Cytological and Genetic Studies on Drosophila algonquin and Related Species Citation Miller, Dwight Dana (1940) Cytological and Genetic Studies on Drosophila algonquin and Related Species. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/adsk-sq92. https://resolver.caltech.edu/CaltechTHESIS:03032025-224902902 Abstract No abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Biology) Degree Grantor: California Institute of Technology Division: Biology Major Option: Biology Thesis Availability: Public (worldwide access) Research Advisor(s): Sturtevant, Alfred Henry (advisor) Dobzhansky, Theodosius Grigorievich (advisor) Thesis Committee: Unknown, Unknown Defense Date: 1 January 1940 Record Number: CaltechTHESIS:03032025-224902902 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:03032025-224902902 DOI: 10.7907/adsk-sq92 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17042 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 04 Mar 2025 19:07 Last Modified: 04 Mar 2025 19:07 Thesis Files PDF - Final Version See Usage Policy. 27MB Repository Staff Only: item control page

CYTOLOGICAL AND GENETIC STUDIES ON DROSOPHILA ALGO NQ.UIN AJ:Jl) RELATED S PEC IES Thesis by Dwight D. Miller In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CALIFORNIA IN3TIT..UTE OF TECHNOL03-Y Pasadena, California 1940 i T .AB.LE OF CONT ENl'S Introduction 1 Chromosomes of Drosophila algonguin 3 Method 3 Structure of the chromosomes 3 Variation in the chromosomes 6 Comparison of the chromosomes with those of other species 13 Mutants of Droso ~)hila algonquin 23 Descriptions of the mutants 23 Linkage relationships 26 Allocation of the mutants to the chromosomes 28 Interspecific Hybrids 32 The algon1uin/athabasca hybrid 32 The affiniVathabasca hybrid 38 The affinis/athabasca isolation mechanism 40 Summary 47 Bibliography 48 Ac lcnowledgements 51 1 INTRODUCTION Drosophila algonguin Sturt. and Dobz. (1936) is one of the American species of Drosophila included in the affinis group. That is, it is one of the six different species into which the old Drosophila affinis Sturt. (1916) has been divided (Sturtevant and Dobzhansky, 1936). The members of the affinis group appear to be close relatives of the western American Drosophila pseudoobscura Frol., which has within recent years proved to be of considerable value in cyt ologica.l and genetic studies of wild poDulations and has given basis for speculation concerning the course of evolution within a Drosophila species and between species (review in Dobzhansk<;, 1939). The members of the affinis group as well promise to be useful in extending our knowledge of the nature of the relationships of forms of Drosophila found in nature, and it is with this in mind that a study of the cyt.ology and genetics of one of these species --D. algon~_g-- has been un::lertaken. In addtition, the research has been extended to include an investigation of c e rtain cases of species hybridization in the affinis group. The known distri but 1 on of Dros O':Jhila algonguin is given on the map of figure 1. The distribution of this species over- lans the distributions of three of the other affinis group species: D. narragansett, which has been found within a rather narrow strip of territory from Wocxls Hole to Wooster; D. affinis, which occurs in addition in places south and east of the localities from which ~algonguin has been ob'§ained; and D. athabasca subsp. mahican, which has been found southeast of the distribution area of D. a:Lgonquin ( Great Smoky LJ: ountains of North Caroline) but occurs mainly north am west of this region. In Texas the FIGURE r.-Distribution map for Drosophila algonquin. The black circles represent places from wh ich strains of this fly were studied as to chromosome structure and variation. The circles with X's in them represent places from which Drosophila algonquin has been collected but from which no strains were available for this study. 2 distribution area of Drosophila algonquin overlaps that of D. oseudoobscura. The three remaining affinis group forms are ap- parently isolated geographically from D. alEonguin: D. athaba~ subsp. athabasca occurs in the northwestern United States, western Canada, and southern A laska; D. azteca has been found within a small area in northern California, but occurs principally in Mexico and Central .,'.merica; D. seminole has been found in a rather restricted area in southern Alabama. 3 CHROMOSOMES OF DROSOPHILA ALG-ON :¾UINJ'· Method The study of the chromosomes of this species has be e n accomplished through the use of aceto-carmine smears of the ganglia and salivary glands of larvae. preparations °\•'! ere used. Both permanent and temporary The first salivary gland preparations, and most of t h os e from which the Standard chromosome drawings (Plate I) were made, were permanent ones, prepared aft e r the alcohol vapor method of Bridges (1937). In the investigation of the various strains of the species for chromosomal aberrations, where a large numb-2r of slid e s nee ded to be made, it was found sufficient to use temporary preparations, which, aside from being temporary, have seemed quite as satisfactory for general analysis as permanent ones. All drawings of these chromosomes have be e n made with the aid of a camera lucida. Structure of the chromosomes Drosophila algonquin has five pairs of chromosomes {Sturtevant and Dobzhansky, 1936). Seen at metaphase in the giant gang- lion cells of the female larva these apDear as a pair of large V-shaped X-chromosomes, a pair of V-shaped autosomes (except when se quence B- 3 i s nresent, see below), t wo pairs of J-shap ed autosornes --the smaller one with an almost terminal spindle attachment--, and a pair of dot-li l-rn au.tosornes (Plate I). male the Y-chromosome is J-shaped. In the In the salivary glan:i cells *M ost of the sub~ect matter of this section has already been uresented in publication (Miller, 1939). Figures 1, 2, 3, 4, and 6 and Plate I are t aken from the published paper. MILLER, CHROMOSOMES OF DROSOPHILA ALGONQUIN LONG UMBOF X-CHROMOSOME l 2 1 I ! I ·. I " • I' 4 3 5 6 7 8 'fTTTr7,--CCrr;-..,.,.__-..-,rr.-r,rrril f GANGLION CELL -~.___.,_._~.li..U;_---JLlu.Jl.JlL.Jl."'----i.,)'---~c.JL..J,.___,.__,.....__ SHOITT LIMB OF X-CHROMOSOME 21 22 ,. ..., META PHASE x~ 23 25 . ,- r 29 30 0 I i '! : ; 10 I 20 l I/a ~5:S- ~ \A MICRA A-CHROMOSOME D-CHROMOSOME - LONG LI MB OF B-CHROMOSOME 56 57 58 SHORT LIMB OF C-CHROMOSOME 61 SHORT LIMB OF B-CHROMOSOME 69 I 70 . 71 86 ~ ' - - - '/ --L...U.l...'..;__JcL;;...,l-'-l,;_.-'-'-LI"-'-"""-'---'-...1L..>UL:.:..iJLC..LILill.Ll..U.l---.l.L.J...ilUUL._lj~JU,;,:J. GENETICS 24: 1 SEPT. 1939 PLATE 1.-The salivary gland chromosomes of Drosophila algonq·u in. The sequence is the Standard used in this study. The black triangles indicate the breakage points of the inversions described in the text. 87 4 of the larva the nucl eus cont a. ins seven long and one very short euchromatic strands extending outw a rd from the chromocenter ( Plate I ). The correspondence between the strands seen in the salivary gland c e lls and t he met a Jhase c h romosomes, although not securely established, is suggested by a number of facts. Two of the strands ap ~e ar paler am thinner in preparations made from male larvae than in those from females. It is concluded tl"lat these repre- sent the arms of the X-chrornosome , which, since it is haploid in males and diploid in females, and since homologous chromosomes are generally paired very closely in salivary gland cells, might be expect ed to ap:)e ar thinner in males than in fem a les. According to 3au er (1936) the chrornocent e r of the salivary gland nucleus of Drosophila consists of the fused heterochrornatic regions of all the chromosomes. If the chromocenter is broken, due to pressure applied on the coverslip of the salivary gland preparation, it seems to have a tendency to cle ave along lines separating the different chromosomes. Assuming this to be true, through observation of a large number of salivary gland cells in which the chromocenter has b ee n broken, one can get an impression of how the various euchromatic strands are attached to heterochromatic masses to form the chromosomes. In Drosophila algon- .9.uiE the longest of the autos omal strands is very frequently seen entirely separate from the rest of the strands, with a s1.:iall mass of rather compact heterochromatin adhering to its base. It is concluded that this association represents a chromosome, one arm of which is wholly heterochromatic, and this chromosome --designated the A-chromosome-- is identified with the short, almost telocentric chromosome of the metaphase plate. Two of the other strands are often seen attached to each other through 5 a small amount of heterochromatin. the B-chromosome. This association is called The two remaining long euchromatic strands are also frequently seen together. However, these arms are se- parated by considerably more heterochromatin than are the arms of the B-chromosome, and the mass of heterochromatin is often either broken in two or is incompletely detached from the rest of the chromocenter. This is called the C-chromosome. The i- dentification of the B- and a-chromos omes with metaphase chromosomes is based on the presence in the B-chrornos ome of an inversion involving both arms, correlated with an altered shape of one of the metaphase chromosomes --the V-shaped autosome (see below). The metaphase chromosome undergoing this change of shape is cal.led the B-chromosome. The remaining large metaphase chromo- some is cal Led the C-chromosorne. The very small euchromatic mass of the salivary g land nucleus ap;)ears to be independent of the rest of the strands and most probably represents the dotlike chr omosome seen at metaphase --called the D-chromosome. The drawings of the salivary gland chromosomes represented in Plate I are composites of a number of camera lucida drawings which seemed to depict the chro mosome strands reasonably well. No attempt, however, has be •m made to record all the fine, scarcely visible bands in these chromosomes, it being considered sufficient for the pur 1 oses at hand to note the relative positions of the most prominent ones. Consequently, one need not conclude that regions represented on these maps as devoid of bands are actually homogeneous and without special structure. For convenience in studying these euchromatic limbs the chromosome comnlement has been divided arbitrarily into 100 numbered sections, after the manner of Bridges (1935). 'Wherever it was possible with out making two adjacent sections too differ e nt in 6, size a rather prominent band or group of bands was used as a section boundary. In the short limb of the x-chromosome an:i in the long limb of C there a T,ears to be a region having a strong affinity for heterochromatin. The result is that in each of these strands the basal section --section 21 in the X, 81 in C-- usually is attached at both ends to the chromocenter. Each case is represented in the map of Plate I as a section separate from the rest of its arm with heterochromatin adhering to the separated ends. It would seem likely that these represent ac- tual insertions of heterochromatin into euchromatic arms. The nucleolus of the salivary gland nucleus appears in a number of figures in association with the Short arm of the Xchromosome. In some especially clear figures the euchromatic uart of the short arm of the X was separated quite completely from the chromocenter --very little if any heterochromatin adhered to it-- but was attached to the nucleolus b y thin strands of rather lightly staining material. This association would seem to be similar to that described for the X-chromosome of Drosophila melanogaster by Kaufmann (1938). Variation in the chromosomes The study of the salivary glarrl chromosomes of a number of species of Drosophila has shown that the chromosome band pattern has a tendency to vary within the species. The nature of this variation is easily deduced from the configurations obtained in individuals heterozygous for pattern, and such differences as have been found ap '.-:iear to be due to inversions of sections of chromosomes. An investi gation of the chromosomes of 2..!,_al- gonguin has shown a considerable degree of this sort of variation within this species. 7 Various strains of D. algonguin from places shown in the distribution map (fig. 1) were available for chromosome study. Each strain consisted of the descendants of a single female fly captured in the wild. An examination of the saliv1;1-ry g land chromo- somes of individuals from the various strains has shown one of the gene arrangements to be present in all the localities from which collections were made --though not in all the strains. This arrangement has been arbitrarily chosen as the Standard Sequence. The nature of the arrangements present in each of the available strains --and in a few wild males-- was investigated by mating flies from it to individuals of the Standard type and then examining the salivary gland chromosomes of the offspring larvae. If a strain contained a sequence different from the St 2, ndard, the charc~cter of the difference was studied through the pairing configuration achieved by the homologous chromosomes in the salivary gland nuclei. No sequence besides Standar d was found for the long arm of the X-chromosome. In the short arm of the X a sequence dif- fering from standard by a single irnrersion was found. The limits of this inversion were determined to be in sections 27 and 34. This new sequence was c·alled X-1, and the configuration foundin the salivary gland nucleus of a Standard/X-1 heterozygote is shown in figure 2A. Se a uence X-1 occurred in strains from Bel- fast and Twin Mountain ( near Franconia Notch) . '.Bwo sequences besides Standard occur in the A-chromos ,)me. One of them - -cal led A-1- - differs from Standard by a single inversion the limits of which are in sections 44 and 48 (figure 2B) . Sequence A-1 has been found in strains from the following localities ( named from northeast to southwest): Twin Mountain, 20 31 10 0 MICRA 47 FIGURE 2.- In version configurations in the salivary chromosomes of individuals heterozygous for the Standard sequence and t hese inversions. A. Standard/ X - 1. B. Standard/ A-IC. Standard/ A-2. 8 Woods Hole, South Amherst, Meriden, Derby ( near Meriden), Varna ( near Ithaca), Ithaca, Chautauqua, and Wooster. The other se- quence --A-2-- differs from Standard by two overlapping inversions (figure 20). A structural heterozygo t e of the constitu- tion A-1/A-2 exhibits a single loop configuration, so that it se e ms that one of the inversions whereby A-2 differs from St andard is identical with the inversion differentiating A-1 from Standard. This suspicion is confirmed through an inspection of the limits of these inversions. The inversion overlapping the A-1 inversion in .A-2 h a s b2:·e akage poi nts in secti ons Li-3 1:. nd 46. Sequence A-2 has been found to oc cur in Belfast, Gray, Twin ~fountain, Woods Hole and Derby. It has been pointed out ( Sturtevant and Dobzhanslcy, 1936a; Dobzhanst :y and Sturtevant, 1938) t h at sequences dif f ering by overlapping inversions are of particular interest because of the insight they g ive into t he probable phy logeny of the chromosome. It has been explained that the most probable relationship of t wo such s e quences is throu 0 h a third sequence differmng from e ach of the form er by a single inversion. That is, although one may not b e able to tell which sequence arose first in the phylogeny of the c hromosome , the conclusion seems most reasonable that one of the two arran_ em e nts d i d n ot ar i se d irectly from the other, but rathe r arose a fter the other had been changed through a single inversion; or, perha ps, the intermed iate sequence came first and by a process of two inversions g ave rise to the se qu e nc es in quest ion. Now, given two chro mos ome arrangeme nts diff ering by ov e rlap -" J i ng inversions, it is p ossible to hypothesize the intermediate sequence. In the cas e of t he A-chromosome sequences S t ands.rd and A-2, however, s equence A-1 represe nts suc h. an intermediate s <J·~ ue nce. The probable rel ations h i p of 9 these three se ~uences is diagrammed in fi g ure 3. Three gene arrangemsnts other than Standard have been fo ,_1nd in the B-chromosome. One of them -- B-1-- involves an inversion in the long arm of B with break points in sectisns 59 and 68 ( figure l~A) . Sequence B- L is found at Gray and ·."oods Hole. Another is B-2, which diff ers from Standa:cd by an inve r sion in the short arm of B (figure 4B). This sequence (inversion limits in sections 70 an:1 76) has been found only in one of the strains from ···oods Hole. Since the inversions that differentiate sequences B-1 and B-2 from Standard are pr e S'-.lmably quite indepen:. ent, se- parable through crossing- over, a name has not been adopted for the B-chromosome arrangen:ent differing from Standard by both the B-1 and B-2 inversions, s u c ::1 a sequence when encountered merely being considere d to be a combination of sequences B-1 and B-2. Se quence B-3 differs from S tandard by two overlapping inversions one of which includ es the spindle attachme nt --and all the heteroc h romatic region of the chromos ome as ~-1elL. The spindle attachment inversion has break points in section 57 of tl1.e long arm of Band in section 75 of the short arm, and the i nversion that overlans it has .limits in sections 64 and 72. The hypothe- tical sequence intermediate between Standard and B-3 ( as A-1 is int e rm ed iate between S t andard and .A -2, above) should differ from Standard by the spindle-attachment inve :, sion alone. 1 a se quence has not been found. Such The supp os e d relationship of sequences Standard, H:; not -:-1etical, and B-3 in the B-chrornosome is diagrammed in figure 3. Sequence B-3 is quite widespread, occur r ing in the fol.Lowing localities: Belfast, Gray, Valmorin, Twin !.r- ountain, Ga.le ?'. iver (near F r anconia Notch), Woods Hole, South Amherst, :r,: eriden, Derby, Varna, and '.,\'ooster. A• B C D E·F G H I J STANDARD STANDARD t A-I A·B C D H G F·E I J A-2 A• B G H D C F• E I J Figure 3. Diagram showing the most probable rela- I * HYPOTHETICAL \ B-3 tionship of gene arrangements differing by overlapDing inversions. Such a difference involves four breakage points. It seems likely that the rearrange- ment has occurred in two steps, two breakages resulting in one of the inversions at one time, two more breakages resulting in the other inversion at a later time. This proces s seems especially likely to have taken place if in addition to the two sequences in question the ~ypothetical intermediate arrangement is found in nature. This is the case with these- quences found in the D. algonguin A-chromosome, to which the labels on the left pertain. The spindle attachment of this chromosome is represented by the dot between A and B. The labels at the right per- tain to the B-chromosome of al3onguin. The spindle attachment of this chromosome is represented by the dot between E and F. lO The B-chromosome of s e quenc e B-3 has the length relationshi p of its two arms c o ns :":. d e rably different from t h at of t he Standard R-chr om os ome. The lo ng arm D/ J ,c ar s longer; th e short arm s h orter, b y virt ue of t h e f a ct ".:hat t h e long arm gains by th e s -oi ndl e att ac tunent invers i on more t han h a lf t he l engt h of the s h ort a rm ( euc hr omatic) --the part prox i mal to the bre a k age point in section 75-- in return for the s :.- wrt s e ction p r oxi .. al to t he b reak ~oint in 57. Suc h a difference seems to b e reflec- ted in the met a-ohas e c h romosome s of t he g anglion c e lls in that wit h the ,Jr e senc e of se qu ence B-3 t :1ere a p :) e ars in the metaphase gr ou ~ a J-sha pe d c hr omos ome not otherwi se p r e s e nt. S t urtevant and Dobzh ans 1<:y (1936) deAcribe the meta phase chromos om e group of D. al? on:1u.in as containtng no V-shaped autosome. -~1•- It is most ·~ ··•f'- likely that th , individu al s ex ami nE,d b • these wo nlcers containe d the B-c hr '.Jmos ome s equenc e B-3 '.:'. nd not the S tandard. Metaphase plat e d :::'awings f or indiv j_duals hom 9zy z o:.J. s a nd het e rozygous for sequ e nc e B-3 ar e sho wn in fi gur e s 4D and 4 E r e spe ctively. An individual het e rozy gous fo r a ttachme nt may b r?. an inversion across the s p i ndle expect ed to g iv e rise to a num be r of inviable offsuring d ue to cro s sing-ov e r in t he reg i on of the i nversion, which s h ould r e s u lt in a d'J.Dlication and d e fici ency in the crossov e r 'Jroduct c l1romos ornes (Sturteva nt and Be adle, 1936). Thus, a sequence ar i sing in a wild popul ation throug h an inversi on ac ross the s p i ndl e attachment mig ht be exptected to be at a disadva nt age in the fac e of natural selection, s ince heterozygotes for it and t he prevaili ng s e que nce ar e re nd er e d p artially st e rile. Howev er, if t he s pi ndL) attachment inv ersion is overlap?ed by anot he r i nve ': Sion, h et e rozygotes for thio new· sequence a nd the ori g inal one might b e e x pe ct e d to be at l e ss of an ev olutionary disadva:1t 9.ge t h an t he a bove menti oned het e ro zygotes, since cross- 0 I 10 I I MICRA 20 79 I B C C~r.XA 7~ ,, , ' / 6~--, - I ' \ " D i:tf \ \ \ X '..... _. r:~A 65 e E FIGURE 4-Inversion configurations. A. Standard/B-r •B. Standard/B-2 C. Standard/B-3. The chromosome may be considered to be continuous through heterochromatin between sections 69 and 56, although in this particular figure the heterochromatin connection is broken. D . Ganglion cell metaphase plate from an individual homozygous for sequence B-3. Compare the shape of the B chromosome here with that of the B chromosome in the metaphase plate shown in Plate r. E. Ganglion cell metaphase plate from a Standard/B-3 individual, showing the two kinds of B chromosome. The scale of micra for this figure pertains only to the salivary chromosome drawings. The ganglion cell chromosomes are drawn on a slightly larger scale. 11 ing-ov er in t he r egion ov erl auoe d by t he two i nver s i o ns (since one i 7.c b d e s t he s 9i ndl e attac h.'Ilent) r e s u lts in products that a re not distrih.1 t e d to egg pronuclei (Sturtevant and Beadle, 1936); t hat is, cros s i ng - over in t h e reg i on of the s p i ndle at- tachme :,t l nv e rsion is vi : :-'tua lly r e du c ed. S uch is th e relation- s h i p of t he Droso·ohil a a l f-". ong u i n B-c h r om os ome s e que nces Standard and B-3; e ach diff e rs from t he ot her by a n i nversion including the s·o:i.!1dle attac l1.ment a nd an overl e_p ·1ing inve r sion. S ta ndard e. nd B-3 ar e a t A:x -are ntly littl e c1is advant age in t h e p r e senc e of each othe r, s Lic e this s tudy s h ows t b at t hey exist toget her in a cons i derable numbe r o f nlace s. Ac c ording t o t h e above r e ason- i nr , a c omb ination of Standard and t he B-chromos ome Hy-;J othetical s equenc e (dif fe ri ng from S ta nd a r d by t he S!Jindle att a c hment inv e rsion alone) mi ght b e -:x pect ed to b e att e nde d wit r1 gr e ater disadvantage. But a ccording to the theor ·, of overlap ,i ng inv e r- si o1:1s , St a ndard a nd Hyp ot hetical must have e x ist e d tog e t her at one t i e. Thu s , in s p it .: -; of t he ap ,are nt disadvantage, it se ems th a t Standard aros e fro m Hy p othe tic al and p ersist e d or that Hy p othetic a l arose from Standard and r emained long enough to give ris e to B-3. The suD--::i osition that this evolutionary st ep is di f -<' i cu lt is swr ort e d by t :-ie fact tl1at u nt i .L now, a s far as th e aut h or is awar e , ::10 inve :c sion across the s p indle attachment h a d b ee n foucmd in wild populations of Droso :Jhila, al t h oug h such c '::1 r omos orn a l aberrati ons ha.ve be e n derived by X-rays ( " EJv. cEmt :c ic i nve rs:i ,:ms !I of Catcheside, 1938). (According to a personal com- munication of Professor Th. Dobzhansk;y, a spirrlle attachment inversion ap p arently exists within DrosoDhila du ncani.) The spindle attachment inversion ( inversion of Hy :p othetical with res pect to Standard) overlaps the inversion of sequence B-2. Thus, Stania.rd fo r ms a s eque nce i ntermed iate between B-3 80 7675 • • 7_2 • 70690 56 57 5 _9 • 6 _4 · · • 6,8 HBv°P~;~--.:~::L~-:--::!::_-~::r~:::_~J __ ___ ____________-__ ----1----------"i3',.-, A HYPOTrTIC AL--➔ B-3 B·I OR B l:' i gu:c e 2· A. Diagra m showing t he inversions with respect to Standard in the B-chromosome of D. a l ~onquin. The broken lines connect the limits of the inversions. B. Probable phylogeny of the B- chromosome, based on overlapping inversions. l2 and Hypothetical on the one hand and B-2 on the other. The in- version overlap , ing the spindle attachment inversion overlaps as well the inve ::." sion of sequence B-1. How2ve1' , the spindle attachment inversion itself is indepenient of the B-1 inversion. Consequently B-1 may be considered to be separated from B-3 lm point of ti me by Hypothetical at least, but possible by Standard, or even B-2 as well, since the inversions that differentiate each of thes ::J sequences from Hypothetical are quite independent of that by which B-1 differs from Hypothetical. The diagrams of figure 5 may serve to make the matter of probable phylogenies in the B-chromosome clear. In the C-chromosome there have been found two arrangements besides St,andard. One of them differs from Standard by an in"".' version in the long arm of C extending from section 82 to the boundary of sections 88 and 89 ( figure 6A). This sequence -- called C-1-- has been found in Belfast, Gray, Twin A:i ountain, and South Amherst. T ;,.,e other non-Standard sequence differs from Standard by an inversion in the short arm with one breakage point at the 95-96 bounjary and the other quite close to the tip in section 99. In structural heterozygotes of the Standard/C-2 type the short region distal to the inve r sion iS generally not paired in the salivary gland cells, and the configuration .r.1.as been illustrat ed from one of' the usual figures (figures 6B and 6C) . The distribution of the gene sequences described above is given in Table I, as well as the total number of strains examined. Since the known distx'ibution of the species follow roughly a line running northeast-southwest, the place names have been listed as they come from northeast to southwest. The impression may ,, '' I '' '' ' '------.. B 0 I ' 10 I MICRA 20 I FIGURE ,:- Inversion configurations. A. Standard/ C-r B. and C. Standard/ C-2 , the same configuration at two different foc i. T his configuration is typical for th is in version heterozygote. Sometimes the last two discs of section 99 pair completely. Table I Locality Strains Sequences besides Standard Belfast 1 X-1, A-2, B-3, C-1, C-2 Gray 2 A-2, B-1, B-3, C-1 Valmorin 1 B-3 Franconia Notch (Twin Mountain, Gale River) 5 X-1, A-1, A-2, B-3, C-.l, Woods Hole 5 A-1, A-2, B-1, B-2, B-3 South Amherst 1 A-1, B-3, C-1 Meriden (Derby) 3 A-1, A-2, B-3 Ithaca (Varna) 5 A-1, B-3 Chautauqua 6 A-1 Wooster 2, 13 males A-1, B-3 Wichita Falls O, 1 male none C-2 Austin ( Aldrich) 43 none San Antonio 1 none 75 14 males 13 be gotten that chromosome arrangement in D. algonouin is more variable in the northeast than inthe southwest, although the numbers of wild individuals studied is rather too small to allow for generalizations of this sort. At any rate the uniform- ity of the Texas populat i.on of the speci e s is stri king. In the fall of 1938 there were examined 39 strains from Aldrich, and in none of the se was a gene sequence besides Standard found. It was suggested (Miller, 1939) that the population from which these strains orig inated might have b e en the result of an introduction of a few D. algonguin individuals from the north. In the fall of 1939 there were received and studied four more Aldrich strains and flies from two new Texas localities --VHchita Falls a nd San Ant Jnio. quence found .. In none of these was a non-Standard se- In view of the contin_;_ ed occurrenc e of Standard ~lgonguin at Aldrich ani the findine; of this species at Wichita Falls and San Antonio, the suggestion that D. algonguin was artificially introduced into Texas seems now rather implausible. A stud y of D. algonquin from the territory between the Texas region and the remaining localities in which the species has b een found should prove most interesting in view of the ap~Jarent ho mog eneity of the Texas population and the marked heterogeneity fou nd in the species elsewhere. Comparison of the chromosomes with those of other species The salivary gland chromosomes of some of the other members of the affinis group have been studied: those of Drosophila azt~ (Dobzhansky and Sokolov, 1939), D. athabasca (Novitski, unpub.), and D. affinis (?aul, unpub.). In addition, the existence of certain inters pecific hybrids between affinis group members has made it possible to make direct comparisons between the chrome- 14 somes of some of these closely related species (Bauer and Dobzhansky, 1937; and see below). In the salivary gland nucleus of each of the affinis group species studied there a p ,)ear to be seven long euchromatic strands and one very s h ort one. (This now se ems to be true of D. azt eca, since the recent discovery in this species of a chromosome arm not included in the maps of Dobzhans'.{y and Sokolov, 1939). An examination of the drawings of these chromosomes, aupplemented by some observations on the chromosomes themselves, shows that, although in gener al the gene arra :1ge m,~-mts chose n as Standard for each of these species have lit:tile recognizable similarity to each other, a few points of lilc ness suggest themselves. For exarm)le, in each of these speci e s one of the arms of the X-chromosome is very long and contains a slightly bul5ing subterminal region devoid of darkly staining bands --depicted in the D. algonguin map in section 20, in the D. azteca map of Dobzhansl{'J and Sokolov ( 1939) in section 23. Whether or not such apnarently si milar regions ar f: actually homologous is dif f icult to determine in t c1is ~,,ay. However, such observations may serve to direct one's attention to ~o osslble homologies when methods better suited to testing homology are available --such as intersnecific hybrid chromosome stud y ( see be low). In D. algonguin, D. athabasca, and in D. azteca, but apparently not in D. affinis, one of the autosome arms is very long, being of about the same order of length as the long arm of the X. In all four of these suecies two of the autosome arms are quite constantly attached to each other through a small amO 'J.nt of het e rochro matin --the B-chromosome of D. algonquin, and of D. azteca. A rather con- stant feature of the short arm of this chromosome among these species is a pattern of bulge s and constrictions near the ti p --shown in sections 78, 79, and 80 of the D. algonguin map, an:i in sections 89, 90, and 91 of the D. azteca map. T h e pattern of s e ction 91 of the long arm of the C-c h romosome of D. a lgonquin as well as the characteristic terminal flare of this limb is found in a chromosome arm of com p arable length in each of the other species studi ed --for examp le, the unLLlustrated limb of D. azteca. (Dobzhansky a nd S ok olov, 1939). Also, the short arm of t h e C-c b..romos ome of D. a l go no.u in ha.s an e.p ~'.: are I:.t l1 omc1o 6 ue in each of th e s e species in that the p attern of i ts dist a l region ,rJp e ars to b e ne arly duplicat ed in a chromo s ome strand of each of them --comDare section 99 of the D. algonguin map with section 99 of the map of the C-c hromos ome of D. azteca. fifth chr ornos -Jme seems Finally, the ~u ite s imilar in the different affinis grou p speci e s studied, so t hat it might ?rell be suspect e d of b e ing homo Log ous t hrough out. Fortunately, it has b ee n found possib l e to cross D. algonquin to a n other affi n is group species, D. at h abasca. (A g e neral account of t his hybrid is giv e n below , see Interspe cific Hybrids.) A number of obser .r ati ·; ns h a ve b e en made on the s alivary gland 0 chromos omes of hy brid larvae. Alt h ough hybrids are quite viable and a :JDare nt ly r obust, the saliv a r y glands of even the b e st of th e se Larvae seem to b e s ma ll er th 2. n is g enerally t h e cas e for eit her s ·>Jec ies, and the chromos omes in turn are s mall and slender and not wel l fitted for examination. Th e salivary g land nucleus of hy brid l arvae h as the ap··:)e ar a nc e of a mas s of coiled strands, most of the m, if not all, unpa ir s d. Howe ver, the r e may be an association of what ma y be t az.e n for h orno1o g o:.1s strands in cas e s eve n wher e no intitf\,., at e synapsis has t a ke n 9 lac e , th e h omologues be ing coiled lo osely a b out each othe r with t he ir bas e s arising 16 from the chromocenter rather close together. In the salivary glands of both D. algon ouin and D. at L'1abasca it is rather a common occurrence fo r def i nitely h omolog ous regi ons to fail to syna,se, although pairing is the general rule. This tendency is e vi de-:1tly chare.ct eristic of hy b r~ ds L -! t hat c ert a in si m::.. l ar re- gions of the c ~1r omosorGe s ha~ve be o n found. both in t£1e J)Cti1-')ed a nd un'Jai red conditi o'.:.S a ~a. still ot he r regions t ::at se e m to b e alike hav e never b ee n s ee n pair ed . The drawings of fi gure 7 have been made from p2eparations of salivary glan:1s of larvae from t he cros s D. algongui !l (.Aldrich) fem o.les ti;r;es D. athabasca (Grays River, Wyoming) males. These d e:Jict r egions t hat we --:- e found pa ired or sufficiently ap :'J os e d so that h omo Logy might be consider e d l i k ely. It is not meant to b e imDli ed that t he se represe :1t the only cas e s of such extensive homology b et we en t he chr omos .Jme sets, it be ing quite probablt1 tha t furth e r study will Show mor e instanc ,, s and wil L give a bet- ter id 1:Ja of t he nat 'J.re of the d °L ffer e:1c es betwe ':m the chr omosome s of the two S:)ect e s. No inst a nces of synaps is heve been observed betwe e n the X-chromosom e s of these species. Neither h a s arzy be e n fo·.md betwee n the -~-c hromes )me of D. alP:Ol17Uin and its suppos8d homologue in D. at h abasca. The long arm of the B-c r1r omo- some has b een fo und to be pair 0J d with a strand of si ruilar length only ne ar the bas e in p art of se ction 56 (fi g ·J .re 7A). The short arm, however, s e ems to be more completely si::;ilar in the t vvo species in that pairi ng has be e n found near the bas e , i n the middle, and towards t he tip ( fi gure s 7B, -C, and -D). The basal section of the long arm of the C-chromosome p airs with a similar re g ion i n D. at habasca ( fi gure 7E) . chromos ome arm ther 0 : Towards the tip of t i:1is i s a ls o synapsis. However, a p oint of oJf- ference ex ists here in t h at the promine nt dark t e rminal band 10 0 I 20 I I MICRA A 75 -- - ~~ ~ ' ~ •• C 80 81 ~ l}J}jfJ ; 78 '(dj}:¼ I . - ~)11rof, , E F D H G Figure 7. 9 Examp les of pairing or close ap proxi- mation of the chro mosomes o f D. algonquin and~ athab asca as seen in t he sal ivar y g lands of hy brids. A. Bas e of t h e l ong arm of t he B-chromos ome . Bas e of the short arm of the B-c h ro mosome . Central r egion of the s h ort limb of B . of the short limb of B. of the C-chro mo some . c. B. C. D . Tip E . T?. a s e of the lo:ng arm ? . T i p of the l ong arm of The at h aba sca homol og u e h a s a he avy t e rminal band appar ently not 'Jr e sent in the algonqu in c h romosome. G . Short li mb o f the C-c h romosome . chromos orne . H . D- 17 of t he D . at hab asc a chromosome seems not to be pres e nt in the D. algonguin homologue ( fi gure 7F) . A.n ex ample of such a ter- minal difference b etwe e n otherwise homologous chromo s omes has b e en point e d out by Ki l<:l<:awa ( 1938) wit h in the speci e s D. ananas s ae. Also, Metz ( 1935) has shown that such terminal differen- c e s occur in the chromos omes of Sciara, where minute variations in chromos)me structure are rat h er commo n. '/:hether the present case indi c ates a duplication or d e ficienc y in one speci e s with respect to the other or some ot h erkind of diffe:cence cannot be determined from the evidence at hand. T ~1e short arm of the 6- chromos ome of D. algona uin finds an almost entire homolog ue in D. athab asca. It is been se e n pair ':',d fr om wit h in section 94 to the distal end with but two small regions of inc o,uplete synapsis iriterv e n i ng ( figu :>e 7G). The small D-chro mos ome has been seen ~aired th r oughout (figure 7H). The analysis of the salivary gland chromosomes of hybrids as a method for determining the e xtent of chromosome difference between speci e s has been made in a rumber of cas8 s in Drosophila: in the D. psaudoobscura rac e hybri d (Tan, 1935; Koller, 1936), the DSeudoobscura/miranda hybrid by Dobzhansky and Tan (1936), th e melanogaster/simulans hybrid ( ? atau, 1935; Kerkis, 1936; Horton, 1939) , the athabasca/ azteca hybrid by Bauer and Dobzhansky (1937), and the D. viri l is subspecies hybrid (Hughes, 1939). Var~r ing degrees of deter minable difference have be en found between the chromosomes entering into these co:nbi nations, from a few inversions to many inversi ons and s mall trans locations. In the athabasca/ al9onq_uin hybrid the avail abl e evidence allows :for very little to be concluded a s to the nature of t he chang e s t hat h ave taken place b et we e n the chromos omes of these s peci e s. ? robably due to the ad d ed difficulty of p air i ng t hat would be 18 expected to atten:l such a rearrangement, not even an example of homolog ous regios.s inv erted with r e s pect to each other h as be e n found in this hybrid. However, that consider able change h a s taken p lace between the chromosome ar r a ~ig8ments of the straL 1S used in deriving the hybrid, probably involv ing many inversions, s e e ms likely from the a p:Jarent differe nces in band pattern betwe en the s De ci e s, as wel l as fro m t he high degree of failure of syna '')Sis i n h ybrids. For the present it se e ms most p rofit- ab l e to c onc e nt r ate attention on t ::1 0s e regi o ns that have remained s irnilar to the extent of b e ing able to pair in hybrids. Such re g ions ar e dmonst r able only in the shortest chromosome arms of these s ~eci e s. Of much i ntere~t in con.Dection with a comparison of the c h ro mos omes of D. algonguin wit h t h ose of closely related species is a n analysis of the salivary g l a nd c h romo s :; me s of a n.ot h er hybrid --that involving D. at habasca and D. azteca. A preliminary account of t h ese chromosom,:, s has been p ublished by Baue r and Dobzhansky ( 1 9 37), but the writer is ind e bted to Prof. Th. Dobzh a nsky for furt her information on chromoSJ me p airing in t his hybrid. The nat ure of the salivary chro mos omes seems _to be quite si :-nil ar to that of the one s fou nd in the a bove me ntioned a.lgon..9.uin/ athabasca hybrid. That is, th e saliva r y g land nucl e us g ene- rally consists of a mas s of coil e d , unp aire d stra nds t he relationshi u of which to each ot h er is dif f ic u lt to make out. a few insta,nces of h omology hav e b een found. However, Here, as well as betwe e n D. at h abasca and '.: ' . a.l g onquin,homolog ies a r e most easily d e monstrab le b etwe en the s h ort e st chr omosome a r ms. Instances of pai r i ng h ave b e en found in both arms of the B-chr omos ome, in the C-c h r omoso rne, and i n the s mal l , dot-lik e D-c h r om o s )me. 19 Particularly striking is the almost comnlete homology of the C-chrornosome of D. azteca with the stra.nd in D. athaba~ which has been found to be larg ely homologous to the s h ort arm of the C of D. algonquin. Also, the D-chromosome of D. azteca seems lik e ly to be entirely homologous to the small chromosome of J2..: athabasca, 7rhich seems in turn homologous to the D-chromos ::nne of Q. algonguiE• On the basis of these observations a number of tentative conclusions D.S to the comparative structure of the chromos omes of D. alg Jnguin, D. athabasca, and D. azt 2ca may be made. The same nJ.rnber of euchromatic strands exi.sts in all three species, and the relative lengths of these arms ap-aear so similar in them that j_t may be concluded that no major int e rchange between arms, 'J:1com--:i e nsat e d for, has occurred between the species. However, the salivary gland chromosome :oatterns of these species seem to differ considerably --although possibly less than has been concluded on the basj_s of impressions of dissimilarity and failure of nairing in hybrids. That the occurrence of inversions in the chr omosomes is largely res p onsible for the interspecific pattern differences seems very lilrnly on the basis of the fact that intra-specific variation in chromosome sequence has been found to be due to this mechanism in D. algonquin , D. athabasca (Novitsk:i. , unpub.), and D. azteca (Dobzhansky and 3okolov, 1939). The --:iossibi1ity that translocations have also take n place is not to be ·::xc luded, though the amount of rearrangement a cc omplishea. in this way is probably small. Dobzhansky and Tan ( 1936) have noted a few instances of interchanges between chromosJme s in the useudoobscura/rniranda hybrid, though these have been called into question by MacKnight (1939). One might suppose that in the evolution of these species 20 the patterns of those regions found alike in each ~pecies were the original ones --at least for a.11 but the most pri c.itive s pe cies in case two of them are descended fro m the other one. Consequently, any intraspecific variation on this pattern could be considered likely to be a recent development. In ~algonguin the Standard sequence of the short arm of the C-chr omos ome is one that is large ly ide n tical wit h a chromos :Jme ar::aan~· ement in D. athabasca and i n D. azt e ca. However, the C-9 inversion app arently eff e cts a r earrangement within this chromo some that has not occurred int e rs ·,J ecifically , and for this r e ason the C-2 sequence might be concluded to b e a more rec ent d evelo Dment in the species than Standard. The salivary gland chromosomes of Droso~hila algonguin b e ar little resemblance to those of specL:s outside the affinis group, judgi ng from the publis hed drawings of the chr ,Jmosomes of D. mela nogaster (Bridges, 1935), D. useudoobscura and D. miranda (DobzhanslDJ and Tan, 1936; Tan, 1937), and D. virilis (Hughes, 1939). Aside from the abse nce of s:9ecial similarity in band pattern, these chromos J:ne s diff er from t h ose of D. algonquin in that in each of these sets ther 3 are but six euchromatic arms --five long ones and one very short-- instead of the eight found in '"j ach of the affinis grou::) S'Jecies so far i nv e sti g ated. In- deed, six salivary gland chromo some arms seems to be c h aracteristic of every Drosophila speci es in which th~se chromos Jmes have b ee n studi ed ( except D. d'.-lncani, according to Prof. Th. Dobzhansky) , as ide fr om the memb e rs of the affinis group . It is of interest to spe c ulate as to t1 ow a c ;iange in t he number of euchromatic c h ro ::.1 os ome limbs could b e ac hi ev ': d in nature. Assum i ng that e -, ,rn hromatin is rearranged, not creat ed or lost, two methcds sugge st themselv e s. A chromos ) me may tie tho u.g~:..t 2l of as acquiring, or losing, euch~· omatic material in an arm other wis e heterochroTnatic --ne glecting the p ossibility cbf its being a telocentric chromosome-- throug h (1) a tra nslocation or (2) an inversio n across the spindle attac tunent. That is, a chromo- ~=oorne may underg o a change in the amount of euchr )matin in one of its arms eit r1er throu gh this arE1' s entering into an exchange wit ';:1 another chr ,:-imos ome or with the ot ::-1er arm of the same chromosome. In ei th e r cas e , ·.ml e ss the tra nslocation were nonrecip- ro c al or the inv e rsion termina l, one vrou ld expect a terminal mass of het cirochrornatin to accompany the production of a new arm. ~ oreover, the loss of a euc hr omatic a rm would require a nonreci ';)rocal translocation or a terminal Lnver~ion, and unless the ;'.'lont e r minal breakage point occurred just at the boundary of het •', roc hromatin and e uchr omatin, an int e rstitial mass of heterochr Jmatin should result. (See figu1~e 8.) It is not p ossible at Jresent to say in w:1ic :.1 direction the process would be the rr.or e difficult. Chr omosomal c rrnng es that seem termina l have b e en re'1orted in Drosophila (for example, the terminal inversion of D. ananassae, Kaufman:'.'l, 1936). Interstitial chromosome arm segments that ':'lppear to have het e rochromatin properties have b e en r e ported in Drosonhila in a number of cas F.: s, of which the D. algonquin inst a nce --s :-i ort arm of X, long ar1n of C-- is one (also, the Y-chromosome of D. mire.nda, MacKnight, 1939; sugg esti '.) ns that suc h segments occur in the X. of D. melanogaster, Kaufmann, 1939) . The the ends of chromos omes consist of heterochro- mat in has seemed likely to a n:.J.mber of workers on the basis of the non-s-oecific attractions of the tips of the salivary gland arms --res'..llting i n occasi::mal attachments of en::ls to each other :.1.nd to t h e chromocenter (for example, the alleg ations of Prokofyeva- 3e l g ovs 1rnya, 1939) . However, no large terminal mass of A B MC MION C , D Figure 8. • ------<-- •u- ••o•••••• : 0 1 ___.,. - S chemes whereby a chromosome co u ld either acquire a euchrorr. atic arm or lose on8 , thereby increasing or decr e asing the number of salivary strands of the species. In each case euchromatin is des ignat e d by a narrow l i ne , heterochro matin b y a t hick , irregular one , the spindle attachme nt b y a circle. t e d lines connect p oints of tr ansfer . T he d ot- A. Ac quis ition of a new euchromati c arm through translocation. B. Ac quisition of such an arm through a spindl e attachment inversion. ,... tr anslocation. D. Loss of a euc l1r omati c arm through V • a spi nd le a tt Gc iment L o ss of a euchromatic arm throug h inversion. 22 het e rochr omatin has be :rn obs e rv ed in Drosophila. What the fate of such an arrangem ent would be --for instance, whether or not the terminal ine rt rnateria.1 would be lost r e adily -- is difficult tO S a.y • Both mechanisms sug 6 ested --translocation and a.pindle attachment inversion-- might be ex :)ected to be at t ende e: with some disadvantag e in e s tabl is h:r:ent in a p opulation (Sturtevant and Be adke, 1936). However, the succ es sf u l e stablishment of a spindle at- tac hrne :1t inv eT si o·:1 in natur e has evi c:. e ntly t ake n ;Jl a ce wit h in D. alp.'.onauin, as desc r ibed above. An example of an intra.speci- fic translocation has be e n r epo r ted by White (1940) in a speci e s of grasshopper. As mentione d above, Dobzhansky and Tan (1936) ~ave noted t he chromos omes of D. n seudoobscura 2, nd D. mira:ala to differ by several small tr a nslocations. Sturtevant ( in press) has ma de a com, arison of the mutants of D. affinis wit h t h ose of a numb 9r of s pec ies outside the af~inis group, particularly D. ns e udoobs cur a and D. mela :.1. op;ast er. A study of the linl-{a.ge r e lationsh i'JS in thes e speci e s ha s l ed to the conclusion that no ma jor tr ::rnsloc e.tion of c enetic material from one chro mosome to another - - a s wou ld be evidenced by a dif f erence of linkage re lati 011Sh i ps of apparently homologous gene s-- has ta ken p lace bet'N er,n the eight-euchromatin-arme d species D. affinis and some of those having only six --that D . useudoobscura a nd D . me lanogast e r were not differentiated by int e rarm ex cha ng es had already l-::9e n cone lud ed by Sturtevant and Tan ( 1937) in a similar study. T ~: i s ev idence would S c'1e m to favor the theory of tr ansfor mation by s p indle attachment inversions rather than bJ trans locations. 23 MDT ANI'S OF DR.0 3 0 PHILl1 Ai.Gor,q,ur N Descriptions of the mutants All mutations so far discovered in this species are recessive characters. ~ach of those that has been retained for ge ne- tic study has very good expression --v e r y little if any overlap"'J i ne; of the normal phenoty ~Je-- and reasona b ly g ood viability and f e rtility. Two sex-linked mutants are lcnown in D. algong_uin --dusky (.9:y) and swollen tarsi (_§wt). The effect of dusky (The mutants dus l(y and rough -see below- v,ere discovered in this s p ecies by Dr. A . H . S turtevant) is to reduc e the size of the wings so that they scarcely cover the abdomen and to z ive them a decidedly dark t exture. This is, the phenotype is very similar to that of the sex-linked mutant dusky of D. melanogaster and of ot her species (Morgan, Bridges, and S turtevant, 1925). The effect of s vrnll e n tarsi, as the name i rr: plies, is p r l. ncipally a distortion of the legs. The l eg segments have a n inflated appearance, particularly noticeable in the tarsus. This mutant also affects the wings, making them somewhat shorter and b roade r than in the wild ty ~e fly. There se e ms to b e a conside rab le lethal eff ect accompanying swollen tarsi und~ r certain conditions , judg i ng fr Jm the proportion of such flies gotten in the linkage experiments (Table V). There also seems to be some reduction of the f e rti l ity of females showing this character·. In D. algonquin there are four autosornal mutants --rough (_;::_Q), cinnabar (en), droop (dr), and brown (bw). Flies of the rough type have compound eyes in which an irregular arrangement of the facets and hairs produces a decidedly unev e n surf ace texture. In addition, eyes of this type occasionally contain dark 24 areas of various sizes, ap 1arently due to erupted ommatidia. 0 These spots are particularly noticeable ·when the character occurs in combination with the bright eye color cinnabar. The presence of black erupted areas on the eye has been noted in the case of the D. melanogaster character roughoid (Morgan, Bridges, and Sturtevant, 1925) . In Drosophila algonquin one bright red eye mutant has been found. This character is particularly striking in flies that have just emerged. As the adu:pts become old, the eye color dar- kens somewhat, so that in flies several days a fter hatching the mutant is not so easy to distinguish from wild type. In addi- tion to the compound eye character flies of this kind have colorless ocel li. Within Drosophila species there occur four bright red eye mutants which are often indistinguishable in appearance --vermilion, cinnabar, scarlet, and cardinal. However, it has been discovered that the first two differ from each other and from the others in the reactions they give in transplantations of imaginal eye discs (Be adle and E7!brussi, 1936). Vermilion and cinnabar eye discs are both altered in the direction of wild tyDe ryigmentation when transplanted into hosts that are neither vermilion nor ciTu.~abar. Moreover, vermilion eye discs become wild type when transplanted into cinnabar hosts. It has been concluded that e.ach of these two mutant types is due to the lack of a diffusible substance which hosts of the proper type can furnish --in the case of vermilion the vermilion plus hormone, in the case of cin~abar the cinnabar plus hormone. A technique for determining whether or not fli8S of a given mutant type produce either or both of these hormones has been d 0veloped by Beadle and Law (1938). This consists of feeding crushed pupae of the type to be tested to D. melanogaster vermilion brown and cinna- 25 bar brown larvae. If the ty pe in question is neither vermilion nor cinnabar -- by virture of producing both the v plus an:i c n plus substances -- , both the vermili.on brown an:l c,innabar brown flies will develop eyes altered from the usual pale pink color to a brownish or brown pigmentation. If the type is cinnabar, the change will be effected in the vermilion brown flies, but not in the cinnabar brown ones; if vermilion, the transformation will take p lace in neither kini of test flies. Thus, a. bright red eye mutant of arry speci e s of Drosophila can "h e identifi e d as vermilion, cinnabar, or something else on the basis of what l{inds of substances are produced in its develop ... ment. In D. algonquin the mutant of this type has been identi- fied as cinnabar on the grounds that puo ae of this chara cter --chosen at a time when eye pigmentation was just beginning to become apparent-- when fed to D. melanoe;ast~ larvae of the v bw and en bw kinds, caused v bw eyes to be altered to brown but did not effect en bw eyes. Wild type D. algonguin pupae of the same age caused brown ,Ji g mentation to a ,_J p 1ar in the eyes of both 0 kinds of test flies. That is, the D. algonguin mutant differs from wild tyne in not producing the cinnabar nlus substance, though the vermilion n lus hormone is produced. A third autosomal mutant in D. algonquin is droop. Flies of this type have wings that are s omawhat shorter and broader than normal. In addition, the t exture of the wings has a thin, shiny ap'Je arance. Usually, t h oug h not always, the wings are conc ave underneath so that the edges hang below the normal level of the wings at r est. A similar effect has been describe d for the mutant curved of D. me lanogast er (Morgan, Bridg e s, and sturt ev ant, 1925) . However, in the D. algonquin mutant the wings are not divergent and uplifted, as they are in curved. 26 The mutant brovrn causes the eyes to have a dull, darlr brownish ap")earance. The dark movable fleck: on the compound eye charac- teristic of wild type is missing, an:i the eye seems to be rather translucent. A microscopic e xamination of the ommatidia s11ow less of an agQ' regati on of the dark pigment gra nules in the dist al regions than is true of wild t yne , and one mi g ht conclude that the tra n s lucent appearance is due to the fact that light may thus enter the ommatidia laterally as well as throug~1. the dist al 'c:: nds. The ocelli of this type are color Less, and the testes, which are bri g ht orange in wild typ e , are unpigmented here. A combination of brown with the above described cinnabar causes the ey es to be of a pale yellowish color which darkens to brown as the flies mature --just as the cinnabar eye itself darlrnns with age to n ,:; arly wild type. here are also colorless. The ocelli and testes Examination of the ommatidia shows the distal regions to contain a nu mber of yellow pigment granules ·.?hich seem to be soluble in hot water. The presence of such granules is not characteristic of the bro --n mutant of D. melanogaster. (The writer is indebted to Dr. J. Schultz for this in- formation.) In addition, the D. algonguin brown mutant S8ems to be associated with a semi-lethal eff e ct, as became evident in tests mad e to determine its linlcage re lations (see ·I 'ab1es I I , I II , and IV) . Linkage relationships The lin~~age relationshi ps of the autosornal mutants of J2..!. algonguin have been studied through making matings of all the p ossible combinations of the se mutants and exam ining the F 2 generation flies for double homozygous individuals. Due to the oc- 27 currence of no crossing-over in the male in Drosophila one expects that two genes that are linked will not be found together in the homozygous conditions in the F 2 of a cross into which they entered. separately. On the basis of this expectation the '8rese·-:ce of a double homozy gote in the F2 of such a cross iS evidence that the genes concerned are not linlrnd. From this sort of evidenc e it has b ee n cone luded that the D. algonquin ~utants rough, cinnabar, and droop are not linked to each other --that is, they form thr e e separate linkage 11 groups 11 • Similarly it 1,ve.s discovered that brown is link ed to neither rough nor cinnabar. However, the F 2 generation of the cross between brown and droop contained no brown dro op individ uals (Table II) (the absence of this class is not in itself strong evidence for linkage in this case, since it might b e sup,Josed that the double ho mozy got e would be inviable, judging fr om the apparent deleterious effect of brown alone) . Since D. alp.: onguin has but four autosomes --one of them very small, and since two of these containing the linkage groups of rough and cinnabar-- were shown not to contain brown, it was thought likely that brown was linked to droop, assuming it unlikely that any of the genes 1fiere located in the v r:1 ry small chromosome. was made. Another test on this p ossibility A brown droop stock was derived fr om individuals of subsequent generations and a cross vrns made between b~own droop and wild type. females. F 1 males were then backcrossed to brovmdroop If the genes res p onsible for these characters were linked, such a backeross should yield only individuals of the ~J arental type s, no recombinations. The results of t h is experi- ment --the absence of the recombination classes (Tab le III)-confirm the suspicion that brown and droop are linked. Thus, the known D. alp;onguin mutants arrange themselves into four lin- Table II. F 2 of a cross between droon and brown. wild typ~ 122 Table III. bw dr 49 14 total = :t85 bw dr ~ X Progeny of the cross wild type 134 dr bw 0 dr bw dr 0 67 bw dr / f bw 0 .. total-== 201 Table IV. Degree of linkage of droop and brown. A. Progeny of the cross E!!/dr ~ x.. bw dr 6' bw 50 dr Tir6 bw dr 34- total -- 358 rec o:rhbi na.ti_ons = 122 % recombinations= 34.08 B. Progeny of the cross bw dr 8 7 bw dr /+t ~ X bw dr d" wild type 152 bw 33 = total 314 recombinations= 75 % recombinations= 23.09 dr 42 + cjf Table V. Degree of linkage of dusky and s wollen tarsi. A. F males of the cross .9:I 'f. x. swt (j1' 2 swt £I 78 200 dy swt 97 wild type 215 total -= 590 recombinations= 312 % recombinations = 52.88 B. F 2 males of the cross dy swt 118 wild type<:? X wild type 222 total == 595 recombinations-= 255 % recombinations= 42.86 £I swt di QI 171 swt 84 28 kage systems: (1) the sex linlrnd mutants dusky and swollen tarsi, (2) rough, (3) cinnabar, am (4) droop and brown. Degrees of linl{age between droop and brown and between dusky and swollen tarsi were determined as percentages of recombinations from both cotnling and repulsion data (tables IV and V). In both instances the values for recombination gotten by the two methods were ap,reciably different. This has been interpreted as being due to the disadvantage with r e s pect to wild type exhibited by the mutants concerned, since in each case the higher recombination value wa s gotten when one of the recombination classes was wild type. S·wollen tarsi and brown seemed to car- ry rather definite lethal effects. Allocation of the mutants to the chromosomes It has b een desire d to find out to which chromos om ,s of D. algon:Juin --as observed in the sc1.livary glarrl-- the l{nown genes sho~.ld be aS'?,igned. A method for determining this was made available through the pr e s enc e in t t1is speci e s of cytological ly observalblbe landmarks the inheritance of which could be studied --namely, the inv e rsions ·;,- ~1ereby the c l.1 r ornos .) me sequ.8 nc es differ. A st oc l{ of D. alP: onguin homozy g ous for rough, cinnabar, and droo·: was made up --each of the thr 8e autos omal linkage ti groups , , " '-~ei ng t hus represented. CytologicaLLy, the stock seemed to car r y thr ,o ughout the Stam.a.rd se quence in the A-chromos .1me and the B-3 s equence in the B-chr ,o mos ome. I:ndividuals of this stoc ~:: wer e mat 0i d to wild tyDe flies of t h e Wocxls Hole strain 7 .Sb, '.Y;.1 i~h was determined to contain the A-2 sequence in the A-chromosome and the B-1, B-2, EL1d Standard se1uences in the B-chromo Bome, but :hot the B-3 arranger:1ent. Consequent.Ly, the 29 progeny of this cross were expected to be hetero zygous for gene se -mence in the ..\- and B-chrorn osomes and to b e heterozyg::n.1.s for rnuta.nts in thr ee differe nt autosomes. Mal e s of t ~1i s generation were the n backcrossed to r ough , cinnabar, droop fem 2.les. In th e a bs e~ce of crossing-over in t he males, it would be exoe cted that the mutant t r :ie ;yco: eny of this cros s 'i'l O'~lld be homo zyg ous 0 with r :s1C-.! Ct to strain for the ent:_re chro r:;os omea carrying the 0 g e n0s exnressed. Consequently, since two of the autosome s were mar1:rnd hy L :.v e rsions, homozygosis or heterozy g osis could be deter mined f or them cytolog ic al y and the pres ence or abse :-1ce of genes in th<js,:J chromoso:tes .1 etected. Thus, for example, if such a m J.t ant i nd iv idua.1 c ocJ.ld b e f ound to b e het ero zygo us for chromo1 2 orne s equenc e in t~e A-chromosome, then it could be concluded that th e➔ gene responsil1le for the mutant was not car ~~·ied in that chromosome . Similarly, a g2n8 could b :,1 ,xcl'.Jd · d from the B-chromo- some, or from both the A- and B-c hr omosoffies. However, if a mu- tant individ u al is found to b e homozygous for sequence in the A-chror:1osome, one might c )ncL.1.d e either that the gene exp:cessed is contained in the A-chromosome or that the A-chromosome carries none of the genes re )resentsc1 in the multiple mutant stock. 1 Th e latter condition seems unlikely on t h':! basis of the small lilrnli ~1ood that any of t he t:1ree lin:rnge locat-::d to the small D-c'.J.romosome. 11 ,; ro'vlpsi; i::o to be al- Si milarly , thus, one should b o able to assi r n,:.c a g ene to the B-chrornosome, or two genes to the A- E, nd B-c hrorn oso mes toe: ether. Determinations of horn ozygo- sis or l1et e rozygosis for c i1ro mos ·J.U8 sequence in the A- and BC:J.ro::-: os ~-m: .-,s wer e made for a number of mutant backcross individuals by mating them to fli es of the wild typ (::: strain and examining the salivary glands of their offspring, het erozygosis --and c Jnsequently exclusion of the gene, or eni:Js, of the mutant t / p e ,( II II I Figure 9. ll I I I II ll I Demonstration that a g ene 11 a 11 is located in chromosome II, not in chromosomes III nor IV, of a Drosophila s pecies in which each of the autosomes, except the very small one, is assumed to contain a gene --designated oy the small letters II II a 11 , 11 b 11 , and c II , re:Jresented on the chromosomes as cross 11 nes. The first cross is made between a multiple mutant stock a..t.bi£_,and a wild type stock differing from it by inversions in chromosomes II and III --s i1own as curved broken lines. A male offspring is backcrossed to the mutant stock. From the Jro g eny of this cross an in- dividual exhibiting only gene 11 a 11 ls mated to an in- dividual of the wild type stocl{. ~:x arnination of the salivary gl a nds of the offspring show all of them to be heterozygous for sequence in I I, half heterozygous and half homozygous in III. 30 fr 1m the chromosome so d etermln:,d-- b e ing assumed if part of the larvae were hetero ::: ygous and part homozygous in e::..ther or both chromoso mes , hJmozygosis being considered likely --and consequently a tentative assig::1rnent of the gene or genes to the chromosome or chromos ome s so determirrd-- if seven or more of the larvae s '.1owed heterozy g osis for chromosome se que nce (the chance that a heterozygous individual would have s c:: ven offspring of one k:..nd in a backcros r➔ being 1 in 128) . . An outline of the method is g iven in figure 9. Conclusions as to the allocation of the genes of D. algon.'.J.~in could be drawn from two mutant i nd ividu a ls. male ~JrOV ': d One rough droop to be hJmozygous fol" both the A- and B-c:1romos ,mes. Thus, it s eemed likely that rou :~ h and droop were located in the A- and B-chromosornes. A droo- J male showed hL~self to be h orno- z:nr,ous for the A-chromos )roe , heterozygous for the B. Cons e quently, droo'J seemed to be cont e_ ined in the A-chromosome, not in the B, and frJm the constitution of the preceding in:3.ividual, rough may be cone bded to b e carried in the B-chromosome. 3y exclu- sion, cinnabar should be assigned to the C-c !'.lrom osome. In con- firm a tion of thes e conclusions, a cinnabar female was founi heterozygous for A and B, thus excluding cirn1abar from A and B; a rough c i nnabnr female was det ,~ rmi n ~,d to b e het 0r o zygous for the A-c hrn iTIOs ome, thus excludinc: both rou 3h :: nd cinnabar from A; a droop fem ale w,:;.s found heterozygous for B, thus exch.1ding droo l from B. I n summary, the followine-: gene-chromosome r e lations se ~Jm to e xist in D. algonguin: A-chromosome --a:llroop and brown, B-c t-1 r om os ome --roug h, C-chromosome --cinna:Jar (excludi ng , of course, th'3 7 ossibility that o ne of these 11 groupsn should b 2 assigned to D. The mutant cinnabar, as described above, is one that can 3l be identifie d in other S"'.Jecies. Conse quently, fr :xn the alloca- tion of this mutant to the C-chr omosome of D. algon .J uin it can be concluded t h at the C-c hr omos )me of this s ~eci es is at least in ,art h omolo g ous to c h romos ome III of D. pseudoobscura and to the ri ght limb of chr ornosome II of D. melanogaster, since a gene with the prope rties of cinnabar has b een located in these "?OSitions in these species (orang e of D. pseudoobscura has been deter mined to be cinn2,bar through the transplantation experiments of Tan and -P oulson -1937-). 32 I NT SRS ?:?CIFIC HYBRIDS The algon1 u in/atha basca hybrid Cert a in matings betwe en Drosophila algonguin females and D. athabasca males have been found to yield pro;· eny (unpublished data of Dr. A.H. Sturt evant and results of the writer). offs ·•)ring, t hO'. l.2_'. h quite normal 1_ n These general ap p e ar s. nee, hav e seemed suffici e:1tLy different fr '.rn D. algonouin to ra.ise the suspicion th2.t they we re h:?bt'ids and not the result of non-virginity. The su:>'J osition that hybrids could be got · en between t riese t wo s oecies h a s 'be <:> n confi r me d in t ~1at a cross between rough D. algo2:3uin femal e s and D . athao asca (Grays River, Wyoming) males .:--ave offs .,,ring th::.t were wild ty '.)e 7iith respect to eye char2.cter. 0 Continu e d interspecific matings have shown that hybridization betwe cm D. algon:-:r..1.in an:!. D. athabasca is rather difficult to obtain. Indeed, th e cross athabasca fe mal es times alE,onguin males has failed so far to y ield hybrids at all. An investiga- tion of intersnecific insemination frequency has been made. In this exDeriment it has be c,• n the practice to mate at a time about ten females of one species to as many males of the other. These flies were then kept tog ether for ten or eleven days. After t his time the females were disse cted and whether or not arzy of them had been inseminated (:0tsrmine :'.'. by observing microscopical 1.y the presence or absence of sperms in the seminal reCe1Jtacles. As a check on "r_)Ossible non-virginity only mutant tyne females have be 8 n used in t hese crosses, the mutants being ones that have been determined to be recessive in hybrids as well as in the pure s ·Jecies. Insemination data from matings yielding matroclinous mutant offsDring have been discarded. A summary of interspecific insemination data d erive d from crosses Table VI. Interspecific insemination, between D. algonguin and n. athabasca. D. algon1u1n9,. D. athabascad'> D. athabascag;<. D. algonquin~ droop X vermilion 3/53 vermilion X droop O/l.o4- droop~ cinnabar* 0/54 vermilion x cinn. 0/36 cinn. 0/25 vermilion~ Aldrich 0/19 ~ vermilion cinn. X Sundance, Wy. 0/12 0/159 3/144 ::- 2.0§% * mahican 33 bet ween algoncmin and athabasca is presented in Table VI. Ap- uarent ly not all cases in which insemination occurs yield progeny. None of the three instances of insemination of droop fe- males by vermilion males was accompanied '-1y the occurrence of hybrids. The nature of this failure of inseminated females to produce offspring is not known, whether due to inhibition of egg-la;y ing or to high mortality of hybrids at early stages or to some other cause. A case in which a much higher frequency of intersnecific insemination is attended with low hybrid production is thatinvolving D. affinis females and D. atl1.abasca males. This has been irrv e stigated in greater detail, as des- cribed below. Algonguin/athabasca hybrids have be e n derived principally from a series of grou D matings. In each of these crosses five D. algow1uin females were ke -ot with five D. athabasca males for about a month, and longer than this if hybrids Viere being produced. The D. algonquin females contained various mutants and combinations of mutants, and a number of strains of D. athabas.£.§! were employed. Out of 342 such crosses 13 yielded hybrids. A record of the se succ <:1 Ssful mating s is given in Table VII (the crosses to which cross-numbers have be e n given). In the case of each of the D. algonguin mutants --cinnabar, droop, rough, and dusky, and the D. athabasca mutant vermilion the character was s :.1',pressed completely in the hybrids, thus indicating the presenc e of the wild ty pe al · elomorphs of th e se genes in both suecies. The algonguin/athabasca hybrids appear to be quite viable and robust. However, the size of families gotten from interspeci- fic crosses is considerably smaller than those generally gotten within D. algonquin, even from pa ir matings. Neither sex seems Table VII. hybrids. D. algonguin~ x D. athabasca cf. cross number Crosses . y1eld4,ng males :eroge~ females total 4 3 7 Aldrich x Sundance 0 1 1 cinn. X vermilion 1 0 1 cinn. 3 4 7 cross rough .,. Grays River 71.1 l( vermilion cinn. X Grays River 71.1 # 6 41 36 77 cinn. X Grays River 71.1 # 16 21 27 48 cinn. X Grays River 71.2 # 13 32 33 65, cinn. 'X vermilion # 4 1 0 1 cinn. vermilion 'II 20 3 8 11 cinn. X vermilion II 21 14 9 23 Grays River 71.2 # 10 16 8 24 droop X Grays River 71.2 # 23 13 11 24 droop ~ ,; rough )( Grays River 71.1 # 12 6 34 40 rough X G-rays River 71.1 # 1-05 0 ]._ 1 rough X Grays River 71.2 # 19 18 32 50 rough X Gravina # 71 1 1 2 dusky X. Gravina # 141 4 9 13 178 217 385 34 to have any obvious abnormalities. The hybrid totals g iven in Table VII show that there have been gotten altog ether f ewer males than females --17 3 to 217. However, this aberration of the sex ratio seems to depend on the data derived from two of the crosses involving rou gh D. algonguin females --Table VII, crosses #12 and #19, from which th e re were gotten a total of 24 males and 67 females. Dis r egard ing the se figur e s, the hybrid totals are 154 males and 150 femal e s. It is possible that the disparity of male and female hybrids derived from the D. algon~ rough fe ma les is due to i 11 otherwi s e u ndetected sex-linked lethal for which each of the-s e fe males wa s het e rozygous. In general a'Jnear a nce the algonguinjathabasca hybrids are very similar to D. algonquin, though in a few res pec ts they show the influence of D. athabasca. D. athabasca is a considerably darlrnr form than D. algonguin and hybrids between these two s p eci es have a tendency to darken with a ge more comp l etely than algonquin, thou gh -probably not a s much so a s athabasca. In athabasca the wings ar e slightly longer and nRrrower relative to the body size than in algonquin, arrl a tendency in this direction has been observed in hybrids. D. athabasca ap :Jears to differ cl e arly from D. algonguin in the size of the s e x comb. D. athabasca mal e s have small sex combs usual Ly consisting of four teeth, while the D. algonquin sex n ombs are quite large, consisting of from six to ten teeth. ( Ac r: ording to Sturtev a nt arrl Dobzhansky -19 36b- D. athabasca has four sex comb teeth, D. algonguin from eight to ten. The writ e r h ;3. s observed a f ew D. atha basca sex combs with but thr8e teeth -Gravina- o. nd some with five -vermilion-, an:i counts made on D. algonquin males have shown the :: e to be in:iividuals with 0 six arrl seven teeth per sex comb -cinnabar (figure lOA) and roug h-). A D 4 5 7 6 7 8 I4 • 3 7 8 6 4.2 E B V a 4 5 6 7 18 I ~ 4 8 9 I en S3 4 C 3S 6 5 5 I a 6 3S I 14 31 + a I ~ 4 5 6 7 8 3 4 5 6 Figure 10. 7 8 9 Histograms representing sex comb teeth counts in D. algonguin, D. athabasca, hybrids, and males of the backcros s of hybrid females to D. algonquin. The abscissa classes represent rumbers of teeth in the sex C:omb on the right prothoracic leg. The number at the top of each column is the number of in:iividuals in the class. A. The histogram at the left applies to atha- basca vermilion males, the one on the right to algonquin cinnabar males. B. Hybrids involving the ~lgon- _guin _2E strain. C. Backcross males, the algonguin strain being c n. D. Part of the males of 11 0" derived from hybrids containing athabasca y. E. The types constituting the backcross males of 11 D11 • 35 Some counts of sex comb teeth are presented as histograms in figure 10. Figure lOA shows the existence in the athabasca ver- milion strain of males having four or five teeth per sex comb (right prothoracic leg) and of individuals varying in sex comb teeth number from 6 through 9 in the algonguin cinnabar strain. Figure lOB s':1 o''!S sex c Jmb teeth number to vary from 5 through 8 in a number of hybrid males in which the D. algono~ was of the cinnabar strain. parent Ac coi"ding to Castle et al ( 1906) the number of teeth in the sex comb in Drosophila melanogaster is quite vari a ble --ranging from 7 to 14--, depending to a considerable extent on the culture conditions. That t.here are ge.ne- tic factors concerned in the variability has been shown by Hoge ( 1915) who succeeded in isolating lines of D. melanogaster with high and low mean sex comb teeth numbers. It is likely that in D. algonquin am in D. athabasca the number of teeth of the sex comb deDends in its variability on factors of the sort that operate in D. :nelanogaster, although a study of this matter has not been made in these s-8ecies. In view of this possibility, it is difficult at ·Jret:tent to evaluate the significance of the differenc~ in sex comb teeth number between D. algonguin and the algonquin/athabasca hybrids. The algonquin/athabasca males are quite sterile. ~ An exami- nation of hybrid testes has shov,,rn them tobe somewhat smaller than those of either species, though the spiral shape common to the testes of these species is retained as much as possible. The rest of the genita Lia seem to be normal. An investigation of hybrid testis structure has been made through a study of acetocarmine smears. S 1ermatogenesis seems to be L ,paired to the extent that no spermatozoa are produced. Scattered through the smears there were seen grouns of rather small cells with large A I .-- ~t B ,,," l >t I C f 0 I 10 20 I I MICRA Figure 11. e,-. ' llt ~ D Structures found in the testes of hybrids. A. Spermatogonia in algonguin/athabasca testes. one at the top has two nucleoli. in algonquin/athabasca testes. an affinis/athabasca testis. B. C. 11 11 The s permatocytes" Spermatocyte 11 in D. Part of a giant syn- ci tial mass found in the affinis/athabasca testes. This body measured about 100 micra across and contained 37 11 nuclei" . 36 nuclei, some of which contained two nucleoli. These were inttlr- "Jreted to be the spermatogonia (figure 11A). In addition, the testes contained grouY.JS of considerably larger cells, of the order of siz0 of the Drirnary sperrnatocytes of normal testes. These cells contained a variable nJ.mber of irregular, darkly staining masses, usually quite vacuolated ( figure llB). It seems likely that these cells are spermatocytes in which the normal maturation divisions have been obstructed and that the dark masses are chromos om,::s or groups of chromosomes which have formed pyknotic pseudo-nuclei. Dobzhansky ( 193L~) has described as the :fbnal st age in spermatogenes is i "l Dros oohila -os eudoobscura race hybrids of the weak B female times weak A male cross the formation of irregular clumps of chromosomes within the spermatocyte body. The algonquin/athabasca females are fertile and have been backcrossed successfully to D. algonquin males. These matings --made as pair matings, each lasting about a month-- have succeeded in somewhat less than half the attempts --46 out of 128. Some attempts at backcrossing hybrid females to D. athabasca males have been made, but none has been succ c:ssful so far. In Table VIII are presented the results of backcrosses of algongui!1/' athabasca females to D. algonquin males. In each case the back- cross was made to the algonguin mutant strain involved in the interspeciflc cross; consequently the progeny were both mutant and wild type. In cross #21 the athabasca sex-linked mutant vermilion was used, so among the backcros2 offspring there were some vermilion males. In this instance it was relatively easy to distinguish vermilion a~1d cinc1.abar individuals since vermilion was much more extreme in s xpression than cinc1abar. The number of progeny per backcross was a':)·Jreciably less than the usual number gotten from interspec ific matings, though, of course, Table VIII. Progeny of algonquin/athabasca females backcrossed to D. algonnuin males. number cross f~ C n/ G. R . 71 . 1 pro~ males females ~-mut. mut. ± totals en 11 521 38:!: 54_g 50 194 en "' .£E/G .R. 71.2 3 1.-1: 7± 9 2 19 201 22,g 26 23.d 91 1 1 3 //6 )t. ,#16 X e,n 10 dr/G. R .71.2 I/lo X dr 1 ro/G.R.71.1 #12 X ro 2 1 ro/G.R.71.2 #19 X !.Q 16 4r;J._ ~'13 //21 X totals 2 32 120 14 1? 99,I totals ~n/y l? en 3 ±. V en 27.l 24! 25 612 3 554 194, --1532 1311 504 471 47 170 males females 29s2 378 13 6742.§ 37 since the interspecific crosses 77 ere made as group matings, it was never certain wh ether mor e than one f em ale wa s participating in a single cross. The mean f a mily sL: e in interspecific mat- ings was 25.1, that gotten in backcrosses 14.4. There was a general tendency towards the production of fewer backcross males than females. Lancefield ( 1929) observed deviations from the 1:1 sex ratio ih backcross ~rogenies of the D. pseudoobscura race hybrids. In addition, there were a number of backcross individuals, both males arrl females, that were definit e ly abnorma L in general appearance --having such abnormalities as rough eyes, missing or reduced bristles, wings held out or up, abnormal abdomen, though none of the se has b ee n found to occur in the hybrids themselves. Some of the abnormal mal 0 s wer e f ound to have small misshapen testes as well. An estimate of the fre- que ncy of these aberrant irdividuals has been included in Table VIII (the n'J.mbers written as underlined exponents being the numbers of abnormal individuals found in the various class e s). Presumably these abnormalities occur as the result of a state, or states of "unbalanc e " accomDanying certain combinations ofnhe alizon '.l e,l in and a t habas ca chromos ornes. How ever, in these cross e s the chromosomes were not suff ic lent ly marked to make it possible to det ermine the natur e of the chromosomal conditions on vl.nich the ::,eculiaritie s d =: p e nde<ih. Among backcross males ther e was even greater variation in the n·1mber of sex comb t eath than was found in hybrids. A few baclccros ~, males had sex combs consisting of o :1ly four t e eth, and i ndividuals wit h 5 or 6 t ef.: th seemed than among the hybrids (f i g. lOC). 11 8lativ e ly mor e frequent In the cros s involving atha- ba~ vermilion, where it was oo s si1.1le to disting uish in the backcross t h ose males bearing the athabasca X-chromosome, there 38 was a marked tendency for backcross males with the athabasca X to have sex c ombs with few teeth (fig. lOE), and the total sex comb teeth frequency dist :i: ' ibution among baclrnross males of this combination seemed to be so:~ ewhat bimodal )n this account (fig. 10D) . It is sugg est ed, thus, that the D. athabasca X-c hr .)IDO- s orn'j has an influence in reducing the number of sex comb teeth in backcross males, thus causing t :1ere to ap ~) ear in the backcross (to al von::iuin) ir:dividuals with smaller sex combs than were found among the hybrids --where the rnal ,:·s carried the algo nguin X. An inspection of Table VIII sho -,- s that in the case where such individ:.:wls could be detec-ged, males carrying the athabasca Xchrornosome were less frequent than t ':~ ose having the algonguin X --24 to 52 ( cn/y #21 tim -:.: s en). It would seem from this that the D. athabasca X-chromosorne has a d e leterious effect in combination with algonguin cytoplasm and chromosomes. However, a Dossible unfavorable infL1.enco of the vermilion g ene in t i1is case must not be discounted. Dobzhansl{y ( 1937) has shown that certain D. nseudoobscura mutants of ordinarily g ood viability ha ,_re semi- ;_ethal eff '. : : cts in backcross individuals of race hybrids. It may '!rell be that reduction of sex comb teeth numb2;r among c e rtain backcross mal e s is ,yrnptomatic of general decrease of viability. At any rate, 11 of ",,.,hysical condition a better l{nowledge of the influence on sex c omb t e eth nu~Dber needs to be gotten before the genetic i ~~ ortanc 0 of data auch as these can be evaluat(;d properly. The affinis/athabasca hybrid As wa s the case b etween D. algonquin and D. athabasca a number of crosses bet ween wi l.d ty p e D. affinis females and D. athabasca m les yielded offspring. To investigate the possibi- Table IX. Hybrids derived from the cross D. affinis <=?~ X D. athabasca rr' d' c·ross progecy females totals 2 3 Sundance, Wy. 1 1 pinkish, X Sundance, ~Vy. tiny bristle 1 1 affinis males athabasca cut veinlet X Sundance, Vvy. white X rue; ose X vermilion net ")( Sundance, rugose X vermilion 1 ct vnlt,~ 1 1 Y{y . 1 gy nandromorph 1 1 2 5 9 1 3 1 39 lity that this combination oc casionally forms hybrids further int e rs ·oe cific matinrs were made u s ing mutant D. affinis females. Thes -~ cross es have given wild type offspring --though in very small rumbers-- indicating the existence of aff'inis/athabasca hybrids (Table IX). An inv e sti g ation of inters pe cific insemination between D. The me t h od employed affinis and D. athabasca has 1:Je :; n made . in this expe riment was the same as that described above for the alr:ronTJ.in/athabasca inters-oe cific insemination study. sults a r r) ryresented in Table X . The :re- In view of the rar ; ne ss of species 0 hybrid s in this instance it is surprising that illll.terspe c ific ins emination was found to b e so frequent, at any r ate in the cross D. affinis females ti mes D. athabasca mal es. A further discussion of this -ohe norrw :'lon is J resent ed be.Low in the next section. Hybrids have not 1Je1:1 n gotten at all from the cross D. athabasca f emales ti mes D. affinis ~ales. The affinis/athab a sca hy brids seem to be ~: uite viabl e , although they h ave been obtained only in small nun.1 bers. These hybrids ap ,, e ar to be more darkly ')igment •::d than D. affinis, which 0 is a li ght, e r for m than D. atha basca. monstrat ed to b e fertile. Neith9r sex has been de- The ovari e s of the femal es , how0ver, a.p--c,ear to be 'l Uite norma l, a nd eggs are deposit ed . An examina- tion of the t e stes of one of the males has been ma.de in an acetocarmine smear. No sperm atozoa were: found. li ~rn eel ls were seen. A f ew spermatogonia- In one re g ion of th e smear there were found a f ew large c s lls of t h e size of s pe rmatocytes. In one of the m thsre occurred an irre gular array of r a the r distinct chromosomes, a 7 ar e ntly mor ,,; than the .di ~)l oid numb e r, t n ough it was n.ot nossible to count them ac cur at e ly (fi g . llC). However, the most striking feature of t ~1.ese testes vrn.s the p r es enc e of a num- T'able X. I nterspecific insemination, between D. affinis an:l D. athabasca. - D. affinisi 1' D. athabasca cft D. athabasca q. X D. affinisd' ,,. vermilion 56/B2 vermilion X rugose 0/102 rugose 'I. cinnabar* 34/54 c·innaba.r-x- x rugose 0/24 rugose 8/10 vermilion X Vict.B, Texas 1/15 98/146 vermilion X Gtlnbg., 11/19 Tenn. rui;i:,_, ose ')( Sund., Wy. -== 67.12% vermilion X Balt, .B, 2/19 Md. 14/l.79 * mahican 40 b 8r of very large round bod i e s co nt c1 i ning ir r egul a r, vacuolated chr omatin ma s se s qu i t e similar to those found in the cytes 11 11 s permato- of the al!tonqu in/athabasca hybrids, as described above. Most of th 9 s e g iant str ,1ct ,1 re s wer e) ap , arently broken in t he -proc e ss of sm e aring a nd the 11 nuclei 11 from th em scattered a bout, but one s uch body wa s found more or Less int a ct (fig. llD). T h is was found to measur e a bout 100 rnicra in diameter and to cont a in 37 c hromatin ma s s es. P rof. Th. Dobzhansky informs me that s imi l a r g i ant struct u r es are to be fou nd in the t e st es of at '.'l ab asca / a zteca hybrids. This p he n ome n::m might b e comp ared as we ll to t h e abnorma l mu lti , li c at i on of nuc l e i ta l{ing place in the 11 s nermatoc ytes 11 of t h e D. p s eudo ob sc ur a rac e hybrids of the c ros s strong B fe ma l e s ti m :s s tron? A rr: a l 8s (Dobzhansky, 00 1934) . The data of Table X show th a t insemi n ation takes place be- twe e n Drosophila affinis and D. athabasca in both directions. Betw e en af f inis rugose fe males and athabasca males mating was rather frequent in t h e t e n day p ericd allo we d --taking place in more tha n two-t hirds of the cases. Yet hybrids betwe en these species have b een but rare l y p rod u ced , and then in only s mall numbers (Table IX). Seve:cal e xplanations of this a v :1 arerrt fail- ure of inseminated femal e s to produce of f s p ring may be offer e d: ( 1) ege: -lay ing of thes e f e:-ne.l es is for some r e ason inhibited, ( 2) mort a lity of hybrids is very hi gh at e arly st ag es --in t h e emb ryo nic p eriod or in e ar ly larvae, (3) s p ermatozoa of one s p ecies are inactivated to the ext e nt t h at they u sually cannot effe ct ferti l ization of the eg 6 s of the other sp e cies. The latter al- ternative may b e divid e d into two parts: (a) i nactivation of sperms takes place in the g~nital tract of the female --after they arrive in the seminal receptacles-- or (b) the sperms are able to enter the eggs but inactivation generally takes place soon after penetration so t ~~,at embryonic development isn't initiated. A rr;J.mber of D. affinis females that had been kept with !2...!_ athabasca males for ten days were isolated individually into vis.ls containing fly food. After three days these females Y:, ere dissected and it was determined vvhich of them had b2en inseminated. From the food of females found to be inseminated it was possible to collect a number of eggs --130 altogether. of the eggs had hatched. and observed again. None The eggs were kept two days longer Still none of the egg s had hatched, and microscopic examination of some of them after the criorion had been removed showed disintegration to be taking place --aggregation of the yolk into large globular masses and concentratior(or the opaque material of the egg towards the center, leaving the ends relatively transparent. It was concluded from this that interspecifically inseminated females of this sort do lay eggs, but that the eggs in general do not hatch. It was next desired to know the extent of embryonic deve1opment in these eggs. After a suggestion of Dr. D. F. Poulson it was decided to stain eggs of interspecifically inseminated females with fuchsin-sulfurous acid according to the Feulgen technique and examine them whole for embryonic structures. D. affi nis rugose females that had been kept with D. athabasca vermilion males for ten days were isolated and allovved to lay eggs over a three hour period at 25 degrees. At the end of this time eggs laid by females determined to have been inseminated were fixed in formol-alcohol-acetic acid ( 5: 15: 1) , the chori ons hav- 42 ing been removed and t he vit ell ine membranes p unctured with a very fine ne edle after the manner of Poulson ( 1937). As a con- trol eggs laid by D. affinis rug ose females known to hav e been inseminat ed by males of their own strain we re prepared in the same way . According to a schedule suggested by Dr. Poulson both kinds of eg :;:-·s were stained and mounted whole for exa mination. The eggs laid by the affi~ fe ma les inseminated by males of their own species were found to contain many nuclei. In most of these the blastoderm stage had be e n re : ;_ ched --that is, in these cases the egg surf ace was lined by many nuclei between which cell membranes had formed or ,vere forming, and in the interior a number of nuclei were suspended in the y olk in protoplasmic islands, as described in the case of the D. melanogaster embryo by Poulson ( 1937). In the eggs of the interspecifically insemi- nated females, however, no nuclei were seen at all. Although eggs obviously SDoiling had been rejected b e for e fixation, some of these eggs seemed to show the beginnings of dis i ntegration. Moreover, barring disintegration, it should have been possible to fini in th ese eggs at least the oocyte nucleus, even though fertilization might not have taken ')lac e . It was concluded that thes e eggs had probably b ee n laid to wards the beginning of the egg laying period and, since development in them had proceed e d either not at all or only a little, they had had ti cc,e to s p oil. ~ggs laid by D. affinis rugose females inseminated by 12...! athabasca vermilion males were next coll ec t e d ani fixe d within an hour after lay ing. Since th s- r e seemed to be the p ossibility that these eggs were beir.g laid without f e rtilization, it was planned to pay ".)articular att ention to the region in which maturation is SU:J, osed to take p l a ce in Drosop hila. Buettner (1924) has described maturation in the Drosonoil 9._~f!:.t};og~ter egg . 43 According to him the oocyte nucleus comes to lie in a relatively clear protonlasmic region at the dorsal surface of the egg c.J.bout a third of the length of the eg g from t h e micropylar end. In eggs ready for fertilization this nucleus has advanced to the first meta~hase of meiosis, and presumably the remainder of the maturation Jrocess does not take place unless stimulated to do so by the act of fertilization. Thus, unfertilized Drosophila eggs might be expected to contain the first maturation spindle with the chromosomes at metaphase and no other ruclear structures. J.n ex amination of the eg gs of the int ,Jrspecifically inseminated D. affinis females showed there to be a number of kinds of confi g urations in the region of matu ·~·ation, though no sperms nor evidenc e of cleavage wer e seen in thes e eg g s. Although a spindle with chromosomes mo:e e or less at metaphase was seen in one instanc e , there were found case s in which the r e were two, and even three, nuclei or groups of chromosomes. The number of chromo- somes in these grou1Js could not be determined with certainty,, but it seemed to be variable. It ap ,Jeared from this that the eggs of interspecifically inseminated females were not fertilized, but that some:tizow maturation had Droceeded bey ond the first rnetauhase. It was realiz e d, however, that sperm heads, being very small, could easily have be 8n ov e rlooked in thes 9 eggs. A number of eggs laid by D. affinis rugos e vir g in females were gotten during the first hour after laying, and preparations of these eggs were examined. ·.·11.at ap'Jeared to be the same sorts of phenoms, na found in the 11 inters9ecific 11 eggs described above were observed here --irregular groups of chromosomes and nuclei. In some instances t wo spindles i"-i8re observed --in one of these one of the spindles was tripolar. It was conclud 9d that in D. affinis eggs maturation proceeds without the influence of ferti- 44 lization, though probably attended with some abnormalities. On the basis of these ob ,.ervations eggs laid by D. affinis femal e s inseminat ": .' d by D. athabasca males have seemed to be not essentially different from eggs 1-cnown to be unfe rtilized. In fi g ure 12A is presented a dr :::wing of the region of maturation in an eg_,,. laid by a virgin D. affinis female, showing two spindles. In 12B is shown a comparable r egion in an egg of an ath~- basca inseminat ed affinis female, containing three ireegular bodies. The orientation of these struct ur es in these eggs has prohably been altered somewhat in the process of mounting the Of the expla nations offer ed above for the phenomenon of faih1r 2 of interspecific inseminatiJn to result in offspring the third se ems to be the most reasonable --that the D. athabasca spe rms are inactivated. Some evidence se e ms to favor the sugg es tion t hat the sperms are inactivated in the seminal receptacl ·::s of the D. affinis fe male s rather th a n within the eggs. Observ ations of D. athabasca sperms in D. affinis seminal receptacles have sug r: ested a sort of in:1.ibition of these sperms. In determining ·whether or not a Drosophila f emale rias be e n inseminat e d the vagina and associated structur,c: s are dissect 8d out in fly Ringer's S:::>lution and the p ::·e~aration examined mlcrosc O':l ically as quickly as -:)OS si 1J l e , the -o res ence or absenc e of S':lerms in the rece-otacles being observ ed directly. If the in- semi nation has bee n intra.specific, the ventral receptacle is usually found to cont a in a rn-1mber of rapidly moving sperms, althou gh the mobility of these sperms is subject to some variation. In cas e s where D. affinis f emal (:S have bee n inseminat ed by£.:_ athabasca males the i m~ ressi:::>n has be 8n g otten that there is less activity of sperms in the receptac Le t ha n is charact c:;1 ristic 0 0 ooO 00 o Oo 0 ooioooo 0 o" 0 0 0 0 oo 0 0 0 00 oO A B 0o Oo 0 0 0 Oo oooo ~ 0 0 oO Figure 12. Structures found in D. affinis eggs stained with fuchsin-sulfurous acid (whole mounts). A. Egg denosited by a virgin femal e , showing two spindles in the region of maturation. B. Egg laid by a female inseminated by a D. athabasca male, showing three da rkly staining bodies in the maturation region, probably the pro:iucts of meiotic divisions. The small figures show the -oositiJns of the objects with respect to the whole eggs. whe n the ins e mi ~1ation has been intraspecific. snerms ap ,e ar to -:.e ,_p ite motionless. Occasionally the That the inactivation of D. athabasca sperm in the g e nita l tract of D. afrinis females is not always com-:_:ilete wou ld have to be concec&ed on the basis of the f B.ct that some i nt e rs pe cific hybri.ds h av e been p roduced. 'l'he i nJ1ibition of spermat ozoa of one species in the g 0nital tract of another s pe ci e s h as b e en r ep ort ed in birds and in mammals,. al tho Jgh not, so far as the writer is a ware, in insects. 1 Sere- brovsky (1935) has report ed the inactivation of sperm of the domestic fowl a nd of t he go ose in the ge nit 8.1 tr act of the duck. Yochem (1929) observ e d that L1 t he c a se of the rat/gu inea pig combinatio n the inactivation and removal of sperms wa s much i:·. o r e rapid after an insemination t hat ·;rns int e::. spe cif:i.c than after 0 an i ntraspecific one. Dobzhansky (1937) has discussed t he vari01..1s me ans b y which the free int e rbreed ing of separa te forms in nat ure ap •ears to be p J: ev ,~nt e d. To thes ,>; phenome na has b 8 0: n ap ) li s d the g8 ner al term "isolating mechanisms". Ex amples of such me chanisms are: geographical isolation and ecolo,~ ical isol a tion, in either of which circumstances hy bridization b e t wee n t wo forms is prevented due to the fact that ind ivid u als of the diffe rent t :ind s do not meet; s exual isolati on, r1hi ch depends on the "dis inc Linatiod' of individuals of separate kinds to interbreed; mechanical isolation, wt1ich consists of a block to inters pe cific ma.ting due to i nc ong ru ence of the g e nita1ia of the two species; inviabi.Lit y of hybrids; and hybrid steri l ity. Betwe en D. affinis an:1 D . at ha bas c a s exual isol a tion i s incomplete; e s pe ci al ly be:twe en affinis f emales and at habasca m1:des matins has been found to talrn ;lace with an 8.JJ 'Jre ci abl e f : .~ e qcwncy. tion --the -o rod u ction of adult offs pring However, hybridizafr o□ this cross-- takes 4-6 place lmt rarely under these conditions. The reason for this nhenomenan s -9ems to lie in an inactivation of D. athabasca sperms to the extent that fertilization of D. affinis eggs is generally not effected. Thus, in the case of these two species a further isolating mech : :,. nism --w~1. ich might be ter med 11 physiolo g ical 11 - - a·:;-,o,ars to O'Jerate to prevent the prod.uction of hybrids even when inters ·>Je cific i n semination becomes ) ossible. 47 S UMNJ.A.c"tY A study ha.s be e n made of the structure and variation of the salivary gland chromosom2s of Drosophila algonguin. criptions and drawings of these chromosoEJes are includ ,~,d. DesThe chromosomes of this soecies are compared with those of other S"'.)ecies of Drosop~ila. A direct comparison of the chromosomes of D. algonquin with those of D. athabasca has been possible in hybrids. Descri"Qtions are e; ivsn of the mutants found in D, algonguin, Association of these mutants into linkage groups has been determined, and a tent ative allocation of these linkage groups to the chrcmos ornes has be9n made. An achount of the hybrid between D. algonquin and D. athabasca is presented. Hybrid males are sterile, and a brief note on spermatog e nesis in these ' males is includ e d. Hybrid females are f c_,rtilfJ and have b.::J en backcrossed successfully to D. algon.9.uin. An account of hybridization betw ':oen D. affinis and D. athabasca is 5iven, including notas on spermatogenesis in t'.1.is hybrid. An investigation has been made of the failure to pro- duce offspring of D. affinis females inseminated by D. athabasca males. Since the eggs laid by t 1ese fe ,;: ales seem not to be fer- tilized, the reason for lack of hybridization has been attributed to inactivation of s -oerms, c onsti tu.ting a is ol,:1.tion mechanism between th9Se soecies. 11 physi ological" 48 BIB.LI 03-RA PHY Bau e r, H., 1936 Structur e a nd a rr :.:ng8!re nt of s a,liv a ry g land c h r oni osomes i n Drosophila spec i es . :?1"oc. Nat . .;cad . Sc i. 22 - - - and Th. Dobzhans lcy , 1937 A c ompa rison of t he ge ne arrangements i n Drosouhila azt e ca and D . at ha.bas ca. Genetics 22 (ab stract) Beadle, G. ":T . a nd B . Ephru ssi, 1936 The differenti ati on of eye J i gme nts in Drosophila as studied by trans Jl antation. Ge netics 21 --- and L . W. Law, 1938 Influenc e on eye color of fe e di ng diff us ible s u bstances to Drosophil a rnel anoga.ster. Proc. Soc. Ex pe r. Biol. & Med. Brid ges, C. B., 1935Salivary chromosome maps. Jo ur n. Hered. 26 --- , 1937 The vapor method of changing reage nts and of dehydr a tion. Stain Techn. 12 Cat cheside, D. G., 1938 The effect of X-ray dos ag e u p on the frequency of induc e d structural c ha n g es in t he c hr-omos ome s of Drosophila melanogaster. Jou nr. Ge net. l§. Dobzhans ky, Th., 1934 Studies on h y brid sterility I. Spermat oc1; -2nesis in pure an:l hybrid Dros ovhil a nseudoobscura. Ze its. f. Zellforsch. u. mikr. Anat. 21 1937a Genetic natur e of species differences. Amer. Nat. 71 , 1937b Genetics and the origin of species. Columbia Univ. Press. New York. 1939 :Ixperiment a l ~tudies on genetics of fr e e-living p o p ulations of Drosophila. Biol. Rev. lL~ --- and D. Sokolov, 1939 Structure and variation of t he chro mosomes in Droso:ohila azt e ca. Jou rn. Hered. 30 --- and A. H. Sturtevant, 1938 Inv e rsions in t he chro mos omes of Drosophila pseudoobscura. Genetics 11 --- and C. C. Tan, 1936 Studies on hybrid st e rility III. A. comparison of the ge ne arrangeme nt in two speci e s Drosophila nseudoobscura and Dros onhil a miranda. Ze i ~ f. indukt. Abst. u. Vererb. 7~ Horton, Ira H., 1939 A comparison o f the salivary g la nd chro mosome s of Drosophila me la :1ogast e r and D. sirnulans. Genetics 24 Huettner, A . F., 1921~ Maturation and f ertilizati on in Dros o:J;:1il a melanog aster. Journ. Mor ph. & ?hysiol. 39 R . D~ 1q,9 An analys is of the c hro~osome s of t he t ~o suos :)eC.1. e s JJroso-on1I a vi riiis vlrilis and Droso0hi.La virilis americana . Genetics 2Lr Kau fm a nn, B . P ., 1936 .A t erminal i nv e r s ion in Droso!)hilo. ananas s ae . Proc. Nat. Acad . Sc i. 22 1938 :-Jucleolu.s-organizing rsg i ons in salivary g land chrorn os omes of Droso ohi la melano aast e r . Zei t s . f . Zellf ors ch . u . mil~r . Ana t . 28 1 1939 Distribution of i nduced brea :rn al o ng the X-c ~1romosome of Droso-_o hila melanogas t e r. Proc. Nat. Acad. Sc i. 25 ~--- -- K9rkis , J., 1936 Ch ro mos .Jme co n jug at i on i n hybrlds betwee nJ2E..9S0Dh i l a melanogas t er a nd Dros 0°J l1.i la s i mulans . Amer . Nat. 70 Ki kkawa , H ., 1938 Studies on the gerE:tics and C'.[ t ology of Drosophi la ananas sae. Genetic a 2Q Koller, P . C., 1936 Structural hy bridity i n Dro s oDhi LE. ~;seudoobscura. Journ. Genet . 2g .:.., enc o fiold , D. t: ., 1929 A genetic study of t wo r a ces or p~ ~iolog ical species in Droso phil a obscura. Ze its. f. induzt. Abst. u. Vererb. 52 Me tz, C. 'N., 1935 Structure of the salivary gland chro::1 osomes in S ciara. Journ. Hered . 26 Ma cKnight, R . H., 1939 The sex-determinir1g me chanism of Drosophila miranda. Genetics 24 ILil le r, D. D., 1939 St ruct u re and v ari ation in the chromosomes of Droso~ohila algonguin. Genetics 24 Morgan, T. H., C. B . Bridges, a nd A. H . S turt evant, 1925 The genetics of Drosophila. Pat au , IL, 1935 Chromos omenmor p hologi e bei Drosophil a rnelanogas t er und Drosonhil a simulans und ihr e genetisc he Bedeutung . Naturwiss. 23 Poulson, D. F., 1937 The embryonic development of Dros o·ohila melanogaster. Hermann & Cie. Paris. ? r okofy ev a-Belgovs kay a, A. A. 1939 - - - Comp. rend. ac ad . sci. TJ .R .S.S . 22 (cit ed fr om Kauf mann , 19 39) Serebrov sky , A . C ., 1935 Hybridization in ani mals . Mo s cow. S turt ev ant, A , H ., 1916 Notes on ~Jorth American Droso p hilidae with descri ptions of 23 ne w spec i es . Ann. Ent . S oc. Amer . .2 --- R.ndG . YI . Bead le, 1936The r slati ons of i nv ersions in the X-chromosome of Drosophila rnela nogaster to c ro s s ingov er and disjunction. Genetics 21 50 --- and T h. Dobzhansky, 1936a Inversions in the third chromosome of wild rac es of Dros oDhila pseud oobscura, and their use in the study of the h'Lstory of the spe ci es . Proc. Nat • Ac ad . S c i . 22 --- and Th. Dobzhansky, 1936b Observations on the s peci e s related to Drosophila affinis, wit h d es cri pt ions of seven new forms. Arner. Nat. 70 --- ar:rl C. C. Tan, 1937 The com parative gene tics of Drosoohila pseudoobscura and D . mel anogaster. Journ. Genet. lE Tan, C. C., 1935 Salivary gla nd chromos()mes in the t wo races of Drosophila pseudoob scura. Genetics 20 --- , 1937 The cytological maps of the autosomes in Drosophila DSeudoobscura. Zeits. f. Zellforsch. u . mikr . A nat. 26 --- and D. F. Poulson, 1937The behav ior of vermilion and orange eye colors in t rans 9 lantation in Drosop hila pseudoobscura. Journ. Ge net. ]1 Yochem, D. E ., 1929 Spermatozoon life in the f emale re-oroductive tract of the gu inea Di g and rat. Biol. Bull. 56 5l ACKNOWLEDGEM~l'Il'S The writer wishes to express his app reciation to a number of persons whose assistance and cooperation have mB.d e this work p ossible. He is particularly ind e bted to Professors A.H. Sturte- vant and Th. Dobzhansky of this ~aborat ory for suggestions and criticism during the course of the research. Li kewise, he wishe s to acknowledge the helpful cooperation of other membe Es of the biology deparment, especially Dr. R.H. MacKnight, Mr. E. Novitski, Mrs. C. H. Paul, and Dr. Jack Schultz. In addition, the writer wishes to thank the following persons of other institutions for contributions of material and for h elpful advice: Dr. c. v. Beers of the University of Southern California, Dr. P. T. Ives of Amherst College, Prof. J. T. Patters.o n of the t;T,, niversity of Texas, Dr. D. F. Poulso n of Yale University, and Dr. w. p :~ s·pencer of the College of Wooster.