mid-parent-offspring correlation is 0.38t0.071. The fraternal correlations (only litter mates are regarded as brothers and sisters) are as follows: Brother: brother Sister: sister Brother : sister

0.31 0.54 0.33

It is seen that sisters are more like each other than brothers are like each other, or than brothers are like sisters. Perhaps sex-linked factors would possibly account for the differences observed. In an attempt t o examine the question as to whether selection was playing experiment, in view of the fact that he was taking a r81e in MCDOUGALL'S two individuals from a litter and leaving the rest, I compared the performances of my own number ones wi!h those of the rest. (The individual first taken from the cage becomes number 1.) There is no difference between them. Total Animals

No. 1 Rest

114 20 1

Total Errors

268 1 4814

Average

23.5 24.0

This is as far as I have traveled and already I have a feeling that I have lost my way. The reason for my doubts is that the behavior of different rats in the tank is so remarkably different. During the past two years, again and again I have thought that I could make some generalization or other and again and again I have been forced to jettison it. After the rats have been through the pre-shock stage, it is possible to group them roughly into different behavior classes-those that rush the platform and those that float-in other words those that seem to regard the shock as the lesser of two evils and those that hate the shock worse than the water. Rats of these

132

PIIOCEEIIINGS O F THE S I X T H

classes are not rapid learners. Both continue to make mistakes until, without any warning, they quite suddenly complete the task. They belong to classes 4 to 6 (17 to 51 errors). S o also do the "hopeless" rats who paddle feebly round the central compartment of the tank, who do not squeak and who have the greatest difficulty in hoisting themselves on to the platform when they decide t o crawl out of the water. The best rat of all is the "eager" little fellow who dislikes both water and shock but fears neither. It is he who actively explores both lateral divisions before leaving the tank by one of the two routes. When he gets his first shock he comes through like a ball off a cricket bat. The next time he touches a platform, he screams although he is on the safe side. But he recognizes either a t once or within one or two days that light is associated with pain and darkness with no pain, and he leaves his task with a score of 1 to 5 or, in exceptional cases, continues to make one o r two a day to give a total score of about ten. I am of the opinion that when I have seen a rat in the water on his first or second day of training I can foretell his score fairly accurately. I have amused myself by so doing. I have seldom been grossly wrong. MCDOUGALL discarded runts and the obviously weakly members of litters. I trained every rat in every litter. I found the runt to be, in every case, better than the average of the litter to which it belonged. But the obviously weakly individuals that cropped up occasionally either gave high scores or else were paralyzed and killed. They have great difficulty in getting out of the water and hold on to the platform too firmly and too long. MCDOUGALL is of the opinion that the behavior in the tank of rats of trained stock distinctly differs from that of untrained or only slightly trained stock and that the differences can be best expressed by saying that the behavior of the rats of highly trained stock is markedly more tentative, hesitating, cautious or exploratory than that of the other rats. I n his experiences the rat of untrained stock, in the early part of its training, loiters very little, hesitates very little, and rushes a t a gangway ; whereas the rats of the trained stock loiter, hesitate, oscillate between the two lateral passages and frequently change their ininds when about to leave by a particular gangway. Even when they approach closely either the dim or the bright gangway they may turn back from it and not infrequently they approach cautiously the bright gangway, touch it lightly with nose or paw, receive a slight shock and then retreat. I agree that these two types of behavior are to be recognized but I can not agree that in this way trained can be distinguished from untrained stocks, for if a hundred rats are taken straight from the colony and trained, it is found that both types of behavior are exhibited and are equally common.

+

INTERNATIONAL CONGRESS O F GENETICS

133

I think that I must try more accurately to classify the rats according to their behavior for I have been greatly impressed by the fact that in one and the same litter one can get two or more different behaviors exhibited and that those individuals who behave alike have scores that a r e very similar. On the other hand one can get litters of which all the individuals exhibit PEDIGREES O F L I N E S A, B, C, G , AND K

A

6

I . . I9 37 48

DEAD.

S EATEN.

134

PROCEEDINGS OF T H E SIXTH

PI

I

::91:6:5

,

,

,

386165.69.532

9s

.

107

Ft.

".:

B t?I 2.7 3 2: 4:

the same habit in the water, all rushing, all floating, all hopeless or all eager, and then it has been possible to identify the litter to which an individual belonged by reference to the behavior in the tank. It may well be that as this experiment proceeds I shall find myself studying the mode of inheritance not of capacity for learning but of a peculiar behavior pattern. For indeed it would seem to be the case that there are rats that persistently leave the tank by one route, rats that turn to right or left indiscriminately; rats that tend to avoid the light, rats that do not mind it; rats that are quick to associate light and shock, and rats that are slow. It must be my task to examine the possible genetic bases of these attributes which certainly distinguish rat from rat. At the end of my story, as at the beginning, I am very aware that it is hopelessly incomplete. When I began this experiment I proposed to examine MCDOUGALL'S conclusions. Now, my sole concern is to formulate conclusions of my own. And this I cannot do. Perhaps I should have remained silent but maybe I am wise in thus imitating our voluntary hospitals which habitually parade their poverty and plead for contributions; LITERATURE CITED

CREW,F. A. E., 1930 Lamarckism (Review of MCDOUGALL'S Second Report on a Lamarckian Experiment). Eugenics Review. 22:S.S-59. DUNLAP,KNIGHT,1931 Standardizing electric shocks for rats. J. Comp. Psychol. 12: 133135. MCDOUGALL, WM., 1927 An experiment for the testing of the hypothesis of LAMARCK. Brit. J. Psychol. (General Section). 17: 4. 1930 Second report on a Lamarckian experiment. Brit. J. Psychol. (General Section). 20: 201-218. T. M., 1931 MCDOUGALL'S Lamarckian experiment. Amer. Nat. 65: 541-550. SONNEBORN,

I N T E R N A T I O N A T , C O N G R E S S O F GkINETICS

135

MENDELISM I N MAN C. B. Davenport, Carl~egieIl~stitutiortof was hi rag tor^, C o l d Spring Harbor, hrew' Y o r k

When, some thirty-two years ago, MENDEL'Slaws were rediscovered and his work became famous, one of the first questions that arose was whether the laws of MENDELwere applicable to man. For it was at once seen that if the prediction of an average result could be replaced by a definite statement of the consequences of a particular mating a great step would be gained. The question was answered differently by different investigators, and, it will be recalled, a bitter controversy arose in England between KARLPEARSON, as biometrician, on the one hand, and WILLIAMBATESGX, as Mendelian, on the other. The first demonstration of the application of MENDEL'Slaws to man then a student of W. E. CASTLE. was made by the late W. C. FARABEE, Very soon papers multiplied, and it became known that the inheritances of eye color, skin color, hair form and many diseases of the skin, eye, etc., clearly showed segregation and were based, in some cases, on simple Mendelian factors. I t is because of this avalanche of facts that PEARSON has ameliorated his opposition to Mendelism. Students of heredity in man soon began to run into difficulties due to evident complications in the laws of human inheritance; there seemed indeed a certain justification for PEARSON'S opposition to the universal applicability of Mendelian laws to human heredity. The difficulties in testing Mendelian theories in man are very great and have been often enumerated. There is, first, the great time which elapses between generations, in consequence of which it is difficult to get precise firsthand information concerning even as few as three generations. The best way of meeting this difficulty is, it has long seemed t o me, to have established one or more repositories of human records where these will be preserved through the generations. I t was with the aim of securing some such preservation that the EUGENICS RECORDOFFICEwas established by Mrs. E. H. HARRIMAN. However, since this office carries data concerning less than two in a thousand of the population of the United States, it falls short of meeting the need of a general repository. For selectively bred cattle, hogs, dogs and especially race horses there are rather complete herd books and stud books; there are grave difficulties (chiefly financial) in the way of creating any such pedigreed list of man. But for genetics far more is needed than a pedigreed list, namely, a quantitative statement of the traits and performance of each individual, and this requirement means immensely increased time, expense, and bulk.

136

PROCEEDINGS O F T H E S I X T H

The second difficulty lies in the small sizes of fraternities, since it is almost impossible to apply Mendelian formulae where there are only one to three children. Tied up with the difficulty introduced by small families is the considerable error introduced by selection of families. If one is collating the incidence of a trait in fraternities of offspring when neither parent shows the trait, one naturally is forced to select just those fraternities that reveal the trait. All those fraternities are omitted that might have shown a member with the trait had the fraternity been large enough, but in which it is (by chance) absent, owing to the small number in the fraternity. By these omissions the proportion of persons showing the trait in the sum of the fraternities is too high. The errors due to the small size of families and to selection of only those families showing the trait can be, in part, corrected by the use of WEINBERG'S method, or some other similar methods; but by applying these methods one diminishes greatly the number of available fraternities. Another difficulty lies in the circumstance that many human traits depend upon multiple factors. This is, doubtless, true of hair color and hair form. I t holds true in the range of skin colors found in man. Only recently I have been working upon two apparently unrelated traits, namely, tendency to goiter and a type of hardness of hearing called otosclerosis, and have found that in each the facts are best in accord with the hypothesis of two genes, one sex-linked and one autosomal. If, as seems not impossible, a trait is due to three or more genes, it will be very difficult if not virtually impracticable with our present methods and with the difficulties inherent in human pedigrees to determine the fact definitively. While in Drosophila the number of traits depending upon multiple factors is proportionately small, in man such characters seem to be predominant. I t is perhaps hardly to be wondered at, since man himself is so complicated in his development and, just because he lies a t the end of a long period of evolution, has accumulated a large number of mutations. Also, it is probable that through his capacity for adjustment to conditions of life, even under bodily and environmental handicaps, more of these mutations have been preserved than would be the case in a species that was less plastic and adaptable. As medical skill and state care of defectives advance it is probable that more and more of these mutations will be heaped up in the population. The same gene will have a chance to undergo further genetical change so that we may expect many an allelomorphic series, such as we have in hair color, to arise and many genes to cooperate in a summation result which we call a human trait.

INTERNATIONAL CONGRESS O F GENETICS

137

Another difficulty grows out of the principle, applicable in the realm of diseases and defects, that clinical entities are not necessarily genetical entities. This is now familiar enough to experimental geneticists in other fields. Thus there are perhaps one hundred genetical types of albinism in corn. In MACDOWELL'S mouse colony there are, or have been, a t least three types of circular movements allied to the whirling of the Japanese dancing mouse. In man, myopia, in some pedigrees, shows a sex-linked factor; in others that factor seems to be absent. Other examples are illustrated in WAARDENBURG'S recent book on the inheritance of eye defects. Still another difficulty grows out of the sensitiveness to environmental changes shown by many human characters in their development. This is probably, in part, due to the circumstance that human development is so prolonged, continuing for years, as contrasted with the development of the Drosophila or Daphnia which takes place in less than a week. This interaction of heredity and environment is well illustrated in the case of goiter where a hereditary insufficiency in the development of the thyroid gland may pass unnoticed in an environment rich in iodine. I t first becomes apparent when the iodine in the water or food falls below a certain minimum. Though, as often stated, man is perhaps the worst species in which to work out the laws of genetics, yet it is of great importance to mankind that the principles established in other organisms should be tested in man. Also there are certain traits, especially mental ones, whose inheritance can be better studied in man than in any other species, partly because of our familiarity with his physiological and psychological variations and the opportunity to get his cooperation in their measurement. I t is, indeed, possible to study inheritance of wildness in rats. But the picture of hyperkinesis in man is infinitely richer in detail-the joviality, the generosity, the briskness, the impulsive actions, the erotic tendency, the flow of speech and other familiar traits give us many criteria of the hyperkinetic state and suggest its complex nature. On the other hand, in his instincts and special capacities man is hardly superior to dogs as an object of investigation in heredity. This is not because his instincts and capacities are not as varied but because, for the most part, mankind has not mated so as to produce strains characterized by such special instincts. Dogs are so mated. However, in some cases matings of similar instincts occur, as between naval families whose social interrelations are such that the young men and women of these families are brought early into contact. Similarly, there is selective mating among biologists, partly because of the existence of summer marine laboratories like those at Woods Hole

138

PROCEEDINGS OF THE SIXTH

and Cold Spring Harbor where the young people associate in informal and intimate fashion. Despite the unassorted mating obvious in most cases there is a good deal of social stratification. Thus we have the scholastic stratification, seen in the matings inside of college communities, the stratification among politicians and statesmen, who sojourn long with their families at the legislative capital, the stratification of artists who tend to live in colonies, the stratification of the deaf who can converse only with other deaf who know the sign language, the stratification of exiled missionaries, the stratification of the farm communities and the stratification, in this country, of the valley communities with their high incidence of feeble-mindedness. These stratifications, or castes, if you will, afford an opportunity for the study of selective matings, even in a species whose matings seem a t first glance so uncontrolled as man's. While the slow development of man increases the interval between generations it gives opportunity for prolonged testing and analysis of mental and emotional traits. The school is, indeed, a place of revelation of instincts, tastes, and special capacities, both mental and physical. The physical and scholastic achievement records of our schools are such as to afford material for the study of inheritance of mental capacity. In one other respect the study of man offers advantages for Mendelian analysis, and that is in the frequency of twins, especially monozygotic twins ; for about one quarter of one percent of human births is a birth of monozygotic twins. Now it is clear that in no other animal can we know in such detail and so quantitatively the physical, mental and temperamental qualities as in man, and, accordingly, human twins can be compared with a detail not possible in other animals. By analysis of twins we can distinguish between traits whose development is more influenced by environmental changes and those in which environment plays a smaller part. The opportunities for genetic analysis in twins are almost limitless. Finally, reference may be made to a characteristic of heredity in the higher vertebrates which, on the one hand, complicates their Mendelian analysis, and, on the other, adds a certain interest to its study. This is intervention of the endocrine glands between the genes and the full development of the trait. For example, the study of the heredity of dwarfism becomes the study of the inheritance of an insufficiency in the anterior pituitary gland. The heredity of fleshy body build with male genital dystrophy becomes that of a defect of another anterior pituitary hormone. The study of hereditary goiter becomes that of the heredity of thyroid functioning, and so on. I t becomes daily clearer that the endocrines form an important link between the genes

INTERNATIONAL CONGRESS O F GENETICS

139

and the finished body. Because of the responsiveness of the endocrine glands to environmental conditions they play an important r6le in adjustment to environment. This interaction of heredity and environment is well illustrated in the case of goiter. Goiter is commonly said to be due to insufficiency of iodine in the water or food. But this cannot be the whole story, since in one and the same valley where goiter is endemic certain families will be quite immune. Also, on the sea coast where iodine is abundant, sporadic families will be found which have always lived in the locality and yet suffer from goiter. Evidently the symptoms of goiter begin to appear where a certain degree of thryoid effectiveness meets a certain degree of noxiousness of environment. Thus, if the thyroid be inefficient, then, despite a good environment, it will fail; the environment becomes more noxious, through reduction in iodine content, and a larger proportion of thyroids will prove to be inadequate; as the iodine content falls to its minimum the incidence of goiter in the population becomes a maximum, I n general a developmental trait requires for its expression at least a minimum of the specific internal impulse combined with a t least a minimum of supporting environmental conditions. As either of these factors increases the expression becomes more marked. Thus the studies of the past three decades in human heredity have, by the application of Mendelian principles, been given a very practical slant such as would never have accrued from the purely statistical studies based on masses. What of the future? Are we in a position to predict the directions that the studies in human heredity will henceforth pursue? Let us try to forecast a few. First, efforts t o complete the investigation of the inheritance of human traits must be made. Expensive as they are, they are of such practical importance for the future of mankind that they must not be neglected. Second, in human traits as in the case of those of other organisms we should no longer be satisfied with factorial analysis but should try to get a clearer insight into the way the genes do their work by tracing the ontogenetic development of each trait. This is the point of view that HAECKER long ago emphasized. KRISTINEBONNEVIE has shown how it may be carried out by her fine researches on the development of the papillary ridges of the finger tips. Some students of the comparative embryology of human races, for example, A. H. SCHULTZ,reveal how early and in what fashion the differential characters between Negroes and Europeans show themstudents of D U N N ,are getting selves in ontogeny. CHESLEYand KAMENOFF, most interesting results from a study of the ontogenetic history of the

140

PROCEEDINGS O F THE SIXTH

crooked-tailed mice; this is outside the human group of studies to be sure but illustrates the method and the interpretations to which it leads. Third, since with humans it is rarely possible to get the earliest stages of embryos that we are certain are going to develop a particular trait, while with the rapidly growing mammals this is often possible, it seems desirable to carry along with human genetical researches strains of mice or rats. There are a large number of morphological and physiological characters that are common to mice and men of which the ontogenesis of the genes can be much more satisfactorily worked out upon the animal. Fourth, since we now know that aberrations in the chromosomal complex are responsible for irregularities of development in both plants and animals, it is reasonable to look for them in man also. Herein may lie the cause o f some profound defects that are clearly familial but the method of whose inheritance is not easily revealed. I t would seem that, if anywhere, we should find such chromosomal irregularities in the group of feeble-minded. Some years ago I was able to assist PAINTER to get some perfectly fresh testicular material of a mongoloid dwarf. But, PAINTER tells me, this material revealed no obvious chromosomal irregularities. However, this negative result should not discourage us from continuing the search for possible chromosomal irregularities in genetically complex defects. Such chromosomal irregularities have, indeed, been found in cancer cells; they are, consequently, not foreign to human tissues, nor, probably, to human gametes. T o sum up: Mendelian studies in man offer an alluring field for future investigation, not, indeed, for the determination of fundamental laws of genetics but for the application of the laws to that species upon whom all progress in science depends and upon whom the social order that makes scientific work possible and even congenial rests. A more precise knowledge of the inheritance of traits will contribute toward an insight into the consequences of particular mate selections and of race crossing. Despite the difficulties inherent in genetical work with this species continued research is justified on man if for no other reason than that in no other can an investigation so well be made in the mental field as in the study of twins and perhaps in the endocrine field, opening up a new method of study of the way the genes do their work. In the future more stress will be laid on the internal control of development by the genes, with the aid of experimental mammalian material. Also the explanation of some irregularly inheritable traits will be sought in chromosome irregularities. Thus the solution of problems in human genetics will require in the future the aid of the embryologist and the cytologist.

INTERNATIONAL CONGRESS O F GENETICS

141

T H E P R E S E N T S T A T U S OF MAIZE GENETICS1 R. A. Emerson, Cornell University, Ithaca, N e w York ADVANTAGES A N D DISADVANTAGES O F MAIZE A S GENETIC MATERIAL

As material for genetic studies, Zea mays has many advantages and not a few disadvantages. Of first rank among the latter is the long life cycle which makes it impracticable in temperate regions to grow more than one generation a year or, at best with greenhouse facilities, two generations. The relatively large number of chromosomes, ten pairs, with a correspondingly large number of linkage groups, adds to the difficulty of genetic analysis. The necessity of guarding the pollination of plants even to obtain selffertilized seed is not an unmixed evil. The technique of pollination is so simple and so many seeds result from a single pollination that it is easily possible to obtain 10,000 seeds from a single hour of work. The wealth of genetic material available in maize is undoubtedly directly related to the cross-fertilization prevailing in this plant. Even relatively weak recessive mutations are long preserved against elimination by natural selection under cover of their dominant normal allelomorphs. No variety of maize not previously inbred fails on self-fertilization to reveal numerous recessive characters. This is in sharp contrast to naturally self-fertilized species. Maize, being prevailingly diploid, has, for genetic purposes, the further advantage over such polyploid species as common wheat that a recessive mutant gene can manifest itself a t once without the necessity of a second identical mutation in a duplicate chromosome. In this connection may be noted the suggestive results of STADLER (1929), showing a much higher rate of X-ray induced mutation in diploid barley, oats and wheat than in polyploid oats and wheat. A not inconsiderable advantage of maize for genetic analysis is the fact that this plant lends itself readily to cytological examination. The marked differences in size, form, and other morphological features between the several chromosomes of maize are of the greatest advantage in cytogenetic studies. I cannot refrain from noting here a very real advantage experienced by students of maize genetics, which is in no way related to the peculiar characteristics of the maize plant. I am aware of no other group of investigators who have so freely shared with each other not only their materials but even their unpublished data. The present status of maize genetics, whatever of

' Paper No.

190, Department of Plant Breeding, CORNELL UNIVERSITY, Ithaca, New York.

142

PROCEEDINGS O F T H E S I X T H

noteworthy significance it presents, is largely to be credited to this somewhat unique, unselfishly cooperative spirit of the considerable group of students of maize genetics. In this connection I want gratefully to acknowledge the help of many persons who have contributed directly or indirectly to this summary statement of the status of maize genetics. Formal credit by means of citations of literature will be given only for the more recent papers. SIMPLE AND COMPLEX INTERACTION O F GENES

During the past twenty years the mode of inheritance of many characters of maize, involving more than two hundred genes, has been more or less successfully studied. The genes studied influence such diverse characters as color of aleurone, endosperm, plumule, scutellum, pericarp, silks, anthers, and leaves, endosperm composition, chlorophyll abnormalities, plant stature, root development, pollen sterility, sex abnormalities, defective endosperm and embryo, branching of ear and tassel, form of leaves, development of silks, rate of pollen-tube growth, carbohydrate metabolism, disease resistance, and even chromosome behavior during meiosis. Most of these characters are simple recessives, but several of them involve complex relations of duplicate genes, complementary genes, and multiple allelomorphs. Although most chlorophyll abnormalities of maize are simple recessives, normal chlorophyll development is an outstanding example of genetic complexity. From the fact that some sixty recessive genes are known, any one of which prevents perfectly normal chlorophyll development, it follows that all sixty of the dominant allelomorphs of these genes are essential to the production of normal chlorophyll, and it is highly probable that, a s yet, we know only a small part of the story of the genetics of chlorophyll development. One of the best examples of extreme complexity of complementary interaction of genes, known to the writer, is that afforded by the purple-red series of aleurone and scutellum colors of maize. Since aleurone color ordinarily does not develop except with the genotype A, A2 C R i, it is obvious that numerous genetically different types of colorless aleurone may exist. It is equally to be expected that very diverse F, ratios of colored to colorless aleurone should be encountered. But when it is recalled that scutellum color develops only with the aleurone color genotype noted above and in addition with S, plus any two of the three genes S,,Ssand S,, it will probably be granted that a really complex genetic set-up has been uncovered in scutellum coloration. Studies of scutellum color (SPRAGUE1932) have a n important bearing

INTERNATIONAL CONGRESS O F GENETICS

143

on an earlier idea of the possibly fundamental difference between the interaction of duplicate genes and of complementary genes. Given the aleurone genotype A, A, A, A, C C R R i i plus the scutellum genes S, Sl, the F, ratio of colored to colorless scutellum resulting from the heterozygous condition of two of the other three scutellum genes, say S, s, S3 s3,is dependent alone on whether the remaining scutellum gene is homozygous dominant, S4 S4, or homozygous recessive, s4 s4. Thus, Sz sz S3 s3 gives with sa s4 a 9:7 ratio and with S4 S4 a 15:l ratio. Obviously the particular ratio depends not alone on the interaction of the genes that happen to be heterozygous but rather on their interaction with the residual genetic mass. It follows, therefore, that there may in other cases be no very fundamental difference between so-called duplicate genes giving a 15:l ratio and complementary genes giving a 9 :7 ratio. This scutellum-color situation, moreover, indicates that a 15:l ratio alone cannot be regarded as critical evidence of duplicated chromosomes. Examples of multiple allelomorphism in maize have long been known. The more important of these involve the P, R, and A, loci. Several combinations of pericarp and cob color are dependent on allelomorphs of P. They present no unusual features, but one of them, P",is a convenient tool for use in studies of variegated and mosaic pericarp. The cross variegated pericarp, variegated cob X white pericarp, red cob, P"X Pwr, produces in F, the following genotypes : pvv

1 - variegated pericarp, variegated cob pvv

PVV 2 - variegated pericarp, red cob

pw

pwr

1

white pericarp, red cob p r

It is, therefore, possible to separate the heterozygous from the homozygous variegated ears by reference to the color of the cob without the necessity of growing progenies of the two classes. The R series of allelomorphs exhibits peculiarities worth noting. R' is dominant for both aleurone and plant color and f l is recessive for both; Rg is dominant for aleurone color and recessive for plant color, while r' is recessive for aleurone and dominant for plant color. and ANDERSON 1932) exhibit no simple The A, allelomorphs (EMERSON

144

PROCEEDINGS O F THE SIXTH

sequence with respect to dominance. Thus, A, is dominant to al for aleurone, plant and pericarp colors. A,b is like A, in its dominance to a, with respect to aleurone and plant color but is dominant to A , with respect to pericarp color. The allelomorph aDis similar to a, in its relation to plant color, like A,b in relation to pericarp color, and has an effect on aleurone color intermediate between the effects of A, or A,b and of a,. Until recently all that could be said about the inheritance of quantitative characters in maize was that the results could be interpreted on the basis of the interaction of numerous genes affecting size, length of growing season, 1931) that a quanetc. Recently, however, it has been shown (LINDSTROM titative character, number of kernel rows, is linked with genes for wellknown qualitative characters. This may well be accepted as proof that quantitative characters are inherited just as are other characters. As yet, however, no thorough genetic analysis has been made of any quantitative character in maize. An attempt by the writer to do just this for a relatively simple quantitative character has been under way for ten years. The results to date do not encourage him soon to undertake the genetic analysis of so complex a character as yield of grain or forage. Doubtless, however, by some date well in the future some progress toward this goal will have been made by somebody. LINKAGE I N MAIZE

Ten linkage groups are now known in maize, corresponding to the ten pairs of chromosomes of the ordinary diploid form of this plant. Somewhat over one hundred genes, or about half of those studied, have been assigned to one or the other of the linkage groups. In some cases the order of the genes and their relative linkage-map distances apart are known. Of other genes the approximate order is known, and of still others our only information at present is that they belong to a particular linkage group. In the accompanying diagram the approximate location of genes is shown for the several linkage groups, those genes not even approximately located being shown below the map line. With a single exception, all these genes have been placed in their respective groups by purely genetic methods. In general the first indication of linkage is obtained from F, cultures. The next step is to examine backcrosses to the double recessive. Since most newly found genes are recessive to their normal allelomorphs, the repulsion phase of linkage is involved in most instances. With such material backcross data are better than F, data. I f , as often happens, plants with recessive characters are weak and double recessives still weaker, failure to recover the double recessive type in an F,

INTERNATIONAL CONGRESS OF GENETICS

145

culture of moderate numbers affords little evidence of linkage. The standard practice, as soon as double recessives are obtained, is to test the genes in backcrosses and from these to obtain material exhibiting the linkage in coupling phase. In case of lethal or near-lethal recessive genes, backcrosses

a--C

ad*

125- -bm2

FIGURE 1.-Linkage maps of maize. Genes whose loci are known only approximately are starred. Genes known to belong to a particular group but whose loci are not even approximately known are listed below the appropriate map line. Rg should have been placed near t.,, chomosome 111, and h near R, chromosome X, and starred.

are out of the question. In general, however, close linkage can be determined readily in F2 progenies; and, with loose linkage, double recessives for backcrosses are easily obtained. Maize affords numerous illustrations of the difficulty experienced in attempts to determine the linear order of genes when only two pairs are involved in any one culture. The variability of different cultures in percentage

146

PROCEEDINGS OF THE SIXTH

of crossing over is such that, for all but relatively close linkages, threepoint tests must be resorted to. T o facilitate the discovery of linkages, maize geneticists have built up sets of linkage testers, each having at least two workable linked genes whose loci are relatively far apart on the linkage map. With such testers there is little chance of failure to locate a new gene in its group at an early stage of the study. And yet only recently two supposedly independent linkage groups have been found, by the discovery of intermediate genes and the use of trisomics, to constitute a single group (BRINK 1931). As linkage groups become better known and linkage testers further developed, such difficulties should disappear. But in any material in which, as in maize, crossing over occurs with approximately equal frequency in both male and female, absolute certainty of the correctness of interpretation by purely genetic methods is scarcely to be expected. In view of such considerations, one should not be unduly disturbed by sporadic reports of more linkage groups than there are chromosomes in this or that organism. There are now available cytological helps to the determination of linkages. Such materials and methods will be noted later. It is enough at present to say that trisomics, reciprocal translocations, and deficiencies are among the most useful of cytogenetic tools. The available trisomics should prove particularly valuable in determining linkages, but as yet they have not functioned prominently other than as a means of proving with certainty the independence of groups previously assumed on genetic evidence to be independent. ASSOCIATION O F PARTICULAR LINKAGE GROUPS W I T H SPECIFIC

CHROMOSOMES

Diploid maize has ten pairs of chromosomes, corresponding to the ten known linkage groups. The importance of determining which linkage group is associated with a particular chromosome is obvious. Fortunately the chromosomes of maize differ in length and in other morphological features. The longest chromosome is more than twice the length of the shortest. The several chromosomes are given numbers in order of their length, the longest being I and the shortest X. The spindle-fiber attachment point, indicated by a constriction or a non-stainable area, is terminal in no case and in none is it strictly median. I n some cases, however, it approaches a median position so that, as in chromosome I, the longer arm is only a little longer than the shorter one. In others, V I I for instance, the longer arm is somewhat more than twice the length of the shorter one. Chromosome V I has a satellite which is attached to the major part of the

INTERNATIONAL CONGRESS O F GENETICS

147

chromosome by a thread of variable length and which is always associated with the nucleolus. In certain strains of maize the ninth chromosome has a conspicuous accumulation of stainable substance, appearing as a knob, at the end of the short arm. Similar conspicuous bodies occur, usually some distance from the end, in certain other chromosomes. Even in the case of chromosomes of nearly the same length it is possible, therefore, by means of these other morphological features, to identify each of the ten pairs. Ten linkage groups and ten morphologically different chromosomes having been identified, the next step was obviously to determine which linkage groups are associated with particular chromosomes. This has now been accomplished. For the most part this has been done by the use of trisomics (MCCLINTOCK 1931). 1 or 21-chromosome plants, were first obtained from a Trisomics, 2n naturally occurring triploid. It is now possible to obtain triploids and consequently trisomics almost at will. A recessive gene, asynaptic, with its locus in the first chromosome (BEADLE1930), which largely prevents pairing of all chromosomes at diakinesis, gives a high percentage of triploids when the partially sterile plants homozygous for it are crossed by normal diploids. Moreover, it has been shown that tetraploids can be produced in considerable numbers by heat treatment of diploid maize (RANDOLPH 1932). Crosses of tetraploids with diploids give triploids which in turn yield trisomics. With random distribution of the three chromosomes of a trisomic group, the resulting gametes of a duplex dominant simplex recessive trisomic, D D d, should exhibit a 5: 1 ratio of dominant to recessive. Similarly a simplex dominant duplex recessive, D d d, should give a ratio of 1 : l . I t has been shown, however, that of functional eggs only about one-third instead 1 and of functioning sperm less than two percent are of one-half are n 1. Approximate ratios of dominant to recessive expected in progenies n of duplex daminant and simplex dominant maize trisomics in F2 and in backcrosses to and by recessive diploids are as follows:

+

+

+

Duplex dominant D D d selfed - 12.5: 1 DDdXdd3.5:l ddXDDd2:l

Simplex dominant DddselfedDdd X ddd d X D d d -

1.7:l 1:1.25 1:2

All but one of these ratios ( D d d X d d = 1 :1.25) are in sharp contrast to the usual 3: 1 and 1 :1 ratios encountered in F, and backcrosses of heterozygous disomics. While the identification of trisomic individuals by means of chromosome

PROCEEDINGS O F T H E S I X T H

148

counts from seedling root tips presents no serious difficulty, it is often possible to use trisomics as genetic tools without making chromosome counts. 1 plants with chromosome V trisomic are easily distinguished Thus, 2n from diploids of the same cultures both as seedlings and through later stages of development. Trisomics which cannot be identified phenotypically can, not infrequently, be made use of without constant resort to chromosome counts. This can be done by keeping the trisomic cultures heterozygous for aleurone, endosperm, or seedling characters of the linkage group identified with the chromosome involved in the trisomic. Trisomic individuals can then be distinguished from their diploid sibs by the distorted ratios shown by their seeds or seedlings. Chromosome deficiencies and deletions, though perhaps not so generally useful as trisomics in identifying a given gene o r linkage group with a particular chromosome, not infrequently afford specific information a s to the region of a chromosome in which a particular gene has its locus. Thus, if a plant homozygous for a recessive gene or group of genes is crossed with X-rayed pollen from a plant homozygous for the dominant allelomorphs of those genes, a few of the resulting plants may exhibit the recessive character of the seed parent rather than the dominant character of the pollen parent. Some such plants are maternal haploids. Others are 2n -1 individuals due to the loss of an entire chromosome of the chromosome complement of the pollen parent. In other cases, cytological examination reveals the loss of the terminal part of one chromosome or even of a small piece from some other part of a chromosome. The inference is clear in such cases that the observed deficiency or deletion has resulted in the loss of a dominant gene carried by the pollen parent and that, therefore, the locus of that gene must have been in the lost part 1931b) it has been of the chromosome. By such studies (MCCLINTOCK shown that the gene lg is very near the end of the short arm of chromosome 11, gene a, near the end of the long arm of chromosome 111, P1 probably near the middle of the long arm of chromosome VI, and Y between Pz and the satellite, and that R is in the long arm of chromosome X. Very unpublished) that j lies near the recently it has been shown (MCCLINTOCK end of the long arm of chromosome VIII. This is the only case so far of the placement of a gene in a particular linkage group by other than pureIy genetic methods. Male sterile-8 (BEADLEunpublished) belongs in the same group. Only one other gene, Ch, has been placed even tentatively in this group (ANDERSON and EMERSON1931). This was done by the process of elimination; Ch shows no linkage with appropriate tester genes in any of the other nine groups. Very recently, however, it has been shown by use

+

INTERNATIONAL CONGRESS O F GENETICS

149

of a trisomic involving chromosome V I I I (BURNHAMunpublished) that C;1 is not in that chromosome. This illustrates well the difficulty of locating genes with entire assurance by genetic methods alone. Reciprocal translocations, as well as trisomics and deficiencies, can be used to identify the chromosome carrying the genes of a particular linkage group. By this means it has been shown (BURNHAM1930) that genes of the P-bl. linkage group are borne by chromosome I. By means of a somewhat unique combination of a trisomic and a reciprocal translocation in1931a) it has been found volving chromosomes V I I I and I X (MCCLINTOCK not only that genes of the C-wx linkage group are borne by chromosome I X but also that C is near the end of the short arm and that the genes of the entire linkage map as now known involve little more than the short arm of that chromoson~e. The simple reciprocal translocations, those involving only two nonhoinologous chromosomes, so far as they have been identified cytologically, are associated with about 50 percent of pollen and of egg sterility. Such semisterility can be used in linkage studies much as it could be if it occurred as a gene mutation, except that it shows linkage with genes of two linkage groups (BRINKand B U R N H A M 1929). The point of breaking and reattachment of the chron~oson~es involved in a reciprocal translocation can be determined cytologically with respect to knobs, fiber attachment points, etc. Senlisterility resulting from reciprocal translocations can be used with two genes in three-point genetic tests as an aid in determining the approximate loci of those genes. The absence of interference to double crossing over between genes on opposite sides of the translocation (RHOADES1931) is a further aid in fixing the approximate location of the genes included with semisterility in three-point tests. MUTATIONS I N M A I Z E

Little need be said about mutations in maize. Most of the genes studied by geneticists are assumed to have arisen earlier as point mutations. Relatively few of the genes with which we deal are known to have occurred in controlled pedigreed materials. Irradiation has induced, in addition to deficiencies, translocations and other chromosome abnormalities, what appear to be typical gene mutations. Whether or not they are minute deletions is not as yet known. But do we know that naturally occurring point mutations are never such deletions? The situation with respect to variegated pericarp of maize has been interpreted by the writer as a case of somatic mutation. The variegated type changes somatically to self color with high or low frequency depending

150

PROCEEDINGS O F THE SIXTH

on the strain involved. Self color then behaves as a simple dominant to variegation. Reverse changes from self color to recessive variegation occur but with less frequency in most strains than the change from variegation to self color. The several strains of variegated maize are conceived to differ fundamentally only iq respect t o the frequency of change to self color. I n other words, they are considered as differing only in mutation rate. Crosses between variegated strains of very low mutation rate and stable colorless races usually exhibit a greatly increased rate of mutation. In F, of such crosses high and low mutability are linked with the gene for pericarp color, P. Colorless segregates from a heterozygous variegation culture of low mutability, when crossed with homozygous variegates of low mutability, do not increase the rate of mutation in F1. The writer is quite unable to interpret these facts on the basis of chromosome non-disjunction or the elimination of chromosome fragments in somatic mitosis. SEX EXPRESSION I N MAIZE

Normal maize plants are usually monoecious, the terminal inflorescence bearing staminate and the lateral inflorescence bearing pistillate flowers. Numerous genes, mostly recessive, however, influence the expression of sex. Thus, in certain dwarf and semi-dwarf types stamens regularly develop in the flowers of the lateral inflorescence, resulting in an andro-monoecious condition. A few types are, barring infrequent sex reversal, wholly staminate flowered and others are wholly pistillate flowered. Of the former are barren stalk and silkless and of the latter tassel seed-1, -2, and -3. A dioecious strain has been produced by JONES, involving tassel seed-2 and silkless. The writer has similar strains involving barren stalk-1 and tassel seed-2 and -3. I n view of the fact that of some animals-Drosophila, man, and certain fishes-the male is the heterogametic sex, while of others-birds, moths, and certain fishes-the female is heterogametic, the status of dioecious maize is of interest. One of the writer's dioecious strains involves barren stalk-1 and tassel seed-2, both recessives. The male is bal bal Te2tsZ,heterogametic, while the female is bol bal te2 t82, homogametic. The other strain involves barren stalk-1 with tassel seed-3, the latter a dominant. Here the male is bal bal tas tea, homogametic, and the female is bal bal T e s tRQ,heterogametic. I t is perhaps 'worth noting that in these synthetic dioecious strains of maize, the two sexes are differentiated by a single pair of genetic factors whose loci on the chromosomes are known.

INTERNATIONAL CONGRESS O F GENETICS

151

MISCELLANEOUS OBSERVATIONS

Complex translocations. Reference was made earlier to simple reciprocal translocations involving two non-homologous chromosomes. These form rings of four chromosomes at diakinesis, as in other materials. By combining different semisterile types, rings of six or more chromosomes or two rings of four each result. Since maize chromosomes are differentiated morphologically and since all the ten pairs of chromosomes carry workable genes which can be used as markers, maize affords excellent material for checking similar situations previously noted in Oenothera and Datura. Crossing over between chromatids. Maize trisomics have been found useful in determining whether crossing over in plants ever occurs after synaptic chromosomes have split to form four chromatids. RHOADES(1932 and unpublished) has shown by this means that crossing over does occur in the double-straid stage. Teosinte-maize hybrids. It has been shown (EMERSON and BEADLE1932) that hybrids of maize with annual teosinte exhibit in general normal crossing over. The only exception to this so far discovered is the entire o r almost entire absence of crossing over between chromosome IX, the C-sh-z~~ chromosome, of maize and its homolog in Florida and in Durango teosinte. Normal crossing over occurs between these chromosomes of maize and Chalco teosinte. I t has been suggested but not proved that this failure of crossing over may be due to inversions in the region of the ninth chromosome in question. ANDERSON, E. G., and EMERSON, R. A., 1931 Inheritance and linkage relations of chocolate pericarp in maize. Amer. Nat. 65: 253-257. BEADLE,G. W., 1930 Genetical and cytological studies of Mendelian asynapsis in Zea mays. Cornell Univ. Agric. Expt. Sta. Mem. 129: 1-23. BRINK,R. A., 1931 Heritable characters in maize. XL. Ragged, a dominant character, linked with A*, T., and D,. J. Hered. 22: 155-161. C. R., 1929 Inheritance of semisterility in maize. Amer. Nat. BRINK,R. A., and BURNHAM, 63: 301-316. BURNHAM, C . R., 1930 Genetical and cytological studies of semisterility and related phenomena in maize. Proc. Nat. Acad. Sci. Washington 16: 269-277. EMERSON, R. A., and ANDERSON, E. G., 1932 The A series of allelomorphs in relation to pigmentation in maize. Genetics 17: 503-509. R. A., and BEADLE, G. W., 1932 Studies of Euchlaena and its hybrids with Zea. EMERSON, 11. Crossing over between the chromosomes of Euchlaena and those of Zea. Z. indukt. Abstamm.-u. VererbLehre 62: 305-315. LINDSTROM, E. W., 1931 Genetic tests for linkage between row-number genes and certain qualitative genes in maize. Iowa Agric. Expt. Sta. Res. Bull. 142: 249-288. MCCLINTOCK, BARBARA, 1931a The order of the genes C, S, and W , in Zea mays with refer-

152

PROCEEDINGS O F T H E S I X T H

ence to a cytologically known paint in the chromosome. Proc. Nat. Acad. Sci. Washington 17: 485-497. 1931b Cytological observations of deficiencies involving known genes, translocations and an inversion in Zea mays. Missouri Agric. Expt. Sta. Res. Bull. 163: 1-30. MCCLINTOCK, BARBARA, and HILL, HENRYE., 1931 T h e cytological identification of the chromosome associated with the R-G linkage group in Zea mays. Genetics 16: 175190. RANDOLPH, L. F., 1932 Some effects of high temperature on polyploidy and other variations in maize. Proc. Nat. Acad. Sci. Washington 18: 222-229. RHOADES, MARCUSM., 1931 Linkage values in an interchange complex in Zea. Proc. Xat. Acad. Sci. Washington 17: 694-698. 1932 The genetic demonstration of double strand crossing over in Zea nzays. Proc. Nat. Acad. Sci. Washington 18: 481-484. SPRAGUE, GEORGE F., 1932 The inheritance of colored scutellums in maize. U. S. Dept. Agric. Tech. Bull. 292: 1-33. STADLER, L. J., 1929 Chromosome number and the mutation rate in Avena and Triticum. Proc. Nat. Acad. Sci. Washington 15: 876-881.

INTERNATIONAL CONGRESS O F GENETICS

153

T H E CON JUGATION O F T H E CHROMOSOMES Ilarry Federley, T h e Uuiversity, Helsingfors, Finland

The conjugation of the chromosomes is a universal process in the whole of organic nature. It occurs not only among all multicellular organisms both in the animal and the vegetable kingdom but is also proved to exist in a great number of unicellular organisms. It is not only an indispensable condition for all bisexual reproduction but also of great importance for what we, in a vague term, call variability, and for all kinds of evolution, for in the meiosis the combination of different genes takes place, and thereby through various aberrations of the normal process of conjugation further favorable conditions for the production of mutations are created. Meanwhile, our knowledge of conjugation and its real nature by no means stands in proportion to its importance and significance. REUTER in his profound and scholarly work says that the cause of conjugation lies in the manner of reproduction of the diploids. Yet neither he nor any one else can offer us an explanation of the power which sets free conjugation. W e speak about the affinity of the homologous chromosomes without being able further to define what we mean by such an affinity, whether we think of it as physical or chemical, o r as an unknown form of biological affinity. And why are the chromoson~esin the soniatic cells independent of this mystic affinity that does not set free the pairing of the homologous chromosomes until meiosis? These are the questions that are still awaiting a satisfactory solution. That the affinity between the paternal and the maternal homologues has nothing to do with the sexuality is evident, although this assumption has been propounded by certain authors. For only the Y (the W) chromosome is bound to one sex, while all autosomes and the X (the Z ) chromosomes sometimes happen to exist in individuals of one, sometimes in individuals of the other sex. That the chromosomes a t conjugation play an active and to a certain extent self existent part, independent of each other, seems t o be fully proved by the fact that in certain cases the autosomes conjugate in the first maturation division, while the allosomes and also the microsomes do so only in the interkinesis. That the structure of the chromosomes must be of decisive importance in this case is evident, just as it plays a very important part in the conjugation. But on the other hand we have also clear evidence of the fact that the general physiological conditions in the cell may determine the process of conjugation. Critical experiments have further shown how conjugation depends on the surrounding influences. Finally,

154

PROCEEDINGS OF THE SIXTH

experimental research has during the last few years discovered special genes which determine the conjugation of the chromosomes. Thus we see that this important process mainly depends on the structure of the chromosomes but also on the genotype of the individual and, finally, on the environment in the cell itself and the surroundings in general. As is the case when a complicated biological process is being investigated, it is not only from a careful examination of the normal course of the process that we can expect to arrive at an explanation of any problem. It is rather the deviations from the normal process which often give us a deeper insight into the real nature of the process. This is true in the case of the problem we are dealing with at present. It is chiefly on four different principles that we try to solve the mystery which still obscures the conjugation of the chromosomes: (1) through a thorough study of the normal meiosis in diploid forms with large chromosomes which can be observed during all the different phases of the conjugation; (2) through observations on polyploid and heteroploid forms; ( 3 ) through systematic investigations of hybrids and the circumstances of conjugation of the different species chromosomes; (4) through research into the special genes, determining the conjugation and the outside factors influencing it. Among the many scientists who have made the normal meiosis the submay be mentioned in ject of their special study, BELLINGand DARLINGTON the first place. They have added to our knowledge on the subject very essentially ; our insight into the origin and importance of the chiasmata has made a step forward, and the phenomenon of crossing over has greatly gained in clarity while the synapsis or syndesis has, on the other hand, been pushed a little into the background. DARLINGTON is of the opinion that the meiosis is an abnormal mitosis in which prophase contraction has anticipated the division of each chromosome into two threads. As at mitosis in this stage, the split halves of chromosomes are constantly associated side by side in pairs; this relationship is restored in meiosis by the association of whole, undivided chromosomes in pairs. DARLINGTON questions the existence of any kind of affinity between the chromosomes at metaphase. The four chromatids are kept together by chiasmas, and these prevent the separation of the chromosomes at diakinesis and metaphase, for, according to DARLINGTON, the repelling powers during these stages, especially during diakinesis, are stronger than the attracting powers. I have no doubt that DARLINGTON is right, and I quite acknowledge his great merits; yet I cannot agree with the theory that the question of the

INTERNATIONAL CONGRESS OF GENETICS

155

mystic affinity is reduced to the question of whether chiasmas are produced. For a condition that the chiasmas should be produced is that the homologous chromosomes approach each other or conjugate, that is, that there exists an affinity between the homologues. What powers are in action when the chromosomes approach each other is still a riddle. That the genic constitution of the chromosomes, or, to speak cytologically, of the specific chromomeres of the chromosomes, is in this case of primary consequence, can be considered sufficiently proved. Yet the structure of the chromosomes does not reign alone, and under certain circumstances other factors succeed in gaining supremacy over the chromosomes and prevent the conjugation, which can also be considered to be experimentally proved. By the study of meiosis in polyploids scientists have succeeded in getting clear evidence of the fact that the linear structure of the chromosomes is of the greatest importance for the conjugation. In certain triploid forms, for instance, Canna, Tulipa, Primula, Hyacinthus and others, trivalent chromosomes have been found besides the bivalent ones while among tetraploid forms quadrivalent chron~osomesoccur, an evidence of the fact that the homologues attract each other regardless of their being either 2, 3, or 4 in number. A more detailed study of the manner of association of these multivalent chroinosomes has given us evidence of the fact that the chromosomes are built up of small units, the chromomeres, which are arranged in a quite definite manner. Thus there never exists among the primaries of trisomics of Datura an association in a triangular form, because three congruent chromosomes can not unite in a triangle in such a way that two homologous ends meet. In the secondaries of the trisomics among which an interchange of one half has taken place in one of the chromosomes, this changed chromosome can be inserted between the ends of the two unchanged chromosomes, and thus a triangle arises. In this connection I also want to speak of the well-known cases in Drosophila in which an inversion of a section in one of the homologues has taken place. Such a n inversion presumably prevents the normal synapsis, which makes it likely that an absolute homology between the chromomeres is necessary for a normal conjugation. Through the inversion the formation of chiasmata is undoubtedly prevented, which makes the pairing less stable. with translocations in certain The experiments of MULLERand PAINTER chromosomes of Drosophila caused by radiation speak in favor of this and STURTEVANT, through a series theory. Only lately have DOBZHANSKY

156

PROCEEDINGS OF THE SIXTH

of systematic crosses between normal individuals of Drosophila and those with translocations in the second and third chromosomes, clearly proved that chiasmata are caused between homologous chromomeres and fail between chromosomes of which one has a series of chromomeres disturbed by translocation. The few haploid organisms of which the meiosis is known also prove the correctness of the theory that the structure of the chromosomes determines the affinity. Among haploids there is, as a rule, no synapsis, and where a tendency to a synapsis is sometimes observed it proves t o be defective. Thus, where homologous chromosomes do not appear in the nucleus, there is not a normal tendency to conjugation. W e now pass on to the third group of experiments, comprising the species hybrids, and I shall, in connection with the second group, the polyploids spoken of above, begin my explanation by pointing out some extremely interesting observations concerning crossings of polyploid species, especially the autosyndesis appearing in such crossings. Why chromosonles which never conjugate with each other among the parent species should suddenly be seized by a desire to enter into union with each other when they are brought together with another set of chromosomes, is more than mysterious; every autosyndesis was, to begin with, even considered impossible, and, when reported, was thought t o be the result of erroneous observations. But the instances of autosyndesis are now so numerous, and behind the observations stand so many conscientious scientists, th?t we need not feel any doubt of their correctness. Attempts have been made to explain the fact in such a way, that if, for instance, among a certain species the two chromosome groups A, A, A, A,, where A1 and A, never conjugate with each other, are brought together through crossing with the groups B B C C of another species, and these chromosomes are not related to A, and A;, the last-named would, of necessity, conjugate the one with the other. The said explanation would thus presuppose a tendency for conjugation, innate in the chromosomes, which must be satisfied a t any price. Yet, apart from the fact that this explanation takes into account a real mystic affinity, we do not advance far with this theory. I t builds exclusively on the relationship of the chromosomes, that is, on a more or less strongly characteristic homology, and considers it to be conclusive for the conjugation. In case this should be a correct supposition, conjugation ought to take place without disturbance among the autotetraploids. But this is in no way the case. Many tetraploids show a defective pairing. In this case recourse has been had to the expedient of declaring the forms to be allotetraploids

INTERNATIONAL CONGRESS O F GENETICS

157

rather than autotetraploids. Yet this expedient does not lessen the difficulty. For if it were correct, then, at least among forms which are proved to be amphidiploids, all the chromosomes ought to pair regularly, and they do not. On the contrary certain irregularities are the rule. For example, it is only recently that LEVITSKYand BENETZKAJA ascertained that, among the wheat-rye amphidiploids with twice 21 wheat and twice 7 rye chromosomes, the number of bivalents in the same individual may vary from 25 to 28. Thus we have here 1 to 3 pairs of absolutely homologous chromosomes that do not conjugate with each other. The reason for this must be looked for in other circumstances than in the incompatibility between the chromosomes. Among many other amphidiploids similar irregularities seem to occur, although they have not hitherto been investigated very carefully. The capricious character of the autosyndesis is perhaps most clearly illustrated by a comparison of the condition of the chron~osomesa t different crossings with the same species. In this respect the crossings made with different species of Digitalis are specially enlightening, as GERTRAUD HAASEEESSELrightly points out. A t a crossing between the closely related Digitalis species, lz~tcawith 24 chromosomes and nzicrantha with 48 (by some taxonomists taken to be forms), it became apparent that the hybrid had 36 bivalents in meiosis and that the hybridization set free autosyndesis in both lutea and ~ ~ i i c r a n t hchromosomes, a but the hybrid is, in spite of the absolutely normal pairing of the chromosomes, absolutely sterile. In the hybrid between lz~teaand pztrp~~rca there is no pairing at all between the chromosomes, and there are only univalents in the diakinesis. This hybrid consecluently belongs to the Pygaera type according to the Tackholm division. And finally, at the crossing of lz~teawith lanata the hybrid shows a variable number of bivalents and must thus be reckoned to belong to the Boreale type. T o summarize the above: Crossing with one species sets free the autosyndesis between the lutea chromosomes, in crossing with another species all the lutea chromosomes remain unpaired, and with a third species a variable number of them begin pairing with the chromosomes of this third species. But also, in one and the same crossing, sister individuals can show quite different conjugation conditions. Thus HAASE-BESSEL describes two hybrids between Digitalis canariensis and grandiflora, of which one hybrid did not show any pairing whatever of chromosomes, while the other presented a variable number of bivalents.

158

PROCEEDINGS O F THE SIXTH

The apparently great capriciousness in the matter of the chromosome conjugation has struck every one who has made the hybrid meiosis the subject of research. Hitherto, however, there have been comparatively few systematically carried out investigations of hybrids between different related species or between a number of individuals of the same crossing, and above all our knowledge restricts itself to the pollen development and spermatogenesis, while the oogenesis, technically so hard to study, has been thoroughly investigated only in extremely rare cases. Having myself, during a number of years, been occupied with investigations of hybrids of Lepidoptera, I take the liberty of here recording some partly unpublished results. I will begin by referring to the table which illustrates the conditions of conjugation among a number of hybrids between closely related species of Sphingidae, all with the haploid chromosome number 29. From the table it appears that meiosis with the hybrid Metopsilus porcellusx Chaerocampa elpenor is practically normal, while the hybrid Deilephila galiiX D. euplzorbiae shows a meiosis with comparatively small deviations from the normal. The former of these hybrids is entirely fertile, and in the backcrossing with the parent species the resulting hybrids show a distinct segregation, both in the larval, pupal and imago stages. The latter hybrid is also fertile in a high degree. I regret to say that I have been able to state segregation only in the larval stage. Yet this reciprocal hybrid has been investigated by BYTINSKI-SALZ and has shown itself to be segregate in all stages. The position is quite a different one with regard to the two hybrids D. euphorbiaeXCh. elpenor and Ch. elpenorXD. galii. For both we state a minimal affinity between the chromosomes of the parents, and we find that the spermatocytes are divided among a great number of different conjugation classes. I t has not been in my power to make backcrossing between these hybrids and their parents. Only the males develop into imagoes, while the females, in consequence of a sub-lethal combination of the sex chromosomes, are not able to pass through the metamorphosis and therefore die as pupae. The hybrid males, on the other hand, develop so late in the autumn that there are no longer any females of the parent species left. But I believe that, without being too audacious, I dare assert that neither of these hybrid males is fertile. The table meanwhile teaches us something else also: the conjugation of the chromosomes in the hybrids between the same parent species can turn out quite differently, and even the sibs in the same cross can show great dissimilarities in this respect. I refer only to the two investigated individuals of

INTERNATIONAL CONGRESS O F GENETICS 4

w

s

q

2

k

8

3

w

s

"Y 00

..

..

..

..

..

..

..

..

.. .. .. .. ..

.. .. .. .. ..

.. .. .. .. ..

.. .. .. .. ..

.. .. .. .. ..

.. .. .. .. ..

.. . .. .. ..

.. . * .. m ..

..

..

..

..

..

..

.

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

..

..

..

.. .. ..

..

.

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

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

.. ..

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

.. . .

.. .. .. .. ..

m *

..

. . . . . . . .. .. .. .. .. .. m . . , , . . .. * . . . . . . . .. . . . . , + . . .. N .. .. .. .. .. ..

m

..

..

8 h "Y

w

"Y "Y) 7

5 "Y 0

-

Y) N

"Y

W h .*

* * *w 00

h

'2

* i

1

2

z

I G - - a d" I , J ,

3

4

$

m

&

z 3 m

.. ..

-

m N m

::

N

*

N m N 00

.

-

rO

.

.

m

.

-

N

..

:

h

z

.

h

)

N

m

*

N

e

d

.

-

.

I

d

m

.

..

..

e

.

..

..

.. .. ..

..

h Y)

a

..

.. .. .. ..

.

.

..

..

.

.

.

.

..

..

N

:

:

:

:

:

..

..

.. ..

.. .. . . .

.. ..

-

z

m m

x

z

N

*

v

N

..

I

o

d

h

.. ..

P

o

m

.. .

.

m d

.

.

s

,

~

: +

a

g V

:

:

:

:

: )

1

:

:

d

*

h N

:

:

:

:

:

:

:

salL~o~.emlads 30 IaqmnN

8,

S

:

160

PROCEEDINGS O F THE SIXTH

the family 16 of the cross yaliiXcirpl~orbiac,of which one shows a maximum nutnber of cells with number 29, characteristic of the parents, while the other has its maximum at 28. The number of examined cells is in both cases so great that there cannot be any doubt of the correctness of the observations. In the last named individual a secondary affinity has thus made two chronlosomes associate into one. A similar case has been observed by me earlier at a crossing between the German and the Finnish form of nicranura &nula, which both have the haploid number 21, while the hybrid has 20. Such a very capricious conjugation of the chron~osomesmay be observed in alnlost all hybrids of which the greatest number of individuals is suhjected to a careful investigation; this clearly proves that it is not only the chromosomal structure which determines the conjugation. From the table one might perhaps be tempted to draw the conclusion that the conditions of conjugation of the hybrids might be used for a taxonomic purpose in order to fix the closer or remote relationship of the parent species. The greater the number of conjugating chromosomes, the closer the relationship. Yet this would be a hasty conclusion, I am sure. Of course, in the present case the supposition that elpenor is more closely related to porcellz~s than to euphorbiae and most remotely related to galii would agree with the opinion of the taxonomists. And it would also be justifiable to consider that yalii and ez~phorbiaeare more closely related to each other than to elpeltor. Other instances also supporting this theory could be mentioned both from the botanical and zoological fields. But the conjugation is not at all a reliable taxonomic criterion. As an instance of its untrustworthiness I desire only to quote an example from the same family as the above named. The East Asiatic form Planus of the European Stllcrintkz~socellata, however, which are considered by all entomologists as being very closely related to each other, and which both possess 28 chromosomes, show, in the hybrid, a very low degree of affinity between the chromosomes. From a iable such as the one dealt with here, we can draw a conclusion concerning the fertility and the sterility of the hybrids. The fewer the different conjugation classes are which a hybrid comprises, the greater its fertility, and with a growing breadth of variation its sterility augments. It does not signify in this case whether the top of the curves corresponds with the haploid or the diploid number. The high and narrow curves are always to be found at these numbers; only in some backcrosses do they lie in their immediate rleighborhood. I should like to conclude my discourse on the conditions of conjugation

INTERNATIONAL CONGRESS O F GENETICS

161

of the species hybrids by giving a record of one case which will, perhaps better than anything else, show you how extremely difficult it is t o solve the problem which confronts us. I have in view the altogether dissimilar conditions which in one case have proved themselves to exist between the male and female of the same hybrid. I mean the reciprocal hybrids between Pygaera pigra, n=23, and curtz~la,n=29, which in the male sex show no conjugation at all, or a conjugation between very few chromosomes, while in the female sex all the 23 pigra chromosomes conjugate normally with 23 curtula chromosomes, and of the six remaining curtula chromosomes the greatest number are eliminated. Here we have the most striking evidence in favor of the hypothesis that the conditions of conjugation cannot be used in the service of taxonomy, but we also arrive a t the conviction that physiological conditions in the cell, and especially in the nucleus, are able in certain cases to assume sovereignty over the chromosomal structure. In the case in question the autosomes are absolutely the same in both sexes of the hybrid, and only the sex chromosomes are different in the reciprocal females. But as the hybrids in the reciprocal crosses act alike, one can hardly ascribe the effect of the sex chromosomes to the quite dissimilar conditions of conjugation in the two sexes. I t only remains t o assume that the cause why a seemingly normal conjugation is brought about under spermatogenesis, while it fails altogether in the oogenesis, lies in the dissimilar conditions during these processes. Yet it is true that they are chiefly regulated by the sex chromosomes. Of course, the parallel with the dissimilarity in regard to crossing over in both sexes of Drosophila imposes itself. The reason why crossing over appears only in the female and not in the male has been taken to lie in the fact that chiasmata are formed only in the oogenesis, not in the spermatogenesis. I t is natural to assume something similar also concerning the Pygaera hybrids. I regret to say that it is impossible here to examine chiasmata, because it is difficult to get sight of them in the Lepidoptera. Yet there are circumstances which speak for the opinion maintained by DARLINGTON that chiasmata a t diakinesis and metaphase play a very important part. In certain individuals of the hybrid the existence of a manifestly short syndesis, also in spermatogenesis, can be discerned, which proves that a certain affinity yet exists, even if it is weak and of short duration, and that consequently it can not be traced at diakinesis and metaphase. Altogether, many observations made by the way indicate that the meiosis can turn out quite differently in the pollen mother cells from that in the embryo sac mother cells and also differs in the spermatocytes from that in

162

PROCEEDINGS O F THE SIXTH

the oocytes, which fact ought to stimulate the cytologists to more careful investigations of the arduous but evidently satisfactory oogenesis. There only remain to say some words regarding the last group of investigations concerning the outward and inward factors which have proved tliemselves to be regulators of conjugation. After having spoken about moths, I should like here to state that during the normal spermatogenesis the metabolism, ruling in the testis, determines the conjugation. The chromosomes of those spermatocytes which first enter into the meiosis all pass through a regular conjugation, while the spermatocytes formed by the end of the spermatogenesis do not show any pairing of the chromosomes whatever but only univalents which divide irregularly. The spermatocytes give rise to the apyrene sperms which do not possess the experiment to inject haemopower to fertilize the eggs. Through MACHIDAS' lymph from older larvae of Bombyx into younger ones and thereby to prove that the formation of apyrene sperms is accelerated, it is clearly shown that it is the quality of the haemolymph which decides whether or not the conjugation is brought about. The influence of temperature on the conjugation of the moths has, on the who has shown that in Lymanother hand, been investigated by KOSMINSKY tria dispar the bivalent chromosomes decrease with a rising temperature, so that in the highest temperatures which can be borne only diploid gametes are produced. The influence of temperature on the meiosis has been very thoroughly studied by the botanists. Thus, SAKAMURA and STOWhave, through influence of heat on bulbs of Gagea lutea during the time of the meiosis, succeeded in calling forth polyploid pollen cells and also pollen cells with aberrant chromosome numbers. These have shown themselves to possess the ability of germinating, and can thus give rise to new forms. BELLINGhas found that individuals of Uvularia that a t night during winter were exposed to too much cold in a hothouse showed a number of abnormalities during the meiosis, in a great measure depending on abnormal conjugations. The potato seems to be especially sensitive to high temperature. Conjugation takes place normally only a t 20°C and leads to the formation of pollen cells with 24 chromosomes. In rising temperature the power of conjugation decreases and according to STOWis altogether suspended at 25 to 30°C, in consequence whereof diploid pollen cells with 48 chromosomes are formed. These are sterile. According to STOWthe lower temperatures are more easily borne. At 8 to 12OC the conjugation proceeds normally, and 24 c h r o m o s o ~ ~ ~can e s be counted but show a strong disposition to a secondary pairing so that, in certain cases, only 12 chromosomes are to be found.

I N T E R N A T I O N A L CONGRESS OF G E N E T I C S

163

HEILBORN has made certain kinds of apples grown in the north the subject of a careful investigation with regard to the influence of temperature on the process of conjugation and found that an already comparatively insignificant rise in the temperature above the optimum prevents conjugation, so that a larger or smaller number of univalent chromosomes can be observed. Different sorts show very different susceptibility. Among certain species as low a temperature as 20°C is injurious, while other kinds can bear without injury as much as 35°C. That the age of the flower can influence conjugation is proved by ROSENBERG'S observations of Hieracium laevigutunz and lacerum, among which a formation of gemini does not generally occur. In old cells, however, ROSENBERG has succeeded in coming across a few gemini. I should like further to state that even attacks of parasites can influence the normal process of meiosis. According to experiments by KOSTOFFand by KOSTOFFtogether with KENDALLin Lyciuln halmifolium and Datura ferox, many kinds of irregularities occur in the meioses because of attacks by mites. And finally, some few words regarding specific genes which regulate the meiosis and specially influence the conjugation. BEADLEhas described a couple of such genes in Zea. Among these, one gene called by BEADLE"asynapsis gene" is of interest in this connection. It is recessive, and its effect is characterized by partial or complete failure of synapsis during the prophase of the first division in the microsporocytes. This results in failure of reduction and production of diploid spores. In Drosophila, GOWENhas discovered a gene which quite prevents crossing over and thus in some way or other must change the conjugation, probably through preventing the formation of chiasmata. This is also indicated by the development of all kinds of chromosomal mutations in such families. They must apparently be ascribed to the fact that synapsis failed partly or completely. L. A. SAPEHINhas found in hybrids between Triticum durunz and vulgare specific genes which determine normal meiosis; he calls them organisators, and their recessive alleles he calls disorganisators. These disorganisators cause all kinds of abnormalities in the chromosomes of the microsporocytes and SAPEHINthinks he can declare that they segregate, although he has not been able to decide on the special type of segregation. Among these disorganisators there also exists one which prevents the segregation of the chromosomes both in the first and in the second maturation seems to be condivisions and thus gives rise t o diploid pollen cells. SAPEIXIN vinced that there are a number of such gene organisators that regulate the

161

PROCEEDINGS OF THE SIXTH

whole meiosis, and he is sceptical about the influence of the plasma on the nleiosis. As appears from what I have said, I have not been able to offer any explanation of the phenomenon we call conjugation of the chromosomes, nor, I hope, have any of my hearers expected one. What I wish to convey t o you is the extremely complicated nature of this universal process and its dependence on a number of inward and outward factors as yet too little subjected to research. And if my discourse succeeds in inciting scientists to take a deeper interest in the subject, and if it urges cytologists and geneticists together to undertake thorough investigations in this field of knowledge, I shall have gained my object.

INTERNATIONAL CONGRESS O F GENETICS

165

T H E EVOLUTIONAKY MODIFICATION OF GENETIC PHENOMENA R . A. Fisher, Rothamsfed Ex@erimental Station, Harpenden, England

The title chosen for our discussion is "Contributions of genetics to the theory of evolution," and that these contributions are of two kinds, somewhat sharply contrasted, is well illustrated by comparing HALDANE'S subject, "Can evolution be explained in terms of present known genetical causes?" with the heading under which I chose to speak, "The evolutionary modification of genetic phenomena." My own address might equally well have been entitled, "Can genetical phenomena be explained in terms of known evolutionary causes?" The one approach, as you perceive, is analytic and deductive. Genetic studies are regarded as revealing the mechanism connecting cause and effect, from a knowledge of which the workings of the machine can be deduced and the course of evolutionary change inferred. The other approach is inductive and statistical; genetics supplies the facts as to living things as they now are, facts which, like the living things in which they occur, have an evolutionary history and may be capable of an evolutionary explanation, facts which are not immutable laws of the workings of things but which might have been different had evolutionary history taken a different course. I can only discuss a small portion of the subject. Genetic phenomena concerning the chromosomal organization, such as the male haploidy of the social hymenoptera, as SNELLhas suggested in an illuminating paper recently published in the American Natusalist, may have an adaptive significance; and I think we may look forward with confidence, as the facts become better and more systematically known, to discovering the significance of such phenomena as male linkage in Drosophila, and of the marvelously intricate chromosomal mechanism which is being unraveled by METZin Sciara. The only two phenomena I can attempt to touch upon are those of dominance put before you a selection of the facts and linkage; and on these I can (many of which I owe to the kindness during the last few days of other members of this congress) which seem to me to supply a good deal of light and guidance in forming an interpretation of the general body of genetic facts. As I am a mathematician by trade perhaps I should explain that I shall use no mathematics, partly because I recognize that the first duty of a mathematician, rather like that of a lion tamer, is to keep his mathematics in their place, but chiefly because I think that mathematics, though well fitted to elucidate detailed points of special intricacy, are after all only a special means of carrying out reasoning processes common to all scientific work,

166

PROCEEDINGS O F T H E SIXTH

and are out of place in a theory covering a wide range of disparate phenomena. I believe that no one who is familiar, either with mathematical advances in other fields, or with the range of special biological~conditionsto be considered, would ever conceive that everything could be summed up in a single mathematical formula, however complex. If I am tempted for brevity to express myself in generalizations, it is not because I think exceptions are unimportant. One of the things about them that is important is that they are exceptions ; and it seems to me that it is only by obtaining an understanding of the body of cases which constitute the rule that we can usefully hope to investigate the special causes which have produced an exception. Dominance modification is a special case of the general fact that the expression or manifestation of a genetic factor, or gene substitution, is conditioned by the genotype in which the substitution is made. Phrases such as epistatic factors, duplicate factors, complenientary factors, etc., showed an early recognition of some special cases of this general fact, which has I believe impressed itself more and more on the minds of geneticists, just in proportion as their work has become more detailed and more thorough. If the interaction of factors affects principally the heterozygote, then the relationship which we call dominance will be affected. For example, since all knowledge naturally starts with Drosophila, the dominant mutant Gull, found by MOHRin the second chromosome, is, like so many dominants, lethal when homozygous. The recessive dachsous suppresses Gull in the heterozygote, while the homozygote remains lethal. In the presence of dachsous, therefore, Gull is a recessive lethal, although without dachsous it is quite an ordinary dominant. Drosophilists could probably supply more than one parallel. Here is one from poultry. Frizzle is a dominant which curls the feathers out in a peculiar manner. The homozygote is viable though delicate through losing much of its plumage. Both LANDAUER and HUTThave a recessive mutant which largely suppresses the frizzle effect in the heterozygote, with only occasional or little effect in the homozygote. The suppressing factor by itself seems to be undetectable except by its effect in shifting frizzle some way toward recessiveness. Now dachsous is presumably injurious to survival in wild conditions, and the same may be true of the modifier of frizzle, though there is no evidence for this; but it is clear that, whatever effect the modifier may exert by itself, yet in a population descended from an ancestry containing any perceptible proportion of heterozygous frizzles, its interaction with frizzle would have given it an increased frequency of survival and have tended to make it spread through the population. The magnitude of this tendency depends chiefly upon the frequency of heterozygous frizzles

INTERNATIONAL CONGRESS O F GENETICS

167

in the ancestry, and this in turn must have depended on the mutation rate by which the frizzle gene was produced and on the viability of the heterozygous frizzles. It is easy to see that the viability of the homozygous frizzle and the effect of the modifier upon it are unimportant items of the calculation. As a typical case, one may take a mutation rate of one in a million in each generation and see how the proportion of heterozygous frizzles in the ancestry depends upon their viability compared to the normal non-frizzle birds. For 99 percent viability the proportion is about one in five thousand, for 90 percent about one in sixty thousand, for 50 percent about one in seven hundred and fifty thousand. The point of this simple calculation is to show that the rate of modification depends very greatly on the level of viability already attained. A seriously handicapped heterozygote will be modified very little indeed, even in periods of time ample to bring a more viable heterozygote up to complete normality. The course of the evolutionary progress of the heterozygote will be a rising curve-the later stages of its modification being much more rapid than the earlier. When a heterozygote has been modified up to complete normality, the factor appears as a recessive; if the homozygote happens to be lethal all progress would seem to have ceased, and we should expect to find, as indeed we do find, an enormous number of mutations hung up in the uninteresting condition of being merely recessive lethals. If, however, when this stage is reached the homozygote is viable, a second stage of progress will commence, directed this time to the improvement of the homozygote and depending as to its speed on the viability of the homozygote just as the first stage of progress depended on the viability of the heterozygote. Examples of the modifiability of the homozygote are almost too abundant. I must however mention, for the sheer beauty of their demonstration, the group of recessive suppressors of vermilion, sable, black, and purple, the existence of which was first suspected by BONNIER,which have been shown by BRIDGESand SCHULTZto be certainly not duplications, as was at first believed. In the presence of the suppressor, the vermilion homozygote is normal, and the vermilion mutation is, as far as is known, undetectable. Whereas the first stage of modification ends in a recessive condition with a lethal, or viable and recognizable, homozygote, the second stage reduces it to a state of obliteration, from which it can only be made to appear as a specific modifier if it happens to be a sufficiently substantial modifier of any mutant which is being studied. I t is important t o consider how frequently these processes are actually occurring and how generally we should expect that the condition observed

168

PROCEEDINGS OF THE SIXTH

Is a stage in a process of continuous modification. The examples I have given of known modifiers have necessarily been factors having a relatively large and regular effect. The study of quantitative characters, however, or of peculiarities having variable manifestation, seems invariably to show evidence of a numerous group of modifying factors having each only a slight effect. The cases in which new mutants are found to be affected by a fluctuating variation having a hereditary basis are very numerous; frequently the mutant type has been found to be modified perceptibly toward the wildtype by the natural selection of modifiers mitigating its expression in competition in the conditions of culture. For this reason I am inclined to think that the large modifiers, such as those which suppress the whole manifestation at a single step, have not been the principal agency of dominance modification in the past history of the species studied. In particular, there are reasons for thinking that the homozygote, on the modifiability of which most of our experience is based, has been modified considerably more slowly on the average than the heterozygote. In Drosophila rnelanogaster the mutants classed as recessives with viable homozygotes are about sixteen times as numerous as the semi-dominants with viable homozygotes. These dominants, being incomplete dominants, may be regarded as being still in the first stages of modification, and the recessives, or at least those of them in which the recessiveness is really complete, must be in the second stage; their relative numbers suggest as an upper limit that the homozygote may take on the average sixteen times as long as the heterozygote to complete the normalizing process. The largest factor in causing this difference is, I imagine, that the homozygotes probably commence their modification a t a lower viability than the heteroz~gotes,for, as I have shown, a moderate difference in viability may greatly retard the rate of selective modification. The possibility of modification of dominance by genetic substitution is, I suppose, now unquestioned; but the conclusion that the condition of dominance now observable is in any case the result of evolutionary modification is an inference subject, like all such inferences, to some such proviso as "unless some unknown cause prevents the process." This is a proviso ALL WRIGHT, if I to which all evolutionary theory is necessarily subject. SEW understand him, has suggested that there is such an obstacle and that very small selective intensities do not, as one would naturally assume, exert effects proportional to their magnitude; but I have so far found it impossible to set up any reasonable scheme of genic interaction which would justify this conjecture. The fact of the evolutionary modification of dominance is, however, demonstrated by HARLAND'S case of the mutation known as

INTERNATIONAL CONGRESS O F GENETICS

169

crinkled dwarf in the new world cottons. This mutant is of frequent occurrence in Sea Island cotton and some of its derivative varieties, but has not been found in large selfed progenies in the Upland group. As crosses seemed has to indicate that the dominance relationship was modified HARLAND introduced crinkled dwarf by five generations of backcrossing into the Upland species and has shown that in that species it is an incomplete dominant. The evolutionary process by which these two species have been differentiated has therefore included the modification of their reaction to the crinkled dwarf mutation in such a way that in the species in which it occurs the mutant has become recessive. The case indicates that whatever the cause of the modification may be it is conditioned by the appearance of the mutant in the ancestry of the population concerned, and that the means of modification is the establishment of a group of modifying factors and not merely a modification of the normal allelomorph at the locus of the mutant. In the case of deleterious mutants the proportion of heterozygotes in the ancestry of the population must generally be small and the process of modification correspondingly slow. With species polymorphic in the wild condition, the heterozygotes for the factors determining the polymorphism are much more abundant, so that in these cases rapid modification is possible. In such polymorphic species, moreover, the mere maintenance of a stable gene ratio requires that the selective actions must be balanced, and its stability requires that the heterozygote must generally be at a selective advantage compared to both homozygotes. The dominance relationships in such cases should be entirely different from those of the simple elimination of a recurrent deleterious mutation. I have only time for one example, where the selective balance is evidently due to opposite action in the two sexes. In Lebistes reticulatus WINGEhas found numerous Y-linked genes affecting the spots and patches of color on the male fish. Some-of these have been found to cross over into the X chromosome. These are all without manifestation in the female, apart from intersexes. The effect on the male can be seen to be dominant, since the phenotypic expression is the same whether the variant gene is in the X or in the Y or in both. There is also an autosomal gene zebrinus which is completely dominant in the male but which has shown occasional manifestation in the female when homozygous. In the female, therefore, it has an occasional recessive manifestation. These rather exceptional phenomena conform with remarkable exactness to what would be expected if the genes responsible for polymorphism are advantageous in the male and disadvantageous in the female. First, we should expect the variant genes to become dominant in the male, and recessive in the female fish. In

170

PROCEEDINGS O F THE SIXTH

the next stage we should expect the entire obliteration in the female of the effects of those genes which are capable of crossing into the X chromosome. Thirdly, counter-selection in the females should make the variants rarer in the X than in the Y in wild populations, whereas without selection crossing will equalize the ratio, or indeed reverse it, if AIDAis right in suggesting that crossing over from Y to X is more frequent than from X to Y. Fourthly, favorable selection in the Y with counter-selection in the X would favor those genotypes in which linkage was closest with the sex determining portion of the Y chromosome, and may thus have built up the closely sexlinked system which is observed. There is, one might think, an evolutionary opportunity for a translocation which would put zebrinus into the Y chromosome. The sex linkage, however, need not be ascribable entirely to translocations, for it is obvious that mutations that occur from the first in the Y chromosome will have the highest probability of establishing themselves in the polymorphic system. On the whole, it is difficult to see how WINGE'S findings could suggest more strongly than they do the modification of both dominance and linkage in the evolutionary process. The view of the selective modification of dominance is thus able to reconcile such contrasting facts as the prevalence of recessiveness among recurrent mutations exposed to counter-selection with the prevalent dominance of the variant forms in polymorphic species, although of these I have had time to discuss only one case. Cases where dominance is imperfect or absent are equally instructive. I will mention five classes of these: ( A ) In multiple allelomorphic series the heterozygotes with the wild-type gene will have occurred with sensible frequency in the population's ancestry, and accordingly the wild-type is generally dominant, but the heterozygotes of two mutant genes will have occurred scarcely more frequently than the homozygotes and should therefore show incomplete dominance. ( B ) As has been pointed out by FORD,DOBZHANSKY has shown that the mutants of the white eye series of Drosophila nzclanogaster, as well as sooty and ebony, while recessive in their major morphological features, are yet incomplete dominants in their small but constant effects on the shape of the spermatheca, a feature which one would expect to be unaffected by natural selection. (C) HARLAND'S case in cotton shows a recessive in one species which is an incomplete dominant in a species in which it has not been exposed to counter-selection. ( D ) A very large number of cases could be cited in which genes that are recessive in the wild-type are incomplete dominants in artificial genetical combinations which do not occur in nature. (E) The same thing is shown by unnatural environmental conditions such as exposure of mice to X-rays until the hair

INTERNATIONAL CONGRESS OF GENETICS

171

falls out, when the regenerated coat in heterozygous albinos, but not in mice homozygous for color, shows patches of white hair. All these five groups of evidence, which I have not time to amplify, indicate that the relationship of dominance is usually conditioned by selection of the heterozygote, and by selection in the special genetical complex and in the special environmental conditions which exist in nature. The theory of the evolution of dominance, like other mutations, is itself liable to modification. I t must, I suppose, be subjected to an evolutionary process, and if it is found to be deleterious it also may end in obliteration. The most promising modifications may perhaps be stated very briefly in terms of the magnificent series of multiple allelomorphs of the vestigial series which MOHRput before us on Friday. First, there are one or two allelomorphs like nick which have no visible effect even when homozygous but which may be detected by a slight manifestation in heterozygosis with vestigial. On my own view the natural interpretation to put on nick would be to regard it as having already reached the stage of complete obliteration. Now HALDANE has put forward a theory of dominance modification which he thought might be more effective than mine and which depends on selection among a multitude of normal allelomorphs of different strengths, by which those are selected which completely dominate the deleterious mutants of the series, such as vestigial. On this view nick might be regarded as one of a group of normal allelomorphs which are incapable of giving a completely wild-type development in the presence of heterozygous vestigial. I think this possible selection among multiple allelomorphs may, in some other cases, be of great importance, though generally speaking selection of multiple factors is, I believe, considerably the more powerful agency. In the present case the would regard chief difference between the two theories is that HALDANE nick, or other allelomorphs like it, as having been formerly widely diffused in the wild population and as having been displaced in competition with the wild allelomorph now prevalent, owing to the inferiority of its heterozygote with vestigial; whereas I should say it was incompletely dominant to vestigial just because it had never been sufficiently widely diffused in the wild population for its heterozygote with vestigial to have been modified up to normality. A t low temperatures the effects of some of these slight allelomorphs such tells me, are enhanced, so that in cultures developed as pennant, PLUNKETT at a low temperature hon~ozygouspennant will show a slight manifestation. I imagine that this may be such a case as MULLERhad in mind in suggesting that dominance might have been acquired as a by-product of the wild-type

172

PROCEEDINGS OF THE SIXTH

gaining stability of manifestation under variable environmental conditions. in relying on mulThis modification of my views differs from HALDANE'S tiple factors rather than on multiple allelomorphs, while, on the other hand, it differs from us both in that dominance in MULLER'Sview would be acquired without the previous prolonged occurrence of mutations of the vestigial series. I believe this view would account for the continued progress of pennant toward obliteration until it is unrecognizable even at the lowest temperature possible. I do not yet see how it accounts for the fact that the heterozygous vestigial more closely resembles the wild than the mutant homozygote. In speaking of the modification of the results of single mutations, I implied that the rate of modification would be negligible for forms having less than 50 percent viability in the wild conditions, and that the lethal forms would be unmodifiable. In such a series as has been found at the vestigial locus, such a static and pessimistic view seems unwarranted. The members of the series that, while not completely normal, have yet a -high viability are doubtless exerting a relatively strong selective action on the modifiers available, and these same modifiers are doubtless in some measure simultaneously improving the viability of all other members of the series. As I judge the situation, they must, as the song says, "all go the same way home," and though some, no doubt, would be quite incapable of progress if left to themselves, yet it would seem that their more viable companions must help them along. Even a lethal is not necessarily beyond such assistance but might be hoisted out of the ditch if the others are numerous and active enough.

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

173

G E N E T I K D E R GEOGRAPHISCHEN VARIATION R. Goldschmidt, Kaiser Wilhelm-Institzlt fur Biologie, Berlin-Dahlem, Germany

Das Problem der Evolution, das nach den phylogenetischen Orgien der ersten Nach-Darwinschen Periode etwas in Misskredit gekommen war, das dann spiiter von der neu begriindeten genetischen Wissenschaft respektvoll zur Seite geschoben worden war, ist seit einiger Zeit wieder im Begriff, ins Zentrum biologischer Diskussionen zu riicken. Und zwar haben gleichzeitig mehrere Disziplinen unserer Wissenschaft begonnen, sich dariiber Rechenschaft zu geben, wie man heute auf Grund der neuen Tatsachenerkenntnisse den Vorgang der Evolution, an dessen TatGchlichkeit niemand zweifelt, sich vorstellen muss. Die Palaeontologie hat begonnen, die Reihen der Fossilien unter neuen Gesichtspunkten zu betrachten ; die Systematik hat den Artbegriff durch eine labilere Betrachtungsweise ersetzt ; die vergleichende Anatomie hat angefangen, von der grossziigigen Vergleichsmethode zu einer minutiosen Betrachtung iiberzugehen, die zu elementaren Gesetzdssigkeiten fiihren soll. Und auch die Genetiker haben nach einer Periode der Vorsicht wieder begonnen, sich daruber Gedanken zu machen, was die Tatsachen der Vererbung fiir den Vorgang der Evolution lehren. Zuerst mit mehr negativer Kritik beginnend, haben jetzt verschiedene geneiische Richtungen versucht, die grossen Erkenntnisse unserer Wissenschaft f u r eine neue Attacke auf das Evolutionsproblem zu verwenden. Bald sind es die Erkenntnisse iiber Mutationen, bald die iiber das Zusammenspiel der Gene, bald die iiber Bastardierung, iiber Chromosomenveranderung, iiber die Entstehung der Kulturpflanzen oder uber die mathematischen Konsequenzen des Selektionsvorgangs, die Veranlassung geben, bestimmte Methoden des Evolutionsvorgangs zu postulieren oder zu verwerfen. Aber neben diesen Versuchen, auf theoretischem Wege dem von der Genetik aufgehauften Material neue Erkenntnisse zu entlocken, Versuche, an denen ich selbst mich schon vor 15 Jahren auf damals neuartigen Wegen beteiligte,' haben nicht allzu viele Forscher versucht, sich dem Problem direkt von der Wurzel her zu nahern. Welches ist diese Wurzel? Verfolgen wir die Arbeiten der modernen Systernatik, also der Forscher, die das in der Natur vorhandene Material am genauesten kennen und durch diese Kenntnis neben den Palzontologen am meisten befahigt sind, das Ergebnis der Evolution-aber nicht etwa die Methoden der Evolution-in den Verhaltnissen der Natur zu beurteilen, so zeigt es sich, dass die Entwicklung dahin gefiihrt hat und zwar vollstandig in der Zoologie, noch nicht Is. GOLDSCHMIDT, R., A preliminary report on some genetic experiments concerning evolution. Amer. Nat. 1917.

PROCEEDINGS OF THE S I X T H

174

so weitgehend in der Botanik, den Artbegriff zu ersetzen durch den Begriff geographischer Rassenkreise. Alle Formen, die iiber ein grosses Areal weg sich gegenseitig in den verschiedenen Teilen des Areal vertreten, aber nie nebeneinander vorkommen, bilden zusammen einen Rassekreis, der dann etwa der Linnk'schen Art entspricht, wahrend die einzelnen geographisch wie erblich getrennten Unterformen als Subspezies oder geographische Rassen bezeichnet ~ e r d e n Fur . ~ manche Systematiker bedeutet nun die Gliederung in geographische Rassen, die als ihrem Habitat optimal angepasst betrachtet werden, den sichtbaren Beginn der Divergenz, die zur Bildung getrennter Arten fiihrt, und tatskhlich gibt es unter natiirlichen Bedingungen kein anderes Material, das so deutlich die Grundtatsachen des Evolutionsvorgangs, Divergenz und Anpassung, einschliesst und daher geradezu auffordert, sie mit genetischen Methoden zu erforschen. Da ich selbst als Folge meines eigenen friiheren geographischen Standorts vielleicht etwas mehr als andere Genetiker mit den Stromungen der Systematik vertraut war, begann ich vor etwa 20 Jahren den Versuch zu machen, eine Art zu analysieren, die iiber ein grosses Areal verbreitet ist und innerhalb dieses Areals geographische Rassen bildet. Die Aufgabe war, rnoglichst alle Rassenunterschiede innerhalb der Art genetisch zu analysieren, festzustellen, ob eine bestimmte Ordnung fbesteht, die irgendwie mit der geographischen Ordnung iibereinstimmt, und festzustellen, wie weit von Anpassungscharakteren gesprochen werden kann, und welches ihre genetische Grundlage ist. Da diese Untersuchung nicht nur die erste ihrer Art war, sondern meines Wissens auch die einzige ist, die f u r ein ganzes geographisches Areal durchgefiihrt ist, so mochte ich mir in Anbetracht der Kiirze der Zeit erlauben, nur von meinen eigenen Ergebnissen zu berichten. Die Art Lymantria dispar ist iiber die ganze palaarktische Region verbreitet, d.h. iiber ganz Europa einschliesslich Russland, siidlich bis zum Mittelmeer vordringend, ja sogar es in Algier und Marokko iiberschreitend; ostlich durch Sibirien bis zum Pazifik vordringend, dabei siidlich den Kaukasus und russisch Turkestan erreichend, an der chinesischen Kiiste etwa bis Schanghai herabgehend und endlich auf den drei japanischen Inseln zu Hause. Das Verbreitungsgebiet schliesst also alle moglichen klimatischen Verhaltnisse im Rahmen der gemassigten Zone ein. Die Untersuchung zeigte dann bald, dass die Formen auf dem grossen eurasischen Kontinent in vielen Eigenschaften eine Einheit bilden, der die Formen Ostasiens als geschlossene Gruppe gegeniiberstehen; ferner, dass sich innerhalb des ostavor allem die Arbeiten von KLEINSCHMIDT und STRESEMANN fiir Vogel, GRINNELL JORDAN fiir Schmetterlinge. Eine zusammenfassende Darstellung, leider mit lamarckistischer Grundeinstellung, gibt RENSCH,B., Das Prinzip geographischer Rassenkreise und das Problem der Artbildung. Berlin : Gebr. Borntrager 1929. S.

11.

a. fiir Siugetiere,

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

175

siatischen Verbreitungsgebiets eine besonders typische Ausbildung von geographischen Rassen fand. Dies Gebiet nun zeigt ausserordentlich verschiedene klimatische Verhaltnisse; etwa die im Winter fast subarktische Insel Hokkaido, die Heimat eines nordlichen B L e n ; die westlichen Teile Japans, mit fast subtropischem Klima, mit Kampherbaumen und Zykadeen ; oder die im Sommer gluhenden, im Winter eisigen Kontinentalgebiete von Korea und der Manschurei. Hier musste also die Moglichkeit gegeben sein, eventuelle Anpassungen an den Standort und ihre genetische Bedingtheit zu analysieren. Diese Analyse ist zwar noch nicht abgeschlossen, aber doch schon so weit gefordert, dass die Hauptresultate sichtbar werden. Die Systematiker unterscheiden ihre geographischen Rassen nach mehr oder minder grossen Serien von Museumsstucken, die von einer Lokalitat stammen. E s sind also hauptsachlich Aussencharaktere, die die Formen kennzeichnen und fur die stillschweigend angenommen wird, dass sie Anpassungscharaktere a n die Umgebung darstellen, ohne dass es bewiesen werden kann. Finden sich an der Grenze zweier Areale Zwischenformen und zeigen diese starkere Variation als die typischen Rassen, so werden sie als Bastarde betrachtet. Von den geographischen Rassen wird als selbstverstandlich angenommen, dass sie erblich verschieden sind und nicht etwa Standortsmodifikationen, obwohl dafiir gewohnlich die Beweise fehlen. E s ist selbstverstandlich, dass in unsereii Versuchen alle analysierten Rassen unter gleichen Bedingungen nebeneinander geziichtet wurden, wodurch das nichterbliche Element ausgeschaltet wurde. Gleichzeitig wurde aber versucht, von den sichtbaren Aussencharakteren, die ja relativ leicht auf ihren Erbgang zu analysieren sind, aber meist keinen direkten Anpassungswert haben diirften und daher fiir die Evolution ohne Interesse sind, auf mehr physiologische Eigenschaften vorzudringen, die als Anpassungscharaktere in Betracht kommen konnen. Da das aufgehaufte Tatsachenmaterial ein ausserordentlich grosses ist, so will ich einige typische Tatsachen nur schnell vorbeiziehen lassen und nur einen Anpassungscharakter in den Vordergrund stellen. Wir konnen bei dem Material der Art Lymantria dispar aus der Natur vier Typen von Erbverschiedenheiten unterscheiden, die fur die Evolution von verschiedener Wertigkeit sind. 1. gibt es typische Mutanten von dem Typ der meisten Drosophilamutanten, die als pathologische Mutanten keinerlei Evolutionswert haben. S o trat Weissiugigkeit als rezessive autosomale Mutante auf und ebenso Weichheit des Fliigelchitin~.~ 2. gibt es 3Erstere Mutante wurde von F. LENZentdeckt und dann von mir analysiert; letztere ist von meinem Mitarbeiter MACHIDA gefunden worden. s. J. MACHIDA:Crossing experiments with gipsy moths. J. Coll. Agric., Imp. Univ. Tokyo 7, 1924.

176

PROCEEDINGS O F THE SIXTH

Mutanten, die nicht pathologisch sind, vielmehr im Rahmen einer normalen Variabilitat liegen und die sich dadurch auszeichnen, dass sie bisher an bestimmte Rassen gebunden waren, ohne aber diese Rassen zu charakterisieren. Hier sind drei Beispiele: Nur im aussersten Norden der japanischen Hauptinsel wurden Formen gefunden, deren Mannchen samtschwarze Fliigel besitzen. A n der gleichen Lokalitiit finden sich aber auch helie Formen. Diese Fliigelfarben sind Glieder einer Allelomorphenserie, von denen das hellste Glied fur den aussersten Norden Japans charakteristisch ist, das dunkelste aber nur im Norden der Hauptinsel und neben der hellen Form vorkommt. Wir konnen also von einer zunachst lokalen Mutation sprechen, die aber nicht durch Anpassungsbeziehungen an die Lokalitat gebunden ist, sondern ihr nur zufallig angehort. Ein anderes Beispiel desselben Typus ist das Auftreten von Raupen mit schwarzem Ruckenstreif, das bisher nur von verschiedenen Teilen Deutschlands bekannt ist. E s beruht auf .einem einfachen dominanten autosomalen Gen,' das von sonstigen Raupenzeichnungen unabhangig ist, wie die Rekombination mit der hellen Zeichnung der japanischen Rassen zeigt. Als drittes Beispiel diene die folgende Mutation der Fliigelzeichnung, bei der die mittleren Zickzacklinien ausgefallen sind, eine einfache autosomale rezessive Mutation, die nur bei japanischen Rassen aufgecreten ist. Derartige Mutanten sind also keine typischen Charaktere geographischer Rassen, konnten aber im Fall der Isolation zu typischen Charakteren einer Lokalform werden, ohne dass sie Anpassungscharakter besitzen. Die 3. Gruppe umfasst solche erblichen Differenzen, die iiber grosse Areale weg die Gesamtart typisch gliedern, aber in einer anderen Gliederung, als die spater zu besprechenden Charaktere. Als Beispiel diene die Farbe der Afterhaare der Weibchen. Alle nordlichen Formen, also in Nordeuropa, Nordrussland und auf der nordlichsten japanischen Insel Hokkaido haben dunkle Afterhaare, alle siidlichen, also die mediterranen, die von Turkestan sowie die von Japan und China haben helle Afterhaare. Der ~ konnte es sich beCharakter wird geschlechtskontrolliert ~ e r e r b t .Hier reits um einen Anpassungscharakter handeln. Nicht etwa, dass die Behaarung selbst fiir die Anpassung in Betracht kame. Aber genau die gleichen Gruppen unterscheiden sich auch durch solche physiologischen Charaktere 'Zuerst von KLATT,dann von BALTZER beschrieben; ich ziichtete die Mutante in allen moglichen Bastardkombinationen, die nichts Unerwartetes ergaben. s. KLATI', B., Keimdriisentransplantationen beim Schwammspinner. Zs. ind. Abstl. 22, 1919; BALTZER,F., Ueber mendelnde Raupenrassen bei Lymantria dispar. Festschr. Basel: Zschokke, 1920. 's. R. GOLDSCHMIDT U. S. MINAMI:Ueber die Vererbung der sekundlren Geschlechtscharaktere. Studia Mendeliana, Briinn, 1923.

INTERNATIONAL CONGRESS O F GENETICS

177

wie etwa die Entwicklungsgeschwindigkeit, die eminente Anpassungscharaktere sind und die sich ja unter anderem auch phanotypisch in Pigrnentquantiften ausdrucken konnten. Die wichtigste Gruppe von Rassencharakteren ist aber die vierte. Hier handelt es sich namlich um solche Erbeigenschaften, die sich in einer typischen Reihenfofge von Areal zu Areal verandern, und zwar parallel mit erfassbaren Veranderungen des Milieus. Von diesen Charakteren seien drei, deren Analyse am weitesten fortgeschritten ist, kurz besprochen. Der merkwiirdigste Charakter ist die S f rke der Geschlechtsgene, durch die die verschiedenen Geschlechtsrassen von Lymantria charakterisiert werden, die in unseren 1ntersexualit;itsexperimenten analysiert sind. Die geographische Ahgrenzung dieser Geschlechtsrassen steht nunmehr auf Grund der Analyse eines ungeheuren Materials fest6 und ist die folgende. Die schwachste aller Geschlechtsrassen bewohnt die Insel Hokkaido. Ihr direkt gegeniiber auf der Hauptinsel finden sich die stirksten Formen, die ganz Nordostjapan bevolkern. Unter den starken Rassen gibt es drei verschiedene Stirkegrade, von Osten nach Westen abnehmend. Ein mittlerer Grad findet sich in dem zentralen Gebirgsland und der relativ schwkhste Grad in einem Grenzstreifen ostlich vom Biwasee. Die schmalste Stelle der Insel zwischen Tsuruga und der Isebucht stellt die Grenze zwischen starken und neutralen Rassen dar. Die ganze westliche Halfte der Insel ist von neutralen Rassen bewohnt. Auch auf der dritten Insel, Kyushiu, finden sich nur neutrale Rassen, aber ihr Stiirkegrad ist geringer, als der der Formen nahe der Grenzlinie. An einer Stelle dieser Insel, bei Kumamoto, wurde sogar einmal der nachst niedere St2irkegrad gefunden. Dieser, die halbschwache Form, zu der ubrigens auch die aus Frankreich nach Massachusetts importierten Formen gehoren, findet sich dann in ganz Korea vor. S o haben wir denn eine ganz typisch absteigende Reihe der Starke der Sexualgene vom Nordosten Japans nach Siidwesten fortschreitend bis Korea. In der Manschurei und China steigt die Starke wieder etwas an zur Grenze von halbschwach und neutral. Das ganze ubrige eurasische Festland wird von schwachen Rassen bewohnt, mit unregelmassigem punktweisem Auftreten von halbschwachen Formen in Turkestan, am Alpensudrand und anderen Stellen. E s ist endlich noch zu bemerken, was eigentlich selbstverstandlich ist, dass an der Grenze zweier Gebiete heterozygote Formen vorkommen. Genetisch aber beruhen die Unterschiede aller dieser Sexualrassen auf einer Serie von mindestens 8 multiplen Allelen der Geschlechtsgene. 's. R. GOLDSCHMIDT: Untersuchungen zur Genetik der geographischen Variation. 111. Abschliessendes iiber die Geschlechtsrassen von Lymantria dispar L. Roux's Arch. 126, 1932.

178

PROCEEDINGS O F T H E S I X T H

Die ausserordentlich typische Verteilung dieser Rassen deutet darauf hin, dass ihr irgend eine biologische Bedeutung zukommt. Soweit es moglich ist, in sie einzudringen, sieht es so aus, als ob typische Geschwindigkeiten von Differenzierungsvorgangen der einzelnen Rassen auch eine entsprechende Koordination der sexuellen Differenzierung erforderten, die ihren Ausdruck in dem findet, was wir die Starke der Sexualgene nannten. Die absonderliche Erscheinung der geographischen Sexualrassen ware dann eine Anpassungserscheinung an die Zeitverhaltnisse des Lebenszyklus. W i r wollen diese schwierige Frage aber nicht naher erortern, sondern uns leichter verstandlichen Rassencharakteren zuwenden. W i r erwahnten schon vorher die Verschiedenheit der Raupenzeichnung europaischer und japanischer Rassen. Die europaischen Formen sind dunkel, die japanischen haben eine helle Zeichnung, und diese helle Zeichnung ist verschieden in den verschiedenen Teilen des Verbreitungsgebiets, wofur das folgende Bild ein paar Beispiele zeigt. Dazu kommt nun ein weiterer Unterschied. Unter den japanischen Rassen gibt es solche, die durch alle Raupenstadien hindurch hell bleiben, und andere, bei denen unter identischen Bedingungen die hellen Stellen mit fortschreitender Entwicklung durch dunkles Pigment bedeckt werden, sodass schliesslich die alten Raupen ebenso aussehen wie die Europaer. Teilen wir das relative Mass der Verdunkelung in 10 Klassen ein und verbinden die in jedem Raupenstadium gefundene Klasse durch eine Kurve, so bekommen wir die mit der Entwicklung fortschreitende Verdunkelungsskurve fur viele Rassen. W i r sehen die Kurven immer heller, immer dunkler and allmahlich sich verdunkelnder Rassen. E s zeigte sich nun, dass der Unterschied dieser Rassen durch eine Reihe multipler Allele bedingt ist. Hell ist dominant, und in F, tritt eine Spaltung in 3+ hell: 1 dunkel ein, und dies fur alle verschiedenen Grade der V e r d u n k e l ~ n g . ~ Dazu kommen aber noch zwei Besonderheiten: 1. tritt in der Heterozygote ein Dominanzwechsel von hell nach dunkel in alteren Stadien ein. Die Analyse brachte die Erklarung dafur, die sich aus den folgenden Kurven ergibt. Sie zeigen, dass der Unterschied zwischen hellen und dunkeln Rassen so aufgefasst werden kann, dass die hellen Rassen eine sehr langsame, die dunklen eine sehr schnelle zeitliche Pigmentierungskurve haben ; dazwischen stehen die Kurven fiir die Rassen, die erst hell, dann dunkel sind und natiirlich auch fiir die Heterozygoten, was wie ein Dominanzwech? E r s t e Mitteilung iiber diesen Fall in der in Anm. 1 zitierten Arbeit; weitere Daten in GOLDSCHMIDT, R. : "Die quantitativen Grundlagen von Vererbung und Artbildung." Berlin : J. Springer, 1920. Genaue Analyse in GOLDSCHMIDT, R.: Untersuchungen zur Genetik der geographischen Variation I. Roux Arch. 101, 1924 und erganzende Daten in desgl. 11. Ibid. 116, 1929. Weitere Daten sind in Vorbereitung.

INTERNATIONAL CONGRESS O F GENETICS

179

sel aussieht. Die 2. Besonderheit ist, dass der Phanotypus durch die plasmatische Beschaffenheit beeinflusst wird. Dies war ubrigens der erste Fall, in dem schon vor langer Zeit mit genetischen Methoden eine plasmatische Vererbung festgestellt wurde.' Die folgenden Kurven zeigen ein aktuelles Resultat die matrokline F,, die 3 : l Spaltung in F, und die Tatsache, dass die ganze Fz Kurve in reziproken Kreuzungen nach der mutterlichen Seite verschoben ist, die plasmatische Beschaffenheit also die Pigmentierungskurve nach ihrer Seite zieht. Wie verhalten sich diese Charaktere nun geographisch ? Wie gesagt, sind alle europaischen Rassen dunkel, wenn wir von gewissen kleinen Unterschieden absehen. In Japan nimmt Hokkaido wieder eine Ausnahmestellung eiti. Hier findet sich eine Rasse, deren Raupen zuerst die hellsten von allen sind, spater aber sehr dunkel werden. Auf der Hauptinsel folgt dann im Norden das dunkelste Allel der Serie, und genau von Nordosten nach Siidwesten folgen die helleren Allele bis etwa zur Grenzzone, die wir von den Geschlechtsrassen her kennen. Von hier westlich haben wir die immer hellen Rassen. Gehen wir dann nach Korea hiniiber, so werden die Rassen von Siiden nach Norden dunkler und ahneln in der Manschurei und China bereits sehr den europaischen. Die folgenden drei Bilder zeigen dies in wirklichen Kurven. Diese ausserordentlich typische Verbreitungsart der Allelomorphenreihe deutet darauf hin, dass es sich wohl auch um eine Anpassungsreihe handelt, bei der der Pigmentierungsvorgang von typisch verschiedener Geschwindigkeit naturlich nicht selbst der Anpassungscharakter ist, sondern der sichtbare Ausdruck eines physiologischen Vorgangs von Anpassungswert, iiber dessen Modus wir Hypothesen aufstellen konnen, aber kaum eine einwandfreie Vorstellung beweisen konnen. Anders aber steht es mit einer weiteren Reihe von Erbfaktoren, die die Rassen unterscheiden und deren Natur als Anpassungscharakter sicher feststeht. Der Lebenszyklus von L ~ m a n t r i aist der folgende: Im Sommer erscheint der Schmetterling und legt seine Eier. Diese entwickeln sich sofort, aber das junge Raupchen verlasst das E i nicht, sondern uberwintert in ihm bis zum nachsten Friihjahr. Nach Erwachen der Vegetation schlupft auch das Raupchen aus und entwickelt sich im Lauf des Sommers zum Falter. E s ist nun klar, dass der normale Ablauf dieses Lebenszyklus nur ~noglich ist, wenn erstens die Raupchen im Friihjahr nicht zu fruh und nicht zu spat erwachen und zweitens die Entwicklungsperiode im Sommer nicht langer dauert, als die Vegetationsperiode der Futterpflanzen. Nun sind diese entscheidenden Verhaltnisse in den verschiedenen klimatischen Abschnitten 'Versuche von 1912-14, ausf iihrlich verijffentlicht 1924 ; S. Anm. 7.

180

PROCEEDINGS OF T H E SIXTI-:

des Verbreitungsgebiets recht verschieden; die Tatsache der Besiedlung so verschiedener Gebiete hat also zur Voraussetzung, dass sich der Lebenszyklus der Form dem Jahreszeitenzyklus des Areals anpassen kann. E s ware natiirlich moglich, dass dies allein durch eine geniigende Breite der Modifikabilitat geschieht; tatdchlich ist das aber nicht der Fall, vielmehr sind die geographischen Formen in diesem Charakter dem Milieu erblich angepasst. Ich will nicht auf die von mir studierten Einzelheiten der Physiologie des Ueberwinterungsvorgangs eingehen.' Fur unsere Zwecke geniigt es zu wissen, dass der Zeitpunkt des Ausschliipfens von mehreren Komponenten bedingt wird. Die wichtigste davon ist eine mit der Zeit fortschreitende temperaturabhangige Reaktion, die zum Schliipfen fiihrt, wenn ceteris paribus eine bestimmte Warmesumme erreicht ist. Dazu kommt ein ebenfalls mit der Zeit fortschreitender, aber temperaturunabhangiger Prozess, der die Schliipfbereitschaft standig steigert. Schliesslich sind noch einige andere physiologische Bedingungen vorhanden, und die Gesamtheit dieser Vorgange bedingt somit die Zeit, die bis zum Schliipfen benotigt ist, die Inkubationszeit. Diese Inkubationszeit ist aber erblich, rassenmassig verschieden, und zwar in solcher Weise, dass der Anpassungscharakter manifest wird. Die kiirzeste Inkubationszeit haben die Formen, die den nordlichen eurasischen Kontinent bewohnen, dazu die Formen von Korea und Hokkaido. Der Anpassungscharakter ist klar: ein kalter Winter, schnelles Friihjahr und relativ kurzer Sommer erfordern es, dass eine relativ geringe Warinesumme ausreicht, um das Schliipfen herbeizufiihren. Umgekehrt haben eine sehr lange Inkubationszeit die mediterranen Formen und die Sudwestjapaner. Diese Gegenden sind durch milden Winter charakterisiert ; es wiirden also die Raupen vor der Vegetationsperiode schlupfen, wenn sie nicht erblich auf eine hohere Warmesumme eingestellt waren. Im iibrigen Japan finden wir aber nun scheinbar ein umgekehrtes Verhalten. Von Siidwesten nach Nordosten verlangert sich die ererbte Inkubationszeit, sie wird also rnit strengerem Winter langer statt kurzer. Die Erklai-ung liefert eine genauere Betrachtung der meteorologischen Daten. E s zeigt sich, dass der Jahreszyklus ein anderer ist als etwa in Europa oder Hokkaido. Zwar ist der Winter kalt, aber die Friihjahrserwarmung kommt schon einen Monat friiher, ohne dass dies auch fur das Ausschlagen der Laubbaume zutrifft. E s ist also in Nordostjapan tatsachlich eine hohere Warmesumme als Anpassungscharakter notig. Wir begniigen uns mit der Aufzahlung dieser grossen Linien. Tatsachlich lassen sich noch eine ganze Reihe von Untergruppen mit dem O Naheres in GOLDSCHMIDT, R., Untersuchungen zur Genetik der geographischen Variation. V. Roux Arch. 1932.

INTERNATIONAL CONGRESS O F GENETICS

181

Klima ihres Wohnorts in Beziehung bringen. W i r verzichten auch darauf, auszufuhren, dass noch andere Erbunterschiede fur den gleichen Vorgang nachgewiesen werden konnen. S o ist der Vant'Hoffsche QIo-Quotient fur verschiedene Rassen verschieden, woraus auf eine erblich verschiedene Einstellung auf verschieden hohe Temperaturoptima geschlossen werden kann. Vom genetischen Standpunkt aus sind also die geographischen Formen in Bezug auf die Inkubationszeit erblich verschieden, und zwar beruht die Verschiedenheit auf verschiedenen Geschwindigkeiten im Ablauf eines Hauptprozesses und mehrerer Nebenprozesse. E s ist also eine polymere Vererbung, wahrscheinlich mit einem Hauptgen und mehreren Modifikatoren zu erwarten. Hier ist nun ein Beispiel der genetischen Analyse bei Kreuzung von zwei Formen mit langer und kurzer Inkubationszeit: F1zeigt unvollkommene Dominanz der langsameren Rasse und ausserdem Matroklinie. F, zeigt die Kurve einer polymeren Spaltung, deren Hauptgebiet auf der dominanten Seite liegt. Eine Entscheidung dariiber, ob die Serie der Rassenunterschiede auf verschiedenen Genkombinationen beruht oder mehr auf multipeln Allelen des Hauptgens, steht noch aus, und von allen Einzelheiten und Komplikationen sei hier abgesehen. Die Hauptsache steht jedenfalls fest, dass eine typische, geordnete Reihe von Anpassungen an die Umgebung mendelistisch aufgelost ist. E s sei auf die weitere Beschreibung anderer Charaktere, die unsere geographischen Rassen unterscheiden, hier verzichtet. Nur mit einem Wort sei auf einen Punkt eingegangen, dessen Erwahnung Sie von mir erwarten, die Zytologie. Samtliche Rassen besitzen haploid 31 Chromosomen, die nicht so deutlich verschieden sind, dass man sie einzeln vergleichen konnte.1° Dagegen finden sich typische Unterschiede in der gesamten Chromatinmenge, wie ein Vergleich einer chromatinreichen mit einer chromatinarmen Platte zeigt. E s wurde nun versucht, diesen Charakter fiir die geographischen Rassen zu vergleichen, und dabei zeigte sich merkwurdigerweise, dass die Chromatinmasse umgekehrt proportional der Starke der Geschlechtsrassen ist. Was das hedeutet, ist eine Spezialfrage, die nicht zu unserem heutigen Thema gehort. Wir haben nun gezeigt, dass eine Serie geographischer Rassen der gleichen Art innerhalb ihres Verbreitungsgebietes sich durch Erbeigenschaften unterscheiden, die in geordneter Weise von Lokalitat zu Lokalitat verschieden sind, ohne dass sich die einzelnen Typen von Unterscheidungscharaklo

Naheres in GOLDSCHMIDT, R., Untersuchungen zur Genetik der geographischen Varia-

tion. IV. R ~ u xArch. 126. 1932.

182

PROCEEDINGS OF THE SIXTH

teren vollstandig ortsweise decken. Wir haben gesehen, dass sich darunter typische Anpassungscharaktere finden, die den Lebenszyklus der Form auf den Jahreszeitenzyklus der Natur einstellen, und ferner andere Charaktere, deren Anpassungseigenschaft in gleicher Richtung vermutet werden muss, ohne dass sie ohne weiteres sichtbar ist. W i r haben ferner gesehen, dass genetisch an der Verursachung dieser reihenweise geordneten Erbunterschiede alle Typen mendelnder Vererbung beteiligt sind, einfache und geschlechtskontrollierte Vererbung, multiple Allele, polymere Gene und endlich auch zytoplasmatische Unterschiede, sowie solche in der Chromatinquantitat. Nunmehr fragt es sich, welche Schlusse aus diesem so vollstiindig analysierten Fall fiir das Hauptproblem, die Evolution, gezogen werden sollen. Die entscheidende Frage ist dabei die, ob die Ausbildung der geographischen Rassen innerhalb einer Art oder eines Rassenkreises als Vorstufe des Artbildungsvorgangs betrachtet werden kann, oder doch wenigstens als Model1 fur diesen Vorgang. Wenn wir uns fragen, was auf Grund unserer Analyse vom genetischen Standpunkt aus den Vorgang der geographischen Variation am besten charakterisiert, so ist es zweifellos das, dass sich durch viele Mutationsschritte teils der gleichen Gene in Form von Reihen multipler Allele, teils mehrerer den gleichen Charakter beeinflussender Gene, also in Form der Ausbildung polymerer Verschiedenheit, quantitativ verschiedene, in einer Reihe anzuordnende Erbtypen einer oder mehrerer Eigenschaften gebildet haben, die fur bestilnmte aussere Bedingungen Anpassungswert besitzen-in unserem Beispiel Lebenszyklus und Naturzyklus koordinierenund damit die Besiedlung neuer Areale erlauben. Trotzdem die so entstandenenverschiedenheiten so weit gehen konnen, dass sie die normaleFortpflanzung zwischen zwei solchen Rassen unmoglich machen, bleiben alle entscheidenden Artcharaktere unberiihrt. Wenn wir uns die Entwicklung der Artcharaktere des Individuums auf Grund seiner Erbanlage so vorstellen, wie es in meiner "physiologischen Theorie der Vererbung" geschehen istll eine Vorstellung, fur die es n~einerAnsicht nach keine Alternative gibt, also als das in ihren zeitlichen Ablaufen genau dosierte Zusammenspiel von Reaktionsketten, die damit zu einem System verbunden sind, in dem sich jedes Glied zwangslaufig in das andere fiigt, dann gibt es nur zwei Typen von Veranderungen in einem solchen System: entweder Veranderungen, die das ganze System treffen und ihm einen neuen, veranderten Gleichgewichtszustand verschaffen; oder Veranderungen, die solche Glieder des Systems treffen, deren Ablaufsgeschwindigkeiten nach der einen " GOLDSCHMIDT, R. : Die quantitativen Grundlagen von Vererbung und Artbildung. Berlin: J. Springer, 1920 und Physiologische Theorie der Vererbung. Berlin: J. Springer, 1927.

INTERNATIOSAI, CONGRESS OF GENETICS

183

oder anderen Kichtung geandert werden kijnnen, ohne dass dadtirch das Gleichgewicht des Systems gestort wird. Natiirlich haben die letzteren Aenderungen aus ihrem Wesen heraus (grossere und grossere Reschleunigting oder Verlangsamung eines Reaktionsablaufs) die Neigung, orthogenetisch zu sein." E s scheint mir nun keinem Zweifel zu unterliegen, dass im Tierreich der Vorgang der Bildung geographischer Rassen cler zweiten Kategorie angehort, der Verschiebung einzelner Keaktionsketten innerhalb des Systems; dass aber die Artbildung auf Grund der ersten Kategorie erfolgt, durch Veranderung ganzer Teile des Systems. Die geographische Variation, deren genetische Grundlage jetzt in einem Fall bekannt ist, ware also weder eine Vorstufe noch ein Model1 fiir den Artbildungsvorgang, sondern nur ein Vorgang, der innerhalb der uniibersclreitbarei Cirenzen des Systems der gegebenen A r t durch relativ einfache genetische Vorgange Anpassungen innerhalb des Artbildes an allerlei verschiedene Umgebungen gestattet. Dass dern so ist, kann mit einern Beispiel belegt werden. Hier sehen wir ein paar nahe verwandte Arten der Gattung Lymantria, namlich die hier behandelte dispar, ferner monacha und mathura. Sie sind als Falter ziemlich verschieden und als Raupen noch vie1 mehr. Alle drei kommen nebeneinander in Japan vor, wenn auch in verschiedener Haufigkeit, und sie haben den gleichen Lebenszyklus, so gleich, dass ich in China einmal am gleichen Fleck Dutzende von dispar und mathura Q gleichzeitig nebeneinander ihre Eier ablegen sah Trotzdem hat es das System der dispar erlaubt, die besprochene Verbreitung durch Rildung geographischer Rassen anzunehmen ; der monacha dagegen nur eine nordliche Verbreitung in Ostasien und Verbreitling iiber ganz ELIropa, der mathura aber eine Verbreitung nur iiber das warme Asien yon Japan bis Indien. Aber keiile dieser geographischen Kassen nahert eine Art der anderen in1 geringsten an. Aher auch noch von einer ganz anderen Seite her kann tiiese Schlussfolgerung bestatigt werden. E s ist klar, dass in einem Entwicklungssystem, das auf dem Prinzip abgestimmter Reaktionsgeschwindigkeiten aufgebaut ist, nichterbliche Verschiebungen, also Modifikationen im Phanotypus, nur so weit moglich sind, als das ganze System dadurch nicllt aus dem Gleichgewicht gebracht wird. Genau das Gleiche aber erschlossen wir fiir die mutativen Veranderungen innerhalb des Artbildes, die allein fiir die geographische Variation in Betracht kommen. Nun ist es eine alte Erfahrung, die wir experimentell auch bestiitigen konnen, dass es bei Schmetterlingen moglich ist, rein modifikatorisch im Temperaturversuch einen Phanotypus zu "Diese einfache physiologische Erklarung tler Orthogenesis, die ich in der Schrift von 1920 (s. otxn 11) gab, ist merkwiirdigerweise nie beachtet worden.

184

PROCEEDINGS O F T H E SIXTH

erzeugen, der mit dem Typus bestimmter geographischer Rassen identisch ist; auch mit den meisten genetischen Unterscheidungsmerkmalen der Lymantriarassen konnen wir den gleichen Versuch erfolgreich machen, was eben zeigt, dass die betreffende Mutation nur die verschiebbaren Reaktionsablaufe innerhalb des Systems trifft. E s sei ubrigens dazu bemerkt, dass dies ebenso f u r die meisten sonst von den Genetikern studierten Mutationen zutrifft ; ich konnte im Temperaturversuch eine grosse Reihe nicht erblicher, rein modifikatorischer Phanotypen von Drosophila erzeugen, die nicht von wohlbekannten Mutanten zu unterscheiden sind. Die Parallele zum Vorhergehenden liegt auf der Hand. Wie aber kommen denn die grosseren Veranderungen in ganzen Teilen des Reaktionssystems zustande, die zu wirklicher Artbildung fuhren ? Bedeuten sie eine immer und immer wieder weitergefiihrte Haufung der kleinen Aenderungen mit allmahlichem Umbau des Systems ? Oder bedeuten sie mutative Aenderungen an kritischen Punkten des Systems, also an embryonalen Segregationspunkten, die mit einem Schlag das ganze System verandern? Oder gehoren dazu noch Aenderungen des Substrats der Reaktionsketten, also des Plasmas? Wir miissen ehrlicherweise gestehen, dass wir uns dariiber allerlei Gedanken machen konnen, aber zunachst im Tierreich keinen Weg zur experimentellen Inangriffnahme sehen. Dass es trotzdem solche Wege geben muss, sei zum Schluss a n einem Beispiel gezeigt. Bei Lymantria dispar hat das 8 langgefiederte Antennen, das 9 kurzgefiederte. In der Entwicklung der mannlichen Antenne kommt ein St'adium vor, in dem die innere Fiederreihe noch kurz ist, die aussere lang. Das gleicht sich erst spater durch Wachstum aus. Bei intersexuellen 8 aber kann dieses Stadium dauernd erhalten bleiben, wenn der Drehpunkt fur die Geschlechtsumwandlung gerade in jenes Entwicklungsstadium fallt.13 Bei intersexuellen 9 kann ferner ebenfalls aus bestimmten embryologischen Griinden die Bildung von Platten anstelle von Fiedern eintreten. Nun ist der fiir das intersexuelle 8 gezeigte Zustand typisch fur das normale 8 einer ganz anderen Art Orgyia antiqua, und ebenso der Zustand des letzten Bildes fur die weit entfernten Cossus-Arten. Was also bei Intersexualitat durch abnorme Storung des Zusammenarbeitens der Reaktionsablaufe als pathologisches Produkt hervorgebracht wird, wird bei jenen Arten typisch durch ihr normales Reaktionssystem zustande gebracht. Daraus folgt: wem es gelingt, eine dem Beispiel entsprechende Veranderung im Ablauf einer Gruppe determinierender embryonaler Reaktionsketten herbeizufiihren, ohne dass das ganze System in Unordnung gebracht wird, der hat eine neue Art erzeugt. -Einzelheiten bei R. GOLDSCHMIDT: Untersuchungen iiber Intersexualitat I-V. Zs. ind Abst.23-54: 1920-30.

I N T E R N A T I O N A L CONGRESS O F GENETICS

185

CAN EVOLUTIOK BE E X P L A I N E D I N T E R M S OF KNOWN GENETICAL FACTS ? J . B. S. Hddane, John Innes Hortict~ltural Instifntion, Alerton. Park, London, England

The facts to be explained are mainly derived from two sources. On the one hand the paleontologists describe successive faunas, and in some cases are able to construct extremely plausible lineages. The best of such lineages, such as the Micrasters of the English chalk, show quite continuous evolution, the populations at successive levels overlapping. Where, as in the case of the horse, apparent discontinuities exist, it is possible to postulate continuous evolution in an unexplored area. But many of the trends displayed by such series are hard to explain. Such a trend is the tendency to evolve certain characters, for example, large size, whose full development was a prelude to extinction. Equally difficult t o explain is the trend we11 shown in certain Ammonite lineages and deduced elsewhere by embryologists, for tachygenesis, that is to say, the appearance at progressively earlier stages in the life cycle of a character at first only found in the adult. Thir; leads to the phenomenon of recapitulation, which could readily be explained on such an hypothesis as racial memory, were any genetical facts available to support such a hypothesis. The systematists, on the other hand, describe groups of closely related species of such genera as Crepis or Partula, each with its characteristic h a b'itat. It is moreover hard to escape from the arguments brought forward by such workers as WILLISthat many of these species have originated quite suddenly within recent times. Thus while the paleontologist often proves, and never disproves, continuous evolution the systematist makes the existence of discontinuous evolution almost certain. It may be that these two types of evolutionary phenomena are due to different causes. It is at least noteworthy that, while the paleontologist deals with species which were very common in their own time, and often lived in a constant environment such as the sea bottom, the systematist is mostly concerned with rare species, often occupying rather small ecological niches. The two types of species, dominant and rare, may therefore evolve under rather different sets of causes. The answer to our original question must fall into two parts. First, are interspecific differences similar to intervarietal, and thus explicable by the accumulation of differences of known types? Second, if the first question is answered in the affirmative, can we account for observed evolutionary trends ? When we are able to analyze the differences between two species we oc-

186

PROCEEDINGS OF THE SIXTH

casionally find cases which are fully explicable by gene differences. Thus CHITTENDEN analyzed the differences between Primula acaulis and P. Jdiae. The yellow flower pigment of the former and the purple of the latter are each determined by separate dominant genes. The F1 possessed both pigments, and approximately one-sixteenth (four out of sixty-eight) of the F, had neither and were therefore white. Other differences seemed to be due to relatively small numbers of genes. In mammals the color differences between related species and subspecies (Mus musculus and M. bactrianus; Mus rattus rattus, M. rattus alexahdrinus and M . rattus tectorum; Cavia porcellus and C. rzlfescens) are found to be due to small changes at the a (black) and e (recessive yellow) loci, less pronounced than those found in mutants of domestic forms such as the black mouse or the cream Mus rattus. This is so even when, as in the case of Cavia, the species are so far removed from one another as to give sterile male F, hybrids. GREEN'Sanalysis of the above mouse cross not merely makes it probable that the size difference between these species is due to genes, but identifies the chromosome carrying one of these genes. In plants at any rate cytoplasmic differences between species sometimes occur, but the work of BLEIERshows that they may be negligible. In hybrids between Lens esculenta and Vicia faba (the latter as pollen parent) the Lens chromosomes are eliminated, and the resulting plants can not be distinguished from Vicia. Commonly, however, the results of species crossing are dominated by cytological differences. These may be in the number of chromosomes (for example in polyploid series), in the arrangement of the chromatin (for example, segmental interchange in Datura species) or both. Both these types of difference can be produced experimentally within a species. Further, by hybridization new amphidiploid species can be made, such as Primula Kewensis, or old ones recreated, as in MUNTZING'S synthesis of Galeopsis tetrahit. More complex results can also be obtained by hybridization, as synthetic Viola amtensis. in the case of CLAUSEN'S W e can to some extent extrapolate this analysis to the case of species which will not cross. This may be done, as by ANDERSON in Dianthus, by the use of a third species as a bridge, by comparing the chromosome arrangement, as in different Drosophila species and so on. There is no suggestion of a qualitatively new type of difference not found in more closely related species. Semisterility of hermaphrodites or of both sexes, and sterility of the heterogametic sex, may be produced between varieties by chromosome rearrangement. Death of all F, hybrids has not yet been observed between

INTERNATIONAL CONGRESS O F GENETICS

187

varieties, but almost all the other phenomena once thought to be characteristic of species crosses have now been observed after varietal crosses. There is no strong reason to doubt that DARWINwas correct in regarding varieties as incipient species. Unless some genotypes are fitter than others, a population in which polymorphism is due to genes is in equilibrium, apart from effects of mutation and random extinction. The latter can be shown, to be unimportant as an evolutionary agent in large populations, but may have acted on small populations (for example, in oceanic islands). Mutation, though needed t o account for initial polymorphism, will not cause a gene to spread through a population in the face of any but the very feeblest natural selection working in the opposite direction. I t would seem that the main burden o i causing evolution must be thrown on natural selection. Observation shows that it occurs. Calculation shows that it would produce changes fast enough to account for the speed of evolution recorded in the rocks. But many difficulties remain. Individual mutations are generally harmful, as is inevitable if a species is in equilibrium. Otherwise the mutant forms would displace the type. And a successful change in such an organ as the eye requires the compresence of several mutant genes (coaptation). GONSALEZ described a case in Drosophila where a double recessive had a longer expectation of life than either single recessive. In such a case more than one stable equilibrium is possible in a population. I t is possible that species of a related swarm may represent different stable equilibria, each having a combination of genes such that changes in only one or two at a time would lower fitness. T o pass from one such equilibrium to another, either an unrepresentative fraction of a population must be isolated, or as the result of a chromosomal rearrangement two genes must come to be so closely linked as to be inherited together. The possibilities of gene recombination are enormous. Assuming 400 loci in Drosophila and only two allelomorphs at each (both underestimates) we should have Z 4 O 0 possible combinations. I t is highly probable that many of these are fitter than the wildtype. The evolutionary appearance of apparently useless characters can often be tentatively explained by the multiple effects of genes (DARWIN'S"correlation") and the fact that apparently irrelevant genes restore a physiological balance. Thus in order to be hairy Matthiola incana must have colored flowers. If arc wing were selected for in Drosophila rnelanogaster the addition of an axillary spot would increase the fitness. Characters actually detrimental to a species may spread as the result of

188

PROCEEDINGS OF THE SIXTH

natural selection when competition is largely between members of the species, that is, whenever there is crowding a t any stage of the life cycle. Thus where a gene for a somatic character slows down pollen-tube growth (as in Zea and Oenothera) such a somatic character will be eliminated even if moderately advantageous. And there will be no compensatory advantage to the species due to the gain of a few hours between pollination and fertilization of the ovules. Competition,between male animals may be expected to lead to an increase in size which may be detrimental to a species for other reasons. Again in polytokous mammals we may expect that genes favoring rapid embryonic growth will spread, as slower growing embryos are a t a disadvantage. Such genes will not necessarily or even generally produce characters advantageous in the adult. W e can thus understand the orthogenetic evolution of characters which make the species as a whole less fit and may lead to its extinction. This tendency will however be in abeyance in a relatively rare species where competition usually occurs with other species rather than between members of the same species. Where characters are very nearly neutral from the point of view of fitness, mutation may be expected to determine evolutionary trends. I t is capable of explaining the slow reduction of useless but not harmful organs, such as the pelvis of the whale. I t can also explain certain cases of tachygenesis, for the following reason. Wild-type genes will tend to increase in pointed out, this dominance for two causes. On the one hand, as FISHER will lead to partial protection against harmful mutations ; on the other hand, mutation in this direction will not be prevented by natural selection. Accorddominant genes produce their effects more rapidly than ing to GOLDSCHMIDT recessives. The characters produced by them will thus tend to appear progressively earlier in the life cycle. While we cannot yet explain all evolutionary phenomena in terms of known genetical facts, the number of phenomena so explicable increases every year, and there is no sign that the possibilities of explanation are reaching a limit. W e may reasonably hope for a fairly complete explanation. It would seem that we must envisage the possibility that there are two rather different types of evolution. The first, primarily studied by the paleontologist, is that of dominant species in a fairly stable environment. Such species change slowly by the gradual spread of genes each with a relatively slight effect. I t is hard to study in practice, but can be theoretically treated on lines laid down by WRIGHT,FISHER,and HALDANE. Such species are peculiarly liable to unfavorable evolutionary changes. The second type is characteristic of species whose members exist in quite

INTERNATIONAL CONGRESS O F GENETICS

189

small and nearly or quite isolated groups. Such a group may undergo a cytological change or a change in several genes a t once. Such changes, while they must ultimately stand the test of natural selection, are not themselves due to natural selection. The new variants so produced are not likely, as in the first type, to be swamped by hybridization before they have a chance to develop into new species. Nor is the probability of harmful orthogenetic evolution due to competition so great. Until recently man was divided into small tribes, and human evolution was predominantly of the second type. Under modern conditions of large communities and increased mobility man is becoming subject to evolutionary influences of the first type.

190

PROCEEDINGS O F T H E SIXTH

ON T H E POTENCY OF M U T A N T G E N E S A N D LVILD-TYPE ALLELOMORPHS Otto L. Mohr, Anatomical Institute, The University, Oslo, Norway

As a starting point for the following considerations we shall use the p e r h a p m o s t familiar of all Drosophila mutants, the school example of a typical recessive, vestigial wing, found by MORGAN in 1911. Most striking in vestigial is the enormous reduction of the wing, due to the trimming away of the marginal and terminal regions. This reduced wing stands out a t right angles. The balancers, especially their terminal segments, are reduced in size, the posterior scutellars are erect, and the viability of the fly is somewhat below standard (BRIDGESand MORGAN1919). Thus, vestigial is a typical representative of genes with manifold effects. Some years ago it was noticed that both the black purple vestigial Lobe and the pure vestigial laboratory stock cultures contained some flies which failed to manifest the vestigial character. In the former stock the exceptional individuals had full-length, slightly divergent wings with large marginal incisions ("ragged" wings) and in the latter wild-type wings with a tiny terminal nick in rare cases (figure 1 ) . The cause of these changes was not analyzed a t the time, and the stocks were propagated in the ordinary way without selection as before. Not until about one year later, when pure vestigial flies were to be used in a mating, was the case again brought to the foreground of attention. Practically all o f the supposedly pure vestigial stock flies were now found to be wild-type, except for an occasional terminal nick in the wing, and when the black purple vestigial Lobe stock was reexamined, not only the vestigial, but also the ragged wing type had disappeared entirely! All the flies had long, normallooking wings, except for distinct terminal notches in quite a few cases. Appropriate tests proved that these changes in the stocks were due to secondary, regressive mutations. In the multiple stock vestigial had mutated to a new allelomorph, denoted as notched (v;'") which in homozygous condition ~)roclucesterminal notches in about 40 percent of the cases. In compound with vestigial this gene gives the ragged wing type, encountered in this stock one year previously. Correspondingly, in the pure vestigial stock vestigial had tnutated to another allelomorph, denoted as nicked (ugni). When homozygous this gene has no somatic effect, but compound nickedhestigial flies have tiny terminal nicks in about 30 percent of the cases. In passing, it may be mentioned that from the first detection of a few exceptional flies it was not long hefore the new allelomorphs had entirely driven out and

INTERNATIONAL CONGRESS OF GENETICS

191

FIGURE 1.-a, Compound nickedjvestigial male. b, Homozygous notched female. c, Compound notched/vestigial ("ragged") male.

192

PROCEEDINGS O F THE SIXTH

replaced the vestigial gene in the respective stocks, striking cases of "natural selection in vitro." Three other supra-vestigial and infra-wild-type allelomorphs have earlier and MORGAN ( 1919), namely, nick (vsn) ,antlered been described by BRIDGES (vga), and strap (wg8), the former found by BRIDGES,the two latter by MORGAN(see figure 2). Stocks of these are no longer in existence. Nick is very much like our nicked; homozygous nick flies are wild-type, while 65 percent of compound nick/vestigial flies have terminal nicks. Antlered produces large marginal incisions and the wings stand out a t an angle of 30". Strap, finally, causes still more pronounced alterations, the widely divergent, "leg-of-muttonv-shaped wings resembling those of the longwinged vestigial flies appearing under special temperature conditions (see below). The balancers are in strap somewhat reduced, and, judging from some illustrations, the posterior scutellars are slightly erect. In addition to these allelomorphs there are about an equal number which produce changes that are farther removed from normal than those present in vestigial. ( F o r stocks as well as for information dealt with below I am indebted to BRIDGES.)Nearest to vestigial comes No-wing, found and studied by MORGAN(1929). This gene which reduces both wings and balancers almost completely is treated by MORGAN as a recessive. But in our analysis of different vestigial allelomorphs (see below) it turned out that a- terminal break in the second longitudinal vein belongs to the vestigial characteristics, and this alteration is present in 70 percent of heterozygous No-wing flies, which also have a terminal nick in the wing in rare cases. No-wing (vgNw) is accordingly here considered as incompletely dominant. Homozygous Nowing flies have erect scutellars. The viability is extremely low. No-wing connects vestigial with a series of incompletely dominant allelomorphs which are lethal in homozygous condition and which when heterozygous remove bites out of the wing margin (MORGAN, BRIDGES,SCHULTZ 1930). As representatives of these allelomorphs, which have not yet been analyzed in detail, may be mentioned vestigial-Beaded (vgBd) found by BRIDGES,Carved (vgC) by DEMEREC,Snipped (vgS") by MULLER(1930) and Depilate (vgD) by BRIDGES. Summing up, we are accordingly confronted with a very extensive spectrum of effects of uni-local genes. At one end subliminal, recessive members are found which closely approach the wild-type gene in potency, at the other dominant members which in homozygous condition cause such great changes that the zygote is non-viable. Between these extremes a graded series of intermediate steps occur. The series illustrates in a striking way that dominance and recessiveness are relative differences only.

INTERNATIONAL CONGRESS O F GENETICS

PROCEEDINGS OF T H E S I X T H

191

For a comparative study of the potency of the vestigial allelomorphs those members were used of which stocks were at hand, namely, No-wing, vestigial, notched, nicked and the wild-type allelomorph. Their relative effectiveness (at 25" ; standard banana-agar culture medium) in producing marginal incisions, divergent wings, rudimentary balancers and erect scutellars was the object of this analysis. When it later turned out that a terminal TABLE 1 Summary of quantitative evidence on the effect o f the vestigial-allelomorphs No-wing (voNw), vestigial (v,), notched (v0no), nicked (voni) and the normal allelomorph ( + w ) . The numbers in parentheses represent totals on which the corresponding percentages are based. INCISED GENOTYPE

WINQB

Percent

DIWBQENT WIN08

8AORTENED

ERECT 8CwTELLAIlE

BALANCER8

VIABILITY IN

2 VEIN8

RELATJON TO 0g

Percent

Percent

.. 0 0 .. 2.6 .. (17 :670) 0 .. 0 .. (0: 765) 73.3 .. (127:176) 8.4 111.8 (64:760) (1062 :950) 0 115.6 (0:757) (267 :231) 100 .. (138: 138) 2.1 104.9 (10:468) (450: 429) .. 100 (156: 156) .. uncontrollable uncontrollable .. uncontrollable ..

0 nicks 0 . 2 (3: 1206) nicks 1 . 3 (10: 794) nicks 27.1 (288 :1062) notches 42.4 (479:1162) scalloped 70.7 (606 :857) ragged 100 (2495 :2495) antler-like 100 (156: 156) stumps 100 stumps 100 absent 100 -

-

--

-

break in the second vein belonged to the vestigial characteristics, data were also secured on the frequency of this alteration in the different compounds. The results are summarized in table 1. I n combination with the wild-type gene, nicked, notched and vestigial have no detectable somatic influence. That the vestigial/wild-type (vg/+ ) combination nevertheless has another potency than the homozygous wild genotype (-I-/+) is evidenced by DEXTER'S(1914) finding that heterozygous vestigial enhances (as do also heterozygous antlered or strap) the effect of the incompletely dominant Beaded, which by itself produces mar-

INTERNATIONAL CONGRESS OF GENETICS

195

ginal incisions in the wing blade. Homozygous nicked flies have wild-type wing margin, and so have also the nicked/notched compound flies. Only in very exceptional cases there occurred in the latter combination a slight and doubtful indication of nicks. The most extreme allelomorph used, No-wing, gave in combination with the wild-type gene terminal nicks in 1.3 percent of the cases, and from here the frequency and degree of marginal incisions increase gradually in the nicked/vestigial, homozygous notched and nicked/No-wing ("scalloped") combinations (see table 1 and figure 3 ) . When notched is combined with vestigial the threshold for constant production of large marginal incisions (ragged wings) is reached, and in this compound, especially in earlyhatching females, we also encounter the first indication of divergent position of the wings, erect scutellars and reduced balancers. In the "antler-like" notched/No-wing compound all these characteristics are more pronounced and constant. In homozygous vestigial and in the vestigial/No-wing compound the wings are reduced to stumps, while finally in homozygous No-wing both wings and balancers are practically absent (see figure 3 ) . All this evidence may be accounted for if we assume that the allelomorphs studied form a quantitative series in the following order of potency: Nowing, vestigial, notched, nicked and the wild-type gene. But a g!ance at the summary (table 1, sixth column) makes it clear that, with respect to their influence on the fifth characteristic under control, the second vein, they behave quite otherwise. Here notched and nicked have changed places in the series, notched approaching more closely the wild-type allelomorph in effect. Vestigial and No-wing have retained their order, but their difference in potency is much more marked than was the corresponding difference in their influence on the other vestigial characteristics. (1925) and with STERN(1929) in his If we, in analogy with WRIGIIT analysis of the additive effect of the bobbed allelomorphs, give the different allelomorphs relative numerical values, which as well as can be approximated cover the experimental evidence, we arrive at the following relative values for the tendency to production of marginal incisions: , t g ~ w ,6; vg, 10;vgn0, 15; van', 22; + vu, 30. When these values are added in the respective diploid combinations the numerical series presented in table 2, third horizontal row, results. The threshold for the first and doubtful indication of nicks lies at 37. With decreasing values an increasing percentage of nicked, notched and scalloped wing types occur, until at 25 the threshold is reached for constant production of the ragged phenotype. At 20 the wings are reduced to stumps, at 12 practically absent. Modal wing-types and balancers from the different combinations are presented in figure 3.

TABLE 2 Numerical series illustrating the effect of the vestigial allelomorphs studied. For exfilanation see text. GENOTYPE

TgNw/toNw

Character of wing

absent

uy,hgNw

g0/ng

stumps

VgnO/UgNw

v ~ ~ tgn+/UgNw ~ / v ~ . J ~ ~vgni/Vg ~ / c ~+vg/UpNw ~ ~

antlerlike

ragged

scalloped

notches

nicks

Ugni/Ugn0

fVO/~g

Ugn'/Vgni

++Vg/~gno

wildtype

wildtype

wildtype

+ug/Ugni

wildtype

fvg/fvg

nicks

nicks

wildtype

1.3

0.2

0

0

0

0

0

37

40

44

45

52

60

Percentage of incisions

100

100

100

100

100

70.7

Numerical value

12

'16

20

21

23

28

Divergent wings

++

++

++

+

+-

-

-

-

-

-

-

-

-

-

-

Erect scutellars

++

++

++

f

+-

-

-

-

-

-

-

-

-

-

-

Rudimentary balancers

+++

++

++

+

+-

-

-

-

-

-

-

-

-

-

-

Vg/vgNw

Dgni/UoNw

+vg/UT

+vg/+wg

GENOTYPE

Percentageof shortened 2 vein Numerical value

DgNw/UgNw

uncontrollable

6

uncontrollable

13

100

17

Up/Up

uncontrollable

20

Ilgno/FgNw

+vg/UoNw

100

73.3

22

23

42.4

27.1

30

32

ugni/ug

8.4

24

36

ugni/Foni

2.6

28

lgn0/UO

fVg/Up

Ulno/Ugni

+vg/~g"i

Ugno/Ugno

2.1

0

0

0

0

0

0

29

30

33

34

38

39

40

198

PROCEEDINGS OF THE SIXTH

For those characteristics which are correlated with the marginal incisions the threshold for the first deviation from wild-type lies considerably lower, namely, at 25 (table 2 ) . From here the degree of these changes increases gradually with gradually decreasing values. But with respect to the remaining characteristic, the vein-shortening, the relative values which correspond to the experimental data are different, namely, V g N w , 3 ; v,, 10; v , n i , 14; v P n o , 19; + U P , 20; and the resulting numerical series as in table 2, last row. The threshold for the first deviation from wild-type lies here at 29 and that for constant production of the vein-shortening a t 22. The above numerical values are used for the sake of illustration in order to bring out more clearly the relative difference in potency among the genes studied, and the series have clearly the character of a rough approximation. But their general trend corresponds well with the experimental evidence. Not until the collected data were analyzed was it realized that the material lent itself to such a treatment as that above. (The data on the vein-shortening in the No-wing compounds have later been increased. These additional data are in line with those presented.) I t then turned out that the vein character of one particular compound, the nickedhestigial, had not yet been investigated. By aid of the numerical value of this compound it was then predicted that it was to be expected that this would have the vein-shortening in a percentage of cases lying between 73 percent and 3 percent. The low percentage actually found, 8.4 percent, is in accordance results (1929) which indicate that the potency curve rises with STERN'S very smoothly when it approaches the threshold for the production of the wild phenotype. This is also the case with respect to the marginal incisions (see table 2 ) . In connection with the above evidence the results of ROBERTS(1918), HERSHand WARD(1932) and especially of HARNLY(1930) on the influence of temperature on the development of the wings of vestigial flies found that a rise in temperature from 29" to should be recalled. HARNLY 31" had a striking effect on the wing length of vestigial males, while a rise from 30" to 31" had an analogous effect in the female. A t the critical temperature flies occur with wings resembling strap, antlered, cut or even Beaded. Correspondingly, the divergent position of the wings and the erect position of the scutellars were in the males changed toward the wild-type. The latter point is interesting since in the different genotypical combinations the correlated vestigial characteristics are generally more pronounced in the females. Thus, a critical rise in temperature modifies vestigial into phenotypes which approach the normal and which parallel those which we, within the

INTERNATIONAL CONGRESS O F GENETICS

199

refractory temperature limits, are able to produce by combination of different vestigial allelomorphs. When we remember the slow development of ordinary vestigial flies, analyzed by HARNLY(1929) and by ALPATOV (1930) this situation falls we11 into line with GOLDSCHMIDT'S conception of Abgestimmte Reaktionsgeschwindigkeiten, and the effect of the vestigial allelomorphs on the correlated characteristics may also be accounted for by quantitative interpretation. But when we also the aid of GOLDSCHMIDT'S consider their influence on the vein-shortening, we are up against the same ( 1930) in his anaIysis of the three Stubble allelodifficulties as DOBZHANSKY morphs, which, according to potency, range in one order with respect to influence on bristle length, and in another.as regards their effect on size of wings and length of legs. As emphasized by DOBZHANSKY, this disproportionateness of the effect on different characteristics seems very hard to bring in accordance with GOLDSCHMIDT'S theory in its simple form, even though we are aware that between the primary reaction and the end result there lies a long chain of intermediate reactions. I n the above the normal allelomorph was given the highest potency value. Correspondingly among the mutant genes those which produce the slightest deviation from standard, that is, which in their effect most closely approach the wild-type gene, have been considered most potent, and vice versa. In order to obtain further light on this point we are going to consider another line of evidence, namely, the behavior of mutant and normal genes in deficiency where we are able to study their effect in single quantity, that is, without any normal or mutant partner gene being present at the same time. That a deficiency represents an actual loss of a chromosome section may be regarded as established in view of the combined genetical and cytological evidence of BRIDGES(1921) from the haplo-IV case and of PATTERSON and PAINTER (1931) from the mottled notched X-IV translocation. In his above-mentioned analysis STERN(1929) bases the evaIuation of the 'normal allelomorph on the "soweit bekannt, allgemein giiltigen Tatsache, dass ein Faktorenausfall (that is, deficiency) auf die Auspragung eines normalen Allels keinen Einfluss hat, dass also die Wirkung eines einzigen normalen Allels noch zur Erzeugung des Normaltypus ausreicht." WhiIe this may be true in the bobbed case, it cannot, as we shall see, be the general rule. I n table 3 is collected evidence from deficiency cases in Drosophila melanogaster studied by different investigators. For quite a few only preliminary accounts are as yet available so the data may be subject to later completion. Still, a comparison of the cases brings out some outstanding facts. Firstly: Most deficiencies are accompanied by dominant character changes.

PROCEEDINGS OF THE S I X T H

200

This may simply be due to the fact that deficiencies with dominant changes are most easily detected, and BRIDGES'(1917) forked-Bar deficiency forms a clear exception to the rule. STURTEVANT'S reverted Bar, which represents a short deficiency at the Bar locus, comes in a special category since STURTEVANT (1925, 1928) has demonstrated that there exists no normal allelomorph at the Bar locus. The dominant changes are either as in Notch, Gull, in BRIDGES'Plexate (MORGAN, STURTEVANT, BRIDGES1927) or in a new TABLE 3 Comparison of deficiency cares in Drosophila naelanogaster. CEROMOBOMB

INCLUDED MUTANT LOCI

A 2, At, A pnt w,4 6 , f a t Az, A ec Left end to ec (14 loci) Spb, fa,

W , S P ~ f,a ,

co

Y, S c , ac

hkr Pr Pt,

(Bi),

E X ~ N T ~

-

f, B

hkt

DOMINANT ~ N Q E S

It

p, to right end (12 loci)

Notch Notch Notch Lethal (Notch) Minute (M-30) Bristle character Gull Plexate Minute (MI) Minute (MI,) Minute (Def. "H") Lethal (Def. "G") Lethal (Pale-Def .) Vein

IV

all IV (7 loci) Itt C i

Minute (Haplo IV) Minute (M IV)

111-chromosome deficiency Vein quite like those typical of any ordinary dominant gene, or they belong to the well-known dominant Minutes. Secondly: The dominant changes are as a rule different from those produced by the included mutant genes. This relation is so striking that it may help in deciding between the alternatives lethal allelomorph or deficiency in cases where only one mutant gene is included (see MOHR1928). For this reason Minute-30 of SCHULTZ(1929), Minute-IV of BRIDGES(MORGAN, STURTEVANT, BRIDGES1926, 1928) and Vein have long been regarded as deficiencies in spite of only one gene being included, BELIAJEFF'S(1931) rotated abdomen in Minute-IV (MORGAN, STURTEVANT, BRIDGES1926, 1928) and BRIDGES'divergent in deficiency Vein. The legitimacy of this

INTERNATIONAL CONGRESS O F GENETICS

201

view was proved when recently both a new recessive, javelin ( j " ) , located half a unit to the left of divergent, and MULLER'Smoirk (illo) were found to be included in Vein, and correspondingly cubitus interruptus in MinuteIV. The latter observation was independently made by STURTEVANT. Without the above criterion the deficiency interpretation remains doubtfuI in the so-called vermilion-deficiency case (BRIDGES1919) and also in the case of the dominant Truncate which, as shown by MULLER,acts as a lethal member of the Truncate allelomorph series (see MOHR1928). Thirdly: All known cleficiencies are lethal when homozygous, and from cases of translocations we know that several are lethal also in heterozygous condition. It is not justified, as has frequently been done, to correlate this heterozygous lethal action with the extent of the deficient section. DOBZIIANSKY'S (1930b) short deficiency "G" (see RIIOADES1931) is lethal when heterozygous, while much longer Notch deficiencies and deficiency Vein (1932b) has demonstrated that a very are not (see below). PATTERSON limited deficiency to the left of prune in the X chromosome is lethal when heterozygous. Clearly, the heterozygous lethal effect is not due to the number, but to the special quality of the genes involved. Fourthly: In combination with the deficiency the included mutant genes are exaggerated in their effect (MOHR 1919, 1923). With increasing evidence, it has been all the more firmly established that the character is in the compound shifted away from, not toward, the wild-type. Occasionally an included mutant gene fails to exhibit exaggeration effects. This applies to yellow in L. V. MORGAN'Syellow-scute-achaete deficiency (MORGAN, Notch 172b and to STURTEVANT, BRIDGES1927), to prune in PATTERSON'S deficiencies "G" and "H" (DOBZHANSKY 1930b, purple in DOBZHANSKY'S MORGAN, BRIDGES,SCHULTZ1930). This relation is readily explained as due to the homozygous effect of the gene in question representing by itself the bottom of the differentiable mutant changes possible at this locus (see MOHR1928). This conception is supported by the fact that in none of these cases have more extreme mutations at the same locus been encountered. T o my knowledge only a single case forms here an apparent exception. But since the stock in question had been lost before the exaggeration phenomenon had been detected, it is not now possible to decide what importance should be attributed to this case. After the establishment of the above rules we may now turn to the question of their explanation. The best known and frequently recurring deficiency is the sex-linked Notch, which has been encountered more than 30 times in untreated ma-

202

PROCEELIINGS OF THE SIXTH

terial and apparently still more frequently in PATTERSON'S ( 1 9 3 2 ~ )X-ray experiments. The majority involve short sections including the facet locus (at 3.0+), but some, such as Notch 8 (MOHR1919, 1923) or Notch 18 found by BRIDGES(see LI 1927, LI and BRIDGES1929), extend far enough to the left to include white, while in still others involved in PATTERSON'S chromosolnal aberrations (see PATTERSOW 1932a, 1932b, 1932c), the entire left end of the X chromosome to and including echinus (at 5.5) has been removed. Such a deficiency involved in the mottled notched transloca1932a) is lethal for the zygote, but somatic mosaic tissue tion (PATTERSON resulting froin elimination of the translocated fragment later in development is viable and Notch. Now PATTERSON (1932b) was able to demonstrate that this zygotic lethal effect is due to the very limited section between scute and prune not being present in duplicate. If the deficiency is combined with a fragment containing the loci for yellow and scute, the zygote is non-viable. But if it, as in Notch 172b, is combined with a Theta fragment which in addition contains the normal allelomorph of broad ( b r ) , then the zygote is viable and Notch (see concludes diagram, figure 4). From various tests of this type PATTERSON that the small section between scute and prune contains a "gene for viability" which must be present in duplicate if the zygote is to survive. A comparison of these Notch deficiencies of very different length (figure 4) leads to the following conciusion: Genetically, the only thing which they have in common is that a short section around facet is not present in duplicate. Still, disregarding the totaI extent of the haploid section, the phenotypical result is essentially the same in all, namely, the Notch complex: the thickened veins with A-like insertions, the notches, the supernumerary, irregularly arranged acrostical hairs and the bristle irregularities (figure 5 ) . This was confirmed by special experiments with four short (N26 to 29) and two long (N8 and N172b) Notch deficiencies. This can only mean one thing: In the short section mentioned there must be present one or more normal allelomorphs which when present in single quantity in the female are not potent enough to produce the wild phenotype. They are haplo-insufficient. Accordingly the dominant character complex results. It is not justified to identify as has repeatedly been done the locus responsible for the Notch character with the facet locus proper. I have re(ROKIZKY1930) cently tested the mutant split bristles ( s p) of DUBININ with the above-mentioned four short and two long Notch deficiencies, and split, located a t 2.7f,proved to be included in all of them. The same was also found to be true of NASARENKO'S (1930) incompletely dominant Ab-

INTERNATIONAL CONGRESS O F GENETICS

203

rupt-X ( A ), located very close to split (no crossing over in more than 2000 flies). Neither split nor Abrupt-X is allelomorphic to facet. Hence, what we know is that when the very short spb-fa-A2region with its contents of normal allelomorphs is not present in duplicate, then the Notch character results. The left-hand adjacent section, to and including prune, and the righthand section, to and including echinus, can not contain analogous haplo-

X

N 28

N 8; 3.8

N 172b,>40

&---.-

br ~n

Ax

=,Pbfa: : !

w

:

y

b r pn

+,

w

I

4

Lethal

I

fa: I

I

1

so

:

I

SC

A

I

,

A

ec

Lethal in zygotes N m m o ~ a ttissue c

hlotdh!

FIGURE4.-Diagram

of the X chromosomes in five different Notch deficiencies. In the three most extensive a loss of the left end of the X, to and including echinus ( e , ) , is combined with attached fragments of different length, as indicated to the left in the diagram. Above: Map of the left part of the X chromosome. The relative location of the closely linked genes spb,fa and A, is provisional. According to later information from NASARENKO A, is located close to the right of fa (1 crossover in about 5,000 flies).

insufficient normal genes since the absence of one representative of these additional sections does not alter the Notch phenotype. Deficiencies in these sections are accordingly not expected to produce dominant somatic character changes. But between scute and prune we encounter another haploinsufficient normal allelomorph, which, if left alone, induces such great changes that the zygote is as a rule non-viable. This is in our opinion the gene for viability. nature of PATTERSON'S An analysis of the exaggeration of the commonly included mutant genes in the same six Notch deficiencies of very different length proved that the

204

PROCEEDINGS O F THE S I X T H

exaggeration is also of the same type and degree in all, entirely disregarding the extent of the section involved. This indicates that the exaggeration can not be due to the difference in ratio of plus to minus modifiers within the lost section as compared with that of the normal X chromosome (BRIDGES1921, 1922, 1923) and gives support to GOLDSCHMIDT'S (1927, 1928) view that the exaggeration is due to the mutant gene being present in single quantity only, with a corresponding shifting of the character further away from the wild-type. This relation which also holds in cases of autosomal deficiency demonstrates that the exaggerated mutant genes in question are more potent in duplicate than singly. I t may of course happen that neighboring included haplo-insufficient normal genes may produce character changes which modify the effect of a particular included mutant gene. An illustration of this relation forms in my (1930) opinion the Notch/Abrupt-X compound studied by NASARENKO where Notch and Abrupt, which separately produce opposite character changes, neutralize the effect of each other so that Notch is suppressed and the exaggerated Abrupt character somewhat less pronounced than in homozygous Abrupt (figure 5). (In NASARENKO'S (1930) linkage tests with A2 there was a reduction in crossing over around the A, locus. This in combination with the exceptional character of the N8/A-, compound led NASARENKO to the hypothesis that the A= mutant is possibly due to a duplication. A,

In our linkage tests W

spb

the linkage values were normal.) An e,

rb

analogous situation is found in the haplo-IV eyeless flies in which the exaggeration of the eye character is almost negligible due to the opposite tendency of the haplo-insufficient normal allelomorphs which cause the enlarged eye of ordinary haplo-IV flies (see MOHR1932). In this special sense the change in balance may modify the end result, but it can not be the cause of the exaggeration proper. Already in the first account of the exaggeration of the included mutant genes the question was raised whether normal genes opposite to the deficient section were not influenced in a similar way, and it was stated (MOHR 1919) : "To some extent this seems to be the case. . . . I t is natural to suppose that these somatic .peculiarities (that is, the dominant Notch character complex) are a result of the modification of one or more of the normal genes in the region opposite to the deficient piece, similar to that which has been demonstrated in the case of the mutant genes. It is superfluous to regard the character Notch as due to an independent specific mutant gene contained in or linked to the deficient region.

INTERNATIONAL CONGRESS OF GENETICS

FIGURE 5.-a, X female.

205

Notch-28 female. b, Notch-28/Abrupt-X female. c, Homozygous Abrupt-

206

PROCEEDINGS O F T H E S I X T H

"It would seem probable that many normal genes are contained in such a piece of the X chromosome as that opposite to the deficient region. The fact that more extensive alterations are not caused when the deficient chromosome is present could perhaps be said to point in the direction that the normal genes must have different potencies. The mentioned mutant genes and some of the normal ones in the region opposite are affected, while other tnorlnal genes, sufposedly present in the same region, are not visibly influenced." The correctness of this at the time necessarily more tentative opinion has been confirmed by the now available additional evidence discussed above. W e may now try the application of this view to some additional cases: BRIDGES'Minute-1 deficiency (MORGAN,STURTEVAKT, BRIDGES1924, BRIDGES1930) covering the arc-plexus section of the second chromosome produces Minute bristles and female sterility. SCHULTZ'Sshorter aIlelomorphic Minute-1, deficiency closely to the right of plexus (MORGAN, BRIDGES,SCHULTZ1930) and overlapping the right-hand end of Minute1 produces Minute bristles, but no female sterility. BRIDGES (1930) accordingly concludes that the latter characteristic is due to the loss of genes located in the left part of the Minute-1 region. Conversely, when BRIDGES combined Minute-1 with the Pale-I11 duplication, a translocated section extending from (and including) plexus to the right-hand end, the resulting MI/ + P,,, flies were deficient for the arc locus, female sterile and nonMinute (see diagram, figure 6). Obviously, in the left part of the Minute1 section there are haplo-insufficient normal allelomorphs which, when left alone, cause the female sterility and correspondingly in the right part analogous haplo-insufficient normal genes which are responsible for the Minute reaction. 11-Y translocaFurther: The hook-purple deficiency in DOBZHANSKY'S and BRIDGES1932) produces a dominant Minute, tion "H" (see SCHULTZ while his longer allelomorphic hook-purple-light deficiency involved in his analogous "G" translocation studied by RHOADES(1931) is lethal when heterozygous. When SCIIULTZcombined deficiency "G" with the shorter duplication "HJJ'the net result was a short deficiency which produces an extreme Minute and acts as a deficiency for light (MORGAN,BRIDGES, SCI~ULTZ 1931) (see diagram, figure 7 ) . Hence both inside and outside the section which the two deficiencies have in common there must be haploinsufficient normal allelomorphs which, when left alone, induce Minute reactions. And if the entire hook-light region is represented in single quantity only, the sum of the action of all the haplo-insufficient normal genes present in this section leads to deleterious character changes.

93

100

"

104

105

106

107

Minute 9 ster~le 80% viability

t

$St.

I I I

103

P

fM1 MI,

102

101

I I

I

I

4

I I I

I

I I

1

pm

PU Def.

Minute normal fertility and v~ability

nupi.

Pale I

I

I

I

Pm DUPI.

%MI

--m

I

a

non - plinute 9 sterile ~iormalviab~lity

$st.

FIGURE6.-Diagram for comparison of the allelomorphic I1 chromosome deficiencies Minute-1 and Minute-1,. Below: the combination of deficiency Minute-1 with the Pale I11 duplication.

Finally, as BRIDGES (1921) has shown, diploid haplo-IV flies are Minute. As pointed out by SCHULTZ (1929) there is reason to assume that the Minute-IV section, a deficiency for rotated abdomen and cubitus interruptus, represents the section responsible for the Minute bristle change. BRIDGES has found that whereas diplo-IV triploids are viable and fertile, haploIV triploids die. And the responsibility for this lethal action has been defiwith the Minute-IV section, since he found that nitely placed by SCHULTZ triploids with two fourth-chromosomes of which one carries the Minute-

Dupl."H Pr *

- -.I

Dupl. "G"

cr

Dup1"H"

I

I

I

Extreme Minute

FIGURE 7.-Diagram for comparison of the allelomorphic I1 chromosome deficiencies "H" and "G!' Above: Map of the corresponding region of the second chromosome. Below : The combination of deficiency "G" with duplication "H."

PROCEEDINGS O F THE S I X T H

208

IV deficiency are also non-viable (see diagram, figure 8). Indirect evidence supporting the same view is presented-by SCHULTZ and may also be derived from BOLEN'S(193 1) analysis of a reciprocal X-IV translocation. Hence, the Minute-IV region contains haplo-insufficient normal genes which, when left alone, in the diploid System cause the Minute bristles and correspondingly in the triploid have a lethal effect. The striking phenomenon that so many deficiencies produce Minute charevidence to indicate that many acter changes seems in view of SCHULTZ'S of the haplo-insufficient normal genes are concerned with early growth processes. In this connection it is also of interest to recall that the haploinsufficient gene between scute and prune in the X chromosome produces lethal changes if present in single quantity during the early Tiiploid I , U ~n

I l~ m I 11 III

1rm

I ~m I lr m

Diploid

r

Inm

Lethal

, AIef.1~

N

Lethal

In m

1llm

TKm

r

v Minute

--

,DeeIV

Minute

FIGURE 8.-Diagram for comparison of the effect of haploidy for the entire fourth chromosome and for the Minute-IV section in diploid and triploid individuals.

growth stages, but not so if the loss of the section takes place during the later somatogenesis. Correspondingly, as shown by STERN(1927), deficiencies involving large sections of an autosome have no lethal effect in somatic mosaic cells, which we know from translocations that they have during the early development of the zygote. For a series of Minutes, inREDFIELD and especially SCHULTZ cluding clear cases of deficiency, BRIDGES, (1929) have found that two doses of a Minute in the triploid have a lethal effect while one dose is completely recessive. In the latter respect the Minutes differ from other dominants such as Curly, Dichaete or Stubble (SCHULTZ 1930). All these dominants are lethal when homozygous, 1929, REDFIELD but in spite of this SCHULTZ found in the case of Stubble that triploid flies with two doses of Stubble are of good viability, and they have very pro1930a). nounced Stubble characteristics (DOBZHANSKY This special behavior of many Minutes seems very natural. If we consider a particular Minute character in the diploid as due to the action of the haploinsufficient normal allelomorph a, then in the triploid with one dose of the

INTERNATIONAL CONGRESS O F GENETICS

209

deficiency a is represented twice, equalling 2a, while in triploids with two doses of the deficiency only one a is present (see diagram, figure 4). In order to make the potency values for the triploid system comparable to those of the diploid, we have to reduce the former by multiplication with 2/3. This gives for the triploid combinations mentioned the values 4/3 a and 213 a respectively. While the potency value 2/3 a is generally so low that the lethal threshold is passed, the value 413 a is, conversely, in most cases sufficient to reach the threshold for the production of the non-Minute, wild phenotype. That two doses of an ordinary recessive gene have as a general rule BRIDGES,STURTEVANT 1925, no somatic effect in the triploid (MORGAN, SCHULTZ and BRIDGES1932) seems also natural since to the effect of the two mutant genes is here added that of the normal allelomorph simultaneously present so that the sum of these combined effects passes the wild3N

- ,

--

Lethal

X.a-Xa FIGURE 9.-Diagram triploid individuals.

--

3 N -non-Minute g.2a-Xa

-

~inute

2N

a

for comparison of the effect of Minute deficiencies in diploid and

type threshold. Conversely, triploids carrying a deficiency plus two doses of an included recessive gene show the corresponding recessive character, but not the dominant character of the deficiency (MORGAN,BRIDGES, SCHULTZ 1930). We have in the above concentrated our attention on certain evidence from multiple allelomorphism and from the action of genes in single quantity, two of the indirect ways accessible for inferences respecting the potency of mutant and wild-type genes. Both these lines converge in supporting the conception that mutant and normal genes are essentially similar and additive in their effect, a result which is in accordance with that arrived at by STERN (1929) and, for certain classes of genes, by MULLER,LEAGUE and OFFERMANN (1932) in the studies of changes in gene dosage. The mutant genes dealt with are less potent than the corresponding wild-type genes, and they are more potent in duplicate than singly. Of the normal allelomorphs some are so potent that the wild-type threshold is reached even when they are present in single quantity only. They are haplo-sufficient. Others, the haploinsufficient normal genes, are less potent, and the latter are responsible for the dominant changes encountered in deficiency. A reasonable corollary from this view would be that there is also an upper potency threshold for the wild- phenotype. If this threshold is passed by

210

PROCEEDINGS OF T H E S I X T H

provision of additional gene doses (possibly also by gene mutation to allelomorphs which are more potent than the corresponding wild-type gene) then again dominant changes will result. Evidence supporting this conception may be derived from cases of duplication. LITERATURE CITED

ALPATOV, W. W., 1930 Growth of larvae in Drosophila melanogaster and its mutant vestigial. J. Exp. 2001. 56:63-71. BELIAJEFF,N. K., 1931 Erbliche Assymetrie bei Drosophila. Biol. Zbl. 51:701-708. BOLEN,H. R., 1931 A mutual translocation involving the fourth and the X chromosomes of Drosophila. Amer. Nat. 653417-422. BRIDGES,C. B., 1917 Deficiency. Genetics 2:445-465. 1919 Vermilion-deficiency. J. Gen. Physiol. i:645-656. 1921 Genetical and cytological proof of non-disjunction of the fourth chromosome of Drosophila melanogaster. Proc. Nat. Acad. Sci. Washington 7:168-192. 1922 The origin of variation in sexual and sex-limited characters. Amer. Nat. 5651-63. 1923 Aberrations in chromosome materials. Eugenics, Genetics and the Family 1:76-81. 1925 Sex in relation to chromosomes and genes. Amer. Nat. 59:127-137. 1925a Elimination of chromosomes due to a mutant (Minute-n) in Drosophila melanogaster. Proc. Nat. Acad. Sci. Washington 11:701-706. 1930 The neutralization of the effects of deficiencies through duplication of the same chromosome materials. Copied from The Collecting Net 5:273. BRIDGES, C. B., and MORGAN, T. H., 1919 The second chromosome group of mutant characters. Pub. Carnegie Instn. 278:123-304. DEXTER, J. S.,1914 The analysis of a case of continuous variation in Drosophila by a study of its linkage relations. Amer. Nat. 48:712-758. DoszaAns~y,T., 1930a The manifold effect of the genes Stubble and stubbloid in DYOSOphila melanogaster. 2. indukt. Abstamm.-u. VererbLehre. 54:427-457. 1930b Cytological map of the second chromosome of Drosophila melanogaster. Biol. Zbl. 50:671-685. GOLDSCHMI~, R., 1920 Untersuchungen iiber Intersexsualitat. 2. indukt. Abstamm.-u. VererbLehre. 33:l-199. 1927 Physiologische Theorie der Vererbung. Berlin 1-247. 1928 The gene. Quart. Rev. Biol. 3:307-324. HARNLY,hl. H., 1929 An experimental study of environmental factors in selection and population. J. Exp. 2001. 53:141-169. 1930 A critical temperature for lengthening the vestigial wings of Drosophila melanogaster with sexually dimorphic effects. J. Exp. 2001. 56:363-368. HERSR,A. H., and WARD,E., 1932 The effect of temperature on wing size in reciprocal heterozygotes of vestigial in Drosophila melanogaster. J. Exp. Zool. 61223-244. LI,J. C., 1927 The effect of chromosome aberrations on development in Drosophila melanogaster. Genetics i2:l-58. LI, J. C., and BRIDGES, C. B., 1929 Deficient regions of Notches in Drosophila melanogaster. Pub. Carnegie Instn. 399:93-99. MOHR,0.L., 1919 Character changes caused by mutation of an entire region of a chromosome in Drosophila melanogarter. Genetics 4:275-282. 1923 A genetical and cytological analysis of a section deficiency involving four units of the X chromosome in Drosophila melanogaster. Z. indukt. Abstamm.-u. VererbLehre. 32:108-232.

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

211

1927 Exaggeration and Inhibition phenomena. Avh. Ilet Norske Vid. Akad, Oslo. I. Mat.-Nat. KI. 6:l-19. 1928 Exaggeration and inhibition phenomena encountered in the analysis of an autosoma1 dominant. 2.indukt. Abstamm.-u. VererbLehre. 50:114-200. 1932 Genetical and cytological proof of somatic elimination of the fourth chromosome in Drosophila melanogaster. Genetics 17:60-80. MORGAN, T. H., 1911 The origin of nine wing mutations in Drosophila. Science 33:4%-499. 1929 Data relating to six mutants of Drosophila. Pub. Carnegie Instn. 399:169-199. MORGAN, T. H., BRIDGES,C. B., and SCHULTZ,J., 1930 The constitution of the germinal material in relation t o heredity. Carnegie Inst. Year Book 29:352-359. 1931 T h e constitution of the germinal material in relation to heredity. Carnegie Inst. Year Book 30:408-415. MORGAN, T. H., BRIDGES, C. B., and STURTEVANT, A. H., 1925 The genetics of Drosophila. Ribl. genet. 2:l-262. MORGAK, T. H., STURTEVANT, A. H., and BRIDGES, C. B., 1924 The constitution of the germ material in relation to heredity. Carnegie Inst. Year Book 23:231-236. 1925 The constitution of the germ material in relation to heredity. Carnegie Inst. Year Book 24:286-288. 1926 The constitution of the germ material in relation to heredity. Carnegie Inst. Year Book 25 :308-312. 1927 The constitution of the germ material in relation to heredity. Carnegie Inst. Year Book 26:284-288. 1928 The constitution of the germinal material in relation to heredity. Carnegie Inst. Year Book 27 :330-335. 1929 T h e constitution of the germinal material in relation to heredity. Carnegie Inst. Year Book 28:338-345. MCLLER,H. J., 1930 Types of visible variations induced by X-rays in Drosophila. J. Genet. 22 :299-334. MULLER,H. J., LEAGUE,B. B., and OFFERMANN, C. A,, 1932 Effects of dosage changes of sex-linked genes, and the compensatory effects of other gene differences between male and female. Anat. Rec. 51:110. MULLER,H. J., and PAIXTER,T . S., 1929 The cytological expression of changes in gene alignment protluced by X-rays in Drosophila. Amer. S a t . 63:193-200. NASARENKO, J. J., 1930 Ein Fall wahrscheinlicher Verdoppelung eines Chromosomstiickes bei Drosophila mela~zogaster.Biol. Zbl. 50:385-392. PAINTER,T. S., and MULLER,H. J , 1929 Parallel cytology and genetics of induced translocations and deletions in 1)rosophila. J. Hered. 20:287-298. PATTERSOK, J. T., 1931 The production of gynandromorphs in Drosophila melanogaster by X-rays. J . Exp. Zool. 60:173-203 1932a A new type of mottled-eyed Drosophila due t o an unstable translocation. Genetics 17 :38-59. 1932b A gene for viability in the X chromosome of Drosophila. 2. indukt. Abstamm.-u. VererbLehre. 60:125-136. 1932c Lethal mutations and tleficlencies in the X chromosome of Drosophila melafiogaster by X-radiation. Amer. Nat. 64:193-206. PATTERSON, J. T., and MULLER,H. J., 1930 Are "prngressive" mutations produced by Xrays ? Genetics 15:495-577. PATTERSON, J. T., and PAINTER,T . S., 1931 A mottled-eyed Drosophila. Science 73:530-531. REDFIELD,H., 1930 Crossing over in the third chromosome of triploids of Drosofihila melanogaster. Genetics 15:205-223.

212

PROCEEDINGS OF T H E SIXTH

RHOADES, M. M.,1931 A new type of translocation in Drosophiia melanogaster. Genetics 16 ~490-499. ROBERTS, E., 1918 Fluctuations in a recessive Mendelian character and selection. J . Exp. Zool. 27 31.57-192. ROKIZKY, P. T., 1930 Uber die differentielle Wirkung des Gens auf verschiedenen Korpergegendcn. Z. indukt. Abstamm.-u. VererbLehre. 57:37-90. SCHULTZ, J., 1929 The Minute reaction in the development of Drosophila melanogaster. Genetics 14:366-419. SCHULTZ, J., and BRIDGES, C. B., 1932 Methods for distinguishing between duplications and specific suppressors. Amer. Piat. 64:323-334. STERN,C., 1927 Uber Chromosomenelimination bei der Taufliege. Naturwissenschaften 15 :740-45. 1929 Uber die additive Wirkung multipler Allele. Biol. Zbl. 49961-290. 1930 hlultiple Allelie. Handb. Vererb. Wiss. 1:l-147. STURTEVANT, A. H., 1925 The effects of unequal crossing over at the Bar locus in Drosophila. Genetics 10:117-147. 1928 A further study of the so-called mutation at the Bar locus of Drosophila. Genetics 13 :@I-409. WRIGHT, S., 1925 The factors of the albino series of guinea pigs and their effect on black and yel!ow pigmentation. Genetics 10223-260.

INTERNATIONAL CONGRESS O F GENETICS

213

F U R T H E R S T U D I E S ON T H E NATURE A N D CAUSES OF G E N E MUTATIONS H. J. Muller, University of Texas, Austin, Texas RECENT WORK O N T H E CAUSES O F MUTATIOKS

The problem of the induction of mutations by irradiation Possession of the capability of transmuting the gene brings with it the obligation of attempting to find an explanation of how this transformation takes place. A t first it seemed quite an understandable result-even one to have been anticipated-that high-energy radiation should change the gene. For the atoms of genes cannot be immune from activation either by the Xray quanta themselves or by the fast-moving electrons released by the latter, and such activation should, in the case of some of the atoms at least, be the prelude to chemical reactions which alter the composition of the molecules in which these atoms lie. I t was tempting, especially for the physicist, to believe in this relatively simple explanation of the induced mutations, particularly since the work of HANSON,HEYESand STANTON (1929, 1931, 1932) and of OLIVER(1930a, 1932), corroborated by that of (1930), of TIMOF~EFF-RESSOVSKY STADLER( 1930), of SEREBROVSKY ( 1931b), of PATTERSON ( 1931 ), and of others, demonstrated so clearly that the frequency of the induced mutations is directly proportional to the total energy absorbed from the high-energy radiation, regardless, within wide limits, of its distribution in time and of the size of the individual quanta. These findings showed that the impacts of the released electrons must be the primary causative agents in producing the mutations, and it was easiest to think that the latter were always the direct results of the former. No doubt they are, in some cases; that is, it is difficult to believe that any gene can be so stable that no well-directed electron can produce a permanent change in it. However, later results give us reason to conclude that the induction of some at least of the mutations is a less direct matter which, though tracing - back on the one hand to the electrons, also involves an intermediary course of events, dependent somehow upon the peculiarities of the biological system, The series of problems is as yet far from being solved, yet I believe it will be useful to present them at this stage, for consideration and further experimentation.

I~terdependenceof chromosome breaks The clearest series of facts comes to light in connection with a study of chromosome breaks, which I believe may have some significance in the

214

PROCEEDINGS O F THE SIXTH

study of gene mutations also. In my first work on induced mutations (1927), I showed that the "mutational" effects of irradiation fall into two main subdivisions, namely, changes of individual genes-"gene mutationsH-and rearrangements of genes, due to breakage of the chromatin, often followed by reattachment of one or more of the pieces in a new order. It was natural to suppose that these two phenomena were interrelated. Since the genes form a chain, the chemical alteration of a gene might a t times be a matter entirely confined to the individual link, and it might, on other occasions, by destroying or breaking the link or its connection with an adjacent link, result in a breaking of the whole chain. According to certain results of ALTENBURG and myself (1930), the two effects ordinarily occur with similar frequency, and OLIVER(1932) has obtained evidence that the frequency of the induced rearrangements, like that of the gene mutations, is proportional to the energy absorbed. Let us then examine further into the mechanism of production of these rearrangements in the hope of throwing further light on the problem of the gene mutations. In so doing, we come upon a significant series of facts in which, hitherto, the forest has usually been ignored on account of the trees. I n the study of the frequencies of induced trans1d;cations above referred to (MULLERand ALTENBURG 1930; submitted 1929) it was suggested that these translocations might frequently involve breakage of both chromosomes concerned, since otherwise it was difficult to explain why the smaller chromosomes shotlld serve less often than the larger for the attachment of translocated pieces of other chromosomes. As more and more cases of translocations have been analyzed, this supposition has been confirmed. Since the finding of the first mutual translocation in Drosophila ("Swoop," in 1929-see MULLER1930a and b ) , more and more of the induced translocations have been proved, on analysis, to belong to this category. Thus, of five translocations between chron~osomes I1 and 111, analyzed by DOBZHANSKY and STURTEVANT (1930, 1931), four (all the induced ones) were shown to be of the mutual type. OLIVER.(1930b, 1932), GLASS(1932), working at the Texas laboratory, and VANATTA (1932), working with OLIVERa t WASHINGTON UNIVERSITY, have made similar findings for induced translocations involving X with 111, and I1 with 111, and BURKART(1931, 1932), in STERN'Slaboratory, supplemented by OFFERMANN in Texas, has found a spontaneous mutual translocation of X, with I1 (called "Blond"). In a t least one of OLIVER'Scases, where the translocation itself was not mutual, there was nevertheless evidence of a mutual breakage, since a piece broken off by one chromosome seemed to have become inserted into a gap made by breakage of the other chromosome.

INTERNATIONAL CONGRESS O F GENETICS

215

I t might be thought that a t least those translocations which involved the tiny fourth chromosome with another chromosome might be cases where only one of the chromosomes-the larger-had been broken. As a matter of fact, breakage of the fourth chron~osomewould usually be very difficult to demonstrate in such cases, even if it had occurred, and so most translocations involving the fourth, reported by myself and PAINTER (1929), by DOBZHANSKY (1929, 1930b, 1931), and by PATTERSON (1932a) may or may not be of this type; the data so far given do not bear on this question. W e have, however, at our laboratory, made several studies of the matter on translocations involving the X and IV, which are somewhat better *adapted for a solution. Of the three cases studied, one, "X-IV 3," investigated by BOLENand myself, is clearly a mutual translocation in which the little chromosome, IV, has itself been broken and a piece of it exchanged for a piece of the X, which was also broken (BOLEN1931). Another, "XI V 1," has been found by OFFERMANN (see OFFERMANN and MULLER 1932) to involve a breakage of the fourth chromosome together with an insertion, into the gap thus made, of a piece deleted from the middle of the X ; the latter chromosome, which had been broken in two places, had its two terminal pieces joined together. The third translocation of X and I V studied, "X-IV 4," is as yet doubtful, though certain peculiarities of its behavior, found by STERN(1931), suggest a mutual breakage. I do not mean to insist that all induced translocations involve mutual breaks; in fact, there are certain phenomena which indicate that this is not a universal rule. Thus, in the induced translocation accompanying scute19, it has been found by LEAGUEand myself that although the left end of the X is attached at or near the left end of the genetic map of 11, there has been no transfer of material from I1 to X (at least, not of any active region), since flies are viable that are homozygous for the chromosome I1 of this translocation but contain only a normal X or X's. If a piece had been removed from 11, such flies would completely lack that piece. It is still possible, but dubious, to assume an insertion very near the end of 11. A similar and MULLER1929). BRIDGES' case is that of "11-111 26" (of PAINTER 1926) and the sponoriginal spontaneous translocation (see HAMLETT and STURTEVANT (1931) also aptaneous case described by DOBZHANSKY pear to involve but one break, with attachment of the resulting fragment to the side of another chromosome. However, in the face of the above findings, including those concerning chromosome IV, in which they are to be least expected, it is evident that the great majority of translocations conform to this rule. Furthermore, in the great majority of cases of induced translocations in which we have been able to get any evidence on the ques-

216

PROCEEDINGS O F THE SIXTH

tion, it has appeared that the fragments have become united together a t their mutual points of breakage, rather than by any of their originally free ends, or by an attachment to the side of another chromosome. T o parallel the above facts concerning translocations, we have a similar series concerning inversions-that is, evidence that they usually involve two breaks, with a reattachment of the pieces a t their points of breakage although in a different direction with regard to each other than the previous one. Of the five inversions (all spontaneous) analyzed by STURTEVANT (1926, 1931) and GRAUBARD (1932) four could be shown to involve two breaks; it is assumed that in the fifth (CE,) one of the breaks is located near the end of the chromosome, beyond the farthest marker used. The comparison. and PLUNKETT of D. sirnulans and rnelanogaster made by STURTEVANT (1926) also shows that a two-break inversion must have occurred in one of them in its evolution. W e have analyzed one spontaneous and three induced inversions of the X ("ClB," "8 49," "ClB reinversion," and "W"'5," respectively) sufficiently to determine this question in regard to them, and find them all to be double-break inversions, while the findings of SEREBROVSKY and of LEVITon the induced scute-8 inversion show these also to have double breaks. Evidently the only difference between most inversions and translocations is the more or less accidental one that in the former the two points of breakage and exchange happen to be on the same chromosome, in the latter on different ones. Deletions, including many of those less extensive genetic changes hitherto known as "deficiencies" of inner segments of a chromosome-which we now have every reason to regard as small deletions (see page 232),-conform to the same rule. That is, they too involve two breaks, with reunion of the pieces a t the point of breakage, only it happens that in these cases, unlike the inversions, the new junction is between the terminal pieces. Apparently it is more or less a matter of chance which broken ends unite with which, so long as the junction is between one broken end and another one. The above array of findings regarding translocations, inversions and deletions leaves only one alternative to the conclusion that the occurrence of one break tends to be associated with that of another one. That alternative is t o suppose that the breaks occur independently of one another, but that only in cases where two breaks have happened to occur does reattachment usually take place. I t is agreed that attachment must ordinarily be by the adhesive broken ends, and it is concluded that in the cases in which reattachment fails to occur the fibreless fragment is lost and the resulting aneuploid individual is usually inviable. This hypothesis would also require the assumption of

INTERNATIONAL CONGRESS O F GENETICS

217

some a s yet unknown mechanism to enable these four distant broken ends to find each other, two by t w o - e i t h e r some force of attraction at a distance or a sort of very thorough groping movement. The alternative hypothesis above depicted breaks down in the face of the evidence that the frequency of gene rearrangements varies approximately in proportion to X-ray dosage, and probably not as rapidly as the square of the dosage. The frequency would vary more nearly as the square of the dosage if the rearrangements required the coincidental occurrence of two independent events (the two different breakages), each of which separately varied as the dosage. Such a mode of frequency variation seems irreconcilable with the results of OLIVER(1932) above referred t o ; the results of (1930), though not so critical in regard to this MULLERand ALTENBURG question, also speak strongly against it. W e would also expect some traces of such an effect on the general frequency of mutations (including those connected with rearrangements), if the square rule held true, whereas the studies and of OLIVERon mutation frequency show a strict, simple proof HANSON portionality with dosage, with certainly no excessive increase at higher dosages. A s direct evidence against the idea of independent breaks we may also mention a study of LEVIT'S (unpublished), which indicated that (although in some regions "simple" breaks occurred readily) in a certain region of the X (of scute-8 inversion) there were actually fewer simple breaks than breaks occurring concomitantly with breakage in another specified region. I f , now, the two breaks involved in a rearrangement are not independent, we must conclude either that one break somehow acts to induce another one, or-what seems a priori more likely-that both are due to a common cause. In either case, the localization of the effect (its confinement to two given spots) would practically require (in consideration of cell structure and mechanics), a spatial propinquity of the two chromosome threads concerned. This brings us, in essentials, back to the interpretation which SEREBROVSKY in 1929 offered for translocations and inversions, which was extended by DUBININto deletions. SEREBROVSKY and DUBININadvocated the interpretation (which we also had had in mind as a possibility) that where the chromosomes crossed each other, no matter whether the same or different chromosomes, there, under the influence of radiation and sometimes of other circumstances, they were likely to stick together and then to become broken apart in such a way that the pieces came to lie in a different order than they had before. This would mean, in many cases, that a double break and exchange of parts occurred, the pieces becoming reattached a t their broken, not their free ends.

218

PROCEEDINGS O F T H E S I X T H

There are still certain difficulties in this scheme, especially that of explaining how the insertion into one chromosome of a piece deleted out of another would take place, for this would seem to necessitate the rare coincidence of three strands meeting a t exactly the same point. Moreover, we cannot at all agree with SEREBROVSKY'S extension of this hypothesis to account for gene mutations as being simply deletions or additions of so minute a portion of chromatin as to be classified as a single gene. The reasons for rejecting this have been given in a previous publication (PATTERSON and MULLER1930) ; and certain minor modifications may also be required-it may be, for instance, that union does not occur, or is not completed, before breakage, and it is highly unlikely that a piece can be mechanically cut out of the side of a chromosome, sausage-like, if the persistent structure in the chromosome is a thread-like chromonema, containing a single-file string of genes. Yet in view of the undoubted tendency for association of one break with another, far beyond what chance would allow, and for a union between the broken ends formed by these breaks, we are practically forced to and DUBININ'Sgeneral scheme, in so adopt the essentials of SEREBROVSKY far as gene rearrangements involving reattachment are concerned.' In adopting this scheme, we accept the principle that the rearrangements occur by a process which is virtually crossing over, except that it is between non-homologous regions and at a time other than the proper synaptic period. The irradiation has somehow (possibly by de-charging them) done away with the repulsion which normally holds chromosome strands apart from one another, and it has allowed the crossover mechanism to operate illegitimately. I t may be noted in passing that the existence of this effect renders quite unnecessary and irnprolxble the hypothesis of BELLING, which seeks to exCertain further tests await the hypothesis. Among these is the determination of whether, in the case of two or more breaks, there is always an exchange of attachments. Suppose, for example, that the chromosome o r chromosome-region containing the chain of genes A U C U broke between B and C, and that chain L M N 0 broke between M and N. If, now, sections A B and L M become attached together a t their broken ends, forming a sequence A R M I*, it is necessary, on the above hypothesis, that if C D becomes reattached a t all, it becomes attached t o N 0, making the sequence D C N 0. If C D o r N 0 became attached at the side o r at one of their free ends (forming C D N 0 o r D C 0 N o r C D 0 A'), o r if C D became attached by its broken end t o the segment formed by a third breakage (forming D C S T, f o r example), some modification of the above hypothesis would be necessary. In the latter case, however, caution would be required in order to make sure that C D had not become broken again (say, between C and I)), that is, that the junction wj:h the segment S T, from the third break, was really a t exactly the first point o i breakage in question (forming D C S T rather than D S T). This matter is also being discussed by GLASS (1932). Such cases await future analysis.

INTERNATIONAL CONGRESS O F GENETICS

219

plain normal crossing over without the assumption of breakage and reunion of chromonemata. T o avoid the assumption that breakage and reunion are possible, he assumes instead that, in the propagation of genes, the daughter genes are at first unconnected by strands, and that the new strands, in growing out, may, in places, make cross connections between homologous chromosomes instead of intra-chromosomal connections. Our facts, however, show that the chromosomes do possess the potentiality of segmental interchange by breakage and reunion, since this can occur, under irradiation, even between non-homologous parts, at a period (the mature sperm) when there is no growth or synapsis. I t becomes highly probable, therefore, that during the intimate union of homologous strands at synapsis this power of breakage and reunion is exercised normally. For our present inquiry, however, the more important point in our conclusion is the following: The locus of breakage of a given thread, under irradiation, is determined somehow by its relation and its proximity to another thread, and the breakage of both threads accordingly has a common cause. O n considerations of size it is easy to show (see discussion below on striking of genes by electrons) that there is no appreciable chance of a chromosome that is touching another one being struck and broken by the same electron as the latter. With all doses ordinarily given, the chance of its being struck by other electrons is much greater. But we have already seen that the breakage of the two chromosomes cannot be a result of two independent hits. Therefore the points of breakage (or at least cf one of the breakages) were determined rather by the relation of the biological structures than by the exact paths of incidence of the electrons. The passage of the electrons in general, or of some one particular electron, must have provoked some more diffuse train of reactions, which in turn made possible, or facilitated, the breakage and reunion a t the point of crossing of the threads. The occasional occurrence of similar phenomena at one jump even in non-treated material (observed in the case of the CIB inversion, the Blond mutual translocation, et cetera), where natural radiation was negligible, fits in with this conclusion. Breakage, then ( o r a t least the "second" breakage), is not determined by the fact that an electron has chanced to hit the broken chromosome thread particularly accurately. Now what bearing has this upon gene mutations? Simply this. If breakages can be proportional in frequency t o the absorbed radiation, and can nevertheless be produced otherwise than by direct hits, then we have no right to assume that gene mutations, which likewise occur a t more or less isolated points, and which obey the same frequency law in

220

PROCEEDINGS O F T H E SIXTH

relation to dosage as the breaks, are probably due to direct hits either. In fact, we should rather draw the contrary conclusion, that they, like the breaks, are probably not usually caused by direct hits but by a somewhat less direct action of the radiation.

On the relation between gene mutations and rearrangements W e have next to consider some evidence bearing more directly on the question of a relationship between the manner of origination of gene rearrangements and of gene mutations. Some years ago I noted the fact that the majority of induced translocations were associated with either dominant or recessive lethal or other phaenotypic effects; the same has proved true also of our induced inversions. The principle has held also in all the four cases of spontaneous rearrangements in which our knowledge of the time of origination of the condition was sufficient to indicate whether or not the two effects arose simultaneously. These comprise: the "Pale" translocation (BRIDGES1919), the CZB inversion (MULLER1922a), the "Blond" translocation (BURKART 1931), and translocation "11-III E" (DOBZHANSKY and STURTEVANT 1931). Spontaneous mutations are so rare that these cases cannot possibly represent coincidences. There are three more o r less alternative possibilities to account for this association (MULLER1930b, MULLERand ALTENBURG 1930), namely : ( 1 ) That one o r more genes directly at the point of breakage had become destroyed or altered in the process of breakage (this generalizes upon BRIDGES'proposal that in his "Pale" translocation not quite all of the piece that had been broken off became transferred, but that a small portion was lost [BRIDGES19231.) ( 2 ) That a gene mutation had simultaneously occurred at another locus, linked to that of the breakage. ( 3 ) That the phaenotypic change was a result of a different mode of expression of the gene, dependent on its being in the proximity of different gene-associates, which influenced its kind or degree of activity. I n this latter case, as I pointed out (MULLER1930a), a reestablishment of the original gene order would automatically reestablish the normal phaenotype. (Likewise a recurrence of the same rearrangement would always reproduce the same phaenotypic change.) My earlier inclination was rather toward what appeared to be the simplest view, which was comprised under ( I ) , that is, that a gene directly a t the breakage locus had been changed in the process of breakage itself. Now, however, evidence has accumulated to show that genes not directly a t the breakage locus, even though very near by, can also become changed, while

INTERNATIONAL CONGRESS OF GENETICS

ni

at the same time loci between those of breakage and of the affected gene may not be changed perceptibly. This bespeaks either ( 2 ) a permanent gene mutation separate from, though no doubt somehow associated in its origin with, the break, or ( 3 ) a position effect.' I do not wish to deny the possibility of a position effect. Whether or not such an effect commonly occurs is a matter of great importance in its bearing on the mode of action of the genes and on evolutionary possibilities ; it should, moreover, be capable of solution in the near future. I believe, however, that the third alternative, that of a real gene mutation, is more probable as an explanation of the phenomenon here in question, especially in consideration of related phenomena. As an illustration of the phenomenon, I may refer to the case of a deleted X chromosome (MULLER1930b), in which the left break was just to the right of the locus of facet, which is a t 3.0. I n this same deleted X there was a "mutation" of the normal gene at the locus of white (1.7) to a mottled allelomorph (".w"8"). The genes that had mutated could be determined by getting the deleted X into flies that had in their entire X o r X's recessive mutant forms of these same genes; if the dominant normal allelomorph was still present in the deleted X, the recessive allelomorph would then be "covered," that is, the phaenotype would be normal in that respect. I n this case, facet, nearer the break, was "covered," but white, though further away, showed as a white-mottled compound. Again, in another of my deleted X's (which I have called number "24"), the left break was found to be slightly to the right of the locus of a certain but (by our usual tests) to the left of broad (locus 0.5). lethal ("I,,"), Now the deleted X "covered" the lethal; nevertheless, when tested with scute allelomorphs it proved to have a very pronounced "mutation" a t the locus of scute, which is still further to the left of the break than the lethal is. The loci in question are, however, all so close together that their seriation would not have been discovered if a new method of analysis, involving the use of broken chromosomes, had not been available. The crossing over test would have been. inapplicable, owing to the minute distance involved. 'DOBZHANSKY (1932) has recently espoused the last interpretation, that of a "position effect," mainly on the basis of GERSHENSON'S (1931) supposed finding that an exact reinversion in the CZB chromosome abolished its lethal action. I have examined the data in question crcically, and, in the light of a definitely analyzed prior case of my own (MULLER and STONE1930) believe it far more probable that the former reinversion, like the latter, was not an exact reinversion, and that the apparent abolition of the lethal was caused later by ordinary crossing over. Coincident with the present paper, DOBZHANSKY and STURTEVANT (1932) adduce further data which they consider to support the "position" interpretation. A detailed study of such cases, for decision between the two possibilities, now becomes urgent.

222

PROCEEDINGS OF T H E S I X T H

It may be of interest in this connection to give a brief account of the method of analysis here used. A certain translocation which arose simultaneously with mutation scute-19 (found by LEAGUEand analyzed by LEAGUEand MULLEK)involves a break close to the right of the scute locus, the left end of the X (containing scleitself) being attached at or near the left end of 11. It is found that this left-hand fragment, when present as "an extra piece," fails to suppress the lethal action of lethal " j J 1 . " On the other hand a hypoploid female having I,, in one X chromosome and having, as was detached, its other X, the right-hand remainder of the X from which sClS but not the left-hand fragment, has I,, covered (that is, it is able to live). Hence j,, is to the right of the break of the sclD translocation. Hence too must be to the right of scute. Now, deleted X 24, when present as an extra fragment, covers &,; that is, it allows males carrying I,, in their entire X to live. Accordingly the left break of deleted X 24 is to the right of jJ1. Nevertheless, deleted X 24 carries an extreme scute allelomorph, which virtually fails to cover any part of the scute character except the socalled "left-hand" part (the dorso-central bristles) . 3 As in the case of the "wW8" containing deleted X, then, so here also, we have an apparent gene mutation occurring at a perceptible, though very minute, distance from the break, wit11 at least one gene between showing no noticeable change. Unless we accept the very improbable, and ascribe their simultaneous origin to mere coincidence, we must admit that two separated phenomena, a break and a gene mutation (or what is in effect a mutation), have both been engendered by a common cause, with a highly localized, yet by no means punctiform influence. Ordinarily, in such cases, we should not have the means of discriminating between the loci of the mutation and of the break, and we should say that the former happened "at" the locus of the latter. But the presumption now is that the ordinary cases of simultaneous mutation and breakage are in essentials like the cases here analyzed. Consideration of various scute allelomorphs found by the Moscow geneticists and others is of interest in connection with the problem of the nature of gene mutations associated with breaks. Many of the scute mutationsa See Acat (1930, 1932) for a discussion o f this deleted X in relation to the problem of scute. I t was natural a t first to suppose that the break of this deletion had probably occurred within the scute gene, and only a f t e r a n analysis of its relations and those of the s~"translocation with III could the surprising finding of a double genetic change, separated by a distinct distance, be made. O n the other hand, STURTEVANT and SCHULTZ'S criticism (1931) that the effect of deleted X 24 o n the scute character could be due to hyperploidy in regard to other loci than scute was (even before the criticism) proved by AWL to be incorrect.

INTERNATIONAL CONGRESS O F GENETICS

223

like those of so1' and of the deleted X 24 noted above-involve breakages at or near the scute locus. Some involve inversions, in which the scute locus has been removed from the genes normally to the right of it and has been placed near a portion of the inactive right-hand region of the X, that I have found to be homologous with the Y. Yet in these cases the scute mutations were different from one another, and some resembled closely, though not exactly, other definite scutes that had occurred without detectable gene rearrangement. Similar facts have been found in the cases of other loci. If, now,. we decide to consider the mutations arising in connection with breaks as true "gene mutations," then we must conclude that, since the position of a break is not decided by the direct hit of an electron, the position of the gene mutation is probably not so decided either. If it is decided by some other local disturbance-the same as that which produces the breakthen it might sometimes happen that this disturbance, instead of producing a break and a gene mutation, produced two gene mutations (or a multiple group). This contingency makes it possible for our problem to be studied from another angle.

On the connection between one gene mutation and another The question of the possible simultaneous occurrence of two gene mutations, that are expected often to be in close propinquity to one another, is rather difficult of approach, partly because of the fact that most detectable gene mutations in Drosophila are lethals. One means of partially avoiding this difficulty is to look for newly arisen visible mutations, by backcrossing the treated individuals to others which already have visible recessives at certain chosen loci, and then to test the visible mutations thus found to determine whether they have lethals near-by. In one such experiment (see PATTERSON and MULLER1930), two visibles occurred at a locus under observation, one spineless and one scarlet, and both were somehow connected with lethal effects. The mutation first named (spineless) proved to have its lethal effect inseparably connected with the visible. Many cases similar to this have been found, both before and since this experiment (see, for instance, PATTERSON 1932c), though such an extensive test for possible crossing over between the visible and lethal has not usually been made. Such cases, including the present case of spineless, are, on a priori grounds, open to any one of three possible interpretations. The first is that the lethality is one effect of the same gene as causes the visible result; this is undoubtedly true in some cases, such as, for instance, ;broad-lethal, and truncate. The second interpretation

224

PROCEEDINGS OF THE SIXTH

is that a real deficiency, that is, a small deletion, has occurred, removing more than one gene; proof of this requires the condition rather seldom met that two or more k n o w genes be involved. The third interpretation is that there has been a simultaneous mutation of two very closely or completely linked genes. This last possibility has usually been ignored. That the last mentioned alternative has a just claim to consideration is indicated by the case of the second visible mutation found in the experiment in question-that of scarlet. This scarlet also seemed to act as a lethal, but after extensive trials it was found possible to secure homozygous stock of the scarlet separately from the lethal, and to establish the fact that onetenth of one percent of crossing over occurred between the genes for these two effects. The small amount of crossing over was due to propinquity, not to some chromosome abnormality, since known third chromosome genes were proved in this case (and also in that of the spineless) to cross over with their normal frequencies. Now the chance that, with the dose of X-rays used, a lethal should have arisen at the same time as a given scarlet, in such close proximity to it, if the two mutations had been independent events (that is, the chance for a mere coincidence), is considerably less than one in a thousand. In other words, among over a thousand cases of scarlet produced in this way, only one should have another lethal so close by. Thus we should hardly find another such case among a thousand visibles detected by this method, if the case has no significance. If, however, there is such a tendency for double mutations, and they are commonly in extremely close propinquity, it may be difficult to get evidence on the question, since the crossover test for their separateness may usually fail (see the possible case of spineless above). Fortunately, for the scute locus we have another test than that of crossing over available, a test which is sometimes capable of deciding this question with a finer discrimination. This method is the same as that which was employed in the analysis of deleted X 24, namely, the use of chromosome fragments to determine whether or not a given gene is "covered." I t happens that in the scute-19 translocation the X chromosome is broken very close to the right of the scute locus (the left-hand fragment being attached to 11). If now, in another scute mutant, having a lethal effect, the lethality is due to another gene, somewhat to the right of scute, then hyperploid males, containing this double scuteand-lethal mutation in their entire X, and possessing also the left-hand fragment from the scute-19 translocation, will have only their scute gene "covered" (by the s o l 9 allelomorph), but their lethal uncovered. They will therefore fail to be viable. On the other hand, females of the converse composi-

INTERNATIONAL CONGRESS O F GENETICS

225

tion, having the scute-and-lethal in one X, and having as their other X only the right-hand part of the scute-19 chromosome, without the left-hand fragment, will have their scute quite uncovered, but their lethal covered. They will therefore be viable. The two results, checking each other, will prove that the lethal is at a separate locus, to the right of scute. If the lethal had been either at or to the left of the scute locus, the viabilities of both of these classes would have been reversed. If the lethal proves to be to the right, however, then by using deleted X's or translocated fragments of different lengths, and determining which ones "cover" the lethal, the locus of the latter may be ascertained more exactly, even though it be so close to scute as to give no appreciable crossing over. An experiment was accordingly undertaken which had as one of its objects the finding of mutations of (yellow and) scute which might at the same time be accompanied by a lethal effect. This involved irradiating (nonyellow) non-scute males on a large scale and crossing them to (yellow) scute females; numerous female offspring were then examined for manifestations of the recessive genes and bred to determine whether a lethal effect was simultaneously present. ( I wish to acknowledge the extensive in this work. Scutes found by her are desighelp of JESSIE JACOBS-MULLER etc.) nated with the superscript J, as scJ1, scJ2, I t appears from the work that a fairly high proportion-perhaps more than a quarter--of scute mutations are lethal, or are connected with a lethal. The relation of the lethal to the scute gene has now been determined in several of them by the method outlined above. It appears that in the very first scute found in this experiment-scute-Jl-a lethal arose simultaneously with the scute mutation, but at a different locus, just a little to the right of scute, but so close as to have given no crossovers. This lethal was in fact the lethall,,, previously referred to, which has been of help in the analysis of deleted X 24. Likewise in scutes-J4, J6 and J7 a lethal or semi-lethal arose simultaneously with scute, to the right of it. In some of these there was also a gene rearrangement; this part of the analysis, which is important because of the possible explanation of such results as "position effects," is not completed. There are from a fifth to a tenth as many visible mutations as Iethals induced by irradiation in the X chromosomes of Drosophila (see MULLER 1928b). Therefore, if there is a tendency for one gene mutation to be accompanied by another near-by, some of the induced scute mutations may have other visible effects besides those commonly ascribed to the scute locus, and even though closely or completely linked their genic separateness might

226

PROCEEDINGS O F T H E S I X T H

be demonstrated by the same test with scute-19 as was above applied in the case of lethals. T o test this possibility, the scutes then available-about a score-were examined for other effects, and it was found that scute-10 (DUBININ),sometimes known as achaete-2, did exhibit another peculiarity-a disarrangement of the ommatidia-which behaved as completely linked with the scute. Tests with scute-19 and with deleted X's then showed that the eye abnormality was due to a different gene than scute, to the right both of scute and of the lethal, lJ1that had been found in connection with s$ but to the left of broad. It can easily be reckoned that, on the principle of random sampling, the chance was less than one in a hundred that another visible mutation should have occurred in any of these scutes at all, in a locus less than half a unit away. When we consider this together with all the other evidence given above, there can then be no reasonable doubt of the tendency of induced gene mutations-or at least of mutational effects, if we still adhere to the "position" possibility-to occur in localized groups."

O n the chance of douible hits by otze electron I t is natural to conclude that these probable group mutations are due to some indirect effect of the radiation, dependent on certain peculiarities of the biological system and not to be explained merely on the basis of the general physics of ordinary materials. But we have first to dispose of an alternative possibility (MULLER1928a). This is the possibility that two nearby mutations (including, in some cases, breaks) may be caused by the same electron, if the electron happens to have a course approximately parallel to the chromonema and makes an effective hit at two near-by spots in the latter. For in cases in which a point, A, is known to be hit, it is more likely that a point B, near-by, will be hit than in cases in which point A is not hit, for in the first class of cases we know positively that at least one electron did pass near to B, while in the second class of cases there is only the usual (or rather, somewhat less than the usual) chance of such electron passages. This increased likelihood of a hit at a point a given distance away from 'The contingency must also be borne in mind that in some cases of double gene mutations (as of chromosome breaks) the two mutations may be causally related in their origin (owing to some transitory proximity?) and may nevertheless be f a r apart in the actual chromosome map. More than suggestive in this connection is the finding (see MULLER1928c, and PATTERSON and MULLER1930, pp. 591-593) of a double visible mutantspectacled, and reversed-forked-involving two distant loci. There were only three demonstrable mutations (including these two) in a count of 2651 flies, yet two were in the same specimen ! The chance of such a coincidence is only 1 in 883 if the events were unconnected.

INTERNATIONAL CONGRESS O F GENETICS

227

a first hit is a matter that can readily be calculated to a sufficiently close degree of approximation, inasmuch as we know the approximate number of fast moving electrons released in a given volume of known material by a given dose, and also the approximate length of path of these electrons. I t will be understood that the stronger the irradiation, the greater will be the chance that point B should be hit anyway, even when point A is not hit, and so the shorter will be the distance from A at which the fact of A being hit will cause a noticeably increased chance of B being hit. Doctor L. M. MOTT-SMITH,of the physics department of the RICEINSTITUTE,and I have made the requisite calculations on this matter, and we find that, with the heavy doses of X-rays used in genetic work, the increased likelihood of a second hit could extend noticeably for only a very minute distance, of the order of size of the molecules of relatively simple, inorganic substances. If, then, the group effect observed for gene mutations and breaks is simply an expression of the path of a single electron, the genes would have to be far smaller than any one had imagined, or else packed together like pancakes with their shortest dimensions length-wise of the chromonema. This would, however, probably make the chromonema too short, unless the genes were packed in groups, with connecting fibres between the latter. There would be additional difficulties in accounting for the fact that effective hits came so close together, for by no means every molecule that is hit, in the sense of having an electron pass through it, has its structure affected by that electron. The points along an electron passage at which the electron takes effect by tearing out other electrons from atoms, that is, by ionization, are much too far apart to allow of the group effect in question, and it would have to be supposed that between each two such ionization points there are usually a great number of other 11itswhich, though not causing ionization, nevertheless can readily alter the physico-chemical structure of genes. As a further test of the "two birds with one stone" hypothesis, I have been carrying on some experiments, in part suggested by Doctor HUGO INSTITUTION, Cold Spring HarFRICKE,physical chemist of the CARXEGIE bor, in which the gamma rays from radium emanation were used instead of X-rays. The electrons hereby produced are much faster. This greatly reduces the likelihood that a given molecule, traversed by an electron, will chance to be changed by it. Hence the distances between effective hits are much greater, under this treatment, and the group effect on mutations should be so reduced as to be imperceptible, if this effect really results from double hits by one electron. Analysis of the results is as yet incomplete, yet it is noticeable that about as many of the visible mutations in this experiment were accompanied by lethal effects, as when X-rays were used; the genes

228

PROCEEDINGS OF T H E SIXTH

in question were mainly yellow and scute as before. Hence it seems at present probable that the group effect occurs after gamma ray irradiation also, and that it is not due to "double hits." If, now, we conclude that one of the two mutations in a group mutation was not caused by a direct hit, there would seem to be little need to assume that the other one was caused by a direct hit either. Thus the idea of a somewhat less direct mechanism of mutation production would be strengthened. ( W e wish here to thank CORPORATION, and Doctor SPERTI,for supplying the RADIUMEMANATION us with this emanation free of charge, at the instance of the Committee on the Effects of Radiation on Living Organisms, of the NATIONALRESEARCH COUNCIL. ) Incidentally, if the above is true, it follows as a corollary that we cannot use the likelihood of mutation in given genes at given doses to measure the size of these genes--or even of some portion of the genes, assumed to be sensitive to hits-as was done by BLACKWOOD, and in some independent calculations of MOTT-SMITHand myself (unpublished).

Mutations otherwise induced If the X-ray effect upon the gene may be exerted via some intermediary chain of processes (perhaps one in which the touching of chromonemata plays some r61e), then it becomes more likely that other influences than radiation may also be able to induce gene changes by affecting this chain of processes at some point. This is in line with the calculations of MOTTSMITH and myself (1930), and those independently carried out by TIMOF~EFF-RESSOVSKY (1931b) and by EFROIMSOK ( 1931), showing that most mutations in untreated material do arise otherwise than as effects of natural high energy radiation. Further, it is in line with the early find(1919, also MULLER1928c), that a modings of myself and ALTENBURG erately raised temperature, applied over a considerable period, induces an increase in mutation frequency, and that certain unknown factors also cause considerable variation in mutation frequency. In a later report on the production of mutations (MULLER1928b)b I mentioned briefly that I had tried a new method of temperature treatment for this purpose-namely, the application of heat in semi-lethal doses, which 1 found to be attained in 40 to 64 hours at 36OC. For it was conceivable that under these abnormal conditions new and more drastically effective chemical ' T h e exact figures were given in a later publication (MULLER 1930a, p. 234), and here attention was called to certain peculiarities of the results, which indicated that factors were at work causing an especially high number of mutations to originate in certain particular individuals.

INTERNATIONAL CONGRESS OF GENETICS

229

processes might be set into operation. In this case adult (newly hatched) males were treated and lethals were looked for by the ClB method. The work was done on a rather small scale (approximately a thousand F, cultures, equally divided among treated and control) but one large enough to show that there was no such enormous raising of the mutation rate as is produced by X-rays. On the other hand, there was a suspicion, not based on statistically significant numbers, that a lesser but positive effect had perhaps been produced, and it was stated that larger numbers would be desirable to settle this question. In the well known work of GOLDSCHMIDT, undertaken in the following year (1929), and in that of some workers following him (JOLLOS1930, ROKITZKY 1930), the same method of treatment was followed, being, however, applied to the larvae. They believe that their results show that in certain cases-not in all-the frequency of visible mutations is raised enormously thereby. On the other hand, the published work of FERRY,SCHAPIRO, and SIDOROFF(1930), and the unpublished results of REDFIELDand SCHULTZ, of TIMOF~EFF-RESSOVSKY, of DEMEREC and of STURTEVANT, are negative. The question thus arises whether the positive results first reported may not be due either to influences of some other kind, in combination with the heat, or to differences between the control and treated series in regard to the genetic composition of the stocks, the degree of inbreeding practiced. or psychological factors in the operators, affecting the detection of the visible variations. The ClB method, applied to numerous PI individuals divided a t random into the two series to be compared, is not open to these objections. This year MACKENSEN, in the Texas laboratory, has repeated on a far larger scale my experiment of applying an almost lethal degree of heat to adult males for from one to several days, and used controls which could differ consistently in no other factor than the heat application; as in my earlier work, the CZB method was used. His results thus far show a rise (3.6 times its probable error) in the lethal mutation frequency. Considering the relatively short duration of the treatment, this rise seems to be somewhat higher than that caused by an approximately equal amount of temperature difference when the latter occurs at lower temperature levels, more normal to the organism. The rise was, however, far short of that reported by GOLDSCHMIDT and his followers, and quite inadequate to account for the latter.' 'Decided effects from extreme heat, but effects of a far lower order than those of GOLDSCHMIDT, were also exhibited or otherwise communicated at the Congress by PLOUGE, EFROIMSON, T I M O ~ ~ F - R E S S O Vand S K YGROSSMAN. Certain of these are not yet clear; others would lead to a conclusion similar to that above arrived at.

230

PROCEEDINGS OF THE SIXTH

The above contradictions, as yet unreconciled, call for further research. I t may be that, while temperature accelerates the mutation rate, it does so not merely as it accelerates many simple chemical reactions, but according to a curve that rises more sharply at higher temperatures. This would indicate a complicated set of reactions in which more than one process takes part. Thus it raises the hope that chemical treatments may yet be found that will affect mutation frequency despite the rather effective protection of the gene from outside influences shown by the negative results of the trials hitherto made. In MACKENSEN'S experiments, the mutants were all tested for significant changes in crossover frequency, and it was found that neither those occurring in the controls, nor those in the heated series, were accompanied by gene rearrangements detectable in this manner. Among a similar number of X-ray mutants, a considerable number would have been associated with inversions or translocations that reduced crossing over markedly. Recent

REARRANGEMENT

I

MUTATION

INFLUENCES

I

HIGH-ENERGY RADIATION tests of ALTENBURG, specifically designed to discover translocations, likewise show a dearth of these, as compared with gene mutations, in material not subjected to X-rays. I t follows that heat and also the influences causing gene mutations in untreated material act differently from high-energy radiation, in that the former produce gene mutations with no where near as high a proportion of gene rearrangements. Yet the character of most of the gene mutations themselves is similar in all these cases, so that we must conceive a similar end-mechanism of mutation to be brought into operation. Probably then, in the production of mutations by heat and "spontaneously," the influences impinge upon the chain of mutation-producing reactions at a more nearly terminal point than do the X-rays, as the accompanying diagram indicates. ( I t is a common error to suppose that all

INTERNATIONAL CONGRESS O F GENETICS

23 1

the influences causing mutations are necessarily influences external to the organism. I t is true only if taken in the most general and ultimate sense, inasmuch as certain conditions-warmth, food, oxygen, et cetera-are necessary in order that life, metabolism, and inner motion in general may occur at all. But whether or not a given mutation occurs must oiten be decided by a complicated set of internal "historical" processes, in part of a submicroscopic nature, in the determination of which "external influences," in the ordinary meaning of the term, play little or no part.) This again would imply that the mutation-producing reactions set into operation by irradiation may not always be so simple and direct as an alteration of a gene by an electron hitting it. What the processes involved are is another question. There would seem to be a field of research here which will some day prove fertile. O N T H E CHARACTER O F MUTATIONS

Methods of attacking the problem of whether mutations are merely quarttitative changes Probably some geneticists would welcome the problematical connection between induced gene mutations and rearrangements, and between the latter and chromosome contacts, as evidence for the view that gene mutations, or at any rate those produced by irradiation, are merely due to losses or transfers-the latter in some cases perhaps involving additions--of chromosome material of a type previously present. They would take it as evidence for a presence-and-absence, or at any rate for a quantitative, interpretation of mutational changes. Perhaps they might now extend the interpretation to parts of genes, or sub-genes, in order to account for cases like the scute or truncate series, but, so far as any given kind of gene material was concerned, they would see in the mutation process only a mechanical loss or diminution of the gene, by subtraction of material from the chromosome, or-as they would have to say in the case of some reverse mutations, for example-an increase of the gene, such as might be caused by its overgrowth or by the attachment to the chromosome of homologous material from a sister or homologous chromatid. Further plausibility is lent such a view by the fact that many allelomorphic series do give the phaenotypic appearance of being quantitative in their basis. Fortunately X-rays provide us with a new tool which helps to shed light on these questions concerning the character of the mutations produced by them and by other influences. That is, we can induce gene rearrangements and so get fragments of chromosomes containing normal or mutant genes a t given loci. W e can then add or subtract such fragments, creating hyperploidy or hypoploidy, and can thus determine what the effects of changing

232

PROCEEDINGS OF THE SIXTH

the quantity of a given gene material really are. These known effects of purely quantitative changes may then be compared with the effects that were produced by the mutations themselves. I t has sometimes been assumed that one can judge the phaenotypic effect o i different quantities of a gene simply by comparison of the appearances of heterozygotes and of homozygotes of the two opposite types, or, as a greater refinement, by comparison of the different grades of heterozygotes in polyploids. However, the situation in these cases is hopelessly complicated by the fact that in the comparison of such types we deal not merely with a difference in the dosage of one allelomorph, but always with a simultaneous and opposite difference in the dosage of the other allelomorph, since we must always reckon with a substitution of one allelomorph for the other, when chromosome fragments are not added or subtracted. We cannot legitimately assume in advance of the evidence that either the one or the other allelomorph is a mere absence, and so we cannot tell to what extent the observed effects may be due to the changed dosage of the one, to what extent to that of the other allelomorph, or to an interaction process. For example, in a comparison of the homozygous eosin-eyed Drosophila, the intermediate colored eosin-white compound, and the homozygous white, it need not be assumed, a prior;, that the eosin gene has the effect of producing color, and produces more in double dose. I t might be assumed instead (or in addition) that the white gene inhibited color, and inhibited more strongly in double dose. I t might even be conceived that both allelomorphs inhibited the pigmentation which genes in other loci tended to produce, but that white was a more effective inhibitor than eosin. STERN(1929) used actual dosage differences of a given allelomorph in his determination that each additional dose of mutants of the bobbed series adds to bristle length, up to a certain limit. In his work, instead of a small chromosome fragment, the practically inert Y chromosome served to furnish the extra doses. MOHRand BRIDGES,in their studies on deficiencies, realized that they might be dealing with real dosage differences, but at that time other interpretations, such as a peculiar sort of chain mutation, were not excluded. In an attempt to answer this question, however, I have examined cases in which there were known to be actual losses of the same region as was involved in the above cases of deficiencies, and find the effects to be the same. Thus, for comparison with the Notch-8 deficiency of MOIIR (1919, 1923), in which a piece near the left end of the X chromosome, extending from the left of white (1.7) nearly to echinus (5.5), is "deficient," we have cer( 1 9 3 2 ~ )produced by X-raying. I n these, a relatain cases of PATTERSON'S

INTERNATIONAL CONGRESS O F GENETICS

233

tively large piece was removed from the left end of the X chromosome, though at the same time the very left end, which he has found (1932b) to be necessary for the life of the fly, was provided in advance, in the form of a fragment (called duplication X 1 or "theta") attached to the right end. These known losses of the w-e, region result in Notch wings, and allow recessives of the w, fa and ec loci, present in the homologous chromosome, to manifest themselves just as they would in a compound having them in 6ne chromosome and the most extreme possible allelomorph of that sort in the other. I find females having apricot in one X chromosome and either white, MOHR'S Notch-8 deficiency, or one, of these known losses in the other, all to be indistinguishable from one another in shade. Again, to parallel BRIDGES'forked deficiency (1917), I have obtained, by X-raying special stocks, known losses in the region of forked, which allow forked in the other chromosome to show to a n exaggerated degree. And OFFERM A N N and I, studying BURKART'S (1931, 1932) Blond translocation, have been able to show that flies can be obtained from it which lack the right end of the second chromosome (this having been transferred to the X ) ; in such flies the recessive speck, if present in the other second chromosome, manifests itself, and there is a plexus-like venation, as in BRIDGES'"Plexate deficiency." There is now some evidence from Drosophila, but more especially from 1931), that the two breaks in cases of double breakmaize (MCCLINTOCK age within a chromosome may be at any distance apart, not being limited in their proximity by any principle of interference as rigorous as that which applies to crossings over. In view of this, and the above parallelisms, there can now be no reasonable doubt that the original proved "deficiencies" were small deletions, that is, actual removals of small regions, and so the studies involving them may now take their place definitely with the dosage studies. Later, I shall again refer t o the results from this source. In the meantime, before the status of these deficiencies was established, I undertook, with the assistance of MISS LEAGUE,purposely to produce fragments containing known genes, and to use these for studying the effects of dosage changes.

Hypomorphic mutations The first locus which we undertook to study was that of white eye. W e chose first flies containing the moderately pigmented mutant allelomorph of white called eosin, in which the color is considerably lighter than the normal red, and is distinctly sexually dimorphic, being much lighter in the male than in the female. By irradiation we produced a deleted X chromosome containing this gene. I t was then found that the addition of this frag-

234

PROCEEDINGS O F T H E S I X T H

ment to a male or female which was otherwise an ordinary eosin caused the eye color to become darker, more nearly like the normal red. This shows that the actual effect of the eosin gene is not to inhibit color, as might have been thought by comparison of it with red, but to produce color, since the addition of more of it results in more color,--only i t does not produce as mwh color as the normud "red" allelomorph does. In the male, the addition of the fragment raises the dosage to two, and results in a color like that of the ordinary eosin female, which of course has two doses, while adding the fragment to the female, and so raising the dosage to three, results in a still darker color. This shows that the sexual dimorphism of eosin is due to the difference in dosage normally existing between the two sexes, and not to a difference in the action of the gene in male and female.7 That the above observed results were not to be explained as effects of the excess dosage of other genes than eosin in the extra fragment was shown by producing a slightly smaller deleted X chromosome, not containing the locus of eosin, and repeating the same tests with it. It was found to have no effect upon the eye color. The allelomorph of eosin known as apricot, which has a similar coloration except that male and female are alike, was then tried in the same way as eosin. It was thought that it might not show a phaenotypic effect of dosage changes, since the female with two doses looks like the male with one dose, but it responded similarly to eosin, additional doses darkening the color. Two doses of apricot in the male, therefore, give a considerably darker color than two doses in the female. Evidently it is the difference in dosage of other genes in the X chromosome of male and female which, interacting with the effect of apricot, causes the color, for a given dosage of apricot, to be darker in male than in female, in fact, just enough darker so that one dose in the male gives about the same phaenotype as two doses in the female. The same is presumably true of most of the other members of the white series of allelomorphs, which, except for eosin and ivory, look nearly the same in the two sexes.' The important thing for us now, however, " F o r this reason, eosin cannot Iegitimately be used as an indicator of sex in such exin which he sought t o demonstrate the female character of periments as those of BRIDGES, haploid tissue. That the haploid tissue was dark eosin, as in a female, was doubtless due to the fact that one dose of eosin, with one dose of all other genes, involves the same ratio as two eosins in a diploid, and was not due t o the tissue being female. I n the present author's opinion haploid tissue of Drosophila containing but one X should in fact be female, but the matter cannot be demonstrated by the use of eosin a s a sex marker. ' I n a recent publication, MORGAN,BRIDGES and SCHULTZ(1931) include cherry among the strongly sexually dimorphic members of the white series. This was certainly not true 1913). The present sexually dimorphic stock, labelled of the original cherry (see SAFIK "cherry Abnormal," contains neither cherry nor Abnormal abdomen, but is doubtless an ordinary eosin that either displaced the cherry by contamination or was mislabelled.

INTERNATIONAL CONGRESS O F GENETICS

235

is that apricot, like eosin, is a mutant gene which produces an effect similar to that of the normal allelomorph, but a lesser effect. That is, it works in the same direction (towards the same superficial end result) as the normal allelomorph, but not so strongly. I t is, in this sense, like a lessernormal. I therefore call it a "hypomorphic" mutant. The above results agree perfectly with the findings of MOHR that if eiiher apricot or eosin is in one X chromosome of a female, and the other X has Notch-8 deficiency, which includes a deficiency for this locus, the color is lighter than in the homozygous female. As mentioned above, the same result was obtained when this part of one X was known to have been removed by X-rays. Thus, one dose of this gene produces an effect less like normal than two, and two doses less than three. Similar tests involving known additions or losses of fragments, or both, were then applied to genes in a number of other loci. A deleted fragment containing the gene scute-1 was first produced and was used to study the effect of increased dosages of scute-1, a gene which is said t o "remove" certain bristles. (See, for example, STURTEVANT in these Proceedings.) As with apricot, eosin, and bobbed, so here, the addition of an extra dose of'scute in male or female made the individual more nearly normal, in this case almost completely normal, while the presence of two extra doses tended to result in slightly more of certain bristles than are present in the normal. Scute-1 is therefore a hypomorph. I t does not "remove" bristles, except by comparison with normal. I t produces them, though not as efficaciously. In line with this conclusion derived from hyperploids, AGOL (1932) found, by the use of a chromosome (from scute-19) from which we knew the extreme left end, containing the scute locus, had been removed, that a female with just one dose of scute-1 has fewer bristles than one with two. The test of the effect of underdoses, as seen in hypoploids, is obviously as valid and informative regarding these problems as the test involving overdoses in hyperploids. What MOHRhas named the "exaggeration phenomenon" shown by deficiencies is, then, in our terminology, the lesser effect of one dose of a hypomorphic gene than of two doses. By this test the other mutant allelomorphs of scute, in which other groups of bristles tend to be absent, are also hypomorphic, as AGOL(1932) found; facet is hypomorphic, cases of known losses ; as shown by MOHR'Sdeficiencies and PATTERSON'S and forked is hypomorphic, as shown by my experiment previously cited. In elucidation of the test for forked, it may be explained that in this experiment females were made up which possessed one entire X bearing forked and having attached to its right end an extra piece consisting of the region

236

PROCEEDINGS O F T H E SIXTH

from Bar to the right end; these females also possessed another X that had contained the scute-8 inversion but that had had the distal ("left") end of this chromosome removed up to a point between forked and scalloped. Hence all regions were present in double dose except a small region between scalloped and Bar, containing the forked locus. These haplo-forked hypoploids were markedly forked, phaenotypically. Tests thus far indicate that most mutant genes (both spontaneous and induced) are hypomorphs, inasmuch as they show "exaggeration" with deficiencies, as MOHRhas pointed out, or a t least give a form having about the same degree of abnormality as the homozygous mutant. The latter relation would be expected in cases like white eye, where the mutant gene had nearly reached the bottom of the scale of effectiveness and hence itself had almost as little normal effect as the deficiency had. This latter type of mutant may, descriptively, be called "amorphic." These hypomorphs and amorphs are just the kind of mutants which the few remaining advocates of the presence-and-absence hypothesis, and the advocates of purely quantitative mutation, require as evidence for their views. I t should be noted, however, that their having a lesser effectiveness than the normal allelomorph by no means proves that they themselves involve material losses. They may consist of partial inactivations, or they may give rise to processes that lead in a somewhat different direction, and hence do not work so effectively in the observed direction, or they may involve conflicting tendencies. Moreover, a given mutant allelomorph (whether spontaneous or induced) may be very hypomorphic, or practically amorphic, in regard to one kind of activity of the normal gene, and normal o r nearly normal in regard to another kind of activity. This is well exemplified in the scute series, in which each different allelomorph acts hypomorphically only in respect to its own peculiar combination of bristles, and is normal or nearly so in its action on other bristles. Since, in a comparison of different allelomorphs, the amount or intensity of effectiveness may vary separately from the types of effect, and both of these in turn may vary separately from the number or extensity of the effects, advocates of the quantitative view would here be driven t o admit the existence of various parts of the gene, and to assume that these parts could vary quantitatively more or less independently of one another. This would be a distinct retreat from the simple hypothesis of quantitative variation of the gene as a whole. Whatever the explanation of hypomorphism may be, it is of interest t o observe that the finding that most mutant genes are of this type conforms to WRIGHT'Scontention (1929; see also MULLER1928b, pp. 259-260) that

INTERNATIONAL CONGRESS O F GENETICS

237

gene mutations should in the majority of cases involve more or less inactivation of the processes governed by the normal gene, and that these less active genes should more often act as recessives to the normal than as dominants. This implies that one dose of the normal gene usually has an effect more nearly like that of two doses than of no dose. Whether the latter principle is a primary one, however, or is due to the past selection of modifiers, is another question.

On the compensation of the effects of dosage differences between the sexes, and on domilu~zce In the above connection, it will be worth while t o make somewhat of a digression, to consider a curious fact that has emerged from the results concerning hypomorphs. hat is, it appears that in the great majority of the cases of hypomorphic sex linked genes, one dose in the male produces about as strong or a t times even a slightly stronger effect in the direction of normality than do two doses in the female. This must of course be due to the interaction of other genes in the X chromosome, whose simultaneous change in dosage affects the reaction.' In some cases at least it has been possible to show, by studies of the effects of different chromosome pieces, ( a ) that genes other than the genes for sex are acting as the "modifiers" in question, ( b ) that the modifiers responsible for the dosage compensating effect on different loci are to some extent different from one another, and (c) that more than one modifier may be concerned for a specific locus.10 I base these conclusions on various results obtained in work of OFFERMANN, who has been especially active in the study, of PATTERSON, and of myself. W e may for convenience call these genes "modifiers," but with the reserW e arrived at our main results and conclusions regarding this phenomenon of dosage compensation in the spring of 1930. Although we communicated our results to Doctor STERNa t that time (prior to the remarks made by STERNand OGURA1931, upon this topic), we withheld our preliminary report (MULLER,LEAGUEand OFFERMANN 1931) until after certain checks had been carried through. lo Judging by certain results recently reported by MORGAX, BRIDGES and SCHULTZ (1931), the second-chromosome mutation Pale (associated with BRIDGES'original translocation) has, in addition to a "diluting" effect, an effect on the different eye colors of the white series similar to that produced by lessening from two doses to one the gene or genes in the X chromosome that are responsible for the dosage-compensation of most members of this series (thus, those allelomorphs of white that are lighter in the male are lightened by Pale, but the others are darkened somewhat). This means that the chemical process affected by Pale is the same as, or in its effect similar to, that affected by the dosage compensator ( s ) of the X ; but, since we have seen that there is no reason t o identify the latter with the gene or genes in the X that decide sex, we have no reason to agree with the suggestion of the above authors that "the translocation (Pale) may be closely connected with the sex-determining reaction."

238

PROCEEDINGS OF THE SIXTH

vation that they may sometimes be as important in the causation of the phaenotypic effect as the "primary" gene whose mutations we have available for study. The essential relation is that, in so far as the amount of phaenotypic effect produced by this so-called "primary" gene depends on its dosage, it does not depend at all on the mere "concentration" of this gene in the cell, nor on the relation of its dosage to that of the other genes in general, s t i l less to that of the autosomal genes, but solely on the ratio of its dosage to that of another specific gene or genes which lie in the same chromosome (the X ) . That a relatively high amount of intra-chromosomal interdependence in regard to dosage expression existed among sex linked genes 1930b) and denoted as "intra-chromowas realized some time ago (MULLER soma1 genic balance." In that work, however, we were dealing with those relatively rare normal genes, or gene-combinations, which have a quite different effect, visibly, in one dose than in two. The present findings go much further, in showing the existence of a far stronger interdependence, and one which applies not just to a relatively few scattered genes but to the great majority of the individual genes in the X which can be sadpled. Now this great system of "modifiers," all acting to give a similar sort of effect, and probably affecting most of the genes of the X chromosome, must have a function. I t cannot be that of giving the male mutant as strong, that is, as nearly normal, an expression of its mutant gene as the homozygous female,mutant has. It must therefore be a system which acts on the normal allelomorph similarly to the mutant, but the action of which is more readily apparent to our eye in the mutant type. In most cases the normal gene gives, so far as our eye can perceive, practically the same effect in one as in two doses. Nevertheless, there must be some difference which, though imperceptible, is important for survival; otherwise this system of genic interaction would not be thus maintained to keep the same optimum degree of effect in both sexes, despite the different doses. I t follows that the dominance of the normal gene over its "absence7' is really far from perfect, physiologically (that is, that one dose is not really as effective as two), though it may seem so to the casual genetic observer, and that by selection a system of interacting genes has become established such that the expression of the one dose in the haplo X type is like that of the double dose in the two X type. Bobbed, being present in double dose in male as well as in female, is, as expected, an exception t o this rule of dosage compensation, in Drosophila melanogaster. In Drosophila simulans, on the contrary, bobbed does show dosage compensation, and here it is found, correspondingly, that the male carries only one dose of the "normal" allelomorph (the Y being

INTERNATIONAL CONGRESS OF GENETICS

239

sometimes neutral and sometimes actually "antimorphic" in effect-see page [I9291 ) . 245-as shown by results of STURTEVANT The question may here be raised: Why were the normal allelomorphs of most of the sex linked genes other than the bobbed of D. melanogaster ever "lost" from the Y chromosome if their absence was so deleterious as to require the subsequent evolution of this complicated compensation system? The case of the Y of simztlans shows that they can be &us "lost," or, better to say, changed in expression like a loss, and this would seem to point to the importance of accidental multiplication, not guided by selection, as an occasional evolutionary process. I t may be, however, that most of the genes in the X, unlike bobbed, never were present in the Y, in anything like their present form, at least; that is, that the male has had but one dose of them from the beginning of their existence as such. In that case, they must have arisen either as duplications, as "neomorphs" (see page 246), or both, after the present sex-determining system had already become established. This question might be answered definitely by genetic analysis in a species in which we knew that a part of the X had been derived from an autosome (for example, D. hydei or "obscura"?). The existence in the X of "modifiers" of such a specific kind that, by their change in dosage, they modify the amount of effect of other sex linked genes to the extent required to make the male and female alike, indicates that specific modifiers of gene action are plentifully available. In the case of some of the sex linked genes arising in the manner last suggested (so as to have existed in different dosages in the two sexes from the start) it is possible that the dosage compensation did not result from the selection of mutations in these modifiers but that the "primary') genes themselves were so selected at the time of their origination as to be, ab knitio, adapted in their action to the other, prexisting genes in the X which we now call "modifiers." And even when the compensation did not thus exist from the start, it is likely that the inter-adaptation of primary gene and modifier did not always occur through changes in the modifier alone but also through changes in the primary gene that made the latter sensitive to the modifier (such changes in the primary gene as would be involved in the mutation of eosin to apricot, for example). Thus, where there is only one modifier causing the dosage compensation of a gene that did not have this property to begin with, the chances seem a priori to be equal that the dosage compensation arose (if by one step) by a further change in the primary gene itself, or by a mutation in the modifier; the greater the number of modifiers, the larger the r61e that their mutations have probably played in the process, as compared with

240

PROCEEDINGS O F THE S I X T H

mutations in the primary gene. W e should also remember that mutations could also take place in other genes, for example, autosomal genes, which would serve to bring the primary gene and the modifier into the reciprocal relation with one another which they now have. But, however all that may be, the results do give evidence of the availability of "modifiers," or, to put it more precisely, of mutations which cause certain specific types of nicely adjusted genic interaction, favorable for survival, and not having this survival value too much obscured by pleiotropic effects. theory The above conclusion would appear to lend support to FISHER'S of the origin of dominance, inasmuch as on that theory, too, specific modifiers (albeit of a somewhat different kind), without important other effects of their differences from their own parent genes, are called for. I t would conalso allow us to adopt to a certain extent the suggestions of HALDANE, comitantly. There is, however, an important difference between the mechanism of selection for dosage compensation here studied and that postulated or by HALDANE for the modification of dominance. For in either by FISHER the former the selective moment, if I may call it so, exists throughout the population, while in the latter it is supposed to be limited to a comparatively small minority. Thus -the difficulty is encountered that the pressure of the selection in question may be too small, as compared with that of mutation, or of the selection for even very weak pleiotropic effects. I believe that the above difficulty can be avoided and a better case made out for the origin of dominance by selection if we assume that this selection has had a somewhat different mechanism from that previously postulated. I prefer rather to postulate that the mutations favoring dominance-the genes or genetic conditions which tend to make the heterozygote like the homozygote-have been selected and are maintained not so much for their specific protection against heterozygosis at the locus in question as t o provide a margin of stability and security, to insure the organism against weakening or excessive variability of the character by other and more common influences-environic and probably also genetic. These modifiers must so affect the reaction set going by the primary gene in question as to cause this gene, when in two doses, to be near an upper limit of its curve of effectiveness,ll that is, in a nearly horizontal part of the curve, not so readily subject to variation by influences in general, including reduction in the dosage ''That is, the curve expressing the relation of amount of phaenotypic effect (the ordinate) to the amount or concentration of gene material (the abscissa)-a curve which must usually, in its right-hand portion, rise with ever decreasing slope, approaching a studies on bobbed and in ours on scute horizontal limit, as seen, for instance, in STERN'S and apricot.

INTERNATIONAL CONGRESS OF GENETICS

241

of the primary gene. This does not mean that the phaenotype is necessarily made any more extreme, for counter-checks can be set up. That is, the level of the curve as a whole and its shape, as well as the region wherein it approaches a horizontal limit, are also adjustable, by means of modifying mutations that reframe the conditions under which the reaction takes place. Such modification (see FORD 1930), and not merely an increase in potency of the "primary" gene, will be necessary in the numerous cases in which the curve of effectiveness did not, originally, approach the horizontal within physiologically acceptable limits (or did not do so at all). It should be distinctly understood that the crux of the above view of the origin of dominance lies in the proposition that, where a change in gene dosage causes a perceptible change in its phaenotypic expression (that is, when it is in a noticeably sloping part of its "curve of effectiveness"), it is likely that the degree of expression of the character will be modifiable to an unfavorable extent by environic and by other genetic changes. This seems reasonable, a priori, inasmuch as some of the disturbing influences would be expected to act by altering the reaction in a way similar t o that whereby the change in gene dosage would alter it, and hence would tend to be similarly effective. But we need not rely on a priori reasoning alone. There is a significant amount of experimental evidence already existing to show that there is considerably more phaenotypic variabiIity in the expression of hypomorphic mutant genes than of their normal allelomorphs. Now, these hypomorphs evidently cause a reaction of a type similar to that of their normal allelomorphs, but a weaker or lesser reaction, one which, unlike that of the normal allelomorphs, is much affected by dosage changes. This variability is true of all the known hypomorphs yet studied: namely, all the hypomorphs of the scute series, the white series, the forked series, and the bobbed series (excluding the amorphs, which afford a converse test of the same proposition). The same variability applies also, as we should expect, to the effect of normal allelomorphs in single dose in those relatively rare cases in which the single dose has a perceptibly different (that is, lesser) effect than two: these cases comprise Notch wings, Plexate venation, and several Minute bristle conditions. All told, the evidence given above may be sufficient to show the truth of our proposition as a usual rule. It is not necessary to claim for it, nor do we believe that it has, the validity of a universal law. In conclusion, we may call attention to the bearing of a further fact, derived from our study of dosage compensation, on the problem of dominance. We have seen that in all probability many of the normal genes in the X

242

PROCEEDINGS O F THE

SIXTH

chromosome have this dosage compensation, despite the fact that, even in the female, there is hardly a perceptible difference between the effect of one dose and of two. This indicates-that, even though the normal gene produces its effect in what appears to us to be a nearly horizontal region of its curve of effectiveness (where changes in dosage produce little discernible effect), nevertheless there is a distinct influence, unfavorable to the organism and perceptible in its survival rate, if the effect is made either slightly stronger or slightly weaker. The disadvantage of a stronger effect is shown by the fact that, in the female, the strength of effect has become fixed at so low a level as to call for dosage compensation. For in a sense the dosage compensation may as rightly be regarded as a means of keeping the female from having too strong an action of the gene as a means of giving the male a strong enough action. If twice as high potencies in the female were biologically acceptable, this relatively simple change should often have been utilized (that is, have survived) whereby the male would automatically have been provided with a sufficiently high potency to obviate the need for dosage compensation. W e must conclude, then, that in the fixing of the conditions determining dominance too, it was not feasible merely to increase the potency of the "primary" gene; instead, the characteristics of its curve of effectiveness had somehow to be altered. Experimental evidence of a different nature, indicating that dominance is not a primary property of genes but must have become developed by selection, is given in the section on neomorphs (see page 248) .I2

We must now return from our digression, which has perhaps helped us to understand why hypomorphic mutant genes usually show dosage changes better than do the normal genes from which they were derived, and are recessive to the latter. The question next arises: are all mutant genes hypomorphic? This can be answered categorically in the negative. Since it has been found that there are, reverse mutations of hypomorphic mutant genes, such as scute, apricot, and forked, both spontaneously and as a result of irradiation, we must regard the allelomorphs thereby resulting not as hypomorphic but as hypermorphic to their immediate progenitor genes. Whether or not such a change involves a real increase of material is a doubtful question, subject to the same considerations as applied, con" I am indebted to Doctor C. R. PLUNKETT for calling my attention to the fact that in a paper (1932) presented independently to the same congress he has espoused what is essentially the same viewpoint regarding dominance as that given in the above section.

INTERNATIONAL CONGRESS O F GENETICS

243

versely, to the hypomorphic mutations. TIMOF~EFF-RESSOVSKY (1929, 1931a), as well as PATTERSON and myself (1930), has discussed in some detail the varying frequency of such changes for different loci and allelomorphs. Now if hypermorphic changes of already mutant allelomorphs may occur, resulting in partial o r complete reverse mutations, there might well be hypermorphic mutations of normal genes also, resulting in changes of a type opposite to that of our ordinary mutations. Usually these would be difficult or impossible to detect, on account of the fact previously referred to that two doses of the normal gene are already a t nearly the maximum po?nt in the curve of effectiveness. Thus such changes would be apt to escape (1930) mutant "abrupt" observation. Very likely, however, NASARENKO'S is a hypermorphic mutation of the normal allelomorph of Notch, for it is a t or near the Notch locus, and it and Notch deficiency counteract each other instead of showing an exaggeration effect.

Antimorphic mutatio~zs What evidence have we for other mutational changes than such as could Ix: explained as mere diminuiions and increases? The dominant (somewhat variegated) allelomorphs of brown eye in chromosome I1 are a case in point. When there is one dose of the recessive brown and one of the normal gene, the latter dominates and the phaenotype is red. But, as GLASSand I have found (see GLASS1932), when to the above complex a dose of the dominant allelomorph of brown is added, the result is a brownish (somewhat variegated) color. I t may be explained that this combination is produced by making up a fly that is a compound of recessive brown and dominant brown, and carries as excess a fragment of the second chromosome derived from BRIDGES'"Pale" translocation; this fragment contains the normal allelomorph of brown. The resulting brownish color shows us that the addition of dominant brown to a heterozygote of normal and recessive brown has a real effect and involves the addition cf some kind of gene material different in its effect from the material in the normal gene. This effect, the color change, lies in the same direction from normal as does that of the recessive brown, as comparison of the colors indicates. This is shown more conclusively by the fact that while a hyper-diploid containing one dose of dominant brown and two of normal has practically normal red eyes, a hyper-diploid otherwise similar to the above but with a dose of recessive brown substituted for one of the normals has brownish (somewhat variegated) eyesthat is, the substitution of recessive brown in place of normal results in a

244

PROCEEDINGS O F T H E SIXTH

better manifestation of dominant brown. But the recessive brown itself acts practically as an amorph, since the addition of a dose of it, as an extra, has practically no effect either on the incompletely brown color of the heterozygote of dominant brown and normal or on the red color of the heterozygote of recessive brown and normal. Hence the dominant brown represents something that differs from normal, in its effect, in the same direction as a loss does, but more strongly. In a sense, it has an actively negative value. More accurately, it has an opposite action to that of the normal allelomorph, competing with the latter when both are present. A similar conclusion may be drawn with regard to the mutant gene ebony, of chromosome 111. For, starting with a hyperploid containing two ebony genes and one normal (derived from translocation II-III~~-PAINTERand MULLER1929), as a basis of reference, we find that the subtraction of one ebony makes the color lighter, while the subtraction of the normal makes it darker. I would term such antagonistic mutant genes, having an effect actually contrary to that of the gene from which they were derived by mutation, antimorflzic. Abnormal abdomen may now h interpreted to be a member of this class, as shown by results in MOHR'Sexperiments with Notch-8 deficiency. In the first place, it is to be observed that the gene for Abnormal produces a change in the same direction as a loss of the normal gene. This is shown by the fact that if we start with a heterozygous fly having one Abnormal and one normal gene (this is somewhat Abnormal in appearance), the substitution of a real loss (Notch-8 deficiency) for the normal gene in it intensifies the Abnormal abdomen character. But the Abnormal gene, though thus producing a change in the same direction as a loss of the normal gene, acts more strongly in this same direction than a mere loss does. This in turn is shown by the fact that homozygous Abnormal flies are still more Abnormal in appearance than are the compounds of Abnormal and deficiency. That is, the degrees of phaenotypic Abnormality, as found by MOHR,were as follows : Ab. f: n;y. ng;. norm. Ab. norm. Since in the first three terms of the series the gene represented below was always the same, the observed differences prove the degree of abnormal effects to be in the order Ab>def>norm. It may be mentioned that a recessive allelomorph of Abnormal has been produced by X-rays. It will be seen that in such cases the recessive mutant, though ~Iassifiableas an amorph or possibly a weak hypomorph, probably

> > >

INTERNATIONAL CONGRESS O F GENETICS

245

involves no mere loss of material, since what is apparently a still greater change in the same direction gives a gene which again has a demonstrably active influence. Unless we make the very improbable assumption that the Y may contain other active genes than bobbed influencing the same character, we may also include among antimorphs the gene existing in the Y of most races of Drosophila sinzulans (see STURTEVANT 1929) which (unlike the bobbed allelomorphs reported upon by STERN1929) actually decreases the bristle length of males containing bobbed in their X. It is also possible that the genetic conditions designated as Minute l2 and Plexate include antimorphs-in fact, such is the conclusion which we should ordinarily draw from a recent report; on the other hand, an earlier report interprets these conditions as deficiencies (see MORGAN,STURTEVANT and BRIDGES1927, and MORGAN, BRIDGESand SCHULTZ1931). Possibly the apparent contradiction is due to the effect of dosage changes of other genes in the added fragments rather than to the genes in question (that is, an intraregional dosage interdependence). Fortunately, this possibility can rather easily be put to the test in these cases (in part at least), since a smaller fragment involving the region in question is available in the Blond translocation, and others can rather readily be manufactured. In the meantime, the "position effect" interpretation is not excluded here, nor is that of gene mutation accompanying breakage. Neomorphic mz~tations Somewhat different from the negatively acting, competing mutant genes, or antimorphs, is the class which I am provisionally terming "neomorphs." A good example is the dominant mutant, Hairy wing, near the left end of the X chromosome. The homozygous Hairy wing female is about twice as hairy as the heterozygous Hairy wing female or the Hairy wing male (this constituting an exception to the dosage compensation rule for sex linked genes). The relatively low grade hairiness of the heterozygous as compared with the homozygous female, in this case, is due solely to the single dose condition of the gene for Hairy wing and not at all to a possible influence of the normal allelomorph in the heterozygote. For if a small piece containing this region be broken off of a normal X chromosome, and added either to the heterozygous or homozygous Hairy wing female o- to the Hairy wing male, there is no diminution of the hairiness. On the other hand, if a small piece containing a Hairy wing gene be added t o an individual otherwise normal, Hairy wing will show. The normal allelomorph thus fails to compete. It itself acts like an amorph, so far as its detectable effect on the

246

PROCEEDINGS O F THE SIXTH

character under consideration is concerned. Yet it is no mere absence; it has a material existence, for Hairy wing has arisen at the same locus several times (including at least twice by irradiation). W e must conclude from the above results that the mutation t o Hairy wing does not result from an addition of material transferred from another locus (since the mutation always reappears a t the same locus). It must rather be a change in the nature of the gene at the original locus, giving an effect not produced, or at least not produced to an appreciable extent, by the original normal gene. If the effect had been produced to some appreciable extent by the normal gene also, then the addition of a dose of the normal to the Hairy wing individual should have actually increased hairiness. The fact that normal genes may thus act as amorphs with regard to a particular character affected by their mutations should serve as another warning against regarding mutant genes that seem to Ix amorphic or hypomorphic as really involving a mere absence or loss of material. The obtaining of reverse mutations from near-amorphs, such as eosin from white, gives further evidence for this conclusion. The same kind of finding as above noted for Hairy wing-namely, lack of effect on the character when extra doses of the normal allelomorph are added-was observed by OFFERMANN in studying the spontaneously arisen dominant, Blond, of BURKART. This interpretation holds only if we regard Blond as having its locus in the X chromosome. This is uncertain as Blond lies near the break of a mutual translocation involving X and I1 (see BURKART 1932), but as Blond follows the sex linked rule of dosage compensation it is in all probability in the X. We are, however, making sure of its neomorphism by testing also the effect of adding an extra dose of the suspected region of chromosome 11. Bar eye is a third neomorph. It is well known that STURTEVANT has considered Bar as having no normal allelomorph, at least none at the same locus (1932), as itself. However, the recently reported finding, by DOBZIIANSKY of a second Bar-like mutation ("baroid"), induced by X-rays at the same locus as the old, indicates to me that this locus normally contains a gene that still is subject to this particular type of mutation, although DOBZHANSKY believes that the normal allelomorph was somehow transported there from another locus, at the time of the mutation. BRIDGES'original Bar-deficiency of 1915 (published upon in 1917), which we may now interpret definitely as a loss, shows that the absence of the Bar-locus in the non-Bar chromosome of a heterozygous Bar female has the same effect on the Bar eye character as the presence of the normal allelomorph itself, and STURTE-

INTERNATIONAL CONGRESS O F GENETICS

247

VANT'S work on chromosomes which have lost the Bar locus by unequal crossing over is an indication in the same direction. (There is a possibility that in the origination of Bar a gene became duplicated in sifu, and that one of the resulting twins mutated at the same time. On this rather special hypothesis the mutation would have been of the neomorphic type. But in that case the normals formed from Bar by unequal crossing over would not represent complete "absence.") On the other hand, increased doses of Bar give the abnormal effect more strongly, just as we find for Hairy wing and Blond, and unlike the situation in the case of hypomorphs. (1929) has raised some objection that we may here While THOMPSON be adding and subtracting only a part of the gene, in getting these effects, this possibility is ruled out in some recent studies of OFFERMANN using a strong allelomorph of Bar ("Super-Bar," B" found by STONE)that exists in a chromosome fragment. The addition of fragments containing the whole Bar gene had the expected effect of increasing the bar-like character of the eye in a clear-cut fashion. OFFERMANN likewise proved that this result could not be due to the excess dosage of other genes in the piece. Bar, then, is a mutation of a normal gene, giving a gene that produces a new effect, foreign to the original gene, and not competing with the latter. I t is Yery probable, however, that the new effect is in some way related to that of the normal allelomorph. For it is evident that Bar obeys the usual rule of sex linked genes, having the male, with his one dose, much more nearly like the homozygous female, with her two doses, than like the heterozygous female (see also the case of Blond, and note the contrast with that of Hairy wing). A recently published mention by MORGAN, BRIDGES and SCHULTZ (1931) of the lack of effect of changes in dosage of a fragment containing the normal allelomorph of Bristle on the degree of expression of this second chromosomal dominant leads to the conclusion that it also must belong in the class of neomorphs. I t might yet be possible to evade the obvious conclusion that gene mutations, including those produced by X-rays, involve qualitative changes, changes in the kind of structure and not merely in the quantity of the gene or its parts. For it might be postulated that in all cases of neomorphs there was an imperceptible rudiment of the part which produced the effect in question, already present in the normal gene, and that this part merely became vastly increased in amount by the "mutation." Or it might be postulated that all such changes were "position effects," caused by gene rearrangements. While there are an exceptionally large number of rearrangements both among known neomorphs and antimorphs, there are cases-Hairy wing, Bristle, Dominant eyeless, Abnormal abdomen-which do not involve

248

PROCEEDINGS O F TIIE SIXTH

such changes, unless we suppose the rearrangement to be on such a minute scale as to escape detection. Both these paths of escape into the ultra-small would, however, be pure speculations, the burden of proof for which would rest upon the advocate thereof. I t does not seem to be a coincidence that more loci have yielded hypomorphs than neomorphs, and that even loci which have yielded neomorphs have done so with relative infrequency. These results, if corroborated by more extensive work, would speak for the correctness of the principle put forward by ~ V R I C H(1929; T see also MULLER192813, pp. 259-260) that mutations having an effect in the direction of losses (that is, those that tend to be disorganizing and inactivating) should in general be more frequent than those causing increased or new effects. But while this principle is necessary theory of dominance, it is not, alone, sufficient for as one basis for WRIGHT'S a derivation of the latter; neither is it contradictory to the general viewpoint that the usual dominance of normal genes has been put forward by FISHER developed through natural selection. I t is to be noted, further, that the hypomorphs tend to be recessive, and the neomorphs "dominant." This again is in view, but it is also in line with FISHER'S (since any given line with WRIGHT'S neomorph originates so infrequently that there has been much less chance for selection to have affected its mode of expression), and it is still more in line with the idea previousIy offered (p. 240), that selection has worked primarily towards the stabiIization of the reactions of the normal, homozygous genes. (In the latter case, even rather frequently recurring neomorphs would tend to be dominant.) When, however, we examine into the type of dominance found, we obtain a result of greater apparent significance. For while the recessiveness of the hypomorphs is usually fairly complete, as generally expected, fhe "dominawe" of the neomorphs is in most cases far from conzplete, being of the "intermediate" type. Now this result is exactly what we should expect if dominance of the nearly complete type has been developed by selection (especially, if by the type of selection advocated on page 240), but it is a considerable surprise, in fact, it seems contradictory to the idea that such dominance is usually a primary property of the gene. It wiIl therefore be important to examine further cases with reference to this question. While we have spoken above of the general trends of the results, it should be emphasized that n o absolute rules can be made with regard t o the dominance of the different classes of mutants. A known loss like Notch-8, Plexate, and at least three known Minute bristle conditions, may be dominant or semi-dominant in its effect, and therefore an amorph or a hypomorph may be likewise. In these cases one dose of the normal gene has dis-

INTERNATIONAL CONGRESS OF GENETICS

249

tinctly less effect than two. On the other hand, neomorphic genes may be so "weak" in their effect that two doses are required before they rise to the level of visible manifestation. This was very nearly true in the case of a certain Hairy wing mutant, and in the case of baroid in the female ; under certain genetic conditions (for example, in the presence of ZELENY'Smodifier, called "emarginate") it was true of Bar itself, and under certain environmental conditions it was true of Abnormal abdomen. For the same reason, we cannot make absolute rules regarding the exaggeration ~f recessives and dominants by deficiencies. If the recessive or near-recessive should be a neomorph, like baroid, it will not show exaggeration by a deficiency; if the dominant should be hypomorphic, as in the case of the absence of coxal bristles in some scutes, it will be exaggerated by a deficiency. But the more usual case is the recessive hypomorph (for example, eosin, facet), which shows exaggeration, the amorph (like white) which shows no effect, and the semi-dominant neomorph (for example, Bar) and antimorph (for example, Abnormal), which show instead an apparent inhibition by a deficiency. On our interpretation of most gene mutations as qualitative structural changes, even the distinction into classes above outlined is not an absolute one, and reflects rather the gene's final behavior than its real structure. So we may expect to find genes, for example, that are hypomorphic in one respect and neomorphic in another. Possible examples of this are scute-8, scute-12, and scute-M-4 (in deleted X 24) ;the two latter show certain semidominant Hairy wing effects, as well as hypomorphic scute characters, but it is as yet uncertain whether these effects are really referable to the same locus or represent group mutation or possibly effects of changed position.

Multiple allelomorphs forming non-quantitative series There are already numerous cases known in which it can be shown that a given hutation has markedly changed a gene only in regard t o certain of the effects which the original gene produced, while another mutation in the same gene changed it more pronouncedly in some other respects. This has been shown par excellence with regard to the various hypomorphic changes possible in the scute locus in the studies on scute allelomorphs carried on by the Moscow geneticists. One of their most important contributions lies in showing the richness of the different patterns of change possible in a given gene, since thus far very few of the numerous allelomorphs are indistinguishable from one another. That the tendency to certain kinds of groupings of effects on the different bristles is partly an expression of cer-

250

PR0,CEEDINGS OF THE SIXTH

tain real features of gene structure, and will. help us to understand the arrangement of gene parts, is also a reasonable conclusion. Attempts to explain the matter in a simple quantitative way, as in GOLDSCHMIDT'S criticisms, or by means of developmental relations, as in the Plunkett-Sturtevant-,Schultz hypothesis of diffusion of influences from a center, fall in the face of the facts. We do not have time to mention the various logical difficulties which the latter hypothesis encounters in its actual working out. Suffice it here to say that a study of numerous gynandromorphs involving various scute allelomorphs has been carried out in our laboratory, chiefly by PATTERSON, and that the results show clearly that the development of bristles, in so far as it is under the influence of the scute gene, is not governed by one or a few centers, but is in its major features autonomous a t the site of each bristle. On the other hand, later work throws grave doubt on the possibility of grouping all the effects into one exact line (this is equally against ,both the unmodified sub-gene hypothesis and the theories of GOLDSCHMIDT, STURTEVANT, et cetera). And the evidence that such a line, if it represents gene parts in a one-to-one correspondence, may be cut without destruction of either piece, is still to be found (see page 222). This still leaves the locus of scute the most suitable yet found for the study of multiple allelomorphism and gene structure, and it leaves the subgene hypothesis, or sQme modification of it, as a possible interpretation, although the way is not as clear and easy as before. I t will, I think, be profitable to follow the method there used, that of concentrating on intensive studies of the different kinds of mutations possible in individual genes, as induced by irradiation and otherwise. Such studies as we have carried out on other loci than scute have shown somewhat similar phenomena, and in some respects amplify our view. For example, the cases now known are fairly numerous in which different recessive mutant allelomorphs of the same locus have effects which are to some extent, or almost wholly, different in their character or in their location on the organism. Thus, mutant allelomorph 1 may affect character A very much and B very little or not at all, while allelomorph 2 affects A little and B much. Such allelomorphs, when crossed, usually form a compound that is more normal than either. For, in respect to each character effect or body region, the more normal effect is usually the more dominant; that is, the compound is usually in each respect more like that allelon~orphwhich has a more nearly. normal effect on that character or region. This was evident, for example, in EMERSON'S (1911) allelomorphs giving different combinations ( = patterns) of red versus white silk, cob, grain, et cetera, in corn. In Drosophila, the first case was that of the truncate series (MULLER1919,

INTERNATIONAL CONGRESS O F GENETICS

25 1

1922b), which concerns not only different regions but different characters, and obeys the same rule throughout. Thus, in this case, the cross of vortex bristles by oblique wings was found to give a compound that was sensibly normal. T o explain those members of this series which showed two or more of the effects a t once, the interpretation of group mutation of neighboring but physiologically entirely distinct genes was early considered but it was rejected, chiefly because studies on the action of modifying genes as well as of "chief" genes a t other loci showed the different developmental effects in question to be physiologically related. In this case, it was also observed that the groupings of effects of different allelomorphs fitted in with no linear series rule. The normal-appearing compound of achaete and scute-1 (found by DUBININ to be allelomorphs) falls under the same category as the vortex-oblique cross. So too may the normal compound of split bristles and recessive notch wings (GLASSand MULLERunpublished), and also (1930a) in the Stubble series o f certain effects observed by DOBZHAKSKY allelomorphs. The list could be considerably extended. There are, however, exceptional cases, in which the compound is not more like the normal in respects in which the two allelomorphs differ. The best case of this is the appearance of leg-like antennae in the compound between aristopedia, which has such an effect, and its allelomorph spineless, which (1929). A few of the missing bristle does not, as found by STURTEVANT effects in scute crosses show a similar tendency; so too does the extra bristle effect in crosses of split bristle and facet-eye (see below). W e now have to report exceptions of the opposite type also, namely, those in which the compound is more like normal in respect to effects in which both allelomorphs are similarly abnormal. One such case is that of lozengeeye in combination with a particular spectacled-eye allelomorph of it. The compound has a practically normal eye but has the female infertility common to both, and their nlutual allelomorphism is further shown by the fact each gives a distinctly mutant eye type when crossed with still other memand MULLER1930, ACOL 1930). Anbers of the series (see PATTERSON other case is that of the ommatidial disarrangement in split Gristles and facet-eye. Both of these mutants cause ommatidial disarrangement, yet (as with spectacled and lozenge) the compound has a normal eye (MULLER unpublished). Their allelomorphism is shown not only by their linkage but by their behavior with other mutual allelomorphs (notches) and by the appearance of extra bristles in the compound, as in split bristles by itself (see above). I n such cases as these, we must draw the conclusion that the two allelomorphs, although acting on the very same body region, and having superficially similar effects on that region, nevertheless attain these effects

252

PROCEEDINGS OF THE S I X T H

through the intermediation of qualitatively different developmental processes. Further studies of the relations in such series are needed. Ultimately, too, we must undertake the still more difficult study of the effects of successive mutations in the same gene, to discover, if possible, principles governing their continued evolution. Such evolution, as I see it, impIies the possibility of qualitative change in the gene as a necessary condition. The foregoing iIlustrations, if taken together, afford, I believe, considerable experimental evidence for the existence of such a phenomenon, both as a natural occurrence and as a result of irradiation. And this conclusion remains Iikely no matter whether the mutational effects of irradiation are of a direct or an indirect nature. For the rest, I fear that the present paper has raised far more questions than it has solved. But if some of these questions may thus have been opened to attack, our time may not have been wasted. The author wishes to acknowledge with thanks the assistance of the Committee on the Effects of Radiation on Living Organisms, of the NATIONAL RESEARCH COUNCILof the United States, in the prosecution of experiments referred to in the foregoing. LITERATURE CITED

AWL, I. J., 1930 Evidence of the divisibility of the gene. (Abst.) Anat. Rec. 47:387. 1932 Das Sichtbarmachen der verborgenen allelomorphen scute-Teile mit Hilfe von Faktorenausfallen (deficiencies). Biol. Zbl. 52:349-367. BOLEN,H. R., 1931 A mutual translocation involving the fourth and the X chromosomes of Drosophila. Amer. Nat. 65:417-422. BRIWES,C. B., 1917 Deficiency. Genetics 2:445-465. 1919 Duplications. Abst. Anat. Rec. 15:357. 1923 The translocation of a section of chromosome I1 upon chromosome I11 in Drosophila. Abst. Anat. Rec. 24:426. BURKART, A,, 1931 Investigaciones geneticas sobre una nueva mutaci6n de Drosophila melanogaster determinante de excepciones hereditarias. Rev, de 1. Fac, de Agron. y Veter. Univ. de Buenos Aires, Ent. 11. Tomo 7:393-491. 1932 (in press in Biol. Zbl.) DDBZHANSKY, T., 1929 Genetical and cytological proof of translocations involving the third and the fourth chromosomes of Drosophila melanogmfer. Biol. Zbl. 49:408-419. 1930a The manifold effects of the genes Stubble and stubbloid in Drosophila melanogaster. Z . indukt. Abstamm.-u VererbLehre. 54:427-457. 1930b Translocations involving the third and the fourth chromosomes of Drosophila melanogaster. Genetics 15:347-399. 1931 Translocations involving the second and the fourth chromosomes of Drosophila melanogarter. Genetics 16:629-658. 1932 The baroid mutation in Drosophila melamgasfer. Genetics 17:369-392. DOBZHANSKY T., and STURTEVANT, A. H., 1931 Translocations between the second and third chromosomes of Drosophila and their bearing on Oenothera problems. Pub. Carnegie Instn. 421 29-59.

INTERNATIONAL CONGRESS O F GENETICS

253

1932 Changes in dominance of genes. lying in duplicating fragments of chromosomes. Proc. Sixth Int. Congress Genetics 2:45-46. DUBININ,N. P., 1930a Issled, stup. allelomorphisma. Tsentrovaya teoria gena achaete-scute. J. Exp. Biol. (Russ.) 6. 1930b On the nature of artificially deleted X chromosomes. J. Exp. Biol. (Russ.) 7~365-368. EFROIMSON, W. P., 1931 Die transmutierende Wirkung der X Strahlen und das Problem der genetischen Evolution. Biol. Zbl. 51:491-506. EMERSON, R. A., 1911 Genetic correlation and spurious allelomorphism in maize. Ann. Rep. Nebraska Agric. Exp. Sta. 24: FERRY,L., N. T. SHAPIRO and B. N. SIDOROFF,1930 On the influence of temperature on the process of mutation, with reference to Goldschmidt's data. Amer. Nat. 64:570574. FISHER, R. A., 1928 The possible modifications of the responses of the wild type to recurrent mutations. Amer. Nat. 62:115-126. 1930 The genetical theory of natural selection. Oxford Univ. Press. 272 pp. FORD,E. B., 1930 The theory of dominance. Amer. Nat. 64:560-565. GERSHENSON, S. M., I930 Phenomenon of reinversion in the sex chromosome of D. melanogaster. Reports to Fourth All-Union Congress of Zoologists, etc., p. 7 (Russ.) GLASS,H. B., 1932 A study of dominant mosaic eye-color mutants in Drosophila. Proc. Sixth Int. Congress Genetics 2:62-63. GOLDSCHMIDT, R., 1929 Experimentelle Mutation und das Problem der sogennanten Parallelinduktion. Biol. Zbl. 49:437-448. 1931 Die entwicklungsphysiologische Erklarung des Falls der sogenannten Treppenallelomorphe des Gens scute von Drosophila. Biol. Zbl. 51:507-526. GRAUBARD, M. A., 1932 Inversion in Drosophila melanogaster. Genetics 17:81-105. HALDANE, J. B. S., 1930 A n6te on Fisher's theory 'of the origin of dominance and on a correlation between dominance and linkage. Amer. Nat. 64:87-90. HAMLETT, G. W. D., 1926 The linkage disturbance involved in the chromosome translocation I of Drosophila, and its probable significance. Biol. Bull. 51:435-442. HANSON,F. B. and F. HEYES,1929 An analysis of the effects of the different rays of Radium in producing lethal mutations in Drosophila. Amer. Nat. 63:201-213. 1932 Radium and lethal mutations in Drosophila. Amer. Nat. 64:335-345. HANSON,F. B., F. HEYESand E. STANTON, 1931 The effects of increasing X-ray voltages on the production of lethal mutations in Drosophila melanogaster. Amer: Nat.

65:134-142.

J o ~ s V., , 1930 Studien zum Evolutionsproblem. I. Biol. Zbl. 50:541-554. MCCLINTOCK,B., 1931 Cytological observations of deficiencies involving known genes, translocations and an inversion in Zea mays. Missouri Agric. Exp. Sta. Bull. 163: 30 pp. MOHR,0. L., 1919 Character changes caused by mutation of an entire region of a cbromosome in Drosophila. Genetics 4:275-282. 1923 A genetic and cytological analysis of a section deficiency involving four units of the X chromosome in Drosophilo melanogaster. 2. indukt. Abstamm.-u VererbLehre. 32:108-232. 1927 Exaggeration and inhibition phenomena. Norsk Videnskaps-Akad. i Oslo. I. m. 6:19 pp. 1929 Exaggeration and inhibition phenomena encountered in the analysis of an autosoma1 dominant. Z. indukt. Abstamm.-u. VererbLehre. 50:113-200. MORGAN, T.H., A. H. STURTEVANT and C. B. BRIDGES,1927 The constitution of the germ material in relation to heredity. Carnegie Instn. Year Book 26:284288.

254

PROCEEDINGS OF T H E S I X T H

and J. SCHULTZ,1930, 1931 The constitution of the germinal MORGAN, T. H., C. B. BRIDGES material in relation to heredity. Carnegie Instn. Year Book 29:352-359 and 30: 408-415. MuH. J., 1919 A series of allelomorphs in Drosophila with non-quantitative relationships. Address before Amer. Soc. of Nat. at Princeton, December 31, 1919. Title in Science, 1920. 1922a A lethal gene which changes the order of the loci in the chromosome map. Address before Genetics Sections at Toronto, December, 1921; title in Anat. Rec, 23 :83. 1922b Variation due to change in the individual gene. Amer. Nat. 56:32-50. 1927 Artificial transmutation of the gene. Science 66:84-87. 1928a The effects of X-radiation on genes and chromosomes. Science 67:82-83. 1928b The problem of genic modification. Z. indukt. Abstamm.-u Vererblehre. Supl. Bd. 1934-260. 1928c The production of mutations by X-rays. Proc. Nat. Acad. Sci. Washington 14:714-726. 1928d The measurement of gene mutation rate in Drosophila, its high variability, and its dependence upon temperature. Genetics 13:279-357. 1930a Radiation and genetics. Amer. Nat. 64:220-251. 1930b Types of visible variations induced by X-rays in Drosophila. J. Genetics 22999334. MULLER,H. J., and E. ALTENBURG, 1919 The rate of change of hereditary factors in Drosophila. Proc. Soc. Exper. Biol. and Med. 17:lO-14. 1930 The frequency of translocations produced by X-rays in Drosophila. Genetics 15:283-311. MULLER, H. J., B. B. LEAGUE and C. A. OFFERMANN, 1931 Effects of dosage changes of sexlinked genes, and the compensatory effects of other gene differences between male and female. Abst. Anat. Rec. 51:110. MULLER,H. J., and L. M. Mm-SMITH,1930 Evidence that natural radioactivity is inadequate to explain the frequency of "natural" mutations. Proc. Nat. Aoad. Sci. Washington 16:277-285. MULLER,H. J., and T. S. PAINTER1929 The cytological expression of changes in gene alignment produced by X-rays in Drosophila. Amer. Nat. 63:193-200. 1932 The differentiation of the sex chromosomes of Drosophila into genetically active and inert regions. 2. indukt. Abstamm.-u. VererbLehre. 62:316-365. MUUER, H. J., and W. S. STONE,1930 Analysis of several induced gene-rearrangements involving the X chromosome of Drosophila. Abst. Anat. Rec. 47:. NASARENKO, I. I., 1930 Ein Fall wahrscheinlicher Verdoppelung eines Chromosomstiickes bei Drosophila melanogaster. Biol. Zbl. 50:385-392. OFF-ERMANN, C. A,, and H. J. MULLER,1932 Regional differences in crossing over as a function of the chromosome structure. Proc. Sixth Int. Congress Genetics 2:143145. OLIVER,C. P., 1930a The effect of varying the duration of X-ray treatment upon the frequency of mutation. Science 71:44-46. 1930b Complex gene rearrangements induced by X-rays. Abst. Anat. Rec. 47: 1932 An analysis of the effect of varying the duration of X-ray treatment upon the frequency of mutations. Z. indukt. Abstamm.-u. VererbLehre. 61:447-488. PAINTER, T. S., and H. J. MULLER,1929 Parallel cytology and genetics of induced translocations and deletions in Drosophila. J. Hered. 20:287-298. PATTERSON, J. T., 1931 Continuous versus interrupted radiation and the rate of mutation in Drosophila. Biol. Bull. 61:133-138. 1932a A new type of mottled-eyed Drosophila due to an unstable translocation. Genetics 17:38-59. 1932b A gene for viability in the X chromosome of Drosophila. Z. indukt. Abstamm.-u. VererbLehre. 60:125-136.

INTERNATIONAL CONGRESS O F GENETICS

255

1932c Lethal mutations and deficiencies produced in the X chromosome of Drosophila melanoguster by X-radiation. Amer. Nat. 66:193-206. PATTERSON, J. T., and H. J. MULLER,1930 Are "progressive" mutations produced by Xrays? Genetics 15:495-578. PLOUGH,H. H., and P. T. IVES,1932 New evidence of the production of mutations by high temperature, with a critique of the concept of directed mutations. Proc. Sixth Int. Congress Genetics 2:156-158. PLUNKETT,C. R., 1926 The interaction of genetic and environmental factors in development. J. Exp. Zool. 46:181-244. 1932 Temperature as a tool of research in phenogenetics: methods and results. Proc. Sixth Int. Congress Genetics 2:158-160. ROKITZKY,P. T., 1930 Uber das Hervorrufen erblicher Veranderungen bei Drosophila durch Temperatureinwirkung. Biol. Zbl. 50:554-566. SAFIR,S. R., 1913 A new eye-color mutation in Drosophila. Biol. Bull. 25:45-51. SEREBROVSKY, A. S., 1929 A general scheme for the origin of mutations. Amer. Nat.

53 :374-378. SEREBROVSKY, A. S., and N. P. DUBININ, 1930 X-ray experiments with Drosophila. J. Hered. 21 :259-265. STADLER, L. J., 1930 Some genetic effects of X-rays in plants. J. Hered. 21:2-19. STERN,C., 1929 Uber die additive Wirkung multipler Allele. Biol. Zbl. 49:231-290. 1931 Zytologisch-genetische Untersuchungen als Beweise fiir die Morgansche Theorie des Faktorenaustausches. Biol. Zbl. 51547-587. STERN,C., and OGURA,S., 1931 Neue Untersuchungen iiber Aberrationen des Y chromosoms von Drosophila melanoguster. Z. indukt. Abstamm.-u. VererbLehre. 58231-

121. STURTEVANT, A. H., 1926 A crossover reducer in Drosophila melanogaster due to inversion of a section of the third chromosome. Biol. Zbl. 46:697-702. 1929 The genetics of Drosophila simulans. Pub. Carnegie Instn. 399:l-62. 1931 Known and probable inverted sections of the autosomes of Drosofihila melanogaster. Pub. Carnegie Instn. 421:l-27. 1932 Paper in these Proceedings, Val. I. 1930 Reciprocal translocations in Drosophila and STURTEVANT, A. H., and T. DOBZHANSKY, their bearing on Oenothera cytology and genetics. Proc. Nat. Acad. Sci. Washington. 16533-536. STURTEVANT, A. H., and C. R. PLUNKETT, 1926 Sequence of corresponding third-chromosome genes in Drosophila melanogaster and D. sirnulatax. Biol. Bull. 5056-60. STURTEVANT, A. H., and J. SCHULTZ, 1931 The inadequacy of the sub-gene hypothesis of the nature of the scute allelomorphs of Drosophila. Proc. Nat. Acad. Sci. Washington 17:265-270. THOMPSON, D. H., 1925 Evidence on the structure of the gene. Amer. Nat. 59:91-94. 1929 The side-chain theory of the structure of the gene. Abst. Anat. Rec. 44:288. TIMOFEEFF-RESSOVSKY, N. W., 1929 The effect of X-rays in producing somatic genovariations of a definite locus in different directions in Drosophila melanogasfer. Amer. Nat. 63:118-124. 1931a Reverse genovariations and gene mutations in different directions. J. Hered. 21:67-70. 1931b Die bisherigen Ergebnisse der Strahlengenetik. Ergebn. d, medii. Strahlenforschung 5:130-228. VAN ATTA, E. W., 1932 Dominant eye-color in Drosophila. Amer. Nat. 66:93-95. WRIGHT, S., 1929a Fisher's theory of dominance. Amer. Nat. 63974-279. 1929b The evolution of dominance. Amer. Nat. 63556-561. ZELENY,C., 1918 Full-eye and emarginate-eye from bar-eye in Drosophila without change of the bar gene. Anat. Rec. 14539.

256

PROCEEDINGS OF THE SIXTH

T H E CYTOLOGICAL MECHANISM FOR CROSSING OVER Karl Sax, Arnold Arboretum, Haward University, Boston, Massachusetts

The genetic analysis of Drosophila by MORGAN and his associates has contributed more to our knowledge of chromosome behavior at meiosis than have the investigations of the cytologists. Any theory of chromosome pairing and crossing over must conform to the rigid requirements of the geneticist. As a foundation for the discussion of the mechanism of crossing over, it is essential to consider the facts which have resulted from the genetic investigations of Drosophila during the past twenty years. I t is known that the four linkage groups in Drosophila melamgmter correspond to the four pairs of chromosomes. The genes are arranged in a linear order in these chromosomes. Chromosome pairing must involve a gene by gene association in order to account for the great precision of crossing over. The frequency of crossing over is not the same for all regions of the chromosome and is reduced in the region of the spindle-fiber attachment. Chromosome pairing and disjunction are regular in both males and females, but no crossing over occurs in the male. The phenomenon of interference is of special importance in any interpretation of the mechanism of crossing over. If one crossover occurs, a second one is never found in adjacent regions of the same chromosome (STURTEVANT 1913). Interference is complete for 10 to 20 units, depending on the region of the chromosome involved (MULLER1916, WEINSTEIK 1918). The first meiotic division in Drosophila is usually, if not always, reductional at the spindle fiber attachment point, that is, the sister chromatids are held together at the fiber constriction at this division (BRIDGES and ANDERSON 1925, ANDERSON 1925, 19-9, L. V. MORGAN 1925, REDFIELD 1930, RHOADES 1931, STURTEVANT 1930). Crossing over occurs at early prophase of meiosis, after the chromosomes have paired, and between only two of the four chromatids at any one locus 1925). It has also been shown that (BRIDGES 1916, BRIDGES and ANDERSON only two of the four chromatids are involved in a crossover in Zea (RHOADES 1932) and in Habrobracon (WHITINGunpublished). Any factor which reduces crossing over is.associated with a reduction in 1929). regularity of pairing and disjunction (GOWEN1928, ANDERSON Crossing over is decreased in individuals heterozygous for an inversion (STURTEVANT 1926). Crossing over is not limited to two of the four chromatids throughout their length, because crossing over is found in more than 50 percent of the

INTERNATIONAL CONGRESS O F GENETICS

257

emerging X chromosomes of Drosophila, and about 75 percent of the third 1931, chromosomes have one or more crossovers (ANDERSON and RHOADES REDFIELD1930). Crossing over occurs more or less at random between homologous chromatids (ANDERSON 1925). There is no evidence that crossing over occurs between sister chromatids (STURTEVANT 1925, 1928). I t has generally been assumed that genetic crossing over is correlated with an actual physical interchange of chromosome segments. This assumpand tion has been confirmed by the brilliant investigations of CREIGHTON MCCLINTOCK(1931) with Zea, and of STERN(1931) with Drosophila. THE MECHANISM FOR CROSSING OVER

Janssens' interpretation of chiasma formation JANSSENS'(1924) partial chiasmatypy hypothesis has been supported ( 1930,1931b), MAEDA(1930a), by BELLING( 1931a, 1931b), DARLINGTON ana others. It is assumed that every chiasma represents a crossover which has occurred between two of the four chromatids at pachytene. As the homologous chromosomes open out at diplotene, only sister chromatids are paired, and the chiasmata indicate the point of interchange in crossing over. The only plausible explanation of how crossing over could occur, on JANSSENS'theory, has been presented by BELLING(1931b). According to BELLING,the homologous chromosomes pair as single threads. Half twists occur in the paired chromosomes before the new chromatids are formed. Each chromomere, or gene, then divides. The connecting fiber between chromomeres may remain with the old chromomere or pass to the new one at random. New connecting fibers are then formed to unite the free chromomeres. At the half twist formed by the original chromatids, the connecting fibers unite the free genes by the shortest path, so that a crossover occurs at random between non-sister chromatids. I t is, of course, also necessary to assume that sister-strand crossovers are very frequent. As the writer (SAX 1932a) has pointed out, such a random assortment of connecting fibers would result in numerous twists in paired chromatids in both somatic and meiotic chromosomes. BELLING'Stheory can be made more plausible if it is assumed that crossovers always involve the two new chromatids, as BELLING(1931a) suggested in his second paper on crossing over. An occasional crossover between sister chromatids would result in apparent random crossing over between homologous chromatids (SAX 1932a). An interchange of segments between two of the four chromatids should invariably produce an asymmetrical arrangement of the chromatids. This

258

PROCEEDINGS OF T H E SIXTH

situation was recognized by JANSSENS,but has received little or no attention from the recent supporters of his theory. The asymmetrical relations of the chromatids, if each chiasma represents a crossover, are shown in figure 1, diagram A. Chromatids which were adjacent at the four-strand stage form the cross at each chiasma. As a result of the sister-strand crossover at the second internode, the genetically detectable crossovers are not confined to two of the four chromatids. Sister chromatids are always paired at early diplotene.

As the chromosomes open out in subsequent stages of meiosis, the chiasmata are often terminalized until, at metaphase, only terminal or sub-terminal chiasmata are found (DARLINGTON 1931, 1932, GAIRDNER and DARLINGTON 1932, and others). I t is assumed by DARLINGTON that the attachment loop, in expanding, pushes all the others together at the end. As a result of such terminalization, a rather complicated association would result where two or more chiasmata are pushed together at metaphase (figure 1, diagram B). The two uppermost chromatids form the cross at the chiasma to the left of the spindle-fiber attachment. If this chiasma were opened out to form the typical cross-shaped figure, there would be a twist in one pair of chromatids, as shown in figure 1, diagram C. If only one chiasma is formed, it is possible that the torsion would rotate the paired chromatids at the

INTERNATIONAL CONGRESS O F GENETICS

259

proximal end so that a symmetrical cross would result; but where a chiasma is formed on each side of the spindle fiber, few symmetrical chiasmata would be expected. Two consecutive reciprodal chiasmata on the same side of the fiber should produce interlocking of chromatids a t metaphase in half the cases if one chromatid passes above or below the other a t random. The chiasmata represented in figure 1, diagrams A to C, are all reciprocal crossovers, which is the type expected most frequently on BELLING'Shy(1932) in Stenopothesis and observed most frequently by DARLINGTON bothrus. These diagrams are based on BELLING'Slatest theory, with a few favorable modifications. If, however, the interchange of segments occurs in the chromatids which form the cross, the results will be the same, so far as the relation of the chromatids is concerned. Either a reciprocal (adjacent) or an equational (diagonal) crossover will result in an asymmetrical relation of the chromatids (figure 1, diagrams D and E ) . Viewed from the end, each chromatid will not lie in the same quadrant a t all loci. In the diagram of the equational crossover, as viewed from the left end at the four-strand stage (figure 1, diagram I?), the chromatid a t the upper right quadrant changes to the lower left quadrant, owing to the crossover, while in the reciprocal crossover the chromatid in the upper right quadrant changes to the upper left quadrant. BELLING'Stheory of crossing over seems to be untenable for several reasons. H e assumes, as does DARLINGTON, that homologous chromosomes pair as single threads a t meiosis. According to KAUFMANN(1931), the chromosomes a t the telophase of the last premeiotic division are two-parted in several plant species. In several Orthopteran species the chromosomes are split longitudinally before pair ing a t meiosis (ROBERTSON 1916, 193la, 1931b, MCCLUNG1928). RANDOLPH(1932) finds a longitudinal split in the early leptotene threads in a haploid Zea plant. These observations can not be reconciled either with BELLING'Stheory of crossing over or with DARLINGTON'S theory of meiosis. The random assortment of connecting fibers between chromomeres, postulated by BELLING,would mean that sister strand crossovers and many twists in paired chromatids would be expected. There is no evidence that sister strand crossing over occurs, and few twists are found in paired chromatids of either somatic or meiotic chromosomes. The third objection to BELLING'Stheory-and this applies to any interpretation of JANSSENS'hypothesis-is the prevalence of symmetrical relations of the chromatids in many genera, even where several chiasmata are found in a bivalent chromosome. It is also difficult t o account

260

PROCEEDINGS O F THE SIXTH

for interference in crossing over and the absence of crossing over in the Drosophila male, on any modification of JANSSENS'theory. DARLINGTON (1932) has recently revived the torsion hypothesis to explain how crossing over occurs. It is significant that he shows no diagrams illustrating how this mechanism might cause exact crossing over between only two of the non-sister chromatids at any one locus. McClungJs theory of chiasima formation According to MCCLUNG,chiasmata are caused by the alternate opening out of pairs of sister and non-sister chromatids. This interpretation has

(1916), WILSON(1925), been supported by WENRICH( 1916), ROBERTSON BELAR(1928), CAROTHERS (1926), and NEWTON(1926). As MCCLUNG (1927) has pointed out, such an origin of the chiasmata should usually result in a symmetrical relation of the chromatids. The apparent cross a t a chiasma is formed by chromatids which were diagonal at the fourstrand stage, and each chromatid may lie in the same quadrant at all loci (figure 2, diagram A ) . The relations of the chromatids at metaphase, when no crossover has occurred, are shown in figure 2, diagram E.

INTERNATIONAL CONGRESS O F GENETICS

261

The writer's theory of crossing over The writer (SAX1930, 1932a) has suggested that crossing over is caused by breaks in two of the four chromatids at a chiasma. A t very early diplotene many chiasmata may be formed in each bivalent chromosome, as shown in figure 2, diagram A. These chiasmata are formed by alternate pairing of sister and non-sister chromatids. Occasional half-twists in paired sister chromatids permit more or less random crossing over between any two homologous (non-sister) chromatids. Between the earliest diplotene stage and diakinesis there is considerable reduction in the number of chiasmata. Unequal opening out of certain internodes will result in the elimination of certain chiasmata by cancellation, as shown in figure 2, diagram B. I t is also possible that some chiasmata may be terminalized to such an extent that they will pass off the end of the chromosome. Crossovers occur only when two of the chromatids break at a chiasma, and the segments reunite in a new association, as shown in figure 2, diagrams A and B. When no crossover occurs, the chromatids a t a chiasma are symmetrical, each chromatid lies in the same quadrant on both sides of the chiasma, and the apparent cross a t the chiasma is formed by chromatids which were diagonal a t the four-strand stage. If the distal ends of the chromosomes open out in the same plane, a symmetrical cross is formed, as shown in figure 2, diagram E. A crossover, or a twist in sister threads, will produce an asymmetrical arrangement of the chromatids, as shown in figure 2, diagrams B and C. A crossover formed between two intact chiasmata will result in interlocking of chromatids a t metaphase (figure 2, diagram D). This theory of crossing over seems to meet the rigid genetic requirements, and is in accord with most of the cytological evidence. The lengths of the internodes, or loops, between chiasmata would account for interference. The reductional loop at the fiber attachment point is usually larger than the others, which would reduce crossing over in the spindle-fiber region. Chiasma formation and normal chromosome pairing can occur without crossing over, as is presumably the case in the Drosophila male. Crossovers between sister chromatids should be rare. Reciprocal and equational crossovers may occur with equal'frequency (resulting in a genetic ratio of 2 : l in attached XX's), but under certain conditions there should be an excess of equational crossovers. A symmetrical relation of the chromatids would be expected where no crossover has occurred, but half twists in paired sister chromatids or a crossover would produce asymmetrical chiasmata.

262

PROCEEDINGS OF THE SIXTH CYTOLOGICAL EVIDENCE

The arrangement of chromatids If MCCLUNG'Stheory of chiasma formation is correct, the chromatids should usually show a symmetrical arrangement at each chiasma if little or no crossing over has occurred. Chromosome configurations which seem to support this interpretation of chiasma formation have been shown in Orthopteran species by SUTTON( 1902), WENRICH( 1916), ROBERTSON ( 1916), JANSSENS ( 1924), BELAR( 1928), CAROTHERS (1926), MCCLUNG (1927), DARLINGTON and DARK(1932), and others. Similar figures have been shown for other groups of animals (WILSON1925). In plant species the individual chromatids are not clearly differentiated until metaphase, and then only when the smear technique is used. This technique has long been used by the zoologists, but has been adopted by pointed out its application to plant cells in botanists only since TAYLOR 1930), 1924. Symmetrical chiasmata have been found in Gasteria (TAYLOR Uvularia (BELLING1926), Tulipa (NEWTON1926), Paeonia (SAX1932b), Secale (SAX1930), and in Larix (H. J. SAX1932). According to NEWTON, the hypothesis "which explains the diakinetic figures as due to the opening out in two planes at right angles of what are originally four parallel chromatids, is adequate to explain the events of diakinesis and division in Tulipa and Fritillaria." Asymmetrical chiasmata have been described in Orthopteran species by JANSSENSand DARLINGTON, but these figures are apparently not typical for this group. As MCCLUNGhas pointed out, the clearest figures shown by JANSSENS show the chromatids in the same quadrants at all loci. DARLINGTON and DARK (1932) believe that most of the chiasmata in Stenobothrus are reciprocal and involve the same two chromatids at successive chiasmata. Of the two figures referred to, one (figure 7, chiasmata A and B) shows clearly that both chiasmata are symmetrical and diagonal, as is the case in most of the figures shown, where the chromatids can be observed. The interlocking of chromatids at metaphase does indicate that some crossing over has occurred. MZDA (1930a) finds a large proportion of asymmetrical figures in Lathyrus chromosomes, as might be expected, since crossing over is known to occur in this genus. Crossing over is also indicated by interlocked chromatids at metaphase. Asymmetrical chiasmata are frequently found in Paeonia chromosomes (SAX1932b). In most cases these figures could be attributed to half twists in sister chromatids, but some figures do seem to support JANSSENS' hypothe-

INTERNATIONAL CONGRESS O F GENETICS

263

sis. These exceptional figures may be due to an association of non-sister chromatids at the fiber attachment point, accompanied by a half twist in one pair of sister chromatids. The prevalence of symmetrical chiasmata in the Orthoptera and in certain plant species seems to indicate that JANSSENS'partial chiasmatypy theory of chiasma formation is untenable. The asymmetrical chiasmata found in both plant and animal species would be expected on MCCLUNG'S theory of chiasma formation if twists occur in sister chromatids, or if crossovers occur in some of the chiasmata which disappear between early diplotene and metaphase stages of meiosis. If no crossovers occur, almost all the chiasmata may be symmetrical, as seems to be the case in some species. Reduction in chias~+za frequency I f crossing over is caused by breaks in the chiasmata, as the writer assumes, then there should be a reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis. Reduction in chiasma frequency can also be attributed to the meeting and cancellation of chiasmata, as shown in figure 2, diagrams A and B. There is also the possibility that some chiasmata may pass off the ends of the chromosomes. At earliest diplotene there are numerous nodes and internodes along the bivalent chromosomes in many species of plants and animals. The frequency of these nodes, most of which are probably chiasmata, may be as high as 7 or 8 in certain chromosomes. Bivalent chromosomes with numerous nodes and internodes at early diplotene have been shown in Stenobothrus by JANSSENS(1924, figure 242), in reptilian chromosomes (1932), and in Ophyotrocha by GRBGOIREand DETON by NAKAMURA (1906). In plant species, chromosomes with an apparent high chiasma fre1904), Lilium and' Allium quency have been shown in Pinus (FERGUSON (BERGHS1904), Crepis (BABCOCK and CLAUSEN1929), Tulipa (NEWTON 1926), Nothoscordum (BEAL1932), and in Callisia ( S A X1932a). I n these genera the chiasma frequency may be reduced from five or six or more a t early diplotene to one or two at metaphase. A reduction in chiasma frequency between diplotene or early diakinesis and metaphase has been found 1932), Lilium (BELLING1931b), in Tulipa (NEWTON1926, DARLINGTON Primula ( DARLINGTON 1931a), Rosa (ERLANSON 1931) , Matthiola (PHILP and DARLINGTON 1932), and and HUSKINS1931), Campanula (GAIRDNER Callisia ( S A X1930). I n these genera the average number of chiasmata lost per bivalent between diplotene and metaphase is somewhat more than one. The loss of chiasmata reported in most cases would undoubtedly have been

264

PROCEEDINGS O F T H E SIXTH

greater if the earliest diplotene stages were favorable for examination of chiasma frequency. In some genera there is probably little reduction in chiasma frequency between early diplotene and metaphase. In these cases the chiasmata should be symmetrical, and little crossing over would be expected. The reduction in chiasma frequency is often associated with an increase in the proportion of terminal chiasmata, and, in many species, only terminal chiasmata are found at metaphase, even though there may be from three and DARK (1932), in disto five chiasmata at diplotene. DARLINGTON cussing chiasma terminalization, assume that "the attachment loop, in expanding, pushes all the others together at the end." But if each chiasma represents a crossover, the accumulation of two or more chiasmata would often result in a complex association of chromatids, as shown in figure 1, diagram B. In numerous species, terminal or sub-terminal chiasmata show no interlocking or asymmetrical relations of the chromatids, even where three or more chiasmata are found at diplotene. If two crossover chiasmata on the same side of the spindle fiber are terminalized without passing off the ends of the chromosomes, they will be reduced to one interlocked chiasma, or canceled, depending on the types of crossovers. If each chiasma is reciprocal and both crossovers occur in the same two chromatids, the two chiasmata will be canceled, o r reduced to one, with equal frequency. If crossing over is at random between homologous chromatids, the two chiasmata would be reduced to one interlocked chiasma in 50 percent of the cases, and reduced to none in 50 percent of the cases. In Primula and Campanula the chiasma frequency is reduced from three or more at diplotene to two terminal chiasmata at metaphase, but there is no evidence of interlocked chromatids in these genera. On the writer's hypothesis, the reduction in chiasma frequency between early diplotene and metaphase can be attributed to cancellation of chiasmata and breaks in some of the chiasmata as the early diplotene loops open out. Two chiasmata might be reduced to none by cancellation and reduced to one if the other chiasma were broken. Both the genetic and the cytological evidence support MORGAN'S (1925) suggestion that crossing over is an accidental by-product of meiosis. SANSOME (1932) has described a case of chiasma formation in a ring of six chromosomes in Pisum, which seems to support the partial chiasmatypy hypothesis. In 62 out of 78 figures one or two chiasmata were found in the "X segment" of the ring, resulting in a figure-of-eight configuration. Occasionally two chiasmata were found in this region, but the exact

INTERNATIONAL CONGRESS OF GENETICS

265

proportions of double and single chiasmata are not given, presumably owing to the difficulty in studying the chromatids in this region. On the writer's theory, two chiasmata must be formed in the X segment if only two chromosomes can pair at the same locus. I f one of the chiasmata breaks, the remaining one represents a crossover. The remaining chiasma would be under a special strain, but it seldom breaks. A much more critical case of the same type has been shown in Zea by BEADLE(1932). Chiasma frequency was obtained at both diplotene and metaphase, and the crossover length of the segment was known. The ring or chain of 4 chromosomes, resulting from segmental interchange, is so constituted that crossing over in one arm is limited to about 12 crossover units between the translocation point and the region where a teosinte chromosome segment has been introduced. When a chiasma is formed in this region, a ring with two free arms is found a t diakinesis and metaphase, and when no chiasma is present, the four chromosomes are arranged in a chain. At diakinesis 20 percent of the associations of four chromosomes were in the form of rings, but a t metaphase only 10 percent of these chromosomes were rings. On the writer's theory, a t least two chiasmata must be formed in this wz-translocation region. K O terminalization of chiasmata can occur because of the change in homology of the distal segments. As the ring opens out, all the chiasmata in this region may disappear by cancellation (compare figure 2, diagram B). If one chiasma breaks, the remaining chiasma represents a crossover. This chiasma can not be canceled or terminalized. The frequency of such a crossover chiasma in the wz-translocation segment a t diakinesis (20 percent) corresponds with the crossover frequency found in the derived Zea-teosinte chromosomes as bivalents. But at metaphase the chiasma frequency was only 1 0 percent. If the frequency of ring formation a t meta~haseis significant, this must mean that two chiasmata were often present at diakinesis and were canceled before metaphase, or that the crossover chiasma was broken. If the crossover chiasma breaks, a double crossover might result, but the two crossovers would probably be too close together to be detected. The effect of a break in the single crossover chiasma on the crossover length of the segment would depend on types of double crossovers obtained. If chiasma frequency is high a t early diplotene, the writer's explanation of chiasma formation in the wz-translocation segment is not improbable. Two chiasmata might be formed. Unless one chiasma breaks, the original chiasmata probably would be canceled, owing to the strain imposed by the

266

PROCEEDINGS OE THE SIXTH

opening of the ring and limitations in terminalization. If a chiasma breaks, cancellation of the remaining chiasma is impossible, and since the remaining crossover chiasma can not be terminalized, it is found in the ws-translocation region a t diakinesis and metaphase.

Behavior of unequal homologues In all recorded cases of pairing of unequal homologues, the long chromatids are paired with long and the short ones are paired with short a t meiosis. With only one chiasma and a terminal spindle-fiber attachment, it is difficult to explain the occurrence of both pre- and post-reduction in the same bivalent (SAX 1932a, b) as found by WENRICH(1916). This be(1932) on the assumption that the havior is explained by DARLINGTON fiber attachment is median, but the assumption is clearly erroneous, as shown by WENRICH'Sfigures of somatic chromosomes in Phrynotettix. On the MCCLUNGinterpretation of chiasma formation, the chromatids must separate equationally at the fiber attachment, where a single chiasma is formed between the fiber and the unequal segments if no crossing over has occurred. This assumption may be valid for some species, although it is apparently not the case in Drosophila or Zea.

Interlocked bivalents Interlocked non-homologous bivalents have been found in Oenothera, ahd Campanula, Tradescantia, and other genera. According to GAIRDNER DARLINGTON ( 1932), three types of interlocking would be expected : ( 1 ) proximal interlocking, where the loops containing the spindle fiber are involved; ( 2 ) distal interlocking, or locking of terminal loops; and ( 3 ) proximal-distal interlocking. If chiasmata are formed by alternate opening of sister and non-sister chromatids at diplotene, interlocking of bivalents could occur only in alternate internodes where sister chromatids are paired. If each chiasma represents a crossover and sister chromatids are always paired at early diplotene, interlocking can occur at any internode. In Campanula (GAIRDNER and DARLINGTON 1932) from two to six chiasmata are found in each bivalent, and the most frequent number seems to be three at diplotene. If each chiasma represents a crossover, distal or proximaldistal interlocking should be at least as frequent as proximal interlocking. theory of chiasma formation is correct, distal or proxiBut if MCCLUNG'S mal-distal interlocking should be very much less frequent than proximal interlocking. These conclusions are based on the assumption that chiasmata do not pass off the ends of the chromosomes but are pushed together a t

INTERNATIONAL CONGRESS O F GENETICS

267

the ends of the chromosomes, as postulated by DARLINGTON and DARK find, in Campanula, that "proximal (1932). GAIRDNER and DARLINGTON interlocking occurs in about 20 percent of nuclei in homozygous groups. Distal interlocking seems to be much rarer." Similar observations have been made in Oenothera by CATCHESIDE ( 1931 ) . Proximal interlocking is common in Tradescantia, but distal interlocking is rare (SAX and ANDERSON theory of chiasma forma1932). These results indicate that MCCLUNG'S tion is correct, or that chiasmata do pass off the ends of the chromosomes before metaphase. According to GAIRDKER and DARLINGTON, interlocking should occur between the chromatids of paired or unpaired chromosomes if the chromatids ever separate equationally at diplotene. This conclusion is obviously untenable ( S A Xand ANDERSON 1932). GENETIC EVIDENCE

Non-disjunction and crossing ovcr

In high non-disjunction lines of Drosophila, crossing over is greatly reduced in both the normal progeny and in the exceptional females (ANDERSON 1929). The genetic evidence indicates that non-disjunction is caused by a failure of chromoson~epairing at the first meiotic division. The segregation of the univalent homologues will result in no-X and X X eggs which give rise to the exceptional males and females. ANDERSON found only 7.3 percent crossing over in exceptional females, and most of these crossovers were a t the distal end of the chromosome. This distribution of crossovers would indicate that the few chiasmata formed are usually near the distal end of the chromosome, and that they either break or are prematurely terminalized, so that univalent XX's result at the first meiotic metaphase. If non-disjunction is due to premature terminalization, it is difficult to account for the crossovers near the forked locus, if chiasmata represent crossovers, because such a chiasma would have to be terminalized almost the entire genetic length, and for about half of the cytological length, of the X chromosome. If such premature terminalization could occur, one might expect frequent non-disjunction in normal stocks of Drosophila, where the formation of a single chiasma would be expected in the distal half of the X in about two-thirds of the meiotic divisions. But non-disjunction occurs only once in about 2000 times in normal stocks. I t would, of course, be impossible to explain crossing over in non-disjunction XX's on the partial chiasmatypy hypothesis if terminal affinity prevents the chiasmata from sliding off the ends of the chromosomes,

268

PROCEEDINGS O F T H E S I X T H

as suggested by DARLINGTON (1932). If only one or two chiasmata are formed, and they break, crossing over would be found occasionally in nondisjunction XX's, even in the region of the forked locus. The writer's theory of crossing over seems to provide a more plausible explanation of crossing over in non-disjunction chromosomes than does the partial chiasmatypy theory, although neither theory is entirely satisfactory (SAX1932a). According to DARLINGTON (193 1) , chromosome pairing is invariably dependent on chiasma formation. If every chiasma represents a crossover, no normal chromosome can have a crossover length of less than 50 units. Little or no crossing over has been found in the fourth chromosome of Drosophila melanogwter, although this chromosome is as regular in pair(1932) ing and disjunction as the X, which is 70 units long. DARLINGTON assumes that this fourth chromosome may have a crossover length of 50 units, but that the number of genetic factors is not sufficient to prove it. Regular chiasma formation might be expected in a short bivalent if the chromatids open out reductionally at one end of the chromosome and equationally at the other, as WENRICH(1916) assumes. I t has been clearly demonstrated that chromosome pairing is not always dependent on chiasma formation (SAX1932b), although in marly species with long chromosomes pairing at meiotic metaphase does seem to be dependent on chiasmata.

Randomness of crossing over The genetic evidence shows clearly that crossing over is not confined to two of the four chromatids (ANDERSON and RHOADES1931, REDFIELD 1930, and others). The first crossover from the spindle fiber in the X chromosome of Drosophila is at random between the homologous (non-sister) chromatids, as shown by the proportion ( 2 : l ) of equational and reciprocal crossovers in attached XX's (ANDERSON 1925) and by the percentage of homozygous recessives at the forked locus in the attached XX's (RHOADES 1931, STURTEVANT 1930). If crossing over i s at random at all crossovers, the homozygosis in attached XX's at the distal ends of the chromosomes should be about 20 percent (SAX 1932a). The percentage of homozygosis found by STURTEVANT was 17.1, and by RHOADES18.6. The deficiency of homozygosis may not be significant, however, owing to the lower viability of the homozygous recessive segregates. I f crossing over is at random for all crossov.ers at all loci, the first crossover in attached XX chromosomes should be equational or reciprocal in the proportions of 2 : l . If the first crossovers are not at random, there

INTERNATIONAL CONGRESS O F GENETICS

269

should be an excess of equationals on the writer's theory, and an excess of reciprocals on BELLING'Stheory. Three types of second crossovers should be found in the ratio of 2:8: 1, as follows: (1) equationals homozygous at the distal end; ( 2 ) equationals homozygous at the proximal end; and ( 3 ) reciprocals. If the second crossover is not a t random, the second class of crossovers will be decreased and the third class increased. Random crossing over can occur, on the writer's theory, only if some half twists occur in paired sister chromatids (SAX 1932a). The genetic analysis of attached XX's in Drosophila indicates that the first crossover occurs at random. The data on second crossovers and the percentage of homozygosis at different loci are not adequate for a critical test of random assortment of chromatids at the second crossover.

Absence of crossing over in the Drosophila male No crossing over occurs in the Drosophila male, but pairing and disT junction of all chromosomes are regular (METZ 1926, G U Y ~ N Oand (193 1 ) assumes that chiasma formation and NAVILLE1928). DARLINGTON crossing over are essential for chromosome pairing. H e assumes that in the male Drosophila there are always two chiasmata in each bivalent, one on each side of the spindle fiber; that the two chiasmata are very close together; that no mutations occur in the region between chiasmata; and that both crossovers are reciprocals and involve the same two chromatids. These assumptions are not only highly improbable, but also they can not be reconciled with either BELLING'Sor DARLINGTON'S explanation of the mechanism for crossing over. The assumption that all crossovers in the male are reciprocal is difficult to reconcile with the fact that both equational and reciprocal crossovers occur in the female. The writer assumes that some chiasmata are formed in the bivalents of the Drosophila male, but that these chiasmata are terminalized without breaking, as seems to be the case in several genera. The apparent differences in the duration of the metaphase stages at meiosis in males and females seem to support this assumption. CHIASMA FREQUENCY AND CROSSING OVER

If each chiasma represents a crossover, it should be possible to estimate the crossover length of the chromosomes in certain species, especially where (1931a) little or no terminalization of chiasmata is found. DARLINGTON attempted to correlate chiasma frequency with crossover length of one of the chromosomes in Primula, but with rather unsatisfactory results (SAX

270

PROCEEDINGS O F THE SIXTH

1932a). Zea should be satisfactory for this purpose, but little is known concerning chiasma frequency at prophase stages. There is sufficient cytological and genetic evidence to warrant a comparison of chiasma frequency and crossover lengths of the chromosomes of Vicia faba. In this species MEDA (1930b) found, at metaphase and diakinesis, an average of 8.1 chiasmata in the long bivalent and an average of 3.5 chiasmata for each of the five short bivalents. SIRKS has studied 26 genetic factors in Yicia faba. Of these 26 factors, 19 were found in four linkage groups, and only 7 were independent. Of the 19 factors, 7 were found in the first linkage group, 4 in the second, 5 in the third, and 3 in the fourth. In no case did any linkage group exceed approximately 50 crossover units. I i each chiasma represents a crossover, there should be more independent chromosome units than there are factors. The long chromosome should have a crossover length of 400 units, and the average length of the short chromosomes should be 175 units. With such a high frequency of crossing over, it would be highly improbable that 19 of the 26 factors would be found in four linkage groups, none of which exceeds 50 crossover units.

According to JANSSEXS' partial chiasnlatypy hypothesis, each chiasma represents a crossover which occurred at pachytene. BELLING has offered the only plausible explanation of how crossing over might occur on this theory, but his explanation is not in accord with certain critical cytological and genetic observations. If each chiasma represents a crossover, it is difficult to account for the prevalence of symmetrical chiasmata in many species and the reduction in chiasrna frequency often found between early diplotene and metaphase stages of meiosis. If, as DARLINGTON believes, chromosome pairing at meiotic metaphase is dependent on chiasma formation, it is difficult to account for the absence of crossing over in the Drosophila male, and difficult to explain the crossover length of the fourth chromosome of D. melanogaster. If each chiasma represents a crossover, it is difficult to reconcile the cytological observations with the genetic results in Vicia faba. The writer's theory of crossing over is based on the assumption that chiasmata are caused by an alternate opening out of pairs of sister and nonsister chromatids. Crossing over occurs only when two of the four chromatids break at a chiasma. This theory is in accord with the following genetic evidence: only two chromatids cross over at any one locus; one

1 l i T E R N A T I O N A L CONGRESS OF G E N E T I C S

271

crossover interferes with a second one; crossing over is reduced in the spindle-fiber region; no crossing over occurs in the male Drosophila; and crossing over is more or less at random between any two homologous chromatids. This theory is supported by the cytological evidence that there is often considerable reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis, and by the prevalence of symmetrjcal chiasmata in many species. LITERATURE CITED A N ~ E R S O E. N , G., 1925 Crossing over in a case of attached X chromosomes in Drosophila melanogaster. Genetics 10:403-417. 1929 Studies on a case of high non-disjunction in Drosopltila melanogaster. Z. indukt. Abstamm.-u. Vererb1.chre. 51:341-397. ANDERSOS,E . G., and RHOADES,M. M., 1931 T h e distribution of interference in the X chromosome of Drosophila. Papers Michigan Acad. Sci. 13:227-239. BABCOCK, E . B., and CLAUSEN, J., 1929 Meiosis in two species and three hybrids of Crepis and its bearing on taxonomic relationship. Univ. California Pub. Agric. Sci. 2 :401-432. BEADLE,G. W., 1932 The relation of crossing over to chromosome association in ZeaEuchlaena hybrids. Genetics 17:481-501. BEAL,J. M., 1932 hficrosp~rogenesisand chromosome behavior in Nothoscordum bivalve. Rot. Gaz. 93:278-295. BBLA~,K., 1928 I)ie cytologischen Grundlagen der Vererbung. p. 412. Berlin: Borntraeger. RELLIKG,J., 1926 Single and double rings a t the reduction division in Uvularia. Biol. Bull. 50 :355-363. 1927 T h e attachments of chromosomes at the rcduc:ion division in flowering plants. J. Genet. 18:177-205. 1931a Chromomeres o i liliaceous plants. Univ. California Pub. Bot. 16:153-170. 1931b Chiasmas in flowering plants. Univ. California Pub. Bot. 16:311-338. BERGHS,J., 1904 Formation des chromosomes hkt6rotypiques dans la sporog6nbse vPg6tale. La Cellule 21:173-188. IJRIDCES, C. B., 1916 Non-disjunction as proof of the chromosome theory of heredity. Genetics : 1:l-52, 107-163. BRIXES,C. B., and ANDERSOX, E. G., I925 Crossing over in the X-chromosomes of triploid females of Drosophila melanogaster. Genetics 10:418-441. CAROTHERS, E . E., 1926 The maturation divisions in relation to t h e segregation of homologous chromosomes. Quart. Rev. I3iol. 1:119-435. CATCHESIDE, 1). G., 1931 Critical evidence of parasynapsis in Oenothera. Proc. Roy. Sot. London 109:165-184. COOPER,D. C., and BRINK,R. A,, 1931 Cytological evidence for segmental interchange between non-homologous chromosomes in maize. Proc. Nat. Acad. Sci. Washington 17 ~334-338. CREIGHTON, H. R., and MCCLINTOCK, B., 1931 A correlation of cytological and genetical crossing-over in Zea mags. Proc. Nat. Acad. Sci. Washington 17:492-497. DARLIXGTOK, C. D., 1931a Meiosis in diploid and tetraploid Primula sinensis. J. Genet24 :65-%.

1931b Meiosis. Biol. Rev. 6:221-264. 1932 Recent advances in cytology. p. 559. Philadelphia: P. Ulakiston's Son and Co.

272

PROCEEDINGS O F T H E S I X T H

DARLINGTON, C. D., and DARK,0 . S., 1932 The origin and behavior of chiasmata. 11. Stenobothrus parallelus. Cytologia 3:169-185. ERLANSON, E. W., 1931 Chromosome organization in Rosa. Cytologia 2:256-282. FERGUSON, M. C., 1904 Contributions to the knowledge of the life history of Pinus, with special reference to sporogenesis, the development of the gametophytes, and fertilization. Proc. Nat. Acad. Sci. Washington 6:l-202. GAIRDNER, A. E., and DARLINGTON, C. D., 1932 Ring-formation in diploid and polyploid Campanula persicifolia. Genetica 13:113-150. GOWEN,J. W., 1928 Mutation, chromosome non-disjunction and the gene. Science 68:211212. G R ~ I R EV., , and DETON,W., 1906 Contribution i I'etude de la spermatogtnese dans I'Ophyotrocha puerilis. La Cellule 23:435-440. GUY~NOT, E., and NAVILLE,A,, 1928 Les chromosomes et la reduction chromatique chez Drosophila melanogarter. La Cellule 39:27-81. JANSSENS,F. A,, 1924 La chiasmatypie dans les insectes. La Cellule 34:135-359. KAUFMANN,B. P., 1931 Chromonemata in somatic and meiotic mitoses. Amer. Nat. 63 :280-283. MAEDA,T., 1930a The meiotic divisions in pollen mother cells of the sweet pea (Lathyrus odoratzu), with special reference to the cytological basis of crossing over. Mem. Coll. Sci. Kyoto Imp. Univ. 5239-123. 1930b On the configurations of gemini in the pollen mother cells of Vicio faba L. Mem. Coll. Sci. Kyoto Imp. Univ. 5:125-137. MCCLUNG,C. E., 1927 The chiasmatype theory of JANSSENS.Quart. Rev. Biol. 2:344-366. 1928 Differential chromosomes of Mecostethus gracilis. Zeit. Zell. v. m. Anat. 7 :756-778. METZ, C. W., 1926 Observations on spermatogenesis in Drosophila. Zeit. Zell. Mik. Anat. 4 :1-28. MORGAN, L. V., 1925 Polyploidy in Drosophila melanogaster with two attached X chromosomes. Genetics 10:148-178. MORGAN, T. H., 1925 The bearing of genetics on cytological evidence for crossing over. La Cellule 36:113-123. MULLER, H . J., 1916 The mechanism of crossing over. Amer. Nat. 50:193-221. NAKAMURA, K., 1932 Studies on reptilian chromosomes. 111. Chromosomes of some Geckos. Cytologia 3:156-168. NEWTON,W . C. F., 1926 Chromosome studies in Tulipa and some related genera. J. Linn. soc. 47 :339-354. PHILP, J., and HUSKINS,C. L., 1931 The cytology of Matthiola incana R. Br, especially in relation to the inheritance of double flowers. J. Genet. 24:359-404. RANDOLPH, L. F., 1932 The chromosomes of haploid maize, with special reference to the double nature of the univalent chromosomes in the early meiotic prophase. Science 75 :566-567. REDFIELD,HELEN,1930 Crossing over in the third chromosome of triploids of Drosophila melanogaster. Genetics 15:205-252. RHOADES,MARCUSM., 1931 The frequencies of homozygosis of factors in attached X females of Drosoplzila melanogarter. Genetics 16:375-385. 1932 The genetic demonstration of double strand crossing-over in Zea mays. Proc. Nat. Acad. Sci. Washington 18:481-484. ROBERTSON, W. R. B., 1916 Chromosome studies. I. J. Morph. 27:179-332. 1931a A split in chromosomes about to enter the spermatid (Paratettix texanus). Genetics 16 :349-352.

INTERNATIONAL CONGRESS O F GENETICS

273

1931b Chromosome studies. 11. Synapsis in the Tettigida with special reference to the presynapsis split. J. Morph. and Physiol. 51:119-146. SANSOME, E. R., 1932 Segmental interchange in Pisum sativum. Cytologia 3:200-219. SAX,H. J., 1933 Chiasma formation in Larix and Tsuga. Genetics 18:121-128. SAX, KARL,1930 Chromosome structure and the mechanism of crossing over. J. Arnold Arb. 11 :193-220. 1932a The cytological mechanism of crossing over. J. Arnold Arb. 13:180-212. 1932b Meiosis and chiasma formation in Paonia suffruticosa. J. Arnold Arb. 13:375-384. SAX,KARL,and ANDERSON, EDGAR,1933 Segmental interchange in Tradescantia. Genetics 18: 53-67. SIRKS, M. J., 1932 Beitrage zu einer genotypischen analyse der Ackerbohne, Vicia faba L. Genetica 13:209-631. STERN,C., 1931 Zytologisch-genetische Untersuchungen als Beweise fiir die Morgansche Theorie des Faktorenaustausches. Biol. Zbl. 51:547-587, STURTEVANT, A. H., 1913 The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14:43-59. 1925 The effects of unequal crossing over at the bar locus in Drosophila. Genetics 10:117-147. 1926 A crossover reducer in Drosophila melanogaster due to inversion of a section of the third chromosome. Biol. Zbl. 46597-702. 1928 A further study of the so-called mutation at the bar locus of Drosophila. Genetics 13:401-409. 1930 Two new attached-X lines of Drosophila melanogaster, and further data on the behavior of heterozygous attached-X's. Pub. Carnegie Instn. 42161-81. SUTTON,W. S., 1902 On the morphology of the chromosome group in Brmhystola magna. Biol. Bull. 4:24-39. TAYLOR, W. R., 1924 The smear technique for plant cytology. Bot. Gaz. 78:236-238. 1930 Chromosome structure in mitosis and meiosis. Contr. Bot. Lab. Univ. Pennsylvania 7 96.5-270. WEINSTEIN,A,, 1918 Coincidence of crossing over in Drosophila melanogaster. Genetics 3:135-173. WENRICH,D. H., 1916 The spermatogenesis of Phrynotettix magnus with special reference to synapsis and the individuality of chromosomes. Bull. Mus. Comp. Zool. 60:57-133. WIISON, E. B., 1925 The cell in development and heredity. p. 1232. New York: The Macmillan Co.

256

PROCEEDINGS OF THE SIXTH

T H E CYTOLOGICAL MECHANISM FOR CROSSING OVER Karl Sax, Arnold Arboretum, Haward University, Boston, Massachusetts

The genetic analysis of Drosophila by MORGAN and his associates has contributed more to our knowledge of chromosome behavior at meiosis than have the investigations of the cytologists. Any theory of chromosome pairing and crossing over must conform to the rigid requirements of the geneticist. As a foundation for the discussion of the mechanism of crossing over, it is essential to consider the facts which have resulted from the genetic investigations of Drosophila during the past twenty years. I t is known that the four linkage groups in Drosophila melamgmter correspond to the four pairs of chromosomes. The genes are arranged in a linear order in these chromosomes. Chromosome pairing must involve a gene by gene association in order to account for the great precision of crossing over. The frequency of crossing over is not the same for all regions of the chromosome and is reduced in the region of the spindle-fiber attachment. Chromosome pairing and disjunction are regular in both males and females, but no crossing over occurs in the male. The phenomenon of interference is of special importance in any interpretation of the mechanism of crossing over. If one crossover occurs, a second one is never found in adjacent regions of the same chromosome (STURTEVANT 1913). Interference is complete for 10 to 20 units, depending on the region of the chromosome involved (MULLER1916, WEINSTEIK 1918). The first meiotic division in Drosophila is usually, if not always, reductional at the spindle fiber attachment point, that is, the sister chromatids are held together at the fiber constriction at this division (BRIDGES and ANDERSON 1925, ANDERSON 1925, 19-9, L. V. MORGAN 1925, REDFIELD 1930, RHOADES 1931, STURTEVANT 1930). Crossing over occurs at early prophase of meiosis, after the chromosomes have paired, and between only two of the four chromatids at any one locus 1925). It has also been shown that (BRIDGES 1916, BRIDGES and ANDERSON only two of the four chromatids are involved in a crossover in Zea (RHOADES 1932) and in Habrobracon (WHITINGunpublished). Any factor which reduces crossing over is.associated with a reduction in 1929). regularity of pairing and disjunction (GOWEN1928, ANDERSON Crossing over is decreased in individuals heterozygous for an inversion (STURTEVANT 1926). Crossing over is not limited to two of the four chromatids throughout their length, because crossing over is found in more than 50 percent of the

INTERNATIONAL CONGRESS O F GENETICS

257

emerging X chromosomes of Drosophila, and about 75 percent of the third 1931, chromosomes have one or more crossovers (ANDERSON and RHOADES REDFIELD1930). Crossing over occurs more or less at random between homologous chromatids (ANDERSON 1925). There is no evidence that crossing over occurs between sister chromatids (STURTEVANT 1925, 1928). I t has generally been assumed that genetic crossing over is correlated with an actual physical interchange of chromosome segments. This assumpand tion has been confirmed by the brilliant investigations of CREIGHTON MCCLINTOCK(1931) with Zea, and of STERN(1931) with Drosophila. THE MECHANISM FOR CROSSING OVER

Janssens' interpretation of chiasma formation JANSSENS'(1924) partial chiasmatypy hypothesis has been supported ( 1930,1931b), MAEDA(1930a), by BELLING( 1931a, 1931b), DARLINGTON ana others. It is assumed that every chiasma represents a crossover which has occurred between two of the four chromatids at pachytene. As the homologous chromosomes open out at diplotene, only sister chromatids are paired, and the chiasmata indicate the point of interchange in crossing over. The only plausible explanation of how crossing over could occur, on JANSSENS'theory, has been presented by BELLING(1931b). According to BELLING,the homologous chromosomes pair as single threads. Half twists occur in the paired chromosomes before the new chromatids are formed. Each chromomere, or gene, then divides. The connecting fiber between chromomeres may remain with the old chromomere or pass to the new one at random. New connecting fibers are then formed to unite the free chromomeres. At the half twist formed by the original chromatids, the connecting fibers unite the free genes by the shortest path, so that a crossover occurs at random between non-sister chromatids. I t is, of course, also necessary to assume that sister-strand crossovers are very frequent. As the writer (SAX 1932a) has pointed out, such a random assortment of connecting fibers would result in numerous twists in paired chromatids in both somatic and meiotic chromosomes. BELLING'Stheory can be made more plausible if it is assumed that crossovers always involve the two new chromatids, as BELLING(1931a) suggested in his second paper on crossing over. An occasional crossover between sister chromatids would result in apparent random crossing over between homologous chromatids (SAX 1932a). An interchange of segments between two of the four chromatids should invariably produce an asymmetrical arrangement of the chromatids. This

258

PROCEEDINGS OF T H E SIXTH

situation was recognized by JANSSENS,but has received little or no attention from the recent supporters of his theory. The asymmetrical relations of the chromatids, if each chiasma represents a crossover, are shown in figure 1, diagram A. Chromatids which were adjacent at the four-strand stage form the cross at each chiasma. As a result of the sister-strand crossover at the second internode, the genetically detectable crossovers are not confined to two of the four chromatids. Sister chromatids are always paired at early diplotene.

As the chromosomes open out in subsequent stages of meiosis, the chiasmata are often terminalized until, at metaphase, only terminal or sub-terminal chiasmata are found (DARLINGTON 1931, 1932, GAIRDNER and DARLINGTON 1932, and others). I t is assumed by DARLINGTON that the attachment loop, in expanding, pushes all the others together at the end. As a result of such terminalization, a rather complicated association would result where two or more chiasmata are pushed together at metaphase (figure 1, diagram B). The two uppermost chromatids form the cross at the chiasma to the left of the spindle-fiber attachment. If this chiasma were opened out to form the typical cross-shaped figure, there would be a twist in one pair of chromatids, as shown in figure 1, diagram C. If only one chiasma is formed, it is possible that the torsion would rotate the paired chromatids at the

INTERNATIONAL CONGRESS O F GENETICS

259

proximal end so that a symmetrical cross would result; but where a chiasma is formed on each side of the spindle fiber, few symmetrical chiasmata would be expected. Two consecutive reciprodal chiasmata on the same side of the fiber should produce interlocking of chromatids a t metaphase in half the cases if one chromatid passes above or below the other a t random. The chiasmata represented in figure 1, diagrams A to C, are all reciprocal crossovers, which is the type expected most frequently on BELLING'Shy(1932) in Stenopothesis and observed most frequently by DARLINGTON bothrus. These diagrams are based on BELLING'Slatest theory, with a few favorable modifications. If, however, the interchange of segments occurs in the chromatids which form the cross, the results will be the same, so far as the relation of the chromatids is concerned. Either a reciprocal (adjacent) or an equational (diagonal) crossover will result in an asymmetrical relation of the chromatids (figure 1, diagrams D and E ) . Viewed from the end, each chromatid will not lie in the same quadrant a t all loci. In the diagram of the equational crossover, as viewed from the left end at the four-strand stage (figure 1, diagram I?), the chromatid a t the upper right quadrant changes to the lower left quadrant, owing to the crossover, while in the reciprocal crossover the chromatid in the upper right quadrant changes to the upper left quadrant. BELLING'Stheory of crossing over seems to be untenable for several reasons. H e assumes, as does DARLINGTON, that homologous chromosomes pair as single threads a t meiosis. According to KAUFMANN(1931), the chromosomes a t the telophase of the last premeiotic division are two-parted in several plant species. In several Orthopteran species the chromosomes are split longitudinally before pair ing a t meiosis (ROBERTSON 1916, 193la, 1931b, MCCLUNG1928). RANDOLPH(1932) finds a longitudinal split in the early leptotene threads in a haploid Zea plant. These observations can not be reconciled either with BELLING'Stheory of crossing over or with DARLINGTON'S theory of meiosis. The random assortment of connecting fibers between chromomeres, postulated by BELLING,would mean that sister strand crossovers and many twists in paired chromatids would be expected. There is no evidence that sister strand crossing over occurs, and few twists are found in paired chromatids of either somatic or meiotic chromosomes. The third objection to BELLING'Stheory-and this applies to any interpretation of JANSSENS'hypothesis-is the prevalence of symmetrical relations of the chromatids in many genera, even where several chiasmata are found in a bivalent chromosome. It is also difficult t o account

260

PROCEEDINGS O F THE SIXTH

for interference in crossing over and the absence of crossing over in the Drosophila male, on any modification of JANSSENS'theory. DARLINGTON (1932) has recently revived the torsion hypothesis to explain how crossing over occurs. It is significant that he shows no diagrams illustrating how this mechanism might cause exact crossing over between only two of the non-sister chromatids at any one locus. McClungJs theory of chiasima formation According to MCCLUNG,chiasmata are caused by the alternate opening out of pairs of sister and non-sister chromatids. This interpretation has

(1916), WILSON(1925), been supported by WENRICH( 1916), ROBERTSON BELAR(1928), CAROTHERS (1926), and NEWTON(1926). As MCCLUNG (1927) has pointed out, such an origin of the chiasmata should usually result in a symmetrical relation of the chromatids. The apparent cross a t a chiasma is formed by chromatids which were diagonal at the fourstrand stage, and each chromatid may lie in the same quadrant at all loci (figure 2, diagram A ) . The relations of the chromatids at metaphase, when no crossover has occurred, are shown in figure 2, diagram E.

INTERNATIONAL CONGRESS O F GENETICS

261

The writer's theory of crossing over The writer (SAX1930, 1932a) has suggested that crossing over is caused by breaks in two of the four chromatids at a chiasma. A t very early diplotene many chiasmata may be formed in each bivalent chromosome, as shown in figure 2, diagram A. These chiasmata are formed by alternate pairing of sister and non-sister chromatids. Occasional half-twists in paired sister chromatids permit more or less random crossing over between any two homologous (non-sister) chromatids. Between the earliest diplotene stage and diakinesis there is considerable reduction in the number of chiasmata. Unequal opening out of certain internodes will result in the elimination of certain chiasmata by cancellation, as shown in figure 2, diagram B. I t is also possible that some chiasmata may be terminalized to such an extent that they will pass off the end of the chromosome. Crossovers occur only when two of the chromatids break at a chiasma, and the segments reunite in a new association, as shown in figure 2, diagrams A and B. When no crossover occurs, the chromatids a t a chiasma are symmetrical, each chromatid lies in the same quadrant on both sides of the chiasma, and the apparent cross a t the chiasma is formed by chromatids which were diagonal a t the four-strand stage. If the distal ends of the chromosomes open out in the same plane, a symmetrical cross is formed, as shown in figure 2, diagram E. A crossover, or a twist in sister threads, will produce an asymmetrical arrangement of the chromatids, as shown in figure 2, diagrams B and C. A crossover formed between two intact chiasmata will result in interlocking of chromatids a t metaphase (figure 2, diagram D). This theory of crossing over seems to meet the rigid genetic requirements, and is in accord with most of the cytological evidence. The lengths of the internodes, or loops, between chiasmata would account for interference. The reductional loop at the fiber attachment point is usually larger than the others, which would reduce crossing over in the spindle-fiber region. Chiasma formation and normal chromosome pairing can occur without crossing over, as is presumably the case in the Drosophila male. Crossovers between sister chromatids should be rare. Reciprocal and equational crossovers may occur with equal'frequency (resulting in a genetic ratio of 2 : l in attached XX's), but under certain conditions there should be an excess of equational crossovers. A symmetrical relation of the chromatids would be expected where no crossover has occurred, but half twists in paired sister chromatids or a crossover would produce asymmetrical chiasmata.

262

PROCEEDINGS OF THE SIXTH CYTOLOGICAL EVIDENCE

The arrangement of chromatids If MCCLUNG'Stheory of chiasma formation is correct, the chromatids should usually show a symmetrical arrangement at each chiasma if little or no crossing over has occurred. Chromosome configurations which seem to support this interpretation of chiasma formation have been shown in Orthopteran species by SUTTON( 1902), WENRICH( 1916), ROBERTSON ( 1916), JANSSENS ( 1924), BELAR( 1928), CAROTHERS (1926), MCCLUNG (1927), DARLINGTON and DARK(1932), and others. Similar figures have been shown for other groups of animals (WILSON1925). In plant species the individual chromatids are not clearly differentiated until metaphase, and then only when the smear technique is used. This technique has long been used by the zoologists, but has been adopted by pointed out its application to plant cells in botanists only since TAYLOR 1930), 1924. Symmetrical chiasmata have been found in Gasteria (TAYLOR Uvularia (BELLING1926), Tulipa (NEWTON1926), Paeonia (SAX1932b), Secale (SAX1930), and in Larix (H. J. SAX1932). According to NEWTON, the hypothesis "which explains the diakinetic figures as due to the opening out in two planes at right angles of what are originally four parallel chromatids, is adequate to explain the events of diakinesis and division in Tulipa and Fritillaria." Asymmetrical chiasmata have been described in Orthopteran species by JANSSENSand DARLINGTON, but these figures are apparently not typical for this group. As MCCLUNGhas pointed out, the clearest figures shown by JANSSENS show the chromatids in the same quadrants at all loci. DARLINGTON and DARK (1932) believe that most of the chiasmata in Stenobothrus are reciprocal and involve the same two chromatids at successive chiasmata. Of the two figures referred to, one (figure 7, chiasmata A and B) shows clearly that both chiasmata are symmetrical and diagonal, as is the case in most of the figures shown, where the chromatids can be observed. The interlocking of chromatids at metaphase does indicate that some crossing over has occurred. MZDA (1930a) finds a large proportion of asymmetrical figures in Lathyrus chromosomes, as might be expected, since crossing over is known to occur in this genus. Crossing over is also indicated by interlocked chromatids at metaphase. Asymmetrical chiasmata are frequently found in Paeonia chromosomes (SAX1932b). In most cases these figures could be attributed to half twists in sister chromatids, but some figures do seem to support JANSSENS' hypothe-

INTERNATIONAL CONGRESS O F GENETICS

263

sis. These exceptional figures may be due to an association of non-sister chromatids at the fiber attachment point, accompanied by a half twist in one pair of sister chromatids. The prevalence of symmetrical chiasmata in the Orthoptera and in certain plant species seems to indicate that JANSSENS'partial chiasmatypy theory of chiasma formation is untenable. The asymmetrical chiasmata found in both plant and animal species would be expected on MCCLUNG'S theory of chiasma formation if twists occur in sister chromatids, or if crossovers occur in some of the chiasmata which disappear between early diplotene and metaphase stages of meiosis. If no crossovers occur, almost all the chiasmata may be symmetrical, as seems to be the case in some species. Reduction in chias~+za frequency I f crossing over is caused by breaks in the chiasmata, as the writer assumes, then there should be a reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis. Reduction in chiasma frequency can also be attributed to the meeting and cancellation of chiasmata, as shown in figure 2, diagrams A and B. There is also the possibility that some chiasmata may pass off the ends of the chromosomes. At earliest diplotene there are numerous nodes and internodes along the bivalent chromosomes in many species of plants and animals. The frequency of these nodes, most of which are probably chiasmata, may be as high as 7 or 8 in certain chromosomes. Bivalent chromosomes with numerous nodes and internodes at early diplotene have been shown in Stenobothrus by JANSSENS(1924, figure 242), in reptilian chromosomes (1932), and in Ophyotrocha by GRBGOIREand DETON by NAKAMURA (1906). In plant species, chromosomes with an apparent high chiasma fre1904), Lilium and' Allium quency have been shown in Pinus (FERGUSON (BERGHS1904), Crepis (BABCOCK and CLAUSEN1929), Tulipa (NEWTON 1926), Nothoscordum (BEAL1932), and in Callisia ( S A X1932a). I n these genera the chiasma frequency may be reduced from five or six or more a t early diplotene to one or two at metaphase. A reduction in chiasma frequency between diplotene or early diakinesis and metaphase has been found 1932), Lilium (BELLING1931b), in Tulipa (NEWTON1926, DARLINGTON Primula ( DARLINGTON 1931a), Rosa (ERLANSON 1931) , Matthiola (PHILP and DARLINGTON 1932), and and HUSKINS1931), Campanula (GAIRDNER Callisia ( S A X1930). I n these genera the average number of chiasmata lost per bivalent between diplotene and metaphase is somewhat more than one. The loss of chiasmata reported in most cases would undoubtedly have been

264

PROCEEDINGS O F T H E SIXTH

greater if the earliest diplotene stages were favorable for examination of chiasma frequency. In some genera there is probably little reduction in chiasma frequency between early diplotene and metaphase. In these cases the chiasmata should be symmetrical, and little crossing over would be expected. The reduction in chiasma frequency is often associated with an increase in the proportion of terminal chiasmata, and, in many species, only terminal chiasmata are found at metaphase, even though there may be from three and DARK (1932), in disto five chiasmata at diplotene. DARLINGTON cussing chiasma terminalization, assume that "the attachment loop, in expanding, pushes all the others together at the end." But if each chiasma represents a crossover, the accumulation of two or more chiasmata would often result in a complex association of chromatids, as shown in figure 1, diagram B. In numerous species, terminal or sub-terminal chiasmata show no interlocking or asymmetrical relations of the chromatids, even where three or more chiasmata are found at diplotene. If two crossover chiasmata on the same side of the spindle fiber are terminalized without passing off the ends of the chromosomes, they will be reduced to one interlocked chiasma, or canceled, depending on the types of crossovers. If each chiasma is reciprocal and both crossovers occur in the same two chromatids, the two chiasmata will be canceled, o r reduced to one, with equal frequency. If crossing over is at random between homologous chromatids, the two chiasmata would be reduced to one interlocked chiasma in 50 percent of the cases, and reduced to none in 50 percent of the cases. In Primula and Campanula the chiasma frequency is reduced from three or more at diplotene to two terminal chiasmata at metaphase, but there is no evidence of interlocked chromatids in these genera. On the writer's hypothesis, the reduction in chiasma frequency between early diplotene and metaphase can be attributed to cancellation of chiasmata and breaks in some of the chiasmata as the early diplotene loops open out. Two chiasmata might be reduced to none by cancellation and reduced to one if the other chiasma were broken. Both the genetic and the cytological evidence support MORGAN'S (1925) suggestion that crossing over is an accidental by-product of meiosis. SANSOME (1932) has described a case of chiasma formation in a ring of six chromosomes in Pisum, which seems to support the partial chiasmatypy hypothesis. In 62 out of 78 figures one or two chiasmata were found in the "X segment" of the ring, resulting in a figure-of-eight configuration. Occasionally two chiasmata were found in this region, but the exact

INTERNATIONAL CONGRESS OF GENETICS

265

proportions of double and single chiasmata are not given, presumably owing to the difficulty in studying the chromatids in this region. On the writer's theory, two chiasmata must be formed in the X segment if only two chromosomes can pair at the same locus. I f one of the chiasmata breaks, the remaining one represents a crossover. The remaining chiasma would be under a special strain, but it seldom breaks. A much more critical case of the same type has been shown in Zea by BEADLE(1932). Chiasma frequency was obtained at both diplotene and metaphase, and the crossover length of the segment was known. The ring or chain of 4 chromosomes, resulting from segmental interchange, is so constituted that crossing over in one arm is limited to about 12 crossover units between the translocation point and the region where a teosinte chromosome segment has been introduced. When a chiasma is formed in this region, a ring with two free arms is found a t diakinesis and metaphase, and when no chiasma is present, the four chromosomes are arranged in a chain. At diakinesis 20 percent of the associations of four chromosomes were in the form of rings, but a t metaphase only 10 percent of these chromosomes were rings. On the writer's theory, a t least two chiasmata must be formed in this wz-translocation region. K O terminalization of chiasmata can occur because of the change in homology of the distal segments. As the ring opens out, all the chiasmata in this region may disappear by cancellation (compare figure 2, diagram B). If one chiasma breaks, the remaining chiasma represents a crossover. This chiasma can not be canceled or terminalized. The frequency of such a crossover chiasma in the wz-translocation segment a t diakinesis (20 percent) corresponds with the crossover frequency found in the derived Zea-teosinte chromosomes as bivalents. But at metaphase the chiasma frequency was only 1 0 percent. If the frequency of ring formation a t meta~haseis significant, this must mean that two chiasmata were often present at diakinesis and were canceled before metaphase, or that the crossover chiasma was broken. If the crossover chiasma breaks, a double crossover might result, but the two crossovers would probably be too close together to be detected. The effect of a break in the single crossover chiasma on the crossover length of the segment would depend on types of double crossovers obtained. If chiasma frequency is high a t early diplotene, the writer's explanation of chiasma formation in the wz-translocation segment is not improbable. Two chiasmata might be formed. Unless one chiasma breaks, the original chiasmata probably would be canceled, owing to the strain imposed by the

266

PROCEEDINGS OE THE SIXTH

opening of the ring and limitations in terminalization. If a chiasma breaks, cancellation of the remaining chiasma is impossible, and since the remaining crossover chiasma can not be terminalized, it is found in the ws-translocation region a t diakinesis and metaphase.

Behavior of unequal homologues In all recorded cases of pairing of unequal homologues, the long chromatids are paired with long and the short ones are paired with short a t meiosis. With only one chiasma and a terminal spindle-fiber attachment, it is difficult to explain the occurrence of both pre- and post-reduction in the same bivalent (SAX 1932a, b) as found by WENRICH(1916). This be(1932) on the assumption that the havior is explained by DARLINGTON fiber attachment is median, but the assumption is clearly erroneous, as shown by WENRICH'Sfigures of somatic chromosomes in Phrynotettix. On the MCCLUNGinterpretation of chiasma formation, the chromatids must separate equationally at the fiber attachment, where a single chiasma is formed between the fiber and the unequal segments if no crossing over has occurred. This assumption may be valid for some species, although it is apparently not the case in Drosophila or Zea.

Interlocked bivalents Interlocked non-homologous bivalents have been found in Oenothera, ahd Campanula, Tradescantia, and other genera. According to GAIRDNER DARLINGTON ( 1932), three types of interlocking would be expected : ( 1 ) proximal interlocking, where the loops containing the spindle fiber are involved; ( 2 ) distal interlocking, or locking of terminal loops; and ( 3 ) proximal-distal interlocking. If chiasmata are formed by alternate opening of sister and non-sister chromatids at diplotene, interlocking of bivalents could occur only in alternate internodes where sister chromatids are paired. If each chiasma represents a crossover and sister chromatids are always paired at early diplotene, interlocking can occur at any internode. In Campanula (GAIRDNER and DARLINGTON 1932) from two to six chiasmata are found in each bivalent, and the most frequent number seems to be three at diplotene. If each chiasma represents a crossover, distal or proximaldistal interlocking should be at least as frequent as proximal interlocking. theory of chiasma formation is correct, distal or proxiBut if MCCLUNG'S mal-distal interlocking should be very much less frequent than proximal interlocking. These conclusions are based on the assumption that chiasmata do not pass off the ends of the chromosomes but are pushed together a t

INTERNATIONAL CONGRESS O F GENETICS

267

the ends of the chromosomes, as postulated by DARLINGTON and DARK find, in Campanula, that "proximal (1932). GAIRDNER and DARLINGTON interlocking occurs in about 20 percent of nuclei in homozygous groups. Distal interlocking seems to be much rarer." Similar observations have been made in Oenothera by CATCHESIDE ( 1931 ) . Proximal interlocking is common in Tradescantia, but distal interlocking is rare (SAX and ANDERSON theory of chiasma forma1932). These results indicate that MCCLUNG'S tion is correct, or that chiasmata do pass off the ends of the chromosomes before metaphase. According to GAIRDKER and DARLINGTON, interlocking should occur between the chromatids of paired or unpaired chromosomes if the chromatids ever separate equationally at diplotene. This conclusion is obviously untenable ( S A Xand ANDERSON 1932). GENETIC EVIDENCE

Non-disjunction and crossing ovcr

In high non-disjunction lines of Drosophila, crossing over is greatly reduced in both the normal progeny and in the exceptional females (ANDERSON 1929). The genetic evidence indicates that non-disjunction is caused by a failure of chromoson~epairing at the first meiotic division. The segregation of the univalent homologues will result in no-X and X X eggs which give rise to the exceptional males and females. ANDERSON found only 7.3 percent crossing over in exceptional females, and most of these crossovers were a t the distal end of the chromosome. This distribution of crossovers would indicate that the few chiasmata formed are usually near the distal end of the chromosome, and that they either break or are prematurely terminalized, so that univalent XX's result at the first meiotic metaphase. If non-disjunction is due to premature terminalization, it is difficult to account for the crossovers near the forked locus, if chiasmata represent crossovers, because such a chiasma would have to be terminalized almost the entire genetic length, and for about half of the cytological length, of the X chromosome. If such premature terminalization could occur, one might expect frequent non-disjunction in normal stocks of Drosophila, where the formation of a single chiasma would be expected in the distal half of the X in about two-thirds of the meiotic divisions. But non-disjunction occurs only once in about 2000 times in normal stocks. I t would, of course, be impossible to explain crossing over in non-disjunction XX's on the partial chiasmatypy hypothesis if terminal affinity prevents the chiasmata from sliding off the ends of the chromosomes,

268

PROCEEDINGS O F T H E S I X T H

as suggested by DARLINGTON (1932). If only one or two chiasmata are formed, and they break, crossing over would be found occasionally in nondisjunction XX's, even in the region of the forked locus. The writer's theory of crossing over seems to provide a more plausible explanation of crossing over in non-disjunction chromosomes than does the partial chiasmatypy theory, although neither theory is entirely satisfactory (SAX1932a). According to DARLINGTON (193 1) , chromosome pairing is invariably dependent on chiasma formation. If every chiasma represents a crossover, no normal chromosome can have a crossover length of less than 50 units. Little or no crossing over has been found in the fourth chromosome of Drosophila melanogwter, although this chromosome is as regular in pair(1932) ing and disjunction as the X, which is 70 units long. DARLINGTON assumes that this fourth chromosome may have a crossover length of 50 units, but that the number of genetic factors is not sufficient to prove it. Regular chiasma formation might be expected in a short bivalent if the chromatids open out reductionally at one end of the chromosome and equationally at the other, as WENRICH(1916) assumes. I t has been clearly demonstrated that chromosome pairing is not always dependent on chiasma formation (SAX1932b), although in marly species with long chromosomes pairing at meiotic metaphase does seem to be dependent on chiasmata.

Randomness of crossing over The genetic evidence shows clearly that crossing over is not confined to two of the four chromatids (ANDERSON and RHOADES1931, REDFIELD 1930, and others). The first crossover from the spindle fiber in the X chromosome of Drosophila is at random between the homologous (non-sister) chromatids, as shown by the proportion ( 2 : l ) of equational and reciprocal crossovers in attached XX's (ANDERSON 1925) and by the percentage of homozygous recessives at the forked locus in the attached XX's (RHOADES 1931, STURTEVANT 1930). If crossing over i s at random at all crossovers, the homozygosis in attached XX's at the distal ends of the chromosomes should be about 20 percent (SAX 1932a). The percentage of homozygosis found by STURTEVANT was 17.1, and by RHOADES18.6. The deficiency of homozygosis may not be significant, however, owing to the lower viability of the homozygous recessive segregates. I f crossing over is at random for all crossov.ers at all loci, the first crossover in attached XX chromosomes should be equational or reciprocal in the proportions of 2 : l . If the first crossovers are not at random, there

INTERNATIONAL CONGRESS O F GENETICS

269

should be an excess of equationals on the writer's theory, and an excess of reciprocals on BELLING'Stheory. Three types of second crossovers should be found in the ratio of 2:8: 1, as follows: (1) equationals homozygous at the distal end; ( 2 ) equationals homozygous at the proximal end; and ( 3 ) reciprocals. If the second crossover is not a t random, the second class of crossovers will be decreased and the third class increased. Random crossing over can occur, on the writer's theory, only if some half twists occur in paired sister chromatids (SAX 1932a). The genetic analysis of attached XX's in Drosophila indicates that the first crossover occurs at random. The data on second crossovers and the percentage of homozygosis at different loci are not adequate for a critical test of random assortment of chromatids at the second crossover.

Absence of crossing over in the Drosophila male No crossing over occurs in the Drosophila male, but pairing and disT junction of all chromosomes are regular (METZ 1926, G U Y ~ N Oand (193 1 ) assumes that chiasma formation and NAVILLE1928). DARLINGTON crossing over are essential for chromosome pairing. H e assumes that in the male Drosophila there are always two chiasmata in each bivalent, one on each side of the spindle fiber; that the two chiasmata are very close together; that no mutations occur in the region between chiasmata; and that both crossovers are reciprocals and involve the same two chromatids. These assumptions are not only highly improbable, but also they can not be reconciled with either BELLING'Sor DARLINGTON'S explanation of the mechanism for crossing over. The assumption that all crossovers in the male are reciprocal is difficult to reconcile with the fact that both equational and reciprocal crossovers occur in the female. The writer assumes that some chiasmata are formed in the bivalents of the Drosophila male, but that these chiasmata are terminalized without breaking, as seems to be the case in several genera. The apparent differences in the duration of the metaphase stages at meiosis in males and females seem to support this assumption. CHIASMA FREQUENCY AND CROSSING OVER

If each chiasma represents a crossover, it should be possible to estimate the crossover length of the chromosomes in certain species, especially where (1931a) little or no terminalization of chiasmata is found. DARLINGTON attempted to correlate chiasma frequency with crossover length of one of the chromosomes in Primula, but with rather unsatisfactory results (SAX

270

PROCEEDINGS O F THE SIXTH

1932a). Zea should be satisfactory for this purpose, but little is known concerning chiasma frequency at prophase stages. There is sufficient cytological and genetic evidence to warrant a comparison of chiasma frequency and crossover lengths of the chromosomes of Vicia faba. In this species MEDA (1930b) found, at metaphase and diakinesis, an average of 8.1 chiasmata in the long bivalent and an average of 3.5 chiasmata for each of the five short bivalents. SIRKS has studied 26 genetic factors in Yicia faba. Of these 26 factors, 19 were found in four linkage groups, and only 7 were independent. Of the 19 factors, 7 were found in the first linkage group, 4 in the second, 5 in the third, and 3 in the fourth. In no case did any linkage group exceed approximately 50 crossover units. I i each chiasma represents a crossover, there should be more independent chromosome units than there are factors. The long chromosome should have a crossover length of 400 units, and the average length of the short chromosomes should be 175 units. With such a high frequency of crossing over, it would be highly improbable that 19 of the 26 factors would be found in four linkage groups, none of which exceeds 50 crossover units.

According to JANSSEXS' partial chiasnlatypy hypothesis, each chiasma represents a crossover which occurred at pachytene. BELLING has offered the only plausible explanation of how crossing over might occur on this theory, but his explanation is not in accord with certain critical cytological and genetic observations. If each chiasma represents a crossover, it is difficult to account for the prevalence of symmetrical chiasmata in many species and the reduction in chiasrna frequency often found between early diplotene and metaphase stages of meiosis. If, as DARLINGTON believes, chromosome pairing at meiotic metaphase is dependent on chiasma formation, it is difficult to account for the absence of crossing over in the Drosophila male, and difficult to explain the crossover length of the fourth chromosome of D. melanogaster. If each chiasma represents a crossover, it is difficult to reconcile the cytological observations with the genetic results in Vicia faba. The writer's theory of crossing over is based on the assumption that chiasmata are caused by an alternate opening out of pairs of sister and nonsister chromatids. Crossing over occurs only when two of the four chromatids break at a chiasma. This theory is in accord with the following genetic evidence: only two chromatids cross over at any one locus; one

1 l i T E R N A T I O N A L CONGRESS OF G E N E T I C S

271

crossover interferes with a second one; crossing over is reduced in the spindle-fiber region; no crossing over occurs in the male Drosophila; and crossing over is more or less at random between any two homologous chromatids. This theory is supported by the cytological evidence that there is often considerable reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis, and by the prevalence of symmetrjcal chiasmata in many species. LITERATURE CITED A N ~ E R S O E. N , G., 1925 Crossing over in a case of attached X chromosomes in Drosophila melanogaster. Genetics 10:403-417. 1929 Studies on a case of high non-disjunction in Drosopltila melanogaster. Z. indukt. Abstamm.-u. Vererb1.chre. 51:341-397. ANDERSOS,E . G., and RHOADES,M. M., 1931 T h e distribution of interference in the X chromosome of Drosophila. Papers Michigan Acad. Sci. 13:227-239. BABCOCK, E . B., and CLAUSEN, J., 1929 Meiosis in two species and three hybrids of Crepis and its bearing on taxonomic relationship. Univ. California Pub. Agric. Sci. 2 :401-432. BEADLE,G. W., 1932 The relation of crossing over to chromosome association in ZeaEuchlaena hybrids. Genetics 17:481-501. BEAL,J. M., 1932 hficrosp~rogenesisand chromosome behavior in Nothoscordum bivalve. Rot. Gaz. 93:278-295. BBLA~,K., 1928 I)ie cytologischen Grundlagen der Vererbung. p. 412. Berlin: Borntraeger. RELLIKG,J., 1926 Single and double rings a t the reduction division in Uvularia. Biol. Bull. 50 :355-363. 1927 T h e attachments of chromosomes at the rcduc:ion division in flowering plants. J. Genet. 18:177-205. 1931a Chromomeres o i liliaceous plants. Univ. California Pub. Bot. 16:153-170. 1931b Chiasmas in flowering plants. Univ. California Pub. Bot. 16:311-338. BERGHS,J., 1904 Formation des chromosomes hkt6rotypiques dans la sporog6nbse vPg6tale. La Cellule 21:173-188. IJRIDCES, C. B., 1916 Non-disjunction as proof of the chromosome theory of heredity. Genetics : 1:l-52, 107-163. BRIXES,C. B., and ANDERSOX, E. G., I925 Crossing over in the X-chromosomes of triploid females of Drosophila melanogaster. Genetics 10:418-441. CAROTHERS, E . E., 1926 The maturation divisions in relation to t h e segregation of homologous chromosomes. Quart. Rev. I3iol. 1:119-435. CATCHESIDE, 1). G., 1931 Critical evidence of parasynapsis in Oenothera. Proc. Roy. Sot. London 109:165-184. COOPER,D. C., and BRINK,R. A,, 1931 Cytological evidence for segmental interchange between non-homologous chromosomes in maize. Proc. Nat. Acad. Sci. Washington 17 ~334-338. CREIGHTON, H. R., and MCCLINTOCK, B., 1931 A correlation of cytological and genetical crossing-over in Zea mags. Proc. Nat. Acad. Sci. Washington 17:492-497. DARLIXGTOK, C. D., 1931a Meiosis in diploid and tetraploid Primula sinensis. J. Genet24 :65-%.

1931b Meiosis. Biol. Rev. 6:221-264. 1932 Recent advances in cytology. p. 559. Philadelphia: P. Ulakiston's Son and Co.

272

PROCEEDINGS O F T H E S I X T H

DARLINGTON, C. D., and DARK,0 . S., 1932 The origin and behavior of chiasmata. 11. Stenobothrus parallelus. Cytologia 3:169-185. ERLANSON, E. W., 1931 Chromosome organization in Rosa. Cytologia 2:256-282. FERGUSON, M. C., 1904 Contributions to the knowledge of the life history of Pinus, with special reference to sporogenesis, the development of the gametophytes, and fertilization. Proc. Nat. Acad. Sci. Washington 6:l-202. GAIRDNER, A. E., and DARLINGTON, C. D., 1932 Ring-formation in diploid and polyploid Campanula persicifolia. Genetica 13:113-150. GOWEN,J. W., 1928 Mutation, chromosome non-disjunction and the gene. Science 68:211212. G R ~ I R EV., , and DETON,W., 1906 Contribution i I'etude de la spermatogtnese dans I'Ophyotrocha puerilis. La Cellule 23:435-440. GUY~NOT, E., and NAVILLE,A,, 1928 Les chromosomes et la reduction chromatique chez Drosophila melanogarter. La Cellule 39:27-81. JANSSENS,F. A,, 1924 La chiasmatypie dans les insectes. La Cellule 34:135-359. KAUFMANN,B. P., 1931 Chromonemata in somatic and meiotic mitoses. Amer. Nat. 63 :280-283. MAEDA,T., 1930a The meiotic divisions in pollen mother cells of the sweet pea (Lathyrus odoratzu), with special reference to the cytological basis of crossing over. Mem. Coll. Sci. Kyoto Imp. Univ. 5239-123. 1930b On the configurations of gemini in the pollen mother cells of Vicio faba L. Mem. Coll. Sci. Kyoto Imp. Univ. 5:125-137. MCCLUNG,C. E., 1927 The chiasmatype theory of JANSSENS.Quart. Rev. Biol. 2:344-366. 1928 Differential chromosomes of Mecostethus gracilis. Zeit. Zell. v. m. Anat. 7 :756-778. METZ, C. W., 1926 Observations on spermatogenesis in Drosophila. Zeit. Zell. Mik. Anat. 4 :1-28. MORGAN, L. V., 1925 Polyploidy in Drosophila melanogaster with two attached X chromosomes. Genetics 10:148-178. MORGAN, T. H., 1925 The bearing of genetics on cytological evidence for crossing over. La Cellule 36:113-123. MULLER, H . J., 1916 The mechanism of crossing over. Amer. Nat. 50:193-221. NAKAMURA, K., 1932 Studies on reptilian chromosomes. 111. Chromosomes of some Geckos. Cytologia 3:156-168. NEWTON,W . C. F., 1926 Chromosome studies in Tulipa and some related genera. J. Linn. soc. 47 :339-354. PHILP, J., and HUSKINS,C. L., 1931 The cytology of Matthiola incana R. Br, especially in relation to the inheritance of double flowers. J. Genet. 24:359-404. RANDOLPH, L. F., 1932 The chromosomes of haploid maize, with special reference to the double nature of the univalent chromosomes in the early meiotic prophase. Science 75 :566-567. REDFIELD,HELEN,1930 Crossing over in the third chromosome of triploids of Drosophila melanogaster. Genetics 15:205-252. RHOADES,MARCUSM., 1931 The frequencies of homozygosis of factors in attached X females of Drosoplzila melanogarter. Genetics 16:375-385. 1932 The genetic demonstration of double strand crossing-over in Zea mays. Proc. Nat. Acad. Sci. Washington 18:481-484. ROBERTSON, W. R. B., 1916 Chromosome studies. I. J. Morph. 27:179-332. 1931a A split in chromosomes about to enter the spermatid (Paratettix texanus). Genetics 16 :349-352.

INTERNATIONAL CONGRESS O F GENETICS

273

1931b Chromosome studies. 11. Synapsis in the Tettigida with special reference to the presynapsis split. J. Morph. and Physiol. 51:119-146. SANSOME, E. R., 1932 Segmental interchange in Pisum sativum. Cytologia 3:200-219. SAX,H. J., 1933 Chiasma formation in Larix and Tsuga. Genetics 18:121-128. SAX, KARL,1930 Chromosome structure and the mechanism of crossing over. J. Arnold Arb. 11 :193-220. 1932a The cytological mechanism of crossing over. J. Arnold Arb. 13:180-212. 1932b Meiosis and chiasma formation in Paonia suffruticosa. J. Arnold Arb. 13:375-384. SAX,KARL,and ANDERSON, EDGAR,1933 Segmental interchange in Tradescantia. Genetics 18: 53-67. SIRKS, M. J., 1932 Beitrage zu einer genotypischen analyse der Ackerbohne, Vicia faba L. Genetica 13:209-631. STERN,C., 1931 Zytologisch-genetische Untersuchungen als Beweise fiir die Morgansche Theorie des Faktorenaustausches. Biol. Zbl. 51:547-587, STURTEVANT, A. H., 1913 The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14:43-59. 1925 The effects of unequal crossing over at the bar locus in Drosophila. Genetics 10:117-147. 1926 A crossover reducer in Drosophila melanogaster due to inversion of a section of the third chromosome. Biol. Zbl. 46597-702. 1928 A further study of the so-called mutation at the bar locus of Drosophila. Genetics 13:401-409. 1930 Two new attached-X lines of Drosophila melanogaster, and further data on the behavior of heterozygous attached-X's. Pub. Carnegie Instn. 42161-81. SUTTON,W. S., 1902 On the morphology of the chromosome group in Brmhystola magna. Biol. Bull. 4:24-39. TAYLOR, W. R., 1924 The smear technique for plant cytology. Bot. Gaz. 78:236-238. 1930 Chromosome structure in mitosis and meiosis. Contr. Bot. Lab. Univ. Pennsylvania 7 96.5-270. WEINSTEIN,A,, 1918 Coincidence of crossing over in Drosophila melanogaster. Genetics 3:135-173. WENRICH,D. H., 1916 The spermatogenesis of Phrynotettix magnus with special reference to synapsis and the individuality of chromosomes. Bull. Mus. Comp. Zool. 60:57-133. WIISON, E. B., 1925 The cell in development and heredity. p. 1232. New York: The Macmillan Co.

256

PROCEEDINGS OF THE SIXTH

T H E CYTOLOGICAL MECHANISM FOR CROSSING OVER Karl Sax, Arnold Arboretum, Haward University, Boston, Massachusetts

The genetic analysis of Drosophila by MORGAN and his associates has contributed more to our knowledge of chromosome behavior at meiosis than have the investigations of the cytologists. Any theory of chromosome pairing and crossing over must conform to the rigid requirements of the geneticist. As a foundation for the discussion of the mechanism of crossing over, it is essential to consider the facts which have resulted from the genetic investigations of Drosophila during the past twenty years. I t is known that the four linkage groups in Drosophila melamgmter correspond to the four pairs of chromosomes. The genes are arranged in a linear order in these chromosomes. Chromosome pairing must involve a gene by gene association in order to account for the great precision of crossing over. The frequency of crossing over is not the same for all regions of the chromosome and is reduced in the region of the spindle-fiber attachment. Chromosome pairing and disjunction are regular in both males and females, but no crossing over occurs in the male. The phenomenon of interference is of special importance in any interpretation of the mechanism of crossing over. If one crossover occurs, a second one is never found in adjacent regions of the same chromosome (STURTEVANT 1913). Interference is complete for 10 to 20 units, depending on the region of the chromosome involved (MULLER1916, WEINSTEIK 1918). The first meiotic division in Drosophila is usually, if not always, reductional at the spindle fiber attachment point, that is, the sister chromatids are held together at the fiber constriction at this division (BRIDGES and ANDERSON 1925, ANDERSON 1925, 19-9, L. V. MORGAN 1925, REDFIELD 1930, RHOADES 1931, STURTEVANT 1930). Crossing over occurs at early prophase of meiosis, after the chromosomes have paired, and between only two of the four chromatids at any one locus 1925). It has also been shown that (BRIDGES 1916, BRIDGES and ANDERSON only two of the four chromatids are involved in a crossover in Zea (RHOADES 1932) and in Habrobracon (WHITINGunpublished). Any factor which reduces crossing over is.associated with a reduction in 1929). regularity of pairing and disjunction (GOWEN1928, ANDERSON Crossing over is decreased in individuals heterozygous for an inversion (STURTEVANT 1926). Crossing over is not limited to two of the four chromatids throughout their length, because crossing over is found in more than 50 percent of the

INTERNATIONAL CONGRESS O F GENETICS

257

emerging X chromosomes of Drosophila, and about 75 percent of the third 1931, chromosomes have one or more crossovers (ANDERSON and RHOADES REDFIELD1930). Crossing over occurs more or less at random between homologous chromatids (ANDERSON 1925). There is no evidence that crossing over occurs between sister chromatids (STURTEVANT 1925, 1928). I t has generally been assumed that genetic crossing over is correlated with an actual physical interchange of chromosome segments. This assumpand tion has been confirmed by the brilliant investigations of CREIGHTON MCCLINTOCK(1931) with Zea, and of STERN(1931) with Drosophila. THE MECHANISM FOR CROSSING OVER

Janssens' interpretation of chiasma formation JANSSENS'(1924) partial chiasmatypy hypothesis has been supported ( 1930,1931b), MAEDA(1930a), by BELLING( 1931a, 1931b), DARLINGTON ana others. It is assumed that every chiasma represents a crossover which has occurred between two of the four chromatids at pachytene. As the homologous chromosomes open out at diplotene, only sister chromatids are paired, and the chiasmata indicate the point of interchange in crossing over. The only plausible explanation of how crossing over could occur, on JANSSENS'theory, has been presented by BELLING(1931b). According to BELLING,the homologous chromosomes pair as single threads. Half twists occur in the paired chromosomes before the new chromatids are formed. Each chromomere, or gene, then divides. The connecting fiber between chromomeres may remain with the old chromomere or pass to the new one at random. New connecting fibers are then formed to unite the free chromomeres. At the half twist formed by the original chromatids, the connecting fibers unite the free genes by the shortest path, so that a crossover occurs at random between non-sister chromatids. I t is, of course, also necessary to assume that sister-strand crossovers are very frequent. As the writer (SAX 1932a) has pointed out, such a random assortment of connecting fibers would result in numerous twists in paired chromatids in both somatic and meiotic chromosomes. BELLING'Stheory can be made more plausible if it is assumed that crossovers always involve the two new chromatids, as BELLING(1931a) suggested in his second paper on crossing over. An occasional crossover between sister chromatids would result in apparent random crossing over between homologous chromatids (SAX 1932a). An interchange of segments between two of the four chromatids should invariably produce an asymmetrical arrangement of the chromatids. This

258

PROCEEDINGS OF T H E SIXTH

situation was recognized by JANSSENS,but has received little or no attention from the recent supporters of his theory. The asymmetrical relations of the chromatids, if each chiasma represents a crossover, are shown in figure 1, diagram A. Chromatids which were adjacent at the four-strand stage form the cross at each chiasma. As a result of the sister-strand crossover at the second internode, the genetically detectable crossovers are not confined to two of the four chromatids. Sister chromatids are always paired at early diplotene.

As the chromosomes open out in subsequent stages of meiosis, the chiasmata are often terminalized until, at metaphase, only terminal or sub-terminal chiasmata are found (DARLINGTON 1931, 1932, GAIRDNER and DARLINGTON 1932, and others). I t is assumed by DARLINGTON that the attachment loop, in expanding, pushes all the others together at the end. As a result of such terminalization, a rather complicated association would result where two or more chiasmata are pushed together at metaphase (figure 1, diagram B). The two uppermost chromatids form the cross at the chiasma to the left of the spindle-fiber attachment. If this chiasma were opened out to form the typical cross-shaped figure, there would be a twist in one pair of chromatids, as shown in figure 1, diagram C. If only one chiasma is formed, it is possible that the torsion would rotate the paired chromatids at the

INTERNATIONAL CONGRESS O F GENETICS

259

proximal end so that a symmetrical cross would result; but where a chiasma is formed on each side of the spindle fiber, few symmetrical chiasmata would be expected. Two consecutive reciprodal chiasmata on the same side of the fiber should produce interlocking of chromatids a t metaphase in half the cases if one chromatid passes above or below the other a t random. The chiasmata represented in figure 1, diagrams A to C, are all reciprocal crossovers, which is the type expected most frequently on BELLING'Shy(1932) in Stenopothesis and observed most frequently by DARLINGTON bothrus. These diagrams are based on BELLING'Slatest theory, with a few favorable modifications. If, however, the interchange of segments occurs in the chromatids which form the cross, the results will be the same, so far as the relation of the chromatids is concerned. Either a reciprocal (adjacent) or an equational (diagonal) crossover will result in an asymmetrical relation of the chromatids (figure 1, diagrams D and E ) . Viewed from the end, each chromatid will not lie in the same quadrant a t all loci. In the diagram of the equational crossover, as viewed from the left end at the four-strand stage (figure 1, diagram I?), the chromatid a t the upper right quadrant changes to the lower left quadrant, owing to the crossover, while in the reciprocal crossover the chromatid in the upper right quadrant changes to the upper left quadrant. BELLING'Stheory of crossing over seems to be untenable for several reasons. H e assumes, as does DARLINGTON, that homologous chromosomes pair as single threads a t meiosis. According to KAUFMANN(1931), the chromosomes a t the telophase of the last premeiotic division are two-parted in several plant species. In several Orthopteran species the chromosomes are split longitudinally before pair ing a t meiosis (ROBERTSON 1916, 193la, 1931b, MCCLUNG1928). RANDOLPH(1932) finds a longitudinal split in the early leptotene threads in a haploid Zea plant. These observations can not be reconciled either with BELLING'Stheory of crossing over or with DARLINGTON'S theory of meiosis. The random assortment of connecting fibers between chromomeres, postulated by BELLING,would mean that sister strand crossovers and many twists in paired chromatids would be expected. There is no evidence that sister strand crossing over occurs, and few twists are found in paired chromatids of either somatic or meiotic chromosomes. The third objection to BELLING'Stheory-and this applies to any interpretation of JANSSENS'hypothesis-is the prevalence of symmetrical relations of the chromatids in many genera, even where several chiasmata are found in a bivalent chromosome. It is also difficult t o account

260

PROCEEDINGS O F THE SIXTH

for interference in crossing over and the absence of crossing over in the Drosophila male, on any modification of JANSSENS'theory. DARLINGTON (1932) has recently revived the torsion hypothesis to explain how crossing over occurs. It is significant that he shows no diagrams illustrating how this mechanism might cause exact crossing over between only two of the non-sister chromatids at any one locus. McClungJs theory of chiasima formation According to MCCLUNG,chiasmata are caused by the alternate opening out of pairs of sister and non-sister chromatids. This interpretation has

(1916), WILSON(1925), been supported by WENRICH( 1916), ROBERTSON BELAR(1928), CAROTHERS (1926), and NEWTON(1926). As MCCLUNG (1927) has pointed out, such an origin of the chiasmata should usually result in a symmetrical relation of the chromatids. The apparent cross a t a chiasma is formed by chromatids which were diagonal at the fourstrand stage, and each chromatid may lie in the same quadrant at all loci (figure 2, diagram A ) . The relations of the chromatids at metaphase, when no crossover has occurred, are shown in figure 2, diagram E.

INTERNATIONAL CONGRESS O F GENETICS

261

The writer's theory of crossing over The writer (SAX1930, 1932a) has suggested that crossing over is caused by breaks in two of the four chromatids at a chiasma. A t very early diplotene many chiasmata may be formed in each bivalent chromosome, as shown in figure 2, diagram A. These chiasmata are formed by alternate pairing of sister and non-sister chromatids. Occasional half-twists in paired sister chromatids permit more or less random crossing over between any two homologous (non-sister) chromatids. Between the earliest diplotene stage and diakinesis there is considerable reduction in the number of chiasmata. Unequal opening out of certain internodes will result in the elimination of certain chiasmata by cancellation, as shown in figure 2, diagram B. I t is also possible that some chiasmata may be terminalized to such an extent that they will pass off the end of the chromosome. Crossovers occur only when two of the chromatids break at a chiasma, and the segments reunite in a new association, as shown in figure 2, diagrams A and B. When no crossover occurs, the chromatids a t a chiasma are symmetrical, each chromatid lies in the same quadrant on both sides of the chiasma, and the apparent cross a t the chiasma is formed by chromatids which were diagonal a t the four-strand stage. If the distal ends of the chromosomes open out in the same plane, a symmetrical cross is formed, as shown in figure 2, diagram E. A crossover, or a twist in sister threads, will produce an asymmetrical arrangement of the chromatids, as shown in figure 2, diagrams B and C. A crossover formed between two intact chiasmata will result in interlocking of chromatids a t metaphase (figure 2, diagram D). This theory of crossing over seems to meet the rigid genetic requirements, and is in accord with most of the cytological evidence. The lengths of the internodes, or loops, between chiasmata would account for interference. The reductional loop at the fiber attachment point is usually larger than the others, which would reduce crossing over in the spindle-fiber region. Chiasma formation and normal chromosome pairing can occur without crossing over, as is presumably the case in the Drosophila male. Crossovers between sister chromatids should be rare. Reciprocal and equational crossovers may occur with equal'frequency (resulting in a genetic ratio of 2 : l in attached XX's), but under certain conditions there should be an excess of equational crossovers. A symmetrical relation of the chromatids would be expected where no crossover has occurred, but half twists in paired sister chromatids or a crossover would produce asymmetrical chiasmata.

262

PROCEEDINGS OF THE SIXTH CYTOLOGICAL EVIDENCE

The arrangement of chromatids If MCCLUNG'Stheory of chiasma formation is correct, the chromatids should usually show a symmetrical arrangement at each chiasma if little or no crossing over has occurred. Chromosome configurations which seem to support this interpretation of chiasma formation have been shown in Orthopteran species by SUTTON( 1902), WENRICH( 1916), ROBERTSON ( 1916), JANSSENS ( 1924), BELAR( 1928), CAROTHERS (1926), MCCLUNG (1927), DARLINGTON and DARK(1932), and others. Similar figures have been shown for other groups of animals (WILSON1925). In plant species the individual chromatids are not clearly differentiated until metaphase, and then only when the smear technique is used. This technique has long been used by the zoologists, but has been adopted by pointed out its application to plant cells in botanists only since TAYLOR 1930), 1924. Symmetrical chiasmata have been found in Gasteria (TAYLOR Uvularia (BELLING1926), Tulipa (NEWTON1926), Paeonia (SAX1932b), Secale (SAX1930), and in Larix (H. J. SAX1932). According to NEWTON, the hypothesis "which explains the diakinetic figures as due to the opening out in two planes at right angles of what are originally four parallel chromatids, is adequate to explain the events of diakinesis and division in Tulipa and Fritillaria." Asymmetrical chiasmata have been described in Orthopteran species by JANSSENSand DARLINGTON, but these figures are apparently not typical for this group. As MCCLUNGhas pointed out, the clearest figures shown by JANSSENS show the chromatids in the same quadrants at all loci. DARLINGTON and DARK (1932) believe that most of the chiasmata in Stenobothrus are reciprocal and involve the same two chromatids at successive chiasmata. Of the two figures referred to, one (figure 7, chiasmata A and B) shows clearly that both chiasmata are symmetrical and diagonal, as is the case in most of the figures shown, where the chromatids can be observed. The interlocking of chromatids at metaphase does indicate that some crossing over has occurred. MZDA (1930a) finds a large proportion of asymmetrical figures in Lathyrus chromosomes, as might be expected, since crossing over is known to occur in this genus. Crossing over is also indicated by interlocked chromatids at metaphase. Asymmetrical chiasmata are frequently found in Paeonia chromosomes (SAX1932b). In most cases these figures could be attributed to half twists in sister chromatids, but some figures do seem to support JANSSENS' hypothe-

INTERNATIONAL CONGRESS O F GENETICS

263

sis. These exceptional figures may be due to an association of non-sister chromatids at the fiber attachment point, accompanied by a half twist in one pair of sister chromatids. The prevalence of symmetrical chiasmata in the Orthoptera and in certain plant species seems to indicate that JANSSENS'partial chiasmatypy theory of chiasma formation is untenable. The asymmetrical chiasmata found in both plant and animal species would be expected on MCCLUNG'S theory of chiasma formation if twists occur in sister chromatids, or if crossovers occur in some of the chiasmata which disappear between early diplotene and metaphase stages of meiosis. If no crossovers occur, almost all the chiasmata may be symmetrical, as seems to be the case in some species. Reduction in chias~+za frequency I f crossing over is caused by breaks in the chiasmata, as the writer assumes, then there should be a reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis. Reduction in chiasma frequency can also be attributed to the meeting and cancellation of chiasmata, as shown in figure 2, diagrams A and B. There is also the possibility that some chiasmata may pass off the ends of the chromosomes. At earliest diplotene there are numerous nodes and internodes along the bivalent chromosomes in many species of plants and animals. The frequency of these nodes, most of which are probably chiasmata, may be as high as 7 or 8 in certain chromosomes. Bivalent chromosomes with numerous nodes and internodes at early diplotene have been shown in Stenobothrus by JANSSENS(1924, figure 242), in reptilian chromosomes (1932), and in Ophyotrocha by GRBGOIREand DETON by NAKAMURA (1906). In plant species, chromosomes with an apparent high chiasma fre1904), Lilium and' Allium quency have been shown in Pinus (FERGUSON (BERGHS1904), Crepis (BABCOCK and CLAUSEN1929), Tulipa (NEWTON 1926), Nothoscordum (BEAL1932), and in Callisia ( S A X1932a). I n these genera the chiasma frequency may be reduced from five or six or more a t early diplotene to one or two at metaphase. A reduction in chiasma frequency between diplotene or early diakinesis and metaphase has been found 1932), Lilium (BELLING1931b), in Tulipa (NEWTON1926, DARLINGTON Primula ( DARLINGTON 1931a), Rosa (ERLANSON 1931) , Matthiola (PHILP and DARLINGTON 1932), and and HUSKINS1931), Campanula (GAIRDNER Callisia ( S A X1930). I n these genera the average number of chiasmata lost per bivalent between diplotene and metaphase is somewhat more than one. The loss of chiasmata reported in most cases would undoubtedly have been

264

PROCEEDINGS O F T H E SIXTH

greater if the earliest diplotene stages were favorable for examination of chiasma frequency. In some genera there is probably little reduction in chiasma frequency between early diplotene and metaphase. In these cases the chiasmata should be symmetrical, and little crossing over would be expected. The reduction in chiasma frequency is often associated with an increase in the proportion of terminal chiasmata, and, in many species, only terminal chiasmata are found at metaphase, even though there may be from three and DARK (1932), in disto five chiasmata at diplotene. DARLINGTON cussing chiasma terminalization, assume that "the attachment loop, in expanding, pushes all the others together at the end." But if each chiasma represents a crossover, the accumulation of two or more chiasmata would often result in a complex association of chromatids, as shown in figure 1, diagram B. In numerous species, terminal or sub-terminal chiasmata show no interlocking or asymmetrical relations of the chromatids, even where three or more chiasmata are found at diplotene. If two crossover chiasmata on the same side of the spindle fiber are terminalized without passing off the ends of the chromosomes, they will be reduced to one interlocked chiasma, or canceled, depending on the types of crossovers. If each chiasma is reciprocal and both crossovers occur in the same two chromatids, the two chiasmata will be canceled, o r reduced to one, with equal frequency. If crossing over is at random between homologous chromatids, the two chiasmata would be reduced to one interlocked chiasma in 50 percent of the cases, and reduced to none in 50 percent of the cases. In Primula and Campanula the chiasma frequency is reduced from three or more at diplotene to two terminal chiasmata at metaphase, but there is no evidence of interlocked chromatids in these genera. On the writer's hypothesis, the reduction in chiasma frequency between early diplotene and metaphase can be attributed to cancellation of chiasmata and breaks in some of the chiasmata as the early diplotene loops open out. Two chiasmata might be reduced to none by cancellation and reduced to one if the other chiasma were broken. Both the genetic and the cytological evidence support MORGAN'S (1925) suggestion that crossing over is an accidental by-product of meiosis. SANSOME (1932) has described a case of chiasma formation in a ring of six chromosomes in Pisum, which seems to support the partial chiasmatypy hypothesis. In 62 out of 78 figures one or two chiasmata were found in the "X segment" of the ring, resulting in a figure-of-eight configuration. Occasionally two chiasmata were found in this region, but the exact

INTERNATIONAL CONGRESS OF GENETICS

265

proportions of double and single chiasmata are not given, presumably owing to the difficulty in studying the chromatids in this region. On the writer's theory, two chiasmata must be formed in the X segment if only two chromosomes can pair at the same locus. I f one of the chiasmata breaks, the remaining one represents a crossover. The remaining chiasma would be under a special strain, but it seldom breaks. A much more critical case of the same type has been shown in Zea by BEADLE(1932). Chiasma frequency was obtained at both diplotene and metaphase, and the crossover length of the segment was known. The ring or chain of 4 chromosomes, resulting from segmental interchange, is so constituted that crossing over in one arm is limited to about 12 crossover units between the translocation point and the region where a teosinte chromosome segment has been introduced. When a chiasma is formed in this region, a ring with two free arms is found a t diakinesis and metaphase, and when no chiasma is present, the four chromosomes are arranged in a chain. At diakinesis 20 percent of the associations of four chromosomes were in the form of rings, but a t metaphase only 10 percent of these chromosomes were rings. On the writer's theory, a t least two chiasmata must be formed in this wz-translocation region. K O terminalization of chiasmata can occur because of the change in homology of the distal segments. As the ring opens out, all the chiasmata in this region may disappear by cancellation (compare figure 2, diagram B). If one chiasma breaks, the remaining chiasma represents a crossover. This chiasma can not be canceled or terminalized. The frequency of such a crossover chiasma in the wz-translocation segment a t diakinesis (20 percent) corresponds with the crossover frequency found in the derived Zea-teosinte chromosomes as bivalents. But at metaphase the chiasma frequency was only 1 0 percent. If the frequency of ring formation a t meta~haseis significant, this must mean that two chiasmata were often present at diakinesis and were canceled before metaphase, or that the crossover chiasma was broken. If the crossover chiasma breaks, a double crossover might result, but the two crossovers would probably be too close together to be detected. The effect of a break in the single crossover chiasma on the crossover length of the segment would depend on types of double crossovers obtained. If chiasma frequency is high a t early diplotene, the writer's explanation of chiasma formation in the wz-translocation segment is not improbable. Two chiasmata might be formed. Unless one chiasma breaks, the original chiasmata probably would be canceled, owing to the strain imposed by the

266

PROCEEDINGS OE THE SIXTH

opening of the ring and limitations in terminalization. If a chiasma breaks, cancellation of the remaining chiasma is impossible, and since the remaining crossover chiasma can not be terminalized, it is found in the ws-translocation region a t diakinesis and metaphase.

Behavior of unequal homologues In all recorded cases of pairing of unequal homologues, the long chromatids are paired with long and the short ones are paired with short a t meiosis. With only one chiasma and a terminal spindle-fiber attachment, it is difficult to explain the occurrence of both pre- and post-reduction in the same bivalent (SAX 1932a, b) as found by WENRICH(1916). This be(1932) on the assumption that the havior is explained by DARLINGTON fiber attachment is median, but the assumption is clearly erroneous, as shown by WENRICH'Sfigures of somatic chromosomes in Phrynotettix. On the MCCLUNGinterpretation of chiasma formation, the chromatids must separate equationally at the fiber attachment, where a single chiasma is formed between the fiber and the unequal segments if no crossing over has occurred. This assumption may be valid for some species, although it is apparently not the case in Drosophila or Zea.

Interlocked bivalents Interlocked non-homologous bivalents have been found in Oenothera, ahd Campanula, Tradescantia, and other genera. According to GAIRDNER DARLINGTON ( 1932), three types of interlocking would be expected : ( 1 ) proximal interlocking, where the loops containing the spindle fiber are involved; ( 2 ) distal interlocking, or locking of terminal loops; and ( 3 ) proximal-distal interlocking. If chiasmata are formed by alternate opening of sister and non-sister chromatids at diplotene, interlocking of bivalents could occur only in alternate internodes where sister chromatids are paired. If each chiasma represents a crossover and sister chromatids are always paired at early diplotene, interlocking can occur at any internode. In Campanula (GAIRDNER and DARLINGTON 1932) from two to six chiasmata are found in each bivalent, and the most frequent number seems to be three at diplotene. If each chiasma represents a crossover, distal or proximaldistal interlocking should be at least as frequent as proximal interlocking. theory of chiasma formation is correct, distal or proxiBut if MCCLUNG'S mal-distal interlocking should be very much less frequent than proximal interlocking. These conclusions are based on the assumption that chiasmata do not pass off the ends of the chromosomes but are pushed together a t

INTERNATIONAL CONGRESS O F GENETICS

267

the ends of the chromosomes, as postulated by DARLINGTON and DARK find, in Campanula, that "proximal (1932). GAIRDNER and DARLINGTON interlocking occurs in about 20 percent of nuclei in homozygous groups. Distal interlocking seems to be much rarer." Similar observations have been made in Oenothera by CATCHESIDE ( 1931 ) . Proximal interlocking is common in Tradescantia, but distal interlocking is rare (SAX and ANDERSON theory of chiasma forma1932). These results indicate that MCCLUNG'S tion is correct, or that chiasmata do pass off the ends of the chromosomes before metaphase. According to GAIRDKER and DARLINGTON, interlocking should occur between the chromatids of paired or unpaired chromosomes if the chromatids ever separate equationally at diplotene. This conclusion is obviously untenable ( S A Xand ANDERSON 1932). GENETIC EVIDENCE

Non-disjunction and crossing ovcr

In high non-disjunction lines of Drosophila, crossing over is greatly reduced in both the normal progeny and in the exceptional females (ANDERSON 1929). The genetic evidence indicates that non-disjunction is caused by a failure of chromoson~epairing at the first meiotic division. The segregation of the univalent homologues will result in no-X and X X eggs which give rise to the exceptional males and females. ANDERSON found only 7.3 percent crossing over in exceptional females, and most of these crossovers were a t the distal end of the chromosome. This distribution of crossovers would indicate that the few chiasmata formed are usually near the distal end of the chromosome, and that they either break or are prematurely terminalized, so that univalent XX's result at the first meiotic metaphase. If non-disjunction is due to premature terminalization, it is difficult to account for the crossovers near the forked locus, if chiasmata represent crossovers, because such a chiasma would have to be terminalized almost the entire genetic length, and for about half of the cytological length, of the X chromosome. If such premature terminalization could occur, one might expect frequent non-disjunction in normal stocks of Drosophila, where the formation of a single chiasma would be expected in the distal half of the X in about two-thirds of the meiotic divisions. But non-disjunction occurs only once in about 2000 times in normal stocks. I t would, of course, be impossible to explain crossing over in non-disjunction XX's on the partial chiasmatypy hypothesis if terminal affinity prevents the chiasmata from sliding off the ends of the chromosomes,

268

PROCEEDINGS O F T H E S I X T H

as suggested by DARLINGTON (1932). If only one or two chiasmata are formed, and they break, crossing over would be found occasionally in nondisjunction XX's, even in the region of the forked locus. The writer's theory of crossing over seems to provide a more plausible explanation of crossing over in non-disjunction chromosomes than does the partial chiasmatypy theory, although neither theory is entirely satisfactory (SAX1932a). According to DARLINGTON (193 1) , chromosome pairing is invariably dependent on chiasma formation. If every chiasma represents a crossover, no normal chromosome can have a crossover length of less than 50 units. Little or no crossing over has been found in the fourth chromosome of Drosophila melanogwter, although this chromosome is as regular in pair(1932) ing and disjunction as the X, which is 70 units long. DARLINGTON assumes that this fourth chromosome may have a crossover length of 50 units, but that the number of genetic factors is not sufficient to prove it. Regular chiasma formation might be expected in a short bivalent if the chromatids open out reductionally at one end of the chromosome and equationally at the other, as WENRICH(1916) assumes. I t has been clearly demonstrated that chromosome pairing is not always dependent on chiasma formation (SAX1932b), although in marly species with long chromosomes pairing at meiotic metaphase does seem to be dependent on chiasmata.

Randomness of crossing over The genetic evidence shows clearly that crossing over is not confined to two of the four chromatids (ANDERSON and RHOADES1931, REDFIELD 1930, and others). The first crossover from the spindle fiber in the X chromosome of Drosophila is at random between the homologous (non-sister) chromatids, as shown by the proportion ( 2 : l ) of equational and reciprocal crossovers in attached XX's (ANDERSON 1925) and by the percentage of homozygous recessives at the forked locus in the attached XX's (RHOADES 1931, STURTEVANT 1930). If crossing over i s at random at all crossovers, the homozygosis in attached XX's at the distal ends of the chromosomes should be about 20 percent (SAX 1932a). The percentage of homozygosis found by STURTEVANT was 17.1, and by RHOADES18.6. The deficiency of homozygosis may not be significant, however, owing to the lower viability of the homozygous recessive segregates. I f crossing over is at random for all crossov.ers at all loci, the first crossover in attached XX chromosomes should be equational or reciprocal in the proportions of 2 : l . If the first crossovers are not at random, there

INTERNATIONAL CONGRESS O F GENETICS

269

should be an excess of equationals on the writer's theory, and an excess of reciprocals on BELLING'Stheory. Three types of second crossovers should be found in the ratio of 2:8: 1, as follows: (1) equationals homozygous at the distal end; ( 2 ) equationals homozygous at the proximal end; and ( 3 ) reciprocals. If the second crossover is not a t random, the second class of crossovers will be decreased and the third class increased. Random crossing over can occur, on the writer's theory, only if some half twists occur in paired sister chromatids (SAX 1932a). The genetic analysis of attached XX's in Drosophila indicates that the first crossover occurs at random. The data on second crossovers and the percentage of homozygosis at different loci are not adequate for a critical test of random assortment of chromatids at the second crossover.

Absence of crossing over in the Drosophila male No crossing over occurs in the Drosophila male, but pairing and disT junction of all chromosomes are regular (METZ 1926, G U Y ~ N Oand (193 1 ) assumes that chiasma formation and NAVILLE1928). DARLINGTON crossing over are essential for chromosome pairing. H e assumes that in the male Drosophila there are always two chiasmata in each bivalent, one on each side of the spindle fiber; that the two chiasmata are very close together; that no mutations occur in the region between chiasmata; and that both crossovers are reciprocals and involve the same two chromatids. These assumptions are not only highly improbable, but also they can not be reconciled with either BELLING'Sor DARLINGTON'S explanation of the mechanism for crossing over. The assumption that all crossovers in the male are reciprocal is difficult to reconcile with the fact that both equational and reciprocal crossovers occur in the female. The writer assumes that some chiasmata are formed in the bivalents of the Drosophila male, but that these chiasmata are terminalized without breaking, as seems to be the case in several genera. The apparent differences in the duration of the metaphase stages at meiosis in males and females seem to support this assumption. CHIASMA FREQUENCY AND CROSSING OVER

If each chiasma represents a crossover, it should be possible to estimate the crossover length of the chromosomes in certain species, especially where (1931a) little or no terminalization of chiasmata is found. DARLINGTON attempted to correlate chiasma frequency with crossover length of one of the chromosomes in Primula, but with rather unsatisfactory results (SAX

270

PROCEEDINGS O F THE SIXTH

1932a). Zea should be satisfactory for this purpose, but little is known concerning chiasma frequency at prophase stages. There is sufficient cytological and genetic evidence to warrant a comparison of chiasma frequency and crossover lengths of the chromosomes of Vicia faba. In this species MEDA (1930b) found, at metaphase and diakinesis, an average of 8.1 chiasmata in the long bivalent and an average of 3.5 chiasmata for each of the five short bivalents. SIRKS has studied 26 genetic factors in Yicia faba. Of these 26 factors, 19 were found in four linkage groups, and only 7 were independent. Of the 19 factors, 7 were found in the first linkage group, 4 in the second, 5 in the third, and 3 in the fourth. In no case did any linkage group exceed approximately 50 crossover units. I i each chiasma represents a crossover, there should be more independent chromosome units than there are factors. The long chromosome should have a crossover length of 400 units, and the average length of the short chromosomes should be 175 units. With such a high frequency of crossing over, it would be highly improbable that 19 of the 26 factors would be found in four linkage groups, none of which exceeds 50 crossover units.

According to JANSSEXS' partial chiasnlatypy hypothesis, each chiasma represents a crossover which occurred at pachytene. BELLING has offered the only plausible explanation of how crossing over might occur on this theory, but his explanation is not in accord with certain critical cytological and genetic observations. If each chiasma represents a crossover, it is difficult to account for the prevalence of symmetrical chiasmata in many species and the reduction in chiasrna frequency often found between early diplotene and metaphase stages of meiosis. If, as DARLINGTON believes, chromosome pairing at meiotic metaphase is dependent on chiasma formation, it is difficult to account for the absence of crossing over in the Drosophila male, and difficult to explain the crossover length of the fourth chromosome of D. melanogaster. If each chiasma represents a crossover, it is difficult to reconcile the cytological observations with the genetic results in Vicia faba. The writer's theory of crossing over is based on the assumption that chiasmata are caused by an alternate opening out of pairs of sister and nonsister chromatids. Crossing over occurs only when two of the four chromatids break at a chiasma. This theory is in accord with the following genetic evidence: only two chromatids cross over at any one locus; one

1 l i T E R N A T I O N A L CONGRESS OF G E N E T I C S

271

crossover interferes with a second one; crossing over is reduced in the spindle-fiber region; no crossing over occurs in the male Drosophila; and crossing over is more or less at random between any two homologous chromatids. This theory is supported by the cytological evidence that there is often considerable reduction in chiasma frequency between earliest diplotene and metaphase stages of meiosis, and by the prevalence of symmetrjcal chiasmata in many species. LITERATURE CITED A N ~ E R S O E. N , G., 1925 Crossing over in a case of attached X chromosomes in Drosophila melanogaster. Genetics 10:403-417. 1929 Studies on a case of high non-disjunction in Drosopltila melanogaster. Z. indukt. Abstamm.-u. Vererb1.chre. 51:341-397. ANDERSOS,E . G., and RHOADES,M. M., 1931 T h e distribution of interference in the X chromosome of Drosophila. Papers Michigan Acad. Sci. 13:227-239. BABCOCK, E . B., and CLAUSEN, J., 1929 Meiosis in two species and three hybrids of Crepis and its bearing on taxonomic relationship. Univ. California Pub. Agric. Sci. 2 :401-432. BEADLE,G. W., 1932 The relation of crossing over to chromosome association in ZeaEuchlaena hybrids. Genetics 17:481-501. BEAL,J. M., 1932 hficrosp~rogenesisand chromosome behavior in Nothoscordum bivalve. Rot. Gaz. 93:278-295. BBLA~,K., 1928 I)ie cytologischen Grundlagen der Vererbung. p. 412. Berlin: Borntraeger. RELLIKG,J., 1926 Single and double rings a t the reduction division in Uvularia. Biol. Bull. 50 :355-363. 1927 T h e attachments of chromosomes at the rcduc:ion division in flowering plants. J. Genet. 18:177-205. 1931a Chromomeres o i liliaceous plants. Univ. California Pub. Bot. 16:153-170. 1931b Chiasmas in flowering plants. Univ. California Pub. Bot. 16:311-338. BERGHS,J., 1904 Formation des chromosomes hkt6rotypiques dans la sporog6nbse vPg6tale. La Cellule 21:173-188. IJRIDCES, C. B., 1916 Non-disjunction as proof of the chromosome theory of heredity. Genetics : 1:l-52, 107-163. BRIXES,C. B., and ANDERSOX, E. G., I925 Crossing over in the X-chromosomes of triploid females of Drosophila melanogaster. Genetics 10:418-441. CAROTHERS, E . E., 1926 The maturation divisions in relation to t h e segregation of homologous chromosomes. Quart. Rev. I3iol. 1:119-435. CATCHESIDE, 1). G., 1931 Critical evidence of parasynapsis in Oenothera. Proc. Roy. Sot. London 109:165-184. COOPER,D. C., and BRINK,R. A,, 1931 Cytological evidence for segmental interchange between non-homologous chromosomes in maize. Proc. Nat. Acad. Sci. Washington 17 ~334-338. CREIGHTON, H. R., and MCCLINTOCK, B., 1931 A correlation of cytological and genetical crossing-over in Zea mags. Proc. Nat. Acad. Sci. Washington 17:492-497. DARLIXGTOK, C. D., 1931a Meiosis in diploid and tetraploid Primula sinensis. J. Genet24 :65-%.

1931b Meiosis. Biol. Rev. 6:221-264. 1932 Recent advances in cytology. p. 559. Philadelphia: P. Ulakiston's Son and Co.

272

PROCEEDINGS O F T H E S I X T H

DARLINGTON, C. D., and DARK,0 . S., 1932 The origin and behavior of chiasmata. 11. Stenobothrus parallelus. Cytologia 3:169-185. ERLANSON, E. W., 1931 Chromosome organization in Rosa. Cytologia 2:256-282. FERGUSON, M. C., 1904 Contributions to the knowledge of the life history of Pinus, with special reference to sporogenesis, the development of the gametophytes, and fertilization. Proc. Nat. Acad. Sci. Washington 6:l-202. GAIRDNER, A. E., and DARLINGTON, C. D., 1932 Ring-formation in diploid and polyploid Campanula persicifolia. Genetica 13:113-150. GOWEN,J. W., 1928 Mutation, chromosome non-disjunction and the gene. Science 68:211212. G R ~ I R EV., , and DETON,W., 1906 Contribution i I'etude de la spermatogtnese dans I'Ophyotrocha puerilis. La Cellule 23:435-440. GUY~NOT, E., and NAVILLE,A,, 1928 Les chromosomes et la reduction chromatique chez Drosophila melanogarter. La Cellule 39:27-81. JANSSENS,F. A,, 1924 La chiasmatypie dans les insectes. La Cellule 34:135-359. KAUFMANN,B. P., 1931 Chromonemata in somatic and meiotic mitoses. Amer. Nat. 63 :280-283. MAEDA,T., 1930a The meiotic divisions in pollen mother cells of the sweet pea (Lathyrus odoratzu), with special reference to the cytological basis of crossing over. Mem. Coll. Sci. Kyoto Imp. Univ. 5239-123. 1930b On the configurations of gemini in the pollen mother cells of Vicio faba L. Mem. Coll. Sci. Kyoto Imp. Univ. 5:125-137. MCCLUNG,C. E., 1927 The chiasmatype theory of JANSSENS.Quart. Rev. Biol. 2:344-366. 1928 Differential chromosomes of Mecostethus gracilis. Zeit. Zell. v. m. Anat. 7 :756-778. METZ, C. W., 1926 Observations on spermatogenesis in Drosophila. Zeit. Zell. Mik. Anat. 4 :1-28. MORGAN, L. V., 1925 Polyploidy in Drosophila melanogaster with two attached X chromosomes. Genetics 10:148-178. MORGAN, T. H., 1925 The bearing of genetics on cytological evidence for crossing over. La Cellule 36:113-123. MULLER, H . J., 1916 The mechanism of crossing over. Amer. Nat. 50:193-221. NAKAMURA, K., 1932 Studies on reptilian chromosomes. 111. Chromosomes of some Geckos. Cytologia 3:156-168. NEWTON,W . C. F., 1926 Chromosome studies in Tulipa and some related genera. J. Linn. soc. 47 :339-354. PHILP, J., and HUSKINS,C. L., 1931 The cytology of Matthiola incana R. Br, especially in relation to the inheritance of double flowers. J. Genet. 24:359-404. RANDOLPH, L. F., 1932 The chromosomes of haploid maize, with special reference to the double nature of the univalent chromosomes in the early meiotic prophase. Science 75 :566-567. REDFIELD,HELEN,1930 Crossing over in the third chromosome of triploids of Drosophila melanogaster. Genetics 15:205-252. RHOADES,MARCUSM., 1931 The frequencies of homozygosis of factors in attached X females of Drosoplzila melanogarter. Genetics 16:375-385. 1932 The genetic demonstration of double strand crossing-over in Zea mays. Proc. Nat. Acad. Sci. Washington 18:481-484. ROBERTSON, W. R. B., 1916 Chromosome studies. I. J. Morph. 27:179-332. 1931a A split in chromosomes about to enter the spermatid (Paratettix texanus). Genetics 16 :349-352.

INTERNATIONAL CONGRESS O F GENETICS

273

1931b Chromosome studies. 11. Synapsis in the Tettigida with special reference to the presynapsis split. J. Morph. and Physiol. 51:119-146. SANSOME, E. R., 1932 Segmental interchange in Pisum sativum. Cytologia 3:200-219. SAX,H. J., 1933 Chiasma formation in Larix and Tsuga. Genetics 18:121-128. SAX, KARL,1930 Chromosome structure and the mechanism of crossing over. J. Arnold Arb. 11 :193-220. 1932a The cytological mechanism of crossing over. J. Arnold Arb. 13:180-212. 1932b Meiosis and chiasma formation in Paonia suffruticosa. J. Arnold Arb. 13:375-384. SAX,KARL,and ANDERSON, EDGAR,1933 Segmental interchange in Tradescantia. Genetics 18: 53-67. SIRKS, M. J., 1932 Beitrage zu einer genotypischen analyse der Ackerbohne, Vicia faba L. Genetica 13:209-631. STERN,C., 1931 Zytologisch-genetische Untersuchungen als Beweise fiir die Morgansche Theorie des Faktorenaustausches. Biol. Zbl. 51:547-587, STURTEVANT, A. H., 1913 The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14:43-59. 1925 The effects of unequal crossing over at the bar locus in Drosophila. Genetics 10:117-147. 1926 A crossover reducer in Drosophila melanogaster due to inversion of a section of the third chromosome. Biol. Zbl. 46597-702. 1928 A further study of the so-called mutation at the bar locus of Drosophila. Genetics 13:401-409. 1930 Two new attached-X lines of Drosophila melanogaster, and further data on the behavior of heterozygous attached-X's. Pub. Carnegie Instn. 42161-81. SUTTON,W. S., 1902 On the morphology of the chromosome group in Brmhystola magna. Biol. Bull. 4:24-39. TAYLOR, W. R., 1924 The smear technique for plant cytology. Bot. Gaz. 78:236-238. 1930 Chromosome structure in mitosis and meiosis. Contr. Bot. Lab. Univ. Pennsylvania 7 96.5-270. WEINSTEIN,A,, 1918 Coincidence of crossing over in Drosophila melanogaster. Genetics 3:135-173. WENRICH,D. H., 1916 The spermatogenesis of Phrynotettix magnus with special reference to synapsis and the individuality of chromosomes. Bull. Mus. Comp. Zool. 60:57-133. WIISON, E. B., 1925 The cell in development and heredity. p. 1232. New York: The Macmillan Co.

PKOCEEDIKGS O F THE SIXTH

274

ON T H E G E N E T I C N A T U R E OF INDUCED MUTATIONS I N PLANTS1 L. J . Stadler, lMissouri Agrictlltural Experiment Station and United Stabes Department of Agriculture, Columbia, Missouri

I t is generally considered that the hereditary variations induced by irradiation are due to two fundamentally different types of change produced within the irradiated cells, (1) breakage and rearrangement of the chromosomes, and ( 2 ) modification of the composition of individual genes. The possibility that the induced gene mutations may be due to mechanical changes analogous to those causing the grosser chromosomal aberrations is rejected chiefly on the basis of the induced reverse mutations in Drosophila which have been convincingly demonstrated by MULLER(1928), TIMOFI?EFF-KESSOVSKY ( 1930), and others. PATTERSON and MULLER( 1930), after a comprehensive experimental and theoretical analysis of this question, conclude that the induced point mutations are changes in the chemical composition of the genes, and that they probably are "endless in their eventual possibilities." This interpretation is applied to the point mutations in general, which are considered a group fundamentally distinct from the variations due to chromatin displacement. The results of genetic experiments with X-rays in plants are not entirely in harmony with this view. Radiation induces gene mutation as well as grosser chromosomal variations in plants as in the fruit fly, and the induced mutations meet all tests of typical gene mutation. The evidence from plants considered alone, however, does not permit any sharp differentiation between the induced gene mutations and various extra-genic alterations which may be expected to accompany the types of chromosonlal derangement brought about by the treatment. This does not imply that mutation as a whole is purely a mechanical process, for it is possible that the mutations affected by radiation are a special class not representative of mutation in general. This special class may be wholly or largely made up of mechanical or extra-genic changes. There is no reason to assume that irradiation could not produce chemical changes within the gene, changes which might well be representative of those involved in the natural evolution of the gene. But since chemical changes in the gene are beyond direct investigation the conclusion that the genes are chemically transformed must be based almost entirely on negative evidence. T o state that an induced variation is a gene mutation is not

' Contri.bution from

1)epartment o f Field Crops, M r s s o u n ~Acnrccr.-rcn~r. EXPERIMENT

STATION, Journal Series 351.

INTERNATIONAL CONGRESS O F GENETICS

275

to explain it but merely to label it. W e do not demonstrate that a chemical change has occurred; we simply infer, since no mechanical explanation can be found, that the variation must be due to this invisible mechanism. The implications of this conclusion are so far reaching, both for genetical theory and for breeding practice, that it is essential to examine thoroughly any possibility of accounting for the presumed mutations a s phenomena subject to direct investigation. In this paper I shall consider the evidence from experiments with plants, chiefly for its bearing on two questions: (1) Are the induced mutations partly or wholly the result of mechanical alterations, and (2) are they representative of natural mutation in general? WHAT IS MUTATION

?

I shall use the term mutation exclusively in the restricted sense of "gene mutation." Even in the restricted sense the term is rather indefinite in meaning. We may define mutation as a transmissible change in the gene. But we identify mutations by experimental tests, and these tests are not such as to establish conclusively, in specific instances, that a change within the gene has occurred. In effect, any Mendelizing variation which can not be shown to be due to a change involving more than one gene is a mutation. In Drosophila the gene variation is distinguished from variations of higher order chiefly by the use of closely linked genes. In plants few closely linked genes are available for use, but the inviability of deficient genomes in the haploid generation serves to some extent as an alternative distinction between mutation and deficiency. In either case the distinction is more or less arbitrary. Deficiencies may occur which do not involve the loci of two known genes, and which therefore cannot be distinguished from mutations. On the other hand, it is possible that variations involving changes at neighboring loci may in some cases be due not to the ioss or addition of a chromosome segment but to simultaneous changes within the two genes affected. Similarly, in plants, there may be minor deficiencies which have no lethal effect when haploid, and which therefore are inherited as mutations, while in other cases transformations of single genes may be lethal to the gametophyte and thus may be eliminated with the deficiencies. Thus the working definition of mutation necessarily differs somewhat from the ideal definition. It is this working definition which must be considered in generalizing from the experimental evidence. The mutations ex-

276

PROCEEDINGS OF THE SIXTH

perimentally known may include not only variations due to alterations within the gene but also variations due to losses of genes or groups of genes, to additions of genes or groups of genes, and possibly also to changes in the spatial relations of genes to one another. It is easy to say that the latter groups are excluded by definition and should not be considered mutations. But in practice it is ordinarily impossible to distinguish them from mutations of the ideal type. Many variations previously classed as mutations have been removed from that class as mechanical explanations were found. I t is safe to assume that other types of variation still classed as mutation are the result of extra-genic alterations not yet recognized. RADIATION-INDUCED MUTATIONS I N

PLANTS

The induced mutations found in barley and maize are typical gene mutations in the sense that they show normal Mendelian inheritance and normal linkage relations. They are fully viable variations recognized by their visible phenotypic effects, and in many instances they appear to be identical with previously known mutants of spontaneous origin. I n the course of experiments on the relation of various factors to mutation frequency in barley, many hundreds of these mutations have been observed. If these mutations are due to the loss of germinal material rather than to some change within the gene, the losses are so slight as to have no lethal effect on the haploid gametophyte. Most of the mutants are the progeny of fully fertile plants, and the homozygous mutants which reach the flowering stage are also fully fertile. All of the induced mutations found thus far are recessive. The most striking difference between the results in barley and the results of comparable experiments in Drosophila is the absence of induced dominant mutations in the plant. Why do X-rays induce dominant and recessive mutations indiscriminately in the fly, and induce recessive mutations only in barley? This apparent difference is due at least in part to the different technique used in distinguishing mutations from grosser chromosomal changes in the plant and in the animal. In the plant the alternation of diploid and haploid generations subjects all of the induced variations to the test of haploid survival before they may appear in the diploid progeny. For example, if the treatment causes the loss of a chromosome segment essential to cell survival (that is, a deficiency lethal when haploid or homozygous), the diploid individual in which the change was induced may survive. But

INTERNATIONAL CONGRESS O F GENETICS

277

half of its spores receive a deficient set of chromosomes, and these are unable to develop through the gametophyte generation to the production of functional gametes. Thus the gametophyte generation filters out the variations due to lethal chromosomal derangements, and the variations observed in the progeny are only those which can survive two or three cell generations in haploid form. These may be assumed to be mutations, though it is possible that even these may include variations due to actual losses of genes which are not essential to the gametophyte. There is no reason to suppose that dominant variations are not induced in plants as commonly as in animals. Their occurrence in maize may be demonstrated in progenies produced by the use of irradiated pollen, which are heterozygous for any changes induced in the male germ cells. Among such progenies defectives of many distinctive types are found. In these plants a t least half of the pollen and ovules are aborted, and the defective type is not reproduced in the progeny. By the use of dominant genes as markers in the treated pollen it may be shown that many of these plants are deficient for some part of the normal chromosome set, and that it is the deficient spores that do not function. The characteristic type of defective development associated with deficiency of certain chromosomal regions is striking. These characteristic defects, associated with specific deficiencies, are typical induced dominant variations of maize. In an organism in which deficient gametes are functional, they would be transmitted to three-fourths of the progeny. One-fourth, being homozygous, would commonly be inviable; the other two-fourths would be heterozygous like the parent and would show the characteristic effect of the deficiency. Thus the defect would be inherited in the same manner as the familiar "dominant mutations, lethal when homozygous," which make up the majority of the known dominant mutations of Drosophila. These are dominants of a different type from those with which we are ordinarily concerned in the cultivated plants. Numerous dominant variations are known in cultivated races of barley, entirely free from lethal effect and inherited in simple Mendelian fashion. These include various plant colors and other characters which could be detected in sectorial chimeras. The approximate distribution of the sectors to be expected is known from chimeras produced by the irradiation of heterozygous seed; and it is certain that they could be detected. The variety of barley used is recessive for a number of such genes. All treated plants have been minutely searched for evidence of the dominant mutation of any of these known genes or of un-

278

PROCEEDINGS O F THE SIXTH

known dominant genes. Some 80,000 heavily treated plants have been thus examined. Each tests the occurrence of dominant mutation not in a single treated cell but in each of several treated cells. No case of dominant mutation has been found. This result is in sharp contrast to the experience of TIMOF~EFF-RESSOVSKY (1930) who tested several mutant races of Drosophila for dominant reverse mutation, and, though he used much smaller populations, found one or more cases of dominant mutation in about half of the genes tested. Whether or not the mutant races of barley will yield reverse mutations has not yet been determined. It is not necessarily true that all of the dominant variations which are lethal to the gametophyte are due to gross chromosomal alterations. As mentioned above, the gametophyte lethals removed in the normal process of reproduction in plants may include some intra-genic as well as extra-genic alterations. If the recessive mutations are due to some change within the gene, it is reasonable to suppose that corresponding changes producing a dominant effect will sometimes occur. But the surprising fact remains that when the variations lethal to the garnetophyte are removed only the recessive variations remain. Any change sufficiently extreme to show its effect in the presence of its normal allelomorph appears to be sufficiently deleterious to be inviable in haploid cells. This apparently quantitative relation between the dominant and recessive variations suggests the possibility that both may be made up chiefly of deficiencies, and that only the deficiencies of very minor extent may survive the haploid generation. But it is very improbable that all types of induced mutation are the result of deficiency-particularly those involved in the reverse mutations of Drosophila. It is possible however that the extra-genic variations of nonlethal effect, which are brought about by the treatment, include types other than mere losses. In order to consider eHectively what types may reasonably be expected we need as clear a picture as possible of the way in which irradiation affects the chromosomes. A study of the grosser changes within reach of cytological technique may yield a basis for inferences regarding events on a genic scale. THE MODE O F ACTION O F RADIATION

Cytologists have long recognized certain typical reactions of the chromosomes to X-rays. There was no basis for a mechanical theory of the phenomenon, however, until the results of combined genetic and cytological investigations became available during the last five years. The evidence is

INTEKNATIOSAL COXGRESS OF GENETIC3

279

not yet sufficient to provide a clear picture of the mode of action of racliation, but it does permit the formulation of definite working hypotheses subject to experimental test. In the genetic investigations with X-rays it soon became evident that breaks and rearrangements of the gene-string were a characteristic effect of irradiation, and cytological investigations showed that these corresponded to actual derangements of the chromosome material. The chromosomal alterations reported have included the loss of a segment of a chromosome (either terminal or internal), the removal of a segment to a new position in the same or in another chromosome, the inversion of a segment in its original position, and the interchange of segments between chromosomes. It was at once recognized that all of these effects might be accounted for as results of the random breakage and reattachment of the chromosomes, and that additional deficiencies and duplications might he expected as a result of genetic recombination of affected and unaffected chromosomes. SEREBROVSKY (1929) has suggested as a mechanism for such chromosomal alterations (and for gene mutations) a tendency of chromosomes to become attached and later to break apart at a different point. Presumably, on this hypothesis, the effect of X-rays is due to an increase in the tendency of the chromosomes to become attached. Mutations are considered minor deficiencies, and they with other deficiencies are assumed to be a by-product of translocation. Other investigators also have noted the possibility that the characteristic genetic effects of radiation may be traced to a basic effect on translocation. The chromosomal alterations induced by X-rays in maize are similar to those found in Drosophila. Although the genetic analysis of the induced changes is exceedingly crude in comparison to that which may be made in Drosophila, these alterations are of interest to us at this point because they have had the benefit of critical cytological study at the period of synapsis. At this stage in the propliase of the first meiotic division, because of the length of the chromosomes and the pairing of homologous parts, aberrant chromosomal conditions may be identified with a precision impossible in the study of the condensed chromosomes. The cytological investigation is entirely the work of Doctor BARBARA MCCLINTOCK, and was made possible by her development of the technique for the study of prophase chromosomes in maize and her determination of the distinctive morphology of the 10 chromosomes in normal material. The cytological observations have been described in detail ( MCCLIKTOC K 1931) . In certain important respects, the chromosomal alterations found dif-

280

PROCEEDINGS O F THE SIXTH

fer from those which might be assumed on the basis of genetic evidence o r of the cytological study of condensed chromosomes. They include deficiency, inversion, and segmental interchange, and various modifications and combinations of these phenomena. No case of simple translocation of a fragment to a whole chromosome has been found. Deficient segments may be either terminal or internal, but inversions are always internal. Many deficiencies are associated with translocation, but there are also some deficiencies which occur in cells otherwise normal. The absence of simple translocations and terminal inversions, if later investigation shows it to be general in irradiated material, will greatly simplify the interpretation of the effects of radiation. Among the induced chromoson~alderangements which have been reported, these are the only types which involve the attachment of a fragment a t an unbroken end. If they do not occur it is unnecessary to assume any increased "stickiness" of the chromosomes under treatment. All of the chromosomal attachments observed may then be ascribed to the tendency of chromoson~efragments to become attached at their broken ends, a tendency already familiar in the normal process of crossing over. It is true that many cases of simple translocation "are known," but the study of meiotic prophase material indicates that cases giving both genetic and cytological results typical of simple translocation may actually be the result of segmental interchanges in which one (1932) has recently described a of the segments is very short. BURNHAM particularly instructive case of this kind. All of the induced chromosomal derangements which have been found in maize are consistent with the hypothesis that the effect of the treatment is merely to break the chromonema a t various points, and that the fragments tend to attach themselves to one another by their broken ends. Among the reconstituted chromosomes and fragments, those which include the spindle node continue to be distributed normally at mitosis, while those which have no spindle node are sooner or later lost. On this hypothesis the occurrence of segmental interchange is due to breaks in two chromosomes, followed.by an exchange of partners in the reattachment of the distal fragments. Similar interchange of two ends of the same chromosome produces an internal inversion. If the distal fragments exchange places not with each other but with the proximal fragments (transverse interchange), the result is the loss of both distal fragments and the formation of a new chromosome with two spindle nodes. (Such chromosomes are usually eliminated in the course of the first few cell generations, but occasionally are saved by the loss of a segment in-

I N T E R N A T I O N A L CONGRESS O F GENETICS

281

cluding one of the spindle nodes). An analogous interchange within the single chromosome will produce either an internal deficiency or a ring chromosome with both distal segments deficient, depending on the location of the breaks with reference to the spindle node. Displaced internal fragments tend to become inserted in an internal position, since the attachment of one end to a broken chromosome leaves the other end free for the attachment of another fragment. Finally, simple deficiency results from a break followed by the loss of the distal fragment. The hypothesis as stated is vague at one important point, the length of the possible interval between breakage and reattachment. As long as the broken ends retain the capacity for reattachment, translocations may occur freely after treatment, as fragments come in contact in the course of intracellular movement. But even if this property is a permanent characteristic of the broken ends, a large proportion of the induced translocations must occur before the next cell division, for the fragments without spindle nodes tend to be eliminated in mitosis. If the tendency to reattachment is assumed to be transitory the opportunities for deferred translocation are reduced accordingly. A t the extreme in this direction we may assume that translocations take place only when the chromosomes affected are in actual contact a t the points of breakage. This possibility is not positively excluded, but the evidence now available is more favorable t o the assumption of a mechanism permitting delayed attachment. Thus all of the known chromosomal derangements induced by irradiation may be viewed as secondary effects of one primary process, chromosome breakage. I t is possible that even the known effects of X-rays on the frequency of crossing over and of non-disjunction may be accounted for, at least in part, as consequences of induced translocation and deficiency. I n what ways would the mechanism postulated lead to changes inherited as mutations? The frequent occurrence of mutations a t points of chromosomal interchange suggests that deficiencies or other changes incident to chromosome breakage may be a frequent cause. But there must be many points of breakage a t which the two fragments broken apart reunite, instead of changing positions with other fragments. Thus the resulting type of mutation, though frequently associated with translocation, should also occur readily at other points. But this is not the only type of mutation that may occur a t a point of breakage, though probably much the most frequent type. Insertion of deleted segments may also occur, and the insertion of short segments, inherited as units because of the absence of crossing over, may simulate mutation also. Further, the extent to which gene displacement

282

PROCEEDINGS O F THE SIXTH

alone, without gain, loss, or transformation of genes, may cause modification of the phenotype is as yet quite unknown. Consequently the fact that some of the induced mutations behave in a manner inconsistent with a deficiency interpretation is not positive evidence that the mutations in general are intra-genic transformations. EXTRA-GENIC ALTERATIONS AS A C A U S E O F MUTATION

If the induced mutations are due to extra-genic alterations, a certain parallel might be expected in the relation of mutation and chromosomal derangement to factors affecting their frequency of occurrence. The correlation will not necessarily be close, for different types of chromosomal alteration may respond differently to the factors applied, and mutation may not correspond to all types of chromosomal change induced by irradiation. Particularly inconclusive are those comparisons which may introduce complications due to differential survival, for we may expect derangements of different types and degrees to differ widely in survival. Nevertheless, the parallel between induced mutations and induced chromosomal derangements is striking. The frequency of induced mutation is directly proportional to dosage (STADLER1928, 1930a). So are the frequency of deficiencies induced by treatment of pollen (STADLER1931) and the frequency of endosperm mosaics induced by treatment shortly after fertilization (GOODSELL 1930). The frequency of induced mutation is unaffected by temperature variations between wide limits (STADLER1931b). This is true also of the deficiencies involved in endosperm nlosaics (GOODSELL unpublished). These similarities may be merely the result of similar but independent response to factors which are without significant effect on the occurrence of either mutation or deficiency. A more convincing parallel is found in the relation to dormancy. Dormancy greatly reduces the frequency of induced mutatioil per unit of radiation intensity, but even in dormant cells mutations are induced at an appreciable rate. If induced mutations are due to chemical changes in gene composition, and induced chromosonlal derangements t o mechanical causes, we would not expect to find this peculiar relation of dormancy to mutation duplicated in the case of chromosomal changes. W e find, however, that the frequency of the chromosonlal derangements resulting in partial sterility is reduced by dormancy to about the same extent as is the frequency of mutation. Presumably, though the material has not been cytologically checked, the chromosomal derangements involved inclutie both

INTERNATIONAL CONGRESS O F GENETICS

283

translocations and deficiencies. Further, the frequency of deficiency alone, as measured by sectorial loss of dominant plant characters, apparently responds to dormancy in the same way. Whatever the cause of the peculiar relation between dormancy and mutation frequency, it apparently applies as well to induced chromosomal derangements. Deficiencies simulating mutation may be cytologically demonstrable in maize. Deficiencies for specific chromosome regions may readily be obtained by treatment of pollen carrying dominant genes as markers. By the observation of the location of the chromosome regions lost in such plants, the regional location of marker genes may be established. Among the genes which have been used extensively those most frequently lost are A, J and L,, which are located near the ends of three different chromosomes.Under heavy pollen treatment each of these genes is lost in about 2 percent of the progeny. In plants deficient for any one of these markers, the length of the deficient segment varies widely. Occasionally a plant is found which has lost so small a segment that the pollen receiving the deficient genome is partly developed. MCCLINTOCK(1931) has described a plant deficient for the gene Ls that had lost only a terminal segment of 4 chromomcres. The deficient pollen was partly developed and contained some starch. Plants that have lost the marker gene J are in some cases normally vigorous, and some of these have defective pollen fairly well developed. This was true of one case in which the deficient region comprised about one-fourth of the long arm of the chromosome. Among a large number of plants which had lost these markers, however, none with wholly normal pollen was found. Among about an equal number of plants which had lost the marker gene A, 3 were found with half of the pollen recognizably defective but approaching the normal in development, and 2 with pollen apparently fully normal. All five of the plants bore ears showing distinctly less than 50 percent sterility. If, as seems probable, the loss of A in these cases was due t o non-lethal or incompletely lethal deficiency, these plants are connecting links between induced deficiency and induced mutation. A deficiency of a segment including A, with no gametic elimination, would be inherited as a mutation of the dominant A to the recessive allelomorph. The two plants with apparently normal pollen may be of this class. An alternative interpretation is that the change involved in these cases was actually an intra-genic mutation of A. This seems improbable in view of the partial gametophyte development

284

PROCEEDINGS O F THE SIXTH

of plants deficient for fairly large, readily visible segments, and the series of intermediate grades of gametophyte development ranging to the apparently normal. A positive determination of deficiency may be possible in these cases by cytological examination in the meiotic prophase, although it is not certain that deficiencies of as little as one or two chromomeres can be cytologically detected in maize even at this favorable stage. The chief reason for questioning the mechanical interpretation of the induced mutations is the occurrence of reverse mutations in Drosophila. If the original mutation is due to a loss, how shall we account for the return of the lost gene in reverse mutation? Apparent reversion may also be found however in cases of typical genetic deficiency. In deficiencies in the endosperm, induced by irradiation of pollen, small areas of tissue showing the recovery of the lost dominant characters are not uncommonly found. If we assume that a mutation which later reverts can not possibly have been due to deficiency, we might with similar logic state that these losses of linked genes in endosperm tissue, since they revert, cannot be considered deficiencies. The notion that a variation which reverts cannot be due t o deficiency is based on the assumption that the loss involved in deficiency is brought about by some sort of instantaneous destruction or elimination of a chromosome or fragment. Apparently deficiency is not produced in this way. As I have previously pointed out (1930b, 1931a) any change which prevents the reproduction or mitotic distribution of the affected segment may have the effect of deficiency, for even if this segment were never eliminated it might be present in only one cell of the mature individual. The simplest hypothesis to account for recovery in the endosperm is that the chromosome segment affected is a fragment without a spindle node, which in most cases would be eliminated in an early mitosis, but which may occasionally escape elimination through a few cell generations and may then be restored to normal distribution by becoming attached to some chromosome with the normal spindle mechanism. Proof that fragments may thus persist and that translocation may be so long deferred is still lacking. Other chromosomal mechanisms resulting in frequent but not invariable elimination, such as certain types of transverse interchange, also may produce results simulating segmental loss followed by reversion. But, whatever the mechanism of recovery, the occurrence in the deficient endosperms of sectors of tissue showing the dominant characters proves that the treated gamete carries with it the chromosome segment which is later found to be deficient.

IPU'TERNATIONAL CONGRESS OF GENETICS

285

Since induced deficiency does not necessarily involve the immediate and complete loss of the genes affected, and since these genes are able to resume their normal activity later when suitable conditions are somehow restored, we cannot wholly exclude the possibility of a mechanical explanation for even those induced mutations which are known to he reversible. The reverse mutations described in Drosophila are not explainable on the basis of the tnechanisms mentioned above, hut it is not obvious that a tnechanical explanation of some kind is im~ossible.In this connection the important recent results of DORZHAXSKY (1932) with the induced mutant "baroid" should be cited. This mutation, apparently an allelomorph of Bar, arose in an irradiated fly at the point of breakage of a chrornosorne. Considered in connection with the induced mutations of Bar to wild-type reported by HANSON(1928), this might be considered a reverse mutation, and at a locus at which there is reason to suppose that the original mutation (Bar to wildtype) represented an actual loss. The analysis of such cases may discover a mechanical basis for the occurrence of reverse mutation. l'OLYPI.OI1)Y

I K RELATION T O M U T A T I O N

The interpretation of mutation in certain plant species is comp!icated by the possibility of polyploid origin. The identification of a Mendelizing variation as gene mutation or deficiency in plants is dependent on the assumption that deficiencies will have some lethal (or at any rate distinctly deleterious) effect on the gametophyte. In those species whose genomes include two or more distinct sets of chron~osomesthere is little ground for this assumption. Even though the different sets of chromosomes are derived from different species, and have undergone a long course of evolution since their union, they may still possess much in common. T o the extent that they carry duplicate genes they may be protected from the effects of recessive gene mutation (and to a lesser extent of deficiency), for the loss of a gene is not likely to be very injttrious in the presence of a duplicate gene. The extent to which this effect will apply is likely to vary greatly in different polyploid species, depending on the amount of differentiation that has occurred in the corresponding chromosomes. From the time the polyploid combination is established it is free to withstand the loss of duplicate chromosome regions, and such losses, if not unfavorable to survival, may become established in the evolution of the species. Eventually some species of polyploid chromosome number may lose almost all of their duplicate genes, while others may have a high proportion of genes duplicated. Since we have as yet no evidence indicating the extent of gene reduplica-

286

PROCEEDINGS O F THE SIXTH

tion in different polyploid species, it is well to be cautious in the interpretation of mutation evidence from any polyploid species. The characteristic effects which may be expected in polyploid species are: ( 1 ) A reduction in the apparent frequency of gene mutations, due to the presence of duplicate genes. For example, seedling mutations are common in Triticz~mmonococcum with 7 pairs of chromosomes, much less common in T . durum with 14 pairs, and rare in T. valgare with 21 pairs (STADLER 1929). (2) A reduction in the frequency of induced partial sterility, due to the survival of deficient gametophytes. According to unpublished data of W. R. TASCHER, of the UNIVERSITY OF MISSOURI, irradiation induces partial sterility with high frequency in T. monococcunz, but the frequency is much lower in T . duurum, and extremely low in T. vulgarc. ( 3 ) The appearance of variations due to the phenotypic effects of deficiencies and duplications which have been transmitted because of polyploidy. If deficiencies and translocations are induced in T. vztlgare as in T . monococcuwz, but sterility does not occur, we must expect types characterized by deficiencies and duplications in the progeny. Some of these may be inherited as if due to mutations, either dominant or recessive. This is the probable cause of some of the speltoid and dwarf types which are found in the progeny of treated plants of Triticunz vulgarc. (4) The appearance of apparent recessive mutations, due to the loss of segments bearing a dominant gene which is present in only one set of chromosomes. For example, if a variety of wheat is homozygous for a dominant, say beardless, in only one of its 3 sets of chro~nosomes,a nonlethal deficiency of this region will behave as a recessive bearded mutation. In another variety with the dominant gene present in 2 or 3 sets this mutation would not appear. This effect of gene reduplication provides a basis for "premutation" in the polyploid species. Typical Mendelian behavior therefore is not t o be considered entirely convincing evidence of the occurrence of gene mutation in the polyploid species. Mutations may occur, but there is no genetic method now available by which they may be distinguished from viable deficiencies. This effect is not limited to polyploid species, although they furnish the most convenient material for its demonstration. Gene reduplication probably has occurred in the evolution of other groups as a consequence of translocation or other chromosomal derangements. S o far as the reduplicated genes are concerned, analogous effects would occur in these species. Is maize a polyploid species? The indication that fairly extensive deficien-

I K T E K N A T I O N A L CONGRESS OF GENETICS

287

cies may be viable in maize suggests the possibility that this species also may be polyploid. The genetic and cytological conditions found in maize have not been such as to suggest polyploid origin. There is no unexpectedly high frequency of duplicate factors; the 10 chromosomes are all morphologically distinguishable; chromosome pairing is normal in the diploid, and the chron~osomesof haploids show no distinct tendency to pair. The chromosonle number 10 and its multiples are common in species of the Maydeae and of the related tribe Androfiogoneae, and (so far as basic chroinosome numbers may be determined by comparative counts) it has seemed reasonable to regard 10 as the basic number for this part of the grass family. However, in the course of his extensive investigations of the chromosomes of various species of the Andropogoneae, BREMER(1925) found one species, Sacclzarum (Erianthus?) narenga, with 15 pairs of chromosomes, suggesting the possibility of an ultimate unit of five chron~osoinesin the ancestry of this tribe. Recently LONGLEY has reported a 5-chromosome informs species of sorghum, Sorghunz versicolor, and P. C. MANGELSDORF me that he has found Coix aquatics, an Asiatic species of the Maydeae, to have 5 chromosomes also. I t therefore seems probable that the 10 chromosomes of maize are derived ultimately from five, though the subsequent differentiation may have been extreme. If maize is polyploid we may expect that some deficiencies may not be lethal to the gametophyte because of chromosomal reduplication, and that apparent mutations may sometimes occur as a result of the occurrence of a viable deficiency including a dominant gene. Consequently even though it may be possible to demonstrate cytologically that the radiation-induced mutations are accompanied by deficiency in maize, there is no necessary implication that this is generally true of indrlced mutations in species which are not polyploid. I t is possible that a certain disparity in the effects of translocation in Drosophila and maize may be due to gene reduplication connected with polyploidy in maize. In Drosophila few of the induced translocations can be made homozygous, since most of them are accompanied by lethal mutations at the point of breakage of the chromosomes. On the contrary most of the induced translocations in maize may be made homozygous without difficulty. If translocation in maize were accompanied by lethal mutation, we should expect the translocations to be eliminated in the first gametophyte generation. The fact that they are not so eliminated and that they may be established in homozygous form may be due to the suppressing action o f duplicate genes.

288

PROCEEDINGS OF THE SIXTH NATURAL MUTATION

The foregoing considerations indicate that the mutations induced in plants by X-rays may be of a special type not representative of typical gene mutation. The most direct way to determine whether this is the case is simply to determine whether the treatment affects the frequency of various typical gene mutations. But what is a typical mutation? W e do not know enough about the mutation of specific genes to be able to select the typical. Practically all of the available evidence on the mutation rate of specific genes applies to genes which were selected for study because of their known high rates of mutation. If mutation is a compound class of diverse phenomena, there is little likelihood that these selected cases are a representative sample. Even the use of genes arising as mutants in culture is somewhat selective, for these will tend to be the more mutable genes. This difficulty can be in part avoided in the cultivated plants by using the characters distinguishing the established races, since these are a sample, almost random as to mutability, of the genes appearing over a much longer period. The determination of specific mutation rates in unselected genes is feasible in maize, at least in the case of genes for endosperm characters. By a simple technique it is possible to determine the frequency of mutation in several genes simultaneously. Since each seed tests one female gamete for mutation it is practicable to determine mutation rates based on many thousands or even millions of gametes. During the last several years we have accumulated data on the normal frequency of recessive mutation of eight genes determining endosperm characters in maize. The genes used are entirely unselected except in their one common quality of affecting the endosperm. None of these genes had previously been known to mutate. Each is entirely regular in genetic behavior. All but one of the eight genes tested yielded recurrent mutations. The frequency of these mutations varied for different genes from approximately 1 to 500 per million gametes. Different families varied rather widely in frequency of mutation of the same gene. The most mutable gene, R, mutated in the most frequently mutating family a t a rate of more than 0.1 percent. The mutations of these genes have no lethal effect on the gametophyte. Among more than one hundred mutant plants examined, none was affected by genetic partial sterility. Closely linked genes do not mutate together. Even in cases which may be regarded alternatively as due to multiple allelomorphs or to completely linked genes, the natural mutations

INTERNATIONAL CONGRESS O F GENETICS

289

seem to affect one unit independently of the others. For example, in the multiple allelomorphic series Rr-R"rr-rg, affecting aleurone and anther color (which may also be represented as being due to two completely linked genes, R for aleurone color and N for anther color) the mutation affecting aleurone color is not accompanied by a mutation affecting anther color. In other words R r mutates frequently to r', or R N to r N , but R r never mutates to @, or R N to r n, If the series is regarded as due t o multiple allelomorphs, mutation of the extreme dominant is always to an intermediate member; if it is regarded as due to combinations of completely linked genes, the mutation which affects one gene never affects the other with it. Deficiency of varying extent, simulating mutation, would be expected usually to simulate mutation to the ultimate recessive. Several interesting questions regarding the nature of the induced mutations may be investigated by determining the effects of irradiation on the mutation of these genes. Are all of the genes caused t o mutate more frequently by the treatment, or is the increase in the general rate of mutation due to a relatively great increase in the mutation rate of some genes with no effect on that of others? In the case of genes affected by irradiation, are the mutation rates increased in more or less uniform proportion, as might be 'expected if the normal processes of mutation are somehow catalyzed by the treatment? O r do induced mutations of all of the genes occur at similar rates, regardless of their normal frequency, as might be expected if the induced mutations are due to mechanical changes induced a t random in the treated chromosomes? Are the induced mutations of Rr, like the natural mutations, limited to one of the component genes or genefractions, or do they, like deficiencies, usually affect both components together? These and many similar questions require evidence from the mutation of specific genes; they are untouched by experiments on the general rate of mutation. I n spite of the technical difficulties, an attempt was made therefore to determine the effects of X-ray treatment on the frequency of mutation of the unselected genes whose normal mutation rates had been determined. EFFECTS O F IRRADIATION O N THE MUTATION O F UNSELECTED GENES

In order to determine whether the normal mutation rates of the unselected genes are modified by X-ray treatment, it is necessary t o apply the treatment during the early development of the ears of the plants in which mutation is to be determined. This seriously limits the amount of material which can be handled, but during the past three seasons we have secured

290

PROCEEDINGS O F THE S I X T H

evidence regarding mutation of seven endosperm genes in treated populations of from 40,000 to 80,000 each. The populations are large enough to yield several mutations of the more mutable genes R and I in untreated material of the families used. They are too small to yield more than a n occasional mutation of the other genes, unless the mutation rates are much increased by irradiation. If the mutation rate for all genes were increased as much as fiftyfold, numerous instances of mutation of all seven of the genes should be found. The treatments were applied in doses sufficient to induce seedling mutations with fair frequency. In the selfed progeny of the plants thus treated several seedling mutations have been found, indicating that the X-ray treatments used produced the characteristic mutational effect. The extent to which the spontaneous mutation frequency was increased is unknown. Mutation of the endosperm genes under test occurred at about the same rate in the treated ears as in the untreated check. In the case of the more mutable genes R and I a considerable number of mutations was obtained in both the treated and untreated groups. Not only was the mutation rate not greatly multiplied by the treatment but it was not even significantly increased. The genes Pr and Y yielded only a few mutations scattered at random among treated and untreated ears, with no significant increase following treatment. S h and SUyielded no mutations. From the natural frequency previously determined the probability of finding mutations of these genes may be computed for different rates of increase in mutation rate. The chances are even that at least one mutation of each of these genes would have been found in the irradiated material if the treatment had resulted in a tenfold increase in their natural mutation rates. For the seventh gene ( W z ) a different result was obtained. Previous trials in untreated material had failed to discover a single mutation among more than 1,500,000 gametes tested. Among 50,000 irradiated gametes tested, two clear mutations of waxy were found. The mutant waxy gene was found to have no detectable injurious effect on the male or female gametophyte and was indistinguishable from the standard waxy gene in its phenotypic effect in endosperm and pollen. Although the previous tests of untreated material were made with various families, the stock used in the radiation experiment has been tested quite extensively for natural mutation of waxy. It has yielded no spontaneous mutations in more than 100,000 tested gametes. LITaxytherefore may be a locus susceptible to induced mutation. If the 7 genes used map be regarded as a representative sample of the

INTERNATIONAL CONGRESS O F GENETICS

29 I

genes of maize they indicate that the specific mutations appreciably affected by radiation may be a rather small fraction of the whole. I t would be hard to prove that the 7 genes tested are a perfectly random sample, and at best they would constitute an inadequate sample inadequately tested. But the results of this necessarily small experiment lend no support to the assumption that mutation in general is "speeded up" by irradiation. Induced mutation does not differ from natural mutation merely in its greatly increased frequency of occurrence; apparently there is a qualitative as well as a quantitative difference. DISCUSSION

The mutations induced by X-rays are typical gene mutations, apparently identical in genetic behavior, and in many instances in phenotypic effect, with well-known mutations of spontaneous origin. Are the germinal changes involved in these mutations representative of those involved in the natural evolution of the gene? If this is true, we may substitute the easily obtained induced mutations for the much rarer natural mutations, and we may reasonably hope for rapid progress in the solution of fundamental problems of gene structure and behavior which have previously, because of the rarity of mutation, been beyond the reach of experimental investigation. But an affirmative answer to this question requires something more than the demonstration that mutations are induced by the treatment. Mutation is not a single homogeneous class of germinal variations. I t is a residual class, in which are included various types of germinal change whose physical nature is undetermined. These may include extra-genic as well as intragenic variations. A treatment affecting only some particular type of mutation may greatly increase the total yield of mutations without necessarily affecting any process involved in the normal evolution of the gene. The induced mutations also may include various types of germinal change, for there are several conceivable ways in which irradiation may so modify a chromosome as to cause the appearance of a Mendelizing variation. The available evidence does not yet provide an adequate basis for final.conclusions as to their physical nature, but it is probable, in view of the considerations discussed in the preceding pages, that most of the induced mutations in plants are due to various extra-genic alterations, chiefly non-lethal deficiencies. This is true in spite of the fact that the variations recorded as mutations in the experiments with plants are exclusively the "visibles," the "lethals" being automatically excluded by the interposition of the haploid gametophyte generation. The possibility that chemical transformations of the genes also are in-

292

PROCEEDINGS OF THE SIXTH

duced by the treatment cannot be excluded, for various chemical reactions are known to be energized by high frequency radiation. The difficulty here is that the assumption of a chemical change can not be directly tested. All germinal variations induced by X-ray treatment could be ascribed to chemical changes in the genes, by analogy with the known chemical effects of X-rays. Since the assumed chemical change in the hypothetical gene could not possibly be disproved, the only induced variations which could be removed from this class would be those for which some other mechanism could be demonstrated. It is hazardous to assume that the known mechanical causes of germinal variation are operative only in the cases in which they are specifically demonstrated, and the hypothetical chemical causes are responsible for all other cases-particularly since the genetic and cytologi? cal criteria of deficiency approach the limit of their range of effective application as the segments involved become small enough for their effects to be transmitted as mutations. Nevertheless the occurrence of reverse mutations in Drosophila is a strong argument for the inclusion of chemical transformations of the gene among the mutations induced by X-rays. Mutation to different allelomorphs of the same series also adds weight to this argument, though our knowledge of the nature of multiple allelomorphism is not sufficiently definite to exclude alternative interpretations. These cases considered alone might plausibly be assumed to be the result of chemical, or at any rate intragenic, mutations. However, in view of the indications of a mechanical basis for the greater portion of the induced mutations, the possibility of some mechanical explanation of even the reverse mutations should be considered. The frequent occurrence in Drosophila of mutations at the points of breakage of the chromosomes suggests that the point mutations in this species as well as in plants may be largely of mechanical origin. This association of induced mutation with chromosome breakage does not necessarily exclude the possibility that the mutations are intra-genic changes, for it is conceivable that some change within the gene may be the cause of the break, and that in some instances the gene may continue to function as a gene after the change has occurred. But the possibility that some mutant characters may be due merely to the changed spatial relations of genes to one another ("position effect") cannot be disregarded. Moreover, as has been shown in the text, the occurrence of reversion is not proof that the original mutation could not have been due even to deficiency. Although the possibility that the induced mutations may include instances of intra-genic transformation must be left open, the identification of the

I N T E R N A T I O N A L CONGRESS O F GENETICS

293

induced mutations in general as a class distinct from the chromosomal abnormalities is not possible in the experiments with plants. If the induced mutations are wholly or largely by-products of the induced chromosomal variations, other treatments which have similar effects on the chromosomes may yield mutations of the same type. SPRAGUE (1930) found that exposure of pollen to an electromagnetic field caused an increased frequency of translocations and endosperm mosaics in maize. I t is interesting to note that in the selfed progeny of these plants "a few recessive mutations were observed." RANDOLPH(1932) found various chromosomal aberrations to be induced by high temperature treatment of ears, including chromosomal deficiencies and translocations. N o seedling mutations were found among 113 selfed progenies examined, but a inore extensive test would be required to detect an effect on mutation rate comparable to that of an equivalent X-ray treatment, if equivalence is based on the frequency of induced translocation and deficiency.' Finally, the basic question, whether the phenomenon of induced mutation is representative of the phenomenon of natural mutation, can be answered only by the determination of the effects of the treatment on specific genes unselected as to mutability. The experiments in which the effect of irradiation is determined from its influence on the general rate of mutation automatically eliminate from consideration the genes which are unaffected by the treatment. The results of these experiments may be generalized only on the assumption that mutations in general are a homogeneous class, of which the induced mutations are a representative sample. Unfortunately, the experimental test of the validity of this assumption involves technical difficulties, for it requires the determination of specific mutation rates which may be very low. An attempt in this direction is reported in the present paper. The results do not support the assumption that mutation in general is affected by irradiation. LITERATURE CITED

BREMER, G., 1925 T h e cytology of the sugar cane. Third contribution. Genetica 7293-322. BURNHAM, C. R., 1932 An interchange in maize giving low sterility and chain configurations. Proc. Nat. Acad. Sci. Washington 18:434-440. DOBZHANSKY, TH., 1932 The baroid mutation in Drosophila melanogaster. Genetics 17:369392. (Note added in proof.) In this connection a report just published by BEADLE(2. indukt. Abstamm.-u. VererbLehre. 63:195-217) is of interest. A recessive gene causing abnormal chromosome behavior in maize was found to increase markedly the frequency of translocation and non-disjunction. In the F2 of crosses of this strain with normal maize, 6 recessive seedling mutations were found among 198 progenies tested.

294

PROCEEIIINGS OF T H E S I X T H

GOODSELL., S. F., 1930 The relation between X-ray intensity and the frequency of deficiency in the maize endosperm. (Abst.) Anat. Rec. 47:381-382. HANSON,F. B., 1928 The effect of X-rays in producing return gene mutations. Science 67 :562-563. LONGLEY, A. E., 1932 Chromosomes in grass sorghums. J. Agric. Res. 44:317-321. MCCLINTOCK,BARBARA,1931 Cytological observations of deficiencies involving known genes, translocations and an inversion in Zea mays. Missouri Agric. Expt. Sta. Res. Bull. 163:l-30. MULLER,H. J., 1928 The production of mutations by X-rays. Proc. Nat. Acad. Sci. Washington 14:714-726. PATTERSON, J. T., and MULLER,H. J., 1930 Are "progressive" mutations produced by X-rays? Genetics 15:495-578. RANDOLPH, L. F., 1932 Some effects of high temperature on polyploidy and other variations in maize. Proc. Nat. Acad. Sci. Washington 18:222-229. SEREBROVSKY, A. S., 1929 A general scheme for the origin of mutations. Amer. Nat. 63 ~374-378. SPRAGUE,G. F., 1930 Some genetic effects of electromagnetic treatments in maize. (Abst.) Anat. Rec. 47:382. STADLER, L. J., 1928 T h e rate of induced mutation in relation to dormancy, temperature, and dosage. (Abst.) Anat. Rec. 41 :97. 1929 Chromosome fiumber and the mutation rate in Avena and Triticum. Proc. Nat. Acad. Sci. Washington 152376-881. 1930a Some genetic effects o i X-rays in plants. J. Hered. 21:3-19. 1930b Recovery following genetic deficiency in maize. Proc. Kat. Acad. Sci. Washington 162714-720. 1931a The experimental modification of heredity in crop plants. I. Induced chromosomal irregularities. Sci. Agric. 11:557-572. 1931b The experimental modification of heredity in crop plants. 11. Induced mutation. Sci. Agric. 11 3645-661. TIMOF~EFF-RESSOVSKY, N. W., 1930 Das Genovariieren in verschiedenen Richtungen bei 1)rosophila melanogaster unter dem Einfluss der Rontgenbestrahlung. Die Naturwissenschaften 18:434-437.

INTERNATIONAL CONGRESS O F GENETICS

295

N E U E R E ERGEBNISSE uBER D I E GENETIK U N D ZYTOLOGIE D E S CROSSING O V E R Curt Stern, Kaiser Wilhelm-Institut fiir Biologie, Berlin-Dahlem, Germany

Die Untersuchungen uber Faktorenaustausch oder crossing over nehmen eine besondere Stellung im Rahmen der Genetik und Zytologie ein. Die urspriinglichen Untersuchungen iiber den Zusammenhang von Genetik und Zytologie, die von SUTTONund BOVERIinauguriert wurden, zeigten zwar, dass Chromosomen und Gene zusammengehijren und vereinigten zwei verschiedene Gruppen von Erkenntnissen. Aber diese neue Erkenntnis, so fruchtbar sie auch war als eine Quelle des Verstandnisses fur das Verhalten der Chromosomen und der an sie gebundenen Gene, brachte uns doch nur wenig neue Einblicke in das Verhalten der Chromosomen oder der Gene. Anders bei den Untersuchungen iiber Faktorenaustausch. Hier diente die genetische Analyse der Koppelungserscheinungen von vornherein als ein Mittel, unsere Kenntnisse iiber das Wesen und das Verhalten der Chromosomen zu vertiefen. Der Titel der Arbeit von MORGAN(1911), welche den Beginn dieses Abschnittes der ,Forschung bildete, hiess: "An attempt to analyze the constitution of the chromosomes on the basis of sex-limited inheritance in Drosophila." Und man kann ohne vie1 Uebertreibung sagen, dass der Hauptteil der genetischen Arbeit, die in den letzen zwanzig Jahren auf dem Gebiet der Faktorenkoppelung geleistet worden ist, weniger Genetik als Zytologie gewesen ist: Eine Erforschung des Wesens und Verhaltens der wichtigsten biologischen Elementargebilde, iiber die wir etwas Sicheres wissen, der Chromosomen. Was das Wesen der Chromosomen anbetrifft, so wissen wir, dass bestimmte Chromosomen Trager bestimmter Gruppen gekoppelter Gene darstellen, dass verschiedene Teile eines Chromosoms verschiedene Gene enthalten, und dass die Gene eine seriale Anordnung aufweisen. Im Besonderen auch diese letzten Erkenntnisse, die urspriinglich als kiihne Hypothesen erschienen, konnen jetzt nach den1 Ausfall entscheidender genetisch-zytologischer Untersuchungen mit Cliromosomenfragmenten als vollig gesichert gelten (Drosophila: STERN, MULLER und PAINTER, ; Datura : BLAKESLEEund BELLING ; Mais : MCCLINTOCK) . DOBZHANSKY Da die Gene, zumindest bei den genetisch bearbeiteten Organismen, nicht optisch beobachtbar waren, so ist es klar, dass die genetische Analyse der Chromosomen bezuglich der Anordnung der Gene bisher die einzig mogliche war. Es wurden Strukturen analysiert, deren direkte Beobachtung nicht moglich war. Der mikroskopische Teil der Analyse bestand dabei darin,

296

PROCEEDINGS OF THE SIXTH

grosse mikroskopisch sichtbare Teile der Chromosomen als Reprasentanten gewisser Gene zu verfolgen. Dagegen gab es zwei verschiedene Methoden der Analyse des Verhaltens der Chromosomen, die normale direkt zytologische Methode der mikroskopischen Beobachtung und die indirekte Methode mittels des Kreuzungsversuches. Diese indirekte Methode kann man charakterisieren, indem man sagt, dass das Verhalten der mikroskopisch sichtbaren Gebilde, der Chromosomen, mithilfe des Verhaltens mikroskopisch nicht sichtbarer Reprasentanten, der Gene, verfolgt wurde. Im Folgenden sol1 versucht werden, die grundlegenden Ergebnisse der Erforschung des Chromosomenverhaltens beim crossing over zu schildern. Dabei werden wir im allgemeinen nur die Ergebnisse der indirekten Methode behandeln, da die direkte Methode ja in dem Vortrage von S A Xausfuhrlich erijrtert worden ist. Die Vorstellung, die allen Arbeiten zugrunde lag, ist die von JANSSENS und MORGAN begrundete Theorie, nach der die homologen Partner eines Chromosomenpaares im Verlaufe der Keimzellengenese homologe Abschnitte miteinander austauschen. Die direkt zytologische Erforschung des Chromosomenverhaltens hat bis vor kurzem kein endgultiges Urteil uber die Richtigkeit der Theorie sind von fuhrenden beibringen k ~ n n e n .Die Interpretationen von JANSSENS Zytologen nicht als zwingend anerkannt worden. Und urn die geistreichen liegen, Deduktionen von BELLINGund DARLINGTON, die im Sinne JANSSENS wird ja jetzt gerade gekampft. Eines scheint jedoch festzustehen: Sowohl BELLINGund DARLINGTON als auch SAX glauben auf Grund mikroskopischer Untersuchungen ( 1) dass die homologen Chromosomen wahrend der Prophase der Reifeteilungen regelmassig Stucke austauschen; (2) dass dieser Austausch eine Paarung der homologen Chromosomen voraussetzt ; ( 3 ) dass die Chiasmata gepaarter Chromosomen eine wesentliche Rolle bei wahrend des dem Austauschvorgang spielen; (4) dass die Chroi~~oson~en Austauschs bereits in Tochterchromatiden gespalten sind, wobei nur zwei von den vier Chromatiden einer diploiden Gruppe in einem bestimmten Chiasma am Austausch beteiligt sind ("partielle Chiasmatypie"). Die Beweiskraft dieser Uebereinstimmung ist allerdings geringer, als es zuerst erscheinen mag, da die verschiedenen Forscher verschiedene und sich teilweise widersprechende Interpretationen zur Stutze ahnlicher Aussagen anfiihren. Doch kann man als vorsichtiger Beurteiler wohl feststellen, dass die mikroskopischen Befunde den angefuhrten Satzen nicht entgegenstehen. Wir werden ferner sehen, dass, wenn auch das "Wie" des Chromosomenstuck-Austauschs noch nicht vollig geklart ist, doch die Tatsache des Aus-

INTERNATIONAL CONGRESS OF GENETICS

297

tauschs jetzt seit einem Jahre feststeht ulid die Gultigkeit der oben unter ( 1 ) und (4) angefuhrten Satze gesichert ist. Da diese Erkenntnisse durch eine Kombination der direkten mikroskopisch-zytologischen mit der indirekten genetischen Methode erlangt wurden, so sollen sie erst spater behandelt werden. Wir wenden uns daher jetzt der genetischen, indirekten Methode zum Studium des Chromosomenverhaltens zu. Sie beruht auf der Tatsache, dass aus zwei Gruppen homologer gekoppelter Gene A B C D E und a b c d e Koppelungsgruppen mit Neukombinationen hervorgehen konnen, etwa A B c d e und a b C D E. MORGAN interpretierte diese Tatsache dahin, dass ein Austausch der A B und a b enthaltenden Abschnitte eines homologen Chromosomenpaares stattgefunden habe. Nimmt man diese Interpretation an, wie das bis vor kurzem allgemein geschah, so gibt uns die genauere Analyse der Genneukombinationen die indirekte genetische Methode zur Aufklarung des Chromosomenverhaltens beim Austausch. Das erste Ergebnis dieser Analyse ist die Bestimmung des Zeitpunktes des Austauschs. Hier zeigte ALTENBURG (1916) an Primula, dass der Austausch sicherlich spater erfolgt, als die Determinierung der Antheren oder Karpelle einer Bliite, da die verschiedenen Pollenkorner und Samenanlagen je eines Staub-oder Fruchtblattes noch sowohl die Nichtaustauschals auch die Austauschkombinationen gekoppelter Gene aufweisen. PLOUGH (1917) fand dann bei Drosophila, dass eine Beeinflussung der Austauschhaufigkeit erst an denjenigen Eiern zu beobachten ist, die bei einer Behandlung von Weibchen mit besonderen Temperaturen etwa sechs bis sieben Tage nach Einsetzen der Behandlung abgelegt werden. Daraus ist zu schliessen, dass der Austausch nicht fruher als sechs bis sieben Tage vor der Ablage eines Eies erfolgt. PLOUGHfand, dass sich die Eier sechs bis sieben Tage vor ihrer Ablage im spaten Oogonien-oder friihen Oocytenstadium befinden. Austausch erfolgt also nicht fruher als auf diesem Stadium. Zu dem gleichen Ergebnis fuhrten ganz andersartige Ueberlegungen und Versuche GOWENS( 1929), ebenfalls an Drosophila. Wenn Austausch auf m gar einem Urkeimzellenstadium ereinem friihen ~ o ~ o n i e n s t a d i u oder folgte, so sagte er sich, dann miissen wegen der spateren Teilungen dieser Keimzellen stets mehrere Eier gleicher und reziproker Austauschkombinationen entstehen. Da aber seine Versuche mit seltenen Austauschkombinationen zeigten, dass solche Kombinationen stets einzeln auftreten, so muss man schliessen, dass der Austausch nach dem letzten oogonialen Teilungsschritt erfolgt. Dasselbe geht aus Angaben von STURTEVANT (1928) hervor. Diese Bestimmungen des Zeitpunktes des crossing over konnten noch

PROCEEDINGS O F T H E S I X T H

298

weiter eingeengt werden. Da in Fallen von non-disjunction bei heterozygoter Konstitution des P-Individuums Gameten entstehen konnen, die ein NichtAustauschchromosom und ein homologes Austauschchromosom enthalten, so folgerten BRIDGESund MULLERbereits 1916, dass bei Drosophila der Austausch auf dem Stadium der gespaltenen Chromosomen und nur zwischen zwei Spalthalften erfolgt. Denn wenn die urspriingliche Konstitution etwa

ABCDE abcde

war und die non-disjunction-Gamete etwa

abcde abCDE

enthielt, so ist dies nach der crossing over Vorstellung nur moglich, wenn die urspriinglichen Chromosomen sich gespalten haben: A B C D E, A B C D E, a b c d e, a b c d e, und dann Austausch zwischen nur zwei Tochterchromatiden erfolgte: A B C D E, a b C D E, A B c d e, a b c d e, sodass dann bei non-disjunction: a b C D E und a b c d e gemeinsam in einen Gameten gelangten. Dieser Schluss von BRIDGESund MULLERist dann in ausgedehnten ahnlichen Versuchen von BRIDGESund ANDERSON ( 1925), ANDERSON (1925), L.V. MORGAN(1 925), STURTEVANT (1931 ) fur das X Chromosom, sowie von REDFIELD(1930, 1932) fur die Autosomen I1 und I11 bei Drosophila gesichert worden. Ganz neuerdings konnte crossing over auf dem Doppelstrangstadium auch fur Zea mays festgestellt werden (RHOADES1932 [Proc. Nut. Acad. Sci. Washington]). Dagegen scheinen die Versuche von F. V. WETT~TEIN (1924) an dem Laubmoos Funaria fur Austausch auf dem Stadium der ungespaltenen Strange zu sprechen; doch ist die Natur der Koppelung bei diesem Organismus noch nicht geklart. Schliesslich haben verschiedene genetische Methoden uns etwas iiber den Zusammenhang zwischen crossing over und Verteilung der homologen Chromosomen bei der Reduktionsteilung gelehrt. GOWEN(1922, 1928) entdeckte eine Rasse bei Drosophila, bei der kein Austausch erfolgte und eine ungeregelte Verteilung der Chromosomen bei den Reifeteilungen auftrat, und ANDERSON (1929) stellte ebenfalls bei Drosophila fest, dass nondisjunction der X Chromosomen mit Ausbleiben von crossing over korreliert war. (Vgl. auch DOBZHANSKY 1932.) Man k ~ n n t edenken, dass diese Korrelationen auf eine gemeinsame Ursache zuriickgefiihrt werden kijnnen, namlich auf ein Ausbleiben der Konjugation der Chromosomen, deren Ergebnis sowohl ein Ausbleiben von crossing over als auch eine ungeregelte, zu non-disjunction fiihrende Verteilung der Chromosomen sein wiirde. Doch werden wir sehen, dass dies nicht notwendigerweise der Fall ist. (1931) es durch eine interessante Analyse Dagegen hat DOBZHANSKY wahrscheinlich gemacht, dass die Verminderung des crossing over Prozent-

INTERNATIONAL CONGRESS O F GENETICS

299

satzes bei Translokationen auf einer Erschwerung - der Paarung - der Chromosomen beruht. Wenn wir z.B. zwei Chromosomen A A und zwei Chromosomen B B verfolgen, wobei ein Teil des einen Chromosoms A an ein Chromosom B verlagert ist, so beobachtete man eine Herabsetzung der crossing over Haufigkeit und zwar nicht nur zwischen dem verlagerten Teil (AS) und seinem Homolog, sondern auch zwischen dem nicht-verlagerten Teil (A2) und seinem Homolog oder zwischen B1 und B2. Dies ist verstandlich, wenn man annimmt, dass die Paarung der Chromosomen eine Vorbedingung fur Faktorenaustausch ist, da offenbar die Paarungsmoglichz.B. wird sowohl von keiten wegen teilweise widerstrebender Krafte-A'

A3 als auch von A2 "angezogen"-gehindert ist. Und eine Stutze fur diese Erklarung ist es, dass die AustauschGufigkeit zwischen B1 und B2 und A3 und dem Homolog erhoht ist, wenn durch Einfiihrung besonderer Verhaltnisse eine Paarung zwischen A 2 und seinem Homolog verhindert wird. Denn dann sind weniger widerstrebende Krafte im Spiel. So hat uns die indirekte genetische Methode eine Anzahl von Ergebnissen gebracht, iiber die auch die zytologische Methode zu ahnlichen, wenn auch weniger sicheren Schliissen gekommen ist: Zeitpunkt des crossing over wahrend der Gametogenese; Crossing over auf dem Doppelstrangstadium zwischen zwei Chromatiden (partielle Chiasmatypie) ; Zusammenhang zwischen crossing over und Chromosomenpaarung. Ueber diesen letzten Punkt hat uns weiterhin eine gleichzeitig mit der indirekt genetischen und der zytologischen Methode arbeitende Untersuchung von BEADLE(Genetics, 1932) Aufklkung gebracht. EMERSON und BEADLEhaben auf genetischem Wege gefunden, dass

300

PROCEEDINGS OF THE SIXTH

zwischen bestimmten Abschnitten von Chromosomen der verwandten Spezies Euchlaena und Zea in den Bastarden kein oder hochst selten crossing over erfolgt. BEADLEhat nun durch zytologische Beobachtung festgestellt, dass zwischen den betreffenden Chromosomenabschnitten zwar eine urspriingliche Paarung stattfindet, die aber nach dem Diplotanstadium nicht mehr aufrecht erhalten wird. Die- Ergebnisse von GOWENund ANDERSON iiber crossing over und non-disjunction, die wir oben betrachteten, entsprechen wahrscheinlich diesen Befunden BEADLES. Die genetische Methode der Untersuchung des Chroinosomenverhaltens hat uns ferner gezeigt, dass ein Chromosom an verschiedenen Stellen Austauschbriiche zeigen kann und hat uns bei Drosophila die Haufigkeit der einfachen, doppelten und weiteren mehrfachen Briiche gelehrt. Dabei wurde gefunden, dass eine Interferenzwirkung besteht, wobei ein bestimmter Bruch im allgemeinen weitere Briiche in seiner Umgebung zu verhindern r die Hiufigkeit mehrfacher Briiche geringer, als nach einem strebt. ~ i h e ist zufallsgemassen Zusammentreffen der einzelnen Briiche zu erwarten ware. Man hat daher natiirlich versucht, Beziehungen zwischen der Zahl und dem Auftreten der Chiasmata und der Zahl der genetisch bestimmten Chromosomenbruchstellen festzustellen. Doch kommt man, je nach den verschiedenen I'heorien iiber die Bedeutung der Chiasmata, zu verschiedenen Schliissen. BEADLESDaten scheinen mehr im Sinne von BELLINGand DARLINGTON als fur SAX zu sprechen. Schliesslich sei noch auf die Unterverwiesen, der an Hand von anderen Forschern beigesuchung HALDANES brachten zytologischen Materials zeigte, dass die Zahl der einfachen und mehrfachen Chiasmata eines Chromosomenpaares eine Interferenzwirkung ja die crossing over der Chiasmata aufeinander erkennen last,-wie Briiche eine Interferenzwirkung aufeinander ausiiben. So sehen wir eine weitgehende Uebereinstimmung zwischen den Ergebnissen der direkten zytologischen Methode und der indirekten genetischen Methode. Dennoch bliebe die Lage unbefriedigend, wie sie bisher geschildert wurde. Denn die gesamte genetische Analyse beruht auf der Annahme, dass crossing over zwischen Faktoren tatsachlich durch Austausch von Chromosomenteilen erfolgt. So wahrscheinlich auch diese Annahme war, so war sie doch letzten Endes noch nicht bewiesen. Und WINKLER(1930) hatte ausserdem gezeigt, dass sie nicht die einzig mogliche Annahme war. Auch die direkte zytologische Untersuchung der vermutlichen Austauschstadien hatte ja bisher nicht zu vollig eindeutigen Ergebnissen gefiihrt, wie wir sahen. Wenn aber auch die vermutlichen Austauschstadien der Beobachtung nicht zuganglich waren, so ergaben sich doch vor kurzem Moglichkei-

INTERNATIONAL CONGRESS O F GENETICS

301

ten, die Tatsache des Austauschs von Chromosomenstucken beim Faktorenaustausch mit Sicherheit aus dern zytologischen Ergebnis des Austauschs zu erschliessen. Wenn zwei homologe Chromosomen Stucke miteinander ausgetauscht haben, so ist dies gewohnlich spater nicht zu sehen. Wenn aber die homologen Chromosomen an zwei verschiedenen Stellen Verschiedenheiten voneinander aufweisen, so muss sich das Ergebnis von stattgefundenem Austausch sichtbarlich zeigen. Solche Falle doppelt-heteromorpher Chromosomen sind nun aufgefunden worden. Beim Mais haben CREIGHTON und MCCLINTOCK (1931) ein Chromosomenpaar verfolgt, bei dern das eine Ende den Unterschied verdickt-unverdickt, das andere Ende den Unterschied verlangert-unverlangert zeigte. Unter den Nachkommen solcher Individuen gelang es ihnen, alle vier Typen von Chromosomen aufzufinden, namlich die Nichtaustauschtypen "verdickt, verlangert" und "unverdickt, unverlangert," sowie die durch Austausch von Chromosomenstucken entstandenen neuen Typen "verdickt, unverlangert" und "unverdickt, verlangert." Die Zahlen, durch welche dieser Austausch bewiesen wurde, sind recht erheblich ; die Individuenzahlen. der vier Klassen waren : 47: 52:37 :28. (Mit Hilfe dieser doppelt heteromorphen Chromosomen haben die genannten Autoren auch das Vorkommen von Austausch auf dern Doppelstrang-Stadium zytologisch nachweisen konnen [vgl. Band 2 dieser Verhandlungen] ). Ganz entsprechende Ergebnisse sind bei Drosophila erzielt worden. Hier dienten zwei heteromorphe X Chromosomen dern Nachweis von Chromosomenstuckaustausch: An dern einen Ende des X Chromosoms war entweder ein grosses sichtbares Fragment des Y Chromosoms angeheftet oder es fehlte, und an dern anderen Ende war entweder die Halfte des X Chromosoms abgebrochen, oder sie war vorhanden. Und in ausgedehnten Versuchen wurden unter den Nachkommen eines Weibchens, welches beide X Chromosomen besass, alle X mit Y-Arm, nicht vier Typen, namlich zwei Nichta~~stauschchromosomen fragmentiert, und X ohne Y-Arm, fragmentiert sowie die zwei Austauschchromosomen: X mit Y-Arm, fragmentiert, und X ohne Y-Arm, nicht fragmentiert, gefunden. Damit war also fur Mais und Drosophila der Beweis fur das regelmassige Vorkommen von Austausch von Chromosomen stiicken geliefert worden. Und da dieser Beweis an den genetisch so gut bekannten Organismen gefiihrt werden konnte, so liess sich der noch vie1 wichtigere Nachweis fiihren, dass crossing over von Genen bedingt wird durch den Austausch von Chromosomenstucken. Hier liegen vorlaufige Angaben beim Mais vor und ausfiihrliche Untersuchungen an Drosophila. E s handelte sich bei den

Drosophilaversuchen um den Nachweis, dass die zytologisch feststellbaren Chromosomenkonstitutionen bei den vier Typen von Nachkommen einer Kreuzung mit zwei Paaren gekoppelter Gene samtlich verschieden

INTERNATIONAL CONGRESS O F GENETICS

303

voneinander waren. Nach ihrem Phanotypus wurden die Nachkommen der Kreuzungen in vier Klassen eingeteilt (vgl. Abb. 2 ) : die Nichtaustauschklassen "Bar carnation" (schmales Auge, nelkenrot) und "Nicht-Bar, nichtcarnation" (rund, dunkelrot), sowie die Austauschklassen "Nicht-Bar, carnation" und "Bar-nicht-carnation." Wenn die Morgan'sche Grundvorstellung iiber crossing over richtig war, dann mussten die Individuen der einen Nichtaustauschklasse ein Nichtaustauschchromosom "X ohne Y-Arm, fragmentiert" besitzen, die Individuen der anderen Nichtaustauschklasse das andere Nichtaustauschchromosom "X mit Y-Arm, nicht fragmentiert," sowie die Individuen der Austauschklassen ein entsprechendes Austauschchromosom "X ohne Y-Arm, nicht fragmentiert" oder "X mit Y-Arm, fragmentiert." Unter 374 zytologisch untersuchten Individuen waren 213 genetisch als Nichtaustauschindividuen klassifizierte, die samtlich ein Nichtaustauschchromosom besassen, 156 genetisch als Austauschindividuen klassifizierte, die samtlich Austauschchromosomen besassen, und nur 5 Individuen, die nicht der Erwartung entsprachen, was wohl sekundar bedingt war. Aehnliche Versuche an Drosophila von Fraulein PHILIP (noch unveroffentlicht) iiber genetischen und zytologischen Austausch in dem doppelt heteromorphen Zustand "ein X Chromosom nicht fragmentiert, mit Anheftung eines zweiten X Chromosoms an einem Ende-das andere X Chromosom ein Fragment, ohne AnheftungM haben ebenfalls gezeigt, dass Faktorenaustausch und Chromosomenstiickaustausch stets zusammen auftreten. Damit ist die Grundlage der indirekten, genetischen Analyse des Chromosomenverhaltens, die Annahme von regelmassig erfolgendem Chromosomenstiickaustausch gesichert worden, und damit die auf ihr aufgebauten Schlussfolgerungen, die wir oben besprachen. (Aus Mange1 an Zeit kann hier nicht auf die mir ungeniigend begriindet erscheinende andersartige Interpretation W I N K L E Ran ~ den zuletzt dargestellten Versuchen eingegangen werden.) E s ist zu erwarten, dass die nachsten Jahre uns den Abschluss dieser fruchtbaren Periode zytologischer Forschung mittels der verschiedenen Methoden bringen werden. (Fiir die Literatur sei auf den Artikel "Faktorenkoppelung und Faktorenaustausch" im Handbuch der Vererbungswissenschaf t verwiesen. )

304

PROCEEDINGS O F THE SIXTH

T H E U S E O F MOSAICS I N T H E S T U D Y O F T H E DEVELOPMENTAL EFFECTS O F GENES A. H. Sturtevant, California Institute of Technology, Pasadena, California

One of the central problems of biology is that of differentiation-how does an egg develop into a complex many-celled organism ? This is, of course, the traditional major problem of embryology; but it also appears in genetics in the form of the question, "How do genes produce their effects?" One of the most productive methods of studying the question in its embryological aspects is that of grafting, as elaborated by LEWIS,HARRISON, SPEMANN and others associated with them. The use of the grafting technique is also possible in studying the question from the geneticist's point of view, and much valuable information is available. The method has, however, certain limitations, some of which can be avoided by the use of spontaneously occurring mosaics, such as gynandromorphs and somatic mutants. There are two questions that one may hope to study through the use of such mosaics, namely in what parts of the body and at what times of development are specific genes effective? The latter (the question of time of action) may also be studied by methods involving spontaneous changes in experiments, or induced the course of development, as in GQLDSCHMIDT'S changes such as those described by LILLIEand J U H N ;the former question (as to the place at which genes act) may also be studied by means of grafting, constriction, excision, centrifuging, or other treatment. If a Drosophila egg starts development as a female, with two X chromosomes, it was shown by MORGAN and BRIDGES that one of these X's is occasionally eliminated from a cleavage nucleus. The cells descended from such a nucleus are male in character, and there results a gynandromorph. I t often happens in experimental cultures that the male parts of such a gynandromorph have sex-linked recessive genes that are suppressed by their dominant allelomorphs in the female parts of the specimen. MORGANand BRIDGES showed that, in general, the sex and the sex-linked characters in such -gynandromorphs are completely autonomous in development. That is to say, each part develops according to its own constitution, producing the same end-result as though it had formed part of an entire individual of that same constitution. This remains the rule in Drosophila gynandromorphs; but a few exceptions have appeared, and these may be described briefly. I have shown that the sex-linked dominant gene for Bar eye is an exception to the general rule. Bar acts to reduce the area of the eye, and the number of facets removed is the most convenient index of its action. Study of mosaics for Bar shows that the area occupied by facets in the normal

INTERNATIONAL CONGRESS O F GENETICS

305

eye may be divided into three regions with respect to the activity of the Bar gene. In the periphery of this area Bar removes facets not only from tissue that is genetically Bar but also from small not-Bar regions; in the central area Bar never removes facets under any condition; in the intermediate zone Bar removes facets unless not-Bar tissue is present, in which case facets develop in regions that are Bar in constitution and that lie close to not-Bar regions. I have interpreted these results as demonstrating a inutual effect of Bar and not-Bar areas on each other. It must be admitted, however, that it remains possible that we are dealing with a time relation, and that the zones differ only in the relative time at which it is determined in development that facets are or are not to be developed. Bar mosaics have been studied by the use of the Minute-n of BRIDGES,which causes eliminations late in development; if a series of early cleavage mosaics for Bar could be studied the question could be settled. Unfortunately there is no convenient method of obtaining such cleavage mosaics in quantity in Drosophila melanogaster, and Bar is known only in this species. I have recently studied several of the allelomorphs of the sex-linked recessive scute in mosaics, and have found that scute is also an exception to the general rule of autonomous development. Here again most of the mosaics studied are due to eliminations occurring late in development. The results show that scute, which has as its effect the removal of specific bristles, does not remove these bristles as frequently when they lie in small patches that are scute in constitution as it does in flies that are wholly scute. Here also there is a possibility that we are dealing with a time relation, though my own interpretation of the scute series of allelomorphs leads me to suspect that the smallness of the areas scute in constitution is responsible. In this case cleavage mosaics are again needed; a few have been found, and these strongly indicate that scute shows autonomous development in large areas that result from eliminations occurring early in development. I t was observed several years ago by several investigators that yellow body-color present in the form of small spots on wild-type flies is usually not as distinctly yellow as the same area in wholly yellow flies. I t has been uncertain whether this incompleteness of the yellow effect was a real one or was due to the presence of underlying not-yellow tissues partly visible through the yellow area, or possibly to some reaction in the eye or nervous system of the observer. Both of these possibilities are rendered unlikely by the recent observation that such small spots of silver body color are regularly perfectly distinct and possess sharper outlines than do yellow spotsthough silver itself is a color similar to yellow but less different from wild-

306

PROCEEDINGS O F THE S I X T H

type. Yellow, like scute, must be classed as a gene whose effects are weakened in small spots but not in large areas of mosaics. The first exception found to the rule of autonomous development in Drosophila was the sex-linked recessive eye color, vermilion. It has long been known that gynandromorphs with one or both eyes genetically vermilion may have eyes wild-type, vermilion, or intermediate in color. The results from mosaics due to elimination late in development were not particularly helpful here; but vermilion has recently appeared in Drosophila simulans, where there exists a method of obtaining early cleavage mosaics in large numbers. A series of such specimens has now been studied, and definite conclusions are available. In such mosaics, as the experiment was carried out, ovaries are always not-vermilion in constitution and testes are always vermilion in constitution. Of the specimens having eye-tissue vermilion in constitution and large enough in area to classify certainly for color, 51 of the 53 flies that had two ovaries were wild-type in eye color; all 20 of those with two well-developed testes were vermilion or intermediate in eye color; those with poorly developed testes, or with one ovary and one testis, fell into any one of the three color classes. These results mean that the vermilion color in such mosaics is suppressed by something produced by the wild-type ovary. Three separate elements in the system may be distinguished: (1) constitution of the facet area itself; (2) constitution of the gonad; ( 3 ) constitution of some other region, or perhaps merely presence or absence of not-vermilion tissue in the fly. The third element is necessary to account for the differences between specimens with like gonads. Tabulation of such specimens indicates that this third element does not lie in the abdomen; but no particular region of the head or thorax appears to be more effective than another. One may surmise that the effective element is some internal organ not very closely related (in terms of cell lineage) to the surface areas. These experiments were carried out by mating claret 9 X vermilion garnet forked 8 . The garnet and forked served to identify the male regions in the 180 gynandromorphs observed. Two controls were also made: claret 9 X garnet forked 8 made it certain that garnet is completely autonomous in development; vermilion claret 9 X vermilion garnet forked 8 gave 86 wholly vermilion gynandromorphs with vermilion garnet areas, from which it is clear that poorly developed gonads alone do not interfere with the development of vermilion eye color. The relation between eye pigment and gonad constitution described above is especially interesting in connection with the results of DOBZHANSKY, who

INTERNATIONAL CONGRESS O F GENETICS

307

studied D. simulans gynandromorphs in which the male parts were genetically white. The eye color in such mosaics showed autonomous development, as has long been known. White-eyed males also have transparent vasa efferentia and testicular envelopes, those of wild-type males being showed that this yellow color bright yellow at emergence; DOBZHANSKY does not show autonomous development. A testis or a vas efferens attached to a female (that is, under the conditions of the experiment, wild-type) duct or ovary will be yellow a t emergence even though it be white in constitution. Furthermore, even if the entire genital apparatus be white and male in constitution, color may still develop, though much more slowly, if wildtype tissue is present elsewhere in the fly. DOBZHANSKY and I have recently studied this slower development of color in somewhat more detail. I t appears that, under the conditions stated, if both eyes of the mosaic are red (in some cases with very small white areas) the testes develop some color about 4 days after emergence; if both eyes are white or with a small spot of red, color does not appear in the testes until 9 to 10 days after emergence. If the eyes have both white and red present in considerable areas, the result is more variable, but color usually develops in the testes 5 to 8 days after emergence. One may conclude that there is a reciprocal relation here; in vermilion the gonads affect the eyes, in white the eyes affect the gonads. I t is clear that in most cases there is a chain of reactions between the direct activity of a gene and the end-product that the geneticist deals with as a character. One may surmise that any valid generalizations about these reactions are more likely to concern the initial links than the terminal ones. However, it is the terminal ones that are usually more open to experimental attack, since the only index to the effectiveness of a given experimental technique is the condition of the end-product. Looked at from this point of view, the type of experiment that I have described may be considered as a beginning in the analysis of certain chains of reactions into their individual links.

308

PROCEEDINGS OF THE SIXTH

MUTATIONS O F T H E G E N E IN D I F F E R E N T DIRECTIONS N . W . Timofteff-Ressovsky, Kaiser Wilhelm-Instituts fur Hirnforschung, Berlin-Buch, Germany INTRODUCTION

BATESON'S "presence-absence" hypothesis is the most persistent basis for the views upon the nature of the gene and of gene mutations. The modern critique of this hypothesis is chiefly based on more or less indirect counter evidence, shown by the phenomenon of multiple allelomorphism and by the whole, general picture of the mutability of Drosophila (MORGAN, BRIDGES, STURTEVANT 1925, MORGAN,STURTEVANT, MULLERand BRIDGES1923, STERN1930b). This criticism has led to some modernized alterations of the original extreme "presence-absence" hypothesis, for instance, to the view that gene mutations are merely quantitative changes of the gene (GOLDSCHMIDT 1928). But the acceptance of each modification of the "presence-absence" hypothesis leads us to the conclusion that the gene mutability is merely a process of degradation and even of loss of the previously present genic material. If so, then we must take the next step and admit that gene mutations have no positive significance at all in the process of evolution and that the mutability of the species as studied by geneticists is a purely "pathologjcal" phenomenon which is permanently controlled and suppressed by natural selection. The possibility of proving the "presence-absence" hypothesis in a more or less direct experimental way would thus be of great interest. I think that the study of reverse gene mutations and the quantitative study of the mutability of single individual genes can give us a conclusive experimental basis for our views upon the general nature of the gene changes. The study of spontaneous mutations in Drosophila melanogaster and in some other species has already shown that different mutational steps are probably occurring with different frequencies, and that reversions of previously mutated genes may occur in some cases (MORGAN 1926, MORGAN, BRIDGES,STURTEVANT 1925, MULLER1920, MOHR 1922, KEELER1931, 1925,1928, ZELENY1921). The most STERN1930a, TIMOF~EFF-RESSOVSKY convictive cases of reversions, excluding contamination, are those in which 1928) and reversions occurred in somatic tissue (TIMOF~EFF-RESSOVSKY the "frequently mutating genes" in Drosophila uirilis and in some plants 1919, EMERSON (DEMEREC 1928, 1929a, 1929b, 1931, BAUR1926, CORRENS 1923, IMAI 1925, PLOUGH 1928). 1917, EYSTER1924, IKENO An exact quantitative study of reversions and of the mutability of single

309

I N T E R N A T I O N A L CONGRESS O F GENETICS

individual genes is in the spontaneous process of mutability possible only in the cases of "frequently mutating genes." And these latter could be interpreted as a special group of mutations. But with the help of X-rays we can hope to get statistically significant data also upon the mutability of different "normal" genes. In this paper an attempt will be made to give a review of our X-ray experiments upon Yeversions and mutational potencies of some individual genes in Drosophila melanogaster and to draw some conclusions relating to the nature of the gene changes.

k l m n o

k l m n o

: h i m n o \

x=,

! ,

,

h

~

"a I-:-:

k l r n n o i :

h l m n o

'4

h l m n o

h l m n o

FIGURE1.-Scheme of crossings in the X-ray experiments upon reverse gene mutations in Drosophila melanogarter. At the left: X-rayed males, containin.: sex-linked recessive mutations are mated to "attached X" females; all reversions induced in the X chromosome of the fathers will manifest themselves in their sons. At the right: males homozygous for several recessive genes in the I11 chromosome are X-rayed and crossed with homozygous females from the same stock; the induced reversions will manifest themselves in F,. The X-rayed chromosomes are represented by solid lines.

The first question to be solved was whether reversions of previously mutated genes could be produced at all by means of X-ray treatment. F. 3.HANSON has shown that reversions from Bar to full-eye can be produced by X-rays in Drosophila (HANSON 1928). But these reversions can be interpreted as gene deficiencies because Bar is probably a quite new gene and the normal flies seem to have no allelomorph of Bar (STURTEVANT 1925). In the early X-ray work of MULLER and of myself some re-

~

~

PROCEEDINGS OF THE SIXTH

3 10

versions o f recessive mutations w e r e o b s e r v e d in D r o s o p h i l a (MULLER 1928b, 1930a, 1930b, TIMOF~EFF-RESSOVSKY 1929a, b a n d d ) In o r d e r to p r o v e e x a c t l y t h e occurrence and t h e "generality" o f t h e p h e n o m e n o n o f reversions induced b y X-rays, special e x p e r i m e n t s w e r e ar-

.

Reversiom of recessivegene mutations i n the X and 111 chromosomes of D7osophilontelanogaster, produced by X-rays (dosages approximately 3600 r and 4800 r ) . TBQ X-RAYED ALLELOMORPEE AND THEIR LOCI I N TEIE

X-RAYED CHROMOBOMEB,

CHROMOEOMEE

CONTAINING TEE LOCI

I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome, I Chromosome,

IChromosome, I11Chromosome, I11Chromosome, I11Chromosome, 111Chromosome, IIIChromosome, I11Chromosome, I11 Chromosome, I11 Chromosome, I11Chromosome, I11 Chromosome,

0 O+ 2 2 2 2 7 16 25 40 51 62 0 26 42 44 48 50 58 62 71 101

NUMBER OF ANALYZED

TYPn AND NUYBER OF REYEREION8

y s, w w

w we

e, c, cr v

g f r, h tn st

# c, ss s, eb c,

Total number of X-rayed chromosomes, containing the above loci and number of reversions Controls

139234

r a n g e d on a l a r g e scale (TIMOF~EFF-RESSOVSKY 1930a a n d b). T w e n t y d i f f e r e n t recessive m u t a n t allelomorphs o f Drosophila hzelanogaster, 10 in t h e X c h r o m o s o m e (yellow, scute, white, eosin, echinus, crossveinless, cut, vermilion, g a r n e t a n d f o r k e d ) and 10 in t h e 111 c h r o m o s o m e (roughoid, h a i r y , thread, scarlet, pink, curled, spineless, stripe, s o o t y and claret) w e r e Xrayed. In o r d e r to raise t h e c h a n c e o f o b t a i n i n g reversions a m o n g a limited

INTERNATIONAL CONGRESS O F GENETICS

311

number of flies, males from poly-recessive cultures (such as X-pl, 111-pl and 16 rucuca") , containing simultaneously several recessive mutations, were Xrayed. The X-rayed males were then mated either with females of the same genetic composition or (if sex-linked genes were studied) with "attached-X" females. These methods of crossing are shown in figure 1. All cases of reversions obtained in these experiments are summarized in table 1. Reversions to the normal allelomorphs were obtained from scute, eosin, crossveinless, vermilion and forked in the X chromosome and from hairy, pink and sooty in the I11 chromosome. Besides this the following reversions towards normal were obtained: from white to eosin, from white to blood and from eosin to blood. Purely phenotypical, occasional non-manifestation or a suppression of the characters in question by specific suppres1928) can imitate a reverse sors (BRIDGES1932, MORGAN 1929, PLOUGH gene mutation. Therefore all flies showing reversions were tested by further crossings, and in table 1 only those cases are mentioned which proved to be true reverse gene mutations. Contamination, as another possible source of errors, is practically excluded in these experiments, since multiple stocks were used; only one of the genes reverted in each case and several other genes, contained in each culture, served as markers. In total, 18 reversions were observed among 289,000 treated mutant allelomorphs. One hundred thirty-nine thousand controls gave no reversions. This difference is statistically quite significant, so that the production of reversions must be ascribed to X-rays. At the loci of scute, white, forked and pink, reversions occurred more than once. Some of the cultures, containing normal allelomorphs which arose as reversions under X-ray treatment, were further X-rayed and in some cases gave secondary mutations back to or toward the original mutant allelomorphs. In figure 2 are summarized all such cases in which mutations in both directions were induced by X-rays directlysone from another. These and MULLER cases together with similar cases obtained by PATTERSON (1930) are of special interest. They show that the action of X-rays upon the genes is in general by no means of a purely destructive but rather of a reconstructive kind, because, as MULLERhas figuratively expressed it, it is highly improbable that "if with one blow we punch the gene out, with the next we would punch it in again." Thus also the gene changes (gene mutations) can not be losses of the specifically differentiated gene material, but, in some cases at least, must consist of some kind of intra-genic rearrangements of a reversible nature.

312

PROCEEDINGS O F T H E S I X T H MUTATIONS I N OPPOSITE DIRECTIONS AT THE LOCUS O F FORKED I N

DROSOPHILA, INDUCED

BY X-RAYS

The above mentioned extensive first sets of experiments had to establish the fact of occurrence of reverse gene mutations in general and to prove whether their origtn could be ascribed to X-rays. These experiments gave positive results and then the next step was to try to get statistically significant data upon the mutability of some definite individual loci.

FIGURE2.-Allelomorphic pairs of Drosophila melafiogaster, in which mutations were produced directly one from another by means of X-rays. I. From a normal allelomorph of the white series eosin was induced, and this latter produced under further treatment a reversion back to normal. 11. A spontaneously arisen eosin gave under treatment a reversion to normal, and this normal mutated under further treatment back to eosin. 111-IV. Mutations produced by X-rays: from normal to forked and from this forked back to normal, and from forked to normal and from this normal back to forked. V. Mutations from pink to normal and from this normal back to pink.

The most practicable loci for such intensive experiments seemed to be scute, white, forked and pink, since in these loci mutations in both directions were repeatedly induced by X-rays. Scute mutations have been intensively (DUBININ1929, SEREBROVSKY and studied by DUBININand SEREBROVSKY DUBININ 1930) ; therefore, I have concentrated my work upon white and forked. I n the meantime the excellent work of PATTERSON and MULLERwas published, in which the authors describe their X-ray experiments dealing with the same problem and based chiefly upon induced forked mutations (PATTERSON and MULLER1930). In table 2 are summarized the forked mutations induced by X-rays in

INTERNATIONAL CONGRESS O F GENETICS

313

the experiments of PATTERSON and MULLERand of myself. The mean dosage of X-rays was about 3500 r in the experiment of PATTERSON and MULLER and about 4800 r in my experiments. The results are summarized on the basis of 4800 r. The calculations are based on the admission of a direct proportionality between the dosages and the induced mutation rates. Both independent sets of experiments gave substantially the same results. The_ frequencies of "direct" and "revers'e" gene mutations at the locus of forked seem to be of the same quantitative order, reversions being even somewhat more frequent (F+f = 11 : 43000, f + F = 15 : 44000). As was already mentioned above (figure 2), direct and reverse gene mutations at the forked locus were produced directly one from another by XTABLE2 "Direct" and "reverse" mutations of the forked gene in Drosophila melanogaster, prodaced by X-ray treatment. DIRE^" MUTATIONS FROM F TO OR TOWARD FORKED AUTHORS

NUMBER OF X-RAYED CHROMOSOkiES

NUMBER OF MUTATIONS

"RE&RBE"

MUTATIONS &%OM

FORKED TO OR TOWARD

. NUMBER OF X-RAYED

OBROM080MEB

F

NUMBER OF MUTATIONE

PATTERSON and MULLER(Dosage approximately 3500 r)

32000

6

20000

8

TIMOF~EFF-RESSOVSKY (Dosage approximately 4800 r)

19000

5

29000

7

Total, on the basis of 4800 r

43000

11

44000

15

rays. In the experiments of PATTERSON and MULLERalso 4 of the forked mutations ( F + f ) were induced by X-rays from a normal allelomorph, which itself derived from forked as an X-ray induced reversion (PATTERSON and MULLER1930). Experiments with forked have shown that at this locus mutations in both directions can be produced with practically the same frequency. The preliminary results of some X-ray experiments with the bobbed locus of Drosophila melanogaster (these experiments are not yet accomplished) are showing that at this locus reversions from an extreme mutant allelomorph bobbedlethal (kindly given me by C. STERN)to or toward normal are probably much more frequent than "direct" mutations from normal to or toward bobbed. This extreme bobbed allelomorph (bobbed-lethal) behaves like a n ordinary sex-linked recessive lethal, the only difference being that it does not kill the males, since they always contain the normal allelomorph of

314

PROCEEDINGS O F THE SIXTH

bobbed in their Y chromosome (STERN 1930b). If the above mentioned results are confirmed by further experiments, they will show that even such "pathological" mutations as the lethals can be reversible and therefore are not always destructions or losses of the "normal" genes (gene deficiencies, PATTERSON 1932). MUTABILITY I N DIFFERENT DIRECTIONS W I T H I N THE WHITE-EYE SERIES O F

DROSOPHILA,

PRODUCED BY X-RAY TREATMENT

The most intensive study was done upon the mutability in the white-eye series of Drosophila naelanogaster. This locus was chosen for the following reasons : 1. Spontaneously and under X-ray treatment it proved to be one of the most mutable loci. 2. A t this locus many allelomorphs are known, so that one could hope to produce many different mutations from and to different allelomorphs. 3. Phenotypically this series of allelomorphs shows a quantitative scale of eye colors, so that one could easily assume that the gene changes leading to the different allelomorphs are in this series also of a purely quantitative kind. A uniform method was used in all X-ray experiments with the whiteeye series. Males from different cultures, containing the white allelomorphs and also some other sex-linked genes as markers (in order to avoid contamination), were X-rayed with the same, rather heavy dosage of about 4800 r and mated with "attached X" females (figure 1). All mutations of the W gene could be detected in the F, males. In order to eliminate those cases in which specific modifiers arising at other loci or some other eyecolor mutations are imitating mutations of the white locus, all newly arisen eye-color mutations were tested by further crossings. The following allelomorphs of the white-eye series were X-rayed: Normal ( W), coral (ztfo), blood (&), cherry (&), apricot (&), eosin (&), buff (wbf) and white (w). The newly arisen mutations were classified by careful comparison with the known ones. Most of them proved to be recurrences of previously known allelomorphs, but some of them showed slight differences and were classified as new ones. Altogether about 185,000 F, males from X-rayed parents were examined and 68 mutations of the white locus were found among them. The mean mutation rate was thus about 1 : 2700. First we will consider the mutability of the white locus only qualitatively. In figure 3 it is shown that white was induced as a mutation from all those

INTERNATIONAL COICGRESS O F GENETICS

315

allelomorphs which were X-rayed. Figure 4 shows that the allelomorphs blood, eosin and buff were induced as "direct" mutations from normal ( W + d , W+M and W + d f ) and also as "reversions" from white (w+d, w+& and w+wbf).In figure 5,are shown the different mutations to and from eosin which were induced by X-rays. Mutations to eosin were , blood ( d + W ) , from apricot induced from normal ( W + ~ ) from

FIGURE 3.-Mutations from different W allelomorphs to white, produced by X-rays in Drosophila melanogaster. (M+zer") and from white (w+&). From eosin mutations were induced to normal (W+?V), to blood (.w"+zd) and to white ( d + w ) . Mutations from eosin to normal and from this normal back to eosin ( d e w )and mutations from normal to eosin and from this eosin back to normal (WFtWe) were directly induced one from another by X-rays (figure 2). The eosin allelomorph, which arose as an X-ray induced reversion from white, has under further treatment given two independent mutations back to white ( w ) . All the above facts show that at the white locus quite different mutational steps and mutations in different directions can be induced by Xrays. At least some of the mutational processes within this locus are reversi-

PROCEEDINGS O F THE S I X T H

316

ble. In this respect the results are in principle the same as those obtained a t the forked locus. On the other hand we can already note some differences. At the forked locus the extreme mutations in both directions (F+f and f+F) were produced. At the white locus the extreme reversion from

wW FIGURE 4.-The same intermediate allelomorphs of the white series in Drosophila melanogaster (blood, eosin and buff), produced by X-rays from the extreme allelomorphs W (normal) and w (white). T h e extreme reversion w+W was never observed.

white direct to normal (w+W) was never observed. But this reversion can be produced in two steps: from white to eosin and from eosin to normal

(w->,We+W). I f we now consider quantitatively the single mutation rates a t the white locus, then the "unordered" general picture of the mutability will change

W-W

t----------

e <

>W

FIGURE 5.-Mutations from and to eosin, produced by X-rays in the white-eye series of Drosophila melanogaster.

its face and further differences from the mutability of the locus of forked can be noted. In table 3 are summarized all mutations induced a t the white locus and the corresponding numbers of treated gametes (F1males) among which they were found. I t is quite evident that different m u t a t i o ~ sare not produced

INTERNATIONAL CONGRESS O F GENETICS

'

317

with equal frequencies. The normal allelomorph ( W ) mutated 37 times in a total of 48,500 tested gametes, while the allelomorph white (w)mutated only 3 times in a total of 54,000 tested gametes. Among 37 mutations from normal ( W ) are 25 white mutations (W+zu) and only 12 mutations to a t least six different other allelomorphs. Eosin (w" seems also to mutate much more frequently to white than to all other allelomorphs, but the total mutability of eosin is lower than that of normal. TABLE 3 Summary of all mutations i n different directions within the white-eye series of Drosophila melanogaster, produced by X-ray treatment (dosage approximately 4800r). "DIRECT" GENE MUTATIONS IN THE

MUTATIONS

W LOCUS

"REVERSE" QENE MUTATIONS IN TAE

NUMBER

NUMBER

RATE OF

OP MU-

OF

MUTATION

T.VI"TONS

FLIES

I N O/OO

o,272

YUTATIONS

wc... . ? d ... .?

w LOCUB

NUMBER

NUMBER

RATE OF

OP MU-

OP

YUTATION

TATIONS

PLllS

IN

O/OO

0 0

In table 4 the different mutations of the white locus are classified and their mutation rates are compared. In a total of 129,000 tested gametes 62 "direct" mutations (from darker allelomorphs to lighter ones) were induced. On the other hand only 6 "reverse" mutations (from lighter allelomorphs to darker ones) were induced in a total of 134,500 tested gametes. The difference between these mutation rates is statistically quite significant. Statistically.significant also are the differences between the rates of mutations from eosin to lighter allelomorphs (15 : 39,000) and from eosin to darker allelomorphs ( 3 : 39,000), and the rates of mutations from normal and from white to different other allelomorphs (W+& = 37 : 48,500 and w+W = 3 : 54,000). W e can group all the tested allelomorphs into three classes: (1) normal, ( 2 ) intermediate allelomorphs (coral, blood, cherry, apricot and

3 18

PROCEEDINGS OF THE SIXTH

eosin) and ( 3 ) light allelomorphs (buff and white). Then the highest mutability is shown by normal, the intermediate allelomorphs have a mutability which is half as high as that of normal, and the lowest mutability is shown by the lightest allelomorphs. The quantitative study of mutation rates thus shows that the mutability within the white-eye series is not unordered and not merely a matter of chance, but rather has some characters of "determinate variation" (VOGT 1929). The "directing principle" is manifested by the following facts: ( 1 ) different mutational steps and mtitations from different allelomorphs are Comparison of diJerent mutation rates within the white-aye series of Drosophila melanogaster, produced by X-ray treatment (dosage apfloximately 4800 r ) . MUTATIONS

All "direct" All "reverse"

NUMBlR OF

kDMBER OP

RATE OP MUTATION

MDTATIONS

PLIES

IN 0/00km

62 6

129000 134500

0.481k0.061 0.045rt0.018

DIPPERENOE OF RATES f m dif,

0.436$_0.063

produced with quite different frequencies ; ( 2 ) "direct" mutations are much more frequent than "reversions"; (3) some of the theoretically conceivable mutational steps are probably not realizable (for instance, w+W) ; (4) mutational end-results which are not realizable at once can be attained in two definite steps (w+zer"W). COMPARISON O F IDENTICAL ALLELOMORPHS O F DIFFERENT ORIGIN AND THE

EXISTENCE OF DIFFERENT "NORMAL"

ALLELOMORPHS OF THE WHITE-

EYE SERIES O F DROSOPHILA

The last series of experiments deals with the question of whether W allelomorphs of different origin, classified as identical according to their eyecolor effects, are really identical. An exact classification of eye colors is dif-

INTERNATIONAL CONGRESS O F GENETICS

319

ficult and a priori it could be possible to admit that practically each mutation is leading to a somewhat different new allelomorph, that the eye colors are really forming a continuous series and that our classification of allelomorphs is artificial. A confirmation of this assumption would be a strong support for the hypothesis of purely quantitative mutational changes at this locus. The first evidence against the assumption of a continuous quantitative series of eye-color allelomorphs is the fact that some of the white allelomorphs are showing specific peculiarities in their eye-color effects: eosin has, for instance, a pronounced sex dimorphism, and blood shows a high degree of fluctuation in its eye color. Another way to solve this question would be by the study of manifold effects of certain allelomorphs of different origin. W e can test whether mutations classified as identical according to their eye colors are also identical in respect to other effects of the gene. Within the white-eye series of Drosophila rnelanogaster the normal, blood, eosin and white allelomorphs can be most easily and exactly identified on the basis of their eye-color effects. Different mutations to these allelomorphs were thus chosen for the study of the manifold effects of identical allelomorphs. The following cultures were used: (1) Four white mutations of difin 1922 ferent origin ( a "spontaneous" white obtained from H. J. MULLER and designated as w - 1; two different W+w mutations, induced by X-rays and designated as w - 4 and w - 7; one w"+w X-ray induced mutation, designated as w - 11); ( 2 ) Four eosins of different origin (the original spontaneous" M - 1; zple - 3 and zel" - 6, induced as W - af mutations ; Zpje - 4, an X-ray induced w+& reversion) ; ( 3 ) Four different bloodcultures (a "spontaneous" zeP - 2 ; two independent X-ray induced W + d mutations designated as zd' - 2 and ZPP - 3 and a M + w b reversion designated as d - 5 ) ; (4) Four normal allelomorphs of different origin ( W A from American cultures, WR from a Russian culture; W X 1and WX2- two X-ray induced reversions from eosin to normal). Besides the eye colors, which served as a basis for the classification of the above mutations, the following effects of the four allelomorphs were tested: 1927), viability of the males (meascolor of the testicle tunic (DOBZHANSKY ured by the deviation of the mutant type from the 1:l ratio in the male progeny of heterozygous females) and fertility of the females (measured by the mean number of eggs laid by 1 female in the first 10 days of the egg-laying period). Before the tests of these effects were made, all the white, eosin and blood cultures were backcrossed through more than 20 generations 61

PROCEEDINGS OF THE SIXTH

320

Manifold effects of w, we, d and W allelomorphs of different origin; the allelomorphs were identified and classified by comparison of their eye-color effects. The viability rates are given i n percentage of the viability of wA(=I00 percent) and are based on the deviations from the 1:1 ratio i n the male progeny of heterozygousfemales. The fecundity rates are given i n the mean number of eggs per female, laid d ~ r i n gthe f i s t 10 days of the egg-laying period. FECUNDITY I N ALLELOMORPHB

COLOR OF THE

VIABILITY I N

MEAN NUMBER OF

TUNIC OF THE

PERCENTAGE, AS CON-

EGGS PER FEMALE

TEBTICLES

PARED WITH

W

REMARK0

I N TEE FIRST

10 DAY0

w- 1 w - 4 w-7 w -11

pellucid pellucid pellucid pellucid

7951.0 67k1.0 82f2.0 6951.5

550f 9.5 503 f 13.5 543 f 18.0 481 k 16.5

Viability rates based on deviations from 1 :1 ratio in the male progenies of heterozygous wA/w (and

Total w

pellucid

7251

5335 17

wA/weor wA/&) females

1 3 4 6

pellucid pellucid pellucid pellucid

8851.0 90f1.5 89k1.0 9251.5

402f 17.0 391 f 19.5 398f 22.0 411k18.5

Total we

pellucid

90+ 1

395 5 23

5

pellucid pellucid pellucid pellucid

83f 1 . 5 76f1.5 80k1.0 84k1.5

418520.0 441k23.0 422f17.0 436f19.0

Total &

pellucid

81 5 1

438 14

wA wR wX1 wX2

yellow yellow yellow yellow

100 108k1.0 9651.0 103+1.5

653rt 18.0 691 k 29.5 622f15.5 6725 20.5

102f1

664f 19

wewewewe-

dd&d-

1

Z 3

-

Total W

where the viability of

wA is taken as equal to 100percent

+

Viability rates based on male progenies of heterozygousw-l/Wfemales.

--

yellow

with a definite pure-bred normal culture. This was done in order to eliminate eventual influences of other genes upon the differences in the above mentioned effects, to be found between allelomorphs of different origin. In table 5 are summarized the results of these tests. White, eosin, blood and normal show very pronounced differences in viability and fertility. But it is evident, that all four eosins of different origin are quite identical in respect to all four tested effects. The same is true for blood. The four whites show greater differences in viability and fertility. This latter fact can be

INTERNATIONAL CONGRESS O F GENETICS

321

ascribed to chance or could be perhaps explained by the assumption that there are different white allelomorphs which are indistinguishable in their eye color. But the results on eosins and bloods of different origin clearly show that definite mutational steps and recurrences of exactly the same allelomorphs really exist. A second important fact shown by the results of the tests summarized in table 5 is that the quantitative gradation of the allelomorphs of the whiteeye series will be different if based on different effects of this gene. The sequence of the four tested allelomorphs (W, d, 7tf and w ) in the direction from "normal" toward the "abnormal condition" is, if based on their eyecolor effects: normal (W)>blood (zvb)>eosin (&)>white (w).I f based on viability rates, the sequence is W > z t f > d > w . And if based on From the above tests it is evifertility rates, the sequence is W>W>W">PJ. dent: ( 1 ) that definite, recurrent mutational steps exist within the whiteeye series and (2) that the different allelomorphs can not be ordered in a simple quantitative series if manifold effects of the gene are taken into consideration. (When these experiments were already accomplished and came to my attention [FRISEN the paper was written, a paper by G. FRISEN 19311. In this paper the author examines the fertility of several W allelomorphs, their compounds and heterozygotes with wild-type and proposes, as explanation of the results obtained, a hypothesis of "chain-mutations." Without being able to give here a full discussion of the matter, I can only briefly mention that my results, showing the identity of similar recurrent mutations, seem to disprove the hypothesis of "chain-mutations.") The last question to be mentioned here is: are all allelomorphs which we designate as "normal" and can not distinguish one from another really identical or not? This question arose in connection with the following facts. In our X-ray work two "normal" cultures of Drosophila nzelanogaster were used, one derived from a Russian wild population and the other from America. In summarizing the germinal and somatic white mutations obtained in the first sets of X-ray experiments, it was found that the "American" flies gave about twice as many mutations a t the white locus as the "Russian" flies (under the same experimental conditions). The data showing this peculiar race difference were not statistically significant. But nevertheless this question was pursued in all further X-ray experiments in order to test this phenomenon on a large scale. Such race differences in the mutation rate of a definite gene can be caused either by the general differences in the two genotypes (in some way controlling the mutability of the single gene) or by the structural difference of

322

PROCEEDINGS OF T H E S I X T H

this gene itself. In order to solve this question, four cultures were produced by crossings (with the help of "markers" in different chromosomes) : (1) a purely "American" culture, ( 2 ) a culture containing the "American" left end of the X chromosome in a "Russian" set of chromosomes (the American W Aallelomorph in Russian "genotypical environment"), ( 3 ) a purely

FIGURE 6.-Scheme of the "American" (WA) and "Russian" (WR) normal allelomorphs of the white-eye series of Drosophila melanogaster in "Russian" and in "American" genotypical environment. a. WA in its own, American genotype. b. WR in its own, Russian genotype. c. WA in Russian genotype. d. WR in American genotype.

"Russian" culture and (4) a culture containing the "Russian" left end of the X chromosome in an "American" set of chromosomes (the Russian W R allelomorph in American "genotypical environment"). Schematical drawings of these four genotypes are shown in figure 6. Males from these four cultures were then used in all further X-ray ex-

INTERNATIONAL CONGRESS O F GENETICS

323

periments. All induced mutations at the locus of white were registered according to the above four genotypes of the X-rayed males and were classified into two groups: 1, white (W+w) and 2, all other colored allelomorphs ( W+wX) . Table 6 shows the results of all X-ray experiments in which the four above cultures were used. The original observation was confirmed by these Mutability, produced by X-ray treatment, of the American ( w A )and Russian ( w R )normal allelomorph of the white-eye series of Drosophila melanogaster. The W gene mutatiolzs produced are cZassi$ed into two grozlps: 1, white (W-+a) and 2, all other, intermediate eye-colors (W-+wx).X-ray dosage= apfroximately 4800 r. NUMBER OF PRODUCED

w MUTATION6

PERCENTAGE6

NUMBER 01 OULTURmS

OF

X-RAYED

W Q ~ N E S W-+w

W+wZ

TOTAL

W+w

OF ILL GENE MUTA-

AMONG ALL

=IoN8

MOTATION6

w GENE

OF

W+wX

AMONG ALL GnNE MUTATIONS

wAin American "geno31000

22

5

27

0.087

81

19

W* in Russian "geno28200 typic environment"

19

9

28

0.100

68

32

41

14

55

0.093

75f5

25+5

typic environment"

wAtotal

59200

f 0.012

wR in

Russian "geno49200 typic environment"

13

13

26

0.053

50

50

26100

6

8

14

0.054

43

57

75300

19

21

40

0.053 +0.008

47f7

53f7

wRin American "genotypic environment"

wRtotal

The difference between the mutabilities of WA and wR=0.040 0.013 percent. The diierence between the relative percentagcs of w and wz gene mutations from 28 k8.5 percent.

+

wAand W R =

results: the American WA allelomorph has given nearly twice as many mutations as the Russian W R allelomorph (55 : 59200=0.093 percent and 40 : 75300=0.053 percent respectively). A second difference between the two normal allelomorphs is that the more frequently mutating American WA gives many more mutations direct .to whitel than to all other allelomorphs (41 WA+w : 14 ITA+&),and the more stable Russian WR pro-

324

PROCEEDINGS OF THE SIXTH

duced white and colored allelomorphs in approximately equal numbers (19 WR+w : 21 WR+ziF). Both differences are statistically significant. The summarized results of all experiments also show that the above mentioned differences in the mutability of WA and WR are not due to different "genotypical environment." The American WA mutates in the same way in its own American and in the Russian genotypes. And the mutability of the Russian WR remains substantially the same in Russian and in American c genotypical environment." I

W e thus had two "normal" allelomorphs differing in their mutabilities but otherwise indistinguishable. CONCLUSIONS

From the results obtained in all above reviewed experiments, some conclusions can be drawn upon three intimately connected questions: ( 1) the nature of the action of X-rays on the genes, ( 2 ) the general nature of muiational gene changes and ( 3 ) the general nature of the gene structure. The following statements concerning the nature of the genetic X-ray action can be made: 1. The action of X-rays upon the genes is by no means purely destructive. A direct proof of this statement is given by those cases in which mutations in both opposite directions were produced by X-ray treatment directly one from another (figure 2). Each mutational change can a priom' be ascribed to total or partial gene destruction, but mutations in opposite directions, as, for instance, F+f and f+F, cannot possibly both be destructions: if one of them is so, then the other must be a gene construction. This point has already been emphasized in the excellent work of PATTERSON and MULLERand in some of my own previous papers. 2. X-rays can, even in a definite single locus, produce quite different mutational changes. The best example illustrating this statement is given by the different mutations from and to eosin shown in figure 5. This property of X-rays is to be expected if we recall the nature of physical action of X-ray quanta upon the matter.'Even monochromatic X-rays can produce in the X-rayed matter electrons of very different speeds; this depends upon the amounts of energy which happen to be given up by the X-ray quanta to the accidentally hit and ejected electrons. These ejected electrons of different speed build a large scale of different energetic quanta which now can be delivered to other molecules and produce quite different effects. Since the kind of reactions which the struck molecules will undergo is also dependent upon the kind of atoms and molecules present at the moment in

INTERNATIONAL CONGRESS O F GENETICS

325

their surroundings, it is clear that quite different end-effects can be produced by X-ray treatment. Some of the reactions can, of course, as MULLER has already pointed out (PATTERSON and MULLER1930), lead t o the "breakdown" or to a "simplification" of the previous physico-chemical structure, but at least some of the others must be of the nature of syntheses. 3. The specific nature and the relative frequencies of different mutations are primarily dependent upon the specific structures of the treated genes and not upon the specific action of X-rays. This becomes evident from the general parallelism shown by the "spontaneous" and the X-ray induced mutations in Drosophila. And this is particularly proved by the general differences in the mutabilities at two different loci (for instance, those of white and forked in Drosophila) induced by the same X-ray dosage. Concerning the nature of mutational gene changes the following can be stated : 1. In general gene mutations are not losses of the specific and probably highly differentiated genic material. This is proved by the fact that at least some of the mutational processes are reversible. But even those cases in which reversions can not be produced are by no means proving the loss hypothesis. The reversion from white to normal (w+W) is probably not realizable, but the occurrence (although only in very rare cases) of reversions from white to eosin proves that white is not "nothing," is not a loss of the gene. True "gene deficiencies" probably occur in certain cases (PATTERSON 1932), but they are surely not the typical kind of gene mutations. 2. In general the gene mutations are not merely quantitative alterations of the previous genic material. The quantitative theories of gene mutations are primarily based on some phenotypical phenomena, chiefly upon the fact that some series of multiple allelcimorphs can be ordered in a quantitative scale according t o their phenotypic effects. But against this view some objections can be made. An important logical objection is that we have no reasons to transfer a quantitative scale of phenotypical effects into the gene structure. Still more important is the fact that even the phenotypes of multiple allelomorphic series can not always be arranged in simple quantitative scales. In some cases different allelomorphs manifest qualitatively different characters (DUBININ1929, STERN1930b). In still other cases the same multiple allelomorphs must be arranged in different sequence on a quantitative scale if different effects of the gene are taken into consideration. This was shown by the results of the analysis of the manifold effects of the white-eye allelomorphs of Drosophila nzclanogasfer. Finally, against a purely quantitative theory of gene mutations the same objections

326

PROCEEDINGS O F THE SIXTH

can be made as against the extreme "presence-absence" hypothesis: the loss of the whole gene o r of a half or a quarter of the gene is equally improbable in all those cases of mutations which are reversible. Thus mutations must in general consist of some kind of intra-genic rearrangements altering the physico-chemical structure of the gene, and purely quantitative gene mutations are (like the "gene deficiencies") conceivable only as a special group of mutational changes. 3. The specific structures of the genes determine the relative frequencies and the direction of mutational changes. At some loci (for instance, at that of lorked) the mutability seems to be, within certain limits, quantitatively unordered: even both extreme allelomorphs (normal and the "spontaneous" forked used in our experiments) show the same degree of mutability. In other cases, as in that of the white-eye series, the mutability is a clearly determinate or directed one: not all theoretically conceivable mutational steps are realizable, and those which are possible show different, specific method of induction of frequencies. V. JOLLOS,applying GOLDSCHMIDT'S mutations (treatment of Drosophila larvae with high temperature), has shown that an almost absolutely directed series of mutational steps can be 1929, induced at the white locus of Drosophila melanogaster (GOLDSCHMIDT JOLLOS 1930, 1931a, 1931b). In our X-ray experiments the mutability a t the white locus was not by far as absolutely directed as in the experiments of JOLLOS.But this difference is not at all surprising if we consider the high power and the heterogeneity of X-ray action. Thus the results of X-ray and temperature experiments with the white locus of Drosophila and especially the finding of differently mutating "normal" allelomorphs of white show that the gene structures can determine the direction of mutability and the relative frequencies of different mutations of a species and even of a race. Some conclusions upon the nature of the gene structure spontaneously result from the above statements. Are the genes fixed quantities of specialized matter, consisting of a definite number of identical physico-chemical units (for instance, molecules) ? O r is the gene itself a physico-chemical unit of some kind ( a large molecule, a micella or a colloid particle of specific structure) ? The first hypothesis is, I think, unacceptable as a general scheme. I t seems to be very improbable that X-rays would simultaneously change in the same way all the identical physico-chemical units constituting the gene. W e would be thus forced to accept the quantitative theory of gene mutation, which, as we have seen above, encounters many serious objections. The most plausible general scheme of gene structures is thus the ac-

INTERNATIONAL CONGRESS O F GENETICS

327

ceptance of the second hypothesis: genes are probably physico-chemical units which can undergo definite reactions, and some of these mutations-reactions are of reversible nature. I think that further details concerning a general scheme of gene structure are now of little importance and might do more harm than good for the progress of experimental work. The mutabilities of the scute (DUBININ1929), white and forked (PATTERSON and MULLER 1930) loci in Drosophila ~lzelanogasterand of the "frequently mutating" genes in Drosophila virilis, the only well analyzed cases, are in some respects so different that we must admit that different genes differ profoundly one from another in their structures. I am at the end of my paper. W e geneticists are in a very happy condition: our science is young, its "developmental curve" is rising rapidly and the future will bring us the most interesting facts and views concerning the gene problem. W e must and can be optimists. And the chief purpose of my paper is to show (besides those theoretical aspects which were mentioned above) that with the help of the X-ray method the mutability of individual genes and the study of their evolutionary potencies can be attacked experimentally. In connection with phenogenetical work the study of mutability can show us the intimate nature of gene mutations. And a quantitative comparztive study of the mutability in related species can elucidate some of the profound problems of evolution. But surely it will furnish us exact empirical materials on which our views upon the method of evolution and the nature of the genes can be based. SUMMARY

1. The following reverse gene mutations were induced by X-rays in Drosophila nzelanogarter: from scute, eosin, crossveinless, vermilion and forked to their normal allelomorphs, from white to eosin and to blood and from eosin to blood in the X chromosome and from hairy, pink and sooty to their normal allelomorphs in the I11 chromosome (figure 2 and table 1 ) . 2. At the locus of forked (in the X chromosome of Drosofhila, melanogaster) 11 direct mutations from normal to or toward forked ( F + f ) and 15 reverse mutations from forked to or toward normal ( f + F ) were induced by X-rays in a total of 43,000 normal and 44,000 forked treated gametes (in experiments of PATTERSON and MULLERand of the present author). This result proves that at this locus mutations in both opposite directions are arising with approximately equal frequencies (table 2 ) . 3. Within the white-eye series of Drosophila mutations in almost all conceivable directions were induced by X-ray treatment (figures 3, 4 and 5,

328

PROCEEDINGS O F T H E S I X T H

table 3 ) . But the frequencies of different mutations are quite different, reversions being much less frequent than direct mutations and the mutability of normal being higher than that of intermediate and especially than that of the lightest allelomorphs (table 4). 4. A study of viability and fertility rates and of the testicle-tunic color of ofh different s origin has shown white, eosin, blood and normal a l l e l ~ m o ~ ~ that: ( a ) allelomorphs of different origin, classified as identical according to their eye-color effects, are really identical, showing no differences in their manifold effects; (b) normal, blood, eosin and white must be arranged in different orders on a quantitative scale if different effects of the gene (eyecolor, viability, fertility) are taken as the basis of classification (table 5 ) . 5. Two distinct "normal" allelomorphs of the white-eye series, showing different stability (mutating with different frequencies) but otherwise indistinguishable, could be detected in the X-rayed material (figure 6 and table

6). 6. Some conclusions upon the general nature of X-ray action and the gene structure were drawn from the experimental results (see Conclusions). LITERATURE CITED

ALEXANDER, J , and BRIDGES, C. B., 1928 Some physico-chemical aspect; of life, mutation and evolution. Colloid Chemistry (N.Y.) 2: BAUR,E., 1926 Untersuchungen iiber Faktormutationen. 1-111. Z. indukt. Abstamm.-u. VererbLehre. 41 : BRIDGES, C. B., 1932 The suppressors of purple. Z. indukt. Abstamm.-u. VererbLehre. 60: CORRENS, C., 1919 Vererbungsversuche mit buntblattrigen Sippen. I. Sitzber. d. Preuss. Akad.d. Wiss. 34: DEMEREC,M., 1928 The behavior of mutable genes. Proceedings of the Fifth International Congress of Genetics, Supplementband of the Z. indukt. Abstamm.-u. VererbLehre. 1: 1929a Changes in the rate of mutability of the mutable miniature gene of Drosophila virilis. Proc. Nat. Acad. Sci. Washington 15: 1929b Genetic factors stimulating mutability of the miniature-gamma wing-character of Drosophila virilis. Proc. Nat. Acad. Sci. Washington 15: 1931 Behavior of two mutable genes of Delphinium ajacis. J. Genet. 24: DOBZHANSKY, TH., 1927 Studies on the manifold effect of certain genes in Drosophila melanogaster. Z. indukt. Abstamm.-u. VererbLehre. 43: DUBININ,N. P., 1929 Allelomorphentreppen bei Drosophila melanogaster. Biol. Zbl. 49: EMERSON, R. A,, 1917 Genetical studies of variegated pericarps in maize. Genetics 2: EYSTER,W. H., 1924 A genetic analysis of variegation. Genetics 9: FRISEN,G., 1931 Kettenmutationen. ~ u r n a lEksperim. Biologii 7:(Russian) GOLDSCHMIDT, R., 1928 The gene. Quart. Rev. Biol. 3: 1929 Experimentelle Mutation und das Problem der sogenannten Parallelinduktion. Biol. Zbl. 49: HANSON, F. B., 1928 The effect of X-rays in producing return gene mutations in Drosophila. Science 67: IKENO,S., 1923 Erblichkeitsversuche an einigen Sippen von Plantago major. Jap. J. Bot. 1:

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

329

IMAI, I., 1925 Genetic behavior of the willow-leaf in the Japanese Morning Glory. J. Genet. 16:

JOLLOS,V., ,1930 Studien zum Evolutionsproblem. I. Uber die experimentelle Hervorrufung und'steigerung von Mutationen bei Drosophila melanogarter. Biol. Zbl. 50: 1931a Die experimentelle Auslosung von Mutationen und ihre Bedeutung fiir das Evolutionsproblem. Die Naturwissenschaften 19: 1931b Genetik und Evolutionsproblem. Verh. Deutsch. Zool. Ges. KEELER,C. E., 1931 A reverse mutation from "dilute" to "Intense" pigmentation in the house mouse. Proc. Nat. Acad. Sci. Washington 17: KOLTZOFF,N. K., 1928 Physikalisch-chemische Grundlage d e r hlorphologie. Biol. Zbl. 48: MOHR,0. L., 1922 Cases of mimic mutations and secondary mutations in the X chromosome of Drosophila melanogaster. Z. indukt. hbstamm.-u. VererhLehre. 28: MORGAN,T . H.; 1913 Factors ant1 unit characters in Mendelian heredity. Amer. Nat. 47: 1926 T h e theory of the gene. New H a v e n : Yale Univ. Press. 1929 Data relating to six mutants o f 1)rosophila. I'ub. Carnegie Instn. 399: MORGAN,T. H., BRIDCES,C. B., STURTEVAKT, A. H., 1925 T h e genetics of Drosophila. Bibl. genet. 2: MORGAN, T. H., STURTEVANT. A. H., MCLLER,H . J., and BXID(;ES,C. R., 1923 T h e mechanism of Mendelian heredity. New Y o r k : Henry Holt Co. MULLER,H. J., 1920 Further changes in the white-eye series of Drosophila and their bearing on the manner of occurrence of mutations. J. Exp. 2001. 31: 1923 Mutation. Eugenics, Genetics, and the Family 1: 1928a T h e problem of genic modification. Proceedings of the Fifth International Congress of Genetics, Supplementband o i the Z, indukt. Abstamm.-u. VererbLehre. 1: 19.2813 T h e production of mutations by X-rays. Proc. Nat. Acad. Sci. Washington 14:

1929a T h e method of evolution. Sci. Month. 29: 1929b The gene a s the basis of life. Address given 1926 to the Intern. Congr. of Plant Science, Ithaca, New York. Proc. Intern. Congr. Plant Science. 1: 1930a Radiation and genetics. Amer. Nat. 64: 1930b Types of visible variations induced by X-rays. J. Genet. 22: PATTERSON, J. T., 1929 The production of mutations in somatic cells of Drosopkila melanogaster by means of X-rays. J. Exp. Zool. 53: 1932 Lethal mutations and deficiencies, produced in the X chromosome of Drosophila melanogoster by X-radiation. Amer. Nat. 66: PATTERSON, J. T., and MCLLER,H . J., 1930 Are progressive mutations produced by X-rays. Genetics 15: PLOUGH,H. H., 1928 Black suppressor-a sex linked gene in Drosophila. Proceedings of the Fifth International Congress of Genetics, Supplementband of the Z. indukt. Abstamm.-u. VererbLehre. 2: SEREBROVSKY, A. S., 1929 A general scheme for the origin of mutations. Amer. Nat. 63: SEREBROVSKY, A. S., and DVBISIK,N. P., 1930 X-ray experiments with Drosophila. J. Hered. 21:

SPENCER,W. P., 1926 T h e occurrence of pigmented facets in white eyes in Drosophila melanogarter. Amer. Kat. 10: STERI~, C., 1930a Kleinere Beitrige zur Genetik von Drosophila melanogaster. 11. Gleichzeitige Riickmutation zweier benachbarter Gene. Z. indukt. Abstamm.-u. VererbLehre. 53 : 1930b Multiple Allelie. Handb. d. Vererbwiss., Nr. 14, Berlin. STURTEVANT, A. H., 1925 T h e effect o f unequal crossing over at the Bar locus in Drosophila. Genetics 10:

330

PROCEEDINGS O F T H E S I X T H

TIMOF~EFF-RESSOVSKY, H. A., 1930 Rontgenbestrahlungsversuche mit Drosophila funebris. Die Naturwissenschaften 18: TIMOF~EFF-RESSOVSKY, N. W., 1925 A reverse genovariation in Drosophila fuiwbris. ijurnal Eksper. Biologii, v. 1 (Russian) and (1927) Genetics 12: 1928 Eine somatische Riickgenovariation bei Drosophila melanogmter. Arch. EntwMech. Org. 113: 1929a The effect of X-rays in producing somatic genovariations of a definite locus in different directions in Drosophila-melanogmter. Amer. Nat. 63: 1929b Somatic genovariations in a definite locus in different directions, produced by X-ray treatment. Trudy Vsesoj. Sjezda po Genetike, v. 2, Leningrad (Russian). 1929c Der Stand der Erzeugung von Genovariationen durch Rontgenbestrahlung. J. Psychol. u. Neurol. 39: 1929d Riickgenovariationen und die Genovariabilitat in verschiedenen Richtungen.1. Somatische Genovariationen der Gene W, w und w bei Drosophila melanogaster unter dem Einfluss der Rontgenbestrahlung. Arch. EntwMech. Org. 115: 1930a Reverse genovariations and the genovariability in different directions. 11. The production of reverse genovariations in Drosophila melanogaster by X-ray treatment.2urnal. Eksperim. Biologii 6: (Russian) and (1931) J. Hered. 22: 1930b Das Genovariieren in verschiedenen Richtungen bei Drosophila melanogmter unter dem Einfluss der Rontgenbestrahlung. Die Naturwissenschaften 18: 1931a Einige Versuche an Drosophila melanogaster iiber die Art der Wirkung der Rontgenstrahlen auf den Mutationsprozess. Arch. EntwMech. Org. 124: 1931b Die bischerigen Ergebnisse der Strahlengenetik. Erg. d. med. Strahlenforschung, Bd. 5. Leipzig: VerLThieme. 1932 Verschiedenheit der "normalen" Allele der white-Serie aus zwei geographisch getrennten Populationen von Drosophila melanogaster. Biol. Zbl. 52: VOGT,C., und VOGT,O., 1929 Uber die Neuheit und den Wert des Pathoklisenbegriffes. J. Psychol. u. Neurol. 38: WHITING,P. W., 1932 Mutations in Habrobracon. Genetics 17: ZELENY,CH., 1921 The direction and frequency of mutations in the Bar-eye series of multiple allelomorphs in Drosophila. J. Exp. Zool. 34:

INTERNATIONAL CONGRESS O F GENETICS

331

T H E PROCESS O F EVOLUTION I N CULTIVATED P L A N T S N . I. Vavilov, Institute of Plant Industry, Leningrad, Union of Socialistic Soviet Republics

One of the most essential factors in understanding the process of evolution in living organisms is the geographical distribution of species and varieties at the present time and in the past. Several regions of the globe are extremely rich a t the present time in numbers of species and varieties of plants and animals. In North and Central America such regions are Southern Mexico, Guatemala and some adjacent small countries to the south. In Europe such regions are the Caucasus, the Balkan States, Italy and Spain. Huge spaces in Northern and Central Asia are on the contrary quite poor in the number of species. On the other hand; Southeastern China, India, Indo-China and the mountainous regions of Persia, Afghanistan, Russian Turkestan and Asia Minor are extremely rich in the variety of existing species of wild animals and plants. The same applies to cultivated plants and domesticated animals. During OF PLANT INDUSTRY in the Union of Sothe last ten years the INSTITUTE cialistic Soviet Republics has carried on extensive botanical and geographical investigations of a great number of cultivated plants on a world-wide scale. The principal purpose of these researches has been to ascertain the diversity of varieties of cultivated plants by applying modern methods of differential taxonomy and botanical geography. Separate plants have been subjected to a detailed cytogenetical and biochemical study. In undertaking this work we have been prompted by the actual needs of plant breeding, which is one of the most pressing practical problems in the large-scale socialistic agriculture in our country a t the present time. It was evident to us that in order to conduct any rational large-scale work in plant breeding it was necessary to master exhaustively'the whole initial varietal potentialities of the world as well as to. learn about the nearest wild relatives of our useful plants. Special attention has been devoted to the regions of the primary origin of cultivated plants, that is, to ancient agricultural countries almost untouched by previous investigators. In this respect the Soviet Union is in an exceptionally favorable condition since a vast initial material of varieties of many of the most important cultivated plants of the Old World is concentrated within the Caucasus and Turkestan, as well as in adjacent oriental countries. Even at present millions of acres of forests may be seen in these regions (especially in the Caucasus) consisting chiefly of wild apples, pears, plums, sweet cherries, wild almonds, and wild pomegranates. Here it is possible to trace the evolution of cultivated plants from wild form step by step. In conformity with a definite plan, numerous expeditions have been sent

332

PROCEEDINGS O F THE SIXTH

to different regions of the Soviet Union, as well as to many of the ancient agricultural countries, including Abyssinia, Aighanistan, Asia Minor, Persia, Western China, all countries bordering the Mediterranean and Mexico, Guatemala, Peru, Bolivia, Columbia and Chile. At present researches are being extensively carried on upon over 300 different cultivated species. The huge new material on cultivated plants and their wild relatives, collected in all countries of the world and consisting of not less than 300,000 specimens, has been subjected to a detailed study at the stations, and many results were published recently in the.forrn of comprehensive monographs containing a survey of the world diversity of separate crops and showing their distribution over the different regions o i the globe as well as the amplitude of their variation in quantitative and qualitative characters. Not only taxonomists are engaged in the study of cultivated plants, but plant breeders, geneticists, physiologists and biochemists have joined in this work. Nurseries have been established containing all varieties collected throughout the worlcl, each nursery in the corresponding climate of the Union of Socialistic Soviet 12epublics. J'xhibits which were sent to the Sixth International Congress of Genetics by our institution will give some idea of the results of these investigations. 'The purpose of this briei con~municationis to formulate on the basis o f our studies some general statements which may be distinctly traced in the evolution of cultivated plants. A S P E C I E S A S A COMPLEX VAIIIABLE A N D M O B I L E S Y S T E M

Knowing the great diversity of cultivatetl plants ant1 of their wild relatives we have been lecl to conceive of Linnean species as of actually existing real complexes. In our judgment they represent rnobile and variable systems embracing categories of different amplitude and connected in their historical development with their habitat and area. At a given geological period and at a given moment of observation, the species have a real existence, some opinions to the contrary notwithstanding. The separation among the species and their divergence are not merely in the investigator's mind. A n intrinsic unity of discontinuity and continuity is a characteristic of the eyelution of living organisms. According to a detailed study of a great number of cultivated plants a Linnean species represents a distinct complex and mobile morphophysiological system bound in its genesis to a definite environment and area. Herbarium specimens of species, on the basis of which the present recognized botanical classifications have been drawn up, are only a first approach

INTERNATIONAL CONGRESS O F GENETICS

333

toward the study of a species. The real species with which the geneticist and the plant breeder have to deal are by far more complicated. Actual investigations of a number of species in their whole varietal diversity and over the whole world have led us to the discovery of great varietal potentialities unknown up to now to the botanist, plant breeder and geneticist. T H E GEOGRAPHICAL PRINCIPLE I N EVOLUTION

DARWINin his researches devoted much attention to the geographical principle in evolution. A mere knowledge of general boundaries of the habitat of a species with which the botanist and the geographer usually have to deal does not satisfy either the plant breeder or the geneticist. For us it is quite important to know the localization of the definite genotypes within the limits of a species and the geographical distribution of genes within the area of a species. It has become at once evident that this principal factor in the understanding of the evolution of cultivated plants had practically been left out of consideration. Even in regard to the most important cultivated plants the botanist, the plant breeder and geneticist have had to content themselves with isolated specimens of species, without knowing the sum total of the varieties constituting the species. A study of one plant after another has led us to the discovery of regions characterized by a striking intraspecific diversity, hitherto never suspected by the botanist or the plant breeder, to say nothing of the geneticist. Even for such plants as wheat we have discovered regions containing quite exceptional riches of new varietal character up to now unknown to the plant breeder. Thus little Abyssinia has proved to contain more than half of the varietal diversity of cultivated wheat found in the world. As a result of our investigations of field and truck crops, as well as of fruit trees and shrubs, a new immense primary diversity of varieties has been found. Moreover, for a series of cultivated plants this study has led to the discovery of a great number of new Linnean species. Thus during the past five years we have succeeded in discovering five new wheat species each represented by a great diversity of forms. These species were found in Transcaucasia and in Abyssinia. OF PLANTINDUSTRY sent out in 1926 to Expeditions of the INSTITUTE Mexico and to Central and South America, which worked there for three years, have revealed that up to the present we have had no real knowledge of the potato. It has been proved by morphological and physiological methods and by cytogenetical analysis that instead of fragments of one speciesSolanurn Tuberosu'm--known to plant breeders, there exist not less than

334

PROCEEDINGS O F THE S I X T H

14 good species of the cultivated potato grown by the natives of Peru, Bolivia, Chile and Columbia. The whole series of polyploid species with 24, 36, 48, 60 and 70 chromosomes was discovered. These species are well differentiated, each characterized by a different number of chromosomes, by their crossability, their morphological and physiological qualities and their area of distribution. Very likely there are even more than these 14 species of potato in the regions of the Andes. The whole group of wild species of potato in Mexico and Guatemala proved to be immune to Phytophtora infestans. Similar facts have been established for several fruit trees in Transcaucasia. Such an extension of our knowledge of species naturally will create a new basis for the comprehension of evolution and make us alter our own conceptions of the species themselves. A detailed analysis of the geographical localization of evolution has not received due attention up to this time. The predominant influence of the methods of archaeology and history in the study of the process of evolution is responsible for the fact that the biological substance in this process has been ignored. A wide field of new, interesting work, full of practical purpose, opens before the geneticist and will lead him to a better knowledge of evolution. The detailed study of the varietal diversity of cultivated plants and their wild relatives has enabled us to establish for a majority of them the regions of their origin. These regions have proved to be confined to comparatively small territories concentrated mainly in the mountains and at the foothills in the subtropics and tropics. At the Fifth International Congress of Genetics in Berlin the author outlined five principal world centers in which the most important cultivated plants have originated. On the basis of a large amount of supplementary data obtained during the last five years, we have been enabled to locate more exactly the regions of the origin of cultivated plants. W e distinguish at present seven principal world centers of the origin of cultivated plants: 1. Southwestern Asia including Transcaucasia and the northwestern portion of India originated soft wheats and rye as well as many grain Leguminosae, alfalfa, Persian clover, etc. Here, especially in the western part of this area, is the home of the most important fruit trees. 2. India is the native country of rice, sugar cane and many tropical plants. 3. The mountains and foothills of Eastern China are the home of many fruit trees, truck crops and the soybean. The vast regions of Central Asia,

INTERNATIONAL CONGRESS O F GENETICS

335

investigated by us in detail in 1929, have proved alien to the primary process ons, of form origination. In spite of some former botanical ~ ~ ~ ~ o s i t iCentral Asia and Siberia have had no influence upon the origin of cultivated plants. 4. Abyssinia, though economically a country of no particular importance with its cultivated area of only several million acres, shows a striking concentration of the diversity of the gcnes of wheat, harley and many leguminous grain crops. 5. Certain countries bordering the Mediterranean are the home of the olive tree, the carol tree, a series of original forage plants and Egyptian clover. The sixth and seventh centers must be sought in America. I n the New World the primary process of form origination is narrowly localized; the regions showing a striking species and varietal diversity occupy comparatively small territories concentrated in Southern Mexico and Central America as well as in Peru and Bolivia. The home of corn and of the upland cotton in all probability is Mexico and Central America, whereas that of the potato is in Peru and Bolivia. These seven centers have developed on the basis of an extremely rich wild flora. Here we find conditions especially favorable to the development of species and varietal diversity. These regions have proved equally favorable to civilizations and of course it is no accident that the map showing the distribution of the chief sources of food plants essentially coincides with that of the distribution of the first agricultural civilizations. The mountain and foothill regions in the subtropics are the most retnarkable places for comprehending the evolution of cultivated plants as well as of many wild species. In these seven regions the beginnings of the evolutionary process rnanifest themselves in a salient way especially when we compare the evolution of different species and genera. The existence of such group evolution of different species and genera facilitates greatly an understandir~gof the evolutionary process. T H E LAW O F IIOMOLOGOUS SERIES I N VARIATION

The phenomenon of parallel variation, that is, of the regular repetition of series of characteristics in varieties in different allied species and genera, is quite general. In our studies we have found thousands upon thousands of examples of this parallelism in different species belonging to the same genus, or in different genera belonging to the same family. Such a parallelism is remarkable, for example, when we compare the varietal diversity of wheat,

336

PROCEEDINGS O F THE SIXTH

barley and rye, or of peas, chick peas, lentils, vetches and horse beans. Every month brings new evidence of this astonishing parallelism which can not be neglected in the study of the evolutionary process. The author has named this parallel variation, "The law of homologous series in variation" (Journal of Genetics, 1922). This law is actually a development of the idea of evolution and helps considerably in the study of the varietal diversity and of the same process of evolution. Of course, this parallelism should not be taken as something absolute. But on the whole it is quite general and manifested by all groups of cultivated and wild species studied by us in detail. For a great majority of the plants we can speak only about phenotypical parallelism, but there are already known many cases of experimentally established facts of genotypical parallelism. A great majority of varietal links missing in the species were found in the primary seven centres of the origin of cultivated plants. REGULARITIES I N T H E DISPERSION O F CULTIVATED PLANTS FROM THEIR PRIMARY REGIONS

The above described primeval regions are characterized by a vast potentiality of genes and not infrequently (though not always) by a great diversity of forins. A study of the genetical diversity of the most important cultivated plants in such primeval regions has divulged the presence of a great number of the dominant genes along with the recessive ones. Many dominant forms have been revealed by us for the first time in these primary regions of the origin of a given plant. Thus, for instance, dominant violetgrained, almost black-grained wheats have been found only in Abyssinia. The regions of Abyssinia, the region enclosed between the Hindu-Kush and the Himalaya, as well as Asia Minor, Southern Mexico and Guatemala are characterized by an extremely great number of dominant forms of cultivated plants. Central Peru is distinguished by an enormous diversity of cultivated potatoes and especially by the presence of dominant genes. An actual investigation has shown that during the spreading of the species toward the boundary of a region chiefly recessive forms are singled out and survive. As we withdraw from the primary regions, a decrease of dominant forms and a predominance of recessive forms are observed. The process of the separation of recessive forms is evidently connected with frequent mutations toward recessiveness which, as is well known, usually take place. For instance, ERWINBAURand his collaborators have established about 300 different mutations in Anterrhinunt Majus, among which only 9 or 10 are dominant, and all others are definite recessives. T o a

INTERNATIONAL CONGRESS O F GENETICS

337

considerable degree this separation of recessives in the process of intraspecies evolution is also a result of inbreedings. The spreading out of forms, their geographical isolation, is particularly favorable for the singling out of recessives. The trend of evolution within a species toward recessives is a common phenomenon. In peas, corn, rice, sweet peas and barley, mostly studied from a genetical point of view, the mutations are almost exclusively recessive in relation to the wild Cilician forms. In maize, far from its Mexican primary centre, we have found exclusively new recessive types like ligules, waxy and ramose varieties as well as fornis with glossy seedlings (N. KULESHOV). The opposite process, that is, the singling out of dominant forms, though observed is of very rare occurrence. Quite frequently it manifests itself in a peculiar way. Thus for example the appearance of such a dominant beardless barley is observed in Eastern Asia and in Japan, far away from the principal centres of the origin of cultivated barley in Southwestern Asia and Abyssinia. The dominant hornless varieties of cattle are known in great numbers in the extreme north of European and Asiatic Russia. Biologically such forms are unquestionably secondary ones, as hornless cattle, awnless barleys, and awnless wheats may exist only under conditions of cultivation. The dominant vegetative mutations occur frequently in the potato ( T . V. ASSEJEVA).I t is possible to say that as a rule cultivated types of plants are mostly recessive. A series of peculiar, as if synthetic, forms have been found in a number of cultivated plants in primeval regions. These forms show an absence of the divergence of characteristics and combine the qualities of different species, showing thereby that this phenomenon is without relation to hybridiza tion. Such forms are of considerable interest in the study of the process of evolution. Thus, for instance, wheats of Abyssinia combine the properties of soft and of durum wheats. THE INCREASE O F THE SIZE O F SEEDS, FRUITS AND FLOWERS I N A DEFINITE GEOGRAPHICAL DIRECTION

In the geographical evolution of cultivated plants interesting general facts have been observed. Thus one of the most noteworthy relationships is a regularly observed, gradual increase in the size of seeds, fruits and flowers from the Himalaya region toward the Mediterranean. In extreme cases, the linear dimensions are consistently increased several times. The varieties of flax which originated in Morocco, Algeria and Cyprus

338

PROCEEDJXGS O F THE SIXTH

are twice as large as those of Persia and Afghanistan. In the Mediterranean countries the seeds of beans, lentils, and chick peas exceed in size several times those of Indian forms. A t any rate the fact of accumulation in the Mediterranean region of exceptionally large-fruited and large-seeded forms of cultivated as well as of certain wild species is incontestable. Perhaps these are extreme cases of recessives, or they may be results of hybridization. A t any rate the genetic nature of this increase in size is as yet unknown, but the fact itself must be taken into consideration by the plant breeder. Another regularity in regard to cereals has been observed in Eastern Asia, where an apparently simultaneous development of naked oats, naked millet and naked barley has taken place. Some regions are characterized by a number of species showing a definite trend toward some particular variation which is of great importance t o those in search of initial material for the purposes of plant breeding. As a n illustration, Peru and Bolivia are characterized by many tuber-bearing species and genera, such as oca, Tropaeolum, Ulluco, etc., besides potatoes. SEPARATION O F ECOLOGICAL GROUPS

The process of differentiation of particular ecological types under the influence of a definite environment may be convincingly established by studying various geographical groups within the limits of a species. A knowledge of such ecological groups gllows one to comprehend the process of geographical evolution, so important for the mastery of the diversity of types by the plant breeder. Not infrequently such groups of plants are characterized by a large number of features in common. Such groups doubtless are determined by a great number of genes. In the long historical process of the migration of species from the primary regions this crystallization of specific ecological types may be discerned clearly. THE R ~ L EO F N A T U R A L SELECTION

Doubtless an important r6le in the evolution of cultivated plants has been played by natural selection. Thus, for instance, some of the most important European crops, such as rye and oats, have been forced irrespective of man's will into cultivation as a result of natural selection. It may be observed in the Caucasus even now that a weed rye gradually supplants wheat. By way of natural selection, a s cultivation of wheat moves northward into more severe climatic conditions, rye from a weed gradually becomes an independent useful crop. Northern countries cultivate now chiefly the former weeds, namely rye and oats.

INTERNATIONAL CONGRESS O F GENETICS

339

In certain specific cultivated plants, for instance flax, the formative rble played by natural selection may be clearly seen. W e have ascertained for flax that the decisive factor in the geographical distribution of this crop is the length of the vegetative period. Due to a shorter vegetative period in the North, earliest forms were selected for cultivation there. The character of earliness in flax is t o a considerable degree correlated with the length of its stem, in the sense that earlier varieties grow taller and vice versa; hence, it is in the North that the bulk of long-stemmed fiber flaxes were selected. As a consequence flax is grown for fiber in the North and for oil in the South. This example is quoted to show that a differentiation of the crop has been above all a result of natural selection. THE R ~ L EO F M A N I N THE EVOLUTION O F CULTIVATED PLANTS

Undoubtedly an important r6le in the evolution of plants has been played by man. The destiny of many cultivated plants is intimately connected with the history of human civilizations and migrations of peoples. In some plants in particular the r6le played by man in their modification has been quite conspicuous. The influence of such great sedentary ancient civilizations as that of China and the Mediterranean on the modification of initial primitive plants was extremely great. Giant forms have been obtained by selection of extreme recessives and mutations, not infrequently surprising in their contrast with wild primitives. I t is interesting to note that some of the most peculiar cukivated types were selected by agriculturists of high, old civilizations. Varieties of rice, barley, cabbage, fruits, and radishes indigenous in China, Japan and the Mediterranean countries indicate considerable creative work on the part of in some European types of cultivated plant breeders. The same is n~tice~lble plants. On the other hand, in the majority of cases cultivated varieties of plants of India and Afghanistan as well as Bolivia and Peru do not differ greatly from the corresponding wild forms. Numerous transitional forms may be readily observed. A systematic study of quantitative variation in a vast material affords facts for the comprehension of the range of differences of the extreme types. Thus the best European pear reaches a weight of seven pounds, while the fruits of wild pear trees in Caucasus weigh but one gram. All quantitative transitions may be traced in certain cultivated plants. Some of the extreme cases biologically are cultivated monstrosities, not vital under natural conditions. It suffices to mention different seedless varieties of fruits and double flowers.

340

PROCEEDINGS O F THE SIXTH LYSSENKO'S DISCOVERY

The remarkable discovery recently made by T. D. LYSSENKO of Odessa opens enormous new possibilities to plant breeders and plant geneticists of mastering individual variation. He found simple physiological methods of shortening the period of growth, of transforming winter varieties into spring ones and late varieties into early ones by inducing processes of fermentation in seeds before sowing them. LYSSENKO'S methods make it possible to shift the phases of plant development by mere treatment of the seed itself. The essence of these methods, which are specific for different plants and different variety groups, consists in the action upon the seed of definite combinations of darkness (photoperiodism), temperature and humidity. This discovery enables us to utilize in our climate for breeding and genetic work tropical and sub-tropical varieties, which practically amounts to moving the southern flora northward. This creates the possibility of widening the scope of breeding and genetic work to an unprecedented extent, allowing the crossing of varieties requiring entirely different periods of vegetation. THE PROBLEM O F GIANTISM I N T H E EVOLUTION O F CULTIVATED PLANTS

The study of the quantitative range of the vast world material has established the limits of hereditary variation for many species of cultivated plants. Several of these species show a striking variability. Thus in cucurbits, in root crops, hereditary variation in regard to the size of fruits and roots may be a hundred- and even a thousandfold. A tendency toward giantism in the evolution of cultivated plants, as well as of animals, may be traced quite distinctly. For comprehending the evolution of cultivated plants, this is one of the most important points on which so far very little genetic research work has been done. The evolution of many cultivated plants essentially consists in an increase of the size of fruits, seeds, and roots. An important r6le in this process has been played evidently by crossing different species and different geographical races. This most important quantitative factor in the evolution of cultivated plants and domesticated animals demands an immediate careful genetical study. Experimentally it can be tackled. This study will undoubtedly do much to further breeding work. PHENOTYPES AND GENOTYPES

In studying evolution, the chief attention of investigators was up to now concentrated on phenotypes. So far the immense diversity of species and varieties has not yet been embraced even by systematic genetic study. A

INTERNATIONAL CONGRESS O F GENETICS

341

boundless sea of experimental work opens before the geneticists. W e have mastered to a certain degree knowledge concerning evolution in space and time, as it finds expression phenotypically. The following stage is the genetics of this phenomenon. An important r6le in evolution has been played by the process of hybridization, by mutation and by inbreeding. Interspecific hybridization has played a great part in the genesis of some cultivated plants. The study of the primary regions of the origin of species has revealed a number of facts pointing to the presence of interspecific hybridization in nature, even among the wild relatives of those species. This is particularly obvious in different fruit trees and shrubs, the pear, the almond, and the blackberry. Some species of cultivated plants are evidently the result of interspecific hybridization. Recent studies conducted in the Union of Socialistic Soviet Republics have proved how great is the r6le of hybridization of distant species in the origin of several cultivated plants. A natural crossing on a large scale was established between wheat and rye, between wheat and different species of Aegilops and between several species of Agropyrum and wheat. European plums very likely originated in this way. Such might be also the origin of Triticum Persicunz. In this way too, very likely, originated tobacco which is unknown in wild nature. The phenomenon of amphidiploidy of sterile hybrids proved to be rather frequent. Hence the fertility of these distant hybrids. At Saratov recently many amphidiploids were found among hybrids of wheat and rye. Several cases of amphidiploidy were produced in hybrids of wheat and Aegilops. Artificially produced tetraploidy in cabbage as was recently shown by KARPETCHENKO in our institute, increases the crossability of distant species. Cabbage and mustard, for instanzc, can be crossed. The crossing of distinct geographical races is in many instances responsible for an increase in size and for a change of many quantitative as well as qualitative characteristics. This crossing of distant geographical races affords vast possibilities for the improvement of the existing varieties. The combination of genes of geographically distant races makes it possible to transcend the limits of the ordinary types. GENERAL CONCLUSIONS

An experimental study of the process of evolution in cultivated plants throws a light on many phases of breeding work. The range of this study is exceptionally wide, as the number of cultivated plants already amounts to not less than twenty thousand. The growing needs of civilized man and the development of industry make the introduction of new plants neces-

342

PROCEEDINGS O F T H E S I X T H

sary. The vast resources of wild species, especially in the tropics, have been practically untouched by investigation. Our studies have made it clear that in order to control evolutionary processes in cultivated plants, preliminary research should be done on a worldwide scale. In order to understand evolution and to guide our breeding work scientifically, even in application to our principal crops such as corn, wheat and cotton, we must go to the oldest agricultural countries, where the keys to the comprehension of evolution are hidden. An actual mastery of the processes of evolution which is the chief aim of genetics can be accomplished only through the combined efforts of a strong international association and through the removal of barriers impeding research in those most remarkable regions of the world. Let us hope that this time is not far off. W e are only at the start of our work in this direction and an enormous field lies ahead of us. There is enough work for all of us.

INTERNATIONAL CONGRESS O F GENETICS

343

T H E NATURE O F SEX CHROMOSOMES 0.Wilzge, Copenhugelz, Denmark On the whole, it may be said that the sex difference between males and females is of genotypical nature, and that the primary sex determination is dependent on sex chromosomes. We distinguish between the two classical main types, first set up by E. B. WILSON:(1) the Protenor type, with 2 X chromosomes in individuals of one sex and 1 X chromosome in the other; and (2) the Lygaeus type, with 2 X chromosomes in one sex, whereas the other has 1 X and 1 Y chromosome. As is well known, in some classes the male is heterogametic, while in others the female has this peculiarity; and in the latter case the sex chromosomes are often called Z and W, instead of being designated as X and Y. In passing, I wish to say it does not seem reasonable to complicate matters by using sometimes the designation X-Y, and sometimes Z-W. This distinction, it seems to me, can serve no other purpose than that of "making it harder." I wish to suggest that the designations X and Y a1w:lys be used. The fact that the males and females of the Protenor type differ only in this respect, that the male has 1 X while the female has 2 X, establishes that unquestionably we are here dealing with quantitative, and not qualitative, differences of the two sexes. Furthermore, the Drosophila experiments have demonstrated that a similar view applies to this species too. The number of X, in proportion to the number of autosomes, is the decisive factor in the sex determination of the offspring. The more the researches into the field of heredity progress, the more convincing is the evidence that the nature of the individual, implying its sex, is dependent upon the preponderance of some gene, or genes, in proportion to the others present. As already emphasized by BRIDGES, each faculty of the individual will be dependent, at least to some degree, upon ;he proportion of balance between the genes, those that accentuate and those that attenuate the establishment of a given property. Finally, the hypothesis on the significance of the gene quantities to the advanced by GOLDSCHMIDT properties of the individual, including the sex characteristics, is quite consistent with the view that the nature of the organism as a whole is determined by the balance between all the genes inherent in the organism. In this lecture the aim is to demonstrate, partly through facts that have already been published, not only that cytological observations and genetic experiments substantiate the view that the sex chromosomes are to be re-

344

PROCEEDINGS O F T H E SIXTH

garded as originating from autosomes, but also that it is practicable experimentally tq alter X chromosomes into autosomes and autosomes into X and Y chromosomes; and, finally, it will be shown that under certain conditions the difference between X and Y is merely of quantitative nature. As a rule, indeed, the X chromosome and the Y chromosome differ morphologically, though this is not always the case. In the fish Lebistes reticulatw, belonging to the Poeciliidae, it is not possible to distinguish between X and Y, and crossing over is often taking place. When, nevertheIess, X and Y keep maintaining their respective nature, the explanation of this is to be looked for undoubtedly in the fact that these two chromosomes differ only on one point, in a single gene: the Y chromosome contains an epistatic male-determining gene, which is not present in the X chromosome. Even after crossing over between X and Y, only one of the chromosomes will contain the male gene, and this chromosome does thus become a Y chromosome. It is probable that, in analogy with the conditions in Drosophila, in Lebistes, too, there are several other genes which might tend to influence the sex of the offspring. 'But the Y chromosome contains a dominant and epistatic male-determining gene which, under ordinary conditions, constitutes the sole decisive factor in the esJablishment of the male sex. Some geneticists have a different view on the sex-determination of Lebistes, especially WITSCHI,who thinks it is just as in Drosophila. I regret I have no time for discussing this today. I only want to add that I find it too dogmatic to assume that all organisms behave just as Drosophila do. In Lebistes you have more color genes in the Y than in any other chromosome, and you find crossing over between X and Y. The last fact makes it obvious that you have a special gene or gene-pair for the sex in the sex chromosomes or, as proposed by MORGAN, a special part of the sex chromosomes has the sole sex-determining effect. In a dioecious plant such as hop, Humulw lupulw, we find a typical pair of XY chromosomes in the male, and a pair of XX in the female. Here the Y chromosome is twice as large as X. Besides, the X chromosome is somewhat constricted in the middle, especially during the prophase prior to the reduction division. A priori one might think, perhaps, that in this species X and Y differ in quality, as they look entirely different. In reality, however, there can be no doubt that the Y chromosome is equal to two X chromosomes which have become united. On microscopic examination of the reduction division (figure 1) in the

INTERNATIONAL CONGRESS O F GENETICS

345

male plant of the near-related Humulus Japonicus, one may find, even in the same field of view, now an X Y pair of chromosomes where Y is twice as large as X, now a Y chromosome that is constricted in the middle, and now 3 chromosomes of the same size, that is, 3 X chromosomes. And

F ~ G U R1.-Variation E in the sex chromosome complement at the reduction division in the male Humulus Japonicus.

sometimes the reduction division proceeds in this way: the median sex chromosome goes to one pole while the two terminal sex chromosomes go to the other pole; sometimes one terminal chromosome goes to one pole, while the two adjoining chromosomes go to the other pole. In brief, one has a very distinct impression that in Humulus Japonicus the Y chromosome is equal to 2 X chromosomes. Thus, in this species too the sex de-

FIGURE 2.-T,he sex chromosome pair in the male Melandrium rdrwrn. The X is twice as large as the Y.

termination is dependent upon unequal quantities of sex-determining genes, upon a difference in the proportional balance between the genes with a female tendency and those with a male tendency. There can be no doubt then that also in Humulw lupulus the Y chromosome is composed of two X chromosomes, conjugated end to end. The

346

PROCEEDINGS OF THE SIXTH

difference between the two species of Humulus is merely this: in one of the species, evidently, the X chromosomes are conjugated permanently into one unit, the so-called Y chromosome. I n 1Melandriu.m dbum and M. rubrum the male plant shows a quite typical pair of X Y chromosomes, but here the X chromosome is twice as large as Y (figure 2). It is true that up to the present it has not been practicable to demonstrate cytologically that X is composed of 2 Y chromosomes; yet,

FICGRE 3.-A normal Melandrium (female) to the right. An "abnormal" to the left. The "abnormal" gene, n, may be present in X or in Y in Melandrium.

the genetic experiments I have carried out with this plant leave hardly any question of this being the case. Amongst the X linked and Y linked genes I have been able to demonstrate in Melandrium (the demonstration material is represented here in the exposition field) there is a recessive gene I have called "abnormal" (figure 3 ) . This gene may be present in both the X and the Y chromosome, and it is inherited in a perfectly regular manner, either X linked or Y linked (figure 4). How, one may ask, is the presence of this characteristic gene in either kind of sex chromosomes to be explained? The assumption that the

INTERNATIONAL CONGRESS O F GENETICS

347

Melandrium X is composed of 2 Y chromosomes (figure 5) offers a very simple explanation of this. Occasionally, on bipartition of X during the prophase of the reduction division in a pollen mother cell, the result will be 3 free Y chromosomes, in "end-to-end" position, and the original Y

X n X ~ norm. 0

Xn Yn nbnd

FIGURE 4.-By crossing an abnormal female Melandrium (X,X,) with a heterozygous normal male, entirely different results are obtained in proportion as the "abnormal" gene is present in the X or in the Y of the male. In the first case all daughters are abnormal, all sons normal. I n the second case the opposite result is found.

may probably enter into a new-formed X. The free half of the X chromosome, on the other hand, constitutes itself a new Y chromosome. It is a special case of translocation. Just as the X chromosomes and the Y chromosomes are here proved to

348

PROCEEDINGS OF THE SIXTH

be of identical nature, so I have succeeded in experiments with Lebistes in altering the sex chromosomes into autosomes and vice versa. The male Lebistes has a pair of X-Y chromosomes that cannot be distinguished one from the other, nor can they be distinguished from the autosomes (there are, all tolcl, 23 pairs of chromosomes). Still, the fact that nearly all the genes which give to the Lebistes males their splendid color schemes are inherited as X linked or Y linked is an indisputable evidence of the presence of sex chromosomes. It is only the Lebistes males (figures 6 and 7) that show color patterns, whereas the females are continuously greyish even though they have some

Melandrium

abnormal

abnormal

FIGURE5.-Illustrating the presence in Melandrium of the same recessive gene, n ( f o r "abnormal"), in both X and Y, X probably being composed of two Y chromosomes. genotypical faculty for coloring. In the males the color genes are dominant, and the recessive allelomorphs are absent. Here the "presence-absence theory" holds good. Figure 8 shows the combined X and Y linked inheritance of 5 genes, Coccineus, Vitellinus, Tigrinus, Luteus and Maculatus. In a couple of the races I have experimented with, however, it has not been an altogether uncommon' finding that individuals have resulted that are essentially females, though they have some tendency for the formation of a gonopod, the copuIation organ characteristic of the males, and a t the same time they have shown the color patterns corresponding to their genetic formula.

I N T E R N A T I O N A L CONGRESS O F GENETICS

349

This applies in particular to the two races characterized respectively by the genes Coccineus-Vitellinus and Tigrinus-Luteus in the X chromosomes, which races have shown a rather marked tendency to produce such masculinized fenlales (see figure 9).

FIGURE 6.-The effect of different color genes in Lebistes males. Vertical hatching indlcates red color, while dotting indicates yellow color.

In order to try whether it might be possible to produce females entirely masculinized, I have made some crossings between these two races. A homozygous female Xco, vi Xco, v i was crossed with a male XTi, Lw Y M n (figure 9). The "Ma" gene in the male Y chromosome designates the gene "Maculatus" which is linked absolutely to the Y chromosome, most likely because it is identical with the very male-determining gene in Y. I t has never shown any crossing over to the X chromosome, not even amongst

350

PROCEEDINGS OF THE SIXTH

many thousands of individuals. All the male offspring of a male Maculatus show the male Maculatus gene-for one thing, they have the characteristic black spot in the dorsal fin. The crossing experiments gave, as was to be expected, chiefly males of the formula Xc,, vi YMa(55 individuals) together with a corresponding number of colorless, or nearly colorless, females; but, in addition, there ap-

FIGURE 7.-For

explanation see figure 6.

peared 3 males, showing the genes Co, Vi and Ti, Lu, whereas they were wanting Ma (the type is shown on figure 9, below, in the middle). Obviously, these exceptional males are XX males, for if they had any Y chromosomes they would present the Maculatus gene. They had merely the color pattern the females would have possessed if the latter had been able to bring forth their content of color genes. Now it could be predicted, and it was predicted, that these exceptional

I N T E R N A T I O N A L CONGRESS OF GENETICS

351

males, which were bound to be X X individuals, would give exclusively female offspring when they were mated with ordinary females. In fact, this proved to be the case. All told, the 3 X X males produced a total of 314 young, and every one of the young was a female; there was not a single male in the lot, while in all other cases I used to have about 50 percent of each sex.

FIGURE 8.-A

combined X and Y linked inheritance of color genes in Lebistes.

T o obtain anew some X X males there was only one way open: backcrossing some of the many daughters with the X X males (figure 10). On such backcrossing, one of them gave 164 young. And they were all femalec too. But one of these 164 daughters was again backcrossed with the XX male, and now it gave 30 females and I male, which proved to be a

X C ~V ,i X T ~LY. ,

332

PROCEEDINGS O F THE SIXTH

FIGURE 9.-Two upper rows: The single effect in the Lebistes male of the genes Coccineus, Vitellinus, Maculatus, Luteus and Tigrinus, and a half-masculinized female showing colors according to her formula. Two lower rows: A cross between the C, Vi-race and the Ti L,-race, giving chiefly C,ViM,-males, as expected, but, in addition, a male type with CoVi TiL,, that is, the female formula, with two X chromosomes and no Y.

This new XX male was backcrossed with its mother, and now the outtome was males and females in equal number, though not so many that it was warrantable to give their sex ratio as exactly 50 percent of either sex. (Addition November 1932.) Counting together, however, all the material showing segregation in males and females in the new XX race, 37 males

I N T E R N A T I O N A L CONGRESS O F GENETICS

353

and 34 females are obtained, namely 5 :2, 4: 5 , 3 :3, 9:4 and 16:20, respectively. If the XX males are crossed to females of a normal race, only daughters are obtained.

FIGURE 10.-A Lebistes X X male crossed to an ordinary female gives only daughters (101). Backcrossing one of the daughters to the XX male gives again only daughters (164). Gackcrossing one of these to the XX male gives 30 females and 1 male. By backcrossing this male to his mother, males and females are segregated, apparently in a 1 : 1 ratio, and the X linked genes a r e now inherited in normal Mendelian manner. F o r further explanation see text.

As will be recognized readily, this means the establishment of a new sex determination. Now the XX chromosoines are to be looked upon merely as ordinary autosomes, identical for individuals of either sex. Accordingly,

PROCEEDINGS OF THE S I X T H

354

the genes Coccineus, Vitellinus, Tigrinus and Luteus are inherited in ordinary Mendelian manner, without being sex linked, and they have no connection whatever with the sex determination. The X chromosomes are modified into autosomes.

Lebistes X and capitals pulling in jemale direction Y and small letters pulllny in male direction

I I I

Co M-X and Ti Lu -X weakly female

New sexdetermination realized

A-a now being the sex chmmo~omepair

I

I

FIGURE 11.-Explanation of the transition from the normal Lebistes with XX in females and X Y in males, to the new type, with Aa in females and aa in males (both sexes having now X X ) . In this new type A must be regarded as a new "Y" chromosome, and a as a new " X . The formulas for the ordinary Lebistes, as distinguished from the c,V and TiLu races, have about the same number of autosomal genes pulling in male and in female direction, while the C,Vi and TiL,, races are supposed to have a surplus of autosoma1 genes pulling in male direction. By accidental accumulation of the male genes in certain individuals the development may be in male direction even in XX individuals. Nevertheless, the mechanism of sex determination is brought forth through the outlined backtrossing to XX males, whereby a new pair of chromosomes have taken over the leading role in the sex determination, a pair of chromosomes which were of no particular significance in this respect as long as the ordinarily effective function of the differentiating XX-XY mechanism was maintained.

INTERNATIONAL CONGRESS O F GENETICS

355

On backcrossing to X X males there is a selection of autosomal genes which have a male-determining tendency; and, theoretically, there is no reason why female heterogamy should not result, so that, in future, the difference between the females and the males is that the females are heterozygous (Aa) as to a female determining gene A, while the males are of the homozygous recessive formula (aa), both sexes having XX. I t is obvious that this will result in a new sex linked ( X and Y linked) inheritance of genes in the A-a chromosome pair (figure 11). However, at present it is impossible to establish whether, in this new pair of chromosomes, the male or the female is the heterogametic. A priori, the chances are about equal. In this way it is evident that it is possible, within a given species, to cross from male to female heterogamy. Thus it is less surprising that Lebistes shows male heterogamy, while the near-related Platypoecilus shows female. heterogamy, as has been pointed out by BELLAMY(1922), by GORDON, and others. Naturally, it is to be expected, when the sex-determining genes of lower order take on the leading role, there will exist the possibility that external conditions may influence the sex determination. In the near-related Xiphophorus Helleri, indeed, the sex determination appears to be very uncertain, so uncertain, even, that it has been claimed it is entirely phenotypical. I know, however, from my own experience with Xiphophorus that this is carrying the point to excess. It has been my main attempt to show how the view concerning quantitative differences may be applied at least in some cases on the X and the Y chromosomes, and to show that even the difference between sex chromosomes and autosomes, as to their sex-determining ability, is of quantitative nature, and that sex chromosomes experimentally can be changed into autosomes and vice versa.

356

PROCEEDINGS OF T H E S I X T H

T H E ROLES O F MUTATION, INBREEDING, CROSSBREEDING A N D SELECTION I N EVOLUTION Sewall Wright, University o f Chicago, Chicago, Illinois

The enormous importance of biparental reproduction as a factor in evolution was brought out a good many years ago by EAST.The observed properties of gene mutation-fortuitous in origin, infrequent in occurrence and deleterious when not negligible in effect-seem about as unfavorable as possible for an evolutionary process. Under biparental reproduction, however, a limited number of mutations which are not too injurious to be carried by the species furnish an almost infinite field of possible variations through which the species may work its way under natural selection. Estimates of the total number of genes in the cells of higher organisms range from 1000 up. Some 400 loci have been reported as having mutated in Drosophila during a laboratory experience which is certainly very limited compared with the history of the species in nature. Presumably, allelomorphs of all type genes are present at all times in any reasonably numerous species. Judging from the frequency of multiple allelomorphs in those organisms which have been studied most, it is reasonably certain that many different allelomorphs of each gene are in existence at all times. With 10 allelomorphs in each of 1000 loci, the number of possible combinations is 10IWO which is a very large number. I t has been estimated that the total number of electrons and protons in the whole visible universe is much less than 10IW. However, not all of this field is easily available in an interbreeding population. Suppose that each type gene is manifested in 99 percent of the individuals, and that most of the remaining 1 percent have the most favorable of the other allelomorphs, which in general means one with only a slight differential effect. The average individual will show the effects of 1 percent of the 1000, or 10 deviations from the type, and since this average has a standard deviation of ~ 1 only 0 a small proportion will exhibit more than 20 deviations from type where 1000 are possible. The population is thus confined to an infinitesimal portion of the field of possible gene conibinations, yet this portion includes some loM homozygous combinations, on the above extremely conservative basis, enough so that there is no reasonable chance that any two individuals have exactly the same genetic constitution in a species of millions of millions of individuals persisting over millions of generations. There is no difficulty in accounting for the probable genetic uniqueness of each individual human being or other organism which is the product of biparental reproduction.

INTERNATIONAL CONGRESS O F GENETICS

,337

If the entire field of possible gene combinations be graded with respect to adaptive value under a particular set of conditions, what would be its nature? Figure 1 shows the combinations in the cases of 2 to 5 paired allelomorphs. In the last case, each of the 32 homozygous combinations is at one remove from 5 others, at two removes from 10, etc. I t would require 5 dimensions to represent these relations symmetrically; a sixth dimension is needed to represent level of adaptive value. The 32 combina-

FIGURE 1.-The combinations of from 2 to 5 paired allelomorphs.

tions here compare with 10'000in a species with 1000 loci each represented by 10 allelomorphs, and the 5 dimensions required for adequate representation compare with 9000. The two dimensions of figure 2 are a very inadequate representation of such a field. The contour lines are intended to represent the scale of adaptive value. One possibility is that a particular combination gives maximum adaptation and that the adaptiveness of the other combinations falls off more or less regularly according to the number of removes. A species whose individuals are clustered about some combination other than the highest would

358

PROCEEDINGS O F THE SIXTH

move up the steepest gradient toward the peak, having reached which it would remain unchanged except for the rare occurrence of new favorable mutations. But even in the two factor case (figure 1 ) it is possible that there may be two peaks, and the chance that this may be the case greatly increases with each additional locus. With something like 1O1Oo0possibilities (figure 2) it may be taken as certain that there will be an enormous number of

FIGURE2.-Diagrammatic representation of the field of gene combinations in two dimensions instead of many thousands. Dotted lines represent contours with respect to adaptiveness.

widely separated harmonious combinations. The chance that a random combination is as adaptive as those characteristic of the species may be as low separate peaks, each surrounded by as 10-looand still leave room for 10800 10100more or less similar combinations. In a rugged field of this character, selection will easily carry the species to the nearest peak, but there may be innumerable other peaks which are higher but which are separated by "valleys." The problem of evolution as I see it is that of a mechanism by which the species may continually find its way from lower to higher peaks in such

359

INTERNATIONAL CONGRESS O F GENETICS

a field. In order that this may occur, there must be soille trial and error mechanism on a grand scale by which the species may explore the region surrounding the small portion of the field which it occupies. T o evolve, the species must not be under strict control of natural selection. Is there such a trial and error mechanism ? A t this point let us consider briefly the situation with respect to a single locus. In each graph in figure 3 the abscissas represent a scale of gene frequency, 0 percent of the type genes to the left, 100 percent to the right. The elementary evolutionary process is, of course, change of gene frequency, a DISTRIBUTION OF GENE FREQUENCIES A.

y = c e 4 ~ S XX4NV-I

W h o l c Species

S~~eols

(I-~)4NU-I

gene frequency

Y C

N 5 V,U

4NU,4NS

4NU,4N5 very

very Icqe

small

probabil~t~ coeffccient popul&.on number s e l e c t i o n coefficccnt m u t a t t o n raTes t o and

,

f r o m gene r e ~ ? e c J ~ u e l ~ , pcr qenernl~on

gene frequency

y

prob6brl;ty

C

coc{ftc;ent p o y u l n t ~ o n number

n 5 X=

o 4nn

2 *cry

larye

FIGURE 3.-Random

1

0

4nm

x medium

r

o

ic r 4nm

m

s e l e c t ~ o n coeff~clent p o p u l a t ~ o n exchange w ~ r h r e s l o{ s p e c i e s

verq moll

variability of a gene frequency under various specified conditions.

practically continuous process. Owing to the symmetry of the Mendelian mechanism, any gene frequency tends to remain constant in the absence of disturbing factors. If the type gene mutates a t a certain rate, its frequency tends to move to the left, but at a continually decreasing rate. The type gene would ultimately be lost from the population if there were no opposing factor. But the type gene is in general favored by selection. Under selection, its frequency tends to move to the right. The rate is greatest at some point near the middle of the range. At a certain gene frequency the opposing pressures are equal and opposite, and at this point there is consequently equilibrium. There are other mechanisms of equilibrium among evolutionary factors which need not be discussed here. Note that we have here a theory

360

PROCEEDINGS O F THE SIXTH

of the stability of species in spite of continuing mutation pressure, a continuing field of variability so extensive that no two individuals are ever genetically the same, and continuing selection. If the population is not indefinitely large, another factor must be taken into account: the effects of accidents of sampling among those that survive and become parents in each generation and among the germ cells of these, in other words, the effects of inbreeding. Gene frequency in a given generation is in general a little different one way or the other from that in the preceding, merely by chance. In time, gene frequency may wander a long way from the position of equilibrium, although the farther it wanders the greater the pressure toward return. The result is a frequency distribution within which gene frequency moves a t random. There is considerable spread even with very slight inbreeding and the form of distribution becomes U-shaped with close inbreeding. The rate of movement of gene frequency is very slow in the former case but is rapid in the latter (among unfixed genes). In this case, however, the tendency toward complete fixation of genes, practically irrespective of selection, leads in the end to extinction. In a local race, subject to a small amount of crossbreeding with the rest of the species (figure 3, lower half), the tendency toward random fixation is balanced by immigration pressure instead of by mutation and selection. In a small sufficiently isolated group all gene frequencies can drift irregulary back and forth about their mean values at a rapid rate, in terms of geologic time, without reaching fixation and giving the effects of close inbreeding. The resultant differentiation of races is of course increased by any local differences in the conditions of selection. Let us return to the field of gene combinations (figure 4). In an indefinitely large but freely interbreeding species living under constant conditions, each gene wilI reach ultimately a certain equilibrium. The species will occupy a certain field of variation about a peak in our diagram (heavy broken contour in upper left of each figure). The field occupied remains constant although no two individuals are ever identical. Under the above conditions further evolution can occur only by the appearance of wholly new (instead of recurrent) mutations, and ones which happen to be favorable from the first. Such mutations would change the character of the field itself, increasing the elevation of the peak occupied by the species. Evolutionary progress through this mechanism is excessively slow since the chance of occurrence of such mutations is very small and, after occurrence, the time required for attainment of sufficient frequency to be subject to selection to an appreciable extent is enormous.

INTERNATIONAL CONGRESS O F GENETICS

361

The general rate of mutation may conceivably increase for some reason. For example, certain authors have suggested an increased incidence of cosmic rays in this connection. The effect (figure 4A) will be as a rule a spreading of the field occupied by the species until a new equilibrium is reached. There will be a n average lowering of the adaptive level of the species. On the other hand, there will be a speeding up of the process discussed above, elevation of the peak itself through appearance of novel favorable mutations. Another possibility of evolutionary advance is that the spreading of the field occupied may go so far as to include another and

A. Increased I\l\utation 0. Increased Selection or reduced Selection

4NU, 4NS very larqe

D. Close

Inbreeding small

4NU,4NS very

reduced M u t o t i o n 4NU, 4NS very larqe

or

E. sli h t

Inbreeding

4 ~ t . ncdivrn 4 ~ ~

C. Qualitative Chanqe of

Environment

4NU,+NS very larqc

F. Division

into local Races

4 n m medium

FIGURE 4.-Field of gene combinations occupied by a population within the general field of possible combinations. Type of history under specified conditions indicated by relation to initial field (heavy broken contour) and arrow.

higher peak, in which case the species will move over and occupy the region about this. These mechanisms do not appear adequate to explain evolution to an important extent. The effects of reduced mutation rate (figure 4B) are of course the opposite: a rise in average level, but reduced variability, less chance of novel favorable mutation, and less chance of capture of a neighboring peak. The effect of increased severity of selection (also 4B) is, of course, to increase the average level of adaptation until a new equilibrium is reached. But again this is a t the expense of the field of variation of the species and

362

PROCEEDINGS O F T H E S I X T H

reduces the chance of capture of another adaptive peak. The only basis for continuing advance is the appearance of novel favorable mutations which are relatively rapidly utilized in this case. But at best the rate is extremely slow even in terms of geologic time, judging from the observed rates of mutation. Relaxation of selection has of course the opposite effects and thus effects somewhat like those of increased mutation rate (figure 4A). The environment, living and non-living, of any species is actually in continual change. In terms of our diagram this means that certain of the high places are gradually being depressed and certain of the low places are becoming higher (figure 4C). A species occupying a small field under influence of severe selection is likely to be left in a pit and become extinct, the victim of extreme specialization to conditions which have ceased, but if under sufficiently moderate selection to occupy a wide field, it will merely be kept continually on the move. Here we undoubtedly have an important evolutionary process and one which has been generally recognized. I t consists largely of change without advance in adaptation. The mechanism is, however, one which shuffles the species about in the general field. Since the species will be shuffled out of low peaks more easily than high ones, it should gradually find its way to the higher general regions of the field as a whole. Figure 4D illustrates the effect of reduction in size of population below a certain relation to the rate of mutation and severity of selection. There is fixation of one or another allelomorph in nearly every locus, largely irrespective of the direction favored by selection. The species moves down from its peak in an erratic fashion and comes to occupy a much smaller field. In other words there is the deterioration and homogeneity of a closely inbred population. After equilibrium has been reached in variability, movement becomes excessively slow, and, such as there is, is nonadaptive. The end can only be extinction. Extreme inbreeding is not a factor which is likely to give evolutionary advance. With an intermediate relation between size of population and lllutation rate, gene frequencies drift at random without reaching the complete fixation of close inbreeding (figure 4 E ) . The species moves down from the extreme peak but continually wanders in the vicinity. There is some chance that it may encounter a gradient leading to another peak and shift its allegiance to this. Since it will escape relatively easily from low peaks as compared with high ones, there is here a trial and error mechanism by which in time the species may work its way to the highest peaks in the general field. The rate of progress, however, is extremely slow since change of gene

INTERNATIONAL CONGRESS O F GENETICS

363

frequency is of the order of the reciprocal of the effective population size and this reciprocal must be of the order of the mutation rate in order to meet the conditions for this case. Finally (figure 4F), let us consider the case of a large species which is subdivided into many small local races, each breeding largely within itself but occasionally crossbreeding. The field of gene combinations occupied by each of these local races shifts continually in a nonadaptive fashion (except in so far as there are local differences in the conditions of selection). The rate of movement may be enormously greater than in the preceding case since the condition for such movement is that the reciprocal of the population number be of the order of the proportion of crossbreeding instead of the mutation rate. With many local races, each spreading over a considerable field and moving relatively rapidly in the more general field about the controlling peak, the chances are good that one at least will come under the influence of another peak. If a higher peak, this race will expand in numbers and by crossbreeding with the others will pull the whole species toward the new position. The average adaptiveness of the species thus advances under intergroup selection, an enormously more effective process than intragroup selection. The conclusion is that subdivision of a species into local races provides the most effective mechanism for trial and error in the field of gene combinations. It need scarcely be pointed out that with such a mechanism complete isolation of a portion of a species should result relatively rapidly in specific differentiation, and one that is not necessarily adaptive. The effective intergroup competition leading to adaptive advance may be between species rather than races. Such isolation is doubtless usually geographic in character at the outset but may be clinched by the development of hybrid sterility. The usual difference of the chromosoille complements of related species puts the importance of chromosome aberration as an evolutionary process beyond question, but, as I see it, this importance is not in the character differences which they bring (slight in balanced types), but rather in leading to the sterility of hybrids and thus making permanent the isolation of two groups. How far do the observations of actual species and their subdivisions conform to this picture? This is naturally too large a subject for more than a few suggestions. That evolution involves nonadaptive differentiation to a large extent at the subspecies and even the species level is indicated by the kinds of differences by which such groups are actually distinguished by systematists. It

364

PROCEEDIKGS O F TIlE SIXTH

is only at the subfamily and family levels that clear-cut adaptive differences JACOT).The principal evolutionary mechanism become the rule (ROBSON, in the origin of species must thus be an essentially nonadaptive one. That natural species often are subdivided into numerous local races is indicated by many studies. The case of the human species is most familiar. Aside from the familiar racial differences recent studies indicate a distribution of frequencies relative to an apparently nonadaptive series of allelomorphs, that determining blood groups, of just the sort discussed above. I scarcely need to labor the point that changes in the average of mankind in the historic period have come about more by expansion of some types and decrease and absorption of others than by uniform evolutionary advance. During the recent period, no doubt, the phases of intergroup competition and crossbreeding have tended to overbalance the process of local differentiation, but it is probable that in the hundreds of thousands of years of prehistory, human evolution was determined by a balance between these factors. Subdivision into numerous local races whose differences are largely nonadaptive has been recorded in other organisms wherever a sufficiently detailed study has been made. Among the land snails of the Hawaiian Islands, GULICK(sixty years ago) found that each mountain valley, often each grove of trees, had its own characteristic type, differing from others in "nonutilitarian" respects. GULICKattributed this differentiation to inbreedhas found a similar situation in the land ing. More recently CRAMPTON snails of Tahiti and has followed over a period of years evolutionary changes which seem to be of the type here discussed. I may also refer to garter snakes by RUTHVEN, the studies of fishes by DAVIDSTARRJORDAN, and gall wasps by KINSEYas bird lice by KELLOGG, deer mice by OSGOOD, others which indicate the role of local isolation as a differentiating factor. Many other cases are discussed by OSBORNand especially by RENSCHin recent summaries. Many of these authors insist on the nonadaptive character of niost of the differences among local races. Others attribute all differences to the environment, but this seems to be more an expression of faith than a view based on tangible evidence. An even more minute local differentiation has been revealed when the demonstrated methods of statistical analysis have been applied. SCHMIDT the existence of persistent mean differences at each collecting station in certain species of marine fish of the fjords of Denmark, and these differences were not related in any close way to the environment. That the differences were in part genetic was demonstrated in the laboratory. DAVIDTHOMPSON

INTERNATIONAL CONGRESS O F GENETICS

365

has found a correlation between water distance and degree of differentiation within certain fresh water species of fish of the streams of Illinois. SUMNER'S extensive studies of subspecies of Peromyscus (deer mice) reveal genetic differentiations, often apparently nonadaptive, among local populations and demonstrate the genetic heterogeneity of each such group. The modern breeds of livestock have come from selection among the products of local inbreeding and of crossbreeding between these, followed by renewed inbreeding, rather than from mass selection of species. The recent studies of the geographical distribution of particular genes in livestock PHILIPTSCHENKO and others are and cultivated plants by SEREBROVSKY, especially instructive with respect to the composition of such species. The paleontologists present a picture which has been interpreted by some as irreconcilable with the Mendelian mechanism, but this seems to be due more to a failure to appreciate statistical consequences of this mechanism than to anything in the data. The horse has been the standard example of an orthogenetic evolutionary sequence preserved for us with an abundance of material. Yet MATHEW'Sinterpretation as one in which evolution has proceeded by extensive differentiation of local races, intergroup selection, and crossbreeding is as close as possible to that required under the Mendelian theory. Summing up: I have attempted to form a judgment as to the conditions for evolution based on the statistical consequences of Mendelian heredity. The most general conclusion is that evolution depends on a czrtain balance among its factors. There must be gene mutation, but an excessive rate gives an array of freaks, not evolution; there must be selection, but too severe a process destroys the field of variability, and thus the basis for further advance; prevalence of local inbreeding within a species has extremely important evolutionary consequences, but too close inbreeding leads merely to extinction. A certain amount of crossbreeding is favorable but not too much. In this dependence on balance the species is like a living organism. A t all levels of organization life depends on the maintenance of a certain balance among its factors. More specifically, under biparental reproduction a very low rate of mutation balanced by moderate selection is enough to maintain a practically infinite field of possible gene combinations within the species. The field actually occupied is relatively small though sufficiently extensive that no two individuals have the same genetic constitution. The course of evolution through the general field is not controlled by direction of mutation and not directly by selection, except as conditions change, but by a trial and error

366

PROCEEDINGS O F T H E S I X T H

mechanism consisting of a largely nonadaptive differentiation of 1ocal.races (due to inbreeding balanced by occasional crossbreeding) and a determination of long time trend by intergroup selection. The splitting of species depends on the effects of more complete isolation, often made permanent by the accumulation of chromosome aberrations, usually of the balanced type. Studies of natural species indicate that the conditions for such an evolutionary process are often present. LITERATURE CITED

CRAMPTON, H. E., 1925 Contemporaneous organic differentiation in the species of Partula living in Moorea, Society Islands. Amer. Nat. 59:s-35. EAST,E. M., 1918 The role of reproduction in evolution. Amer. Nat. 52:273-289. GULICK,J. T., 1905 Evolution, racial and habitudinal. Pub. Carnegie Instn. 25:l-269. J A ~A., P., 1932 The status of the species and the genus. Amer. Nat. 66:346-364. JORDAN, D. S., 1908 The law of geminate species. Amer. Nat. 42:73-80. KELMGC,V. L., 1908 Darwinism, today. 403 pp. New York: Henry Holt and Co. KINSEY,A. C., 1930 The gall wasp genus Cynips. Indiana Univ. Studies. 84-86:l-577. MATHEW,W . D., 1926 The evolution of the horse. A record and its interpretation. Quart. Rev. Biol. 1:139-185. OSBORN, H. F., 1927 The origin of species. V. Speciation and mutation. Amer. Nat. 49:193239. OSGOOD, W. H., 1909 Revision of the mice of the genus Peromyscus. North American Fauna 28:l-285. PHILIPTSCHENKO, J., 1927 Variabilitat and Variation. 101 pp. Berlin. RENSCH,B., 1929 Das Prinzip geographischer Rassenkreise und das Problem der Artbildung. 206 pp. Berlin : Gebriider Borntraeger. ROBSON, G. C., 1928 T h e s,pecies problem. 283 pp. Edinburgh and London: Oliver and Boyd. RUTHVEN,A. G., 1908 Variation and genetic relationships of the garter snakes. U. S. Nat. Mus. Bull. 61:l-301. SCHMIDT, J., 1917 Statistical investigations with Zoarces viviparus L. J. Genet. 7:105-118. SEREBROVSKY, A. S., 1929 Beitrag zur geographischen Genetic des Haushahns in SowjetRussland. Arch. f. Gefliigelkunde, Jahrgang 3:161-169. SUMNER, F. B., 1932 Genetic, distributional, and evolutionary studies of the subspecies o f deer mice (Peromyscus). Bibl. genet. 9:l-106. THOMPSON, D. H., 1931 Variation in fishes as a function of distance. Trans. Ill. State Acad. of Sci. 23:276-281. WRIGHT,S., 1931 Evolution in Mendelian populations. Genetics 16:97-159.

I N T E R N A T I O N A L CONGRESS O F G E N E T I C S

APPENDIX T A B L E OF C O N T E N T S

PAGE ARTOM,CESARE,Autopolyploidism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 M. L., Hybrids of Aegilops and Triticum . . . . . . . . . . . . . . 370 BLARINGHEM, BONNEVIE, K. E . H., Hereditary anomalies in mice descending from stock raised (1921) by LITTLEgnd BAGG. . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 DORSEY,M. J., The morphological expression of dioeciousness in the grape 372 GHIGI,A., Heredity in guinea fowls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 JVCCI,CARLO,Inheritance of cocoon color and other characters in silkworms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 LELIVELD, J. ADOLPH,Cytological studies in the diploid offspring of a haploid Oenothera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380 NACHTSHEIM, HANS,Die genetischen Beziehungen zwischen Korperfarbe und Augenfarbe beim Kaninchen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 ROSI~SKI, BOLESLAW, Does the environment cause genetical change in man? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 SANDERS, J., Epilepsy, twins and heredity . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 SNYDER, LAURENCE H., The inheritance of two types of taste deficiency in man . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 VANDENDRIES, R., Essai d'analyse photographique d'une sporke tetrapolaire de Pleurotus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 WLISSIDIS,THR.,Sur un cas d'albinisme gCnPral . . . . . . . . . . . . . . . . . . . . . . 389

.

EXHIBITS TABI,E OF C O N T E N T S

Coleoptera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Lcpidoptera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .390 Livestock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Mice and rats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 I'ouItry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

. . .

' T h e following section contains the condensed articles and descriptions of exhibits received too late to include with volume 2.

368

PROCEEDINGS O F THE SIXTH

PLANTS LOWERPLANTS PAGE Ferns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Fungi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 HIGHERPLANTS Gossypium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oryza .......d................................................

393 394

BY T H E SCIENTIFIC INSTITUTES OF THE ADDITIONAL EXHIBITSPRESENTED UNIONOF SOCIALISTIC SOVIETREPUBLICS

Quantitative variation in different plants . . . . . . . . . . . . . . . . . . . . . . . . . . Barley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Potatoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wheat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

395 395 395 3% 396 396

INTERNATIONAL CONGRESS O F GENETICS

369

AUTOPOLY PLOIDISM Cesare Artom, R. Universitir di Pavia, Italy

Allopolyploidism occurs very commonly in plants. I n the botanical species, in fact, the process of fertilization allows the meeting of extremely heterogeneous genoms. Nevertheless, during the maturation of the sexual cells of the hybrid the perfect stabilization of the gametes may take place. Thus are realized entirely new genetic constitutions that must be considered a s veritable syntheses of heterogeneous genoms. Polyploidism among plants can thus very frequently be traced back to phenomena of allopolyploidism, a consequence of hybridization. Among animals, on the contrary, the few polyploid species owe their origin to mutation phenomena. Whenever the affinity between the homologous chromosomes suddenly ceases, the reductional phenomena do not take place. Thus diploid gametes are produced instead of haploid gametes, the former determining with their union a new tetraploid constitution. Or polyploidism may be brought about by means of nuclear fusions between the ovarian nucleus and nuclei derived from the polocytes, or, finally, fusions between nuclei derived from the first segmentations of the egg may occur. A confirmation of such statements is afforded first by the Solenobia studied by SEILER,in which tetraploidism is brought about either by means of nuclear fusions between egg nucleus and second polocyte or by means of nuclear fusions between nuclei derived from the second egg division. In the Artemia of S ~ T EI, have been able to point out some facts which confirm my views. I t is a parthenogenetic diploid biotype genetically linkable with the amphigonic diploid of CAGLIARI.In the egg of the Artemia of S&TE the number of chromosomes appears in a haploid condition (21 tetrads). The diploid condition is reached in three different ways, which are analytically examined in a previous paper. Sometimes it is possible to observe in the biotype attempts at attaining tetraploidism. These attempts are either represented by nuclear fusion between the first polocyte and two haploid nuclei, or by a fusion of the first polocyte with the egg nucleus, both of them being diploid, or else by fusion of four haploid nuclei. A student of mine, E. STELLA,has recently brought to this question a contribution of some importance. She established, first of all, that the amphigonic diploid biotype in localities very far from one another and geographically unlinkable is uniform and therefore altogether stable. Furthermore, STELLAhas been able to demonstrate that the parthenogenetic diploid biotype (morphologically different from the amphigonic diploid) is

370

PROCEEDINGS O F T H E SIXTH

identical with the parthenogenetic tetraploid, the latter being but an enlarged image of the first, with larger somatic nuclei and cells. My previsions based on cytological facts have been confirmed: In the tetraploid biotype (resulting from the parthenogenetic diploid biotype by means of automictic phenomena) the factorial complex can only be found duplicated. The tetraploid biotype must then, as far as its origin is concerned, be considered as an autopolyploid biotype. HYBRIDS O F AEGILOPS AND TRITICUM M. L. Blaringhem, Paris, France

Aegilops ventricosa Tausch, castrated in the moment of opening of the spikelets, without injury, is pollinated by Triticum turgidurn L. ; the product, a wheat with sterile stamens, is pollinated by turgidum and gives wheats partially self fertile in the F, generation. The author obtained in the descendants ( a ) some lines nearly sterile and with slight indices of the ascendant Aegilops ventricosa but with the general aspect of a wheat, (b) numerous lines of spelt (Triticurtz Spelta L.), tall and productive, classified among the ancient varieties, and ( c ) some lines with the structure of spikelets and grains of the spelt but with non-imbricated spikes, with 3 to 5 grains in each spikelet, and free in the glumes (or palea). The new wheat is a new section of spelt, Triticunz Spelta polycoccum Blaringhem. HEREDITARY ANOMALIES I N MICE DESCENDING FROM STOCK RAISED (1921) BY LITTLE AND BAGG Kristine E. H . Bonnevie, University of Oslo, N o w a y

An investigation of about 700 embryos of all stages of this strain of mice has first of all confirmed the results already reached by LITTLEand BAGGwith regard to the recessiveness of the eye- and foot-anomalies in question, and also those of BAGGwith regard to the manifestation of the anomaly in late embryonic stages as blood clots, and in somewhat earlier stages as clear blebs, on the dorsal side of the feet. My investigation has further shown that the localization of the blebs at this place is a secondary one and that in young embryos 7 to 8 mm long clear blebs first appear in the neck region of the embryos and later extend under or perhaps within the very thin epidermis toward the head and back. A formation of large blebs is thus very often taking pIace on one or both sides of the head, especially round the eyes and above the nose, and in many cases also across the shoulder region and on the hind part of the back. From the shoulder region the fluid will, at further stages, nearly

INTERNATIONAL CONGRESS O F GENETICS

371

always be removed along the dorsal side of one or both forefeet, at the distal end of which the existence of blebs has already been stated by BAGG. From the hind part of the back the bleb fluid is generally removed, not toward the hind limbs but along the dorsal side of the tail. In many cases the blebs may also persist at this place, their contents being gradually resorbed. Of special interest also are the minute "border-blebs" which are seen to occur in large numbers in embryos of 8 to 10 mm in length, especially on the tibia1 border of their hind legs. The fluid is here generally being resorbed a t a relatively early stage, but before that time the blebs may already have caused characteristic abnormalities. The bleb fluid seems as such to be of no harm to the embryo even if through its pressure some blood capillaries may be disturbed. Only when coming into conflict with developmental processes, such as, especially, those of eyes and extremities, may the blebs cause characteristic types of persisting abnormalities. Convincing facts seem to prove that the bleb fluid has its origin within the young medullar tube and is extruded, in embryos 7 mm long, through an open area (foramen anterius) in the anterior part of the roof of the myelencephalon. The existence of such an area as well as the extrusion through it of some cerebrospinal fluid seems, as first shown by WEED (1917), to be of normal occurrence, the abnormality in our strain of mice consisting in an augmentation of the fluid extruded. The further distribution of the bleb fluid, after its exit from the medullar tube, seems to be governed by merely mechanical forces, the elasticity of the bleb-covering epidermis trying to remove the fluid, especially along the concavities of the embryonic surface relief. I n full accordance with this course of development the recessive abnormality of our mouse tribe is highly varying with regard to its expression ( T I M O F ~ E F Fand ) , even more so in born individuals than in embryos, even if its penetrance is upon the whole very high. After they have been outcrossed with other mouse races, norr:lal with regard to the factors in question, we find however a remarkable change of specificity causing changes in the occurrence of abnormalities on forefeet and hind feet as well as changes with regard to the symmetry relations, especially of forefoot abnormalities. Modifying factors seem, therefore, to have been introduced by the various out-cross races. A thorough analysis of the abnormalities in embryonic as well as postembryonic stages has shown that the final types of anomalies are, each of them, due to specific localizations of the embryonic blebs. A change of

372

PROCEEDINGS OF THE SIXTH

specificity of the abnormalities should therefore be supposed to be due to a corresponding change of localization of the foregoing embryonic blebs, or, in other words, if our supposition of a mechanical distribution of the bleb fluid on the embryonic surface is correct, we should expect to find as one of the first manifestations of the modifying factors a series of changes in the embryonic surface-relief. An investigation on enlarged photographs of a large number of embryos from the out-cross races as well as from groups of extracted abnormals has in fact also proved that different mouse races are characterized by specific variations of their surface-relief in young embryonic stages. A direct correlation has, further, also been found to exist between the statistical occurrence of an embryonic concavity across the shoulder region and the number of forefoot abnormalities within the various groups of extracted abnormals. Summing up, we have before us a monohybrid recessive gene causing, in the first instance, an abnormal augmentation of cerebrospinal fluid in embryos less than 7 mm in length. This surplus of fluid, after being extruded through the foramen anterius of the medulla oblongata, causes a bleb formation under the embryonic epidermis, which may further cause persistent abnormalities of eyes and feet. Modifying genes, introduced by outcross races and influencing the surface-relief of young embryos, are causing a change of localization of the embryonic blebs and through this a change of specificity of the final abnormality. T H E MORPHOLOGICAL EXPRESSION OF DIOECIOUSNESS IN T H E GRAPE M . J. Dorsey, University of Illinois, Urbanu, Illinois

The appearance of the perfect flower i n the grape bearing both functional pollen and pistil is an event of considerable genetic and horticultural importance. The different species have usually been classified as dioecious by botanists, but in the horticultural varieties the hermaphroditic type of flowers is quite common. Since there seems to be some contradiction as to how far development proceeds in the pollen and pistil of the dioecious flowers, this study had as its objective the clearing up of that point. Extensive germination tests were made of a large number of pollen grains taken from the reflexed type of anther, and over a three-year period none were found to germinate. In the pollen of the reflexed stamens studied germ pores were not present. These results would seem to throw open to question the validity of self-pollinations or crosses between varieties bearing the reflexed type of stamens and seem to run counter t o the idea of the "mixed" pollen of some of the early investigators.

INTERNATIONAL CONGRESS O F GENETICS

373

A n examination was made of the extent of pistil development in a large number of staminate and intermediate flowers. In the early stages no distinction could be made between the strictly pistillate flower and the staminate or intermediate types. A t bloom, however, great variation was found in the development of the embryo sac and attendant structures in the pistillate and intermediate flowers. In some instances there was abundant stigmatic tissue, but a n aborted ovule and embryo sac, while in others there was a n apparently normal embryo sac but no stigma. Fruit-setting in these types may therefore be prevented by two quite distinct causes. This condition should be taken into consideration in classifying seedlings as to flower type. It also shows how closely pistil development approximates the hermaphroditic condition in the intermediates and the occasional "wild vine" which sometimes sets fruit. HEREDITY IN GUINEA FOWLS1 A. Ghigi, University, Bologna, Italy

I n this paper I deal first with the factors which determine the color differences in each of the domestic breeds of guinea fowls; I then study the results of crossings between several wild varieties and domestic breeds of the genus Numida and state the first results I have obtained by crossing several wild varieties of Guttera. Guinea fowls have been widely raised in Italy for about a century, several breeds being known which differed only in color. They are the following: (1) Common Grey, corresponding to the wild Numida meleagris galeata of Guinea, a variety which differs from all others in that the feathers at the base of the neck are uniformly purplish violet ; (2) Lilac, very much like the preceding save that the fundamental hue of its body is a very pale lilac; ( 3 ) Pavonated or Violet, a variety free from pearly spots except on the long feathers at the sides and along the rachis of the feathers of some other parts of the body. Its fundamental hue is very dark, almost black, abundantly tinged with violet, which, under the action of the sunlight, befrom comes brown. A specimen of this breed obtained by REICHENOW I Africa is the Numida zechi type. Through the kindness of STRESEMANN have examined this specimen and found that it is exactly identical with the domestic Pavonated breed. (4) White, a thorough albino, both on account of the depigmentation of its skin and of the whiteness of its feathers on which no trace of pearly spots is to be found. In France instead of the White breed they raise a variety which I have called Spotted Light Buff. The skin of this variety is pigmented, having a

' A summary of

this article is given in Italian in volume 2, page 60.

374

PROCEEDINGS O F T H E S I X T H

furldamental yelIowish-white color, extremely light, on which one readily notices conspicuous pearly spots. My first experiments were concerned with the behavior of the breeds 1, 2, and 3 in comparison with the White (albino). The albino is completely recessive with respect to any color whatsoever, but the behavior of this factor is curious. The entire F1 generation is uniform and grey with a white spot in the center of the breast. In the F, generation a systematic segregation of the pure factors of the ancestors takes place; in the heterozygotic individuals, however, the white behaves so as to seem due to multiple factors, for it extends to all the lower parts, including the primaries and the adjoining secondaries, which, when not in motion, remain in close contact with the white abdomen. It is an interesting fact that not a white spot is ever to be found on the back; one might say that color predominates in the upper parts, while white is more or less preponderant in the lower parts, having an area of diffusion which radiates from a central point on the breast. k establish the genetic differential value of the Later on I ~ ~ n d e r t o oto colors which characterize the breeds 1, 2, and 3. I found that the Common Grey variety completely dominates the Lilac ( 2 ) , and the Violet ( 3 ) varieties. The hybrid F,of Grey and Lilac produces in F, three-fourths of Grey and one-fourth of Lilac, and the hybrid F, of Grey and Violet produces in F, three-fourths of Grey and one-fourth of Violet. I have concluded, therefore, that in these three breeds two pairs of antagonistic factors exist, which I have set forth as follows: M , margarogen or pearl-producing factor, is present in the Con~monGrey and Lilac breeds. In the Pavonated it is replaced by nz, which inhibits the pearl formation on the greater part of the body. I determines the fundamental dark color in the Conlmon Grey and in the Violet (zechi). It is replaced by i, in the Lilac, the fundamental color of which is pale. These four factors are cotnbined as follows: MI in the Common Grey, ?nZ in the Pavonated (zechi) and Mi in the Lilac. The combination mi could be foreseen ; as a tnatter of fact, I had obtained it in 1924 by crossing the Pavonated breed ( n z l ) with the Lilac ( M i ) , getting the following results:

P mw1ZIX M M i i F, MmIiX MmZi phenotypically common grey F, 9 M1+3 m1+3 Mi+l~qti The last of these varieties, doubly recessive, is the Ghigi Blue guinea fowl, in which the pearly spots are confined to the sides as in the zecki. Its funda-

INTERNATIONAL CONGRESS O F GENETICS

375

mental color, however, is not pale, as in the Lilac, but decidedly tinged with blue. So long as the crossings take place within these four breeds, the behavior of the factors is regular, and the results are those usually obtained with dihybrids. During the two-year period, 1928-29, I sought to find the value of the Spotted Light Buff breed. This breed possesses the margarogen factor (M). As for its fundamental pale brownish-yellow color, I had observed that it varied in intensity, some specimens being somewhat darker and more differentiated and others so light as to seem white. This difference is noticeable even at birth, for some chicks show very conspicuous longitudinal stripes of a tawny hue, while others are lightly colored with a dirty white. Furthermore, the greater intensity of this characteristic seems to be !inked to the female sex. In fact, while I have obtained females of various shades of color, the males have always been very light, almost white. Therefore, by crossing the Spotted Light Buff breed with the Violet of formula MII have obtained in the F, an atavistic result: all the chicks came grey like the wild guinea fowl and like those obtained by crossing the Lilac and Violet breeds. The result in the F, was the following: Pearly Grey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 7 7 Lilac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Violet (zechi) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Blue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Spotted Light Buff with pearls . . . . . . . . . . . . . . . . . 49 Spotted Light Buff without pearls . . . . . . . . . . . . . . 17 The last group represents a new homozygotic combination, in which the yellowish (light-buff) color is associated with the factor nz, which inhibits the extension of the margarogen factor. There seems to be no doubt that the factor for the pale brown (which for the time being I shall designate by the symbol X, meaning thereby that it is an undetermined factor) is recessive with respect to (I), since it (that is, factor X ) in the table given above does not exceed by much one quarter of the specimens possessing the factor of intensity ( I ) . However, having crossed this year the two doubly recessive varieties, the Blue of formula mi with the new form Light Buff without pearls, which should have the same factors plus X, we find in the progeny 15 Pavonated, in which the factor (I) is undoubtedly present, and four Blue chicks. The fundamental color X is,

376

PKOCEEIIINGS O F THE SIXTH

therefore, recessive with respect to the blue color and might be a mutant, but I can not account for the rise of the combination for the Pavonated which has the factor (I).On the other hand, in 1931 a certain number of specimens from the Light Buff group were obtained that were almost white, with a greyish tinge, and that bore a resemblance to their parents. These specimens are now being studied. It still remains to be seen how the Pavonated, produced by crossing the Blue with the Light Buff, will behave in the F2.

This second new breed of a pale yellowish-brown color without spots, as well as the grey-tinged 193 1 specimens, exhibits the same characteristics as the Spotted Light Buff: ( 1 ) variability of the intensity of the fundamental color and (2) linkage of a greater intensity with the female sex. During the last few years I have repeated the crossing of the Numida rneleagris donzestica with the Nztmida ptilorhynca, which I had done as early as 1912. This time, however, I have been substituting the Lilac breed for the Common Grey, in the domestic variety, and the subspecies Sorndiensis for the typical ptilorhynca, in the wild variety. The findings concerning the behavior of the bristle-like nasal papillae of the latter have been confirmed. This behavior, considering the absence of the papillae in the nzeleagris, is exactly that of a multiple character. In the segregation in F, many specimens are produced which show nasal excrescences similar to those formerly designated as papillae. I have also produced a homozygotic strain for the lilac color in which the facial characteristics, although still oscillating, are strongly accentuated. W e have in this case a correlation between the specific characters and those of a recent mutant occurring in the domestic variety. During the last three years I have also crossed Nzi?~tidanzeleagris with N . nzitrata, obtaining in the F, intermediate and uniform hybrids which have caused in the I;, the greatest segregation concerning the few differential characteristics. The wine-colored ring at the base of the neck of the llzeleagris reappears in many specimens as a fundamental color separating the white transversal stripes. These stripes in the mitrata have a greyishblack background. The shape of the wattles causes many and varied combinations. However, while in some specimens we find a reversion to the form and size of the nteleagris, none shows the characteristic thinness of the mitrata. An analytical study of these characteristics can be made only late in the fall after growth and molting have stopped. The most important fact is the complete fecundity of the hybrids of the three wild breeds, nzeleagris, nzitrata, and ptilorhynca, which segregate, variously correlated, the characteristics of the parental species.

INTERNATIONAL CONGRESS O F GENETICS

377

Therefore, one feels authorized to modify the systematics of the genus Numida, which experiments have shown to be constituted of one single species separable into many geographic breeds to which it is wrong to attribute any specific value. I have obtained the same result as concerns the oriental varieties of the genus Guttera (crested guinea fowl). A female hybrid, G. liridicollis X G. barbata, laid this year, for the first time, eleven eggs which were proved to have been fecundated by a G. pucherani cock. My previsions of some time ago c~ncerningthe genesis of these varieties localized in the damp forests of equatorial Africa have, therefore, been confirmed by experiment; this experiment also proves that in this case hybridization is one of the greatest factors in the production of varieties which the systematists are accustomed to consider as distinct species. INHERITANCE O F COCOON COLOR AND OTHER CHARACTERS IN SILKWORMS Carlo Jucci, R. Universitci di Sassari, Sasswi, Italy

Silkworms (Bombyx rnori) constitute very fine material for the dynamic study of inheritance since it is possible t o follow the characters in all the stages of their organic development in order to find out how the characteristic genetic potentiality transmitted by the parents develops when once a t work in the new organism. In respect t o the growth curve the three-molting breeds differ sharply from the four-molting ones, and the bivoltine breeds from the univoltine. In comparison with the four-molting ones the three-molting breeds show less growth because they economize one of the molts by deferring as long as possible the rejuvenating process of the organism, which finds itself compelled to advance a t a slower rate. The bivoltine breeds also show less growth than the univoltine ones, a growth less not in its relative value, since the growth capacity is the same or almost so in both, but in its absolute value, because the hatching weight is lower, in view of the fact that the eggs of the bivoltine breeds (which give two generations a year) are smaller. The number of the molts behaves as a Mendelian factor; the voltinism certainly depends upon many factors, and the problem of its genetic behavior merges in the general problem of inheritance of metabolic type. The developmental capacity of the egg depends upon the constitution of the egg itself, determined in turn by the physiological constitution of the maternal organism. So the developmental capacities of the fertilized egg and of the virgin egg (capacity for sinechepidosis and capacity for partheno-

378

PROCEEDINGS OF THE SIXTH

genesis) are an excellent index of the physiological capacity of the individual and of the breed. I t is possible to follow growth capacity in the reciprocal crosses in the successive stages of larval development. As regards the weight at the end of the larval period, there appears to be a decided predominance of the maternal character. The initial advantage given by the cytoplasmic supply of the egg is so great that it is not lost in the course of the postembryonic life. Opaque skin is a Mendelian factor that is dominant over transparent skin, and in F, segregation occurs in the classic ratio 3 :1. The transparent-skin character seems to be caused by the deficiency in the hypodermic cells of the urate crystalline granules t o which the skin of the common breeds owes its characteristic opacity. When the opaque-skinned worms get ripe, the urates contained in the hypodermic cells redissolve, so that the uric content of the blood rises to a level about three times higher. Nothing like it takes place in the transparent-skinned silkworms. Therefore, in physiological terms, the opaque-skin factor consists in the capacity of the hypodermic cells to extract urates from the blood, accumulating them in the cytoplasm in the form of crystalline concretions. In Bornbyx nzori there occurs a migration of the pigments from the blood of the silkworm to the secretion of silk glands; this is the origin of the yellow color of the cocoon (carotinoids, carotin and xanthophyll, derived from the mulberry leaves). The "capacity of migration" behaves as a Mendelian unit character dominant over the "incapacity of migration" (silkworms with yellow blood and white cocoon). The two races Chinese Golden (Oro Chinese) and European Yellow (Giallo I n d i g h e ) both possess capacity of migration, but the migrations occur at a different time, that is, before maturation in the Golden and after in the Yellow race. So, notwithstanding the fact that the total quantity of pigment is the same or about the same for the same quantity of silk, the cocoons of Golden are a deeply golden yellow while the cocoons of Yellow are pale yellow; but inside the reverse is true and the pigment distribution is just inverted. In the F1 of the hybrid ? GoldenX 8 Yellow (and reciprocal) the pigment distribution in the cocoon's layers and the "degree of precocity of pigment migration" from blood to silk are intermediate between those of the parental races. If the two races Golden and Yellow are crossed with the same White Japanese race, the F1 hybrids get the "migration time" characteristic of the parental blood colored race; so the two crosses ? GoldenX $ White and ? WhiteX 8 Golden produce a Golden cocoon, and the two crosses 9 YellowX 8 White and P WhiteX 8 Yellow produce a Yellow one. I n hybrids having a White mother the

INTERNATIONAL CONGRESS O F GENETICS

379

blood is colored no less than in the reciprocal crosses with Golden or Yellow female or in the pure blood-colored races (very strong and most precocious dominance). In addition to carotinoids the silkworms also absorb from the mulberry Ieaves pigments of the flavone group. Generally these Bavones pass through the chrysalis blood to the eggs, contributing t o the pigmentation of the yolk (while the cocoon is colored by carotins and xanthophylls migrating from the blood of the ripened silkworm to the silk glands). In some races, however, the flavones pass through the blood to the silk; so in the "Japanese Green" the cocoon pigment "bombiclorina" is a flavone. The "green" character is therefore, in physiological terms, "the capacity of the silk glands for absorbing the flavones from the blood." Whether or not this capacity is a Mendelian character as is the "capacity of the silk glands for absorbing carotinoids from the bloocl" only the F, of the hybrids can prove. In the F,the "green" is dominant over "white" (that is, permeability for "flavones" dominates over "impermeability") and is recessive to lemon, orange, rose, European yellow (flesh-colored) and golden. The dominance, however, is far from perfect; often the F,cocoons demonstrate little uniformity, varying from the type of the yellow race to the type intermediate between the parentals. AIso, in the cases in which they are of the yellow type (golden, lemon, orange, flesh-colored, rose) the cocoons contain flavones. The cocoons exposed to KH, vapors became characteristically 66 rusty" owing to the formation of an ammoniacal salt of the flavones. Therefore in the hybrid the two parental characters "capacity of the silk glands for absorbing carotinoids" and "capacity for absorbing flavones" are both developing, if the "green" keeps chron~aticallyconcealed. The reaction in the F,o i the "rusting" by NH, is expected to be valuable as a means of establishing the ratios of dissociation between the parental characters, ratios which should not be easy to establish by any other means because the "permeability for flavones" is evident only in absence of "permeability for carotinoids." The lack of uniformity of the F1 of crosses between green and yellow races is probably due to the presence in some individuals of the green race of some factor concerned in the "carotinoids permeability," or to the presence in some individual of the yellow races of some factor concerned in the 66 flavones permeability" (being not sufficient, when alone, to produce its effect, the gene would keep potential so the ordinary selection could not remove it).

380

PROCEEDINGS O F THE SIXTH

CYTOLOGICAL STUDIES I N T H E DIPLOID OFFSPRING O F A HAPLOID OENOTHERA 3. Adolph Leliveld, Botanical Imtitzlte, Amsterdam, Netherlands

Of late years crosses made between Oenothera species have sometimes resulted in a yield of haploids. STOMPS(1930) reports that the reduction division in these haploids may give in a single case a normal pollen grain or embryo sac with seven chromosomes. I t may be understood that in this way a self-pollination yields a new diploid plant. As a matter of fact, such a diploid plant may be presumed to possess two quite homologous groups of seven chromosomes and, correspondingly, the cytological behavior should be that of an entirely homologous type. I n the year 1931, some fixations were made of the Fz of such a new diploid Oe. franciscana. The haploid from which it originated was the result of a pseudo-cross between Oe. franciscana, the cytology of which had been studied earlier, and Oe. longiflora. The original Oe. franciscana had seven loose pairs of chromosomes. The cytological results, however, did not correspond to the expectations. Diakinesis was the first stage remarkable by its deviations: chromosome pairing by two's occurred, but this was seldom the case. By far the greater part of the riuclei studied showed the pairs locked up by two's, or even an interlocking of three or more pairs was seen. Besides these constellations chains of four and even once a chain of six chromosomes were noticed. Prometaphase proved to be in the possession of a still considerable number of linkages: the interlockings became more rare; there seemed to be a tendency to loosen the connections and to come to an entire pairing-up of the chromosomes. Thus, a large part of the nuclei showed one to three pairs lying apart from the others: they possessed only one terminal connection between the two chromosomes. In almost all nuclei the last-mentioned type of pairs was present, but they were not always apart from the other pairs. Early metaphase was interesting on account of the progressive diminution of the number of linkages or interlockings. Part of the pairs were ringshaped, part of them possessed one terminal connection, and these last ones in a considerable number of cases lay apart. According to a number of investigators of the present time, ring and chain formations and interlockings would be the result of structural heterozygosity. Here, however, any heterozygosity whatever of the homologous chromosomes is excluded. The question arises whether the pairs with the single terminal connection do not arise from the loosening of the interlockings.

I N T E R N A T I O S A L COSGRESS O F GENETICS

38 1

When one keeps it in view that the behavior of the chromosomes during diakinesis and prometaphase equalizes to a normal tnetaphase, the question arises anew whether the said behavior in diakinesis and the following stages, that is, the mean average of the configurations with its corollary deviations, may not be taken as a fact inherent to the species instead of being the direct result of the genetical tnake-up of the type. This idea has already been a d vocated by some of the Oenothera investigators. D I E GENETISCIIEN BEZTEHU-NGES Z W I S C I I E S K0RPERI;AKRE U S D AUGENFARRE REIM KANINCHEN Hans Nachtsheim, Institut fiir Vererbungsforschung der Landu~irtschaftlichenHochschule. Berlin-Daltlem, Germany

Der Vortragende berichtet iiber von ihm in den letzten Jahren ausgefiihrte Versuche mit Kaninchen, deren Ziel die Klarlegung der zum Teil sehr innigen genetischen Reziehungen zwischen Korperfarbe und Augenfarbe ist. Zunachst wird die CVirkung der verschiedenen Farbfaktoren auf Haarund Irisfarbe betrachtet. Die Kombination der dominanten Faktoren liefert die fiir das Wildkaninchen charnkteristische sogenannte Wildfarbung und dunkelbraune Irisfarbe. Je mehr rezessive Farbfaktoren an die Stelle der donlinanten treten, um so starker wird das Haarkleid auigehellt. Mit der Aufhellung des Haarkleides geht im allgerneinen einc Reduktion der Irispigmentierung Hand in Hand, doch zeigt es sich, dass nicht alle Farbfaktoren H a a r und Iris in gleichem Masse beeinflussen; eiriige Faktoren, die das Haarkleid stark beeinflussen, sind auf die Irispigmentierung nur schwach oder iiberhaupt nicht wirksam. S o unterscheidet sich ein schwarzes Kaninchen von einenl wildfarbigen durch das Fehlen des Wildfarbigkeitsfaktors. Zwischen dcr Augenfarbe eines schwarzen und der eines wildfarbigen Kaninchens ist jedoch kein Unterschied. Eine Besonderheit bietet das sogenannte marmorierte Auge mancher Chinchillakaninchen dar. Hier scheint ein Faktor im Spiele zu sein, der die Irispigmentierung reduziert, ohne auf das Haarkleid einen entsprechenden Einfluss zu haben. Bisher wurde das marmorierte Auge nur bei Chinchillakaninchen beobachtet. Ob dies in einer engen Koppelung zwischen dem Faktor fur marmorierte Augen und dem Chinchillafaktor seine Erklarung findet, oder ob noch innigere genetische Beziehungen zwischen den beiden Merkmalen bestehen, bedarf noch der weiteren Priifung. Neben den eigentlichen Farbfaktoren zeigen die Faktoren fiir Leuzismus und Scheckung besonders interessante Beziehungen zur Augenfarbe. Englische und hollandische Scheckung sind zwei genetisch verschiedene und

382

PROCEEDINGS O F T H E SIXTH

auch phanotypisch leicht zu unterscheidende Scheckungstypen. Die Englische Scheckung ist dominant, die Hollanderscheckung ist mehr oder weniger rezessiv gegeniiber Einfarbigkeit. Beiden Typen gemeinsam ist eine starke Variabilitat; das eine Extrem ist ein Tier mit sehr vie1 Pigment und nur geringen weissen Abzeichen, das andere Extrem ist ein fast weisses Tier mit nur ganz kleinen Pigmentzentren. Bei der Englischen Schecke entspricht die Augenfarbe, gleichgiiltig wie der Scheckungsgrad ist, immer der Haarpigmentierung, d. h. eine schwarz-weisse Englische Schecke hat dunkelbraune, eine braun-weisse braune, eine blau-weisse graublaue Augen usw., auch bei extrem-weissen Englischen Schecken, wo sich die Haarpigmentierung auf schmale Augenringe und gefarbte Ohren beschrankt. Bei der Hollanderschecke hingegen entspricht die Augenfarbe der Haarpigmentierung nur bei den mittleren Scheckungsgraden und den extrem-gefarbten Tieren. Bei den extrem-weissen Tieren mit nur kleinen Pigmentzentren um die Augen und am Schwanz sind die Augen blau ohne Riicksicht auf das Haarpigment (sogenannte Glasaugen) oder auch-bei etwas starker pigmentierten Tieren-zweifarbig, d. h. blau und dunkelbraun, braun, graublau usw., entsprechend der Haarpigmentierung (Heterochromie) . Das blaue Auge der extrem-weissen Hollander entspricht morphologisch vollkommen dem blauen Auge des leuzistischen Weissen Wiener-Kaninchens, d. h. im vorderen Teil der Iris fehlt das Pigment, Retina und hintere Irisschicht aber sind pigmentiert, und nach dem Prinzip der halbdurchsichtigen Medien kommt so die blaue Augenfarbe zustande. Die Hollanderfaktoren, die das blaue Auge hervorbringen, sind andere als der Weisse Wiener-Faktor, doch zeigt sich ein Zusammenwirken dieser Faktoren im Kombinationsprodukt. Wird der Weisse Wiener-Faktor in einfacher Dosis in Hollander gebracht, so treten blaue Augen und Heterochromie auch bei Hollandern mittleren Scheckungsgrades und extrem-gefarbten Tieren auf. Der Weisse Wiener-Faktor zeigt diese Wirkung auf die Irisfarbe aber nur im Zusammenwirken mit Hollanderfaktoren, in Englischeri Schecken ist es in einfasher Dosis ebenso unwirksam wie in ganzgefarbten Tieren ohne Scheckungsfaktoren. Bei einer Verbindung der beiden Scheckungstypen ist die Englische Scheckung epistatisch iiber die Hollanderscheckung, soweit die Haarfarbe in Frage kommt. Beziiglich der Augenfarbe ist die Hollanderscheckung epistatisch, d. h. Englische Schecken konnen Glasaugen und Heterochromie aufweisen, wenn sie $eichzeitig extrem-weisse Hollanderschecken sind, und kommt noch der Weisse Wiener-Faktor in einfacher Dosis hinzu, so spielt er die gleiche RoIle wie bei der Kreuzung mit reinen Hollandern. Schliesslich wird noch das Zusammenwirken des Weissen Wiener-Faktors

INTERNATIONAL CONGRESS O F GENETICS

383

rnit den verschiedenen Allelen der Albinoserie besprochen. Bei der Kreuzung von Weissen Wienern mit Albinos erhalt man in F1 gefarbte Tiere und in F2neben den anderen Typen doppelt-rezessive Individuen, "Weisse WienerAlbinos," die im Phanotyp Albinos sind. Der Albinofaktor ist also epistatisch iiber den Weissen Wiener-Faktor. Bei der ICreuzung von Weissen Wienern mit dem nachst-hoheren Typ der Albinoserie, dem Russenkaninchen, erhalt man wiederum in Fl gefarbte Tiere und in F, als doppeltrezessive Individuen "Weisse Wiener-Russen." Diese sind im Phanotyp weder Weisse Wiener noch Russen, sondern vollkommene Albinos rnit roten Augen, es sind synthetische Albinos ohne Albinofaktor. Die Paarung dieser phanotypischen Albinos mit "normalen" genotypischen Albinos liefert nicht wiederum Albinos, wie sonst jede Paarung albinotischer Tiere, sondern Russenkaninchen. Die Versuche mit Weissen Wienern und den hoheren Typen der Albinoserie haben noch nicht zu endgiiltigen Ergebnissen gefiihrt, doch kann als sicher gelten, dass im Zusammenwirken des Weissen Wiener-Faktors mit den hoheren Allelen der Albinoserie ein Wechsel in den Epistase-Verhaltnissen eintritt, der bei den "Weissen Wiener-Russen" bereits eingeleitet ist. DOES T H E ENVIRONMENT CAUSE GENETICAL CHANGE I N M A N ? Boleslaw Rosiriski, L d w , Poland

The paper I am going to present does not deal with the inheritance of any particular character in man in a strictly genetical sense. I t is a statistical work on the cephalic index. I was interested as to whether the cephalic index of man will undergo any changes after he has lived for one or more generations in different conditions of environment. Stated specifically, I studied the cephalic index of the Polish people in their native country and compared it with that of the offspring born in Texas to Polish immigrants. Many research students who have investigated the European people o f the present and past have stated that they are composed of distinct anthropological types. According to the theory of CZEKANOWSKI there exist four fundamental elements in Europe: Nordic, Laponoid, Armenoid and IberoInsular. From these elements there are derived the six other types (secondary types) in this way: Northwestern from Nordic and Ibero-Insular, Subnordic from Nordic and Laponoid, Dinaric from Nordic and Armenoid, Alpine from Laponoid and Armenoid, Preslav from Laponoid and IberoInsular and Litoral from Armenoid and Ibero-Insular. Therefore an intermingled European population can be represented by the following equation:

384

PROCEEDINGS O F T H E S I X T H

In this equation a indicates the Nordic, 1, the Laponoid, h, the Armenoid and e the Ibero-Insular element, whereas 2al, 2ah, 2ae, 21h, 21e and 2he indicate the secondary types. Having this equation, we can define the percentage of the fundamental elements of which the different populations are composed. Among the American people of Polish origin born in Texas we have established the following anthropological types : Men Anthropological tyge

Nordic .............................. a2 Northwestern . . . . . . . . . . . . . . . . . . . . . . .2ae Subnordic . . . . . . . . . . . . . . . . . . . . . . . . .2al Laponoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Preslav . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21e Litoral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2he Dinaric . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..2ah Alpine ............................. .21h Amount

Women

Number

Percentage

Number Percentage

63 53 124 38 31 8 37 13

17.2 14.4 33.8 10.4 8.5 2.2 10.1 3.5

134 36 37 3 31 38

12.8 10.9 36.6 9.8 10.1 0.8 8.5 10.4

367

100.1

366

99.9

47

40

From these percentages of the anthropological types we can derive the percentage of fundamental elements of which our population is composed. These are a+ follows: Men Elements

Nordic . . . . . . . . . . . . . . . . . . . ..- . . . . . . . . . . . . . . . . . . . . .a Laponoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I Armenoid ........................................h Ibero-Insular .................................... .e

Percentage

41.4 37.0 8.3 13.4

Womeri Percentage

35.8 40.9 10.3

12.9

According to the anthropologica1 characteristic of fundamental elements the theoretical cephalic index of the skull of the Nordic element can be regarded as equal to 76.0, that of the Laponoid and the Armenoid to 88.0 and that of the Ibero-Insular to 68.5. For the European people of the present time the average of the theoretical calculated cephalic index will be in agreement with the average of actual cephalic index obtained by the measurements if we accept the dominance of the cephalic index of Armenoid and Laponoid over that of Nordic and of Ibero-Insular. Similarly the cephalic index of Nordic dominates over Ibero-Insular. In order to answer the question of whether the environment changes the dominance of cephalic index in man I have calculated the theoretical cephalic index of Polish offspring born in Texas and compared it with the actual

INTERNATIONAL CONGRESS O F GENETICS

385

index obtained by the measurements. The average of the actual cephalic index of American people of Polish origin born in Texas equals: Men Women

A 83.66k0.13. v=4.23k0.10 A 84.42k0.13. v=4.23k0.11

There is a perfect agreement between the theoretical and actual average of cephalic index if we accept for the Texas born offspring the dominance of the same anthropological types over the others as we have in Poland. The theoretical calculated index under these conditions equals: Men Women

M-83.08 M-83.89

The small difference between the actual and the theoretical indexes can be explained by the fact that the theoretical index refers to the skulls and the actual index is based on measurements of living people. It is known that the skull index is lower than the head index. This result seems to justify the conclusion that the environment does not change the dominance of cephalic index in man. EPILEPSY, TWINS AND HEREDITY J. Sanders, Rotterdam, Netherlands

I had the opportunity to investigate 3 pairs of identical and 1 pair of nonidentical twins who had epilepsy. I will begin by telling you something about these twins. WILHELMINAB. and NEELTJEB. were born July 24, 1915. At the age of 4 years MIEN had a convulsion. Some days later NEL had one. Sudden unconsciousness was followed by shock of the extremities and the head. MIEN had a convulsion of one-half hour's duration, NEL one of one hour. After these convulsions the children had more small ones. A t the age of 14 years NEL had a bad attack. MIEN had a bad attack at the age of 15% years after a fright caused by a fire which also caused NEL to have an attack. The children are both right-handed. They were the same a t school. NEL helps the mother at home because she cannot go into a business in consequence of the severe attacks; MIEN sews at a shop. The father as a child had some convulsions as did also a sister of the mother. This sister could not learn and has now sometimes a strange feeling of changed consciousness. The physician considers these attacks of this mentally deficient woman to be epileptic.

386

PROCEEDINGS OF THE SIXTH

The skin of MIEN and NEL is nearly the same. The teeth and the jaws differ, due in great part to the bad condition of the teeth. Investigation shows that psychologically the twins are not nearly the same. The I.Q. of MIEN amounts to 80 and that of NEL to 90. They belong to the group of the backward-normal persons. Anthropometrically there is little difference. MIEN is a little taller and heavier than NEL, but the indices are nearly the same. They both have strabismus convergens manifestus, but it is worse in MIEN. Objectively there is a great conformity between the eyes of MIEN and the eyes of NEL. Dactyloscopically they differ. The formula, in accordance with HENRY,is 25 o o

17 m o for MIENand --- for NEL. 3 0 0 27 m o

The second pair of twins were born August 15, 1919. At the age of 3 years J A N had the first epileptic attack; COR'Scame one-half year later. They had one each month until 2 years ago. Since 1930 they have had more severe attacks. The epilepsy of Con is more severe than that of JAN. The mother has had 9 children, all alive. Two children are mentally deficient and one has also had some attacks. The father had epileptic attacks from the time he was twelve until he was thirty years of age. The teeth and jaws of J A N and CORhave much conformity with each other. Psychologically the twins are rather similar. The I.Q. of J A N is 81, of COR84. They belong to the backward-normal children. They also conform anthropometrically. Objectively the eyes are nearly the same. Both are right-handed. There is difference in the fingerprints. The formula according

The third pair of twins, the children G., were born December 3, 1921. NINI G. had many convulsions in the second year of her life. She has had epileptic attacks since her eighth year. JOOPJEhad the first convulsion the same day as NINI but has been free from them since her fourth year. NINI now has many very severe attacks. In the last 3 years she has been unable to go to school. Before that time she was number 4 at school, JOOPJE number 2. NINI could learn very well. JOOPJE is an intelligent child with an I.Q. of 115. NINI is now very slow and has the intelligence of a child of 6 years. She has been growing very fast during the last 3 months, so that her weight is 8 K.G. higher than that of JOOPJE. The epileptic attacks have injured her mind.

INTERNATIONAL CONGRESS O F GENETICS

387

The children conform very much anthropometrically, in the eyes and dactyloscopically. The formula according to HENRYfor the fingerprints 1Ui of both children is -. In the family there is no epilepsy or other psychosis. 1Ai The fourth pair of twins is not identical. They were born November 1, 1928. These children are not of the same sex. JOPIEM. had the first epileptic attack at the age of nine months. Next day MARYhad a similar attack. The number of attacks of MARYis 5 daily, of JOPIE,one. The grandmother on the mother's side was a melancholiac. I have here been describing very briefly 4 pairs of twins, all with epilepsy. In the literature I found published the following cases of epilepsy of twins. Identical tzeins Investigator

Concordant

Non-identical tm'ns

Discordallt

Concordant

Discordant

With my own cases I get the following numbers: Concordant

16

Identical twins Discordant

6

Non-identical twins Concordant Discordant

2

15

S o the hereditary index of epilepsy amounts to

This index points to a hereditary factor. But the numbers are small, so the result is not very reliable. We must wait for our final conclusion until many more cases have been published. These must not include identical twins only. The non-identical twins are of the same importance for any conclusion. T H E INHERITANCE O F T W O T Y P E S O F T A S T E DEFICIENCY I N MAN Laurence H. Snyder, Ohio State University, Columbus, Ohio

The taste deficiency for phenyl-thio-carbamide is clearly shown to be a simple autosomal recessive character. One thousand families are presented in support of this statement. The data are treated statistically, and all deviations are less than their probable errors. The taste deficiency for a second

388

PROCEEDINGS O F THE SIXTH

compound, di-ortho-tolyl-thio-carbamide, is shown by means of 200 families. This deficiency is not a simple recessive character, as evidenced by the fact that in many families in which neither parent tastes the compound some of the children do. An additional fact noted is that while many individuals taste both compounds, many taste phenyl-thio-carbamide only, and many taste neither compound, no one seems to taste di-ortho-tolyl-thio-carbamjde only. These two facts taken together point to an epistatic relationship of the factors. A hypothesis which satisfies all the facts and meets all the statistical requirements is offered. This involves two pairs of factors with peculiar epistatic relationships. These taste deficiencies, adding as they do to our list of common unit factors in man, provide suitable material for linkage studies in man. ESSAI D'ANALYSE PHOTOGRAPHIQUE D'UNE S P O R P E TPTRAPOLAIRE D E PLEUROTUS R. Vandendries, Rixensart, Brussels, Belgium

Nous avons essay6 de rendre compte, par la photographie, des rCsultats obtenus en confrontant, deux i deux, vingt-six haplontes de Pleurotus columbinus, espece tgtrapolaire. Nous disposions d'un tableau de croisements traduisant par des signes les rPsultats d'une analyse microscopique complete. Quatre groupes sexuels, portant respectivement les facteurs sexuels ab', a'b', ab' et arb, peuvent donner matikre 2 dix pairages diff6rents: 1. Deux de ces combinaisons sont fertiles et produisent des mycCliums diploides, porteurs d'anses d'anastomose. Ce sont les combinaisons abxa'b' et a7bXab', oii les deux individus confrontCs ont leurs deux facteurs sexuels diffCrents. Nos photographies de cultures sur disques d'agar permettent de reconnaitre morphologiquement ces croisements fertiles. 2. Les confrontations d'individus de mOme formule sexuelle, au nombre de quatre, savior: abXab, a'b'xa'b', a'bXa'b et ab'xab' montrent que les deux myckliums confrontCs croissant l'un dans l'autre comme le font deux boutures d'un mtme individu. Nos photographies sont Cminemment suggestions i ce sujet et demontrent cette analogie. 3. Les confrontations d'individus ayant un facteur sexuel commun au nombre de quatre: abxab', a'bxa'b', abXa'b et ab'xa'b'. Ces quatre groupes de combinaisons sont stCriles. Mais ici une discrimination est possible et parfaitement rendue par la photographie. Entre les individus de formule ab et ceux de formule ab' se manifeste une r6pulsion caract6risCe par un barrage net; pared barrage se manifeste entre les individus du groupe arb et ceux du groupe a'b'. I1 resulte de l i que la cause dCterminante du barrage est la presence du mOme facteur soit a, soit a', attach6 aux facteurs

INTERNATIONAL CONGRESS O F GENETICS

389

diffCrents b et b'. Nos photographies donnent une dCmonstration trPs nette du phknomPne et de son caractere ginCral. Ce rCsultat confirme celui de OORT sur Coprinus fimetarius. Au contraire les combinaisons a b X a r b et ab'Xarb' oG le facteur b ou b' est commun et les facteurs a ou a' sont diffCrents ne donnent pas lieu au barrage. Les individus qui y sont confrontks croissent l'un dans l'autre, sans manifestes de rkpulsion, mais ils ne se conjuguent ni ne s'anastomosent, la culture reste stkrile. Dans quelques cas nous avons pu observer, comme le fit OORT,une action d'inhibition de l'un organisme sur l'autre. Cette action serait donc due B la diffkrence entre les deux facteurs a et a' attach& soit au facteur commun b soit au facteur commun b'. Ces cas de st6rilitC apparaissent B toute Cvidence sur nos photographies. SUR U N CAS D'ALBINISME GBNBRAL Thr. Wlissidis, Atltines, Greece

Dans une note prCcCdente prCsentCe au dernier congrPs internatio~albiologique sur I'hCrkditC qui a eu lieu a Berlin le I1 SCptembre 1927, nous avons expose une analyse sur l'arbre gCnkalogique jusqu' B la cinqiPme gCnCration des ancitres d'un cas d'albinisme gCnCral observk sur deux jeunes frPres Ctudiants. De cette analyse rCsulte, que le phknomkne est recessif et il se limite au sexe mhle. Nous avons pousd la recherche plus loin et nous avons constatk que le pPre de ces deux jeunes gens eut aussi d'une autre m6re une fille, qui n'a donnC aucun signe d'albinisme.

ANIMAL EXHIBITS COLEOPTERA H E R E D I T Y I N T H E X A N D Y CHROMOSOMES O F PHYTODECTA Elhibitor, A . de Zulueta, Museo Nacional de Ciencias Naturales, Madrid, Spain

Mounted specimens and drawings (made by SERAPIOMARTINEZ, Madrid) of Phytodecta variabilis are shown. These represent several different color patterns whose genes are transmitted both through the X and the Y chromosome. The patterns include striped ("de lineas"), yellow ("amarillo"), red ("rojo"), black ("negro") and two others not yet fully investigated. This Chrysomelid beetle lives in Spain in the shrub Retama sphaerocarpa, the various patterns being found together but in different proportions in different localities. The distribution of patterns between the sexes also varies greatly. In Madrid the striped pattern is almost confined to the females, while in Granada it occurs with almost equal frequency in both sexes.

390

PROCEEDINGS OF THE SIXTH

LEPIDOPTERA (FIRST SECTION) Organized by John H. Gerould, Dartmouth College, Ha~zover,N e w Hampshire

GENETIC INVESTIGATIONS W I T H SILKWORMS Exhibitor, Carlo Jucci, R. Univcrsitci di Sassari, Sassari, Italy

Various characters have been studied both in their physiological development and in their genetic behavior. Specimens, pictures and graphs demonstrate the chief results obtained in each subject: growth curve, moult number, developmental capacity of eggs, growth capacity, skin characters (yellow and white, opaque and transparent skinned worms), and color of the blood and of the cocoon. With reference to the color of the blood and of the cocoon there are shown: ( a ) the pigment migration from blood to silkworms "fully grown" and "ripe" of several pure breeds and crosses; blood samples in capillary tubes, alcoholic extracts from blood and from silk glands; liiigration curves in hybrids and in parent breeds; ( b ) distribution of the pigments in the various layers of the cocoon in Chinese Golden and European Yellow and F, hybrids of these two breeds and of each breed with White; ( c ) F1 cocoons of various crosses between the "Green" breed and White and Yellow breeds, both normal and "rusted," exposed to NH, vapors (showing the presence of flavones also when the green color is concealed, as recessive) ; ( d ) Green Japanese cocoons normal, found "rusty," made "rusty" with KOH or NH,; ( e ) main cocoon exhibit. I n the main cocoon exhibit there are: ( 1 ) pure breeds, lemon, orange, rose, green, golden, white, flesh-colored; ( 2 ) F, reciprocal hybrids between golden, flesh and white breeds; (3) F, hybrids (segregation 3 Yellow : 1 White) ; (4) F, Golden, Yellow and White hybrids and backcrosses; ( 5 and 6) Fl reciprocal crosses of Green with White and Yellow l~reeds;( 7 ) F1 hybrids of recessive white (Awojiku) with various breeds; (8) F, hybrids of dominant white (Bagdad) with various breeds; (9) various F, crosses ; ( 10) commercial strains.

LIVESTOCK DOMESTICATED ANIMALS Organized by N. I. Vazilozl, Institute of Plant Industry, Detskoe Selo, Union of Socialistic Soviet Republics

The ACADEMY OF SCIENCES of the Union of Socialistic Soviet Iiepublics sent the photographs of the results of studies of the primitive domesticated animals in Mongolia and Turkestan. IMPORTANCE O F PROVED S I R E S I N DAIRY CATTLE BREEDING Exhibitor, Alzi~nalHtrsbandry Departnzcnt, Cornell University, Itlzaca, N e w Y o r k

INTERNATIONAL CONGRESS O F GENETICS

391

SUB-LETHAL RECESSIVE GENES I N CATTLE Exhibitors, 0. L. Mohr and C. Wriedt, The University, Oslo, Norway

T h e demonstration comprised photographs and skeleton material from calves homozygous for recessive genes that cause malformations or abnormalities that lead to death closely after, or in some cases before, birth. Material from the following six cases was demonstrated: (1) bulldog calves (Achondroplasia congenita) , Norwegian Telemark breed ; (2) amputated calves (Akroteriasis congenita) , Holstein-Friesians, Sweden; ( 3 ) hairless calves (Hypotrichosis congenita), Holstein-Friesians, Sweden; (4) short-spine calves, "Elk-calves," Oplandske mountain breed, Norway; ( 5 ) calves with rudimentary and completely anchylosed lower jaw, Lyngdal breed, Norway; (6) calves with congenital contractions, Western Norway. Data demonstrating the recessive inheritance were presented. In all the cases the spreading of the gene in question occurred through a limited number of prominent sires that were heterozygous for the undesirable gene. I N H E R I T A N C E O F S I I O R T E A R S I N GOATS Exhibitor, S. A. Asdell, Cornell Universily, Ithaca, New York

A female goat of unknown history has ears projecting one inch from the head. When she is bred to a normal prick-eared rnale her progeny have intermediate ears. The F, ( % l~lood)by a normal male approach the normal but are somewhat variable. The backcross of son by normal male to the original goat gives progeny (3/4 blood) whose ears closely resemble those of the dam. The condition is one o i multiple factor inheritance without dominance. T h e number of factors is believed to be small. The h o n ~ o z ~ g o s i tor y otherwise of the dam is as yet unknown.

MICE AND RATS T A I L M U T A T I O N I S MICE Exhibitor, N. Dobrozlolskaia-Zavadskaia, Unlz'ersity of I'uris, Paris, France

The object o i this exhibit is to show the result of selection of some particular forms of the tail mutation in our mice. The following five forms are represented by the living animals exhibited: (1) Three different lines of mice (XXiX, XiX and agouti) are tailless or have a filiform tail (these two forms could not as yet be separated from one another). They breed true ~vliencrossed intcr se in the limits of each line. They segregate ior normals and abnormals when crossed with the normals and in crosses between the lines XXiX and agouti. T h e following lines are not I-reeding true; they segregate in crosses inter

392

PROCEEDINGS OF THE SIXTH

se and with the normals, but the abnormals reproduce all or in a considerable proportion one definite form. (2) One line has a compound "kinky tail" ; the tail of the abnormals bears always 2 or 3 kinks. Neither taillessness nor filiform tail have been seen for a long time; only a tendency to helicoidal twisting has not yet been eliminated. ( 3 ) Another line is characterized by a short straight tail with "filiform tip" (attenuated) ; small single kinks are not rare a t the end of the bony part or between the skeleton and the filament. (4) There is a line with frequent manifestation of a "deficiency" of the skeleton (interrupted). This form has not yet been obtained in such a n isolated state as the preceding ones, but cases of deficiency are observed much more frequently than outside of this strain, and transmission from parent to offspring of a similar deficiency is not rare. A father and a daughter, both with a deficiency in the middle of a short tail, are exhibited. ( 5 ) The fifth form is illustrated by a male with a tail in the form of a "pendant." H e gave descendants with "deficiency" tails, but no one similar to himself. The other cases of "pendant" that were observed in the laboratory behaved in the same way.

POULTRY MORPHOLOGICAL CHARACTERS A N D G R O W T H T H E DEVELOPMENT I N VITRO O F T H E FOWL BLASTODERM AND O F T H E EMBRYONIC FEMUR RUDIMENT (MOTION PICTURE) Exhibitors, Canti, H . B. Fell and C . H . Waddington, Strangeways Research Laboratory, Cambridge, England

PLANT EXHIBITS FERNS E F F E C T O F X-RAYS ON F E R N S Exhibitor, Lewis Knudson, Cornell University, Ithaca, N e w Y o r k

FUNGI Organized b y S . Satina, Carnegie Institution of Washington, Cold Spring Harbor, N e w York

ASCOMYCETES (SECOND S E C T I O N ) TYPES O F SEGREGATION O F S E X AND SELF-STERILITY FACTORS I N T H E ASCOMYCETOUS FUNGUS DIAPORTHE Exhibitor, D. M. Cayley, John Innes Hortirultzwal Institution, Merton, England

This is an exhibit of experimental results showing the inheritance of the phenomenon of intra-perithecial aversion (a peculiar form of self-sterility)

INTERNATIONAL CONGRESS O F GENETICS

393

and the different types of segregation of sex and four self-sterility factors in Diaporthe perniciosa. (Marchal) . The occurrence of intra-perithecial aversion or no aversion depends upon the type of segregation of the four self-sterility factors, X Y , A B , during meiosis in the parent ascus. The haplont colonies show mutual aversion, irrespective of sex, when the allelomorphs of the self-sterility factors contributed by the two haplonts are not balanced, for example, X Y A b and X Y A B , Xyab and x Y a B . The colonies show no aversion, irrespective of sex, when like meets like (that is, haplonts of the same genetical constitution), for example, X Y A b and X Y A b , xyaB and xyaB; when haplonts carry one pair of factors (either X Y or A B ) in common, and the allelomorphs of the other pair are balanced, for example, X Y A b and X Y a B , Xy A B and x Y A B ; when the haplonts have no factor in common, for example, X Y A B and xyab. Two haplonts can fuse to form a zygote only when they show no mutual aversion, have no factor in common, and are of opposite sex. The haplont mycelia carrying intra-perithecial self-sterility aversion factors fall into four groups. All the members of the same group show no aversion, but through lack of balance the members of one group show aversion to all the members of all the other groups. But all the groups have the same factors differently combined; hence they all have the same zygote in common, and when certain types of segregation occur one group can give mycelia of another in its progeny. Segregation can occur at either the first or second division in the ascus: sex and the self-sterility factors segregate independently of one another at either or both of the meiotic divisions. PHYCOMYCETES BASIDIOBOLUS Exhibitor, 2. Woycicki, University of Warsaw, Poland

GOSSYPIUM (COTTON) INTERSPECIFIC HYBRIDIZATION WITHIN T H E GENUS GOSSYPIUM Exhibitor, S. S. Kanash, Agricultural Experiment and Plant Breeding Station, Uzbekistan, Union of Socialistic Soviet Repubtics

A MORPHOLOGICAL STUDY O F T H E CHROMOSOMES O F T H E COTTON PLANT Exhibitors, P. A. Baranof and K. A. Miklzailova, Agricultural Experiment Station, Uzbekistan, Union of Socialistic Soviet Republics

394

PROCEEDINGS O F THE SIXTH

ORYZA (RICE) HAPLOID PLANT O F RICE Exhibitor, Toshitaro Morinaga, Kyushu Imperial University, Fukuoka, Japan

In September, 1930, sixteen flowers of a normal variety of rice, Dekiyama, were cross-pollinated with the pollen of a dwarf variety Bunketu-to, producing 13 well developed F1seeds. All the F, seeds germinated. Of those 13 plants, 12 have grown up very vigorously, showed perfect fertility, and were proved to be true hybrids by their morphological characters. The remaining plant was markedly small compared with the other F, individuals. This dwarf I?, plant, though much reduced in size, resembled Dekiyama, the maternal plant. The plant was highly sterile and produced only one small germinative seed. Both Dekiyama and Bunketu-to possess 24 somatic chromosomes, while the dwarf F, plant showed only 12 chromosomes in its roottips. The reduction divisions of the dwarf' F1 plant were not observed, but the pollen grains produced were very irregular in size. Thus there is no doubt that the small F, individual is not a true hybrid but a haploid plant containing only the maternal set of chromosomes. The haploid plant is now growing in the greenhouse.

GENETICS OF WILD SPECIES Organized by Edgar Anderson, Arnold Arboretum, Boston, Massachusetts Exhibitor, Arne Muntzing, Svalof, Sweden

This exhibit consists of herbarium specimens and photographs illustrating work on the genetics and cytology of Galeopsis: pure lines, species crosses, and the artificial synthesis of the tetraploid species G. Tetrahit. Exhibitors, E. M. Marsden-Jones, Potlerne, England, and W. B. Turrill, Kew, England

Herbarium specimens are shown which illustrate studies on the species problem in three genera: Centaurea, natural and artificial hybrids; Ranu51culzts acris, sex expression; Silene, genetics of S. vulgaris ( a very polymorphic species) and 5'. nzaritiulza. Exhibitors, Karl Sax and Edgar Anderson, Arnold Arboretum, Boston, Massachz~setts

Maps and charts illustrate the distribution and distinguishing characteristics of the species of Tradescantia native to the United States. Photomicrographs illustrate points of particular interest: triploidy in T. bracteata; chromosome interchange in T. edwardsiana and in T. refEexa; chromosome structure in T. refZexa.

INTERNATIONAL CONGRESS O F GENETICS

395

Exhibitor, I . W . Gregor, Scottish Plant Breedilzg Station, Corstorphine, Edinburgh, Scotland

Photographs, drawings and charts illustrating cytological and genetical m and relationships in wild and cultivated types of timothy ( P h l c ~ ~alpinurn Phleuvn pratense) are shown.

PLANT PATENTS Organiged by R. C. Cook, American Gejzetic Association, IVashingto~z,District of Columbia

ADDITIONAL EXHIBITS PRESENTED BY T H E SCIENTIFIC INSTITUTES OF T H E UNION OF SOCIALISTIC SOVIET REPUBLICS Orgaizized by N . I. Vavilov, Institute of Plant I~zdustry,Detskoe Selo, Union of Socialistic Soviet Republics

A special group of exhibits was prepared to show the present organization of plant breeding work in the Union of Socialistic Soviet Republics. Two booklets prepared especially for the SIXTHINTERNATIONAL CONGRESS OF GENETICS have been published in English on genetics and plant breeding in the Union of Socialistic Soviet Republics. A series of charts showed the chief results of the studies on the problem of the origin of cultivated plants. The new map of geographical centers of the origin of the most important cultivated plants was shown. QUANTITATIVE VARIATION 1N D I F F E R E N T PLANTS

Special attention was devoted during the last year to the study of quantitative characters. Charts and maps showed the results of these studies for Cucu;bitacex, Graminex, and vegetables. A new summary of the homologous series for Gramineae was presented. There were exhibits on variation found in wild fruits such as apples, pears and apricots. BARLEY

Materials for the new ecological classification of barley (ORLOV)were prepared tables illustrating the results of shown. G. D. KARPETCHENKO crosses between different geographical groups of barley that indicate a great difference in genetic constitution. MAIZE

KULESHOV (Leningrad) presented two tables showing the world distribution of varieties of Indian corn and their classification according to the number of leaves.

396

PROCEEDINGS O F THE SIXTH

OATS

E M M Epresented results of species crosses in oats. In the same section are INDUSpresented new species of oats described by the I ~ S T I T U TOFE PLANT TRY (A. MALZEV, Leningrad), as well as the standard varieties of oats cultivated in the Union of Socialistic Soviet Republics. POTATOES

The INSTITUTE OF PLANTINDUSTRY (S. M. BUKASOV, Leningrad) presented the results of cytological and morphological studies of new species of potatoes discovered recently in South America. There are altogether 14 species of potatoes cultivated by natives in Peru, Bolivia and Columbia instead of one species (Solanum tuberosum) which was supposed to have been the only species cultivated in these regions. WHEAT

Exhibits were prepared illustrating the new classification of wheat. In this classification several new species discovered in the last few years have been included. The laboratory of the ACADEMY OF SCIENCES (LEPIN)presented the results of the study on the inheritance of quantitative characters in wheat.

GENERAL E X H I B I T S GENERAL CYTOLOGY Organized by N . I. Vavilov, Institute of Plant I n d w t r y , Detskoe Selo, Union o f Socialistic Soviet Republics

CYTOLOGY

G. A. LEWITSKY presented the results of a new method used by him and his school in the study of the morphology of chromosomes.