AN A N A L Y S I S OF T H : g E E W I L D P O P U L A T I O N S OF D R O S O P H I L A S U]3OBSC URA

BY CECIL GORDON, ~ IIELEN SPURWAY AND P. A. R. STREET

Deyart,;ne~t of Biometry, U~versity Cotlege, London

(With Plate II and Five Text-figures)

C O N T E N T S PAGE

i. In:groduc~ion 37

2. ~a%er[~l and method 38

3. "i~xperimenM] procedure '÷ 4]

No:go on culture oonditions . 42

4. Descr[p%ion of %he 21 and F. segregations 4:2

5. Effects due 4o single ~utosoma.] mu:ga,tions 45

6. 9"alse mutan%s ... , . 80

7. ltfa~formed abdo.msn . 61

8. Vein ubnormalities, the inheritance of which is unprecHctable 63 ~)escriptbn of pheno:gypes 63 The F~ seg,'egations 66 The E2 s~greg~t.ions 66 Iden~i%y tests . . 73

9. ]~o.ughs 79

~0, Estimation of the frequency of muta, ted loci, and the comparison of populations 82

II_ Discussion 84

12. S u m m a r y 88

~eferenees 89

Expla.ua%bn of Plate II . 90

1. INTRODUCTION

~[NDIVIDUALS he~erozygous for recessive l~lutations have been found in wild populations of many species of animals, e.g. DrosopAih~. species (Tschetverikoff, 1928; Timofgeff-~essovsky, 1927a; Balkaschina and Romaschoff, 1935; CJordon, 1936), Dermestes w@i,us (Philip, unpublished commnnieation), Ga.m~narus chevreuxi (Spooner, 1932) and Perov~yseus (Sumner, 193@ Similar heterozygotes also exist among plants, as is shown by the increased variabilRy, but average decrease, of the growth rate and productivity, the segregation of sterility factors and abnormal forms snch as chlorophyll defects, produced by inbreeding plants from

1 Carnegie Teaching Fellow, Depar tment of Na:gm'tfl History, Aberdeen UniversRy,

38 Analysis of Three Wild Popv2ations of D. s u b o b s c u r a

species that are normally cross-fertilized, e.g. red clover (Williams,1937 a, b) and S',inc~Tis cdbc~ (Sal-tykovsky & Federov, 1936).

The frequencies of flies he£erozygous for various mutations have been used to compare wild popuIations of D~'osophi~c~ mela'r~ogaster (Dabinin et al. 19.34, 1936), D. ~seudoobseurc~ (Sturtevant, 1937; Dobzhansky & Queal, 1938) and this paper reports the analyses of samples of /). s~&- obsc.z~ra from three localities in the south of England. Stocks existed in the two for m.er species which enabled autosomal lethals to be detected and thelr frequencies measured. ~.n D. subobse~,ra, however, we have had to use a method of analysis by which only sex-linked lethals, and mutatious producing a visible effect when homozygous, could be d.egected.

The work was designed by Gordon, and the inbreeding, extracting, and testing for identities was done jointly in the season 19.36-7. In August 19;37 Gordon went to Aberdeen, leaving the linkage of the vein characters with Spurway. An analysis of the Slough sample by Street and of the New Forest and Studland samples by Spuzway, were accepted by the University of London as theses for the degrees of ~faster of Science and. Doctor of Philosophy respectively. This paper has been re- written by Spurway from the two theses.

Oomplete records of the inbreeding results have been given to the Statistics Archives of the. British Museum (Natural History).

~, ~ATE2%IAL A27D METt~OD

The three populations analysed came from Slough in Buckingham- shire, from Eingwood in the New Forest, Hampshire, and from Stndland in Dorset (Fig. 1). As the flies were trapped they may have flown from great distances, so it is not known over what areas the populations ex- amined extend.

The Slough sample came from the grounds of the Biological Field St.atiml of the Imperial College of Science and Technology, where Gordon (1936) had caught his flies in 1933 and I934. Traps were put down. in two places in an area of 8 acres, the altitude of which was 180 ft., on the grass- less patches under apple trees in an orchard, and white willows (Se/i;s alb@ on the banks of a stream.

The New Forest flies were trapped in. two areas at gingwood.. The first of about 19~- a,ores surroun.ds a private house, and is 138 ft. above sea-level, on the Barton beds. It contains a gravel 5ii1 covered with gorse and. heath descending to a stream which flows through pasture land. with sallows (&diz ci~zerea and, auri~e), marshes, and a small oak and birch wood. The second[ area is a wood of 9 acres, half a mile from the

C!EOIL O0ICDON, ~{ELEN SI~T_rRWAy _~ND P. A. R, STREET 39

i

.q

I

~T ~ -<

F

i bJ~.

O

o

F, ©

8 %

S

o

.4

40 A'~zcdys'is of Three WiZd Popzdc, t ior~s of D. s u b o b s e u r a

first on a clay soil 2741%. above sea-level. The trees are oak, ash, al~d sweet cllesbnut, with a sprinkling of birch and. aspen, and a hazel under- growth.

At Studland flies were caught over a much larger area, which was only 10 t'~. above sea-level. The soil consists of Bagshot beds covered with blown sand, and fl~e vegetation is add heath with scattered, marsh and shrub. The sample was caught in two litde woods of birch and sallow shrub 1600 yards apart. One of these was over SO acres ir~ area and the caller only just over 8 acres.

The ~rapping was done from June go November 1936. As Naehtsheim (1928) and Dubinin (I928) have shown ~11a$ only the last fertilization is effective, the offspring of eada wild female caugh~ are assumed to be the offspring of a single male. Dominant mutatimls will be revealed in the ~,ild fly, or in the 2g], if they had been present in the mate. The ex- pectations for the segregations of single recessive factors are given in Table I from Gordon (1936).

TABLE I

Possible ty~Jes de.ri,ved from si,~.qle fe.mc~es fert£ized in tt~e ~oild Wild parents F1 F.~ cultures AA? x A a ~ [ 25 °2/~5 all _&A, 50 % (IAA. : IAa), A s 9 x AAc~j I ~ A : 1As % (1/kN :2As : laa) Aa9 x A a ~ 1 ~ : 2Aa : ] a a Not requkred SS? × syc? as? Sy~ INS? ; INs?, iSyd : lsyc~ Ss~ :< Syc~ lss9 : ISs?, lSy~ : isyc~ No~ required

A and S denote normal autosomal and sex-linked genes respectively, a and s theh" recessive ~llelomorphs.

Effects dee to several factors might be detected. Less than a quarter of the flies in a culture segregating for such an abnormali ty will show it, IPor example, if mutations at two loci were equally essential a ratio of

15 : 1 would be expected in. one out of six~een cultures. Autosomal lethals cannot be looked for in this species as the material

for the necessary testing stocks has not yet been collected, but Gordon (19.36, q.v.) has explained some of his abnormal segregations of auto- somal recessi.ves as being due to their linkage to a lethal. He describes the segregations expected from el1 possible distributions of the visible and. l.ethal recessives a~d their normal aibbm.orphs between the four homo- logous c h r o m o s o m e s o f t h e t w o wild parents.

0ECIL GO1%DON, H E L E N SPUIR%¥AY AND 1D. A. R . STREET 41

3. ]~XI~F~RIMENTAL PI%OCEDURE

The wild females trapped were presumed to have been mated, and were put singly into cultua'e vessels.

Thlrty-two males of the E 1 of each family were examined for mutant characters, and the females emerging at the same time were counted. 8ex-Hnked lethal factors would be detected by this method.. I f any abnormality was observed among the males, ~he females were examined, to distinguish between sex4inked and aatosoma.I mutations, and com-- plete counts were made of the culture. I f no abnormalities were observed the Y z was discarded when the thir ty-two males had been examined.

From the/~1 flies, sixteen F~ rantings were made up in pairs. Sixty~ four flies were usually examined for all mutants in a series of F~ cultures )~ in number. The number of cultures examined for all mutants was theoretieally eight, but whei1 less than eight cultures in the av 2 produced flies, ]~ is less than eight, and when one or more of the first eight cultures failed to yield sixty-four flies, one or more extra ones were examined in detail, and ]c is greater than eight.

I f an abnormaIi V suspected of being genetic was detected, the culture was not discarded when sixty-four flies had been examined, bu t counted to the end, solely for the segregation of tha t mutant. Similarly, cultures ill excess of /c were examined and counted for the segregations of all mutants found in the first ~ cultures. All abnormalities not observed in the first sixty-four flies examined in one at least of the ]~ eult~u.es have been ignored.

Any character thought to be due to a mutation was carried to the F3 by mating abnormal flies together, by mating them to their wild- type sibs, and by mass marinas of those wild-type sibs. If the abnormal x abnormal crosses did not produce offspring, and confirmation depended on abnormal x wild-type and mass wild-type rantings, a negative was not assumed to have been proved until an _F 4 and sometimes an Fs had been examined. No detailed records were kept of these crosses, and the number of cultures and flies examined depended part ly on the tempera- ment of the investigator.

When a mutan t was confirmed, the information which was gained about its ~dability, fertility and penetranee from the three kinds of cross made up with the ~ies fl'om a segregating F~ culture was considered, and it was decided whether or not a stock of it should be kept. Theoretically, all mutants were kept until their identiV, or allelomorphism, with other mutants tha t resembled them phenotypically could be tested. These

42 A'JaaZysi.3 of Three PViZd Popule~*iooa.~ of D. s u b o b s c m ' a

tests were made by mating the two forms together and examining their offspring. I f the parent forms were not idml~.ical or allelo,aaorphic, all the offspring would be wild tyl?e. Bearing this in mind, no :form that was sterile, and had to be kept by heterozygote × heterozygote rantings, or any that appeared in small numbers in the ~v and t~ 5 were cultured. These latter always appeared inftue~eed by modifiers, and since the petxetrance could be small, examples might occur where, though the two stocks contained the same main gent, they might differ so muel~ in genotypic milieu that it would be diNcalt to demonstrate it.

Chromosome maps are tieing constructed using the mutants collected in this investigation, and thus further facts are known about the genetic behaviour of some of them. These facts wilI be given in the description of the mutant concerned.

Note on c.ttgure co,~uZ~tions

Two food media have been used, the ordinary maize-meM ~nd mo- lasses D.roso2k.da food, and a similar food of which about 12 % by weight of the ingredients is dead yeast. This food is always used in vials, and with the exception of some F~ generations all experimental cultures are reared on it at a temperature of 18 + 1 ° C. The life cycle in these condi- tlons is 28 days. Stocks are kept on the ordinary food in bottles in a temperature v~rying }aetween 1.2 and 22 ° C. according to the season. At 1.8 ° C. the life cycie is about .-35 days.

5Iany characters were found which were strongly pe~aetrant in vials at 18 ° C. and very weakly so in bo~ties at 15 ° C. Noue was leered which was more penetrant at 15 ° @. than al, 18 ° C. Bag this may be because, under the conditions of this work, the latter, if detected at all, would be considered too poorly penetrant eo be preserved.

The flies were mated for 5-8 days before being put into the vessels in which their offspring developed.

~. ])lgSCRIPTION 01~ THJIS ~ I AND i# z SEGREC$~kTIO~TS

The offspring of" forty-sevea females from Slol~gh, forby-two from. New ]~orest, and fifty-five from Studl.and were examined.

No sex-linked lethals were observed in the populations, .for though a deficiency of tomes suggestive of a lethal was oceasi,)nMly observed in. an /~'~, the geae was never confirmed by similar sex ratios appearing in half the F2 cultures. The number of maIes in a poor culture is known to be less than the number of femMes.

CECIL G O1%DOIRT, I-IELEN SPUIRWA.Y AND P. A. ~ . STR.EET 43

G-roups of files showing various abn.ormalitJes appeared in the F1 and p~ cultures. The ratios with which these abnormals appeared did not seem incompatible with the hypothesis that they were due to recessive mutations at autosomal loci, though the expected 3 : 1 Mendelian ratio was very rarely observed.

There were, however, three immediate exceptions to this generaliza- rich. The first, M_c~lfo~'~ed abdo~e~, in the F~ of S~udland 54, is described on p. 61. In the second form, which was from the P~ of Slough 42, the wings of some females were shortened and withered. This character is sex-limited to the female and is due to a mutation at a locus on the X- chromosome. This form~ called w.it;ered (}el), will be considered in a separate paper. The third was short b~istle from the F~ of Studland 55, whiel~ segregated nineteen abnormM to eleven wild type. I t was not inherited.

The offspring of several abnormalities were all wild type, even though the parent tties had appeared in great m~mbers, and sometimes in the expected 3 : 1 ratio. Their morphology will be described later (p. 60).

The phenotypic effects that segregated most frequently in the _F~ were vein additidns and deficiencies, and flies showing these abnormalities were recorded from three F~ enltures. As these characters can be attri- buted to the action of single genes, oNy if other hypotheses are in~oked, the effects wilt be considered separately from the autosomal recessives (p, 63). 1~ueh less frequent but still commoner than other abnormalities were derangements of the facets of the eye, called ~'oughs. Their original 3' 2 segregations can be arranged in a series ranging from that oharactero istic of vein abnormalities to that expected for autosomal recessives (p. 79).

The remaining abnormalities appear to be due to single genes, though only a few of them can ever be assigned to the five "ranks" in which D. meZa~oyc~ste.r genes are arranged. None of them appeared in the -FI cultures.

The F2 segregations are shown in Tables I I - Ig . Their phenotypic effects and behaviour will now be described. A character is said to behave normally when it is completely recessive in an F~, when a 3 : 1 ratio is

observed in an x~, and half the flies in a baekeross show it. The crosses giving these segregations were those made to test linkage and estimate the distance between loci. The figures will be published shortly with the autosome maps of the species.

44 Analysis of 59hree Wild Populations of D. subobsoura

Slough poy~datio.n.

TABLE Ii

F~ segregations of "autosomal recess.ires" No. of No. of

cuRures Na, of flies in all No. of examined cultures segregating abndrmM

Family ~ ~futant for mafi~nt segregating ctfl~ures flies A 5 thin bristl~ A 5 1 63 I t B 8 b~bb&u 8 2 109 17 D 10 tbrk~dD 11 2 305 38 :[~] 8 reduced bristle 9 2 185 13

.roo]' 9 2 179 13 tipped .wi'ng 9 2 117 5

G 8 shaven 9 2 321 t4 K 8 curled1; 8 l 124 S

8 5 cut 5 2 130 1~ 1I 8 poppy 12 6 344 39 t3 8 ~utspread 8 2 373 110 14 8 po~j)y 8 :~ 46 8 18 7 pa~x~less 7 2 i16 5 34 S lbr/'ed~ S 1 76 7 4"4: '7 pointed~ 7 1 86 14 45 12 outspread~ t3 6 427 49 50 l l /)at 1t 2 ~7 4 51 7 twisted 7 2 155 35 91 8 incomplete ~roesvein 12 5 829 150

Thirty families yielded no instants in the F~, the ]~ ~alues being 3(9), 16(8), 6(7L 4(6) and 1(5}, and o£ ~he 226 culture~ 179 contained over 50 Kiss, 40 confabbed between 20 ~nd 50, ~nd 7 contained uader 20.

New ~orest po;oulat.ion.

TABLE III

F., segregations of" autosoma~ 'recessives " No. of h'o. of

caRures No. of flies in all No. of examined cultures segregating abnarmal

Fam~y ~ h'[utan~ for mut.aut segregating cultures flies 1 10 r~urled 10 2 132 11 2 9 thin brlstle e 11 I 92 3 3 8 thi~ 8 3 315 58 7 9 ruby 11 4 513 89

11 9 bro~:en bristle 10 2 20~ 38 12 7 po2py 7 2 i73 38 13 8 forl.;ed~.~ 9 1 I15 27 16 8 frosted 5 1 102 22 17 4 interru2ted vein 4= l 130 14 20 8 eyeless 8 ~ 145 12 22 8 cur&d 9 1 106 8 25 5 tomato 5 I 88 16

zurled 5 2 95 4 26 9 blac~ bdly ] i 4 512 46 42 6 short anterior 6 I 3!) 5

c?'oss'b'e~'~. 55 10 plexus" 10 4 362 82

Twenty-seven fhnfiHes yielded, no mutants i.n ~he Jr._., th~ [c values being 1(12), 3(9), 15(8), 1(7), 3(6), :~(5) and 2(4), and of the 202 cultures i68 contained over 50 flie~, 23 contained between 20 and 50, and 11 contained u~der 20.

CECIL GORDON, ~-~ELEN SPUR~rAY AND ~ . A . I~. S T R E E T

T A B L E IV

;S'tudh~nd population. F. segregations of "autosomal recessives"

No. of No. of cultures No. of flies in all No. of

examined cultures segregating a.bnormal Family k ~[u~ant. for mutant segregating eult~lres flies

7 8 shc~ve~ 8 1 88 14 8 8 cinnabar 10 1 55 13

10 8 maroon 9 ~ 213 54 2t 7 ho~ry 7 1 68 5

bubble2i 7 2 108 l 9 2').?, 15 bubble~ 15 3 85 5 23 12 "forl~ed" 1~ 4. 303 46 26 9 crinkled 9 5) i22 18 30 8 crumpled 1 t 5 369 40 32 9 baioT~ 9 2 I56 14 33 5 dwarf 5 1 47 1 39 9 jaunty 10 1 92 1 43 9 wispy 11 2 162 12 50 8 bubble~o II I 99 2 52 8 a, bdo~i~ial sider tO 2 150 28

shave~s~ 10 5 395 58 54 8 b~bbles~ 8 2 139 6 58 8 bubb~s~ 9 3 191 5 .60 8 J~vdn.~ 9 2 146 18 63 8 ~Vsta, ngi~zy wi'~g 9 2, 376 32

45

Thirgy-ssven f~miHes showed uo mutant in ~he F., ]the k vMues being 4(10), 7(9), 19(8), g(7);0(6),-2(5) and .1(4), and of the 297 evJ~m'e~ 230 contained over 50 flies, 57 c'6n~Mned.between 20 and g0 and I0 contained under 20.

5. EFFECTS DUE TO SIN~,LE AUTOSOMAL MUTATIOlgS

(1) Eye p{ymentation

The subsequent segregations of the eye-colour genes show that with one doubtful exception t~iey behave normally, as would be ezpected on the analogy of those in D. melanog~ster.

The wi Id4ype eye co]our of D. subobscu~'a is otd rose. This varies con-

s iderably in br ightness dur ing the ]fie of the fly. The br ight papa l eye

darkens very quickly after emergence and remains dark while the wings are expanding and the body colour is developing, It then tightens again so t h a t the eye is a br ight crushed s t rawberry colour when the fly is

about, a day old. After t h a t the co]our gradual ly deepens, becoming old

rose, and f inally a cr imson purpl ish brown. There is a wel l -marked re.-

flexion centre. The testes have the br ight orange co]our of those of D. pseudoobscura and like them are ao t coiled.

popl)y (.pp), Slough 11 and 14, New Forest t2. A br igh t scarlet eye

colour resembl ing vermilion in D. mdanogastsr. The reflexion centre is

very dark and clearly outl ined. There is some evidence t h a t flies homo-

zygous for this gene develop more quickly t h a n flies con ta in ing its wild-

46 A~aalys.is of Zl~ree WileI Popzdatio~a,s o f D. s u b o b s c u r a

type alldomorph, as .2)o~)py flies must be removed from cuRm'es more frequen~iy to ensure their v~rginity.

The segregating cultures in the Slough 11 Fz were:

+ ~ o ~ p y 99 17 50 3 60 5 39 3 35 1 22 10

Thus in five out of the six segregating cultures there is a deficiency o-f

i~)o})i~y flies. As poppy is a normal recessive, this deficiency is perhaps due to a lethal gene linked to poppy, the wild female being heterozygous for both mutated loci which were on the same chromosome (Gordon, 1.936). Using Gordon's formulae and the figures observed ir~ the Slough 1t segregations, and ignoring the heterogeneRy among the deficient cultures, tlhe genetic distance between the visible and lethal recessives is 28 traits ignoring double crossing-over, and the expected proportion of segregat- ing cultures deficient in flies showing the visible mutant character is one- halt', whi.ch is not confirmed by observation.

As 81ough 11 and Slough 14 were both caught on the same day, i t is possible that these females or their mates were closely rda ted flies: This is made likely by the suggestion of a similarly situated lethal in Slough 14, though the figures are ~oo small to be significant:

16 4 t 4 1

8 3

The infertility observed in the Slough, 14 cultures persisted in the stock of poppy, isolated from them, until, on being ]proved identieM wRh "~he other two stocks, it was discarded.

d~z~zeba, r (c~), Studland 8. This is a darker, duller all.elomorph of poppy. The eye colour of the offspring of a ei~7,)~c~bar "< ~po~ppy cross is mid way between those of the parents.

to~c~to (to), New Forest 25. A russet or tomn,to-coloured eye, resem- hlin.g the combination scer~et-elm'e~ in D, rr~elc~zoqo, ster. The eye is opaque and there is uo ret~exion spot. The fertilRy .is decrea,sed and the stock is mait~tained by homozygote x heterozygote mat, lugs. The testes sheaths are colourless.

maroo~, 0~La), Studland 10. A dark opaque eye cotour. In young-flies t;he eyes resemble da.ret w-heu seen with ebo',~y in D. mel~e~oqc~ster. The colour is always sligh.tty darker than d~'reg of a corresponding age, and in

C E c r 5 0 o ~ . n o>~, tt~zE>~ SpuI<w,gY A>~) P. A. t~. Sa 'gzEw 47

01difHes becomes an opaque brownish purple with an apparent surface texture reminiscent of graphite. The testes sheath is lemon yellow.

~'uby, New Forest 7, merely caused the wild4ype eye colours at any stage,6o be cleaner and brighter. I t was very di~Fmnlt to score, and could. oedyy be recognized by comparing the eyes of flies wi{h identical body coiourg ~ud therefore presmaably of the same age (the body colour also ~arkerdng as the fly grows older). Some workers were unable to recogmze it at all, and tinting the light made scoring no easier. The stock has now b e e n discarded. T h e iv 2 segregat ions were:

+ ruby 184 30 108 ]7 61 8 71 34

i.e. the ~bllormal flies are deficient in three cultures al~d a}e present in excess in the fourgh if a $ : 1 ratio is expected. There is no obvious explanation for this, and the figm'es are probably not scorn'ate enough to be discussed. Phenotyt)ically-the ruby eye and the wild-type eye looked as though they contailled the same pigment which in the former de- veloped more slowly. If this is so, the mutant differs from the other eye colours, ~nd cannot necessarily he expected to behave normally.

(2) Wing carriaye, texture, aug st~a?e

A chemical alteration would be expected to be more directly depen- dent on the primary chemioat action of the gone than a structural m a t formation, and therefore irregularities in the manifestation would be more likely to occur among characters of the latter type. 0nly a very few structuraI .forms have been found that behave in the theoretically orthodox way, i.e. wigh one of the pair of homologous loci mutated, the fly is wild type; with both of them mutated, it shows the characteristic abnormality. And as more work is done with the characters, which means that ghe gone is put into more and more genetic en,drom.nents, the number about which this can be said decreases.

The map distance of outspread (ot2) from severn1 other loci on the same dlromosome has been measured, and the character has always been manifested normally in the crosses. In this form, found in Slough 1.3, the wings are extended laterally, the costal margin usually beillg at an angle of about 40 ~ to She median line of the body. in young flies when the wings are newly expanded they would appear in section like rams' horns, being held dorsally to th.e body at the base and curving downwards to a

48 A.'racdysis of Tta~'ee Wild Pop,ulc~tfo,~7,s of D. s u b o b s e u r ~

curled tip which is in a plane ventral to the abdomen. As the fly ages the wings uncurl, and when two days old they are straight and point slightly downwards. They always retain a cockled surface. Sometimes the wings d.iverge from the body at at1 angle which is less than ,10 °, and occasionally they merely fait to overlap over the abdomen. The animMs cannot fly. The abdomen is slightly flattened dorsoventraIly.

Though less work has been done with bat from Slough 50 no abnormal- ities have been observed in its penetrance. In this form the positio~ of the wings makes the animal resemble a butterfly set and pinned in a collection. They are extended laterally, the costa being at right angles to the vertical median line of the body or sloping backwards very slightly. The wing membrane is never folded or crumpled, but expanded, showing every dotal1 of venation. Whe~ the wings slope backwards they may occasionally slope upwards at an angle of about 30 = to 5he horizontal. The fly is slightly larger than the wild type and the abdomen is flattened dorsoventrally. S~ooks containing it are diffmult to culture as the flies live for little more than a week. They canno~ fly.

The eu.rled(c~) that appeared several times in the New Forest popula- tion is always completely penetrant, bul~ occasional heterozygotes show a marked concavity of the wing. They can be clearly distinguished from the homozygotes so that the utility of the character is not impaired. In the homozygote the point of the wing is tilted upwards and the texture is opaque and. rough.ish. In old flies the wings may be dried and frayed. In most specimens the tilting is exaggerated into an upward curl, the wing being lifted away fl'om the abdomen. Both the margins are curved up- wards so that the wings are concave. The curvature of the imzer margin is frequently very pronounced, oausin.g this concavity to be turned out- wards. When the curvature ~akes this form the wings diverge posteriorly. The character resembles oz~rged in D. 'meZa,swgc~st, er= and the animals cannot fly.

This phenotypic effect was %und in. New Forest 1., 22 a>d 25. !n 1 a~ad 25 it was shown to be due to a mutation at the same locus by a mating test. The character from 22 was Iost before it oou].d be tested.

The c~trleds from all three Nmilies appeared 100 ~; pe~,letrant in the. Fs, and, with a good viability in 1 and 25. There:fore in both of these the tow penetranoe was most probably due to a, linked lethal. New 1?'crest 22 bred true, bat appeared very inviable and ind:ertile in stock, so assure ing tha t the mutation was at the same locus as in tlhe other two families, the low manife.statioa may have been due to a linked semi-lethal or some

CECIL GORDON, HELEN SP-UT4WAY AND I.D.A. IP~. STI~.EET 49

disease. Bu t treating' the five segregg$ing cultures statistically R is found ~hat they are l~omogeneous:

+ curled

New ]~orest 1 54 7 1 67 4

22 98 8 25 27 2 25 64 2

X-"- =3-75, ,t~=4; p between 0.5 ~,nd 0.7.

Therefore it is possible ~hat the cu.rled on the three occasions in which it was obtained was linked $o ~he same lethal which would be 20.72 units away, ignoring double erossing~over.

I f this interpretation is coffees (and ~he reduced xdabilRy observed in

J ]~ig, 2, T~visted.

ghe New Forest 22 may contradict it), R is more probable that all three females were fertilized by the same male, or ~hat ai1 three females were sisters (or all three males were brothers), thg~ that the two genes are ex~remely common in the population.

In Slough 5I, twisted (Fig. 2), also belie~ed to be a normal recessive, ~he expression of the character varies with the length of life cycle. When this is abo~t 2S days there are shallow ]ongRl~dinal furrows in the wing. At lower temperatures the folding of the wings becomes more pro- nounced and they fail to expand at the base. In the mos~ extreme form ~he wings are fluted and convex, like cockle shells, a]ad twisted at the base so that they cross each other dorsally ~o the animM's abdom.en. This is ghe only abnormality discovered in this investigation in which tl~_e

Journ. of Genetics xxxv r~ 4

50 A.,j,Jd7/,~.,i.~. of Tb'~e l, Vg.d Po2>~dcLfio'~,.s' o/" D. s u b o b s c u r a

e.xpz'esslon is i~zcz'eased by a ]?~'oiougat.ion of the li.fe cycle. When the wing is expaLad.ed but f~a'rowed, flight is reduced to fleadike hops.

All the remaining abnormatit.ies of wing texture and position, may be impenet:ranl,, i.e. flies kt~ow.~ I:o be homozygous for t i e mutated gent may not d~ffe.:r ,Jbserva, bly if'ore the wild-type fly.

.l~'~g. 3. Hoary.

hoc~.r~y (h.o) (]Fig. 3), obtained from Studlau.d 2-i, was recognized by a longitudinal folding of the whtgs. When l~b.e expression [s poor, a small length of one or other o [ ihe margins is fokl.e.d back o1.1 to the Wi]]g sra'fac% causing the rest of ~lle ma,.['gh/ to be puckered_ I~l ttle typica] ~orm ~he wi,~gs are cru.m]?led longRudhmI]y ms tho u~h the}" h.ad been grabbed and cru,shed in the hm.zcl, while s~ill damp. Scored by this character it s,ppea:red

CECIL G0~DON, HELEN SPUPaVAY AND P. A. ~. STR~EET 51

~dth a penetranc, e of 29 ~ . A stock has now been selected in which the penebranoe is over 80 %. The hairs between the eye facets are blanched, ms!dug the eyes appear hoary, This second phenotypie effect is thought to appear in aI1 flies homozygotts for the mutated locus, and therefore may be deduced as being at the end of a shorter reaction c~aiu from the primary gone action than the crushing of the wings. This ibrm is an exact mimic of the sex-linked lm, le (Chffstie, 1939) which has now been

lost. Sever~I bubble (b~) wings were fonnd resembling bubbZe in D. moistm-

gaster. I1~ the young flies one or occasionally two large drops of fluid appear between ~he laminae of a wing, separating ~hem, and producing a heavy globule preventing the animal from flying. Through breakage of these the flies may be stuck in the culture vessels. As the animal dries, these blisters either burst, or more likely the fluid is withdrawn, leaving patches of dry, distended and crumpled wing tissue. In these patches the veins appear unpigmented.

In that recorded from Slough 13 where the blister was distal the pene- ~rance in the F~, was 60 °/o , but after selection it rose to 90 ~/o- In tha t from Studland 22 the blister was near the inner margin of %he wing. In some flies i~ was filled with a dry blackish foam, bat in most of them it was emp'ty. The crumpled tissue of the blister caused .the inner margin of the wing to be curled upwards, producing a superficial resemblance to roof in D. mek~nogaster. This appeared wRh a penetrance of 24 °/o bat was selected until a. penetrance of 100 ~/o was obtained~ which w~s of course lost after outcrossing. These two forms Were not identical. In Stadland 58 the actual drop of liquid persisted for some time before drying up. It appeared with a penetrance of t0 %, and after 12 months' selection the penetrance found in a stock bottle was 8 %. I~ was never outerossed or crossed with the o~her bubbles. The bubbZes found in Studland 50 and 54 were discarded and no later counts were made.

Two forms were found in the New Forest population which had sparkling opaque wings resembling frosted glass. One called frosted, and found in family 16, was behoved to be a normal recessive. I t has now t~een lost. The other from family 42, called sho~ a~d,e~'ior c~ossvei~, after the otlier character which it showed, was incompletely penetran~. The effect., though continually selected., gradually became rarer in_ mass cultures until a f t e r eight generations it disappeared completely. The phenotypioally normal stock has now beela cEscarded.

uj)stc~ndi W "w.f~W, Studla,nd 63. The wings of' the fly were extended dorsally a.~ right angles to both axes of the body, like those of a dead fly.

52 A'~c@j,~'is of TA,re¢ WJ, d Popvffatio'~s of D. s u b o b s c u r ~

The anterior margins diverged and posterior margins converged, The ~aimals could not fly tuld frequently lost their bMance while walking.

Three cattnres segreg~ted:

1 7 I 20 ] 10 ].0

63 .2

The stock after bei~lg selected for penetraz~ee for M)otet ~velve genera- tions, and iu m~controlled conditions was 27 % penetrant in Site male and 01 ~/o in the female, and combined about 38 ~/o. If we except the 3rd culture and consider that the segregation differs from the ex~ petted 1/~ solely because of the !ow penetranee of the mutant, the penetra~ce in the first two cultures is 39 %. Th.e small segregation in the 3rd culture may be due to the influence of suppressor% though the behaviour of this character does not suggest that it is influenced by modifiers. The files may have been heterozygotes, but the stock was discarded without being outerossed, so it is not known whether the character was ever dominant. A [inked lethal factor may h~tve been present, ])at gee small number of cultures examined makes it impossible to die, guess where the let]]M was situated. The most probable situation, hewer-st, is that both ~u.2,ta'~rdi~ 9 and the lethal were present in the wild female, one on each of the homologous chromosomes. F this is so, the expectation (Gordon, 19:36) is that in 0~-/4 - cultures c/.3 of the flies will be mutants, and in (1 -c~)/4 cultures 1/~ will be mutants, when 100 c is the genetic distance between the lethal and ~])8~c~.~di~zg. This is unsatisfactory, as by equating c/3 with 2/6g x 100/39 and substituting the value so obtained in the expectation of cultures it appears that only just over 1 ~/o of the cultures showing ~he rear, ant should show it in the reduced prm portion.

eri~dcled, Studtand 26. The wings were crinkled as thou.gh they had dried defectively. This character was completeIy impenetrant in bottles, and in vials its penegran.ce and expression fluctuated greatly.

c','~U)~ed, Studlan.d :30 (Fig. 4). The phsnotypic effect of this form. was also dependent ou the gsnotypic m].iiem When the expression was good there were two well-marked fomns. The first was a downwa,rd curvature of the tips of the wing over the abdomen. Tits margins and wing surfaces were perfectly smooth. In the second, shown in the il]ustrat~on, %h.e wings were furrowed longieudinMly, and the margin, was bent and folded. Ia stock th.e psnetrance was 100 %, but dropped oonsidera.bly on ont- crossjn.g. The last two mutants, though ea.sily scored, were discarded a,s impractic~J)]e for linkage Costs.

CECIL GoBDON, HELEN SPUIRI,VA.~7 AND P. A. ~ . STRt~ET 53

The roof and t.i~ped fl'om Slough ]~ and thejca~zty from Studland 39, all of which resembled very c]osely their namesakes in D. ~zdc~zo2c~s~e'r, were so frequently impenegrant that no stocks of them were extracted.

Finally two semilethals were found. A e~o-[ed was found in Slough K in which She wings were furrowed longitudinally and ~he bristles sligh~sly bent, and a.n m~ts~rec~d c]osely resembli~ag Slough 13 o~tffl)read appeared in Slough 45. ]froth these forms were reobl~ained among the o/~spring of wild4ype sits of the F~ abnormals, but no culture could be extracted of either. I t is no~ known whether the deficiency with which both of them

[Fig. 4, Crumpled.

occurred in the -~2 segregations was due to reduced viability or whether the characters were also impenetrant, as the cultures foI c~r~ed, were discarded in the Ps, and o~ts)~read failed to rea.ppear in later generations. It is probably Similar to the o~tsbre~d~ed, recorded by Gordon from the same population (]~o6), and dlagnosed by him as viable, subletha.l or lethal in effect, according to the genetic environment.

Two genes were found affecting the actual shape of the wing. e~t, Slough 8. One or occasionally two large ~rcs were removed from

the wing margin, usually on the inner side. The ]?enetrance of this cha,raeter was much influenced by modifiers, selection raising it from 23 to 65 ~/o. i t has now been lost.

poi~ted~4 (2~4), Slough 4~ (Fig. 5). Two wing shapes are found in eultz~res of th.is gene. The first observed is narrower than the normal wing. In ~he second form, which appears at high temperatures, t.he wing is broader and less tapering tha.n the wild type, and instead of ending in a pointi t has a convex bay in it, resembling ~rwazcccted in D. ~,e2c~.,wyc~ster. This polymorphism is reminiscent of that of poi~zged (Gordon, 1936), though both effects are less pronounced, and flies of bo~h kinds are

'm

54 A~~cdyais of Three W,2d Pop~Za, t*ogaa of D. s u b o b s o u r g ~

less vialsle f2t~n wild t, ype. The two loci are nob, however, on the sa,me chromosome.

(e)

Fig. 5. Polnted.~,~: (a) wild-eype wing: (~) form whioA1 s%regated in F~ fl:om Slough 44; (c) "t.runeated fom~t"' observed a~ high temperatures. ~

(3) Ve~,ati.o~z Three of gee numerous segregagioas of alJ~ormalities in the wing

vena,tion are cousidered to be due to single recessive m u t a t i o n s at, aa to- somal loci.

yZez~s @g), New Forest 55. This w~s one of %he few characters t h a t segregated in tb.e expected ratio of 3 : t. I~ h~m been worked, wi~h ex-

tensivel]~ and is thought to be a normal autosomal recessive. An extra vein runs from the junction of the 2nd longitudinal vein and the costal vein, t~owsrds the 3rd. longitudinal vein, sometimes meeting it. This extra vein. is splayed a, nd uneven in thickness, frequendy bz'anched, and often arising from a delta formation or splod.ge of vein tissue. The posterior crossvein is abnormMly long, joining the 5th longitudinal vein at ~n acute angle very near the margin. This vein is also thick and uneven, splaying out at its junction with both the 4th and 5th longbudinM, veins. The short distance of the 5th longitudinal between the junction of the crossvein and the margin has the same olin.ratter, as occasionally have the anterior crossvcin and l~he 4th longitudinpJ. This splayed and uneven form of vein is called in all suJ~sequent descriptions "plexus tissue". Spots of vein tissue znay appear between the 2nd longitudinal and costal veins, parMlel go them both.

More doubt is felt about the propriety of separating the i.)mo~pbte o~'ossvei~z (icy) found, in Slough 91 from the other crossvei~e~ses. The posterior crossvein is absent or incomplete, ~ small length remMning at its junction with the 5th longitudinal vein. The wings are broader and shorter than the wild-type wing, and in some crosses a darkening and coarsening of the veins has been observed. The character segregated in f ive c u l t u r e s o u t o f t w e l v e :

+ i~momf£~t~ crossve;rz 149 28 221 48 132 ,30

84: 25 98 I9

It is now 100 % penetr~nt in stoc]~, bui~ o11 outcrossing and inbreeding the form is recovered in proportions simi]a.r to those of the F2 cultures. Very occasionally heterozygotes show a shortening of the erossvein, one ha~dng been observed in 3.51 flies, ha.If of which were hetsrozygons for tile mutated locus. This comparative constancy of penet.ranoe and the alteration in the wing shape are not observed in any other c~'ossve.#dess stock.

In one culture of' She ilve examined in New Forest 17, a forum with interrupted veins was observed. The longitudinal veins had gaps in them.. No cultures to confirm this character were examined, and whether it is a recessive is ~hersfore ~mknown.

In the segregation + ~7.gen'~pted ~e'4~ 116 14

the abnormMs are deficient but phenogypieally i~ resembles the rank I

56 A'ncUy,s'is of Th,ree Wi ld Pop.~dat,io'~tb' of I). , subobsenr~

recessive i.~sterr~p~ec~ vei~s (Ohrisbie~ 1939) and was unlike any other vein abnormality found.

(4) B.r,ia'tZe form and price,nee When the almormal bristle forms are considered, ag'Mn only a few o:f"

them can be considered to be normal reeessives. Though comparatively little wm'ked with, fo'deed~a (fra) from ~ew

Forest 13 is thought to be oar of these. The thoracic bristles are ~tubby and gnarled. They are often forked at the tips and frequently scythe- shaped, The tips become broken with age.

Iu shaven~o (sh.~o), 8mdland 60, any of dm notopleural, presu~llral, supralar, postalar, dorsocentral and seutellar bristles may be absent, The brisdss, if they are present, are reduced in size and may be double. The expression varies considerably and is most pronounced in old flies, but scoring is always easy. As this is thought to be a norm.aI recessive, the low penetrance with which it appeared at first must have been dtte to a linked lethal.

Ud.n (Us} from New Fores~ 3 has been worked with exeensively, being considered to l~e a normal recessive ~nd used in making chromosome m~ps. Recently, however; i~ has been found that in some environments it shows a marked dominance. Phenotypically the bristles are shorter than wild type, shapeless, straight, wiry in appearance, and. thinner at the base. The tips may be carved or bent. When the character first appeared the bristles were twisted and deformed, bah this form has never reappeared though the character has been introduced into a great many genetical environments.

Another semidominant is brok'en b.r.istZe (br), from New Forest 11. The thoracic bristles are thin, s~.apeless and wirelike. They may be bear or broken off at any poin.g. The number broken and the shortness of the remaining stumps itmreases with the age of the fly, showing that the bristles are brittle. In very old .flies the abdominal bristles are absen% the costa, appears shaggy, and only a small stump remait~s of the arista. The expression of the character therefore appears variable, but is probably nat so. Tlqe two culttzres th4 segregated in the ~'z were:

+ hrolcen b'istIe

107 12, 57 26

i.e. the abnormal flies are significa~dy deficient in o~7,e and in. excess ia the other. As malay flies showing the character appear ia an E~. of' this mutant, a].l abnormal flies in the firs~ eulbure and some of those in the second m.ight }lave been heterozygotes.

CECIL ~O~DON, HELEN SPU!%WAY 3_ND P. A. [[~. STI%EET 57

in ,redz~ced brfstIe i¥om S]ough E some of the main thoracic bristles are thi~ and short. The smaller bristles are freqttent]y missing. The higher fhe temperature the more p.ronounced the expression, Assuming ~]].st no lethals segregated in the IQ. cn]ture, ¢(~ ~/o of the flies assumed to be homozygous for the mutated gene showed the character. After mass selection for 12 generations the character appears to be 70 °/~ peeetrant in a similar environment to tha.t in which it was found, i.e. it is also influenced by the genetic enviromuent.

for#e@ (f~) and for£eda,~ (f~,~), obtained from Slough D and a([ respeotively, though resemblit~g each other very closely phenotypically, ~re due to mutations at diff'erez~t loci. The brist]es are bent acutely near th.e body. When the life cycle is reduced to a month every fly has these abnormal bristles, but as tee temperature drops and development is prolonged, more and more flies become normal. As the manifesta.tion of each character is dependent on the external snvirmmaent, it may also be influenced by the genetic milieu. The appare~at low penetranees in the -F z se~egations therefore may be due to suppression by modifiers, fob'Ice@, however, in the few crosses made with it, appears a normal recessive, if development takes place at the optimum temperature for the manifesta- lion of the character.

In the -/~ of Stndland 23 a fomn called. "forI~ed" segregated in four out of twelve cultures. When extraction was attempted two forms, thi~z~3 and tlzidc, were obtained separa.tely. For the estimation of frequency only one form has been assumed tic be present. In thh~aa the thoracic bristles were both thinner and shorter than normally. In stock the character was 100 % penetrant. Thic# w~s a marked shortening and thickening of the thoracic bristles so that they became pointed spines. Thldc x gMdc. rantings never produced offspring. The character was kept for some time by continually backcrossing heterozygotes of both sexes. ~ile this stock was being extracted some apparent intersexes were observed, but were not found again. Both stocks have now been lost.

The bristles when present in the st~eqz 7 (shT) from Studlalad 7 are reduced in size bnt are never double. This character is not thought to be influenced by modifiers. It is r~ot allelomorphic with sh~ve~6o.

In the ~]~eve%~ (sl~,s~) from Studland 52 the bristles when present axe normal in size and shape and the expression is very variable. ~reqaently there is an apparent overlap with wild type when only i]aconspicuous hrist.les a.re affected. The character is ~hought to be 100 ~o penetra.nt in stock, but this is due to modifiers collected by selection and is lost on oatcrossing.

58 A,Bagy,~.is of ~Fhree I~Fild Pcq),zdatio',s,s' o f D. a u b o b s c u r a

A slzave~,, was repo:rl;ed from Slough C which appeared in two oul, uf tline cultures. The counts of these cultures were:

179 13 12~ 1

iz~ the Fa one in seventy-five flies thought to be homozygous :for sh.c~ve.?~. showed the character.

The ghi'~, bristle~ fl'om New Forese 2 segregated in one culture: + thin brislte

89 3

and a sing!e fly caliecl]b.d.:ed appeared in a culture of ni~lety-six. Eleven cut~ures were examiaed. The main thoracic bristles were thimaer thee wild type. The mutant was confirmed but no stock was set up.

Two characters were found that were so inviable tha t they could not be extracted. The smal~ numbers in which they were :first observed is probably due to this reduced viability. They were ald,)~ bristle A from Slough A, where the thinness of the thoracic bristles made them curl, and wis~)y from Studland 43. h~ this the large thoracic bristles had lost tlleir bristle. . character a:td had l~ecome ~hm' and straggly at the tips. The wines did not appear to expand properly and remained opaque and slightly crumpled as though damp. The veins were thin and did not appear as dark as usual. The morphological similarity between two forms with ~imiiar weak viabilities may not be a coincidence.

(5) Mor?ho~oyy of cq)pendages a~zd abdomen O:f the eb.aracters tha t cause a].teragions iu tire anatomy of the fly

only one, eyeless (ey), f,'om New Forest 20, is sufficiently penetran~ to be serviceable for linkage tests. Even .so it is dffticult to score at~d. the penetrance, wlaich is about 90 ~/o, is thought to be subjective. The area of the surface of the eye is :reduced. The character is very variable, the extremes being normal, and. completely eyeless. The actual shape of the eye is not a:[Ceeeed, a~ad the size is reduced, considerably before ~,]ae facets are disarranged.,

I)aI,~gess, S~ough 18. Oi~e or both of ~he pulps were missing or occasiou~ ally mint~te. Seven F~ cultures were examined and the character was observed, in %we of them. The connts were:

+ pall)less 54 t 53 4

and Lbe pen.ctl;a.R~ec did act seem go he influenced by modifyi.ng genes. baton, S~adland 32. Tiffs form resembled bato'J~ i.n~ D. ,mdc~wcTaster

CECIL GOIRDON, [-~EL[P.N SPUiR1,VAY .AND P . A . ~o,. S T R E E T 59

(Gordon, 1936), i.e. elongated body, genitalia puslied backwards, and ~he ~anal aperture closed[, producing female sterility. Pupae showing the characteristic abdominsl elongation were observed in cuRares from the wiId-~ype sibs of the ba~o~ flies. The flies did ~not emerge.

(0) D~oa,g One dwarf fly was found in Studland 33. Not only were the flies

small, but the eyes were small J n proportion. The ]hairs between the ommatidia were wea)ly and reduead, and blisters were sometin~es observed in the wing. The form was a semiletha], being very iaviable end completely sterile.

(7) Abdomi,ncJ, 2i9.me~tation

abdo~ff~,d sFol, (as), Studland 52 (Plate II e). There are small areas of lacquerdike black pigment on the ventrsl plates of the abdomen, which appears freckled. The hairs on a region so pigmented are hypertrophied. The freckles, which may overlap on to the surrounding membrane, are never larger than half the plate, and frequently are so small that they are only recognized by a giant hair. This character appeared in the P~ with a penetra.nce of 70 }~. In subsequent generations the penetrance has been erratic and frequently drops to below 10 °,/o.

tn blacfc 5eZ~y (bit), New Forest 26 (Plate I I f ) , the edges of the ventral plates of the abdomen are blackish, and in some flies the whole plate is pale grey. The chitin itself appears to contain the pigmen% though small painted dots such as appear in ebrlomi~zcd spot are occasionally seen. As the ventraI pigmentation is slight, the character cannot be recognized in young files before i t has deveIoped, or in old flies where the folded empty abdomen throws too much sh.adow for the shading of the sterna to be seen.

Four F_~ cultures segregated: + 5~aelc beZly

t89 10 121 6 74 2

112 28

The difference in penetrance can be explained by different genetic environments. There is no suspicion of a linked lethal as the penetrance was only improved very slowly by selection.

In all further discussions the forms described in this section will be referred to as "autosomal recessives'. With a few doubtful exceptions

the 3" were recessive in the /v 2 segregations. The inverted eomma.s are ~sed

60 A%c@#~ of Tl~'re~ -W.i~d PolJ~&~,tio'~~s of D. ,~ubobsour~

because several have been shown to be semidomi~ants. "-ks these are among those that were used for ][nhge at~d t.hus put into numermw genetic environments solely because I;heh' beh~viour was apparently conventional, it is probable that dominance would also have been. ob- served among the poorly and m'ra, ticMly peaetrant characters if ~hey had been worked wi~sh as extensively.

~. F_<LSEI MUTANTS

A character was judged to be non-hereditary when:

(c~) ~he progeny of abnormal fifes were all wild type, or (b) ~.he inbreeding of She progeny of abnormalx wild-type sit did

no~ produce a segregation of the character concerned, or (c) the progeny of wild-type sib x wild-type sib did not stow a

segregation. Sometimes a N.rther generation from this wild-type progeny was bred.

All tes~.s were made by mass matings if possible. Wing vein abnormalities, by whatever test they were unconfirmed,

will be considered in She .section devoted to those phenotypic effects (p. 6.31.

Pheno~ypes which did not reappear itl tests (b) and (c) have been ignored. Nest of these were observed in single flies, which were probably non-genetic monstrosities due to accidents in development. The poor penetrance of some "autosomal recessfves", however, shows that their confirmation was largely a matter of luck, and it is probable that a few genetic abnormalities are Unrecorded.

With the exception of the very frequently recurring failure of th.e wings to expand, all the abnormM.Ries which were unconfirmed by test @) are here described.

2'lie most conspicuous of these, both m.orpbotogieatly and numeric- ally, was Udcl~ ~u'gstc~ from the Fs of Studland 14. The arista and its branches i.nstead of beiug thiu and feathery were thick bIaok bristles. The form segregated in two out of eight cultw:es:

49 7 56 19

2'c~,~zbT/s were recorded from Slough 10 and. Slough 15. They resembled very closely ~he'for.m of this ha, me foun.d in D. ~,elc.~%oGa.s~e.r. In Slough 10, eleven f~ cultures were examined and ~wo segregated:

193 3,i~ 120 9

CECYL GO~DON, I-IELE~ SI~UI%WAY AND P. A. ~. STI%EET 61

In Slough 15 the segrega%ing c~ffture was:

+ jau~y 88 I0

Seven _//'~ cuItm'es were examined. In ~he Fs of two New Forest families one otflture showed forms cMlsd

short bristle. The ssgrega~:ions were:

F~mily + short bristle 39 67 I9 43 81 37

In the second famiIy where She abnormals were in excess of the expected quarter, some of the flies were shaven and some had forked bristles. Two mat, antS m~y have segregated in this family and the l~a culture examined was:-W:tl.cl.~ype bedat~se aR the flies were het-erozygous for both of them.

Similar/dzor~, hri~tlessegrega~sed in the P~ cultures of two Studlatzd fg.nSilies,}an!l Wg~e"not inher!ted though severM supposedly pure cultures were:exgmin6d:fo£~hem;. The first of these has been mentioned before:

]~amily + s/~ort ~ristle 55 ii 19 58 21 8

The 2~ of Stucltand 16 segregated for a non-hereditary form in which the tips of:the bristles were not broken, but sharply bent:

+ bent bristle 95 15

7. MA LFOr MeD AI3DO3fElV

The Fj of 8tudland 5~ consisted of ten males and eleven fema,les a.lI wild type in appearance.

P, ight e±tures were examined in the Fs and the character segregated in atI of them-

+ ~/~Jormed ~bdsmsn 29 43 53 26 28 30 7~ 20 28 20 37 30 7 19 4 15

~lies sh6wimg the character ,seem comp]eSely fertile.

The pet~etranee is quite unprsdieta.ble. After about twelve genera.- ~ions of mass selection for the cha,ra,cter when the pene.trance made this

69.2 Analysis of Three Wild I°olovJgl, iogas of D. subobscm'a

]?ossible, m a l e s s h o w i n g t h e c h a r a c t e r were o u t e r o s s e d to f e m a l e s f r o m

f o u r m a r k e r s tocks , sca.rl, et, rouS~, po2p.y, a n d cv, rl~ed. Th.e F z c o u n t s w e r e .

'~ 3fagdbrmed abdor~e~z By scarlet 56 - -

87. 1 :By rough 50 - - -

35 o ~

a I

By poppy 35 i By curteg 77 - -

27 21 95 I.

The p a r e n t s of t h e )% w e r e p h e a o t y p i c a l l . y n o r m a l flies. T h e / e c o u n t s a r e

i r m o m p ~ e t e b u t t h e flies e m e r g e d as fo l lows :

+ Maltbrm~g + + ma.rkm' + abdomen Double recessive

(~c~riet) 58 17 - - - -

82 26 - - - - 31 15 1 - -

3 6 9 - - - -

3 3 5 -- - -

28 6 - - - - 1S 7 - - - -

(2)<vpy) 49 18 - - - - 52 I2 - - - - 28 ii 2 - - 25 I:~ 1 -- 15 2 - - --- 41 I0 - - --

(curled) 64 19 5 1 43 3-1: I --

7 4 41 15 49 6 - - - -

7 4 57 17 - - - - 1 7 6

Further work is necessary before these results aud the ratios in the

i.nbreedi:ig F~ caa be explMne&

t n y o u n g flies t h e a b d o m s u a p p e a r s c r a s h e d ] a t e r M l y a n d o f t e n

t w i s t e d . T h e tors io , ] M:fects tlae exterua.1 g e n i t M i a . T h e v e n t r a l p laSes a~re

v e r y w e a k ] y p i g m e n . t e d a a d s e e m p r e s s e d o u t of s h a p e . T h e d o r s a l p[a~tes

a r e v e r y ~nisslxapen a n d i n c o p a p l e t e , a n d t h e p i g m e n t a g a i n u n e v e n . T h e

a b d o m i n a l b r i s t l e s m a y b e r e m o v e d , p r o d u c i = g 90. e f fec t sh i l l ! a t t o t h a t of

e x t r e m e bobbed in D. m&~,~wge~a~er. T h e r e is f i r equea t l .y / a b l o o m o n flies

whiel t , as t l l ey ~ row o l d e r azad b e c o m e d a r k e r , g ives t.hem, a c o b w e b b y

a p p e a r a n c e ( P l a t s I I c~ a.n.d o).

CECIL GORDON, HELEN r SI~UICWAY A~D F . A. ~ . ST}ZEET 63

~. VEIN ABIqORMALITIES~ THI~ INI~IEi%ITANCE OF

w~Ic~ ~s U~a~OIO¢,X~LE

Description of l)h.enoty2es

With the exception of the three previously described, as autosoma.I reeessives, the segregations of vein gbnorma.Iities in ~he three samples are given in Table V, which also contains some relevant totals. The three _~ se~'egations are given in Table VI and the av~ segregations in Tables VII

IX.

TABLE ¥

Vei~ ab~or~naEtiss, the inheritc~nee of which is unpredictable. Totals of al~ three pol~ulatio~s

Slough

~X~o. of wikt ~ examined 47 No. of f z segregations 1 No. of fli~s examined in iv 1 1390 No. of fli~s ~bnormal in 2~ 7 ~o. . Of ~ s e ~ e g ~ i o n s 37 No: of wild ~ showing 0 F2 segregations 15 :No...of M i d 9 showing I f~ s e ~ e g ~ i o ~ s 28 ~ o . of wild o showing 2 E~ segregg~io]~ 3 No. of wiId O showing 3 F . segregations l ~o: ofE~ s e ~ s g ~ i o n s of crsssvsi~less 15 ),%. of 2~ segregations of folexus, twig ~nd sT~ot 8 No. of' ~W= se~ega~fons of Jzort 9~ 5 No. of ~ segregations of ,short 4 3 No. 6f F~ segregations of short vein 3 No. of segregations reco=mnized by 1 fly 10 To~al ,-m. of ~ultures h~ segregating F . 322 No. of segregating c~m 'es 85 ~o . of flies examined iv. all Zc cul~tu'es 23440 iN-o. of ~bnormal flies in ~v~ 568

New Fores~ S~udland 42 55

0 2 838 I567

0 S 27 34 23 29 12 19

6 6 1 I II i~ II 9 2 8 3 3 0 0 9 8

2~4 300 65 87

21899 27508 324 317

These abnormalities were of five phenotn~es. In those called m~oss - vei~ess the posterior orossvein is either broken or absent, The vein may be broken in the centre ox at its ]unctions with tile 4th and 5th longL tudina.1 veins. Occasionally the an%erior orossvein is missing or incom- plete. The way in which the Cl'OSsvein is broken is ustmlly characteristic of the stock.

The form ddtct erossvei~ varies in spec.ifieat~on considerably. There may be a slight thickening of the posterior crossvein a¢ its junction with the 4th longibudinal vein, and sn~a,ll branches of vein tissue may radiate from it. Or the posterior crossvein may branch into two at some ]point in its length, and both branches join the 4th longitudinal, enclosing a small triangular wing cell. One of the ])ranches may be ])rolcen or absent, and one or both of t h e m m a y carry lil:de secondary branches. De~c~ crossvefn

64 A)t.al, y~.fs of Th,ree Wi[d Pol)u/,crgo~.~" of D. s u b o b s c u r a

is assumed to be a modiKcation of crosavei~zbss even though some forms listed as a~fy were called dsZm when tSey segregated in the F~.

Three stocks of deltc~ were exl:.raeted and seleeted uutil the pen.strauss was 100 %. in tits/;'s of Studland 40, only one fly showed the degta form, all the other abn.omnals being crossvei.~less, but this like tke croesveSgess from Stu.dland 49 was extracted as ddtc~. In St~dland 47 the segregating aE_, cultures showed only a c~dte c,rossvei~, and a stock was obtained of it.

This stock was outcrosssc[ to the second chromosome marker s'caget and then inbred. The Fx males were also baekcrossed to seerlet (see p. 75 for description of linkage cxperhuents). The two ~ cultures were:

÷ c.ros~ve~?~le6"~

96 0 67 I

In the baekoross tke mutant appeared in three out of the six cultures examined:

+ + ~cc~rle~ + + m'os~v~i~tl~s~ scafle~-aros~ve.i'n~es~

30 40 i --

36 64 -- i

26 2 8 -- 1

Thus in all the cultures whets the mntant would be acting as a dominant if it showed, the only form tha t might be at tr ibuted to itwas c,o~'svs4nbss.

Eleven cultures were examined in the F~ : + ero.ssveir~- scr*'rlet acm'le~

+ + scc~'Z¢$ + + delta Zes6 deZtc~ c,'osavei~zleaa

72 17 7 - - 2 104: 22 5 - - ~ - -

70 17 3 1 - - - - 48 1,5 5 -- i --

60 2.8 29 - - i --

35 10 1 I I 17 12 3 - - -- --

79 24 7 -- 4 --

66 26 3 5 - - 2 71 1S 1.4 -- 3 --

63 20 4: -- -- --

If the c.rossvs~dess flies are here hetsrozygotes there exists some detain.ames combined with a very ]ow homozygots penetranee.

The forms called, l)Ze~al.s, twig or el)or are a[l various expressions of a similar phenoOype. In spot a small crumb of vein tissue sppears hJ the submarginal cell, Sometimes this is reduced to being a small ]mot or smudge in. the lamina of the wing. tn a0ig this piece of vein [issue is joined to tlle 2ad longi/,u.dinal vein at or near Its junction wit]s the costal. Th~s extra branch is fr%tuent, J.y forked, one secondary branch rutming towards the 3rd losgitudinal vein and. one forward paralM to the ind. In this form, small br~nches of the posterior crossveia freqL~endy appear. In

CEC]-L GOI~DON, H]~LEN SI~][TIR~VA-Y AND F. A. 1R. STREET 65

4she p~sz~ forms, the junction of the 2nd longRudinal and the costal eins is eo~pletely surrounded by a, plexus of vein tissue resembling

/ s x ~ s in D. ~zdc~ogc~s~e'r. In Q~ese forms the posterior crossvein joins the 5th I01agitudinal very near ~he margin of the wing and frequently in a shadowy delta. In the most extreme form., that from the/Q of Studland 28, the 2nd and 3rd longitudinal veins are nearer together than usual and are either joined by ext~'~ crossveins towards l~heir distal ends or fuse in a plexus. All the veins may have the "plexus" form and be slightly branched. The costal vein between the junctions of the 2nd and the 8rd longitudinal veins is grossly thickened and malformed. In most specP mens extra vein tissue is found in ~he,marginal cell, ths narrow sub- marginal cell, the di.scal cell and in the 2nd posterior cell, especially in ~he region of the posterior erossvein which is frequently unrecognizable. It never appears in the ls~ posterior eeII. All extra vein tissue in these forms is of the "plexus" type.

All stocks show at least two of these forms, bug one is usually charac- teristic of them. The amount of '~Plexus'' tissue is generallyless in a hetero~ zygote than in a homozygote, but the expression varies so much ~hat Q~e genogype of indi~ddual flies cannot be diagnosed from their phenotypes.

In short 2 the 2nd longitudinal vein fails ~o reach and join the costal. The actual length missing varies considerably; half the vein may be absent or there may be merely a defective junction. Sf~.rs sf~ort 2 from the 2' 2 of Studlaad 38 had wings broader and less pointed than those of wild-type files. A similar form was also extracted from the/~s of Slough i0, which gradually disappeared in cultures though it was continually seleeted.

In shor~ 4L the iSh longitudinal vein is incomplete, the amount missing varying from about one-eighth of the distance between the margin and the posterior erossvein, toalmost the whole of R.

Several veins are incomplete in the forms called sJ~oft vei~. In that from Slough D parts of the posterior crossvein and the 4:~h and ~th longRudinals are missing. In those from Slough 1 and Slo~gh 5, 2 and 4 are defective, and. in the latter 5 is.affected as well. The two flies in the first culture of New Fores~ 9 ero~ssei~les,~ had the 2nd longitudinal~zein incomplete but no subsequent flies showed this character.

In oNy one stock in these sam.ples is the 3rd. longitudinal vein in- complete. After ten generations of mass selection, flies in whicla B and 6 furled to i'eaeh the margin were found i~ the stock of Studland 4.3 s/~.o~~ 2.

@ultures have been made of these forms which were continually selected for penetrance. In some of ~hem R improved comparatively

Journ. of Ge~etics xxxw~r

66 Ant@sis of Three Wil.d Popzd.atio'~s of D. s u b o b s c u r a

rapidly and is now 100 ~o, while in others it still remains less than 10 ~/o. The penetranee after twelve generations ofmass selection is fairly cm~. s~a.nt in. any partie,:Llar stock. The penetrance of femM.cs has been found ~o be significantly in excess of ghag of tomes in the majori ty of stocks examined. This is probably because more females are tlecessary f.or pug~ t in t up cultures than males~ and stocks were only kept of those tha t could be selected wRhoug difficulty.

The penetrance of two crossveinZesses, Slough 36 slid New Forest 9, varies considerably with the rata of development. The shorter the life cycle the greater the penetrance. This variation in penetrance, influenced by b o t h t h e external environment and the genetic milieu, shows that these wing-vein abnormalities rescuable those in D. f~zebris studied by Timof6eff~i%essovsky (1927 b, 19.34).

.The $'~ se2regat.io.ns

Three F~ cuRures segregated for malformations of the re,nation. Crosses were m ads with the abnormal flies, a.n.d/72 marinas were set up with their wild-type sibs. The complete counts are given in Table VI. All three characters were shown to be inherRed,.an.d two of them re- appeared in the -~s cultures. There is no evidence that stay of them were sex linked. The preponderance of females observed in the .F~ counts of SIough 5 crossveingess has been reobserved subsequently.

As the phenotypically wild-type parents of the F~ cultures same from a segregation, they are not comparable genotypieal.ly with. flies from an entirely wild-type culture. Therefore these reappearances cannot be considered to be/v~ segregations and are not included in the totals.

Some of t,~e F~ cmmts were so small that they cannot be considered to have been examAned for segregations, but the question of how many flies must be. examined in a oultm'e before it can be assumed to be non- segregating will be considered iza t[he 1next section.

Th~ ica segregetions

AlI abnormalities recorded in tlts Pc have been con.sidered gerletic whether they were eonf~rmed or not, as it now seems [ikety tha t some phenotypio effects, though produced by abt~orma] genotypes, may nee reappear in an F:~ o r / ~ .

Examination of tables VI I - IN shows that on twenty-seven occasions the presence of a mutated locus has 1)sen inferred by the appearance of a single defective fly. Several abnormaIit~es recognized by sash evidence have been confirmed and. extraceed. Even when more than one fly that

C+r+C~ GoP~Do~'% } I ~ L E 5 ~ ~S-PT_.,-P-,,WAY AXD P. A.

+,c,.~.

=j

s+ o

74 ~ ,-o i "e

©

x x m+ ( # c..,'3. ~ .~..~

~ + ~

,-.#

,_j I

~ 4 + + + +-:~ I x

f . -~-"+- ..,+ c + ~ + ~ I

, . , , + ~r 1

× ~, 'r'- o ( ' C d - ~ :3 'D • t - - ' ~ c'~

+ +

~#+- -+ o

1 .~ ~ I +.++ . o ,

e.O

r~

B

+..,.) - %.

,-.,3

"'S

C.]

,~, ~ +'0*

>+ X X vXDei".+ O3 ~ I . "

,'d- ~ +.~..+"~ .~c- ' : t

" ~ +"d++

O "-< O

X X X +J ~ ¢,tt SO 2 + ~ &,'JJ-+

L~ ¢','t ,30 ,..-4

%++-+,.,~0 +

R,. STZ.}::J:'+T 6+

7+ ?d: ~ "~

2

0i,

b.J0

c,

e+

R

k

+0

.g,+

#

,,B

:= & o

.~4

~ o 5 - 2

.v+

- . - a

~2 . g

P -r,a. ,~

m

E~

>

c~

2 ~

68 A%c,Z~s'iu o/77++'+e ++;.+;Z++ Po,p,+tlat,+o?+~ o / D . subobsem'+

~ l i ~ I I I ~ I

. p

II I l l ] i I~ ' I '~ I I I +~+-+ l i I ~

o

o

<,

~?°'~I ~ ' -~ °' I !: . . . . . . I ++~

g

I I ! [ t l l I l l 1 ° t i I

I l l l i l I l l i .._+

I ( +.-+

I I I I I ] I g i J=l~"

t--- . - - :

I ! +" c-I L--

I ~ I J.r= ~ I~,++'I+" I ++'+

c : .

T i l l

+

4

; 2 ,

o

g

~ 4 ~d

2

4 +~,r_ + " m . +

+

2~

i

~ o ~

+ ~ . . . . t ~ , . - . . K . . . . . . . . . - ,-~ ,+..,

,.-:++ ++

. . . . . . . . . . . . . . . . . . . ++

m +

o

~ . k 8 a ~ 4

~ ~ ~.

o

-~ ] I I I [ f I 1 I I~o . r

• ~ . . .~ ~ ? c-~ ~" "~ ~

~5 i-- ~ G-~ . . a ~ ~ , .~ . ~

~ . ~ ~ ~o ~ ~ ~ °£ r~ . . . . . ~ ~ ~ 7~ ' ~ ~ ~ E

70 A;Dc~lymCm of ClVu~'~e WTLcZ Po,p. a(,Zu~,Zo~,m of D. sl:~bobsou~:'~

]L l l L i ~ l l l l l

~ I ~ I I ~ I I I I

! L i I I t, "~

I . . . . i . . . . I

-'d

ro

~D

~ ~ 1 ~ . . . . f ~ . . . . . ~ i

~ i I t ] i I I i ~c a...

_.~

y . - .

~5

0

c~ I J i I I I ] ~r

L ~

I I

i ° 1 ~ 1 I_., I I~.~1~ I ! I

J i I I o ~ i ~ 1 ~ ° 1 I I ~ . ~ ~ .~ ~ ~ ,

. ~ r~

CECIL GOI~DON, ~][t~LEN SPUI~IVAY AND 1 ) . A . ~ . oe . . . . . ~**,. ~eT, 71

showed t h e cha rac t e r was found, the n u m b e r of abno rma l flies in the segregat ing mfltores was much less t h a n a quar ter , with ~he except ion of

the cu l tu res g iven below. Thus flies homozygous for a m u t a t e d locus

f r equen t ly fai l to show the eharaet .er:

+ Abnozmat Slough D vei~desa 79 22 Slough D veinless 71 36 Slough 5 c.rossve{nle~s ¢4 20 Slough 49 t, wig ] 16 38 Slough ¢2 twig 65 23 Slough 4S ~.wig 34 13 New Fores~ 7 pl~ccus 99 44 New l¢ores~ 11 sleet 90 29 New ~orest t l spot 2,3 I0 Stud-land 43 s2oot 66 21

The excess flies in some of these t en cul tures suggest t h a t flies he tero- zygons for the m u t a t e d locus m a y show the vein abnorma l i t y . I f th is is

so, a b n o r m a l flies will occur among the offspring of

+ m u t a t e d locus - - x + +

as we1[ as a m o n g m u t a t e d locus m u t a t e d loc~s

. . . . X

+ +

mat ings , and there m a y be an excess of segregat ing cul tures over the expec ted quar te r . I n the to ta l s given in Table V ~he n u m b e r of cul tures in a f a m i l y has been counted once for eve ry charac te r {ha t was found in

t h a t f ami ly , e.g. Slough 38 eo~ta ined d f i r t y 4 h r e e cultures, s ix of which con ta ined segrega.±ioes. A v e r y s l ight excess of segregat ing cul tures is observed in each of the samples b u t is no t s t a t i s t i ca l ly s ignif icant :

:Pamfly )¢" n 2 Slough 0,33 ] 0.70-0-50 ~rew leorest 0.36 i 0.70-0.50 Studiand 2.56 1 0-20-0. I0

Some i n d i v i d u a l families however do oonta i~ a s t r i ~ n g excess over e x p e c t a t i o n (Table X).

T A B L E X

No, of G No. of IG cultures Proba, bfli%, when

Family Character olfltures- segregatfl]g expectation is 1/4 Slough 45 2~lexus 18 8 0-0047 Slough 92 crosavai,des~ l] 9 0"000]2 New ]tores~ 27 o'oae~ei*desa 5 5 0"0010 New :Foresb 3i cro*avein*e*s 11 7 0"0064 New .:Fores~ 31 twig l 1 7 0"0064 Stuclland 31 ,l, og 8 V 0"00037 Studland 37 cros~vei.n~ess I1 7 0"0064 Studland 47 A crossvei~ 8 6 0.0038

72 AnaT, ysis of Th'!ee Wild Pop.Mat~o;'~s of D. s u b o b s c u r a

In eight of the 98 segregations, dm probabilRy of observing the number of segregating cultures found is less than 7 in 1000, m: 1 in 98. Thns the number of segregating culttu:es to be expected seems to be more them a quarter.

Inspection of the actual segregations in these families show~ that though the cultures are not homogeneous it is impossible to distinguish between a

mutated locus mutated locus X ÷ +

mating, the genotypie expectation of which is mutated locus mutated locus

1~+ :2 : I + + mutated locus' and a

matiag giving

+ mutated locus - - > ( + +

1 --+ " 1 mutated locus + +

This means that the homozygo~e penetrance of dtese characters is not perceptibly greater than the heterozygote penetrance.

On the twenty-seven occasions that a character was recognized by one fly in one cu ra re all the cultures contained at Ieast ~wenty-eight flies and twelve of them contained over sixty-four files. As on seventeen of these occasions eight or more cultures were examined, it is more probable that the single fly was a homozygote and the offspring of a

matat.ed locus mutated locus X + ÷

mating, than that it was a heterozygote, offspring of a

nmtated locus + X - - + +

mt~ting, and no heterozygote by heterozygote mating was examined. (Ou the four occasions when fonr or five cultures only were examined there is less certainty whether the fl.y w~s a homozygote or a hetero- zygote.) Therefore any frequency of wing-vein characters calcul~ted from these figl~es is only an approxlm.ate minhuum.

Si~oe a mutated, locus m.tttated loces

X ÷ -F

x ~ mating, it is necessary to test mating may be mistaken for a -~ -,-

Ct~CIL GORDO~XT, ~Z][EL[I~N SPUI~.WAY AND ~. A. R. STREET 73

whether the frequency of abnormals in the F 1 sample, which is the product of the wild mating system, and the frequency in the F~ sample, which co'hsists of the offspring of ol~e generabiou of inbreeding, are significantly

different. Usi~g the figures given in Table V the X and probabilities of the Yl

and F2 belo~ging to the same statistical population are given below for all three biological populations:

x n /2

Slough ~[.532 ] 10 -~ :New [Forest a.ags 1 o.ooos¢ Studl~nd 2.227 1 0.012

In the Yz of the Slough population abnormal flies are nearly 6 times as common as in the 2~ but the frequency is only increased 2.} times by inbreediug in Studland. In New ~'orest as no abnormals were found in the _F~ no estimation of increase in frequency can be made.

That the frequency is increased shows that recessive genes are in- volved. That it should be increased so 1Rile shows that the manifestation of these wing-vein abnormalities is more conveniently considered as dependent on multiple factors: which are concentrated by inbreeding, than on recessives which the brother-sister matings cause to segregate in a Jaomozygous condition.

Therefore it seems impossible to make any guess about ~h.e frequency in the populations of the genes involved in these phenotypic effects, but this must at least exceed the frequency which would[ be deduced if each was assumed due to one gone only.

Ide~t*i*.j ~es*s

Though the ratios obtained show that mauy genes are concerned in ~he manifestation of these characters, the majority of these were thought to be modifiers, and attempts were made to find if" any of the principal genes were identical.

Timofdeff-lRessovsky (193¢) h~s shown, tha t the specification of a character is dependent on the genetic milieu, and therefore ~ao identities among the crossvei~~.Zesses and plex~ses can be recognized by inspectiom

F. lies from stocks haviaag similar phenotypes were mated together in pail's on the assumption that in the F~ the penetrance modifiers would be in a heterozygous condition. I f these were recessive in effect and differed in the two stocks, the penetrance of the character examined might be less than in the parent cuRures,

Using this method, short ,veiu~ (Gordon, 1956) was found to Be identical wRh the short 2 characters from Slough 52, Studlat?d 7, 9,6 and 57, with

74 d.naI?/6'i.s q f Three lV3d Pcg).zdations of D. subobscur~b

the short 2 which was the only form extracted from the crosses set up for i G stmrt 4 in SiokGh 15, a,nd with d~o~'t vein from Slough 4. No other crosses

were made betweeE stocks with ~his p henotype. Seve~-~ pie:sis forms were tested iri this way and the resttlts are given

in Table Xl. Only one culture of each was examined. [nspection-'of this

TABLE XI

lde.nt.it~ crosses between seven stochs of plercus, twic 2 c~nd spot fo'r.m.s Sgudland S~udiand New .Forest

Sloe, six 52 Slough 8 i0 Slough 50 39 L8

N ew Fores t 55 36 0 47 30 9 56 63 3 39 20 Slough 52 18 5S a l 18 - - --

Slough 8 - - - - 58 7 3 27 16 35 St ,udGnd 10 26 53 18 47

table shows among other absurdities that New Forest 55 is identioM with Slough 8, Slough 8 is idendcM with Slough 52, but Slough 52 is not identical with New Forest 55, Unless we consider the ratio 5 8 + 7 pPe~cus, ill the cross Slough 8 by Studland 10 (extracted from the F3), indicates identity, New t crest co is identical with Slough 8 and Studiand 10, bat they are not identical with raze another. Further work with these forms has shown that New Forest 55 is completely penet.rant and re- cessive while the others show v~riabte homozygote and heterozygote penetrance.

When the s]mrt 4 stocks were tested several cultures were counted of some of them ([['able XII) . In addition to the same contradictions,

TABLE Xll

[de~,t,ity crosses between four stocks of a sIzo.rt 4th ~onyit,udi'nal vein paron~ . , . Slough 7 S[ougk 9 S~u&iand 2 .1 . S~udhmd 70

O i j a , r en t .a_, s/zort 4: ±~ 8her r ,t 4- stw.rt 4 -F 8]~ort

Slough 7* - - - - 31 13 24 22 24: 4: - - - - ( 5 2 14: 31 19 30 0

. . . . 3 21 - - - - Slough 9* 45 8 . . . . . 7 £ 1 45 2

- - - - - - 1 0 9 1 7 4

. . . . . 8 5 Sl~uc~and 21 I" 10 2[ 15 26 -- -- 7 39

27 3 4 4 -- 2 46 Sgudland 70 4:I 12 40 4 22 34 -- --

44 13 8 7 . . . . .

* F a m i l y no t inclucled in anMysis. "t" Appeared in I,\~.

hete.rogeEei~y among crosses between the same two characters added ~o the confusion. Therefore ~his method was discontinued.

CECIL GORDON, ~-IELEN 8PUIR%¥AY :~hND P. A. [[:{.. STR.EET 75

Experiments were then. conducted on %be theory that the dominance which probably all these for:ms show on occasions, could be used to discover oIt whJctl cb.romosome they were placed and it might even enable their distance from certain markers %o be determined. Non-identical charac ors could be separated f}om one another ])]; this method, but the identity of two loci could never be established~ as the linkage distances obtained fro' them would always be approximate, and. loci, mutation at which causes simi].ar phenotypic effects, are known to be near together on the chromosomes ((~riine]gerg, 1937).

The character to be exanfiued was crossed to recessive markers. The ~1 was examined for dominance and the males backcrossed to the marker. Any double recessives found in the progeny of this cross were thought to show conclusively the independence of the loci concerned.

As dominance may be negligible so tha t no data could be obtained from ~he backcross, the f l flies were ~aated together and the 27.,~ flies examined. Though all four classes were examined so that some idea of the dominance of the vein abnormality coNd be known, only those in which it showed were necessary for testing linkage. If the two loci were inde- penden% a quarter of the flies that showed tlle vein character would show the marker. I f the two loci were linked, the only flies that could ~e double recessives would be crossovers and they would be heterozygous for the vein abnormality. Therefore except when the domina.nce and the crossing over were both very large at the same time, wNeh would be rarely, this would be conspicuously less than a quarter of the abnormal flies. And if dominance of the wing-vein character were dependent on recessive modL f~.ng genes in its own stock, though it would not show in the backcross, the segregation of the mod~ers would prevent it being deceptively large in the i~'~_.

i f the ratio in which the double recessives were obtained was not obviously interpretable, they could be outcrossed and again inbred. When the loci were independent the ~Y~ results were expected to be re- peated in the _F 4 . ~ e n the two loci were liM~ed the Ea double reeessives would be heterozygous for the wing vein locus and there would be four kinds o f / ~ enltm% appearing in equal numbers:

(1) 1/~ showing no flies with the abnormal vein. (2) I/4 showitlg the wing vein and double recessive in equal numbers. (3) t/4 showing the wing vein alone slightly more commonly than the

double recessive. (4) 1/,t showg_lg the double recessive greatly in excess of the single

recessive.

76 A~7,cd~,_~is of Tt~'ree l.l:'"i~d IFo2.~.t~g.io,~s of -1). s u b o b s e u r a

In (I), (2) and (:3) eke flies showing the vein ab~orlaality would be hegerozygotes, and in (6) homozygotes would, appear. The vein segrega- lion woaId, be the same as thee expected in the inbreeding J~',, and t?e di~gculty of interpretation encountered there might recur.

When this work was done only three of the £ve augosoi-aes arried markers which did not obscure wing venation: scc~r~et (s) (Gordon, 19;6), po2'~y (,22p) and ,~o~9]~ (r) from Studtand 11 (see p. 79). iEo,~tyh was dis- covered to be dominant during ttiese experiments. Ke~erozygoas rov4/h [ties are ~hought to be rare, but as they are indistinguishable phenotypie- ally from homozygous flies, erie of them by appearing in a ba.ckeross could have caused a locus linked to 'ro~g]~ t.o be considered indepe~iden$ of it. The chromosomes have througlloab been designated by the names of the markers.

The Fa crosses were mass rantings but the F 2 and backeross cultures were made in pairs. Six or more of these F., and backcross cultures were examined until finislaed, or uetil a double recessive fly was observed in the backeross when the counts were discontinued unless sonic anomaly had been observed.

The results obtained are summarized, in Table XI I I . All the 2~ cultures counted have been added together, but only tile backcross cultures sl~owing the wing-vein abnormality have been considered.

These figures have been added together without the application of homogeneity tests. I-Iomogeneity among the individual cNtures of a cross was not expected or required in the experiment and a cursory inspection showed that. in most of the data it was not found. Similarly, the presence of an unknown number of modifiers prevents any expect.a- lion regarding the segregation ratios. Therefore tests of deviation from expectations are not pertinent. I f the difference between a linkage and an independent segregat, ion ratio was not glaring, they were ~zot tlloug}tt to be distingt~ishable by statistical metllods in these experiments. The dedat iou in tb.e segregation of the marker to wild type in tllc individual cultures and i,z the totals hav-e been ignored, with oa~e exception. They can be explained, by different vial eondRions, differential viability, different speeds of developmenl; and th_erefore times of ha%t in t , at~d oceaslonaily perhaps by lack of drginity i.n the mother. Tlaey do not aft:'ect any conclusions which it is permissible to draw from these results.

Examination of the numbers of files found in the two kinds of cross between Slough 36 crossvei)zZess and pop.?oy show clearly a d.[ffict:dey in interpretation which was unexpected, and which finally caused Ll,e experiments to be abandon.ed.

CEC~L GO~aDON, HELEN SPt~aYVA~ AND P. A. g . STR.EET 77

The backero,ss shows independence conclusively. The F~ ratios, howe*:er, eouugerfeig very closely those expected for linkage. The very

TABLE XIH ;S'w~v;na'ry of wing-vein lin/,:age data

Wi~h scrtrlet

F,, ]3 a,ekeross _ _ _ _ _ _ _ A 2 .

+shl s. sh; 4-~hl. ~, sht 4- + s + ve{~ vei~ + + e "1- v e ~ vein Nutan~

Short 4

Slough 9 Studland 2t s tndl~nd 70 New Yores~ S

Crossvein~e~s'

Sioagk 36 Sbu~h 38 New'Forest 9

~lutanl Shor~ 4,

Sbugh 9 378 Stud!and 21 524 Studland 70 594 New Forest 8 363

Cros~vein~ess

Slough 36

59 24 17 5 8 20 -- l 319 i$0 187 1] 452 422 45 3 231 Jig 61 25 55 31 3 -- 202 74 61 10 ~4 34 3 2

136 37 28 7 214 183 12 3 522 153 117 16 124 115 $ 3

With rough . . . . . .& _ _

-g. Baokorosa 2

K

+.shg r. sA~ + sht r. s]~f + + r + vein vein + + ~+ + vei?~ veiT~

I89 169 2 78 36 4 190 61 3 No short 4 235 175 --- 16 27 3 96 28 2 No aho~

Slough 38 395 111 95 3 147 120 14 I0 New Forest 9 160 37 1S 4 96 98 1 30

Wiflh ~ o p p y .......... 2 - _ _

Y.. :Ba, ckeross f 5 r 2 , h

+ sht. p.~J. sI~f + sht p p . s]g ~iuta.n~, + + ~2 + veb~ vein + + 292) + vein vein

Short 4

SIough 9 623 186 251 80 24~ 232 2 8 Sgudland 2I Not exa.mfiaed Studland 70 150 47 47 18 21 21 5 3 New Forest 8 10¢ 32 4 1 18 12 - - 2

Crosswbdes~

Slough 36 258 99 159 9 57 95) 18 2 Slough 38 Not exa, mined 7 4 2 I New Forest 9 49 II 18 I 46 43 4 2

few doable recessive flies, which o c.cur in fo ur only out of the six eul~u res, eould be explained by the dominance of the erossvsingess, which is ob- served ill the backeross, t.he F 1 a~ld the F.. Out of the twenty erossvein- less flies in the baekeross, sixteen, including ~he two dou.ble recessive< occur in o~e culture, d~oagh five were ex~;mined.

78 A~aa,~ysi,~ qf Three Kild .Po.,p~dc~tio,~as of D. sabobscuv~

Th es e res ul~s can b e explained e o mple~ely b y pc s Be] at i nga c.rossvegn le,~'s suppressor linked to poyd~y, This suppresses both the heterozygote and the homoz}-got.e, and as s o m e double ~:ecessives w e r e obtah2ed, crossi~g over e.akes place between it, and pOFlJy, or i~ is not 100 % penetra:,.t.

k similar suppressor will explain the similar rest,lts obtained2in the crosses between Studland 21 sTzort ~i and scarlet, and Slough 38 crossvein- le~s a~>d re.ugh. Though in this iaBter it is possible t;ha, t the double feces- sires observed in t h e / ~ and the baekcross were heterozygous ro.Wh..

Such ratios can only be explained, without posgulating such a .modi- fie]:, by crossing over in the male which occurs extremely rarely in D.rosophiga, and is always due to some external agel~t, the appearance, viability and fertility of such doable recessives as do appear, belying any suggestion of a rigorous differential xdabJHt.y removing most of them.

The apparently widespread presence of such suppressors made i~ impossible in the absence of baekeross da{a to distinguish, by consider8~ tion of ~he "repute!on" /?., alone, between linkage and independence. The only test was to enforces the doable reoessive8 and inbreed their progeny. The expectation., should the two visible loci be linked, has a~ready Been given. But as a linked suppressor has perverted the Fe

e segregations ~t would be expected to pervezt uh Y., ratios in a similar way. If ~he Fz double recessives were heterozygous for the suppressor, in a quarter of t h e / ~ cul.tures, very much less than a quarter of the ab~ normally veined flies would show both characters, tn the remaining cultures the 3 : 1 segregation would be observed. ~If the suppressor was homozygou8 but impenetrant in the F: double reeessives, all tt~e cultures wouId show fide reduction in the number of flies showing both characters. Neither of these segregations could be mistaken for the "coupl ing" ra~,ios expected .in a quarter of gh.e cultu.resif the two visible mutants were linked.

The Facounts of Slough 9snort4 roWh male outerossed and inbred were : + + rouph + b abort 4: 'rouf,,h ,~'ho.rt 4

({~) 31 47 } 0 12 0 1 ] 4: (b) ~6 68 10 I7 / L 0 8 (c) 30 38 9 12 0 2 0 2 (d) 38 BO 12 6 o ~ l ] (e) 38 50 11 7 0 1 0 0 (.f) 57 49 6 F.~ 2 2 o 1 (g) ] t 2.'.3 2 :3 0 ,5 0 2 (A) 2.3 26 1 7 6 i1 7 7 { i l a 1. .54 9 s ~ :~ o 1 (j) s8 98 it'., ]:,. . . . . (/:) :}9 a8 6 .5 . . . . . (/,) 25 26 9 8 . . . . . ( is) ] 1 27 2 :3 . . . . . . . . Tot~i. 443 604 99 1 [3 [0 27 9 26

CECIL GOR.DON, HELEN SPURWAY AND P. A. P~. STI~.EET 79

The figures are not large enough to be significant, but it appears that (~b) .and (b) belong to class (~), p. 75. (c)-(i) may belong to classes (3) and (g), and (j), (k), (1) and (~,) conld belong to class (1), i.e. t ha t Slough 9 shor¢,~ is linked to ~'ov, yl~. The conspicuous excess of females over males also a,ppears in some stocks of shor~ ~. I t has never been investigated. ]inspection also shows that the resemblances anticipated to the inbreed- i~g f~ segregations have bee~ obtained.

The presence of such suppressors makes suspect not only positive P.,. data, bu tany confirmatory evidence for linkage pro~dded by a baekeross, as it would prevent the appearance o:f double recessives, even tho~.gh a large number of flies show~ng the wing-vein character alone showed it to be very strongly dominant. No backeross has bee~. obtained which Js thus doubtful of Jnterpreta.tion.

Thus no positive evidence for linkage can be obtained from either ~he f~ or the backoross when the two mutants are introduced on different chromosomes. 'Similarly no two mutants can be deduced as in.dependent by consideration of "repulsion" figures alone, because as a suppressor can make independent characters appear to segregate in linkage ratios, so a lieteroz)Note intensifier could make linked mutants appear ~o behave as ~hough independent.

The baokcross bet~een New ]?crest 9 c~'os~vei~ess and ~'m~gt~ shows that such intensifiers are present. The behaviour of this intensifier is now being investigated. I t is certain tha t ~'o~yh behaved as a partial dominant in this cross.

Thus unless a double recessive was obtained in the backcross, which meatus that the wing vein must be reasonably dominant and independent of the ma}ker, the crosses would have to be carried to an F~. This is not only unwieldy, but the diNculty of distinguishing homozygote from heterozygote penetrance would still make the interpretation of results doubtful. Therefore the method was discontinued.

9. _Ro~-ess

Several segregations were observed in all three populations of ~ maI- formation of the eye surface, in which the omma~idia are not arranged in straight lines. The sides of the individual facets are irregular in length, and there may be any number from four to eight, Sides m a y be absent and sometimes the angles between the sides are so obtuse tha t the walls of the faee~s appear curved.

The segr%atio.D.s in which these appear aJ:e ~iven in Table K I g . In Studland t l , 13 and 16 the facets and hairs were disarranged over

S0 A;~cdy,~is qf Three Wi ld Pop.zdat.ion.s qf D. subobsou ra~

very large areas of l;he eye surface, and ~5ere was oftett a slight downward curving of ~he wings in the male, where the eye-st~rface effect was also more pronounced.

In Studland 13 the segregation is significantly less than I/4, ard the s~ook was poorly viable and. lost before its id.eatity wit:h the '~'/i~9]~, of Sgudland 11, which it closely rescrnSted phenotypically, could be tested. I t may have been a different mttt~nb, with a different viability, or the sa.me rough may have been linked closely to a sublethal factor.

TABLE X I V Negatdve e ~ g ~ r e u

Y~ segrega t ions Over ]3clew {- •

F a m B y ,h~raete~ ~ ~ 50 20-50 if0 A

Studl~nd 11 rough u 19, 3 . . . . 5 2 - - 13 faugh 47, 4 . . . . 2 2 - - 16 rauyh~ 7~, 16 . . . . 5 2 - -

Slough K -rou~]h . 36, 1 0 1 0 3 , 7 - - - - - - 5 l - -

93 p~exv.s.re~gh I0~, 7 105. 7 -- -- -- 9 1 - -

E rough 84, 15 78, 2 - - - - - - 5 2 - - S tnddaad 40 rough 94, 19 43, 8 64, 3 - - - - 7 - - - -

New Fores~ 33 rouqhoid 146, 43 73, 4 - - - - - - 5 l - -

Stadia.rid 23 rouqhoid 24, 6 50, 5 67, 4 121, 2 - - 5 3 - - hTew :Forest 19 ~o~.~ghoid 33, 9 67, 7 28, -i 23, 3 6, 2 I 1 1

Studla.nd 16 was proved go be identical with Studland 11 and the stock was disca.rded,

Sgudland 1t ~'ough (f") behaved as a normal recessive, and was used a.s a marker for ~he .3rd chromosome. After i t had been used for some t ime i twas noticed tha t out of 130 flies in the F~ of a cross rouyh x Studlan.d 21 short 4 one fly, a male, appearing in the fifth count had a phenotypioa.lly rough eye, indistinguishable from hom ozygous rou,qh. This male was out- crossed to normal females, and She progeny of two of them were counted :

+ rou.q~

35 46 4 1 27 29 - - - -

Therefore in some genetic milieux rough can show dominance. The ba.ekeross da~a of the cross between New ]?'crest 9 and this ro%qh

given in Table X I I I was made up of two cultures. Some domina~c.e mec]3a.n.ism sgJl.! on}y part ial ly u~derstood is ~cting i.n the second culture.

+ + tonga + + cro~svein&.~s rouyh crossvei'n~e.~s 2. . . .

d ? d 9 d ~ d .~ 17 i0 1.8 t5 - - - - I. - - 24 "L5 30 35 - - 1 27 - -

CECIL CORDON, HELEN SPUI%4VAY AND P. A. ~{. STR.EET 81

t~ is com~eoted with the X-chromosome and m~kes cross~ei%Zes~ dominm~ only in phenotypicMly ~'o~.9A flies. A good proportion of Lhe douMe r~essive flies were p~mTed heterozygous for ro¢~.gIt,, as would be expee[ed from.;$he excess of~'m~g]~ over wild type. ]% is possible th~[ the dominance of ro~J~, is caused by the same mechanism as causes the dominance of

crossvef~Iess. Slough tt. The rough-eyed flies were s~erile. Powerhfl suppression,

clominence or variation, in vJM~ilifly seemed present in the F~. Slough 93, l£ez~s ~o%9A ('lr). The roughening of the eye surface w~.s

extremely well marked, the eye i'aoets being emerged as well as dis arranged. The wings ~ere less pointed tlmn in the wild-type flies, and die veins coarsened and spla.yed. In snbseguent generations rite roughening of ~he eye surface gradually became less pronounced while a.l~ extreme /j~e.~.',%s appeared in the, wi~gs, which became narrower and straighter than those of normal flies. After outs:tossing and inbreeding either or both forms may reappear, d~e eye abnormality sometimes being so pro~ nounced that the eye shews a reduetioz~ reminiscent of o2/~l, lm, bt.ol)ed.ic~. In these forms the wings ere withered and reduced in size. This character shows a reduced viability, and is kept by heterozygo[e × homozygote madngs. If the life cycle takes more tlaa.n a month the penetranee al~d expression are bot!a impracticably sma.tl.

The forms found in Slough E and Stndland 40 were both pheno t)qpicMly well marked, but the progeny of ~ow]~ x roz,/jh matings were all p henotypically wild type, though heterozygote >: ~'o%gl~ ma.~ings gave the expee.ted I : 1 segregation. This phenomenon was o]~served repeatedly and was not influenced by ~he length of the life cycle, k p~re ~'oz~S~. × ~'ough stock, phenotypica.lly wild type, of S~udland 40 was kept ]:.y subeulturing. Af*er 6 months the stock ~va.s re-examined and[ rough-eyed flies were found in it. Since then the penn%ranGe in a pt~re stock has bee~ practically 100 %.

The forms recorded from New Foresfi 19 and 33, and Studland 23 were all fo~@.ofc~. Only a. few face~s of l,he eye were displaced. I~ New Forest I9 these were always in the corner of the eye, but i~a [he otl;er two, any region could be affected. The va¢iable nnmbers with wMch the abnorma.1 flies occurred suggest t]~e influence of various genetic milieux, but there is probably some dominance as is suggested by five cub of eight cultures segrega~,:ing in New Forest 19. The stocks showed incomplete pe~mtra.nce, but. were discarded before i~ was seen imw mue]~ this would h.nprove wit, h selection.

The segregation of these ]?henotypic effects makes it possible thai, Journ. of Genetics xxxvr~r 6

$2 A ~ , a l p i s ,Jf Th,re, e, Wild Pop~dcrtio~s of D. s u b o b s o u r a .

ro,~g]~ eyes are due [o genes belonging t.o [l!e same series as those respon- sib]e tbr the al)aormalities in. w{ng venation. The close connexion between ~he manifesb~@on of brew Forest 9 c.'rossve.f.nles,~ ~nd Studland 11 'rov/h an.d the ak.srnative pfe¢us form of Slough 93 rot~gh makes [,his hypo_jnesis very plausible. Unlike t.he vein abnormal.Rice, however, the behav our of some of [he rough.s makes them ascribable to a single main gone modified percep~ibly but ¢omparat ivdy slighdy by the genetic milieu. Studland i]. was used for a year as the marker designate for a chromosome and no anomaly was observed except those mentioned. Probably some ro~.wDs like Slough 91 crossveinIess and New Fm:est 55 pZe.~'u.s approach a com- plste recessive state which may be a limRing condition of the phenomenon observed in th.e others.

The segregation on two occasions of a 'rough, the presumed homo- zygotes of which gave entirely wild-type progeny, cannot be explained on any hypothesis at all.

]0 . ESTI~fATION" 01" TKE FIAEQUENCY OF 5~UTATED LOCI,

-~h=D '2~E (JO.%'I)?AI%I,%ON OF POPULATIO~-S

As the great majority of these mutants are only recorded once from the popula.dons, any calculation of their individual frequencies would hays very little meaning. Therefore the frequeilcy with which a haploid set of au~osomes, or gamete, contains a recessive mutant , is estimated.

Four sets of autosomes were examined in the A of each wild female. Then if the number of wild females whose progeny were inbred was N, and the number of autosomal recessives recovered from them was/k, the desired freqnsncy, called F, is R/~N. This assumes tha~ every possible recessive was recovered from the s~m.ple.

As the number of F2 cultures and the number of flies in them were not infinite, a statistical correction should be made. But when the nature of the characters, the frequency of which is being calculated, is examined, i t is seen that the doubt surrounding the interpre~,a~iou wil.1 introduce errors into any figure obtained, besides which the st~¢tis~ical corrections are negligible.

The irregrdar vein characters have been shown to be due to several genes, and therefore calmer be included. On ~he evidence of some shnilarity of behaviour sh.oatd the ro~(ghs also be excluded'~ But Studtand l 1 .ro~.~gh, though in some enviromnents R shows a startling detain.ante, in most crosses including the original F.,, behaves like a 100 ~/o pens[ran% 100 °/G recessive character. I f R is c a r t e d it is irradonM to iuehJde the great mi.mber of poorly )el.tetran% probably dominant forms, no stocks of

C E C I L GOIZDON, H ~ f L E N SI~Ulg~'~L4-Y A N D P . A , }~. STPdZET 8 3

whieh were kept, and of which the probability that. they might have been missed is extremely high. In. one form, New Forest 11 bro£e.n 5~gsttc, th~¢e is fairly convincing evidence that some of the _F~ abnormals were heter<-gygotes- Similarly the realization of how low peaetrance can be, makes questionable whether failure to confirm even an apparent "auto- soma] recessive" justifies it being ignored as non-genetic.

Finally on.fled being linked to what is apparent]y the same let]~al in the 2~ of New Forest 1, 22 and 25, makes doubtful the basic assumption that the genetic constitution of two wild flies is examined in eve.ry .F> and the number of flies tested i~ a sample is twice the number of females, the progeny of which were ,inbred.

Therefore it. has been thougtR unsuRable to give the frequencies calculated in Table XV a spurious appearance of' precision by any statis-

tical corrections.

1%. of wild 9 examined (N)

Frequency of "a.utosomal recessives ' '

Frequency of "~u%osom~] reees~ives '~ and ¢ougl~s

TABLE XV

Slough ~ew Forest ¢7 42

A! 1o ~ g~ 19 0"10 t6 0-10

2B 0-12 18 0'II

8~uclla,nd 55 ,<

R 2 20 0.09

25 0-12-

The same frequency of recessives was observed in the three samples, about 10 % of all gametes containing a mutated locus, and since with three doubtful exceptions no character was found more than once in a sample, it is impossible to characterize a population by its ]atent, genetic heterogeneity, and consequently no comparisons can be made between them.

A mutation at the same locus in different populations was only found once in %he samples, foo~2y occurring twice in the Slough population and once in the New Forest, and a less pronounced allelomorph, cinnabar, appearing once in Studland.

Considering the veii~ abnormalities, the ~v 2 segregations ate equally frequent in the New ~'orest and Studland samples, and the greater frequency observed in the Slough sample is not statistically signiflea.nt, X a= 0-673, n : 1, p is between 0-5 and 0.3. But the f~equency of ab::ormal files in the F~ from Slough is 1~[- times tha t in the New Forest sample: and twice tha t in the Studland. There is also a significant difference between the proportion of abnormals in the two ls,tter samples, X ~ = 10-189, ~z = 1, and I~ is less tha.u 0"01, In the Slough sample o.rossve, i.~des.s, and in

6-2

84 i.~,rd:!/si.~ qf Tt~rs.e l:Ti~.d Pop~&~on , s of O. s u b o b s e u r a

tl~e New Forest~, s~)o~, t~;@, a~d )~ez4~s are commoner tha~l in l;b.e other populaKons. But the e~oess is not signific'.aut compared with the sample in which the form is next most common. The respective X ~ are 0.758 ~k.d 0"718 allc[ the probabilRy of bo~h whe~x .~. = 1 is betweelt 0.5 and @J'.3.

Consequently, though it is not yet possible to trs,~xstat.e these dt:ffereut freq~l.encies into terms of geaes, there does appear to be some genetic difference between the three samples.

11. D.tsoussms

The results obiaiaed from this analysis are in agreement ve-ith those of M1 previous workers, witlz the exception of Timofdeff-gessovsky (19BYe). One generation of inbreeding shows that a large proportion of the wild flies are heterosygous for recessive mutants. The presence of a hetero- gametic sex makes the negative selection pressure on a sex-linked abnormalRy much greater than t.hat on an autosomal character. There- fore it is in accordance with expeet.a.tion that a preponderance of the recessives sho~td be sutosomal. The one sex-linked mutan t discovered, is an exception that proves the rule, the ]?henotypic effect being confined to the homogametie sex.

The autosomal segregations have been considered to 1oe of three ]duds. Tim first consists of characters that, were not reobserved in the Fs or F 4. As only hereditary variations were soug}Jt, abnormalities were not. described on their first appeara,~ee. Consequently it is impossible to distinguish now between probable phenocopies (Goldsehmidt, 19.38), single monstrosities, somatic injuries and forms with arrested develop- men% probably clue to an inimical external en<.ronment. A few forms found in the f~ and Fs, tha t were almost certainly phones.epics, havre been described as "false mutan t s" . These abnormalities are p~:obably the sam e as the non hereditary traits, or morphoses, which. :form a large part of the aberrant polymorph.ism of wild .populations of D. m, elc~,~zWcaster or fasc~sga (DubinJn st, a/, 1.937). These workers did not find such abnormat flies in gj. ouRures bred in tlhe laboratory, and therefore 4t-c.ibute them to the influence of the ez~vironment, which certainly is not cm~sta,ub in natural eonditio,as. The varying yields we obtain frm~ our cn]turc~ shove- tha t we calmer consider tha t our flies are kept i,~ standardized conditions. Therefore these abnormalities here also can be a t t r ibuted to external influences. Over 1000 wild flies were examined in this i~vestigation and no polymorphista of any kind was fol.md. Compared wi.th the .5 °/o ob- served by Dubinin and his co-workers this absence is signif.ioa.~t.

The second kind of segregatioa was {:lm.¢ of the '"autosomal feces-

C~CIL GO~DON, HELEN SPU~WA¥ A~D P. A, ]%. STREET 85

-sires", or those characters which are apparendy due to a single gene, and the homozygous penetrance of which is much larger than the hetero- zy~ous penetrance. The latter is usually zero.

The third group contains forms in the manifestation of which many genetfl~', factors seem involved. Almost all the abnormalRies in wing venation, some ab least of the eye-facet derangements, and ]perhaps some of the poorly penetrant wing text~u'es and bristle forms confirmed and immediately discarded, were of this nature. These characters seem to be Biologically distinct from the "autosomal recessives" though there are intermediate forms. Previous workers have not distinguished between the two, and therefore our frequencies are not comparable with theirs,

The "autosomaI recessives" are the only forms which the experi- mental method was designed to detect and on them only can any com- parison of populations be based. As was shown in the last section, they appear to be equally common in all the populations and there is no evidence on which to characterize and compare the variation in the three samples qualitatively. As they did not segregate in the F1 no evidence for inbreeddng has been detected. This agrees with the findings of Dubinin and his co-workers (1936) in D. meZanogaster.

Some of the flies known to be homozygous for many of these characters are phenoty]pieally indistinguishable from wild type, and therefore may have no biological disadvantage compared with it. Thus even when the character concerned is unfavoarable in the environment in which the flies are living, the mating together of two heterozygous individuals may not mean the destruction of half the mutated loci among their offspring.

Considering the wing-vein abnormalities the proportion o f abnormal flies is different in the three populations. This may mean tha t there is a qualitative or quantitative genetic difference between the samples, but it is necessary to know more about the inheritance of the genotypes which determine these effects. Only general suggestions can be made about this until we understand the interrelationship of the external environment and the genetic constitution in the abnormal developmental processes which become apparent in a small excess or deficiency of vein tissue.

Inbreeding in a sample produces a comparatively small increase in the frequency of abnormal flies. Therefore several genes must be responsible for the manifestation of many of the characters. Are these genes divided into a main gene, necessary for the production of a phenotypic effect, and modifiers, which alone cause no visible change? In the F~ generation of many families, the proportion of segregating cultures makes it likely that in some of them the abnormal flies are all heterozygotes. I t is impossible

86 Analysis of EI~ree Wild Populations of D. subobscu ra

to diagnose by inspection which these cultures are. Thus genetic environ- ments, and perhaps unknown differences in the culture conditions, must have so influenced the manifestation of the main gent that its hetaro- zygous and homozygous penetrances are very similar. The mair, gene therefore appears to play a subordinate par~ i= the, develogmentat processes which show themselves by an abnormal venation--that is to say, there is no reason to consider it as a main gene.

If we attribute the characters eo several recessive genes, all equally important in the manifestation of the character, the deiMency of ab~ normal flies is explained, but the excess of segregating cultures in many families is still more remarkable. To explain these we must assume that several different genotypes can produce the same phenotypic effect. Some of the confirmatory tests showed that this surprising hypothesis is probably correct. Since no numerical records were kept of these tests, only conspicuous deviations from expectation were noticed. Neverthe- less in several crossvein~ess extractions it was fonnd that crossveinZess × erossvsinlsss cultures yielded less crsssveinless flies than matings of crossvein~ess flies by wild-type ribs. Similar to this is the failure to confirm several forms though their F2 penetranee was good. The most conspicuous example of this is the aoi 9 from Slough 42 where a twig x twig mating yielded entirely wild-type tties in the Fa. ffudg flies have since been observed in stocks of withered from the same family.

If the same phenotype can be produced by differing genotypes in flies of the same family, the results observed in the identity crosses can be explained. The stocks tested in these crosses were descended from most of the F z flies that showed the character and had been kept only in mass cultures for not more than five generations. They were selected for pone- trance which had not reached I00 %. Homogeneity conld not be ex- pected either in the stocks or in crosses made between them.

The excess of segregating cultures both in individual sets of ~v cultures, and in the total segregating families of a sample, and the frequency with which apparent suppressors and intensifiers were recog- nized in the linkage tests, show ~hat genes influencing the wing venation must be very common in all the populations. Indeed it seems probable that there is no fly that does not contain some of them. I t is not known whether all the allelomorphs at these loci, which seem to be scattered over all the chromosomes, interact with one another, or whether one or more is dominant over the others. The frequency with which the phenotypes are produced prevents any one condition of the gent being characterized as wild type. Since there is no inbreeding among wild flies most of them Will

CECIL GOaDON, tIEL~N SPURWAY Aa,TD P. A. t~. S T ~ T 87

be heterozygous for any particular condition of one of these loci. The

effects caused by the genes balance one another and norma] wings are p~c|uced. Of coarse occasional matings will be expected to produce abnL'rmal files. Three of our F~ cultures are offspring of such random crosse< Similar flies were recorded by Dubinin and co-workers (1937) in wild samples of D. mdanogaster or fi~sciata. Inbreeding will cause many flies to be homozygous for some of these atlelomorphs. Some of the geae constellations thus. produced may cause no phenotypic deviation from wild type. Others cause the abnormalities recognized. Tl~at these should be of six kinds may mean that the normal wing development has six thresholds, either ia time or in space, that this system of genes can influence.

Vein abnormalities appeared frequently in the F s and subsequen~ generations, especially on inbreeding tither in pure lines or arm' crossing two unrelated stocks. A very high mu{ation rate must be postulated for ~hese genes unless it is assumed that effective genotypes can be produced by a reshuffling of gent constellations, which themselves do not produce a visible effect. ConverseIy, effective genotypes must be broken up when characters fail to reappear in some f~ cultures after outcrossing.

In the "false linkage" ratios a genetic re;lieu, probably a single gent, linked to the marker, is responsible for producing an ineffective, or normal envixonment, and some of oar forms, like those in D. f~nebris studied by Timofgeff-Ressovsky (I927 b, 193¢}, seemed to be caused by a main gent influenced by apparently subordinate modifiers. This kind of genotype is a special case of the muMple gent condRion, and it gaffs into the Mendelian recessive. The same venation is probably produced by all three ]finds of gene~ic constitution in different flies.

As Balkaschina & l~omasehoff (1935) have pointed out, the greater part of the variation in many species of Drosophi~c~ consists of characters of this nature: These form a background against which autosomaI recessives are comparatively rare. They have been recorded from melanogaste~ by Timofgeff-Ressovsky (1927 a), Gordon (1936) and Duhinin et al. (1937); f~nebris by Romaschoff & BaIk~schina (1931), and Timo- Neff-Ressovsky (1927 b, 193~); in phalerata, transeersa and vibrissina, Balkasehina & Romasehoff (t935); virilis, Metz et al. (1923); hydei, Spencer (1928), and probably from obscu~'a by G.ershenson (1934). They are therefore of considerable phylogenetie age.

88 Analysis of Three Wild Population, s of D. s u b o b s c u r a

12. S v ~ A ~ z

Three populations in the south of England were examined by in- breeding the offspring of witd females.

The incidence of each "kind of segregation in the three popuiations is shown in Table XVI.

TABLE XVI

No. of wild ? 'examined segregations of "fahe mutants"

~ segregations of vein abnormalities with irregular inheri~aee

-Pz segregations of "~u~osomal reeessive8" 19 _F~ segregations of "false mutants" 2 F~ seg-rega~ions of vein o, bnormaIi~ies with 87 irregular inheritance

F~ segregations of roughs 4 /~ segregations of sex-~aked recessives 1 3~ se~mregations of Malformed ~bdsmen 0

Slough New Fores~ Studland 47 42 55

0 0 3 1 0 2

16 20 2 i

27 34

2 5 0 0 0 I

Judged by their inheritance the autosomM abnormalities were of three kinds:

(a) The "autosomal reeessives" seemed due to a single gene, and had a homozygous penetrance much larger than the heterozygous penetranee. They had no speciaI morphological character.

(b) Almost all ~he vein abnormalities appeared to have a heterozygous penetrance of the same order as their homozygous penetrance. The ~'ough8 appeared intermediate between them and the "autosomal reeessives ".

(c) "False mutants" were forms which resembled (a) in the F z or F 2 but were not inherited. With the exception of two jauntier they all affected the bristles.

The "autosomal recessives" were the only forms which the experi- mental method was designed to detect. As they were absent from the F t no evidence for inbreeding was obtained.

Considering the "autosomM recessives" alone, about l0 % of the gametes, in all three populations, contained a mutated Iocns.

The vein abnormalities, ig is suggested, are due fo numerous factors which are concentrated by inbreeding, different genotypes producing the same phenotypes. The proportion of abnormal flies was greatest in the Slough population, and least in the Studland. There was no significant difference between the populations in the frequency of segregations.

We are very grateful to Prof. J. B. S. Haldane for his helpfulness throughout this investigation, to Colonel C. Diver, Dr J. H. tIaines, and

C~CIL GO.POX, I t ~ L ~ 8PV~WAV A~D P. A. P~. 8 ~ a ~ T 89

Dr O, W. ]%iohards for the i r grea~ assis tance in t r a p p i n g the wi ld flies,.

to D r U. Philip., and Dr H. Gr t tnsberg for thei r i n t e re s t and advice, to ]~[ ~ . :L D. W h i t e for t a k i n g the pho tographs , and to Mrs F . Gordon for

doing much of the technical work.

I%EFERENCES

D.~LX~Olt~A, E. I. & ~o~so~o~l~, D. D. (1935). "Genetisehe S~rnkfim" der_Droso- 2hila-2opulationen. I. 8wenigoroder Popnla~ionen yon D. phal~rata l~eig., transveraa F~ll. m,d vibrisalna Dttda.." Biol. Zh. 4, 81-106.

Caa~sT~,A. L. 1~1. (1939). "The effect of X-rays on sex in Drosophila eubobseura, and a¢~ aecom~t of some sex-linked cMract, ers." J. geneL (in the Press).

Do~zmt~sxx~ Tm & Q ~ , M. L. (1938). "Gene~ios of ha.rural populations. IL Genie variation in populations of D~vsophila 2seudoobscura inhabiting isolated moun~a.i~ ranges." Go, reties, 23, ~g344.

DU~z>TI>*, iW. P, (1928), "The influence of consecutive fertilizations on the offspring in D. melanogaster." Biol. Zh. pp. 131-59.

Po~oss~¢, E., 8 m , ~ t ~ , 8. W., S~oo~ov, B. N., F ~ Y , L. W. & T s w r ~ , M:. G. (1934). "Experimental s~ady of the eeogenotypes of JDro~ophita metano- garter. ~ar~s 1 and ~." Biol. Zh. 3, 166-216.

Dvm.~z~, N. P., H ~ , M. A., D ~ o v ~ , Z. A. & Da~c~Kov~, L. I. (1936). "Genetic ~onstitution and gene-dyn~mios of wild popul~tiol~ of Droao2hila melanogaster." Biol. Zh. 5, 939-76.

D~ncr~, N. P., ~o~sxov, D. I), KE~, ~i. A. & D~DOV~, Z. A. (1937). "Aberrant po[ymorphism in Drosophila fascists Melt. (sya. me~anogas~er ~ig . ) . " Biol. Zh. 6, 311-54.

G~zcsK]~cso..-v, S. (I934). ': iViutant genes in a wild population of DrosopMla obseura :FAR." Amer. Nat. 68, 569-71.

GOLDSC~r~O% R.. (1938). Physiological Genetics. McGraw-HLll :Book Company, Inc. Go~o~, C. (1936). "The frequency of heterozygosis in free-living populations of

.Drosophila welanoyaster and Drosophila subobscura." J. Goner. 33, 25-60. Gath¢~t~t~% K. (1937). "Gone doublets as evidence for ~djaccnt small duplications in

Drosophila." Nc~ture, Long., idO, 932. ]H~z, C. V., Moses, M. 8. & M_~so~, g . D. (t923). "Genetic studies in ~Dro~ophila

viri~ia." Publ. Carneg. tn~t.G no, 328, pp. 1-94. N~o~s~a~.~, H. (1928). "Eine ~Iethode zur Prffftmg der Lebensdaaer geno~ypisch

verschiedener Spermien bei Drosophila." Proc. 5th Int. 6'ongr. Goner. pp. 1143-7. ~O~L~SOHO~, D. D. & B ~ _ ~ s e ~ a , E. I. (193t). "Beitr~.ge zur Genetik der Droso-

phila fun&ris." Z. ir~ukt. Abstamm.- u. gererbLehre, 58, 1-80. S~;.¢Y~ovs~Y, A. I. & F~Dd~ov, V. 8. (t936). "Chlorophyll abnormalities ia white

mustard (£'inal)is alba)2' Bull a221. Bot. Set. 2, 9, 287-305. Si~n~e~l~, W. P. (1928). "Five autosomal mutants in Drosophila hydei." Genetics,

t 3 , 4a~9. Seoo~1n,~, G. ~ . (1932}. "An experiment on breeding wild pairs of Gamma.rue zhev-

reuxi at ~. high tenlperature wi~h an accoun~ of two new recessive types of red eye." J. ~Iar. biol. As~. 18, 337-54.

90 Analysis of Three Wild Populations of D. subobscura

S~[n~T~v~(~,~, A. ~. (1937). "Au~osomal lethals in wild populations of Drosophila 2seudoogscura." Biol. 2ulL Wood's Hole, 73, 542-51.

S~J~N~!¢, :P. B. (1932}. "Gene~ie, distributional and evolutionary sbudi~s of the ab- species of deer mice [Peromyscus)." Bibtiogr. genet. 9, 1-116.

T~mF~E~F~ESS0VS~-Z, H. A. & N. W. (1927a). ~'Genetische Analyse ei_uer frei. lebenden Drosophila mdanogaster-Population." Roux Arvh. Entw. Mech. Organ, 1 0 9 , 70~I09.

Tr~OF£~.F~-I~Essovs~Y, N. W. (1927 b). "Sgudies on the phenotypic manifestation of hereditary factors. I. On ~he phenotypie manifestation of ghe genovaria~io~ radius incomplegus m Drosophila funebris." Genetics, 12, 128-98.

- - (I934). "Uber den Eil~tiuss des genotypischen Milleus ~md der ~ussenbedin- gungen auf die l%eMisa.tion des Geno~yps." Nachr. Ges. }Kiss. Gb'ttingsn, 6, N.-~'. 1, Nr. 6, 53-106.

TS0HE~JV]~m:KOF]r, S. S. (1928). "Uber die genetische ]~eschaffeMaei~ wilder Popula- tion." Proc, 5th Int. Congr. Genet. pp. 1499-1500.

Wm~a~s, ~. D. (1937a). "He~erosis in red clover. Effee5 of iabreedimg in Ya and 2~ populations." Welsh J. Agric. t3 , I7~-90.

(19375). "Genetics of red clover and its bearing on practical breeding." Re2, 4th Int. Crrassland Congr. Abergslwgth, Great Britain, pp. 238-50.

E X P L A N A T I O N O F P L A T E I I

a ~nd e. Malformed abdomen o, iu the l~tter ~he wings are cut shor~. b ~nd d. Wild type ~.

e. abdc~mi~al spot o, f. black belly o.

j:~UBNAL OF BENETI08, VOL. XXXVlfl, NO8. 1 8_rid ~2

• ~i~•:•:,:~•:~ii;:~•i ;i•i•:~•~•~:~i:i i~ i'~•~••••!i~•,: •̧ : • •~

: i~ ~ ::i I~I i i:; ::i ~'i .¸¸I:I :

• . : : . i . : : : : ~ : i ~ . . :.̧ ̧ . : i . : ¸ ¸ ¸ ' •

i: : : : ~:; ~i :ii~;i~:'ii! i ~ :~i, ~': :~,: ~ ::

PLATE 11

• ..~:.....,~ . . . . ..... .: 'il ̧:~̧ • :~: :::~:;i ::i .::::i: ~.::: i:.: i i. ii :i i:.~ii:~:i:.i ii. :ii! ~::: !i:i :::~ :.i

L : . z ' . , . . • • . . , : . . i : : " : . : . . /.~ : . : ' : : : : . .

i ~~: III iii~::~:i~: iiii~ iii~ I:~~:~~ ~:~ :~iii:~ii~ ~ii~.ii~: ~ .~! :::~!~:ii.::!~i ! ~i~ ~I :!~"!i~"~:: ii~i!i~i~!~ ~i~: '~:~i!i~~~~!iii: ' • " i:ii~iiii::~i'i:i~i! ¸~,: ~.:: i: ..~ ~:i.. ::. i :::.:.i::.:.:.' ..i..i.::.::~ :!..:.i .:i .i..~i .i..::i!~i:: ;~.; i: !:i:i . i "

~ i ii~ ~i ' ~i :~ ~i~ .i ~..~.. ~..i ~ ~~. ~~. . ~ .-

.i~ ~i ~:~ ..:~:i:~i.~i~ :~!~ !~i~:: ~ ~ !~ ~ "7 .' . . . . . . . . . • . . . . . . • . . . . . . . . . . .

,f