Temperature Effects on Scute in Drosophila melanogaster Citation Ives, Philip Truman (1938) Temperature Effects on Scute in Drosophila melanogaster. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/wn1t-db75. https://resolver.caltech.edu/CaltechTHESIS:11262024-211809036 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 Thesis Committee: Unknown, Unknown Defense Date: 1938 Record Number: CaltechTHESIS:11262024-211809036 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:11262024-211809036 DOI: 10.7907/wn1t-db75 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16882 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 10 Dec 2024 23:45 Last Modified: 11 Dec 2024 00:15 Thesis Files PDF - Final Version See Usage Policy. 53MB Repository Staff Only: item control page

TEMPERATURE EFFECTS ON SCUTE IN DROSOPHILA MELANOGASTER Thesis by Philip Truman Ives The Rufus B. Kellogg University Research Fellow of Amherst College In Parti a l Fulfillment of the Requirements for the Degree of Doctor of Philosophy CALIFORNIA I NSTErUTE OF TECHNOLOGY Pasadena, California 1938 Acknowledgements It is a pleasure to acknowledge here my gratitude and obligations to the following people: To Professor Alfred H. Sturtevant, under whose direction this work was done, for many helpful suggestions and criticisms through the prosecution of the experiments and the preparation of this thesis; To Dr. James Bonner for help with the Moldex-yeast experiment, and for a number of fruitful suggestions; To Miss Eugenia Scott for aid in the formal preparation of this thesis; To many others of the Biology Department of this Institute for suggestions and aid through the course of the experiments; To Dr. George P. Child of Amherst College for criticisms and suggestions concerning the experimental work; To Professor Harold H. Plough of Amherst College for the use of the facilities of the Biology Department of Amherst College through the summer months of 1935, 1936, and 1937; And to Amherst College for the Fellowship which has made this study financially possible. i TABLE OF CONTENTS In trod u cjJ..Q.Q section I. - A Review of the Literature on Temneratureeffective Periods in Droso phi la. A. Introduction. 3 B. Reduplic at ed Legs. 3 c. Bar Eye and Its Alleles. 5 1. Introduction. 5 2. Temper a ture response. 6 3. Sexual domorphism. 6 4. Position of the TEP in development. 6 5. Rate of change in facet number. 8 6. Rel a tion of facet number and temperature 9 7. Nature of the Bar reaction. 10 8. Correlation of the TEP with visible processes. 13 D. Eyeless. 13 ~. Vestigial Wing and Its Alleles. 15 1. Introduction. 15 2. Temperature response of vg. 16 3. Position and duration of the TEP . 17 4. Wing-size temper a ture relation. 19 5. The modifier, 19 6. The allele, 11 dimorphos 11 • pennant 11 • 20 7. The allele, ;;No-vving 11 • 21 8. The dm-vg/vgP combination 22 9. Pattern in the vg series of alleles. 23 11 ii 10. Nature of action of vg and alleles. 23 11. Relation of the TEP to wing embryology. 23 12. The TEP of the wild-type wing. 26 13. Temperature effect on vg in D. virilis. 26 27 F. Bent Wing. G. Unequal Wings, 11 l 7b". 28 H. Dumpy Wing. 28 I. Dichaete Bristles. 29 J. Scute Bristles. 30 1. Introduction. 30 2. Viable temperature response and TEP. 31 3. Temperature change during the TEP, 32 4. Interpretation of the TER 32 5. Relation of the TEP to bristle embryology. 33 6. , Pattern in scute. 33 7. Inadequacy of the sub-gene hypothesis. 34 8. Temperature relations of sc-1 and sc-5. 35 9. 'rhe TEP at high temper a.ture. 36 Section II - The Experiments with Scute and Wild-type. A. Introduction. 37 B. Ma terials and Methods. 39 1. Stocks. 39 2. Relation of control and experimental flies. 39 3. Incubators and temperature measurement. 40 4. The bristles and their classification. 42 5. Calculation of bristle frequency. 43 6. Tabulation of differences in bristle freauencies .. 44 7. Statistical formulae. 44 iii page c. The Effects of Environmental factors and Genetic Modifiers 1. The independence of bristles. 47 47 2. Effects of crowding of larva e. 48 3. Effects of Moldex. 53 4. Effects of genetic modifiers. 57 5. Summary and conclusions. 60 D. The 40°c. Experiments with SC Males. 61 1. General plan of the experiments. 61 2. The data from sc 62 x v car. a. The embryonic period. 73 b. The first instar. 74 The second instar. 74 .• C • d • The first half of the third i nsta.r. 74 e. The second half of the third instar. 75 f. The early pupal period. 75 g. Summary. 75 h. Discussion. 78 i. Conclusions. 79 3. The data from SC X 80 sc. a. The embryonic period. 80 b. The first instar. 84 c. The second instar. 84 d. The first half of the third instar. 84 e. The second half of the third ins tar. 84 f. The early pupal period. 85 g. Sum111ary. 85 h. Discussion. 87 i. Comparison with sc j. Conclusions. x v car. 88 89 iv 4. Summary and Conclusions from 40°c. Experi ments . E. The 36°C. Experiments wi th sc Ma les 92 1. Plan of expe riments. 2. The data from sc x 90 92 v car. 92 a. The first instar. 96 b. The second ins t a r. 96 c. The first half of the third inst a r. 96 d. The second h a lf of the third instar. 96 -e • The early pu pal period. 97 f. Summary. 98 g . Discussion. 100 h. Comparison with 4o 0 c. a.The data from sc x SC X V car. sc. 102 103 a. The first instar. 104 b. The second inst a r. 104 c. The first half of the third ins t&.r. 108 d • The second half of 108 the third instar. e. The early pupal p eriod. 109 f. Summary. 109 g . Discussi on . 109 car. h . Comparison with 36°C. SC X V i . Comparison with 4o 0 c. SC X sc. 4. Su:rnrnary a nd Conclusions from 36°C. experiments. F. Other Temperature Experiments with s c lVIa le s. 112 112 114 116 1. Effects of 40 a nd 36°C. on F 1 ma les from treated parents. 2 . Experiments with lo wer tempe r e_tures. G. Temp e r a ture Effects on Dominan t Manifestation of sc. 1. Introduction. 116 117 118 118 V 2. The effects of 4o 0 c. on sc/v car females. 119 3. The effects of 36°C. on sc/v car females. 121 a. The data. 121 b. Discussion. 123 c. Comperison with sib males. 124 d. Comparison with 4o 0 c. sc/v car females. 124 4. The effects of 36°C. on wild-type flies. 125 a. Introduction. 125 b. The male data. 126 c. Discussion. 126 d. Comparison with 36°C. sc males from sc v car. x 128 e. The female data. 129 f. Discussion. 129 g. Comparison with wild-type males. 131 h. Eompc:.rison with 36°C. sc/v car females. 132 5. Effects of heterozygous SC at 36°C. 6. Summary and Conclusions. H. Interpretation of the Experimental Results. 133 138 140 1. The meaning of the change in bristle frequency. 140 2. N&ture of temperature effect on a bristle. 142 3. Meaning of differences in bristle responses to temperature. 143 4. Develo~mental history of sc. 144 5. :Meaning of the viable temperature TEP's. 145 6. Activity of the sc *ene. 146 Digest of the Thesis. 148 Bibliography._ 149 TEMPERATURE EFFECTS ON SCUTE IN DHOSOHULA :i\filLANOGASTER Introduction During the past twenty-five years a considerable amount of literature has appeared in biological journals dealing with the effects of temper a ture upon the expression of Mendelian characters in Drosophila melanogaster. It has been generally noted that the expression of many, if not of most, mutations is conditioned to some extent by the temperature at which the fly develops. It has been shown further in each of a number of instances that the temperature effect on the adult character is produced during only part of the total developmental period, the position and length of this part depending upon the mutation concerned. This part of the period has come to be known as the "temperature-effective period" (TEP) of the character. Studies dn TEP's have been undertaken largely with the hope of unveiling the mechanism of gene action in development. Although they have not realized this hope extensive- ly, these studies have brought to light many important observational facts, as well as some interesting speculations concerned with their interpretation. Practically all of the studies thus far reported have been made upon the so-called "viable range" of temperatures, by which is meant the temperatures at which D. melanogaster can develop successfully from egg to adult. The -2study reported here, suggested by Professor Sturtevant, was made at temperatures above this range, using a mutation whose responses to viable temperatures were at that time in the process of being determined. The results prove to be of in- terest both by themselves and in contrast with the results of the subsequently publis~ed lower temperature studies. I propose first to review the TEP literature, and then to present and discuss my own experimental data. SECTION I. A REVIEW ill' THE LITERATURE O:N TEMPERATURE-EFFECT IVE PERIODS IN DROSOPHILA -3- A. Introduction No general review of the literature from TEP studies on Drosophila mutants has been published. Accord- ingly, this review is entirely from the original papers. The mutant characters are discussed in approxima tely the chronological order in which their respective TEP's were first studied. It will be noted that the bulk of the literature is from studies on Bar eye and its alleles, and on vestigial wing and its alleles. Most of the work on the Bar series, and much of that on the vestigial series, has been done under the supervision of Professor Charles Zeleny of the University of Illinois (Urbana) or of Professor Charles Plunkett, Washington Square College, New York University (New York City). The studies on eyeless and on scute bristles were also done under the supervision of Professor Plunkett. B. Reduplicated Legs The first report of temperature effect on the expression of a Mendelian character in Drosophila was that of Miss Hoge (1915) on the sex-linked recessive, "reduplicated legs 11 • She found that cultures of a pure line of this mu- tant manifested up to 100% of the character when maintained at a tempera ture of 10 to 12°c. well into the pupal period, as compared to a 5 to 10% manifestation in cultures raised entirely a t room temperatures (about 22°c.). Miss Hoge ma de two series of experiments to show - 4when the temperature effect was produced in development. In the first series, cultures began development at room temperature and were transferred at intervals to the low temperature. The results indicated that the low temper- ature effects were strongest when the transfers were made in the egg stage, and were progressively weaker as the time of transfer was delayed. When development had proceeded for more than four days at room temperature, there was no apparent effect of subsequent low temperature on the manifestation of the character. This indicated that the low temperature wa s effective only durin g the egg-la rval period and not in the pupal stage. In the second series of experiments, the cultures began development at the low temperature and were transferred at intervals to room temperature. From this series, it was learned that the effect of low temperature did not manifest itself until the first f ive or more days of development had elapsed. The effect of the low temperature increased the longer the transfer to room temperature was delayed. The effect was greatest1when cultures were kept at the low temperature until the appearance of the pupae. I In both of these series, the temperature effect seemed to be directly proportional to the amount of egglarval development which took place at the low temperature; and the TEP seemed to occupy the egg-larval period of development. Miss Auerbach (1936) showed in her studies on Drosophila imaginal discs that cert ain ly from the early -5- larval period the limb buds are in visible process of growth. She suggested that the effect of "reduplicated legs" is strongest on the embryonic processes leading to the formation of these buds, since the most notable effect of the low temperature manifested itself during the embryonic period. The TEP, however, was not closely enough determined to make profitable an exact analysis from Miss Hoge's data. c. Bar Eye and Its . Alleles. 1. Introduction. Seyster (1919) was the first to report an effect of temperature upon the appearance of the dominant, sexlinked, eye-reducing gene, Bar (B). He showed that facet number in B eye is inversely related to the developmental temperature. As the temperature was lowered through the 0 viable range from 29 to 15 c. the facet number increased, more rapidly in females than in males. Males, however, ap- peared to have more facets than the females at all temperatures. He found a van't Hoff Q1o value of 2.6 in males and 3.5 in females. From this he concludes that B acts as a chemical inhibitor of facet formation, and that its effect is stronger at high temperatures than at low. His studies on the TEP indicat~d it to be contained within the larval developmental period. Since Seyster's report, many temperature investigations have been made on Band its alleles. The reports have come from l<rafka (1920), Hersh (1924 - 1934), E. c. Driver (1926,1931), Luce (1926, 1931), Olive Driver (1931), -6- Margolis (1935 - 1936), and Margolis and Robertson (1937). In general these studies h ave been made in the viable range of temperatures, from 15 to 31°c., and have been made on several B stocks as well as several alleles. These alleles include "ultra-Bar" (Bu) which manifests a i more extreme effect than B, and "infra-Bar" (B) which produces a less extreme effect than B, as well as the wild-type eye. The studies also include work on the several heterozy- gotic combinations of these genes. 2. Temperature response. Band Bu, and their heterozygotes with each other and with wild-type, show an/. inverse response to temperature changes. Bi shows a direct response to temperature changes, the facet number being higher at high temperatures than at low. l~ile the B/Bi and BU/Bi flies show facet num- bers intermediate to the two alleles in each case, the temperature response_ is inverse, as in Band Bu. 3. Sexual dimorphism. In all of the alleles and their combinations there is a clear sexual dimorphism in facet number, the males consistently having a few more facets than the females at all temperatures. There does not a.ppear to be any notable dimorphism in respect to their temperature responses. 4. Position of the TEP in development. TEP's were determined for Band its alleles by transferring cultures from one temperature to another at graded intervals in development. There appeared to be some -7i~riation in TEP from stock to stock of B. ..CZ. leles ther/1 was a much wider variation. Among the al- In all cases, the TEP seemed to occur within the first half of the third instar period throughout the viable temperature range. The data of E. C. Driver, Luce, and Margolis &re the most instructive concerning the position of the TEP at different temperatures. Driver found that in B the begin- ning of the TEP is probably constant throughout the viable temperature range, and that it probably coincid.es with the beginning of the third instar. The end of the TEP changes at some temperature between 20 and 22°c., and changes the duration of the TEP. U:p to 20°c., the TEP covers 25% of ll=lrval life; from 22 to 30°c. it covers only 16%. He found that in Bu, on the other hand, the duration of the TEP is consistently the same at all temperatures, and consists of approximately 17~{; of egg-larval development. The position of the TEP, however, lies progressively later in the first half of the third instar as the temperature increases. Luce showed that in Bi there is a slight shift in the beginning point of the TEP which gradually shortens its duration from 31% of egg-larval life at 17°c. to 24% at 27°C. The heterozygote, B/Bi, shows to a lesser extent both the effect of Bi on the beginning point and the effect of Bon the end-potnt of the TEP. The duration of the TEP in B/Bi de- creases from 24% of egg-larval life at 17° to 13% at 27°c. Margolis worked out a method for calculating the TEP of the individual larvae from that of the entire larval :population. He found, as might be expected, that the TEP of the individual is much less than that of thepopulation. -8- The population is made up of individuals which vary in the degree of their respective developments at a given time. The TEP of the population is the sum of the TEP's of the constituent individuals, and its length is determined by the oldest and youngest larvae present. Margolis found that in his B stock the duration of the individual was considerably less than half that of the population TEP. From the reports of the investigators as a group it is apparent that the position and duration of the TEP vary slightly, probably because of different sets of modifying genes in the different stocks and because of differences in technique and in environmental conditions. 5. Rate of change in facet number. The rate of change in facet number per degree change in temperature varies between alleles and to some extent between stocks of the same allele in the B series. It is gener- ally not greatly different between the sexes of a g iven stock. Driver found in Ba 10% increase in facet number per degree drop in temperature, between 30 and 22°c. Between 22 and 20°c. the rate changed, paralleling the change in position and duration of the TEP. From 20 to 15°C. the rate of increase in facet number was 7% per degree. Driver's data on Bu indicate a fairly constant rate of increase in facet number throughout the tempejature range. The rate was between 4 and 5% per degree drop in temperature. In 1926 Luce reported in his Bi studies that the rate of increase in facet number per degree increase in temperature, over the entire viable temper a ture range, was 8.3% in - 9- males and 7.3% in fema les. In his 1931 data. these rates had dropped to 7.2% and 6.6%, respectively. A comparison of the data. of Seyster, Krafka, Margolis, and others shows similar small changes in rate between stocks of Band alleles, at different points in time, and under different conditions. Hersh (1927) reported that in reciprocal crosses of Band Bu, while the facet number was nearly exactly intermediate between the alleles in both crosses, the rate of decrease in facet number followed the maternal allele in each case. Further studies (1930} showed that this was true only at lower temperatures; that the re,tes approach each other at higher temperatures; and that at 2a 0 c. the rates in reciprocal crosses are practically the same. In B/Bi Luce found that the rate was only slightly lower than in B. 6. Relation of facet number and temperature. There does not seem to be any general agreement among the investiga tors of Band alleles as to the nature of the rel a tion of change in facet number to change in temperature. Between 17 and 27°c. the relation may be plotted either as linear or logarithmic. While the behavior at ex- treme temperatures seems to favor an exponential viewpoint, it has been pointed out by those favoring a linear relation that irregularities in the linear curves at the extremes can be accounted for on the assumption that they represent irregularities consequent to the increasing developmental disturbances of the extreme temperatures. In the case of B, the change in rate between 20 and 22 ° c . i s such that two straight lines g ive a better fit -10- to the data, whether the calculation be on a linear or ex- ponential basis. ?. Nature of the Bar reaction. Following Seyster, Krafka (192~ calculated van't Hoff Q. 10 values for facet number change in the temperature range 15 to 31°C. this range. In BU, ~10 varied from 1.60 to 2.86 in In B, it varied from 1.77 to 4.43. In both cases it was low in the middle tempera.tures and high at the extremes. Krafka concluded that, on the whole, his da.ta sup- ported Seyster's hypothesis that the B reaction represents a unimolecular chemical reaction. E. c. Driver (1931) and Hersh (1934) have carried their calculations further, using the Arrhenius equation. Instead of giving a straight line as demanded by the SeysterKrafka hypothesis, the plotted u values give a smooth curve throughout the tempera.ture range. Driver, following Crozier (1924)> interprets this to mean that the B reaction is not a "master reaction"; that we are concerned, instead, with an equilibrium between two systems. One of these systems in- volves the reactions of Band its alleles; the other, the reactions of the rest of the gene complex as a vn1ole. The differential temperature coefficients of these two systems, he believes, account for the change in facet number with change in temperature. The Band Bu systems can be considered as having higher temperature coefficients than the system involving the rest of the genes concerned with development. Consequently, their restraining influence on facet formation -11- is greater at higher temper atures. B influences both the rate of facet formation and the duration and position of the TEP; Bu influences the rate of facet formation and to a less extent the position of the TEP. The Bi system can be considered as having a lower temperature coefficient than the system of development as a whole, and consequently affects the facet number less at higher temperatures. The effect of Bi is largely on the rate of facet formation, less on the position and duration of the TEP. Margolis (1935 - 1936) reported studies on the temperature response of Bin vestigial (vg) winged flies. He found the B vg has fewer facets at nearly all temperatures than the B stock from which it was derived; and that B vg also has a slightly longer TEP than the B stock. Since vg is known to increase the len~lt\ of larval life, heargued that the effect of vg on B facet number can be formally explained as an increasing of the time in which the B gene can affect facet number. Krafka (1920) was unable to detedt any effect of temperature on facet number in the wild-type eye. Rosalie Hersh (1924) reported a small steady decrease in facet number with increasing tempera ture, equivalent to a 2.5% decrease per degree of increase in temperature. attempt to locate· the TEP. She did not Margolis and Robertson (1937) found that the effect of temperature was not constant at different points in the developmental period. There seemed to be a general increase in facet number during the embryonic stage, and a g eneral decrease during nearly all of -12larval life, the change in no case being large. In both egg and larval periods, however, there were changes in direction in the facet-temperature curve, and they concluded from the nature of these changes that there are at least four reactions responding at these different times to temperature changes. From the results of the studies with B vg and with wild-type eye, Margolis extends the above theories of the nature of the Bar alleles and of the mechanics of their effects on development. He postulates three reaction sys- tems which lead ultimately to the formation of ommatidia and facets. The first of these systems forms A, which de- termines when the region containing facet-forming material shall differentiate into ommatidia. The second system forms F, the facet-forming material. The third forms D, which catalyses the destruction of F by a chemical reaction which he denotes as and which figures prominently in Bar activity. This third reaction system is of relatively negligible importance in the wild genotype and is in fact part of the second system. In Band its alleles it has become of much more importance in the determination of facet numbEr. From the TEP data of the wild type eye the production of F seems to begin early in embryonic development. From the TEP data on Band al- leles the reaction between F and D occupies the early part of the third inst ar. The onset of the reaction Margolis .assumes to coincide with and indeed to be caused by the be- -13ginning of the production of D by the third ~a ction system. The ending of the reaction comes at the time at which A determines that the remaining F shall become ommatidia. The effects of the different alleles of Bare upon the rate and duration of the reaction between F and D. 8. Correlation of TEP with visible embryolog ical processes. Krafka (1924) reported a brief study of larval embryology of the imaginal eye disc in Drosophila. He de- cided that the TEP of B coincides with the joining of the prospective eye disc with the larval brain ganglia. Chen (1929) reported a much more detailed series of studies on larval embryology. It is generally to his work that subse- quent TEP investigators have referred. He located the pros- pective eye material in the late first instar, when it appears in the form of paired frontal sacs invaginated from the pharynx. He followed the subsequent growth, differentia- tion, and development of these sacs into ommatidia in early pupal life. Studying the effects of B, he was able to dis- tinguish Band wild-type eyes in the mature larvae when the eye bud and ommatidia. number are smaller in B. proximately a day later than the TEP. This is ap- Driver (1931) concluded that the TEP occurs at the time when the frontal sac, having migrated back to the brain ganglia, differentiates into the optic and antennal buds. D. Eyeless The temperature response of the genotype of eyeless (ey) has be en reported by Baron (1935). The ey pheno- type is variable from 11 no eye" to nearly normal eye. In -14- highly selected and i'nt,r ed stocks a large proportion of the population is without eyes. Baron found this proportion varied with the developmental temperature. In a stock se- lected for low frequency of eyes, an increase in temperature decreased the number of eyes. In the higher temperatures of the viable range practically no eyes were present. In a "standard 11 ey 2 stock, inbred but selected for a high frequency of reduced eyes, the situation was entirely different. Both the size of the eye and the number of eyes present increased with the tli:mperature, excepting at the highest temperatures. Nearly all the flies raised at 27°c. had eyes. It is evident that the ey phenotype is much more dependent upon modifying genes than is B. Using the "low-frequency" stock, Baron found that the TEP occurs roughly between l½ and 2½ days of larval development at 25°c. The end, he believes, corresponds rough- ly with the time of appearance of the optic'buds. Chen (1929) found that at this time ey larvae have only rudimentary buds. Baron studied a combination of Band ey highfrequency stock. The facet number was consistently lower than in the B ·stock; but the fEP seemed to correspond entirely with that of B. Baron has advanced an hypothesis to account for the behavior of low-frequency ey. He considers ey as respon- sible for the production of a variable amount of a substance, E, which reacts with the products, F, of the rest of the genotype to produce facets, f. Fis produced in constant quantity at all but the extremes of viable temperatures. - -15He shows from his data that the relation of E to f is constant at least at medium temperatures. Up to an f value of 400 facets the relation of E and f is linear; from f values of 400 to 880 (upper limit in the normal eye), the relation is logarithmic. He interprets this to mean that Eis an enzyine acting on the substrate, F. Be- low the 400 value off, the relation of E to f depends, practically, on the amount of E present, since that amount is small compared to F. Above 400, the amount of E increases and F decreases to the extent that the E - f curve now becomes the product of two variables, E and F, and is consequently logarithmic. E. Vestigial Wing and Its Alleles. 1. Introduction Roberts (1918) first reported an effect of temperature upon the appearance of vestigial (vg) wing in D. melanogaster. He found that vg flies which developed at temperatures between 27 and 34°C. had larger wings than those which developed at room temperatures. The effect was produced during the larval development and not in the pupal stage. More extensive reports have appeared from the work of many investigators in the 20 years since Robert's report, and have included studies on several vg alleles as well as vg itself. thors: Papers have been published by the following au- Stanley (1928, 1931, 1935); Harnly (1930, 1936); Hetsh and Ward (1932); Riedel (1934); Harnly and Harnly -16- {1935, 1936); Li and Tsui (1936); Gable (1938); and Lashin Nadler (1926) has reported a study on vg in D. vir- (1938). ilis. 2. Temperature response of vg. It has been found to be general in D. melanogaster that vg wing increases in size with an increase in developmental temperature. The increase is at a relatively slow 0 rate over the range, 16 to 29 c., and does not involve a marked phenotypic change. The rate of increase is rapid from 30° to 320, which is the limit of the viable range for vg flies. It is accompanied by striking phenotypic changes in the wing. range. A clear sexual dimorphism exists in this upper The "critical" temperature at which male wings first begin to show the rapid increase in rate of change is near 30°c; the critical temperature for the female wing is near 31°c. The female wing at 30°c. is still small, like the male wing at 29°C. The female wing at 32°C. is still larger, and exceeds the size of the male wing, which does not increase in size with the change from 31 to 32°C. From 16 to 29°C., the vg phenotype is practically "vestigial". But as the wing grows larger above the critical temperature, it passes through stages resembling all of the less extreme alleles of the vg series. illustration of alleles.) (See Mohr 1932 for At 32°c., vg wing is notched at the tip; but otherwise it is normal in appearance. In area, it is still smaller than the wild-type wing at 32°c. Accompanying the changes in the phenotype of the vg wing are similar changes towards normal in the elevation -17- ot the posterior scutella r b ristles, and in the size and appearance of the b a lancers. At 32°C., these appear normal in vg flies. 3. Position and duration of the TEP. Stanley and Harnly have repoted most of the extensive TEP studies in vg; Stanley at 17, 27, 30, and 31°c., and Harnly at 30, 31, 32, and 33°c. Their results are in fairly close agreement, and will be discussed together. 0 At 17 c. , the TEP begins apparently with the second instar and extends probably throu gh the early part of the third. At 27°c., the TEP begins possibly before the larva hatches from the egg, and appears to end midway through the second instar . There do not appear to be significant sexual differences in the length and position of the TEP at these two temperatures. At each temperature the TEP occupies approximately 30% of egg-larval development. Beginning at 3o 0 c. a strong sexual dimorphism manifests itself in the TEP, paralleling the dimorphism in the critical temper a tures of phenotypic response. At all the higher temper2.tures the TEP appears to begin in both sexes with the molt that closes the second instar. At 30°c., the TEP ends, f or males, about half way throu gh the instar; for females, probably within the first quarter of the inst a r. At 31°C., the TEP includes the entire third instar for males, which is appr oximately 48 hours; for females, it includes only the first half of the instar, or 24 hours. At 32°c., puparium formation is greatly delayed, and only rarely do flies withs t a nd th e temperature long enou gh to pass through - 18- this stage into pupation. The TEP ends in 48 hours for the males, as it did at 31°c. In females, however, the TEP at 32°c. shows no sign of completion at the end of 70 hours from its beginning. exposure for vg flies. That is the limit of viable At 33°c., the TEP in both sexes lasts the limit of the viable exposure, which is 50 hours, and does not appear to have ended at that time. Harnly pointed out that the sexual dimorphism in phenotypic response to supra-critical temperatures can be formally explained on the basis of the similar differences in TEP length at these temperatures. There is a slight but consistent dimorphism in the 16 to 29°c. range, stronger in Stanley 1 s data than in Harnly's. slightly larger wings than females. Males have This appears to be due to one or more mo d ifiers in the X chromosome,as pointed out by Stanley and Riedel. Li and Tsui, in their studies at 25 and 31°c., produced evidence that the TEP at 31°c. is not the only time when developmenta,l activity of importance to the vg phenotype is taking place. They found that vg flies v1hich spend the first larval instc.1.r, as well a.s the TEP, at 31°C. have considerably larger wings than do those flies which spend only the TEP at 31°c. Flies which spend the first instar at 31° c. and the TEP at 25°c. have the same sized wings as those which develop throughout at 2 5°c. The usual technique employed i n TEP studies, therefore, is not sufficient to establ i sh all the effective period of the vg phenotype. - 194. Wing-size temperature relati on. Hersh and Ward, consid e ring but three temperatures, concluded that the vg wing-size temperature relation is exponential. Harnly stated, however, that when one considers the entire temperature range, at least two logarithmic lines are necessary to fit the data. When the cases of combinations of dimorphos-vg and vg-pennant are also considered (Harnly and Harnly), several such lines are neces sary. Harnly concluded that the loga- rithmic relation is not general in vg and its alleles. 5. The modifier, 11 dimorphos 11 • Harnly and Harnly reported the temperature re- sponse of the sex-linked mod·ifying gene, with vg (1935), and of vg and its allele, 11 dimorphos 11 (dm), 11 pennant 11 (vgP), (1936). The gene,dm, does not manifest any visible effect by itself; but its presence in the genotype increases the size of vg wings at all temperatures. Its name origin- ates in the fact that it affects the male wing much more than the female wing at tempera.tures below 32°C. The cri- tical tempere.ture for phenotypic response to temperature is shifted down to 25°c. in males and to 2s 0 c. in females. In addition, a new and less extreme critical temperature appears for the males, probably between 16 and 1s 0 c. At 32°c., the wings are of equal size in the two sexes, and 16% of them are like the wild-type wing, in both size and appearance. -20- 6. The allele, "pennant". The wings of pennant (vgP) flies are normal in appearance at all temperatures. A few of them have ter- mina: 1 nicks and all of them seem to have weaker distal wing sections. In old flies, the distal portion of the wing is broken and ragged. The response of vgP to temperature is the same as that of wild-type throughout the viable range. Several investigators listed above have shown that the wild-type wing responds inversely, though only slightly, to temperature changes; and that females have somewhat larger wings than males at all temperatures. In both of these respects, the wild-type is the opposite of vg; and in both of them vgP resembles the wild-type rather than vg, from which it arose as a mutation. At all temperatures, too, the abso- lute size of vgP wings closely approximates that of the wild-type wing. The response of vg/vgP, however, is quite different from that of vg/wild-type. As shown by a number of investigators, heterozygous vg is phenotypically wild-type at all temperatures. The rate of decrease in wing size with temperature increase is retarded from that of the wild genotype. As a result, at 32°C. vg/wild-type is larger in area than the wild~type wing. The wings of vg/vgP are intermediate in phenotype, resembling "strap 11 , at all but the higher temperatures. There is practica l ly no change from 22 to 2s 0 c.; but from there to 32 °C. there is a sharp increase in si§e and a progressive change in phenotype. At 32°C. vg/vgP is fully the equal of wild-type or vg/vgP in wing area; but a persistent terminal nick is present. (Since th is terminal nick also appeared in a vgP stock into which the modifiers of the vg stock had been crossed, it does not seem necessary to conclude with Harnly and Harnly that this effect is due to the vg gene.) In the combination of vg/vgP, penna..._,,nt seems to have shifted the original critical temperature of vg downward, and to have introduced a new and oppositely directed critical temperature at 22°c. A strong sexual dimorphism exists in vg/vgP. Be- low 3o 0 c., the males have the larger wings; above 30°c., 1he females have the larger wings. This is analagous to the situation in vg/vg. 7. The allele, 11 No-wing 11 • Miss Gable (1938) has recently reported in abstract the temperature response of another allele of vg, 11 No-wing 11 (vgNw). This is a 11 No-wing 11 which I found among flies whose grandparents had been subjected to 36.5°c. temper a ture during part of their larval development (Plough and Ives 1 935). It is more extreme than vg in itw pheno- typic and physiological effects at room temperatures. Heterozygous vgNw flies usually have terminally notched wings. Homozygous ones have apterous wings, s1naller than vg; and the females are sterile. The homozygous 11 No-wing 11 wing does not vary in appearance over the viable range, according to :Miss Gable. The vg/vgNw wing is vg in appearance; but, like vgNw/vgNw - 22- and unlike vg/vg, it does not vary in appearance over the viable temperature range. The vgP/vgNw wing is intermed- iate in phenotype, and responds inversely to temperature changes throughout the range, the phenotype changing gradually from "strap" to vg. In this case, the r e sponse to temperature is that of vgP. The vgNw/wild-type terminally notched wing is like wild-type in having an inverse relation to temperatures of the viable range; but it is consistently smaller than wild-type in area. 8. The dm-vg/vgP combination. Miss Lashin (1938) has reported in abstract form the response of the dm-vg/vgP combination to temperatures in the range, 16 to 2s 0 c. Wing length in this combination bears an inverse relation to temperature similar to that of wild-type and vg/vg. In males, the area of the wings is like that of vgP/vgP. In females, however, the area curve is U-shaped as in vg/vgP, the area values being approximately twice those of vg/vgP wings at corresponding temperatures. In phenotype the male wings may be wild- type; but they are usually nicked. type is "antlered" or "strap" . In females, the phe no- From the appearance of the area and length curves, and from the changes in phenotypes, she concludes that 28°c. is critical for dm-vg/vgP flies. The TEP's of dm-vg, vgP, and vgNw and their heterozygotes have not been worked out. Nor have extensive embryological studies been made. One can not profitably s uggest further relations from the above data alone. - 23 - 9. Pattern in the vg series of alleles. Ha rnly and Harnly stated that the order of succession of wing types appearin g a t increasing temperatures was essentially the same in v g, dm-vg, and vgJ'vgP . Accord- ingly, they support the sugg estion of several others that the win g develops according to a cert a in pattern, under the influence of the vg a llelic series of g enes. Li and Tsui also divided the wings which a ppe a red in their w6rk into several types, pro g ressing from v g to normal type. But they empha,12. sized that the types did not necessarily succeftd e a ch other as the mumber of hours spent by the larvae a t 31°c. increased. They found no real evidence for a pattern of development in their vg studies. 10. Nature of action of vg and alleles. Stanley adva nced the hpothesis that vg produces an inactivating substance which acts upon wing-forming substance but is destroyed a t hi gher temper a tures. In vg/vgP, however, the effect of temperature appe a red with decreasing as well as with increasing temper a tures. Harnly and Harnly concluded from t h is observa tion that v g does not produce such a substance. Goldschmidt's (1935,1937) embryological studies indicated that at least the less extreme alleles of vg produce their effects appa rently in the -pupal stag e by destroyin g win g tissue. In these alleles, the wing seems to be normal at the time when the pu pal wing-sheath appears, early in pupal development. Subsequently, de ge ner a tion sets in. The a mount of scalloping manifested in the adult wing appears - 24 - to depend upon the time of the beginning of this process. From this, he advanced the hypothesis that the more extreme vg alleles, and finally vg itself, push the beginning point of degener a tion further and further back in development, so that there is more time for destruction of wing material before final differentiation has been accomplished. It is also his opinion that temperature acts upon the ~eginning point of this process similarly, moving it in one direction or another, as the phenotype indicates. Li and Tsui counted the cells in a measured region of vg wings which had spent measured amounts of time at 31°C. in the TEP. They found that corresponding to the increase in wing size with increased time at 31°C., there was an increase in the number of cells in the wing, and a slight increase in the size of the cells. They a.}.so observed an increase in the size of the dotsal mesothoracic imaginal disc (from which the wing disc develops) in 31°c. vg larvae as compared to 25°c. larvae. From this they concluded that temperature may produce in the case of the vg wing a direct increase in growth as well as the degeneration process observed by Goldschmidt. 11. Relation of TEP to wing embryology. Chen (1929) in his embryological studies on Drosophila reported that the dorsal mesothoracic bud first appears in the second instar period. In the third instar, it is already smaller in vg than in wild-type wing. The wing bud seems also to be delayed in vg in its differentiation from the dorsal mesothoracic bud. These effects were con- -25- siderably more pron ounced in a 11 No-wing 11 stock. Stanley and Harnly stated that since the TEP of vg at the higher temperatures appears to begin early in the third instar, the effect of vg is probably upon the growth processes taking place at the time when the wing-forming area divides from the thoracic-forming area of the dorsal mesothoracic bud. The observations and conclusions of Li and Tsui are in accord with these views. Auerbach (1936) has made a penetrating embryological study of the development of the same buds. She w~s able to show that the small dorsal mesothoracic bud : is already present in newly hatched larvae. Poulson (1937) found the beginning of visible formation of the same bud about three-quarters of the way through embryonic develoopment in histological sections of the 17 to 18 hour egg. M$ss Auerbach traced the development of this bud throughout larval life and up to the formation of the wing sheath in the pre-pupal period. The first conclusive morphological difference at 26°C. between vg and wild-type wing she fou~ nd to be just previous to puparium formation, when the wing begins to grow out from the wing-forming area of the dorsal mesothoracic disc. She admits the possibility of fol- lowing the difference further back by histological methods. Miss Auerbach made the important observation, simple though it is, that one does not have to associ ate the TEP with a contemporary and visible morphological process; that reactions may indeed be occuring which do not manifest themselves morphologically until as much later as the actual differentiation of the wing-forming area just -86- previous to pupation. At the same time, her observations and those of others, that the imaginal wing begins visible differentiation sometime previous to the hatching of larva from the egg and continues throughout larval and into pupal life makes it possible, for those who desire to do so, to associate practically any wi ng TEP they find with some visible, if not also differential, morpholo gical process of development. 12. The TEP of the wild-type wing. Stanley (1935) also determined the TEP for the wild-type wing. He found that it begins late in the third instar and continues probably to the middle of the pupal developmental period. This period is clearly later than the TEP of vg, and the duration is also considerably longe r. It corresponds in time to the visible differentiation of the wing. The position of the TEP does not vary greatly over the viable temperature range. 13. Temperature effect on vg in D. virilis. Nadler (1926) reported that in D. virilis, as in D. melanogaster, vg increases in size with increase in developmental temperature. The rate of increase seemed to follow the van't Hoff law for temperature effects on a chemical reaction, with an average ~16 value of 1.98. In the range, 12 to 20°c., ~lo was 2.44; and from 20 to 30°c. it was 1.51. appeared. Further studies of vg in this species have not F. Bent Vlin.g_. Metz (1923) reported briefly a study of the temperature response of the bent-wing phenotype in Drosophila melanogaster and D. virilis, in the range 9-12°, 16°, 23°, and 25°c. At the low temperatures, the effects of bent on the virilis eye ( 11 rou ghening 11 of the facets) were considerably enhanced. Scutella.r bristles were also misplaced or missing, and other new modifications appeared over the fly. The effects of the gene on the wings and legs seemed to be less extreme at the lower temperatures. In melanogaster, bent did not affect the eyes at normal temperatures; but at the low temperatures a roughening effect appeared, less extreme than in virilis. In other respects the behavior of bent was much the same in the two species at the temperatures tested. The TEP seemed to be located in late larval and early pupal life in both species. Wild-type flies did not show any of the modifi- cations of bent at these temperatures; and offspring of bent flies which manifested the extreme effects of cold were only ordinary bent at room temperatures. From this Metz conclud- ed that temperature had only enhanced the effects of bent, and that modifying ;factors were probably not responsible for the phenomena. The time of the TEP corresponds to the time when extensive visible differentiation and expansion is occu4ng in the imaginal disc (Auerbach 1936). - 28- G. Un eg ual WiBE S 17b . Miss Auerbach (1936) studied the temperature response of "unequal wings 17b", a third chromosome mutHnt with variab le manifestation. The percent of manifestation falls with the temperature of development, and is lower in males than in females. mal wings. At l6°C. nearly all flies have nor- The TEP was determined by transfers between 16 and 27°C. and was found to occur late in the third instar, probably just preceding the prepupal period. From her stud- ies on wing development, Miss Auerbach concluded that the deci s ive action of 11 17b" consists probably of an alteration in the process of wing-pouch formation. H. Dumpy Wing. Blanc and Child (1938) have reported in abstract 0 the effects of brief exposure to 36.5 c. upon the expression of "dumpy" (dy) wings. In homozygous condition this gene produces a truncate wing at the usual room tempe r atures, and is not effective in heterozygous condition. Blanc and Child report that dy is an effective dominant when puae are exposed for 12 hours to 36.5°C. during the first day after pupation. Under these conditions~ dy is more effective in males than in females. They observed in addition that wild-type [l ies show a clear but much weaker tendency to produce truncate wings under similar conditions. The TEP for this reaction corresponds to the time of dumpy wing degeneration in the pupal wing sheath as observed by Goldschrnidt (1935). -29I. Dichaete bristles. Plunkett (1926) reported studies on Dichaete (D) in D. melanogaster. This gene prevents the formation of certain bristles of the scutellum and thorax. He was able to show by mean bristle numbers, by association constants between the bristles, and by observations on individual flies, that the effect of Dis a regular pattern which centers at the presutural bristle and spreads in all directions. The lack of left-right correlation between the bristles of the pattern indicated that for the individual fly the pattern of the left half varies independently of the pattern of the right half. The simplest explana- tion seemed to be to assume that the D gene is somehow related to a substance which diffuses from the pattern center on each side of the fly during early development, and which, under the influence of D, prevents the formation of bristles. Because of the high correlations between bristles of the pattern at 25°c., Plunkett confined his observations at other temperatures to one pair of these bristles, the posterior dorso-centrals (p de). He found that the mean number (M) of p de decreased steadily as the temperature increased from 15° to 30°c. duration between 24 and The TEP varied but little in 3o 0 c., and it extended from early larval life until late in the pupal stage, close to the time when the bristles first become visible through the pupa case, a day or so before emergence of the imago. The change in M of Pde appeared to be proportional to the time spent at a given temperature at any time during this extended TEP. -30- When the Arrhenius u values of the thermal incr ement of the D reaction were plotted, they gave a str&,i ght line throughout the temperature range, with an average u value of 34,600 units. This value suggested the he a t destruc- tion of enzymes. From further studies on the M p de, Plunkett advanced an hypothes i s to explain the action of Don the bristles of its pattern, and the effect of temperature upon this reaction. Extended theoretical kinetic consider- ations showed that the data from the p de bristles at the several temperatures and from the pattern of bristles at 25°c. supported the hypothesis with admirable goodness of fit. Later unpublished data by Plunkett (information from Professor Sturtevant) on the other bristles at different temperatures indicat•d that these bristles did not behave similar to the p de, and indicated the probability that the hypothesis developed from the p de was much too simplified to be of use for the whole pattern at all temperatures. J. Scute Bristles. 1. Introduction. Payne (19 80) first observed scute (sc), a sexlinked gene removing certain bristles of the scutellum, thorax, and head in D. melanogaster. He isolated it in the course of selection experiments on bristle number. He reported that a student of his, Froemming , found that the number of scutellar bristles present in the mutant was influenced by the developmental tempera.ture. Flies raised -3 1 - at 10 to 13°c. had more scutellar bristles than those raised at 23 to ~6°C. The effect of the tempercture seemed to be limited to the larval period. 2. Viable temper2.ture response and TEP. Child (1935, 1936) has published the results of an extended series of experiments with sc at tem1Jeratures ranging from 14 to 31°C. He considered all the bristles of the scutellum, thorax, and head. It was appar- ent not only that the mean frequencies of these bristles varied with temperature; but that the differences were not the same for each of the several bristles, either in amount or in direction. Some bristle frequencies reached maxima (100% present); some reached minima (0% present); some showed sharp rises in a relatively small temperature range; and some rose through part of the range a.nd fell through the rest of it. The TEP for all bristles, howevet, when a population of flies was considered, occupied the same relative po sition in egg-lar-val development; namely, the last half of the third instar. Chil d was able to show experi- mentally and by theoretical considerations that for the individual fly the TEP is a much shorter period. For this study he used the ocellar bristl e, whose response to temperature changes was the strongest of all the sc-affected bristles in his experiments. The TEP for this bristle proved to lie entirely between 89 and 97% of egg-larval life. It appeared to ~nd just before the onset of puparium formation. Child did not conclude that the TEP is necessarily of exa ctly the same length for every bristle; but he believed -32- it to be apparent that for them too the TEP for the individual larva is much less than the TEP for the population of larvae in a given culture. 3. Temperature change during the TEP. Child found that when larvae were transferred from one temperature to another during the TEP, the bristle freouenciea were generally intermediate to the standard frequencies for the temperatures concerned. The frequencies of the anterior notopleural (an) and certain of the orbital (or) bristles were exceptions to this rule. When the change was from a low to a high temperature, the an bristle frequency was higher than either of the standard frequencies. When the temperatures were on opposite sides of 2s 0 c., the or frequency was lower than either standard frequency. There wa.s no apparent explanation for either of these anomalies. 4. Interpretation of the TEP. Since the bristle frequency-temperature curves in Child's data vary so widely from bristle to bristle, and between the sexes, there is a strong suggestion that many reactions are involved. Child, however, pointed out that such a view is not the only possible one. The TEP's for all of the bristles occur in general in the sarae time period of development; and we are concerned, in each case, with the formation of a bristle. From this he argued that it is natural to assume that the same reaction is being effected at each bristle sight. He points out that it may be that for some bristles a temperature increase increases -33- the rate or duration of this reaction, relative to the rate or duration of development as a whole; while for other bristles a temperature increase decreases the rate or duration of the reaction, relative to the r a te or duration of the development as a whole. If such differences in rate and duration can be shown, then the differential bristle-temperature curves can be formally explained; but Child's data are not critical on this point. 5. Relation of TEP to bristle embryology. Robertson (1936) studied the pupal embryology of D. melanogaster. He found tha.t pairs of trichogenic (liypo- derma.1) cells, which later form bristles, first appear in the hypoderm about 27 hours after pupation. into bristles within a very few hours. They develop Child's TEP is prior .Q.. to this period by about two days, and is therjore net related to any visible and strictly bristle embryolo g ical process. 6. Pattern in scute. Contrary to Plunkett's demonstration of a pattern in Dichaete, Child found no evidence of a bristle pattern in sc; and this in spite of the fact that the frequency of a given bristle is constant in an inbred population as long as conditions remain constci.nt. Observations on individual flies, and correlation and association coefficients for the several bristles showed clearly that the bristle frequencies are independent of ea.ch other over the viable temperature range. The temperature effect did not appear to be that of a pattern, either. The mean bristle frequencies changed -34- independently of each other as the temperature was changed. Child pointed out that in the light of these facts, the seriation of sc-affected bristles according to their mean frequencies, as proposed by Sturtevant and Schultz (1931), is of no real significance; and that the complex five-division diffusion pattern proposed by Goldschmidt (1931) does not exist in sc-1. It is apparent from Child's results that when it is used in reference to sc, the term "pattern" carries no further meaning than "the bristles affected by sc 11 • ?. Inadequacy of the sub-gene hypothesis. Dubinin (1929), Dubinin and Friesen (1932),and a number of other~~ € have studied the extensive series of sc alleles in reference to the relation of their bristle patterns. From these studies ca.me the well known theory of the centers of the gene. It is possible to arrange all the bristles affected by the sc alleles in a linear seriation of such an order that under a g iven set of conditions each of the several alleles affects a certain number of the seriated bristles, in each case consecutively. The entire seriation was considered by these workers to represent the map of the basi-gene of the scute locus. Boundaries of the alleles, as indicated by the bristles thE:>7affected, were assumed to indicateµ the liimits of subdivisions or centers of the basi-gene. A study of these alleles revealed twelve such centers, or sub-genes, in the sc basi-gene. Each allele contained one or more consecutive centers. Child showed that when sc-1 flies are raised at -35a lower temperature than was used in these studies of Dubinin and others, sc-1 affects bristles which are in some cases several centers removed from the supposed limits of sc-1, and fails to affect bristles in centers in which the sc-1 effect is strongest at higher temperatures. the same thing true in sc-5 (Child 1936). He found From this he concluded that the whole hypothesis of basi-gene and subgenes, which had already been shown inadequate by the studies of Sturtevant a.nd Schultz (1931) on the effects of extra sections of the X-chromosome on the bristle frequencies, i1 of no significance for an explanation of the relation of the sc alleles. 8. Temperature relations of sc-1 and sc-5. Child (1936) reported studies with sc-1, sc-5, and their heterozygote at temperatures of the viable range. He found, as might be expected, that the bristle frequencies in sc-5 and sc-1/s c-5 showed v~rying types of changes along the temper ature range. For each bristle, however, the di- rection of frequency change with pro gressive temperature change was the same in each of the three stocks, so far as it could be measured. And at all temperatures, for all bristles whi ch differ in frequen cy between sc-1 and sc-5, the heter ozygote showed intermediat E bristle frequencies. A preliminary study indicated the probability that the TEP's and the duration of development differ between the three stocks. - 36- 9 . The TEP a t hi gh t emper a ture. In 19 35, I pub lished a prelimina ry account of the results of a first series of experiments with sc-1, in which the data of the ocellar bristles showed that for brief exposures to 40°C., this bristle of sc-1 has a TEP which extends from early in the embryonic st a ge into the early pupal period. These exper i ments ha ve been extended and will be presented a nd discussed in Section II of this report. It is sufficient to note here that they indicate the probability that there are indeed reactions going on throughout the bristle developmenta.l period which affect the bristle frequency in sc, and which can be altered by a change in developmental temperature. By this fa.ct, the data of these experiments make necessary a strictly qualified interpretation of the generc,l meanin g of viable temperature TEP studies relative to the time of differentiation of ima g inal organ-formin g material. SECTION II THE EXPERIMENTS WITH SCUTE AlID WILD-TYPE -37- A. Introduction. The general results of Child's viable temperature studies on sc were discussed in the last part of section I of this report. In addition, a brief summary was given of an earlier report on the first of the high temperature experiments which will be discussed in this section. After a preliminary investigation in the fall of 1934 had indicated that a short exposure of sc larvae to 40°c. produced changes in the male bristle frequencies of the subsequent adults, I planned a series of experiments to determine the bristle effects of such exposures during any one of a number of developmental time periods. These 40°c. experiments involved sc males from each of two matings; sc females and sc females x x vermilion carnation (v car) males, sc males. A second series of experiments tested similarly the effects of 36°c. on bristle frequencies in sc males. A third series of experiments, mostly at 36°C., tested the effects of such exposures on the bristle frequencies of females heterozygous for sc, and on the bristle frequencies of both sexes in a closely related wild-type stock. In connection with these three series of experiments, data were also collected at 25 to 26°c. on the effects of envi~onmental factors, such as larval density a nd the food preservative, Moldex, and of probably genetic modi- - 38fiers, on the b ristle fre q uencies in sc ma les. The Jresentation and discussion of these data will follow a discussion of the experimental materials and technique. -39- B. Materials and Methods. 1. Stocks. The stocks used in these experiments were the standard scute-1 (sc) and vermilion carnation ( v car) stocks of this laboratory. Each was inbred and selected for six generations before the beginning of the experiments late in 1934. The sc stock was selected for the absence of the bristles considered in these experiments (see below); and the v car was selected for the presence of the same bristles. Inbreeding consisted of using one or two pairs of selected sibs as parents in each generation. The wild-type stock was derived form these two mutant stocks. generations. It was inbred without selection for 13 After the period of inbreeding and selection, these stocks were continued generally by 20 pair matings and without extensive selection through the duration of the experiment. This amounted to three years for sc and v car, and to one and one-half years for the wild-type. 2. Relation of control and experimental flies. Selection and inbreeding made each stock approximately isogenic. In order to mimimize the effects of residual heterogeneity and of possible new mutations, the experiments were planned so that comparison co u/g;i\ be made between control and experimental flies which were sibs. A group of 15 to 20 pairs of flies, themselves sibs, were allowed to oviposit in each of a number of cultures. The cultures were then divided into control and experimental -40- lines. From 3 to 9 groups of sib parent flies were carried along simultaneously in this manner as one series. The general procedure was to age the parents two days to attain high fecundity; to transfer them, then, to new cultures three times a day for the next six days; and finally, to use two of each day's cultures in experimental lines while kee ping the third as control. This made it pos- sible to compare sibs that were raised to gether in ti~e and under similar environmental conditions. were always made on that basis. First comparisons However, when their bristle frequencies showed no significant differences, control cultures of two or more consecutive series were grouped as common controls for each of their experimental lines. 3. Incubators and temperature measurements. The incubators used in the experiments were of the standard insulated type designed by Bridges (1932), using a toluene-mercury thermostat, employin g a constantly running fan to minimize temperature variations, and containing pans of water to maintain a high humidity. The control incu1)ator was capab l e of maintaining a temperature of 25°c. with a constancy of !o.1°c. for any one shelf. Because of a difference between shelves, control and experimental cultures in each series were raised intermingled on the same shelf. Variation in room temperature and in the number of cultures in the incubator caused fluctuations in the incubator temperatur e which, throu gh the three years of the experiment, with a few exce ptions, amounted to not more than ~0.5°c. This made necessary a const ant check of con- - 41 - trol bristle frequenci es. \Vhe r e the r e was a d if f ere nee in the controls of two series which had to be compared, allowance has been made for the difference. The temperature in the control incubator was read from the air on standardized thermometers. Tests showed that the culture temperature was generally above the air temperature, by as much as 0.5°C. when the incubator was crowded with cultures. In extreme cases, this dif- iculty was overcome by lowering the air temperature, and by critical use of the control bristle frequencies discussed above. The exposures to hi gh temperatures were made in two smaller incubators, similar in construction to the control incubator. Temperature was read from a thermometer placed in the middle of the culture food, where the larvae were observed to congregate. The thermometer remained in the food throu ghout the exposure and frequent checks of the temperature were made. Because of rather wide variations in temperature conditions in different parts of the incubators, it was possible to place the cultures in a warmer position until the food temper a ture had reached the desired level; and then to shift them to a cooler position to maintain the temperature. Under these conditions, the tempera- ture rose from 25°C. to either 36 or 40°c. in about one hour. The number of cultures exposed together varied between three and five in the 40°c. experiments and between six and nine in the 36°C. experiments. The temperature was read from one or more cultures of each exposure group. Tests -42- showed that there was a va riation in the temperature between cultures of such a group. This amounted to about 0.5°c. in a group at 4o 0 c. and to about 1° in a group at 36°c. The range of variation was larger than this for the many groups of an experiment. In all the groups of the sc x v car 40°c. experiment this variation was between 39.5 and 41.5°c. In the sc x sc experiment, it was between 39.5 and 40.5°C. In a=ll of the 36°c. experiments, it was between 35 and 37oc. At these higher extremes the food temperature tended to be lower than the air temperature. The differ- ence was usually about 1° in the 4o 0 c. series, and about 0~5° in the 36°C. series. This, it will be noted, is 9p- posite to the tendency at 25°c. 4. The bristles and their classification. Flies were classed for seven pairs of bristles, each pair being represented on the left and right sides of the fly. The bristles, with their abbfeviations and loca- tions, are as follows: posterior scutellar {ps) and anter- ior scutellar (as) on the scutellum; anterior notopleural (an) on the anterior lateral region of the thorax; postvertical (pv) on the medial posterior dorsal part of the head; ocellar (oc) on the top of the head, between the ocelli; and anterior and medial orbital (or) on the anterior dorsal rim of the eye. Because of an iwpossibility of distinc- tion between the anterior and medial orbital bristles in sc:l, those two bristles were classed together as "orbital" (or). In the sc classed. x sc experiments, only sc males were In experiments with sc females x v car males, -43- and in experiments with wild-type, bristles were counted in both sexes. The sc males were classed for the frequency of presence of the above bristles. The sc/v car females and both sexes of wild-type flies were classed for the absence of the same bristles. Each fly was recorded individ- ually with respect to the presence or absence of each of the above bristles. 5. Calculation of bristle frequency. As in Child's experiments, practically no correlation was found between the two bristles of any one pair, in both sc and wild-type flies, when the conditions were properly controlled at 25°c. Accordingly, since each half of the fly can have one bristle of each pair, and since the desired bristle frequency is that of the ratio, total observed number of x bristles total possible number of x bristles -,,;;;..-'--------------------..;.....=-,,;;....,;;a...-......,,.-=~..;;..;-'-~=----' the "half-flya unit, proposed by Plunkett (1926) and adopted by Child, has been used in these experiments. As an example we may consider the oc frequency. Each fly represents two observations upon the oc bristle, one on the left side and one on the right side. Each half of the fly, then, is one obser~ation upon the oc bristle; and the oc bristle frequency per half fly in a sample is that of the ratio, the number of oc bristles. twice the number of flies The maximum frequency is 1 and the minimum frequency is The frequency therfore, can be stated as a percentage by multi plying the above ratio by 100. o. -44Al l mean f r eq uencies have b een calculated in the above manner and t ab ulated as percentages. 6. Tabul a tion of differences in bristle fre q uencies. Certain of the t a bles of d a ta which will be discussed in this report re p resent summaries of b ristle freouency changes produced by temper a ture tre a tment. As a convenient method for tabulating these summa ries, I have chosen the value of the ratio, difference between control and exnerimental freguenc_y ___ control frequency which I h a ve named, the "Coefficient of Change". The sign of the coefficient indicates the direction of the change. It can be seen that when the control fre q uency is O and the 5 experimental frequency i/ g reater than O, the v a lue of the coefficient is infinity. Since the purpose of the coefficient is to make possible an easy and direct comparison of degrees of chan g ~ in bristle fre qu encies, in the one or two instances where such a situation as the above has arisen in the data and the increase appeared to be of some si g nificance, the control frequency has been argitrarily put at 0.1. Infinity, however, represents the max imum of incre as e as expr e ssed by the coefficient of change. Similarly, -1 represents the max- imum of decrease, and is observed when th e experimental fre• quency is O and the control freo_uency is greater tha n O. '.!lb.en there is no difference between the control and the exp erimen~- al frequencies, the v a lue of the coefficient of chan g e is 7. Statistic a l formulae. Standar¢d error has been employed in all data considerations for estimating the pro babil ity of differ- o. -45- ence between mean bristle frequencies. The approximate standard error (s) of a mean bristle frequency is given by s = ~v p(l-p) n where pis the mean bristle frequency and n is the number of half-fly observations. The standard error of a differ- ence (sd) between two mean bristle frequencies, p 1 and p , 2 is given by 52 il where s 1 and s 2 are the standard errors of p 1 and p 2 respectively. In general, a difference has been considered significant only when it equals or exceeds twice its standard error. A difference which is smaller than that amount will occur by chance alone more often than once in twenty comparisons between samples drawn from a uniform population. For an estimate of correlation (r) between bristles I have used the formula r = S (xy) where x and y refer to the bristles under consideration. The "t" test of Fisher (1928) has been used to estimate the significance of the deviation of r from o. The value oft is determined from the formula t = r :::=::::::::;:. • Vn 2 1/1 - r 2 where n is the number of pairs of observa tions ( in this case the number of flies) in the sample, and r is the cor- -46- relation coefficient. When the size of the sample ex- ceeds 100 flies, r may be considered as significantly different from O when t equals or exceeds 2, since 2 is approximately the "l in 20" level. The more that t ex- ceeds 2 , the lar ger the chance of significance in the deviation from o. -47- c. The Effects of Environmental Ractors and Genetic Modifiers. 1. The independence of bristles. It has already been noted that Child found a complete absence of pattern relationship in the sc-affected bristles. He found this by observations on individual flies, by absence of correlations between the bristles, and by the small size of calculated association coefficients. So far as they go, my data are in complete agreement with his on this point. Checking of the indi- vidual flies indicated clearly that only in the special case of the anterior and medial orbital bristles was there any association between the bristles with which I have concerned myself. Child reported a complete negative as- sociation between these two bristles and ovserved no instance of their simultaneous appearance on the same side of the fly. In some 500,000 half-fly observations, I found only one such instance. Extensive correlation coefficient calculations were made from the male control data of the 1936-1937 experiments with sc x sc and sc x v car. These correla- tions included all the two-by-two combinations of the six bristles: ps, as, an, pv, oc, and or. The value of r ranged between -0.610 and •0.040 in the sc males from sc x sc; and between -0.007 and 40.074 in sc males from sc x v car. While the t values exceeded 2 for the larger positive correlations, because of the large numbers of -48- flies observed, the correlations are too small to be of significance for these experiments. There is not the ten- dency for covariance among the bristles that one would expect if a definite pattern realationship obtained among them. of 2. Effects of crowdingAlarvae. Plunkett (1926) included a table of data to show that increasing mean number of larvae per culture caused a decided decrease in the mean number of posterior dorsocentral bristles in the subsequent Dichaete adults. He at- tributed the effect to a change in food such that the later larvae from eggs oviposited over a period of several days were raised under different conditions than were their earlier sibs. The effect disappeared when an egg-laying per- iod of one day or less was used. Child publis~ed no data on crowding effects in sc. He used short laying periods, however, and kept the larval density of a culture low. Early in the course of my experiments it was noted that larval density affected bristle frequency, even when the egg-laying period did not exceed 12 hours. In addition, it was apparent that not all bristles were affected similarly. In Tables 1 through 7 are presented samples of the control temperature data from sc males to demonstrate these facts. Each table is a summation of a number of con- secutive series which did not vary significantly within themselves. The cultures comprising each table have been -49- TABIE 1. v car o'c>' - Nov. 1934 Obser. ps as an pv oc or 1274 0.1 !O.l 1.7 !!!l. 4 0.6 :!0.2 0.9 :!O. 3 4.2 :!:0.5 7.9 :!:O. 8 147 (108 -190) 4398 0.7 ~0.1 2.8 :!0.2 0.3 :!0.1 0.7 :?O.l 3.9 :rn. 3 4.2 :to. 3 239 (209 - 290) 2874 0.7 !0.2 3.7 !0.4 0.5 :!:O. l o.7 !0.2 3.4 :!0.3 3.0 :!0.3 SC Size 76 (41 - ~9 X 99) 25°c. TABLE 2. SC '?Y X V car dd' - May 1935 25°c. Size Obs er. ps as an pv oc or 141 1388 0.3 !O.l 2.6 ±0.4 0.2 ±0.1 0.4 ~0.2 3.4 !0.5 7.1 ±0.7 255 (203 294) 2762 0.65 !0.15 4.3 !0.4 0.4 !0.1 0.5 !O.l 2.6 !0.3 4.6 !0.4 375 (315 421) 1848 1.2 !o.3 4.7 !o.5 0.9 !0.2 0.65 !0.2 3.3 !o.4 4.3 !o.5 (72 - 189) - -50- TABLE 3. SC 'f ~ X SC d'd' - Oct. 1936 25°c. Size Obs er. ps as an pv 00 or 125 1916 0 0.4 :!:O.l 0.6 :!:0.2 0.6 :!:0.2 3.9 :!:O. 4 17.0 :!:O. 9 226 275) (183 3388 0.06 :!:0.04 0.3 :!O.l 0.5 :!1J.l 0.7 :!O.l 4.4 !0.3 12.3 :!0.6 315 366) (295 2788 0.04 !0.04 0.4 !O.l 0.3 !O.l 0.8 !0.2 2.1 !0.3 6.6 !0.5 pv QC or (57 - 182) - - --TABLE 4. 1937 25°c. X scd'd" - Feb. Size Obs er. PS as an 112 2096 0 0.5 + -0.1 o.s :!:o. 2 + 1.4 -0.3 3.1 + -0.4 + 173 (143 - 199) 2102 0 .2 .-0.1 0.8 '!0.2 0.4 !0.1 1.4 !o.3 2.8 !o.4 11.0 !o.7 248 (201 - 302) 5636 0.04 :!0.03 0.8 !o.l 0.8 '!O.l 0.9 !O.l 3.5 !0.3 9.4 '!0.4 SC 9 'f (72 - 139) 16.3 -0.8 -51- TABLE 5. ~~ SC X SC o'o" - March 1937 25°c. 1% Moldex pv Size Obser. ps as an 110 1360 + e.1 -0.1 1.1 + -0.3 0.8 + -0.2 + 0.9 -0.3 171 198) (140 6190 0.1 +0.04 1.0 •0.1 1.0 •0.1 242 2 96) (200 59?8 0.2 :!:0.06 1.2 !O.l _, (82 - 139) - oc or + 2.5 -0.4 9.9 + -0.8 1.1 •0.1 2.8 •0.2 6.9 •o.3 l.4 !0.2 0.8 ±0.1 3.8 ~0.2 4.8 -0.3 25.5°c. o.7% Moldex ~ TABLE 6. SC <?-9 SC X d' <3' - April 1937 Size Obser. ps as an l)V oc or 108 3538 0 0,1 !0.05 o.s !0.2 0.7 :to.1 4.2 !0.3 13.5 !0.6 171 (142 196) 9122 0.09 !0.03 0.3 !0.06 0.8 !0.1 0.7 !O.l 5.0 :!:0.2 7.8 :!:O. 3 240 (198 286) 7282 0.04 !0.02 0.6 !o. 1 0.9 !0.1 0.4 :!:O. l 5.0 !o.3 5.1 :!0.3 (68 - 134) - - TABLE 7. SC 9 9 X v car c:fd" - May 1937 2 6°c. 0.7% Moldex Size Obser. ps as an PV oc or 196 (152 218) 1316 0.8 :t0.2 4.3 :t0.5 1.6 :t0.3 2.0 :t 0.4 10.6 :t0.9 3.3 :to. 5 278 (268 288) 1918 1.0 :!0.2 6.6 , 3.0 to.4 1.4 !0.3 10.6 t0.6 to.? 2.3 !0.3 365 (324 420) 1764 l.? :!:0.3 8.2 !0.7 3.6 !Q.4 1.1 :!0.3 11.8 :!0.8 2.7 !0.4 - - -5 2 - grouped according to 11 size 11 ; i. e., the number of flies reco~~ded per culture. Included in each table are: the mean culture size, including in parentheses the sizes of the smallest and largest cultures in the group; the number of male half-fly observations; and the mean percentile bristle frequencies, with their standard errors, for each of the six bristles consider~d in this work. At the head of each table is a statement of pedigree which includes the mating, the date in year and month, and the temperature and food conditions (discussed later) of the included cultures. An examination of Tables 1 through 7 shows that s.orne of the bristle frequencies changed as the mean culture size increased. A considerably larger amount of con- trol data from conditions similar to those of Tables 1 through 4 amplify the data of these tables. From a con- sideration of all the data one dan make the following general statements concerning the relation of the several bristles to larval density. In the matings of sc x v car, the ps, as, and an bristles tend to increase in frequency as the mean culture size increases; the or bristles tend to decrease; and the pv and oc bristles show practically no change with change in size of culture. In the sc x sc matings, the only clear change is a decrease in the or frequency with increasing culture size. Left-right correlation coefficients were calculated for e a ch of the six pairs of bristles in all of the control data. In no case was there any tendency for a change in the value of r for any pair of bristles as the -53mean culture size increased. This indicates the probabil- ity that in these experiments the effect of increasing larval density was on the entire culture population, rather than on a part of it. This is probably due to the short egg-laying periods used in these experiments. Correlation coefficients were also calculated for each of the two-by-two bristle comparisons in these data. In no case was there a correlation above 0.1. It seems probable therefore that there was no significant tendency for covariance among the affected bristles. The data represented in these tables indicate clearly that larval density is a factor which influences some of the mean bristle frequencies in sc males. Accord- ingly, in the data to be presented on the effects of exposures to 40 and 36°C., I have made comparisons between experimental and control series with a,s nearly the same mean culture size as was possible. 3. Effects of Moldex. In February of 1937, the food formula of this laboratory was changed to include a mold preventitive known by the trade name, "Mold ex" ( Glyco Products Co., New York). is a methyl ester of parahydroxybenzoic acid. It In this lab- oratory it is kept in a stock solution of 10 g. of Moldex in 90 cc. of 95% ethyl alcohol. This solution of Moldex constituted 1% of the food for a month, after which time the proportion was lowered to 0.7%. The cultures of Tables 5 through 7 contained Moldex, as indicated in the table headings. Tables 4 and 5 are -54- expecially instructive concerning the effect of Tuloldex on the bristle frequencies. Table 4 cultures conta.irJed no Moldex, while Table 5 cultures conta.ined the 1% Moldex formula. So far as is known other conditions were alike for the two groups of cultures. The main effect of the Moldex, as revealed in these two tables, is on the or Bristle, whose frequency is clearly lower in the Moldex series. In this mating, Moldex does not appear to affect the other bristles significantly. The data in Tables 6 and 7 show the effects of the 0.7% Moldex. There are differences in the temperature conditions in these two tables which make necessary certain corrections before the data ca.n be compared with those of lower temperature series. Occasional small series of the non-Moldex experiments had temperatures comparable to these. From them it appears that in sc x sc the effect of the slightly higher temperature was mostly on the oc bristle, and that the main effect of the 0.7%, Moldex was to decrease the frequency of the or bristle. There does not appear to be a significant difference between the effects of the two strengbhs of MoJ.dex. The effect of Moldex on the males from sc x car is more extensive than it is on the males from sc This is brought out in the data. of Table 7. v x sc. From series raised at simil2.r temperc,tures in the earlier non-Moldex experiments it was noted that the effect of the temperature is manifested in decreases in the ps, as, and pv bristles; and in increases in the an, oc, and or bristles, the oc - 55incre a se b ei ng mo s t mar ked . A compa rison of the d a t a of Table 7 with those of Tab le 2 , its ne a rest control in point of time, r eveals tha t the Mol d ex seems to have incre a sed the ps, as, and pv fr equencies, and to have decreased the or frequency. In each ca se, t h e fre quency chang e is the opp osite from what would have been expecte d f rom th e effects of the temper a ture difference alon e . the sc x Thi s prob ably mea,ns tha t in v car ma ting , the effe c ts of 1Ioldex a re even more pronounced than indicated in these da ta of Table 7. Left-ri ght correl a tion coefficients of the affected bristles ',n ere compa red in Mold ex and non- Moldex seri e s. There were no consistent differences between the two series. This indicates that the effect of the Iioldex wa s upon the entire popul a tion. Two-by-two correlation coefficients of the dif- ferent 'cristles wer e calcul a ted and were g enerally less than 0.1, indic a ting that the bri s tles did not tend to v ary together under the influence of Moldex. With a vie w to in d ic a tin g a possible exp lan a tion for the effects of Moldex upon the bristle frequencies of sc, a rou gh test wa s ma de of the effects of Moldex upon the growth of yeast. A g lucose-ye a st extract media was made up to contain, in one culture, 1% of the Moldex solution, and in another culture of the same volume, 1% of 95% alcohol (the b a se of th e stock k oldex solution). then The cultures were innocul 8 te d with equ a l araounts of a yeast sus pension. At the end of non-l\!Ioldex 10 days the dry wei*ht of the yeast in the cultures WEJ-S sli ghtly more than twice that -56in the l i~ 1do l d ex cu l ture, ind i cati ng that Tuioldex decreases the growth rate of yeast. This test suggests the possibil- ity that the effect of Moldex on the bristle frequency is indirect, and that the lower per capita supply of yeast for larval consumption may be the causal factor. The effects of Moldex and larval density, it can be seen) are in the same direction on at least all but the pv bristles. Accordingly, we may offer as a suggestion the possibility that the effect of larval crowding is likewise due to a decrease in the per capita supply of yeast for larval consumption. Child and Albertowicz (1937) reported in abstract that larvae developed ~more slowly'' on food treated with Nipagen, which is ethyl para hydroxy benzoate,and, as can be seen, is closely related to Moldex. Increasing the con- centration of Nipagen lengthened the time of development. They found that vg flieB which were raised on Nipagen developed larger wings resembling the less extreme alleles of vt raised under normal conditions. The size of the wing increased as the time of development was prolonged by the use of Nipagen. No measurement has been made of the effect of Moldex on the length of larval development, excepting to note that there did not appear to be any large difference in the time of pupation when Moldex was added to the food. The data are not critical, however, for differences of less than half a day. One is tempted to suggest, even in the absence of data, that the effects of Moldex and Nipagen may both be -57- upon the yeast, a lower per capita quantity of which delays larval development and gives the vg and sc genotypes d- correspondingly longer time to affect the adult characters. carefully controlled experiments should indicate whether or not the situation is as simple as this. 4. Effects of genetic modifiers • ..9-.. In his se\ction experiments with low and high bristle frequency strains of sc, Payne (1920) found at least three, and probably more, genetic modifiers of the effect of sc on the scutellar bristles. One modifier was sex-linked (near miniature) and the others were autosomal. In the data of Tables 1 through 7 there is evidence for the presence of genetic modifiers in my experimental stocks. Additional evidence will be pointed out from the 40 and 36°C. series. A comparison of the data from sc those of sc x x v car with sc in Tables 1 through 7 shows that three of the bristle freouencies differ between the two matings. In the sc x v car mating, the ps and as frequencies are higher and the or frequency is lower than comparable frequencies in the sc x sc mating. No attempt has been made to localize the factor or factors responsible for these differences. Since both sc and v car were selected and inbred for several generations previous to the beginning of the experiment each stock should have been practically homozygous for its modifiers of the selected bristles. The cal- cul ati on of left-right correlation coefficients showed tha: J -58- this condition had been realized in the sc stock and for all bristles excepting the as bristle in the v car stock. In the sc male as bristle data from the sc x v car mat- ing there was a persistent though small positive left-right correlation in both the 1934-1935 and the 1937 series. The value of r ranged between .. 0.1 and -£0.2 (with t values of from 4 to 9) in all the series of this mating. not the case in the sc x This was sc mating where the value of r was sometimes positive and sometimes negative, and rarely as large as 0.1 in any case, It seems probable therefore that in this one case the v car stock was not homozygous for a sc modifier. Since the stock did not appear to change with time in respect to the proportion of flies carrying the modifier, the fact of presence does not complicate the comparison of control and experimental data. It is important to know if the sc and v car stocks remained fairly const~nt genetically during the couise of the experiments. Data bearing on this question are those of Tables 1 and 2, and 3 and 4. represent series of sc eral months a.part. sc x x In the first case, the data v can flies which were raised sev- In the second case, the data represent sc flies which were raised several months apart. It appears at first that in both cases there was a small change in the stocks with regard to the or bristle frequency. The or bristle frequency appears to have increased with time in the sc x v car mating, and to have decreased with time in the sc x sc mating. If this were due to a change in genetic constitution of the stocks, one would expect, assuming the change to have been caused by a mutation - 59- or mutations, that only a relatively small proportion of the stock in each case would be responsible and that, accordingly, the left-right correlations would show evidence of the change. As a matter of fact, the left-right cor- relations are not significantly different between the two series in each mating. The r values for the or bristles are •0.010 and+0.015 for the data of Tables 1 and 2, respectively; and •0.056 and +0.026 for the total data of Tables 3 and 4 respectively. It seems probable, there- fore, that the small changes in the or frequency in each case represent some slight and unknown change in environmental conditions which has affected the entire population. The data of the Ivioldex series of Tables 6 and 7 are not directly comparc:,ble with each other because of the difference in developmental temperature discussed above. But when allowance is made for that difference, these data indicate that the bristle frequency differences between the two matings are in the same direction as in the non-Moldex series. There is the additional pos sib ility that the dif- ferences were accentuated by the Moldex; but more critical data than these are necessary to establish this point. -60- 5. Summa.ry and Conclusions. (1) There is no evidence in the data from the 25 to 2s 0 c. series of these experiments that a pattern relationship exists between the bristles affected by sc, or that they tend to vary together in frequency of presence. (2) Larval density influences some of the bristle frequencies of sc males under different constant environmental and genetic conditions, increasing the ps, as, and an frequencies in sc in x both SC v car matings, and decreasing the or frequencies X V car and SC X SC matings. (3) The presence of 0.7% or of 1% of Moldex in the food increases the frequencies of ps, as, and pv bristles in the sc x v car mating and decreases the fre quencies of or bristles in both SC X V car and SC X SC ma tings. (4) The differences in bristle frequencies in sc males from sc x v car as compa red to those in sc males from sc x sc suggest that the sc and v car stocks differ genetically in modifiers of sc-affected bristles. (5) The sc and v car stocks do not appear to h ave changed significantly in genetic constitution during the course of the experiments, so far as the bristle fre quencies are concerned. (6) The results of these studies on environmental and genetic factors demonstrate the necessity of controlled conditions in experimental work with sc bristle frequencies. - 61 - D. The 40__:>_g_, ExI2,_eriments with sc Males. 1. Genera l plan of the experiments. Two experiments h ave been c a rried out to test the effects of brief exposures to 4o0 c. on the bristle frequencies in sc ma les. The ma ting s of these series were, in the first experiment, sc fema les x v car males, a nd in the second experiment, sc fema les sc males. x The gen- eral control of conditions h as already been discussed, and sample control da t a h a ve been presented from each of these experiments (Tables 1 through 6). The only signific a nt difference in condition b etween control a nd experimental lines wa s in the exposure to the hi gh te mpera ture a t a given period in development. The exposures to 4o 0 c. were ma de durin g six development a l periods: 0 to 10 µours, 20 to 30 hours, 48 to 58 hours, 72 to 82 hours, 96 to +06 hours, and 120 to 130 hours after oviposition. The ten hours in ea ch p eriod is the maximum sum of the egg-laying a nd exposure periods in the expe riment a l cultures in ea ch ca se. :No a ttempts were ma de to determine ana tomic a lly the exa ct morpholo g ic a l period re presented in e a ch of the ab ove ti me periods. It was noted, h owever, tha t the ma jor- ity of the l arva e of t h ese series, in not too crowd ed control cultures (under 200 l a rva e), pup a ted between 110 and 120 hours after ovi posit i on. A number of investi gators, among them Dobzhansky a nd Dunc a n (1933) in this l aboratory, h ave re p orted a pupation ti me for D. me l a no gaster more or -62less simil a r to the above a~temperatures of 25 to 26°C. On the basis of these investigations on morphological developmental periods, the above time periods of my experiments may be designa ted approximately as: the embryonic period; the first larval instar; the second instar; the first half of the third instar; the second half of the third instar; and the early pupal period. From the studies of Robertson (1936) it a ppe a rs tha t these periods cover practically the entire developmental period of the bristles. A graded series of exposures to 40°C. were made during the ab ove periods. The hour necess a ry to raise a culture temperature from 25°c. to 4o 0 c. will be referred to as 11 a 11 • Were •. 11 Including this hour, the series of exposures used a 11 , 1 -2 , 1, 1-21 ' 2 ' 2-21 ' and 3 hours • In the sc x v ca r experiment all of these ex- posures were used in the l a rva l developmental periods; but in the embryonic a nd young pupal periods the h a lf-hour ex- posures were omitted. In the sc x sc experiment, genera lly only the 1, 2, and 3 hour exposures were used. 2. The data from sc x v car. Tables 8 through 13 present the bristle d a ta of sc ma les from t h e mating , sc fema les x v car males. The experiments were done in the school ye ar of 1934-1935. At the he a d of each t able is a statement of the ma ting, the exposure tempera ture, a nd the development a l period, morphologically and in hours from oviposition, during which the exposure was made. Included in the tables are: the length -63- TABLE 8. SC XV car - 40°c. - Embryo, 0 to 10 Hours Development Group Size Obser. PS as an pv oc or "a 11 ;1; 2 42 1216 0.1 !Q.l 2 .4 :!0.4 0.8 :to.3 0.6 :to. 2 11.3 :to.9 12.8 :ti. 0 Control 14 1 1388 0. 3 !0.1 2 .6 !0. 4 0.2 !0.1 0.4 !0.2 3.4 :!0.5 7.1 !0.7 -0. 2 ·, !O. l -0. 2 !0.6 +0. 6 :!0. 3 +0. 2 !0.3 +7.9 :!l. 0 +5.7 !1. 2 Difference -64- TABLE 9. X SC V car - 40°c. - First Ins t&,r; 20 to 30 Hours - Group Sj.ze Obser. ps as an pv oc or llall 192 4286 0.5 !0.1 2.5 :!0.2 1.1 :!0.2 0.5 :!O.ih 5.2 :!0.3 4.7 :!0.3 Control 203 4658 0.5 .-0.1 3.0 + -0.3 +0.9 -0.l 0 .,3 + -0.1 +4.2 -0.3 3.3 + -0.3 0 .;.o. 5 -0.4 . "'O 2 -0.2 +o + • 2 -0.1 +1.0 -0.4 . "l + •4 -0.4 Difference ½ Control . . 133 1844 0,3 ~0.1 3.4 ~0.4 0.2 ~0.1 1.1 :!0.2 9.8 !0.7 7.5 :!0.6 147 4398 0.7 :!:O. l 2.8 :!0.2 0.3 :!O.l 0.7 :.to.1 3.9 :!0.3 4.2 :!0.3 -o.4 !0.1 •o.6 "-0.4 -0.1 !O.l +o.4 :!0.2 "5.9 :?o.s +3.3 ?o.7 Difference 1 78 1514 0.2 !0.1 2.6 :!o.4 0.6 !0.2 0.8 !0.2 7.0 -!.o.7 7.4 '!o.7 Control 76 1274 0.1 ~O.l 1.7 _•o.4 0.6 ~0. 2 0.9 ~0.3 4.2 ~0.5 7.9 ~0.8 +0.1 ... o. 9 0 -0.1 :tO.l :te,. 6 J.2.8 :!0.9 -e.5 :tl.O Difference (Continued on next page) :t0.4 - 65- TABLE 9. (Cont.) Grou p 3 iz e 0b ser. }2 S as an J2V oc or l½ 114 144 0 0. 3 ! O.l 2 .0 ! 0. 4 0. 9 !0. 3 0.4 :!0. 2 5. 8 !0.6 9. 2 :!0. 8 Control 11 2 567 :-l. 0.4 :! 0 .l 2 .3 ! 0 .2 0.5 :!0.l 0. 8 :! 0 .1 4 .0 :!o. 3 6.0 !0.3 - 0 .1 ! 0 .1 -0.3 !o.5 ... o. 4 ! 0. 3 -o. 4 ! 0. 2 ... 1.8 !0.7 .. 3 . 2 ! 0. 9 Difference 2 127 1474 o.5 ! 0. 2 3 .7 ! o .5 0 0.1 :!: 0 .l 4 .7 :!:o. 6 5. 8 :!: o . 6 Control 14 7 4398 0 .7 :! 0 .1 2. 8 :!:0. 2 0. 3 ! 0 .1 0 .7 :!:0.1 3. 9 !0.7 4. 2 :!: 0 .3 -0. 2 ! 0. 2 "' o . 9 :!o. 5 -o. 3 :! 0 .1 -o.6 !O.l "'o. s :!0. 9 ... 1. 6 :! 0 .7 Difference 2½ ;3 1 23 1540 0. 2 ! 0 .1 0 .7 ! 0. 2 1. 0 ! o. 3 0.1 - 0 .1 15. ... 5 -0. 9 ...- 80 .1 .7 Control 16 6 22 14 0. 3 ~0 . 1 0. 6 ! 0. 2 1.1 :! 0 . 2 0.1 ! 0 .1 8. 8 ::!o. 6 5. 1 :! o . 5 ·0 .1 :!0 .l .10.1 • J .l :!0. 4 0 Di f fer en ce :!0 . 3 1' .J.6. 7 J. 3 . 0 :!l. 1 :!0. 9 -66- T.ABLE X SC V car 10. - 4o c. - Second Instar; 48 to 58 Hours 0 Group Size Obs er. ps as an PV oc or II all 131 894 0.3 ±0.2 1.5 !0.4 0.2 !0.2 0.2 !0.2 3.1 !0.6 !l.O 0.4 !0.1 1.7 :!0.3 0.3 !0.1 0.4 !0.1 2.8 !o.4 5.1 !o.5 -0.1 ~0.2 -0.2 :!0.5 -0.1 :!0.2 -0.2 :!0.2 o.3 ... .. o.7 +4.2 :!l. 1 Control 158 1800 Difference 1 9.3 2 139 1474 0 1.4 :!0.3 0.1 :!:O.l 0.1 :!0.1 5.2 :!0.6 8.7 !0.7 Control 158 1800 0.4 :!O.l 1.7 :!:o. 3 0.3 :!O.l 0.4 :!:O .1 2.8 :!0.4 5.1 :!:Q. 5 -o.4 !0.1 -0.3 :!Q.4 -0.2 to.1 -o.3 :!O.l "'2. 4 :!Q.7 "'3. 6 to. 9 Difference 1 118 978 0.3 !0.2 1.5 :!0.4 0 0 5.3 :!Q.7 9.3 :!0.9 Control 158 1800 0.4 :!0.1 1.7 :!o.3 0.3 to.1 0.4 :!O.l 2.8 !0.4 5.1 :to.5 -0.1 :!0.2 -e.2 -0.3 :!O.l .ao.4 !O. l ... 2. 5 :!0.8 .. 4.2 Difference to. 5 (Continued on next page) :!l.O -67- TABLE 10. - l½ S:i:ze Obs er. 145 860 Control 141 < Group 1838 Difference (Cont.) ps as an pv oc or 0.1 1.2 :!0.4 0.1 :!O.l 0 :!O.l 10.2 :! 1. 0 :!l. 0 0.4 !0.1 2.7 ::!o. 4 0.6 !0.2 0.2 :!0.1 4.2 ::!o. 5 6.0 :!o.6 -0.3 !0.1 -1.5 !o.6 -0.5 !0.2 -0.2 :to. 1 "'6. 0 ~1.1 3.5 :! 1. 2 9.5 4 2 142 2352 0.2 :!0.1 1. 2 !0.2 0.4 :!0.1 0.04 :!0.04 13.3 :!0.7 9.1 :!0.6 Control 150 3638 0.4 !0.1 2.2 :!0.2 0.4 :!0.1 0.30 :!0.09 3.5 !0.3 5.6 :!0.4 -0.2 :!0.1 -1.0 0 :!0.3 -0.26 :!0.10 +9.8 ... 0.8 +3.5 !0.7 0.6 ~0.2 3.0 !0.5 1.0 !0.3 0.3 !O.l 8.8 !0.8 !o.s +0.4 -0.1 2 .7 .-0.4 ,.o.6 -0.2 0.2 + -0.l +4 • 2 -0.5 +6.0 -0.6 •0.2 :!0.2 +0.3 '!0.6 .. 0.4 '!0.4 .. 0.1 '!O.l +4.6 -t3.3 !l.O :!1.0 "1. 1 -0.3 0.3 + -0.2 0.1 .-0.l 7.2 + -0.7 8.4 .-0.8 Difference 2½ 150 Control 141 1290 1838 Difference 3 130 1244 0 Control 150 3638 Difference 9.3 + 0.4 -0.l 2.2 0.3 0.4 3.5 5.6 .-0.1 .-0.4 .-0.2 .-0.1 .-0.3 -o.4 ~O.l -1.1 :!o.4 -- -0.1 :!0.2 -0.2 !0.1 "3.7 '!0.8 •2.8 ~0.9 -6 8 - TABLE 11 SC V X car - Third Instar; 72 to 82 Hours - 40°c. Group Size Obs er. ps as an pv oc or "a" 95 920 0.4 !0.2 2.0 "0 " • 0 0.8 :!o.3 0.1 -0.1 " 11.8 :!1. 1 8.8 :!0.9 Control 117 134 2 0.3 ~O.l ~ 1.9 -0.4 0.2 :!0.1 0.2 :!0.1 6.1 :!o.7 7.2 :!o.7 .ao. 1 :!0.2 "0.1 :!o.6 .-o.6 :!0.3 -0.1 :!0.1 .. 5. 7 :!1. 3 .-o.6 :! 1. 1 0.3 1.5 :!Q.4 1.1 :!0.3 0 :!0.2 10.4 :!0.9 7.0 :!0.8 0.3 !0.1 2.3 ~0.4 0 •"'') :!0.1 0.2 :!0.2 4.6 :! o .5 6.5 to.6 0 - o.s ~0.6 "0.9 :!0.3 - 0. 2 :!O.l "5. 8 :!1.0 "0.5 :! 1. 0 Difference l 2 105 Control 117 1116 1788 Difference 136 832 0.1 3.0 0.7 .-0.l .-0.6 .-0.3 0 ,.8.4 -1.0 2 .8 .-0.6 Control 134 1834 0.3 ~0.1 2.6 ~0. 4 0.2 io.l 0.4 :!0.1 3.3 !o.4 6.8 !o.6 -o ,&. • 2 -0.1 :o.4 -0.7 !0.5 -0.3 -o.4 ~0.1 5.1 ~1.1 -4 0 -0.9 1 Difference (continued on next page) 4 . . -69- TABLE 11 (cont.) Group Size Obs er. ps as an pv oc or 1-} 103 916 0.3 :!0.2 1.7 !0.4 0.9 :!0.3 0 9.3 :n.o 5.4 :!0.7 Control 117 1342 0.3 :!0.1 1.9 :!0.4 0.2 !o.l 0.2 ::!0.1 6.1 !0.7 7.2 ::! 0. 7 0 -0.2 !o.6 .ao.7 -•0.3 -0.2 :!O.l "'3. 2 :! 1. 2 -1.8 :!1.0 Difference 2 105 1280 0 .,:.,') :!O.l 1.3 :!Q.3 0.8 :!O. 3 0 11.0 :!o. 9 5.4 :!0.6 Control 117 1342 0.3 !0.1 1.9 :!0.4 0.2 ' !O.l 0.2 :!O.l 6.1 :!0.7 7.2 :!0.7 -0.l + -0.1 1 -0.5 -0.6 ... e. 6 -0.2 -0.1 . .. 4.9 ~1.2 .-1.8 -0.9 0.4 :!0.2 9.2 + -0.8 5.5 .-0.6 o.3 + -0.l 3.2 0.9 .-0.5 .-0.3 0.2 1.9 .-0.4 !0.1 0.2 :! o. 1 6.1 '!.o. 7 7.2 ~0.7 -0.1 ~0.1 "1.3 !o.6 ~0.7 :!o.3 -t0.2 :!0.2 -t3. 1 :!1. 1 -1. 7 to.9 Difference 2{- 111 1290 ... 0.2 -0.1 Control 117 1342 Difference ..- 0 • 0'Z 3 137 1242 . . o. 2 .:.0.1 2.9 :!o.5 0.4 ~0.2 0 10.l -0.9 . 5.1 .-0.6 Control 117 1342 0.3 :!0.1 1.9 !o.4 0.2 :!0.1 0.2 :!0.1 6.1 :!o.7 7.2 :!o. 7 -0.1 ;0.1 :rn. 6 .. 1. 0 .. 0.2 :!0.2 -0.2 .-4.0 :!l. 2 -2.1 :!0.9 Difference :!O. 1 TABLE 12 SC V X car - 4o 0 c. - Third Inst2.r; 96 to 106 1::0urs - Grou-o Size Obser. DS as an DV oc or fl r, II C< 152 1180 0.3 !0.2 1.2 ~0.3 0.5 :!0.2 O~l :!O. 1 3.5 :!0.5 4,:3 :!0.6 Control 158 1800 ... 0.4 _0.1 ... 1.7 -0.3 0 3 -0.1 . . 0.4 + -0.l 2 8 -0.4 . . 5.1 :!o.5 -0.1 ~0.2 '~o. 4 .1.0.2 ~0.2 -'0.3 :!0.1 "-0.7 + -0.6 .-0.8 -0.8 0 1.4 ...-0.4 0 .2 0.2 .-0.1 .-0.1 ... 5.3 -0.7 3.4 .-0.6 0.4 :!O.l 1.7 :!0.3 0.3 :!O.l 0.4 :!:O. l 2.8 :!0.4 5.1 :!:O. 5 -0.4 :!O.l - 0 • .J r,• :!0.5 -0.1 :!O.l -0.2 :!O. 1 .&2.5 :!0.8 -1.7 ' :!0.7 Diffe:rence l -2 Control 179 158 1058 1800 Difference -0.5 1. 194 1116 0.1 :!0.1 2.0 :?0.4 0.1 !Q.l 0.3 :?0.2 4.8 !0.6 3.1 ::!O. 5 Control 194 3930 0.6 !0.2 2.7 :? o. 3 0 • .J !O.l "L 0.5 !Q.l 2.5 :?o.3 4.4 !0.4 -0.5 :to. 2 -0.7 :!0.5 -0.2 :!O.l -0.2 :!0.2 .. 2.3 '!.O. 7 -1.3 :!0.6 Difference 1-!- 148 870 0.3 !0.2 2.4 :!:o.5 1. 2 :!:0.4 0.1 :!:O. l 7.2 :!:0.8 5.9 :!:o. s Control 158 1800 0 4 -0.l .. 0 1.7 -. . , o. 4 .) . 0.3 -0.1 -0.l .-0.48 .. Q r - • D -0.1 . . e. 7 :!o.6 :!o.4 -0.3 :!0.2 +4.4 :!o.9 +0.8 :!o.9 Difference "' • :!0.2 '2 ... o. g (continued on next page) q ,:;,,. 5.1 -71- TABLE 12 (cont.) Group Size Obser. ps as an PV oc or 2 163 984 0.6 :!0.3 ,. 2 .0 -0.5 0 0 .. 7.5 -0.8 i 6. 8 -0.8 Control 158 1800 0.4 ~O.l 1.7 :!b.3 0.3 :!O.l 0.4 :!O.l 2.8 !0.4 5.1 :!Oo5 +0.2 :!o.3 .io.3 :!0.6 -0.3 :!O.l -0.4 :!0.1 -'4.7 :!0.9 -1. 7 :!0.9 Difference 2-} 156 1334 0.4 ~0.2 +2 • 8 -0.5 1.4 + -0.3 0.5 + ' -0.2 +6.6 -0.8 +5.2 -0.6 Control 145 3634 +0.4 -0.l 9 2 +~· -0.3 +0 • 3 -0.1 0.4 -0.1 + 3• 1 -0.3 ,.5.9 -0.4 0 .-o.6 :!0.6 "l. 1 ~0.3 "0.1 ~0.2 "3. 5 ~0.9 -0.7 ~0.7 2.4 0.3 :!0.2 0.8 :!0.3 2.9 :!:O. 5 4.4 :!0.6 0 .2 -0.1 0 .4 -0.1 3.4 .-0.5 .· -0.7.17 J0.4 -0.5 :!o.7 Difference + 3 108 1058 0.1 :!O.l :! o. 5 Control 141 1388 + 0.3 -0.1 +2 .6 -0.4 + + -0.2 !0.1 -0.2 ~0.6 -'O.l :!0.2 :!o. 3 Difference -2.7 !o.9 -? 2 - TAB LE 13 SC X V car - 4o 0 c. - Young Pupae, 1 20 to 130 Hours - Group Size 0bser. -ps Control 141 1388 0.3 -0.1 "a" 151 888 0 Difference 1 134 1260 Difference 2 1616 164 114 1010 Difference All oc or 2.6 .-0.4 . 0.2 .-0.0.42 -0.1 3.4 :!o.5 ?.l - 0 .? 1.4 + -0.4 0.1 !0.1 0.9 !o.3 6.4 !o.8 6.9 -0.9 -0.3 !0.1 -1.2 :!0.6 -0.1 .. o. 5 :!0.l :!0.4 .. 3.0 :!0.9 -0.2 -:!l.l 0. 2 ~0 .1 2.3 :!o.4 o.5 :!0. 2 0.4 :!0. 2 6.9 4.8 to.? !0.6 - 0 .1 "0.1 -0.3 "0.6 -40.3 0 "0. 2 .. 3.5 "0.9 - 2 .3 "0.9 0.1 :!0.1 2.4 !0.4 :?:0.2 0.4 :!0.2 6.4 !0.6 5.2 :!0.6 ,:::, -0.2 :!o.6 +0.2 :!0. 2 0 :!0.1 T3.0 !o.s -1.9 :!o.9 0. 2 -0.1 1.5 .-0.4 0.3 ~0.2 -0.1 ~0.1 -1.1 "O.l :!:o. 6 .. 0.1 -0.05 -0.2 ..._0.1 . -0 ." Difference 3. pv 140 4774 Difference ( 1t: + as an 0.4 . . .a 1.0 -0.3 7.7 5.8 .-0.9 .-0.7 :!0.?. " 0 .6 :! 0. 4. +4_3 :! 1. 0 -1. 3 :! 1. 0 ~2 .0 -0.2 ,.o.3 -0.1 .. o.6 -0.1 ,. 6. 8 -0.4 * 2 ,.5. -0.3 -0~6 ~0.4 .. 0.1 ~0.1 .. 0.2 ~0.2 +3.4 ~0.7 -1. 9 Sum of 1, 2, a nd 3, only.) ~0.R -73- of the exposure in hours ("group"), the mean culture size in the series, the number of male half-fly observations, and for each bristle the percentile frequency of bristles present with the standard error of the frequency. For each exposure group the appropriate control data are given, and beneath them the difference between the experimental and control frequencies with its standard error. a. The embryonic period. Table 8 contains the data which show the effects of 4o 0 c. on the bristle frequencies in sc males when the exposure is made during the embryonic life of the fly. The effect of the high temperature is extremely lethal during this period. Even in the 11 a 11 exposure series the culture size was small, and in the 3 hour series only an occasional adult developed. The lethal eff~ct was probably at the time of exposure, and upon the embryo. In the 3 hour series, in cultures which developed no flies, no larvae were observed and the eggs were clearly visible in the food for several days. The data of the 11 a", 1, and 2 hour exposure groups did not differ between themselves, and are therefore presented together. No contemporary controls of similar size were available. The difference in size probably accounts for the difference in the or frequency in the two groups of data. The increase in the oc _frequency, however, is cer- tainly significant, and the increase in an frequency is probably significant. -74- b. The first inst ar. The da ta showing the effects of 4o 0 c. during the first instar of larva l life are those of Table 9. The high temperature was lethal to some of the larvae, expecially in the longer exposures. The dat a show that the main effect of the temperature on the bristles was an increase in the oc a nd or fre quencies, especially the oc frequency, at nearly all exposures. The longer exposures appear also to h ave decre a sed the frequency of the pv bristle. c. The second inst ar. In Table 10 are the da ta from the series exposed during the first h alf of the second instar. The high temper- ature did not appear to be lethal during this period. The data show tha t the temper &ture increased the frequencies of the oc and or bristles a nd decre a sed the fre que ncy of the pv bristle. d. The first h a lf of the third instar. The data of Table 11 show the effects of high temper a ture durin g the f irst h a lf of the third inst a r. The tempera ture effects were not noticeably letha l during this period. They a re evident, however, in incre a ses in the a n and or frequencies. The l a st begins with the one hour ex- posure group a nd continues through the hi gher exposures. The chang es in the ps and a s frequencies, while nearly a ll in one direction, are pro bably not si gnific a nt. -75e. The second half of the third instar. The Table 12 data show the effects of 40°c. during the last half of the third instar - the only period in which viable temperatures, according to Child, influence the bristle frequencies. The 3 hour exposure during this period was considerably more lethal than any of the shorter exposures. The exposure was repeated twice on a smaller scale with similar results. The data indicate a clear in- crease in the oc frequency for most of the exposures. Its return to the control level following the 3 hour exposure was born out in the smaller later test$. The changes in the other bristle frequencies are irregular, but indicate probable decreases in the pv and or frequencies. f. The early pupal period. The data in Table 13 are from flies which were exposed to 40°C. during what was the early pupal period for the majority of the individuals. vae present, too. There were a few lar- The 3 hour exposure in this period was more lethal than the shorter exposures. The significant temperature e-ffects appear to have been on the oc and or bristles. The oc frequency is above the controls in each of the four exposure groups. ed by the 11 The or frequency was unaffect- a 11 exposure, but was below the control frequency in the longer exposures, singly and combined. g. Summary Table 14 is a summary of the significant temperature effects on bristle frequencies demonstrated in the data of Tables 8 through 13. Each entry is the mean 11 coefficient -76- TAB LE 14 SC X V ca r - Summar y of 4o0 c. Se ri es - _pevelo pment a l Period ps as an pv oc or to 10 hours 0 0 +3 . 0 0 ♦ 2 .3 2 0 20 to 30 h ou rs 0 0 0 - 0 . 55 .. 0 . 64 .. o.44 48 to 5 0 h ours 0 0 0 -0 .7 4 ,. 1. 3 6 .. 0 . 63 72 to 82 h ours 0 0 .. 3 . 0 - 0. 71 .. o . 89 - 0 . 32 96 t o 1 06 h our s 0 0 0 - 8 . 39 +0. 8 9 - 0 .1 ~ 12 0 ta 1 30 hou r s 0 0 0 0 ~1 . 00 - 0 . 87 0 -77- of change" in bristle frequency in the experiment a l flies when compared to their reapective controls. (See page 44). T h e ~ coefficient of change is the average of the sum of the coefficients from the exposure groups of a given period, beginning with the shortest exposure to show a significant and including all the exposures longer than that one. The use of the mean coefficient of change is justified by the data of the above tables, which show that, generally speaking, a temperature effect upon a given bristle, once it has been established, is not consistently increased by longer exposures. An entry of O in this table indicates that there was no significant frequency change in that instance. The data. of Table 14 show the following general effects of exposures to 4o 0 c. during the indicated developmental periods. The ps and as frequencies were not affected during any period of development. The an frequency increased in embryonic and early third instar periods to approximately four times the control frequency. The pv frequency decreased during the larval periods to one-ha lf or one-qu arter of the control frequency. The oc frequency increased in all the periods to approximately twice the control frequency. The or frequency increased in the first and second instars to approximately one and one-half times the control frequency; and decreased to approximately three-quarters of the control frequency during the third instar and young larval periods. -78- h. Discussion. The data indicate clearly that the exposures to 40°c. have efftcted changes in the bristle frequencies of the sc males. We may nov, ask if the effect was on all or only a part of the population in each case, and if the bristles showed tendencies for covariance. The first of these questions may be ans-..vered by exajnina.tion of the left-right correlation coefficients. These were calculated for each pair of bristles in each of the developmental periods. In no case did any coefficient appear to be different from its corresponding control. It seems probable, therefore, that the temperature effect was on the whole population. The question of covariance is normally answered by two-~y-two correlation coefficients. experiment (see sc x They were not calculated in this sc experiment below). The data them- selves, however, suggest that the bristles behave independently in this experiment, since, on the whole, the bristles do not respond alike to tem:perature treatment. -79- i. Conclusions. ' -4~~ We may ~e the following general conclusions from the above data: (1) The TEP for scat 40°C. extends throughout the developmental period for cert a in of the bristles. It varies in duration and position for the several bristles. (2) The effect of 4o 0 c. is not the same on all bristles. It increas es the fre quencies of some of them, and decreases the frequencies of others. Some bristles appear to be unaffected. (3) The effect of 4o 0 c. may not be in the same direction for a given bristle at different periods in development; but instead it may increase the frequency at one period and decrease it at another. (4) The effect of 4o 0 c., once est ablished, is genera.lly mot changed by increasing the length of the exposure. -80- 3. The dat a from sc x sc. The 40°c. experiment with the ma ting sc fema les x sc males was done in the school ye a r of 1936-1937 on the same general plan as the sc discussed. x v car experiment already The use of Moldex be gan with the s~cond series in the se c ond instar (see b elow) and wa s continued throu gh the third inst a r a nd youn g pup ae series. The data of the ma les a re given in Tables 15 through 21, in a manner similar to that used in the data tables of the sc x v ca r experiment. The number of obser- vations in e a ch exposure series wa s less tha n in the sc v car experiment, and fewer series were completed. x It wa s observed tha t the temper a ture effect wa s independent of the length of the exposure period, once it had begun. Therefore, in order to lower the st a nd ard devi a tions, the several exposure series h ave been grou ped for compar ison to their controls. a. The embryonic period. Table 15 contains the data from flies which were exposed to 40°c. during their embryonic developmental st a ge. More than one hour of 40°C. proved to be f a tal to a ll but a very fe w of the embryos. The da ta show th a t one hour wa s sufficient to produce increases in the an and oc frequencies. The apparent incrEL_,ase in the or frequency is probably due to the consider ably sma ller mean size of the experiment a l cultures. - 81 - TABLE 15. - SC X SC - 4o 0 c. - Embryo, 0 to 10 Hours Development - GrouP Size Obser. :es as an 1 67 7 20 0.1 :!0.1 1.0 :to.4 2.5 1.5 :!0.6 !o.5 0.1 ~O.l 1.1 + -0.3 -:!0. 2 o.s 0 -0.1 :! 0 .5 1.7 :!0.6 110 Control 1360 Difference 4 oc o:r ?o. s 4.7 17'~ 5 :!l. 3 0.9 ,& -0.3 2.5 ,& -0.4 :!o.s +o.6 :! 0 .6 :!0.9 pv •2. 2 9.9 +3 .6 :! 1. 5 TABLE 16. SC SC X - 40°c. - :1:' irst In s tc1r, 20 te 30 Hours Deve lo iJmen t - Group Size Obs er. PS as an DV oc or 1;2;3 137 17 92 0 0 .7 !o. ~~ 1.5 !0.3 1.0 !0.2 8.0 !0.6 80.9 0 0 .4 :!O.l 0.6 :!0.2 0.6 :!0.2 3.9 :! 0 .4 17.0 :!0.9 0 f0. 3 ..- 0 • kJ •) +0 .9 .-0.4 +0.4 ~0.3 +4.1 :!o. 7 +3.9 .-1.3 0.4 :! 0 . 2 1.5 :!9.3 0.8 :!0. 2 4.? :! 0 . 6 23 .4 :!1.1 2 .1 !o. 3 13.3 !0.7 +2.6 .. 0.7 +10.1 .. 1. 3 Control 1 25 1 9 16 Difference !l.O ----------------------------------------------------------------97 2;3 Control 110 Difference 1430 2 096 0 0 0 o. 3 0.5 o.s :! 0 .1 ! 0 .2 ! 0 .2 -tO.l +l.O 0 •0. 2 .. 0 .4 -82- TABLE 17 x SC sc - 40 OnLl. - Second Instar, 48 to 58 Hours Development - Group Size Obs er. PS 3 91 684 _, 0 • .:.> !0.2 Control 110 2096 0 as 'Z 0 .o .. 0 •J 0.4 pv oc 01' 0.3 4.1 -.. 0" . .-0.8 16.1 + -1.4 2.1 13.3 1-., ~0.3 0.3 :!0.1 0.5 0.8 !0.2 :!:o. 2. :!o.s -0.1 :!o. 4 -o.5 :!o. 3 :o.s 1.0 0.2 :!0.1 6.3 :! 0. 7 15.0 :!1. 0 - "'O. 3 Difference an • 0 :!o. 2 .::, ,L2. :!:1. 6 0 "'2.8 =1.6 ---------------------------------------------------------------108 2;3 Control 108 1196 3538 DiffP-rence 0 0 0 0.4 :!0.2 :!0.3 0.1 :!0.05 0.7 ±0.2 4.2 13.5 :!0.2 :!0.3 :!:0. 6 -t0.3 .. 0.2 -0.5 .. 2.1. :!0.2 !o.4 :!0.2 :!o.s 0.8 4 1 • 0r:!1. 8 TABLE 18 sc x sc - 40°c. - 'l'hird Instar, 72 to 82 Hours Development - Group Size Obser. 2;3 163 2700 1)S as an -pv oc or 0.3 1.6 :!0.2 4.8 :!o.6 ' !0.2 1.3 5.74 :!o.45 :!o.6 1.1 !O.l !0.19 7.5 !0.3 •3 • 0 -0.5 ... 4. 4 .-0.? :!0.1 Control Difference 155 7550 0.1 1.0 1.0 :!0.04 :!O.l :!O.l "'0.~ 6 -to ,•. .s.o. -'3.8 !0.1 -. ~ . -0.4 ... -'O ,. ~ -0.2 2.78 4 11.9 - 83 - TABLE 19. SC X - 4o 0 c. - Third Ins tar, 96 to 106 Hours Development - SC Grou:i2 Size Obser. 12s as an 12v oc or 2;3 150:! 112 6 0.3 :!0. 2 2. 2 :! o .4 3.9 :!0.6 1. 2 :!0.3 9 .7 :!0.9 6.0 :!0.7 Control 150 2932 0.3 1.6 !0. 2 1.5 ~0. 2 0.5 ~O.l 5.6 ~0.4 5.9 ~0.5 -t0.6 +2.4 :!0.6 +0.7 :!0.3 +4.1 :!l. 0 :!0.9 :! O.l Difference 0 :!0.4 +O.l TABLE 20. SC X SC - 4o 0 c. - Young Pu pae , 120 to 130 Hours Development - Group Size Obs er. ps as an pv oc or 1; 2 ;3 i:::5 0:! 13 24 0 0.8 ~0.3 1.3 ~0.3 0.6 ~0.2 4.5 ~0 .6 ... 5.1 -0.6 Control 240 728 2 -0.04 !0.02 0 .6 0 .9 0.4 .-0.1 .-0.1 .-0.1 5.0 ~0.3 5.1 .-0.3 - 0 .04 :! 0 . 02 40.4 :! 0 . 3 -0.5 :!0.7 0 :!0.3 Difference .f0. 2 -4 0 . 2 :! 0 . 2 TABLE 2 1. SC X SC - 4o c. - Older Pupae, 1 44 to 154 Hours De velo pmen t 0 Group Size Obs er. ps as an pv oc or 1;2;3 20 0:! 1518 0 O. 'l :to. 2 0.5 :to. 2 0.6 :to. 2 6.5 :t o .6 7.7 ~0.7 Control 199 4380 0 .5 0. 3 Dif fere nce 0 .0 9 ::! O. 05 !O.l 1.2 !0. 2 !O.l 5.5 !0.3 6 .1 ! 0 .4 - 0 .09 :!0.05 -+0. 2 !0. 2 -o.7 :! 0 . 3 "0.3 :! 0 . 2 -+1.0 !o.? :! o. s Tl. 6 -84- b. The first instar. Data from flies exposed during the first part of the first instar are those in Table 16. The two sections of the table re~present two series carried out several months apart. The data indicate increases in the an, oc, and or bristles. c. The second instar. The data from flies exposed during the second instar are presented in Table l?. The "3 11 hour group was raised on non-Moldex food; the "2;3 11 hour group was raised on :Mold ex food several months later, and at a slightly higher control temperature. The Moldex does not appear to have influenced the effect of the exposure to 40°c. on the bristle frequencies. The data indicate an increase in the oc and a decrease in the pv frequencies. When the data of the two tables are summed, the increase in the or is large enough to indicate probable significance (twice its standard error). d. The first half of the third instar. The data. from flies expos~d to the high temperature during the first half of their third larval instar of development are those of Table 18. They indicate in- creases in the frequencies of the an, oc, and or bristles. The slight increase in the other frequencies is probably not significant. e. The second half of the third instar. Table 19 contains the data from flies exposed to the high tempera ture during the second half of the third -85- instar. Many of the individuals died in the pupal stage. The culture size wa s estimc. ted on the basis of a few rough counts of dead pupa e, along with the fly counts. The data indicate increases in the an a nd oc bristle frequencies. f. The early pupal period. Because of the arowded conditions of cultures used in series testing the effect of temper2.ture during the pupal stage, there was a considera.b le variation in the time of pupation. Some stragglers had failed to pupa te even after 144 hours at 25°C. In addition, the mortality was very high among the pupae, many of which progressed to the imago stage but lacked the strength to emerge. Those that did emerge in the older pupae series were phenotypically abnormal to the extreme, and had very few hairs and bristles. They were classed for bristles by the presence or absence of the trichopore, out of which the bristle normally grows. Many pupae were on the sides of the culture; others were on the ;(paper in the food. Between these two positi6ns there was a difference of at least 1°c. For all of these reasons, the d a ta from the pupae series which are given in Tables 20 and 21 are not as satisfactory as the larval data. They do not, however, indicate any strong ef- fect of the temper a ture treatment on the bristles. g. Summary. The data of Tables 15 through 21 are summa rized in Table 22, where each entry is the coefficient of change in bristle fre qu ency. Table 22 indic a tes the following general facts about the response of the several bristles to -86- TABLE 22. SC X SC Summary of 40°C. Series - Developmental Period ps as an pv oc or 0 to 10 hours 0 0 ~2.1 0 -'0.88 0 20 to 30 hours p 0 • 1. 5 0 •1.15 .. 0;50 48 to 58 hours 0 0 0 -0.67 .. 0.73 .. 0.18 72 to 8~ hours 0 0 .. 3.s 0 ,. 1. 08 .. 0.59 96 to 106 hours 0 0 +1.6 0 +0.73 0 120 to 130 hours 0 0 0 0 0 0 144 to 0 0 0 0 0 0 154 hours -87- the temperature treatment. The ps a nd as bristles were un- affected throughout development. The an frequency increased in all periods of egg-larval life excepting the second instar. The pv frequency decreased in the second instar but was unaffected during the other periods. The oc bristles increased in the first and second instars, and in the first half of the third instar. None of the bristles appear to have been af- fected during the pupal periods. h. Discussion To determine whether the temperature effects on sc ma~les from sc x scare on part or all of the popu- lation, the lef~right correlation coefficients were calculated for each pair of bristles in each developmental period. In the embryo and first instar series, there waB no indication of correlation in any of the bristle pairs. But in the older larvae, an increase in bristle frequency was accompanied by a small but probably significant increase in leftright correlation. In the control flies these correlations ranged from .0.02 to .o.08 with t values of 2 or more for only the larger coefficients. In the treated flies the coef- ficients ranged from .0.10 to+0.20 with t values of 3 to 5. This indicates that in these older series some of the larvae were probably unaffected by the exposure to high temperature. It is quite :probable that these larvae were able to escape the extreme temperature by crawling into the ;'paper in the middle of the food. The temperature was known to be a degree lower there than in the food, due probably to a higher evaporation of moisture from the paper. -88- To determine the possibility of covariance among the affected bristles in sc males from sc x sc, the two- by-two correlation coefficients were calculated for each of the possible combinations of the six bristles considered. These coefficients proved to be practically identical in exper$mental and control series, and in no cases were they large enough to suggest a tendency for covariance among any of the bristles. i. Comparison with SC X V car. The summarized data of SC X SC in Table 22 (page 86) are comparable to the summarized data of V car in Table 14 (page 76). SC X It is apparent at once that the outstanding facts of similarity in the temperature effects on the two matings are the absence of appreciable effects upon the ps and as bristles and the increase in oc frequency through practically all of development. The dif- ferences between the an and pv fequency c~anges are largely matters of duration of the sensitive period (TEP). The an bristle displays a considerably wider sensitive period in the sc x sc mating, while the pv bristle has a broader sensitive period in the sc x v car mating. The differ- ence in the or frequency changes in the two matings involves both a difference in duration of the TEP and a difference in direction of bristle frequency change. In sc x sc, the change in frequency is an increase through a shorter portion of development. In sc x v car, the change is first an increase, then a decrease in frequency, over a considerabley longer period. -89j. Conclusions. I (i-< \\ t.\;· We may mak-e the following general conclusions from the above considerations: (1) The data of sc x sc support those of sc xv car in establishing a TEP for sc which extends for some bristles probably throughout the developmental :period of the bristles, and differs in one way or another from bristle to bristle. (2) They indicate, in addition, that the duration of the TEP and the nature of the effect produced vary between selected stocks. -90- 4. Summary and conclusions f rom the 40° c . experiments. (1) Dat a a re presented which show the effects of a graded series of exposures to 40°c. during six developmental periods, covering practically the entire bristle development period, on the frequencies of certain of the bristles in sc males from the ma tin g s sc fema les sc fema les x x v car males and sc males. (2) The ps and as bristles a re unaffected in both matings. (3) The an frequency increases to from 2 to 4 times the size of the control frequency, in both matings, through a part of the development a l period. greater in sc x sc than in sc The TEP is consider ab ly x v car. (4) The fre quency of the pv bristles decre a ses in both matings to from one-qu arter to three-quarters the size of the control fre quency, through a pa rt of the developmental period. The TEP is much broader in sc than in sc. SC X x v car (5) The oc bristles incre a se in fre quency to a pprox ima tely twice the size of the control fre ouency in both matings, through all of egg-l arva l development and into pupal development in the sc x v car ma ting. (6) The or fre quency increases in sc v ca r in the first x and second inst ars to approx im2.tely l ½ times the size of the control fre quency, a nd decre a ses to ap prox imately threequarters of the size of the control fre quency in the third inst a r a nd e arly pupal periods. In sc x sc the or fre- quency incre a ses in t h e first, second, and early thir d inst ar periods to a pproxima tely l ½ times the size of the con trol fre quency and is un a ffecte d in other develo pmental periods. -91- (7) The TEP of sc males for nearly lethal exposures to 40°c. extends throughout the bristle developmental period for one or more bristle~. (8) The duration of the TEP, and the extent and direction of the change in bristle frequency varies more or less from bristle to bristle, from stock to stock, and from one developmental period to another. (9) A change in bristle frequency effected by a given expos- ure to 40°c. within a given devlopmental period is gener- ally not increased by increasing the length of the exposure period. (10) A lack of correlation between any two of the bristles in the sc x sc experiment indicat$s that the 40°c. treat- ment affects the bristles independently of each other. -92- E. The 36°c. Experiments with sc Ma.J..es. l. Plan of experiments. During the spring and f a l l ~ of 1937, two experiments with 36°C. were carried out, similar in plan and method to the above 40°C. experiments, but less extensive in the number of tests made. Rough experiments in the early months of 1935 had indicated that nearly lethal exposures to 36°C. produced very different effects on the bristle : frequencies than did nearly lethal exposures to 400c. The main difference in method in the two temperature experiments was the use of much longer exposure periods in the 36°c. experiment, and hence the testing of much broader developmental periods. The entire first insta.r con- stituted one test period; the entire second instar constituted another; the third instar was divided into two test periods; and the early pupal period constittuted a fifth period. In addition, the embryonic period was tested in the sc sc x experiment. 2. The data from sc The 36°C. sc 1937. x x v car. v car experiment was done in May, It was necessary to carry out the experiment at a con- trol temperature between 25.5 and 26°C. The 26°C. control data are those of Table 7 previously discussed. The data from the sc males of the experimental lines are presented in Tables 23 through 27. These tables ate set up in a manner similar to that used in presenting the 40°c. data. -93- TABLE 23. X SC V car - 36°C. - First Instar, 24 to 53 Hours Development Group Size Obs er. ps as an PV oc or 16;22 171 1400 2.4 :to.4 14.0 :!0.9 1.9 :!0.4 1.1 :to.3 9.9 :!0.8 '!.Q. 7 0.8 :!0.2 4.3 :!o. 5 1.6 :!o.3 2.0 :!o.4 10.6 :?o.9 3.3 :?o.5 .. 1. 6 :!0.4 4-9.7 .f.O. 3 :to.5 -0.9 :to.5 -0.7 +3.5 :!l. 0 :!l. 2 :to.9 Control 196 1316 Difference 6.8 TABLE 24. SC X V car - 36°c. - Second Instar, 44 to 82 Hours Development Group Size Obser. ps as an pv oc or 24 211 1618 1.7 :!o.3 8.2 :to.7 1.1 !0.3 0.4 :!0.2 12.9 :!0.8 4.0 :!0.5 Control 842 1848 1.1 :!0.2 5.4 :to.5 1. 7 :!O. 3 1.5 :!0.3 6.5 :!O. 6 1.8 :!0.3 •0.6 :!0.4 "'2.8 !0.9 -0.6 !0.4 -1.1 !0.4 "'6. 4 :!l.O !0.6 Difference .a2.2 ~04 1280 1.5 !o.3 10.5 ~0.9 1.9 :o.4 0.6 ~0.2 22.5 :1. 2 6.2 :o.7 Control 196 1316 0.8 :!0.2 4.3 :!0.5 1.6 :!0.3 2.0 :!0.4 10.6 :!0.9 3.3 :!0.5 -t0.3 -1.4 :!0.4 -tll.9 :!l. 5 :!0.9 30 Difference "f0.7 "'-0.4 -t6.2 -1.0 " :!o. 5 -t2.9 -94- TAB LE 25. SC X V car - 36°c . Group Size Obser. 24 228 1322 Control 23? - 'fhird Inst ar, 68 to 105 Hours Development - 3234 Difference 30 204 Control 196 Dif feren ce ps as an pv oc or ?.3 '!. O. 7 12.9 :! 0 . 9 2 .6 :!0. 4 0.5 :!0. 2 10.4 :! 0 . 9 6 .4 :!0.7 0. 9 ::! o . 2 ;o. 4 5. 6 2.4 :! o . 3 1.6 :! o . 2 10 . 6 :! o .5 2 .? :!o. 3 "'7. 3 i1.o " 0. 2 io.5 -1.1 t0.3 -0.2 "' 3 .7 :t. O.? tl.O io.8 16. 4 :!l. 0 22 . 2 11.1 2 .3 .& -0.4 o .7 :! 0 . 2 11.5 .& - 0.9 4.8 .& -0.6 0. 8 :! 0 . 2 4. 3 ! 0 .5 1. 6 :!0. 3 2 .0 :!:0.4 10.6 :!:O. 9 3 .3 :!:O. 5 15. 6 :!LO "'17. 9 .. 0.7 :!0.5 - 1 .3 :!0.4 +0 . 9 :!]t.3 +1.5 :?o.s "' 6 . 4 1326 1316 4 :!l. 2 -95- TAGLE 26 . .x sc - V ca r - 36 ° C. - Th i r d I ns t a,r, 90 to 1 28 Hours Develo pment - Grou:Q Size Obs e r. 12s as an pv oc or 24 ?. ?? 1708 12. 4 :! 0 . 8 1v e O :! 0 . 9 3.5 :! 0 . 4 0. 5 :!0. 2 7. 5 :!0. 6 2 .8 :!0.4 Control ~7 8 19 18 1. 0 ..- 0 .-~ ..- ·0 • 0"' 6. 6 3. 0 .- 0.4 1. 4 ! o.3 10 . 6 !o.7 2. 3 ! o .3 :! o . 5 Dif fe renc e 23 7 30 Control 23 7 73 4 3234 Di ffe r en ce . 1::;; {:; •11. 4 ~0 .8 ... 8 . 9 -- o .5 ~1.1 :! o. 6 - 0. 9 :! o . 4 - 3 .1 :! o.9 ;::;0. ...- 1.54 22 . 8 2. 2 .- 0.5 "'- o "-'•) 0. 3 . 14 . 9 -... 1 ,)" . ... 8 . 3 0.9 ! 0. 2 f 5• 6 ... 1. 6- 10 . 6 - 0. 4 2. 4 ~0. 3 -0. 2 : o.5 ... "-'{) . 7 "'19 . 3 ~1. 5 "' 17. 2 :n . 7 - 0.2 :! 0 . 6 -1. 3 :! 0 . 3 !1. 4 +l - .o,,. -' 4 . 3 .a. o. 5 -1.0 - 0.3 .. 5.6 :! 1. 0 TP,B LE 2 7 . SC X V car - 36°c . - Yo un g Pupae , 11 6 t o 148 Hour s Developmen t - Grou n Si ze Obs er . DS as 8..n ov oc or :C:5 237 :! 832 2. 0 :! O. 5 4 .7 :! 0 . 7 1. 8 ±o.5 0 6. 2 :! 0.8 6. 6 :! 0.9 Cont r ol "t!.-3 7 3234 0. 9 5.6 ! o. 4 2. 4 ! 0. 3 1. 6 ! 0. 2 10 . 6 :! 0 . 5 2.7 !0.3 - o. 9 :! o . s -o.e :! o . 6 -1.6 :! 0 . 2 - 4. 4 :!o.9 "'3.9 :!1. 3 :! O. 2 Di ffer e nc e "'l o 1 ! o. 5 -96a. The first inster. Table 23 contains the data from flies exposed to 36°C. during the first instar. An earlier series with a 24 hour exposure to 36.5°c. failed to hatch many flies; but 16 and 22 hours of 36 °c. were, not markedly lethal. There were no differences in the bristle frequencies of the two series, and the data are presented here together. It can be seen that there was a substantial increase in ps, as, and or bristles, and a decrease in the pv bristles. b. The second instat. The data of Table 24 are from flies exposed dur- ing the second instar. The exposure to 36°c. appears to have reduced tne pv frequency. The difference between the 24 and 30 hour exposures is possibly reflected in a somewhat larger increase in the ps and as frequencies in the 30 hour group. c. The first half of the third instar. The data of Table 25 are from flies exposed to 36°c. during the first half of the third inst2r. The ps and as freauencies are very much larger in the experimental series, and considerably more so in the 30 hour exposure than in the 24 hour group. the e.:;qJerimental series. The pv frequency is smaller in The or frequency is larger. d. The second half of the third instar. The data of Table 26 are from flies heated during the last half of the third instar and in early pupal life. -97Many larvae had pupated at the time exposure was begun. None appeared to pupate during the exposure period; but nearly all the larvae pupated within a few hours after removal from the high temperature. Only a. few flies failed to emerge from the pupae cases in the 24 hour group; but a majority failed to emerge in the 30 hour group. The data are not unlike those of the first half of the instar in Table 25, excepting a curious behavior of the oc frequency,. In the 24 hour exposure group, the oc frequency is below the control frequency by more than three times the standard error of difference. In the 30 hour groups,' it is above the con- trol frequency by approximately the sanie amount. A deviation as great as either of these between two means would be expected to occur by chance alon~in the order of once in 300 trials. It is probable, therefore, that this is a true temp- erature effect,, and not a random variation. e. The eurly pupal period. The data of Table 27 are from flies exposed to 36°C. during the early pupal period. Only a few individ- uals were still larvae at the time of exposure. Most of the flies failed to emerge from the pupae cases. As in earlier pupae series, those which did emerge were markedly abnormal phenotypically. Bristle classification in many cases was by means of the trichopore. The data indicate an increase in the or frequency and a decrease in the oc and pv f~quencies. The small ps frequency increase was probably due to a.n effect on the few larvae present ( discussed below). -98- f. Sunnnary. The da ta of Tables 23 through 27 are summarized in Table 28 . As in previous summaries, each entry is the -- coefficient of change in bristle frequency. The data of this table show the following general observational facts. The ps frequency incre a sed throughout larval development, and very ma rkedly in the third instar. The effect of 30 hours was gre a ter than the effect of 24 hours in at least the two third instar periods. The as frequency likewise increased throughout larval development, and probably somewhat more in the third instar than in the first two instars. The effect of 30 hours was considerabley gre a ter than the effect of 24 hours. The an bristle was unaffected in all developmental periods. The pv frequency was reduced to from one-half to one-quarter the size of the control frequency in every developmental period. The oc frequency in- creased in the second instar to twice the size of the control frequency. In the late third instar period it was re- duced by the 24 hour exposure and increased by the 30 hour exposure, neither change biting large, but ea.ch being statistically significant. In the early pupal period the oc fre- quency was reduced to six-tenths of the control frequency. The or fre quency increased in all developmental periods, generally to about twice the size of the control frequency. There does not appear to have been any significant difference between the effects of 24 hours and the effects of 30 hours upon the fre quencies of the pv and or bristles, and upon the frequency of the oc bristles in the second instar. - 9 9- T AB LE 28 SC X V car - Summary of 36°C. Series - Period Exposure PS as an pv oc or 24 to 53 16;22 _. 2. 0 +2.3 0 -0.45 0 ~1.1 24 ♦ 0.5 .. o . 5 0 -0.73 .. ,1 .0 .. 1. 2 30 -t0.9 '1.4 0 -0.70 '1.1 .a.o.9 24 .f7.l +1.3 0 -0.69 0 -tl.4 30 "l~.5 "-4. 2 0 -0.65 0 .a.o. 46 24 '11. 4 '1.4 0 -o.64 -0. 2 9 .. 0.22 30 •21.4 .. 3.1 0 -o.s1 .. 0.41 .. 2.1 25 0 0 0 -1. 0 - o .4o +1. 4 48 to 8 2 68 to 105 90 to 128 116 to 148 -100g. Discussion. To determine if the temper2.,ture effect was on part or all of the population in a culture, lef~right correlation coefficients have been calculated and are presented in Table 29. Included under each coefficient is its t value. It can be seen that in the control series of this experiment there was a small :positive correlation in all but the oc bristles. In earlier series with this mating the an correlation was generally less than here, and the oc frequency was generally positive, between 0.02 and 0.08. In the first instar series the coefficients do not appear to have changed significantly from the controls. Be- ginning with the second instar there appears to have been an increase in several of the coefficients. This is especially notable in the ps and as coefficients in the two third instar periods, which was also the time of greatest increase in the ps frequency. It seems probable that the temperature effect was not the same on the whole population in these older larvae, and that in the cases of the ps and as bristles more larvae were affected in the 30 hour exposures than in the 24 hour exposures. The possibility must also be rec- ognized, however, that the difference between the two exposures may have resulted from the extra. six hours of exposure upon the same larvae, in the 30 hour group. These data are not critical for this interpretation. The relatively very large increase in ps frequency in the third instar, with accompanying large leftright correlation coefficients, suggests for this bristle -101- TABLE 29. Left-right Correlation Coefficients in 36°C. SC X V car ps as an PV oc or "0.06 2.8 "0.13 6.3 "0.10 5.1 "0.04 2.0 .. 0.01 0.5 ""O. 04 1.7 16;22 .. 0.04 0.9 .. 0.05 1.3 -0.02 0.5 0 .. 0.02 0.5 -+o.og 2.2 24 +0.06 1.5 •0.17 4.9 0 0 "0.08 2.4 "'0.06 1.5 30 0 "'0. 22 5.5 0 0 "0.11 2.7 "'0.12 3.0 24 +0.30 7.8 "0.23 6.0 +o .og 2.4 0 .. 0.14 3.6 "0.19 4.8 30 +0.29 7.5 ~0.27 7.0 +0.11 2.9 0 +0 .06 1.7 +0.11 2.9 ~4 4 0.53 15.0 +o.32 9.2 •0.14 4.2 0 •o. 09 2.5 "0.13 3.7 30 +o.53 10.0 .f.0.19 3.5 0 0 -f0.11 2.0 -,0.16 3.0 116 to 148 25 "0.4fi 4 0.11 2.3 " 0 .12 2.4 0 9.4 "0.14 2.8 "0.20 4.1 Period Ex12osure - Control 24 to 53 44 to 82 68 to 10 5 90 to 128 - -102an especially sensitive period which is more or less similar in position in development to Child's TEP at viable temperatures. These data are not suitable for establish- ing this as a fact. In the case of the relatively large left-right correlation coefficient of the ps bristles in the 116 to 148 hour period, there was only a small increase in ps frequency. This shows that the temperature effect was probably limited to a very few individuals. Since there were a few larvae present at the beginning of the exposure period, it is probable that they, rather than some of the many pupae, were affected. Accordingly, this increase has not been interpreted as indicating a sensitive period in the ·pupal stage. Correlation coefficients were also calculated for the two-bJ-two comparisons of the six bristles in this experiment. There was no consistent correlation between any of the bristles in any of the developmental periods. This indicates a probable independence of the response of each of the several bristles to the temperature treatment. h. Comparison with 40°c. sc x The summarized 36°c. sc v ca.r data of Table x v car. 28 may be compared with the similarly summarized 4o 0 c. sc x v car data of Table 14 (page 76). The most notable dif- ference is in the effect upon the ps and as bristles. The 4o 0 c. treatment did not affect these bristles; the 36°c. treatment .increased their frequencies more than any of the others. The an bristles increased in frequency in two per- iods in the 40°C. experiment, but were unaffected in the 36°c. -103- experiment. The pv frequency was affected over a some- what broader period in the 36°c. experiment. The oc fre- quency was affected over a much broader period in the 40°c. experiment. The or frequency, which increased in every per- iod in the 36°C. experiment, showed a change in direction of response in the 40°C. experiment, where it increased in the first two instars and dedreased in the third instar and early pupal period. Thus, every bristle responded more or less differently to the two temperatures. That this is due in the main to the difference in temperature treatment and not to the Moldex food of the 36°C. experiment is evidenced by the fact that the preliminary 35 to 36°c. experiment previous to the use of Moldex showed essentially the same phenomena, especially in regard to the ps and as bristles. 3. The data from sc x sc. In February, October, and November of 1937 a 36°C. experiment was carried out with the mating sc females x sc males, using practically the same methods as in the 36°C. sc x v car experiment. The pure sc individuals could not with- st a nd as long an exposure to the high temperature as could the hybrids. Accordingly, in this experiment the exposures were, for the most part, 12 to 24 hours in leng\g)~. In the February part of the experiment, tests were made on the embryonic and early larval developmental periods. Flies were allowed to oviposit for 6 to 8 hours in a culture, and after their removal, the culture was immediately expose::) to 36°C. The embryos proved incapable of withst anding 6 hours of this temperature. Twenty such cultures gave practically -104no adults; nor were lar¥ae visible in the cultures which did not hatch flies. Another series of cultures in the time period of 15 to 34 hours of development did not appear to be adversely affected by 6 or 12 hours of 36°c., but could not take 24 hours. While many of the younger individuals in these cultures were undoubtedly still in the eggs during at least part of the exposure period, most of them had probably completed embryonic differentiation, according to the data of Poulson (1935). This period is therefore considered practically as a test of young larvae rather than of embryos. The rest of the data of this experiment is from the series run in October and November of 1937. The data are presented in Tables 30 through 35. a. The first instar. The data of the early first instar are given in Table 30. Because there was no difference in bristle fre- quencies in the 6 and 12 hour exposures the two series have been summed in this table. The data show a small increase in the frequencies of the an, oc, and or bristles. The data from larvae further along in the first instar1 are given in Table 31. There were apparent differ- ences in the 12 and 24 hour exposure groups. Together they show increases in the as, an, oc, and or frequencies. b. The second instar. The data from the second instar series are those of Table 32, in which again the 12 and 24 hour groups are pres~nted together because of lack of difference between them. -105TABLE 30. SC X SC - 36°c. - First Instar, 15 to 34 Hours Development - _group Size Obs P.r. ps as an pv oc or 6;12 153 2388 0.04 :!0.02 O.? :!0.2 1.5 :!0.3 0.8 :!0.2 6.9 :!0.5 16.7 :!0.8 0 .09 :!0.06 0.4 :!O.l 0.6 :!O.l 0.7 :!0.2 5.5 :!0.5 13.2 :!0.7 -0.05 :!0.06 --o.3 :!0.2 --o.9 "O.l .a1.4 .a3. 5 :!0.3 :!O. 2 :!0.7 :!l. 1 Control 149 2216 Difference 1 TABLE 31. SC X SC - 36°C. - 1!'irst Instar, 24 to 57 Hours Development - Group Size Obser. 12;24 117 Control 162 PS as an pv oc or 1298 0.1 :!O.l 1.2 :!0.3 1.7 :!0.4 0.4 :!0. 2 5.0 -:!0.6 17.0 !l.O 1812 0.1 :!O.l 0. 2 :!O.l 0.6 :!0.2 0.6 :!0.2 3.1 :!0.4 8.3 :!0.6 '.'1.0 .f.l. 1 :!0.4 -0.2 :!0.3 +1.9 :!0.7 :! 1. 2 0 Difference :!0.3 +8. 7 TABLE 32. sc x sc - 36°C. - Second Instar, 48 to 80 Hours Development - Group Slilze Obser. PS as an PV oc or 12;24 117 1424 0 0.35 :!0.17 1.5 :!o.3 0.2 :!0.1 4.4 :!o.5 7.9 .-0.7 Control 146 1434 0 0.07 !o.o7 0.7 !0.2 0.4 !0.2 2.2 :!o.4 8.9 !0.8 "0.28 :!0.17 .. 0.8 !0.4 -0.2 :!0. 2 •2. 2 :!0.6 -1.0 !1.1 Difference - 106- T.AJ3LE 3 3 A. X SC SC - 3 6°C. - 'l'hird Inst ar, 7 2 to 9 2 Hours Deve lo pme nt - Gro up Si ze Obser. ps as an pv 12 214 1 220 ') 1 • "~0 .3 5 .8 ~0.7 0.5 :! 0 . 2 0. 5 :! 0 . 2 Co ntrol 1 8 6 1 538 0 .1 ~0 .1 0. 6 ! 0. 2 .- 01.. 38 0. 5 ! 0. 2 -+1.1 :! 0 . 3 ... 5 . 2 :! 0 . 7 -1.3 :! 0 . 4 0 Diff e re nce oc or .- 80 .. 83 .- 06 .7. 6 .- 06 .. 65 1-. 20 .9. 6 ,., 1. 8 :!l. 0 - 6 .0 :!1. 2 T ABLE 33 B . sc x s c - 36°c . - Th i rd Inst a r, 7 2 to 104 Hours Develo pment - Gr oup Si ze Obs e r. 24 19 1 1164 ps as a.n PV oc or 7. 8 16 . 8 !1.1 0. 6 :! 0 . 2 0. 4 :! 0 . 2 4. 9 :! 0 .6 5. 5 :! o . 7 0. 6 ~0 . 2 1. 8 ~0 . 3 0 .5 ~0 . 2 6 .5 ~0 . 6 1 2 .6 ~0 .9 .-1 .2 ...l6 . 2 .- 0 . 87 -t -1.1 _o . 4 - 0 .1 ... 0 ~ .-- 01 .. 69 -1. .-7. 21 :! 0 . 8 Con t rol 1 86 Dif fe r en ce 1 538 0 .1 ~O.l -t 7 . - •0 - 107- TAB LE 34. sc x -Group 12 sc - 36°C. - Th ird In s t a r, 9 6 to 116 Hour s De velo pme nt - Si z e Obs e r. ps as an nv oc 200 ! 902 0. 2 :! O. 1 4.0 "!. 0 .7 4.0 0.1 :! O.l 3. 7 4. 9 "!. 0 . 7 ! 0 .7 0. 1 ! o . n2 o. 3 :! 0 .1 0. 7 :! 0 .1 . :! O.l "!. 0 . 6 2. 9 :! 0 . 3 :! 0 . 5 -. 0 .1 .. 3 . 7 :! 0 .7 .. 3 .7 ! 0 .'7 !0.l .. 0 . 8 ! 0. 7 - 3. 2 !0. 9 Control 1 95 343 0 Dif f erence ! 0 .l 0 r· :::) - 0. 4 or 8 .1 TABLE 35 . sc x s c - 3 6°C. - Youn g Pupae, 1 2 0 to 15 2 Hou~s Develo pment - Grou p Size Obs er. ps as 12 ~o o:t 6 76 0 .1 ~O.l :! o.~ 343 0 0.1 :! 0 . 02 0.3 0 .7 0.5 :!0.1 :! 0 . 1 :! 0 .l 0 0 .. 0 . 2 :! n . 4 - 0. 2 ~0 . 2 Control 195 Diff erence 0.3 an 0.9 :!o.4 pv oc or 0. 3 3. 3 :!n.? :! o. s 2. 9 :!0.3 8 .1 :! 0 .5 .. o. 4 - 2. 0 ~1.1 :!o. :?. ~0 ,7 6.1 -108- The hi gh temper r,.ture incr eased the fre quencies of as, an, and oc bristles. The cha nges in the as a nd an fre quencies are on the ~order line of statistical significance. c. The first half of the third instar. The data from the first half of the third instar are given in Tables 33 A and 33 B. There was a marked dif- ference in the temper a ture effect on the ps and as frequenoies in the 12 and 24 hour exposures. given separately. These h ave therefore been The da ta show decreases also in the fre- quencies of the pv and or bristles, without any a pparent difference between the two exposures. The 24 hour exposure pro- duced much greater increases in the ps and as frequencies than did the 12 hour exposure. The 12 hours difference in time, it will be noted, was entirely a t the upper time limit; i.e., the larvae were exposed 12 hours latet in the third instar in the 24 hour exposure. d. The second half of the third instar. The da ta from the l a ter third inst a r series are given in Table 34. The 24 hour exposure was almost entirely lethal, and the 12 hour exposure was lethal to many individuals. As in the earlier experiments, death appeared to come in the pupal stage. Some larvae had pupa ted a t the time of the exposure; but most of them had not. The majority of the larvae appeared to pupate within a very few hours after removal from the high tempera ture. The da ta of the table in- dic a te incre a ses in the as a nd an fre quencies, a nd decreases in the pv a nd or fre quencies. -109- e. The early pupal period. The data from the early pupal period are those of Table 35. The 24 hour exposure was highly lethal, and the 12 hour exposure was about 50,% lethal. The flies that emerged were generally extremely abnormal phenotypically, as in previous pupal series. The data of the table indi- cate no significant bristle fre quency changes. f. Summary. The data of Tables 30 through 35 a re sumrnarized in Table 36. As in previous summaries, each entry is the coefficient of change in bristle frequency. The following summary of effects can be made from this table. The ps fre- quency was affected only in the first of the third instar series, where it increased markedly, especially following the 24 hour exposure. The as frequency increased through all but the early part of l a rva l life, and showed a much larger increase following the ~4 hour exposure during the third insta r. The an frequency increased in the first two instar periods; decreased in the first of the third instar series, and increased again in the late third instar. pv fre quency decreased in the late third instar. frequency increased in the first two instars. The The oc And the or frequency increased in the first instar, and decreased in the third instar. None of the frequencies changed signif- icantly in the early pupal period. g. Discussion As in the previous experiments, left-ri ght correlation coefficients were calculated to determine whether -110- TABLE 36. SC X SC - · Summary of ?i6 0 c. Series Period Exposure ps as an pv oc or 15 to 34 6;12 0 0 .. 1. 5 0 .a.O. 3 +0.3 24 to 57 12;24 0 +5.0 ... 1.8 0 ... 0.6 ... 1. 0 48 to 80 12;24 0 -'4.0 "1.1 0 .a 1. O 0 72 to 92 12 "11 -'8. 7 -0.72 0 0 -0.48 72 to 104 ----------------------------------------------.. 77 .. 27.0 0 0 -0.56 24 -0.67 96 to 116 12 0 +12.0 ... 5. 3 -0.8 0 -0.40 120 to 140 12 0 0 0 0 0 0 -111the temper a ture effect was on all or only a part of the population in each period. There appeared to be differ- ences between the control and experimental coefficients only in the ps and as bristles in the third instar series. In the case of the ps bristles, the only apparent leftright correl a tion was in the 24 hour exposure of the 72 to 104 hour period. value of 8. The coefficient was .o.35 with a t The correlation coefficient was approximately O in the control seties. In the case of the as bristles, the control coefficient was +0.12 with a t value of 3.4. In the 12 to 24 hour exposure groups of the 72 to 104 hour period, and in the 12 hour exposure of the 96 to 116 hour period, the coefficients were +0.39, .0.40, and .o.54, respectively. 9 to 11. The t values for these coefficients ranged from From this it appears that although the temperature effect was more or less on the whole popul a tion with re-spect to pv and or bristles in these third instar series, it was on only a part of the population with res pect to the ps and as bristles. The marked differences between the l~nd 24 hour exposures in the 72 to 104 hour period with respect to their effects on the ps a nd as bristles is probably due to more larvae having been affected during the 24 hour exposure. in the sc x As v car experiment, the magnitude of the increase in these bristles in this period and in the as bristle in the following period suggest that there probably exists, well in the third instar, an especially sensitive period for these bristles, which may correspond to Child's TEP for all the -112- bristles at lower temperatures. Two-by-two correlation coefficients were not calculated for the bristles in this experiment; but there is no reason to suppose that any more correlation existed here than in the 36°c. sc v car experiment. x h. Comparison with 36°c. sc x v car. A comparison may be made between the summarized data in Table 36 with the similarly summarizt9d 36°c. data from sc x v car in Table 28 (page 99). The data differ to some extent for each of the six bristles and indicate the presence of effective modifiers of sc-affected bristles in the two stocks. The data are alike in indicating an especially sensitive period for as bristles, and to alesser degree for as bristles, in the third instar. i. Comparison with 40°c. sc x sc. A comparison may be made between the summarized data of Table 36 and the similarly summarized 40°c. sc sc data in Table 22 ( page 86). x The most notable differ- ence between the two sets of data is in the response of the ps and as bristles, which was also the case in the comparison of the two similar sc x v car experiments. There appear also to be differences between the other bristles in these two sc x sc experiments. Curiously enough, the response of the or frequency to 36°C. in the sc response to 40°. in the sc x sc mating resembles more its x its response to 4o 0 c. in the sc v car mating than it does x sc mating. In both -113- the 36°C. sc x sc and the 4o 0 c. sc x v car experiments there is a reversal in the direction of or bristle frequency change beginning in the third instar. Since much of the 40°c. sc x sc experiment was done on Moldex food similar to that used in the 36°C. experiment, it seems probable that the differences between the two experiments reflect differences in bristle/ tesponse to the two temperature treatments. -114- 4. Summary and conclusions from the 36°c. experiments. (1) Data are presented from sc males from the mating sc v car and sc x x sc which show the effects of nearly lethal exposures to 36°C. on the frequencies of six of the scaffected bristles. (2) Each of the bristles differs in some respect in its response to 36°c. in the sc x v car mating as compared to its response to 36°c. in the sc x sc mating. (3) The ps frequency increases throughout larval life in sc x v car, and especially in the third instar. x sc it increases only during the third instar. In sc (4) The as frequency increases practically throughout larval life in both matings, but relatively more in sc x sc. The increase is especially marked in the third instar in SC X SCe (5) The an frequency is unaffected in sc x v car. It increases in the first, second, and late third instars, a nd decreases in the first half of the third instar, in sc x sc. (6) The pv frequency decreases throughout larva l and early pup~l development in sc third instar in sc x x v car. It decreases in the late sc. (7) The oc frequency increases in the second instar, decreases following 24 hours exposure and increases follo wing 30 hours exposure in the late third instar, and decreases in the early pupal period in sc x the first two instars in sc sc. x v car. It increases in (8) The or fre quency increases throughout larval and early pupal development in sc x v car. It increases in the first -115instar and decreases in the third instar in sc x sc. (9) The effect of 30 h ours exposure is gre a ter than that of 24 hours on the ps a nd as fre quencies in the third inst ar of sc x v car. The ef f ect of 24 hours is gre a ter than the effect of 12 hours on the ps and as fre quencies in the third instar in sc x sc. These differences do not extend in general to the other affected bristles. (10) The greater effect of exposures durin g the third instar upon the ps frequency in both matings, and upon the as frequency in sc x sc, suggests a period of hi gher sen- sitivity to the temperature treatment, possibly related to Child's TEP for all the bristles a t lower temper a tures. (11) The lack of correlation between any two of the bristles in the sc x v car experiment indic a tes that the 36°C. tre a tment a ffected the several bristles independently of each other. (12) The da ta confirm those from the 4o 0 c. experiments in establishing a TEP for sc which extends for one bristle or another practic a lly throu ghout the developmental period, a s it relates to bristles. (13) In one or both matings, each of the several frequencJes responds differently to 36°c. than it does to 4o 0 c., the difference being most pronounced in the ps, as, and oc bristles. (14) The nature of the change in bristle fre quency in sc males following a nearly lethal exposure to high temperature depends upon (a) the temper a ture tjeatment, (b) the bristle modifiers present in the sc stock, (c) the bristilie fre quency under consider a tion, and (d) the developmental period dbring whic~ the treatment takes place. -116F. Other Temperature Experiments with sc Males. 1. Effedts of 40 and 36°c. on F 1 males from treated parents. Bristle frequencies were determined in F 1 males from treated sc flies of the 40 and 36°C. experiments discussed above. The parent flies were from the longest one or two of the exposures in each developmental period tested in these high temperature experiments. Each developmental period was represented by from 2000 to 5000 observations. In no case did the bristle frequencies of these F1 sc males deviate significantly and consistently from the respective freq\§\u~cies of their cousin control F 1 raised under similar conditions of food, laival density, and temperature. At the time of exposure to the high temperature, these F 1 individuals were probably primordial or developing germ cells in the gonads of the larvae or pupae which were being exposed to the high temperature. Fro~ the results it seems certain that the high temperature treatment which is able in some way to effect a change in adult bristle frequencies of sc larvae is not able to effect such a change during the germ cell stage. -117- ~. Experiments with lower temp er a tures. During the winter and spring school terms of 1936 I attempted a series of experiments designed to test sc male bristle fre quency changes following a three hour exposure to one or a nother of a graded series of seven 0 temperatures between 30 and 40 c. The exposures were made, in one experiment, during the first instar, and, in another experiment, during the second instar. females x The mating was sc v car males as in the 1934-1935 experiment. Un- fortun a tely, it was not possible to secure environmental conditions constant enough to make the data suitable for detailed analysis. sented here. For that jeason, the da ta are not pre- It appeared to be probable, however, that such a short exposure to temperatures below 2.pproxima tely 390c. during these two developmental periods does not effect signific ant changes in sc ma le bristle fre quencies. -118G.Temperature Ef_fects on the Dominant Manifestation of sc. 1. Introduction. The gene sc is normally classed as a recessive gene because of its marked effect on the bristles when in homozygous condition. At ordinary room temperatures it does not show an appreciable dominant effect on these bristles, with the exception of the coxal bristle on the third pair of legs. This last bristle shows a high frequency of ab- sence in sc/non-sc females. It was noted in the 1934-1935 4o 0 c. sc x v car experiment reported above that sc/v car females do occasionally lack a bristle of those which were being tabulated in the sc males; and that this phenomenon seemed to occur more frequently in treated than in control females. The prelim- inary experiment with 36°C. in the early months of 1935 revealed a profound decrease in frequency in certain ~bristles in sc/v car females. sc x As a result, in all subsequent v car experiments, at both 40 and 36°C., the bristle frequencies were tabulated from the females as well as from the males. The conditions and procedure of these experi- ments have been described in connection with the male data. A 36°c. experiment was also performed with the synthetic wild stock derived from the sc and v car stocks, in order to determine the effects of temperature upon non-sc modifiers of bristle frequencies. The results make possible an estimate both of the probable dominant effect of the sc gene itself and of the probable effect of 36°C. upon the sc reaction. -1192. The ef fe cts of 4o 0 c. on sc/v car females. Tabulation of sc/v car bristle da ta in the 40°c. experiment was limited to part of the 2½ and 3 hour exposure groups in the 96 to 106 hour developmental period, and to all of the exposures in the embryonic (0 to 10 hours) and early pupal (120 to 130 \ hours) periods. The data are given in Table 37 in a form similar to that used in the tables of male bristles. In this case, however, each entry represents the percentile frequency of bristles absent per half-fly. Since the usual hybrid fly displays all of these brisltes, the only measurable effect of temperature was a decrease in bristle frequency. Because there were no apparent differences between the exposure groups with respect te bristle frequencies, the several exposures are presented together in each of the periods in this table. The data show a small but signifi- cant temperature effect upon the ps, as, pv, and oc frequencies in the Oto 10 and 96 to 106 hour periods. In the 1io to 130 hour period, the effect upon the pv frequency was less marked and there does not appear to be any difference between control and experimental or frequencies. Although there are no data from the intermediate development~l periods, the impression sustained through those series wa,s that there was a slightly larger frequency of absence, especially oftbe pv bristles, in the treated females. It is probable, there- fore, that the tempere.ture treatment produced a small effect upon heterozygous female bristle frequencies throughout the entire developmental period of the bristles. 11 -o. 00? :!o.oo? .. 0.073 :!0.031 •0.092 :!0.032 .. 1.02 !0.16 0.092 + -0.032 1.65 .-0.14 0 0.080 + • -0. 030 0.092 + -0.032 8736 ,.o.o~ _0.02 il.33 -0.15 0 •0.063 !0.035 0.30 :!0.15 ,.o.o7 -0.03 0.09 ~0.04 5782 a 11 ;l; 2;3 Difference from Control - Total - 120 to 130 0.07 l0.07 2.43 :!o.42 0.13 !0.09 0.029 !0.015 oc 0 0.07 :!O.O? 1358 96 to 106 0.63 :!o.o? pv 2.13 !0.29 0.00? :!0.007 an 0 2½;3 "a";l;2 ... o.13 -0.09 0.007 :!0.007 0 . as ps 0.13 -0.09 13,648 Obs er. 1596 Exposure Oto 10 Control Period sc/v car - 4o 0 c. - Embryonic, Larval, Pupal Exposures - TABLE 37. 0 0 0 0 0 0 or I I 0 I:\) I-' -1213. The effects of 36°C. on sc/v car females. Preliminary experiments in 1935 showed that exposures of one to two days to a temperature of 35 to 36°C. during larval development produced a marked decrease in bristle frequencies in sc/v ca r females. the 1937 3 6°c. sc x Accordingly, in v car experiment the heterozygous sc females were recorded for frequency of absence of br .i stles. a. The da ta. The data of sc/v car females from thl§/~ experiment are g iven in Table 38, in a form similar to that used in Table 37. They sh·ow that the temperature treatment affected the ps, pv, and oc frequencies in the later larval and early pupal periods. The temperature effects are strongest on all bristles in the thiid instar and early pupal period exposures. There are differences between the several bristles, however, with respect to the time of maximum sensitivity. ps Thus, the and as frequencies appear to be most strongly affected by the 30 hour exposure during the 68 to 105 hour period; the an frequency change reaches a decided maximum in the 30 hour exposure durin g the 90 to 128 hour exposure; the pv frequency change is rel a tively very large in both exposures of the 90 to 128 hour period and the one exposure of the 116 to 148 hour period; t he oc bristle is most affected by the 30 hour exposur~ of the 90 to 128 hour period and by the exposure in the pupal period; and the or fre quency chan ge is most prominent durin g the pupal period exposure. It appears, then, that in sc/v car females, the temp er a ture effect upon the severa l bristles differs significantly from bristle to TABLE 38. sc/v car Period Exposure Control - 36°C. - Larval and Young Pupal Exposures Obser. ps as an pv QC or 7752 0.013 .. 0.013 0 0 0.077 .. 0.032 0.013 .. 0.013 0 20 to 53 16;22 1602 0.125 :!0.090 0 0 0.69 :!0.21 0.062 :!0.062 0.062 :!0.062 44 to 82 24;30 4652 0.086 ~0.043 0 0.022 !0.022 0.37 !0.09 0.086 !0.043 0.022 !0.022 24 1304 0.230 '!0.133 0.077 '!0.077 0.077 ~0.077 0.230 ~0.133 0.154 ~0.109 0 30 1140 1.49 :!0.36 1.49 :!o. 36 o.53 =0.21 1.32 =o.34 o.44 :!0.20 0 24 1866 0.81 :!0.21 1.18 :!0.25 1.23 :!0.26 13.89 :!0.80 1.61 :!0.29 0.32 :!0.13 30 872 0.69 :!0.28 o.57 :!0.26 4.59 :!0.70 9.18 :!0.98 4.82 :!0.72 1.72 :!0.44 24 844 0.47 :!0.24 0.47 !0.24 0.83 :!0.31 13.88 ':!1.19 4.62 :!0.72 9.00 !0.98 68 to 105 90 to 128 116 to 148 I I-' l,\J l,\J I -1 23- bristle, either in ti me or in amount of ef f ect, or in both of these ways. h. Discussion. To det ermine whether the temperature effect was upon pa rt or a.11 of the population in this experiment, leftright correlation coefficients were calculated for each pa ir of bristles. The bristle frequencies were very low in (E;~~ control series and in the first and second inst a.r experimental series. No females in these series l a cked both bris- tles of any pa ir. Beginning with the 30 hour exposure in the 68 to 105 hour period, such females a ppe a:=red in the experimenta l series, and indicated marked left-ri ght correlations for several of the bristles in this and succeeding periods. These coefficients were as follows: pv, ~o.564; oc, ~o.270; and or, ~o.452. an, ◄ 0.634; The t value for each coefficient was very l arge, indicating clearly a deviation from 0. From the size of the coefficients it seems certain that the temper a ture ef fect was not the same on the entire popula.tion, and that only a part of the population was affected by the temperature so far as the bristles are concerned. .Q. It is possi b le that her, as in the case of the A male ps and as bristles, we may be dealing with a special sensitive period like Child's TEP. In these females, how- ever, the TEP dififers from his in that it does not appe a r to h ave exactly the same position in l a te develo pment for any two of the bristles. Correl a tion coefficients were also ca lcul a ted for each of the tw6-by-two combin a tions of the an, pv, oc, and or bristles in thsi experiment. There a ppe a red to be no dif- -124ference in the rough data between the 90 to 128 and 116 to 148 hour periods; and they were therefore grouped for these correlation calculations. The an bristle showed a positive correlation with the oc bristle but no correlation with the pv and or bristles; the pv bristle was correlated with the or bristle. All of these coefficients were of approximately the same size, ranging from +0.19 to +0.22, with t values between 8 and 9. They indicate that these bristles tend to be affected together in the sc/v car females. In this respect these data differ from those of their sib males which did not show such tendencies. -e. Comparison with sib males. The data of the 36°C. sc/v car females in Table 38 may be compared with the sib male data in Table 28 (pa.ge 99). They are alike in showing probably the same general period of high sensitivity to the temperature treatment in respect to the ps and as frequencies, and a decrease in pv frequency in every developmental period. The frequency change, however, is oppositely directed in the ps and as bristles between the sexes. And in the effects upon the other bristles the data are opposed to each other. The bearing of these observations on the problem of sc dominance will be discussed later. d. Comparison with 4o 0 c. sc/v car. These data may also be compared with those of the 4o 0 c. sc/v car females in Table 37 (page 120). The two ser- ies differ in conditions in that the 36°C. series was raised -1 25 - on Moldex food. That may be the reason for the slight · difference in pv frequencies in the controls. In their comparable developmental periods the data resemble each other to the extent of showing, in each case, affected ps, as, and pv frequencies. But otherwise they differ markedly in bristles affected and in amount of effect. Since the preliminary experiment on non- Moldex food showed comparable results in the 35 to 36°c. temperature range it seems probable that the differences between the 40 and 36°C. data are due to the difference in temperature treat- ment. 4. The effects of 36°c. on wild-type flies. a. Introduction. In December of 193; I tested the effects of nearly lethal exposures to 36°C. on the bristle frequencies of the wild-type stock synthesized from the sc and v car stocks, considering the same bristles as in previous experiments. This synthetic wild stock contained the nor- mal allele of sc from the v car stock and the normal alleles of v and car from the sc stock. The rest of its genes were in unknown proportions from the two stocks. The Y-chromosome was from the v car stock; and the original egg cytoplasm was from the sc stock. This wild stock was inbred without selection for 13 generations, and then continued by mass matings without selection for an additional year and a half before this experiment. The conditions and plan for this expetiment were the same a s those in the other 36°C. experiments. The wild -126~tock, however, could not wi thstc.nd 30 hour exposures to 36°C. satisfactorily. Exposures were 12 to 24 hours in length. b. The male data. The datajshowing the effects of the 36°c. treatment on the synthetic wild-type males are given in Table 39. The temper a ture effected a decrease in ps and as fre- quencies throughout larval life. An especially pronounced decrease followed the 24 hour exposure during the 72 to 107 hour period. The an frequency was probably unaffected. (There was but one missing bristle in all the treated males) The pv and oc bristles were affected mainly during the 72 to 107 hour period. The or frequency was affected during that and the following period. c. Discussion. During the periods of pronounced temperature effect, there were positive left-right correlations between the affected bristles in wild-type males, indic a ting that the temperature did not affect all of the popula.tion similarly. In the case of the ps bristles there was no correl- ation in the control and first and second instar series. In the remaining periods the correlation w~s as follows; 72 to 94 hours, +0.366; 72 to 107 hours, +0.336; 96 to 124 hours, +0.189; and 120 to 154 hours, +0,142. The only correlation in the as bristles appeared in the 72 to 107 hour period, when the coefficient was +0.336. The pv bristles showed a correlation of +0.15 in the same period. The or bristles showed a correlation in the combined 7 2 to 107 and 96 to 124 12;24 12 24 12 12;24 48 to 81 72 to 94 72 to 107 96 to 124 120 to 154 934 1328 1008 768 · 2442 1592 12;24 24 to 55 Obs er. 3936 Exposure Control Period 1.39 !0.38 4.06 :!0.54 50.50 !l.58 ¼3.80 -1.25 2.94 :!0.34 3.83 :!0.48 1.20 ,& -0.18 ps 1.07 !0.34 4.82 !0.59 39.40 :!1. 54 6.14 !0.87 2.42 :!0.31 1.38 :!0.29 o.46 ♦ -0.11 as 0 0 0.10 :!0.10 0 0 0 0 an 0.11 ~0.11 0.38 :!0.17 2.38 :!Q. 48 0 0.08 :!0.06 0 0 pv ~ 0 0.23 :!0.13 0.69 :!0.26 0 0 0 o. 10 -0.05 + oc Wild Type CC - 36°C. - Larval and Young Pupal Exposures - TABLE 39. 0.11 !0.11 0.68 :!0.23 1.29 :!0.36 0.13 :!0.13 0 0 0.03 + -0.03 or I -.::z z-0 I I-' -128hour periods of +0.18. ranged from 3 to s. The t values of these coefficients The an and oc bristles showed no left- right correlation. There proved to be positive correlations between some of the bristles in the 72 to 107 hour period. This amounted to ~0.265 (t value of 6) between the ps and as bristles; to •0.112 (t value of 2.5) between the as and oc bristles; and to •0.104 (t value of 2.3) between as and or. These coefficients suggest a tendency for these bristles to be affected together in this experiment. The period of marked effect of the temperature for all the bristles w~s in the 72 to 107 hour period. Presum- ably the 24 hour exposure affected more larvae than did the 12 hour exposure; it may also have affected some larvae through a longer portion of a hyper-sens/itive period, such as appeared in the earlier 36°c. experiments. It may be that the hyper- sensitive period is shorter for wild-type males and more like that found by Child in scat viable tempera.ture. d. Comparison with 36°c. sc males from sc x v car. The data from the 36°C. wild-type males may be compared with the summarized data from the 36°C. sc males from sc x v car in Table 38 (page 122). It is te be kept in mind that all of the frequency changes in the wild-type are decreases and that the direction of change in sc males is indicated by the sign of the coefficient of change. It can be seen that the greatest differences in the two experiments are in the oppositely directed effects of the temperature on the ps, as, and or frequencies, the first two of which are -129especially strongly affected in both~ experiments. The comparison of these data will be discussed later with reapect to their indications concerning the responses of sc and modifiers to the temperature treatment. e. The female data. 0 The data from the 36 c. wild-type females are given in Table 40. They show that the bristle frequencies were affected to some extent in every developmental period, and notably following a 24 hour exposure during the 72 to 107 hour period. The main effedt appeared in the ps and as frequencies, which showed a maximum decrease during the above period. The an bristle was unaffected at all periods. The pv fwequency was affected only slightly, during the third instar and early larval periods. The oc was also only slight- ly affected, during the 72 to 107 hour period. The or fre- quency was affected during the pupal period onlJ. f. Discussion. Left-right correlation coefficients indicate that the temperature effect was not on the entire population with respect to some of the bristles. The ps bristles showed a correlation of ~6.30 (t value of 7) in the 72 to 107 hour period, and a correlation of ~o.514 (t value of 12) in the 120 to 154 hour period. The high coefficient in the latter period, coupled with the low frequency, suggests that the temperature effect was upon the few larvae ptesent rather than upon some of the many pupae. The as bristles showed a correlation of ~0.22 (t value of 5) in the 72 to 107 hour 24 12 7 2 to 107 96 to 1 24 12 ; 24 12 72 to 94 1 2 0 to 154 12 ; 24 48 to 8 1 1. 2 7 :! 0 . 35 ,. 0 .36 - 0 .1 4 1 93 4 1022 10.13 - 0. 97 . :t0. 2 8 0 . 49 0.1 8 :!0.09 0.07 ~0.07 0 ps 96 6 6 08 220 0 149 8 12 ; 24 24 to 55 Obs e r. 383 2 Ex posure Control Pe riod lw , 1 0 . 39 :! 0 . 20 ..o.1s - 0 . 09 - 0 .54 ... 2 . 90 0 0.05 + -0. 0 5 0 .1 3 ~0.09 0 as .. 0.10 -0.07 0 . 29 :! 0 .17 0 -0.10 ... 0.10 0 0 0 0 .03 !0.0 3 l)V 0 0 0 0 0 0 an 0 0 0 0 oc 0 0 ... 0 .21 -0.15 . ld T ype ++ oc - 36°c • - La rva l and Young Pupal Exposures - TABLE 40. 1.17 :!0.34 .. o. 2s - 0 .1 2 0 0 0 0 0.0 3 !0.03 or I I I-' vi 0 -131period. The or bristles showed a correlation of ~o.38 (t value of 12) in the 96 to 124 hour period, and a correlation of ~0.16 ( t value of 4) in the 120 to 154 hour petiod. The high value of the 96 to 124 hour coefficient for the or bristles, coupled with the low frequency, suggests that the temperature effect was upon the few pupae present in these cultures, rather than upon a few of the many larvae. g. Comparison with wild-type males. A comparison of the female data of Table 40 with the wild-type male data of To.ble 39 (page Li?) reveals a sexual dimorphism both in the control and in the experimental bristle frequencies. The males have fewer bristles (i.e., have more bristles missing) both at 25°C. and in the series 0 exposed to 36 C. In the conttol series, the difference is evident only in the ps and as bristles; but in the treated series it extends to the other bristles in the period of greatest effect. This dimorphism is typical of many sex-linked characters. Bar and alleles (see Review of Literature) show such a behavior; and Child's data show it clearly for sc. Payne and Plunkett had already reported that sc females have more bristles than do sc males. Although no bristle counts were made in the females in my experiments, it was obvious that sc females showed greater numbers of the bristles used here. It was frequently necessary to examine many flies to find one female with all the bristles missing. The dimorphism present in this synthetic wild stock suggests -132- that part of the cause of missing bristles is due to one or more sex-linked genes. No attempts have been made to locate the gene or genes concerned. There is an additional difference between the male and female data which is of interest; na1nely, in the sensitive period of the or bristles. It has been argued that this period lies in the pupal stage in wild-type females. In the males it appears to lie in the third instar. Al- though no attempts were made to determine accurately the relative lengths of the egg-larval periods in the two sexes in this race of flies, it is reasonably certain that there was no sucu great difference as there is between the maxima of decrease in or frequency. h. Comparison with 36°c. sc/v car females. A comparison may be made between the wild-type female data of Table 40 and the 36°C. sc/v car female data of Table 38 (page. 122). It is evident that while the two sets of data resemble each other in respect to the general period of highest sensitivity late in the third instar, they differ wiaely in relative amount of effect. The ps and as bristles are relatively more affected in the wild-type females; the pv, oc, and or bristles are relatively more affected in sc/v car females. There appears to be no signif- icant difference between the two types of females at 25°c. The bearing of these observations on the problem of the dominant effect of sc will be discussed in the next part of this report. -1335. Effects of heterozygous scat 36°C. It has been noted that a t 25°c . sc shows practically no dominant effect in respect to the six bristles considered in these experiments. The bristle fre quencies are practically identic a l in sc/v c a r a nd wild-type fema les. There are differences in the bristle fre quencies between these two types of females at 36°C., however. By comparing these differences with the differences b e t ween the wild-type males and the sc males from sc x v car, it is possible to sug- gest the direction of the dominant ef f ect of sc at 36°C. To make this comarison possible, it is necess a ry to assume that flies from the sc x v car cross and flies of the wild- type race derived from that cross carry essentially the same bristle modifiers, so that the only effective difference between the two types is with respect to the sc gene itself. The data of the sc males in Table 28 (page 99), of their sib sc/v car females in Tab le 38 (page 122), t" ~ the wild-type males in Table 39 (page 127), and of their sib wild-type females in Table 40 (pa ge 130) are presented in a much sitjlplified surmna ry in Tab le 41. Ea ch entry in this tab- le represents the direction of change in bristle frequency in these 36°c. flies when compa red to their respective 25 to 26°C.controls, the change being evident in one or more developmental periods. A plus sign indicates an increase, a minus sign a decrease, in bristle frequency. In the c a se of the oc bristle in sc males, the entry of both si gns indicates tha t there was an increase in one period and a decre a se in another. When the two types of flies of the same sex showed a similarly directed fre cuency cha nge, the lesser of -134- TABLE 41. Reactions of SC and Modifiers to 36°c. Table Type of Fly 39 wild o' 28 SC 40 wild ~ 38 sc/v car~ d ps + as . an pv 0 (-) 0 0 (-) (-) (-) oc or + + - (.;.) ( ) -135- the two has bee n e nclosed by parentheses in Table 41. In wild-type males, the ps and as frequencies decreased greatly at 36°c. In sc males they increased gre a tly. in wild-type females and in sc/v car females they decreased; but the decrease was less in sc/v car than in wild type. This suggests that the 36°C. change in sc effect, evident so strongly in males, i$ prob ably also responsible for the lesser decrease in sc/v car females. In other word, sc acts as a slight dominant in heterozygous females in the s:ame direction in which it acts as a recessive in sc males, so far as these bristles a re concerned. The an frequency did not change in either type of males nor in wild-type females. sc/v car females. It decreased, however, in If a decrease, proportionately as small, had occurred in sc males, it could easily have escaped attention, because of the already very low an frequency. This decrease in sc/v car females can therefore be interpreted as a dominant effect of scat 36°C. which is not necessarily different in direction from its honozygous effect in sc males at 36°c. The pv frequency decreased in all four types of flies; but it decreased considerably more in sc males and sc/v car females than in wild type flies. The dominant ef- fect of sc in this case seems· to be in the same direction a s its homozygous effect. The oc frequency decreased very sli ghtly in thet third instar in both sexes of 36°c. wild~type f lies. In the sc/v car females it decreased sli ghtly in the first, second, a nd first h a lf of the third inst a rs; and considerably more in -136- the l a st h a lf of the third in s tar ~nd in the ea rly pupal period. In sc males, the oc fre quency increased in the second instar; and did not change in the first half of the third instar. In the second half of the third instar it decreased following a 24 hour exposure and increased following a 30 hour exposure, both changes being small but statistically si gnificant. It decreased in the early pupal period. With respect to thi- bristle frequency, the effects of heterozygous sc, following the 36°c. treatment, seem to be different from the effects of homozygous sc in males. The differ- ence is more pronounced in the second instar; it may disappear entirely in late development. The or frequency decreased in both sexes of wildtype flies, somewhat earlier in males than in females, following the 36°C. treatment. In sc males it increased in every developmental period. In the heterozyg ote, it de- creased late in development, considerably more strongly than in the wild-type females. In this case, too, the dominant effect of sc seems to differ from its male homozygotic effect, but this time late in develo pment, rather than early. From the above considerations it seems probable that the ultima te effect of sc is so altered by nearly lethal exposures to 36°c. that it affects the bristle fre quencies when in heterozygous condition; but that it does not necessarily affect them all in the same direction as when it is in homzygous condition in sc males. The differences in direction of effect, apparent in the or bristles and suggested in the oc bristles, may be due to the difference in sexes. -137Child, however, did not find a difference in direction of effect of viable temperatures on any of the several bristles when compa ring males with females. Another fact of interest suggested by these comp a risons is that the temperature effect is upon the sc-dep endent reaction, or reaction system, and not j l§U)t upon modifier reactions which the presence of sc makes visible. The data from sc and wild-type males suggest also that in the cases of the ps, as, and or bristles, the effects of sc following the hi gh temperature treatment are opposite to the effects of the modifiers; and that the sc effect is much the dominant one in each case. -138- 6. Summa ry and conclusions. (1) During embryonic, late larval, or early pupal development, nearly lethal exposures to 4o 0 c. effect decreases in ps, as, and pv bristle fre quencies in sc/v car females. The oc frequency is also reduced during the embryonic and late larval periods. (2) During any one of several selected periods in larval and early pupal development, nearly lethal exposures to 36°c. effect a decrease in one or more of the bristle frequencies in sc/v car females. (3) The ps, pv, and oc frequencies decrease in all tested periods. (4) The as and an frequencies decrease in the late third instar and early pupal periods. i'r\stcir (5) The or frequency decr~ases in the late third and in early I\ pupal periods. (6) All the bris(J}t:es show periods of maximum effect late in development; but they differ among themselves as to the periods or exposures which produce the maxima. (7) During any one of several selected periods in larval and early pupal development nearly lethal exposures to 36°c. effect a decrease in one or more bristle freE1uencies in males and females of a wild-type stock derived from the sc and v car stocks used in earlier experiments. (8) The ps and as frequencies decrease throughout larval life in both sexes. (9) The an frequency is unaffected in both sexes. (10) The pv frequency decreases in the third instar in both sexes. -139- (11) The oc frequency decreases only slightly in both sexes, the decrease appearing in the third instar series. (12) The or fre quency decreases in the third instar in males, and in the early pupal period in females. (13) In all bristles the decreases ap pear to be gre a ter in males than in females. (14) All frequencies (with the exce ption of that of the female bristle) show a decrease in the third instar. (15) The response of sc males to the 36°C. treatment is opposite to that of wild-type males with respect to the ps, as, and or frequencies. (16) The effects of the 36 6 c. treatment are more pronounced in sc/v car females than in wild-type females with respect to the an, pv, oc, and or frequencies; and are less pronounced with respect to the ps and as frequencies. (17) The da ta are discussed with reapect to a dominant effect of sc in sc/v car females. (18) At 25°C. sc is not dominant in any of the bristles considered. (19) Following the 36°c. treatment sc is probably dominant to some degree in all of the bristles considered; but its effect as a dominant may not always be in the same direction as its homozygotic effect in sc males. (20) The effect of the 36°c. treatment is upon the effect of sc as well as upon the effects of modifiers. (21) In respect to the ps, as, and or bristles in males, the effe c t of sc following the 36°c. treatment is dominant to the effects of the modifiers. - 140- H. Inter pret a tions of the Experiment a l Results. In presenting the da t a in previous sec t ions of this report, little discussion has been made of their probab le theoretic a l signi f ic ance. There are a numbe r of questions which may be asked along this line, and we may turn now to the answers to some of them. 1. The meanin g of the change in bristle frequency. One of the first questions to be considered is, "What does a. bristle fr e quency mean?u Child (1936) has attempted to answer that question, and his reasonin g applies here as well as~ his experiments. He p ictures bristle forma tion as an instance of the clas s ical 11 all-or-none 11 principle. The bristle is either formed or not formed, in any given in s tance; pa rt bristles are not found. This suggests that a cert a in threshold concentr a tion of bristle-forming su bstance wus t be present to insure the formation of a given bristle. An amount just beneath this threshold value will produce no bristle; an amount any dist a nce above it will result in the formation of but one brist l e. In any in d ividual fly the presence or absence of a g iven bristle can be thought of as being determined by the presence or absence, at the time of bri s t l e r orma tion, of an amount of bristleformin g subst a nce equal to or exceeding the threshold concentration. The bristle fre quency, howev er, is determined by observa tions on a popul a tion of individuals. It is Child's view tha t in a popul a tion, the am ount of bristle- f orming substance avail able for a given bristle is not the s ame in -141every fly. Instead it is distributed at random around a mean concentration, in greater or lesser amounts to the many individuals, the distribution following the probability-integral curve. The mean bristle frequency of a population represents the point on the distribution curve which coincides approximately with the threshold concentration of substance necessary for bristle formation. The frequency indicates the pro- portion of the population in which the amount of bristleforming substance available for the bristle equals or exceeds the threshold concentration. From this reasoning it can be seen that a race of flies which manifests 100% presence of a given bristle is one in which the mean amount of substance availagle for bristle formation is so far above the threshold level that normal variations from individual to individual do not produce any individuals with a sub-threshold concentration. Similarly, the complete absence of a given bristle in a population indicates that the mean amount of substance available for bristle formation is far below the threshold. t A race of f, lies which shows a mean bristle frequency between O and 100% for a given bristle is one whose mean concentration of bristle-forming substance is in that case nearer the threshold concentration. On this view it is logical to interpret the effects of both environmental and genetic factors on mean bristle frequencies as representing changes in the riean concentrations of bristle-forming substances. The gene sc in homozygous con- dition shifts the mean concentration considerably below the threshold level in the case of the bristles considered in the experiments discussed in this re~ort. It seems probable that -142- the effect of larval density, Moldex, genetic modifiers, and temperature is to shift these same mean concentrations nearer or farther away from the threshold, depending upon whether the indicated change is a.n increase or a decrease in frequency of bristles present. In the experiment with wild-type flies, the mean concentration was so far above the threshold level that the only practically measurable change was a decrease in mean concentration which resulted in more individuals with missing bristles. 2. :Nature of the temperature effect on a bristle. A second question which may be asked is, "What is the nature of the temperatur~ effect on a bristle?" A large amount of literature has been published on the effects of temperature on living matter (see Belehradek 1935). The general indications are that temperature affects the chemical and physical processes which constitut~ the life functions of the individual. This it does, not by creating new chemical reactions, as a rule, but bY increasing the rates of reactions already taking place to a greste1or lesser extent, and by destroying substa. nces such as enzymes, proteins, etc., so as to make impossible the reactions in which they have an integral part. If the temperature erfect on sc bristles reported here wer• simply a change in the rate of a reaction which is of some importance to the ultimate amount of bristle-forming substance to be produced, then it would be expected that an increase in the length of the exposure woudd produce a measurable increase in the observed effect. With the possible exception of the results from 350c. exposures on the frequencies -143- or ps and as bristles, this is not the case in these experimental reaults. Rather, we are confronted with a tempera- ture effect, which, once it has manifested itself, does not appear to change with increasing length of exposure. It seems probable, therefore,. that we are not dealing with a temperature effect on a simple chemical reaction. If the temperature effect on sc bristle frequencies results from a destruction of an enzyme or some other necessary substance, then it might be expected that the effect would appe ar after a given exposure, more or less as an,~ instantaneous shock. Providing the effect is complete/in all • members of the population, it would not ap_pea.r to increase with an increase in length of exposure. This is, in fact, what is observed in these ex periments. It seems logical, therefore, to offer destruction ·of an enzyme or some other necessary substance as a more probable explanation of these high temper a ture effects. This interpret a tion of the results carries more weight by reason of the fact tha t 36°C. and 40°c. are in the temperature range in which such destructive processes are known in some cases to take place. 3. Meaning of differences in bristle responses to temperature , A third question which may be asked is, uWhat is the meaning of the differences between the response of the several bristles to temp8rature treatments?" Child (1935) suggested that differential temperature effects on the s@ale chemical reaction or substance in different parts of the fly - due to differences in length of ~EP or of rate of reaction, or to both of these - could produce -144all the differe n ces observed in his experiments. The unity ' of the development a l pos;t ions of the short TEP's for all of the bristles ma de such an interpretation plausible. 0 In my experiments with 36 and 40 c., however, the many v ariations in position a s well as in duration 6f the TEP's make such an interpretation, in general, too simplified. Only the ps and as bristles parallel each oth er sufficiently in res ponse to the tempe ra.ture treatment to a llow such a n in- terpretation. While they may h ave a more identic a l history than the other bristles, there is no conclusive evidence that such is the case. Cert a inly, ~n the whole, these temper a ture data suggest an independence of development a l history. And they do also suggest that critic a l reactions t a ke place a t many periods in development for ea ch of the several bristles, possibly differing in na ture from bristle to bristle. 4. Developmental history of sc. A fourth question which may be asked is, 11 \Vhat do these results at 36 and 40°C. indicate concerning the develo pmental history of sc?" I ~has already been suggested that the results of these experiments indic a te the prob ab ility of sever a l reactiops being involved a t different times, possibly different for the several bristles. Powsner (1935) has developed the theory of Plunkett (193~) that development consists -of a long series of int e r a cting chains of reactions, leading from the emb ryonic pe riod throu gh the larval _p eriod and finally to more specific and less inter ~rel a ted reactions which form the imag in a l cha r ac ters. These reactions a re dependent upon -145- environmental conditions as well as genetic constitution. In a very real sense, every adult organ is dependent upon the successful completion of some reaction at any given time in development; and it is dependent in such a way that the failure of this reaction, either because of environmental or genetic influences, or both, results in an alteration of the subsequent adult organ. This is an extremely general picture; and it provides the basis for a description of practically any type of observed phenomena, as Powsner realized. But the temperature effects on sc bristles, demonstrated in my experiments, are so diversified that for the present, at least, only such a broad and general interpretation c an, I believe, be made of them. It appears, then, from the temperature effects, that the normal development of sc-affected bristles is dependent upon the successful compietion of one or more reactions taking place at any given time in development, from early embryonic life until the visible forma tion of the bristles from the trichogenic cells observed by Robertson (1936) in early pupal development. 5. Meaning of the viable temperature TEP's. A fifth question which may be asked is, "What do the results of these experiments suggest concerning the general concept of TEP as developed from viable temperature studies on Drosophila mutant chara.cters? 11 In view of the considerable contrast between the results of Child's experiments and those of the experiments reported here, in respect to the position and duration of the TEP of the sc genotype, it seems probable that the results of -146 v iabl e temp e ra ture studies on Dr os oph i la mut a nts in g eneral (see Revie w of Lit e r a ture) c a n not be int e r p reted to in- d ic a te the period of most important d e velo pmenta.l history for the cha racter in q uestion; a nd certainly not necessarily the on l y period when the cha r a cter in question can b e influenced by a chan g e in developmental temperatures. The via~le temper a ture TEP's can be said to indicate in each case the presence of a reaction or re a ctions which a re of imp ort a nce to the or g an in question and which h a ve temper a ture characteristics differing from those of development as a whole in the viable temper a ture range. In the cases of Bar, vestigial, a n d b ent win g , wher e the si b wild types h a ve been worked with in e a ch case, the viable TEP 's me.y also be considered to indicate a. tempera ture influence on the ultima te effect of the g ene on the ad ult character. cide !i.,n Tha t t h e TEP's h app en to coin- some cases with visible embryolo g ical processes does not necess a rily me a n a direct connection b etween them, a s was p ointed out by Miss Auerb ach (1936). The r e sults of these hi g h temperature expe riments with sc suggest that exp eriments si mil a r to these with cha ract e rs such a. s Ba r or vesti g ial mi ght s h ow as much bro a der TEP's for these ch a r a cters as has b een s h own for sc. 6. Activity of the scute g ene. A sixth q uestion which may b e asked is, 11 V/ha t do these resutts sugg est conc e rning the a ctivity of the sc g ene? 11 These d a ta a re not critical on the direct a ction of the g en e . Nor are any of the TEP data critical on this point. It can not be determined fro m TEP da ta if a g iv e n g ene acts before the time of the TEP , so t hat the effects of temp er a ture -147a re on cert a in of its p roducts; or if the g ene acts following the TE P and hence upon temperature affected materials; or if the temperature effect is upon the direct action of the gene itself during the indicated TEP. In the case of sc, the comparative data of the 36°C. experiments with sc and wild-type males which have probably similar modifying genes do suggest strongly, however, that the high tempere"ture treatment influences the ultimate effect of sc itself and not just the effects of modifiers which the presence of sc makes visible. -1 48 DiRest of tbe Thesi s (1) The literature concerned with temperature-effective period studies on a number of Drosophila mutant cha r a cters is reviewed. It indicates, in g eneral, a specific period in de- velopment in which changes in temper2ture can a lter the effects of a gi ven mutant gene. ( 2 ) Data from scute males given nearly lethal treatments with 0 supra-viable temperatures of 40 or 36 c. show that the phenotypic expression of this gene may b e changed to some extent at practically any time in development, from the layin g of the e gg to the app earance of the bristle cells early in pupal develo pment. Tbis is in conttast to the brief period of sensi- tivity found for this phenotype b y Child at viable temperatures. The n a ture of the effect depends upon (a.) the temperature tre a tment g iven, (b) the bristles considered, (c) the bristle modifiers p resent in the scute stock, and (d) the developmental pe riod in which the treatment is given. (3) Da ta from scute males, from heterozyg ous scute fema les, and from b oth se x es of a closely rel a ted wild-type stock, all g iven the 36°c. tre a tment, show tha t scute has a dominant effect on some of the b ristles under these conditions; and that the temper a ture alters both the effects of the scute gene and of the non-scute mo~ifying g enes on the bristles considered. (4) These d a ta su pport the theory of Plunkett, et al, tha t the phenotypic expression of a Mendelian cha racter is dependent up on the successful comp letion of a series of re a ctions extending throughout develo pment. -149Bibliography 1. Auerbach, Charlotte. 1936. The develo pment of the legs, wings, and halteres in wild type and some mutant strains of Drosophila melanogaster. Trans. Roy. Soc. Edinburgh 58:787-815. 2. Baron, A. L. 1935. Facet number in Drosophila melano- gaster as influenced by certain genetic and environmental factors. 3. Belehradek, J. J. Exper. Zool. 70: 461-490. 11 1935. Temperature a.nd Living :Matter••• Berlin: GebrJder Borntrager. 4. Blanc, R., and Child, G. P. temperature on the Bridges, c. B. culture. 6. Chen, T. Y. The effect of high trunca.te rea.ction 11 in Drosophila (Abstract). Genetics 23:141. melanogaster. 5. 11 1938. 1932. Apparatus and methods for Drosophila Amer. Nat. 26:250-273. 1929. On the development of imaginal buds in normal and mutant Drosophila melanogaster. J. Morph. and Physiol. 47: 135-199. 7. Child, G. P. 1935. Phenogenetic studies on sc 1 of Drosophila melanogaster I. The associations between the bristles and the effects of genetic modifiers and temperature. Genetics 20:109-126. -15 0 8. Child, G. P . 1935. Pbeno g enetic studies on sc 1 of Drosophila melanogaster II. fective period. 9. Child. G. P. 1936. Genetics 20:127-155. Phenogenetic studies on §_£_J:. of Drosophila melanogaster III. ature on sc 5. 10. The temperature-ef- The effect of temper- Genetics 21:808-816. Child, G. P., and Albertowicz, T. 1937. The effect of Nipagin on the wing size of vestigial in D. melanogaster. 11. Crozier, w. J. (Abstract) . Genetics 22: 188. 1924. On biological oiidations as func- tions of temperature. 12. J. Gen. Physiol. 7:189-216. Dobzhansky, Th., and Duncan, R. N. 1933. Genes that af- fect early developmental stages of Drosophila. melanogaster. Wilh. Roux' Arch. F. Entw. mech. Organ. 130:109-130. 13. Driver, E. c. 1926. The temperature-effective period - the key to eye-facet number in Drosophila. J. Exper. Zool. 46:317-332. 14. Driver E. c. _Q. Drosophila. 15. Driver, Olive Temperature and gne expression in 1931. W. :A J. Exper. Zool. 59:1-28. 1931. Temper a ture-effective periods in reciprocal crosses in Drosophila melanogaster. J. Exper. Zool. 59:29-44. 16. Dubi ni n, N. P . . 1929. melanogaster. Alle lomor phentreppen bis Drosophila. Biol. Zbl. 49:328- 339. -15117. Dubi nin, N. P., and F riesen , H. 1932. Die Unmoglicheit ,, einer Erklarung des Fall s d er Treppenallele scute vam Standpunkte der Goldschmidtschen physiol u Theorie der Vererbung. 18. F isher, R. A . 1928. Workers". 19. Gable, Jeanne. 11 Biol. Zbl. 51:507-526. Statistical :tv:ethods for Research Second edition. 1938. ischen London: Oliver and Boyd. Temp e r a ture effects on the wings of vesti g i a l- Newing (vg 33 f 13 ). ( Abstract) ~ Genetics 2 3:147-148. 20 . Goldschmidt, R. 1931. Die entwicklung sphysiolog ische Erkl~rung des Fal ls der sogennanten Tre ppenallelpmorphe des Gens scute van Drosophila. Biol. Zbl. 5 2 :147-162. 21. Goldschmidt, R. Zbl. 22.- 1935. Gen und Aussencha rakter III. Biol. 55: 535-554. Goldschmidt, R. 1937. Gene and cha racter rv. Further data on the development of wing mutants in Drosophila. U. of Calif. ?u b. 41: 2 77- 282 . 23 . Goldschmidt, R. 1937. and v g a lleles. 24. Harnly, M. H. 1930. Gene and ch8r a cter VI. Dominigenes U. of Calif. Pub. 41:297-312. A critical temperature for leng then~ng of the vesti g ial wings in Dr oso phil a mel an ogast er with sexually dimorphic effects. J. Expe r. Zool. 56:363-368. -15 22 5. Harnly, M. H. 1936. The temp e rature-ef f ective periods a nd the g rowth curves f or len g th and a r ea of the vesti g ial wing s of Drosophila melanogaster. Genetics 2 1:84-103. 26. Ha rnly, M.H., and Harnly, Marie Lecuyer. 1935. The ef- fects of temper a ture on the wings a nd the eyes of the dimor phos vesti g ial combin a tion in D. melanogaster. 27. J. Exper. Zool. 72 :75-99. Harnly, M. H., and Harnly, Marie Lecuyer. 1936. The ef- fects of the gene on g rowth and differentiation as shown by the temperature responses of pennant and i t'I\ heterozyg ote in D. mela.nogaster. J. Exper. Zool. 74:41-59. 28. Hersh, A.H. 1924. The effect of temper a ture upon hetero- zygotes in the Bar series of Droso phila. J. Exper. Zool. 3 9:55-71. ~9. Hersh, A. H. 1927. Tempera ture effects in reciprocal crosses of the Bar series in Drosophila. J. Exper. Zool. 47: 2 27- 2 50. 30. He rsh, A.H. 19 28. Organic correlation and its modifi- c a tion in the Bar series of Drosophila. J. Exper. Zool. 50:~39- 2 55. 31. Hersh, A. H. 1930. The facet-temper a ture rel a tion in the Ba r series of Drosophila. J. Exper. Zool. 57:283-306. - 153 32 . Hersh, A. H. The ti me c urve of f ac e t di ,,, ermin a tion 1934 . in an ultrabar stock of Droso phila melanogaster. J. Gen. Physiol. 33. 17: 487-4 98. Hersh, A. H., a nd Ward, Esther. 1 93 2. The ef f ect of temp- erature on win g size in reci proc a l heterozyg otes of vesti g ial in Droso phil a melanogaster. J. Exper. Zool. 61: 22 3- 2 44. 34. Hersh, Ros a lie Karrer. 19 24. The effect of temperature upon the full-eyed race of Drosophil a . J. Ex per. Zoo 1. 39: 43-53. 35. Hog e, lh ldred A. 1915. The influence of temperature on the develo pment of a Mend elian cha racter. J. Exper. Zool. 18: 2 41- 2 97. 36. Ives, P. T. 1935. The temp era ture-effective period of the scute-1 phenotype. ? roe. Nat. Acad. Sci. Washing ton 2 1:6 46-650. 37. Krafka, J. 1920. The effect of tempe r a ture upon facet number in the Bar-eye d mutant of Droso phila. J. Gen. Physiol. 2 : 4 09-432. 38. Krafka, J. 19 2 4. Develo pment of the comp ound eye of Drosophil a melano g aster and its Bar-eyed mutant. Biol. Bull. 47:14 3 -14 8 . 39. Lashin, Be a trice J. 19 38 . Effect of Te mp erat u re on the wi ng s of di morphos ves t i g i a l/p ennant in Droso phila me lano gaster. ( Abstr a ct). Genetics 23 : 155. -15440. Li, J. C., and Tsui, Y. L. 1 93 6. The development of vestigial wings under hi gh temper a ture in Drosophila melanogaster. Genetics 2 1:248- 2 63. 41. Luce, w. M. 1926. The effect of tempEr&.ture on infra- Bar, an allelomorph of Bar-eye, in Droso phila. J. Exper. Zool. 46:301-316. 42. Luce, W. M. 1931. The temperature-effective period in infra-Ba r and its heterozygotes. J. Exper. Zool. 59:467-498. 43. Mar g olis, o. s. 1935a. Drosophila I. Studies on the Bar series of The effect of the gene vestigial on facet number in Bar. 44. Margolis, ~. s. 1935b. Drosophila II. Genetics 20:156-171. Studies on the Bar series of The temperature-effective p eriod for facet determination in Bar and Bar-vestigial. Genetics 20:207-222. 45. Margolis, o. s. 1936. Drosophila III. Studies on the Bar series of The facet relations in dorsal and ventral lobes of the eye. 46. Margolis, O. Genetics 21:817-831. s., and Robertson, c. w. the Bar series in Droso phila IV. 1931/ Studies on The temper a ture- effective periods for facet determination in the wild type. Benetics 22:319-331. - 155 47. Metz, C. W. 1 92 3. A n ote on t h e effects of temperature in the mutant character 11 bent 11 in Drosophila virilis a nd Drosophil a. melanog a ster. Med. 48. Proc. Soc. Ex p er. Biol. 20: 305- 310. Mohr, 9). L. 1932. On the p otency of mutant g enes and wild-type a llelomor phs. Proc. S ixth. Intern at. Cong. Genetics 1:190- 2 12 . 4 9. Nadler, J. E . 19 2 6. Effects of temper a ture on len g th of vesti g i a l wing of Drosophila virilis. Genetics 11:584-589. 50. Payne, F . 1920. Se l ection for hi gh a nd low b ristle num- ber in the mutant strain, 11 reduced 11 • Genetics 5:501-54 2 . 51. Plou gh, H. H., a nd Ive s, f. T. 1935. In duction of mu- t a tions by hi gh temp erature in Droso phil a . Genetics 2 0:4 2 -69. 52 . Plunkett , C.R. 19 26. Intera ction of genetic a nd environ- ment a l f a ctors in development. J. Ex per. Zool. 46:181- 244. 53. Plunkett, c. R. 1 93 2 . Temp er a ture a s a tool of research in pheno g enetics: methods and results. ( Abstr a ct) . Proc. S i x th Intern a t. Cong . Genetics 2: 158. 54. Poulson , D. F . 1937. The emb r y on ic developme n t of Droso phil a mel a no gas ter. e t I ndu strielles 4 98 :1-51. Actu a litj d s Scientifi ques -15655. Powsner, L. 1935. The effects of temperature on the duration of the developmental stages of D. melanogaster. 56. Physiol. Riedel, Hilda. 1934. Zool. 8:474-520. Der Einfluss der Entwicklungstemper- 11 ,, atur auf Flugel- und Tibialange von Drosophila melanogaster (wild, vestigial, und die reziproken Kreuzungen). Willl.h. Roux' Arch. F. Entw. Mech. Organ. 132:463-503. 57. Roberts, E. 1918. 1nuctuations in a recessive Mendelian character and selection. 58. Robertson, c. W. 1936. J. Exper. Zool. 27:157-192. The metamorphosis of Drosophila melanogaster including an accurately timed account of the principal morphological changes. J. Morph. 59:351-399. 59. Seyster, E. w. 1919. Eye facet number as influenced by temperature in the Bar-eyed mutant of Drosophila melanogaster (Ampelophila). 60. Stanley, w. F. 1928. Biol. Bull. 37:168-182. The temperature coefficient and temperature-effective period for wing size in Drosophila. 6l.. Stanley, W. :b'. (Abstract). Anat. Rec. 41:114. 1931. The effect of temperature on ves- tigial wing in Drosophila melanogaster with temperature-effective periods. .?hysiol. Zool. 3:395-408. - 1 5762 . Stanley, w. F . 1935 . The eff ect of t empe r a ture up on wi ng si z e in Dro s ophi l a . 63 . J. Expe r . Zool. 69: 459- 4 95 . St urteva nt, A. H., an d Schultz, J. 19 31. The inade quacy of the sub- g ene h yp othesis of the n a ture of the scute a llelomor phs of Dr os ophil a . Wa s h i ng ton 17: ~65- 270. Proc. Nat. Ac ad . Sci.