I Copies of this publication may be procured from the GENETICS FOUNDATION, DEPARTMENT OF ZOOLOGY The University of Texas, Austin, Texas 78712 PROFESSOR OF ZOOLOGY • studies zn GENETICS III. Morgan Centennial Issue THOMAS HuNT MORGAN (1866-1945) pioneered in the discovery of the unique features of Drosophila flies for genetic research. The multi­plicity of problems which are being studied today with the use of Drosophila is well illustrated by the articles in this edition of Studies in Genetics. In this centennial year we feel, therefore, that it is most appropriate to dedicate this issue to his memory. THE UNIVERSITY OF TEXAS AUSTIN The benefitsof education and of useful knowledge, generally diffused through a community, are essential to the preser­vation of a free government. SAM HOUSTON Cultivated mind is the guardian genius of Democracy, and while guided and controlled by virtue, the noblest attribute of man. It is the only dictator that freemen acknowledge, and the only security which freemen desire. MIRABEAU B. LAMAR PUBLISHED BY THE UNIVERSITY TWICE A MONTH. SECOND CLASS POSTAGE PAID AT AUSTIN, TEXAS. Contents I. Genetic Studies of Natural Populations of Drosophila. II. Pacific Island Populations . WILSON S. STONE, MARSHALL R. WHEELER, FLORENCE D. WILSON, VIRGINIA L. GERSTENBERG AND HEI YANG 1 II. The Nearctic and Neotropical species of Scaptomyza Hardy (Diptera; Drosophilidae) . MARSHALL R. wHEELER AND HARUO T AKADA '37 III. A study of speciation in South Pacific populations of Dro­sophila ananassae . DAVID G . FUTCH 79 IV. The mode of migration of Drosophila ananassae under com­petitive conditions TAKASHI NARISE 121 V. Mating behavior of D. ananassae and ananassae-Iike flies from the Pacific . HERMAN T. SPIETH 133 VI. The influence of light on the mating behavior of Drosophila JOSEPH GROSSFIELD 147 VII. Modification of induced genetic damage in Drosophila melano­gaster by oxygen and argon treatments between two doses of X-rays FLORA T. ELEQUIN 177 VIII. Descriptions and notes on Hawaiian Drosophilidae (Diptera) D. ELMO HARDY 195 IX. Courtship behavior of endemic Hawaiian Drosophila. HERMAN T. SPIETH 245 X. Male genitalia of some Hawaiian Drosophilidae HARUO TAKADA 315 XI. The relationships of the endemic Hawaiian Drosophilidae LYNN H. THROCKMORTON 335 XII. Preliminary report on the karyotypes of Hawaiian Drosophilidae . FRANCESE. CLAYTON 397 The University of Texas Publication XIII. Chromosomal races of Drosophila crucigera from the islands of Oahu and Kauai, State of Hawaii . HAMPTON L. CARSON 405 XIV. Cytological studies on some species of the tripunctata group of Drosophila . CosTAS D. KAsTRITSis 413 XV. A study of the royal jelly gland cells of the honey bee as revealed by electron microscopy . THEOPHILUS S. PAINTER AND JoHN J . BrnsELE 475 XVI. Sulfoxide protection of bacteriophage against X-irradiation inactivation LLOYD S. LocKINGEN 501 XVII. Dimethyl sulfoxide treatment of Drosophila. MARY L. ALEXANDER 505 XVIII. An operational classification of Drosophila esterases for species comparisons F. M. JOHNSON, CARMEN G. KANAPI, R. H. RICHARDSON, M. R. WHEELER, AND W. S. STONE 517 XIX. Notes on the Drosophilidae (Diptera) of Samoa MARSHALL R. WHEELER AND MICHAEL P. KAMBYSELLIS 533 ACKNOWLEDGMENTS The research articles presented here are contributions from the faculty, staff, pre­and postdoctoral students and investigators connected with the research programs of the Genetics Foundation, Department of Zoology, The University of Texas, or the program on endemic Drosophilidae of Hawaii under the general supervision of Pro­fessor D. Elmo Hardy, Chairman, Department of Entomology, The University of Hawaii. Professors Stone and Wheeler and their colleagues in the Genetics Foundation have long been interested in island populations, both in the Caribbean and the Pacific areas. We have been especially interested in the Hawaiian Drosophilidae since Dr. Elwood Zimmerman began publishing on the insect fauna of Hawaii (Zimmer­man, E. C., 1948, Insects of Hawaii. I. Introduction. University of Hawaii Press, 206 pp.). Professor Hardy has, for many years, been studying the taxonomy of the Hawaiian Drosophilidae and his first major publication is now available (Hardy, D. E., 1965, Insects of Hawaii, Vol. 12, Drosophilidae. University of Hawaii Press, 814pp.). A second article, describing additional Hawaiian species, is included in This Bulletin. After several visits and discussions with Professor Hardy, we developed a program of research on the evolution of these most interesting endemic forms involving, in addition to Professor Hardy and his colleagues at the University of Hawaii and Professors Stone and Wheeler and their associates at the University of Texas, a num­ber of collaborators from other Universities interested in this area of research. The other senior investigators who are engaged in this program are Professor Herman Spieth, Chairman, Department of Zoology, University of California, Davis; Pro­fessors Hampton Carson and Harrison Stalker, Department of Zoology, Washington University, St. Louis, Missouri; Professor Forbes Robertson, Institute of Animal Genetics, Edinburgh, Scotland; Professor Frances Clayton, Department of Zoology, University of Arkansas, Fayetteville; Professor William Heed, Department of Zoology, University of Arizona, Tucson; Professor Lynn Throckmorton, Department of Zoology, University of Chicago; and Dr. Malcolm Brown, Botany Department, University of Texas, Austin. The project has also involved a number of postdoctorals, graduate students, and technical assistants in the several laboratories. We wish to thank Professor Newton Morton, Department of Genetics, Pacific Science Center, University of Hawaii, for the use of space and equipment to begin the project. We are deeply indebted to Professor Hardy and his colleagues in the Department of Entomology who obtained additional space and equipment for the research from the University of Hawaii, where the project is now well housed and cared for. We also wish to thank the Administration of the University of Hawaii for approving the joint development of this research effort and helping us get this large, long-term investigation underway. At Austin, the basic genetic laboratories have been provided and supported by the University of Texas for many years, with added research support from the Rockefeller Foundation. The Rockefeller Foundation support is now ended and we take this opportunity to express our lasting appreciation for the early and extensive support which made a strong development in genetics possible at the University of Texas. Our major research support is currently provided by several federal agencies. Pro­fessor Hardy, through the University of Hawaii, has a grant from the National Insti­tutes of Health, Public Health Service, GM 10640, which supports several local and guest investigators and their research on Hawaiian Drosophilidae. The Robert A. Welch Foundation has supported some of our work. The National Institutes of Health, Public Health Service, have supported the major part of the work at the University of Texas under Professors Stone and Wheeler with grant RG 6942, followed by GM 11609. The National Science Foundation supported the Hawaiian project with GB 711 to Professors Stone and Wheeler, and by GF 152 to Professors Wheeler and Throck­morton. The latter grant was part of the U. S.-Japan Cooperative Program in Science, and permitted an investigation of the possible relationships between endemic Hawaiian and endemic Japanese Drosophilidae. Work in Japan on this project has been most pleasant due to the extensive help and cooperation of Drs. Moriwaki, Okada, and Kurokawa in Tokyo, Professors Makino and Momma in Sapporo, and Professor Takada in Kushiro. The United States Atomic Energy Commission, through contract AT-(40-1)-2952, supported Professor Stone's studies on radiation and Pacific island populations through December, 1965, including the atomic and thermonuclear tests on Christmas Island. Professors Stone and Wheeler and their colleagues are greatly indebted to a number of people who have helped us in our studies on various Pacific Islands. Prior and after the Christmas Island tests, the United States Atomic Energy Commission field personnel and the Holmes and Narver personnel supported our field work most effectively. Dr. Harold J. Coolidge, Executive Director of the Pacific Science Board, was very helpful in making arrangements for us to work on a number of Pacific islands. Dr. B. A. O'Connor, then Senior Entomologist, Department of Agriculture, Koronivia, Fiji Islands, helped us on several occasions as have others at this station. The New Zealand Department of Island Territories helped to provide quarters and laboratory space for our work on Aitutaki, Rarotonga, and Niue; additional help was provided locally by Resident Commissioners, for example, Mr. L. A. Shanks on several occasions on Niue Island. We are also grateful to Dr. Harry Nemaia who allowed us to collect Drosophila on his property on Niue, and aided us in other ways. Mr. and Mrs. Alan Berry's kind hospitality on Niue will always be remembered. Her Majesty's government on Tonga supported us in our collecting in the Tonga Island group through the Department of Agriculture. In addition they provided the small ship, Hifofua, for special transportation that made our collections on Niue possible in 1963. Our friend, Sir Ernest Marsden, brought to our attention the high level of natural radioactivity on Niue and has aided in many ways our studies of this interesting island population. Governor H. Rex Lee, Acting Governor Owen S. Aspinall and assistant to the governor, Mr. James Flannery, were most helpful to us on American Samoa, as were members of the Agriculture Department. Chief Le'iato and many other Samoans were most hospitable to us. The Government of Western Samoa, especially Mr. Mike Watt and other persons in the Agriculture Department, helped us with our work. Dr. Dexter Hinckley, Ecologist with the UN-SPC Rhinoceros Beetle Project at Apia, is a friend of long standing who helped us on Upolu in many ways. Pastor Vaialia and his family's hospitality at Aopo, Savaii, is but an example of the kindness and hospitality of many of the Samoan and other Pacific island peoples. MARSHALL R. WHEELER, Editor WILSON s. STONE, Director, The Genetics Foundation I. Genetic Studies of Natural Populations of Drosophila. II. Pacific Island Populations1 WILSON S. STONE, MARSHALL R. WHEELER, FLORENCE D. WILSON, VIRGINIA L. GERSTENBERG AND HEI YANG We have studied certain genetic attributes of several Drosophila populations of various sizes, degrees of isolation and environmental stress, including fallout radiations. The species studied included Drosophila ananassae (light form and dark form) , Drosophila melanogaster, Drosophila simulans, and Drosophila nasuta.2 We reported earlier on Mainland populations of Drosophila pseudo­obscura (e.g., Stone, Wilson and Gerstenberg, 1963) and on a series of studies of ananassae populations collected in the North Pacific from several of the Marshall Islands and from Ponape in the eastern Caroline Islands (Stone, Wheeler and Wilson, 1962; see also other references there). The collections in 1962 were made in connection with the U.S. Atomic Energy Commission test series near Christmas Island. In March, before the tests began, we attempted to collect Drosophila on the island but the island proved to be too dry and barren to support a Drosophiw population. We, also, made collections on Palmyra Island, north of· Christmas Island, and on Tutuila Island, American Samoa-the latter to obtain a control population. In July and August, after the test series, we made collections on Tutuila, Rarotonga (Cook Islands) and Pal­myra, none of which received an appreciable amount of fallout from the tests. During July and August, 1963 and/ or 1964, we collected on several islands in the South Pacific including Tongatapu, Viti Levu, Niue, Tutuila, Rarotonga and Aitutaki. We, also, obtained samples of Drosophila populations from Tutuila, American Samoa, and from Upolu and Savaii, Western Samoa, in July and August, 1965, but the analyses of the last samples are incomplete. Table 1 lists the islands where collections were made and gives some information about the various populations which were studied. We are studying several important components of natural selection and evolu­tion, especially viability (measured as the percent of eggs that develop to adults) , fertility (percent of pairs fertile), and fecundity (number of eggs laid per day by young fertile females). In addition, we are studying the effect of the physical separation inherent in these widely scattered island populations. In some cases populations on certain islands have been subjected to radioactivity. The first group in the Marshall Islands, Bikini (11° 37' N, 165° 33' E), Rongelap (11° 6' 1 This investigation was supported, in part, by Public Health Service Research Grants No. GM-06492 and GM-11609 from the National Institutes of Health, and by contracts AT-(4-0-1 )­1323 and AT-(40-1)-2952 from the U.S. Atomic Energy Commission. 2 Editor's note: Studies in progress indicate that several species may be involved in the name "nasuta"; the form characteristic of Hawaii and Samoa is apparently different from that from Ponape (Caroline Islands) which is in turn different from that from Okinawa (Ryuku Islands). On the other hand, the species setifemur, from Australia, seems to be phenotypically the same as the form of nasuta from Samoa. The University of Texas Publication TABLE 1 Islands where collections were made, 1962--1965, and Drosophila species tested Latitude and Type of When Drosophila Population Island longitude island collected species tested size estimate Christmas, 1°50'N, Dry atoll March, 1962 none Drosophila Line 157°20'W absent Islands Palmyra, 5°52'N Atoll March, 1962 ananassae small Line 162°04'W July, Aug., 1962 nasuta Islands Tutuila, 14°18'S, High March, 1962 ananassae medium to American 170°42'w (wet) July, Aug., 1962 nasuta large, but Samoa Island July, Aug., 1964 melanogaster melanogaster- July, Aug., 1965 small Rarotonga, 21°13'S, High July, Aug., 1962 ananassae ananassae- Cook 159°47'W Island July, Aug., 1964 nasuta medium, but Islands simulans nasuta and simulans-small Viti Levu, 17°40'S, Large July, Aug., 1963 ananassae medium Fiji 178°00'E high island, old group Tongatapu, 21 °10'S, Atoll July, Aug., 1963 ananassae small to Tonga 175°10'W nasuta medium Islands Aitutaki, 18°52'S, Atoll July, Aug., 1964 ananassae small to Cook 159°47'W medium Islands Niue 19°00'S, Raised July, Aug., 1963 ananassae small to Island 169°52'W coral atoll July, Aug., 1964 nasuta medium Upolu, 13°55'S, High July, Aug., 1965 ananassae medium Western 171°45'W island nasuta Samoa Savaii, 13°40'S, Large July, Aug., 1965 ananassae medium Western 172°30'W high nasuta Samoa island N, 166° 58' E) and Rongerik (11° 25' N, 167° 27' E), suffered from the fallout (and some direct radiation on Bikini) associated with testing of atomic and ther­monuclear devices in the U. S. Atomic Energy Commission Pacific Proving Ground. Majuro (7° 20' N, 171 ° 10' E), in the southern Marshalls, and Ponape (6° 47' N, 158° 14' E), in the Eastern Caroline Islands, served as controls, having little fallout radiations. The population on Niue is subject to the radiations from Radium and its decay products (Marsden, 1964; Marsden et al., 1958; Fieldes et al., 1960) which occur on Niue at a higher level than is found for any other known environments of Drosophila. Although it was rumored that there were areas of heavy radiation on Aitutaki and Rarotonga, none were found in survey­ing these islands with radiation counters, even in the local areas where the radia­ tion was rumored to occur. The material from the Pacific Proving Ground has been reported earlier, but Stone et al.: Pacific Populations of Drosophila the studies of these additional non-irradiated populations from various kinds of islands (high and large, flat and large, or atoll) provide added comparisons of population differences and further support our conclusions on the serious genetic damage to Drosophila from direct radiations and fallout, especially on Bikini. MATERIALS AND METHODS In 1962, samples of the populations were collected for a week each on Palmyra and Tutuila (but only two hours on Rarotonga) and were shipped or carried in fresh food in a cool picnic-type ice-box back to the laboratory in Austin. There were only a few hundred flies from Rarotonga but several thousand were col­lected on both Palmyra and Tutuila. In the laboratory we isolated a number of individual females (Po) that had been fertilized by males of the same population on the island or in the food vials in transit. The offspring (designated P1 ) from these wild females were tested in pairs (all subsequent tests are pairs unless stated otherwise) in the following ways: ( 1) inbreeding brother X sister (sib matings), (2) random matings be­tween males and females from different P0 wild females (random matings), (3) crossing males with females from different islands (cross matings), (4) mating male sibs to unrelated females to produce cousins. The next generation ( 5) the cousins were mated (cousin matings) , and ( 6) the F1 flies from the P 1 sib matings and from the P1 random matings were again sib mated or crossed at random (off­spring from different pairs of sib matings were crossed-these are designated sibs from sibs, random from sibs, random from random, and sibs from random). In some years other crosses were made as designated. These crosses allowed us to test the effect of different degrees of inbreeding or past inbreeding on three im­portant components of Darwinian fitness: viability (development), fertility, and fecundity. Fecundity was measured as the number of eggs laid per day over a four to six day period by young mature fertilized females. To insure that the females were fertilized, eggs were included in the test counts only after eggs laid the previous day hatched. Fertility was measured as the number of pairs fertile. These may not coincide with the number of pairs tested for egg development since some pairs proved fertile although they did not produce fertile eggs until after the test period. Viability (development) was measured as the percent of eggs that developed to adults from those laid during the four to six day laying period by the young mature fertilized females. The populations of ananassae and nasuta (the name spinofemora was used formerly for the Hawaiian form ) on Palmyra, a flat atoll island, were small. They live in the few coconut groves in fallen coconuts which have been chewed open by the many crabs and rats. Wheeler estimated the population of each species to be less than 5000, judging from the reduction in numbers we could capture in new runs after we had netted one to two thousand. The populations of these two species on Tutuila are very much larger, but that of melanogaster is small and appears to be restricted to the towns. Tutuila is a large high island with over 200' inches of rainfall per year and has a lush rain forest growth over much of it. Rarotonga is, also, a high island, although smaller and a little drier than Tutuila. We collected only a few hundred flies in a two-hour stopover in 1962, consisting The University of Texas Publication for the most part of ananassae and simulans with a few nasuta. We could not estimate the size of these populations from that collection although ananassae must be medium in size on such an island. Wheeler's collections in 1964 sub­stantiate this opinion. Both simulans and nasuta are small. The light and dark ananassae from Tutuila, which are in fact separate species, form very interesting populations. Although they were collected together at most localities, the isolated P0 wild females usually bred true, that is, if dark they pro­duced dark offspring and if light only light offspring. The number of P0 females collected and tested which produced only light or only dark offspring is perhaps ten times greater than the number listed in the Tables. Because of this marked sexual isolation in natural populations, a number of further tests were made. These are reported in the Tables and by Futch (this Bulletin). A number of crosses are included since some pair matings between the forms are successful when the flies are given no choice. RESULTS The results of our tests of Drosophila collected from March, 1962, through July-August, 1964, are given in a series of tables and figures. For convenience the detailed tables are placed together at the end of the paper and only summaries and pictographs are used in the text. Several different species were tested: ananassae from each island and each collection, nasuta from Palmyra and Tu­ tuila, melanogaster from Tutuila, and simulans from Rarotonga. The average number of eggs laid per day by young, mature fertilized females is a species specific character. It is related to the ecological adaptation rather than size-for example, some of the very large Hawaiian Drosophila lay only a few eggs per day under laboratory conditions. Mature fertilized Drosophila hydei may lay no eggs or up to nearly 300 in one day, depending on conditions and history. There is no significant variation between stocks from different islands, or from year to year, see Tables 2-5 and summary Table 6. The range for ananassae is usually 20 to 30 eggs per day although a few lower values occur. The average values of eggs laid per day are: melanogaster, 25.5; simulans, 19.9; nasuta, 20.4 and 24.2 for the stocks collected two different years. D. nasuta is an appreciably larger fly than the other species but does not lay more eggs per day. The results with ananassae collected in July and August, 1963, are interesting. That year the average number of eggs laid per day by the stocks was much above the average of crosses between stocks. The low fecundity crosses involve Niue. All tests within island strains (Table 4, 1-36) are above 19.0. If we take 19.0 as the dividing line, we find that all crosses (numbers 37-68) giving a frequency of eggs per day below this involve Niue. Furthermore, all of them are recombination tests and involve a Niue hybrid as one or both parents, except two Niue crosses giving 18.8 and 18.6 eggs per day (numbers 39 and 44). Not all Niue hybrids gave such low fecundity, see Table 4 for details. This does not correlate with de­ velopment, Tables 2-5 and 8, nor with fertility, Tables 2-5 and 7. In that year the fertility increased each generation in stocks from Niue, Suva and Tonga, Table 7, and also in the crosses. Perhaps food conditions in the laboratory changed or the stocks (and crosses) were all suffering over several generations from a Stone et al.: Pacific Populations of Drosophila microbial infection or adverse environment. However, fertility does not show the difference between stocks and crosses that fecundity does that year. No other year showed such a trend in fertility. Summary Table 7 shows that fertility varied widely between species and be­tween different island populations of the same species as well as from year to year. The light ananassae from Tutuila is undoubtedly a different species from the other ananassae, Tables 2-7, as is evident in the article by Futch (this Bul­letin) . It is strongly sexually isolated from dark ananassae from Tutuila. They coexist and are collected together but retain their identity with consistent cyto­logical differences. Even in pair matings without choice, they often fail to cross­mate. This does not extend to their offspring which inbreed and backcross or can engage in three-way crosses (three different island stocks), see Tables. The major isolation is the sexual isolation between species, see Tables for fertility and de­velopment from F1 and F" hybrids. This initial isolation contributes to a reduced fertility and egg development in the initial cross test. Because of fluctuations, not often attributable to bad culture conditions or such variables, the average number of eggs laid per day and especially average fertilities represent a considerable range of values. The values for eggs per day were more consistent. Of the 214 tests of all species only 8 averaged more than 30 eggs per day and only about 64 showed less than 20 eggs per day. The values of egg development varied much more consistently and directly with thedegree of inbreeding. The fluctuation of the fertility valued due to undetermined causes makes com­parisons between pairs of crosses difficult to evaluate. We chose to compare the effect of having the parents unrelated or choosing sibs as parents. For ananassae, using only F1 and F2 tests, and omitting the results from Rarotonga, 1962, for rea­sons discussed in the next section, the values are: random from random (20 tests) = 61.5% of pairs fertile; sibs from random (11 tests) = 59.9% fertile; random from sibs (10 tests) = 43.3% fertile. It is clear that if the parents of the flies tested were sibs, factors became homozygous which reduced the fertility measured as pairs fertile. On the other hand, if the parents were unrelated, sib matings were about as fertile as random matings. Thus, we see that inbreeding reduces the fer­tility of the offspring, which would be expected if the flies carried an appreciable number of recessive sterility factors. The results on viability, or rather developmental viability from egg to adult, are much more consistent and comparable from test to test and year to year. Fur­thermore, the percent development relates directly to the degree of inbreeding and to the recent history of inbreeding, that is, there are appreciable paternal and/or maternal effects from inbreeding. The data are given in Tables 2-5, sum­mary Table 8, and Figures 1 and 2. Table 8 compares the reduction in percent development for the different inbreeding histories, combining similar tests in each year's run. The pedigrees in Figures 1 and 2 show typical runs. In these figures the wild female, P0 , is assumed to have bred at random with a male (or males) in the same population so that P1 flies are all from random matings. In P1, F1 and F2, we have random and other matings from random (including brothers x unrelated females to obtain cousins for mating) and matings in which The University of Texas Publication Tutuila light wild 9 random~X uLelat~sibs 86.6 62.5 /81.9~ I random cousins sibs random 89.4 78.9 68.0 74.1 I I random random 86.3 85.5 Tutuila dark wild 9 random ~rsX }nrelat~ sibs 89.4 88.2 74.9 . /I~ I random cousins sibs random 84.5 80.1 63.0 86.2 I I random random 76.4 90.2 Palmyra wild 9 random~rsX u!rela~sibs 86.1 /81-9~ 718 random cousins sibs random 87.1 76.0 70.5 81.7 I I F random random 76.5 87.1 FIG. 1. Pedigrees showing development (%) selected from Drosophila an.anassae collected on Tutuila and Palmyra, July, August, 1962. Stone et al.: Pacific Populations of Drosophila Po Niue wild 9 pl random~X uLela~sibs 89.3 70.2 /8,.1~ F1 random sibs random cousins sibs random sibs / ""' /""' 91. l 67.8 86.2 88.1 73.3 82.3 64.9 /~ /~ /~ F2 random sibs random sibs random sibs 91.0 65.7 90.9 73.4 76.5 60.9 Tonga wild 9 pl random ~sX uLelated~sibs/3Z /81"5~ ;oz random sibs random cousins sibs random sibs 870 70.0 84.1 83.6 69.6 87.l 61.4 /~ G random sibs 89.l 72.7 Suva wild 9 random~ ---------sibs 90.4 633 /~ I ""-. F; random s1bs random sibs 88.8 67.7 82.l 62.7 /~ F; random si bs 85.0 66.1 Fm. 2. Pedigrees showing development ( % ) selected from Drosophila ananassae collected on Niue, Tonga and Suva (Fiji), July, August, 1963. The University of Texas Publication the parents or grandparents were sibs or cousins. Ran,dom from sibs are matings between the offspring of different unrelated sib matings. We obtained only a few flies (only two hours collecting) from Rarotonga in 1962, Table 3. The simulans which were collected there seemed normal but the ananassae were either injured in shipping to the laboratory, poisoned or possibly infected by some microorganism since the egg development was very low in the P1 which were mated at random. The F1 which were sib mated, also, gave low development when compared to other similar tests. The F2 were mated at random (random from sibs ) and gave a reasonable but still rather low egg development. The 1964 collection at Rarotonga was tested more thoroughly and gave egg de­velopment in the normal range, Table 5. Table 8 shows these results clearly. Omitting the data from the Rarotonga ananassae collected in 1962, the average development from egg to adult in ran­dom from random matings is 86.9%, column 1, with a range from 81.6 to 93.9%; Rarotonga, 1962, gave only 65.9% egg development. The average for random from sibs, column 2, was 81.3% egg development for ananassae with a range from 74.1to87.1 %. The value for Rarotonga of 79.7% is reasonable compared to other average values in this type of test. The average values for cousins is 79.9%, column 3, with a range from 72.1 to 88.0%. The matings sibs from random aver­aged 67.7%, column 4, (omitting Rarotonga, 1962) with a range from 61.3 to 74.9%. The value of 56.4% for Rarotonga in 1962 is below the range of all other such matings. The average from sibs from sibs was 63.0%, column 5, with a range from 61.4 to 64.9% development. These values, as well as the average values, are less than most values and average values for sib from ran,dom matings. We made one mating of sibs from sibs from sibs, cross 34, Table 4, which gave 60.9% egg development, lower than all other sib mating values. In contrast, ran­dom from random from sibs with egg development values (matings number 8, 17, 26; Table 3) of 85.5, 90.2, and 87.1 %, with an average of 87.6%, are similar to the usual random from random values, see Figure 1. We can make a general statement about the effect of inbreeding, or of inbreed­ ing of parents even when the eggs and progeny counted are themselves hetero­ zygous from :r:andom matings. Inbreeding clearly reduced egg development, the greater the degree of inbreeding, the greater the reduction in egg development. If parents are the result of inbreeding of sibs, fewer eggs develop even though the offspring are themselves heterozygous, showing that there is a maternal and/ or paternal effect which reduces egg development in these cases. There is, also, a cumulative inbreeding depression, probably due in part to the same causes, in tests of sibs from sibs. In fact, if we compare the several tests from the same stock shown in Table 8, or the averages, we find the following order in egg development: ran,dom from random = random f ram random f ram sibs > random from sibs >cousins > sibs from random> sibs from sibs. We omit the data for Rarotonga, 1962, for reasons discussed above. Column 6, Table 8, shows the difference in percent of average egg development from random from random (1) and sibs from random (4). D. ananassae shows an average difference of 19.2% (omitting Rarotonga, 1962). D. nasuta, with a much smaller population on these same islands, has an overall Stone et al.: Pacific Populations of Drosophila difference of 10.4%; but in fact, the difference in 1962 collections from Palmyra Island are 5.8 and 6.5%, and the populations of both ananassae and especially nasuta are small on that island. D. melanogaster has a very small population on Tutuila and shows an inbreeding depression of only 8.1 %.* Figures 1 and 2 show that similar tests over several generations are quite con­sistent, and in addition show the effect of inbreeding depression on egg develop­ment. In addition to the intrapopulation tests, a number of interpopulation crosses were made between the same or related species from different islands. D. nasuta from Palmyra and Tutuila collected in 1962 were crossed. They were highly cross fertile. Although the percent of pairs fertile in the cross P <;> X T ~ in the March collection was especially low in one cross (Table 2; tests 17, 19; crosses 22, 23), both reciprocal crosses were highly fertile from the July-August collec­tion (Table 3, compare tests 34, 35, 40, 41, 42, 47 with crosses 48 and 49). The egg development was as high as that from random matings within a population. There was no evidence of reproductive isolation between these two island popula­tions. Ft from crosses are, also, quite fertile. D. ananassae dark and ananassae light from Tutuila are undoubtedly different species with consistent cytological differences in the major chromosomes (see Futch, this Bulletin). Although they occur together, they rarely crossmate in nature, i.e., most isolated wild females breed true. They are sexually isolated in nature although they will mate if given no choice but the other species in the laboratory environment. Spieth (this Bulletin) and Futch (this Bulletin) have studied the behavior, isolation and cytological differences between a number of strains of the ananassae species subgroup of Drosophila. The ananassae strains from the other islands reported here belong to the same species as "dark" ananas­sae from Tutuila. The crosses reported here as well as Spieth's and Futch's studies all show that genetic differentiation is occurring in these isolated populations. The collections made in 1962 were made in March on Tutuila and Palmyra and from Tutuila, Palmyra and Rarotonga in July and August. The March, 1962, stocks of the sibling species, ananassae Tutuila light (test 1 of Table 2) and Tutuila dark (test 3) and the strain from Palmyra (tests 5 and 6) were inter­crossed (crosses 9-14) . The percent fertility of the stocks from Tutuila was low although Palmyra was better. Crosses 9 and 10 show differences in fertility in reciprocal crosses. The Palmyra stock was better but the crosses of Palmyra fe­males to either Tutuila light males or Tutuila dark males were only rarely fertile (crosses 12 and 14). Reciprocal crosses went well. However, only the reciprocal crosses between Tutuila dark and Tutuila light gave low egg development (crosses 9 and 10). The tests of the Drosophila strains collected in July and August, 1962, are * Dr. Archie Allen, now at Texas Technological University, Lubbock, checked the lethals on chromosomes 2 and 3 in this population of melanogaster. He found (personal communication) 10 lethals on chromosome 2 (10.8%) and 30 lethals on chromosome 3 (32.3 %) from the March, 1962, collection, with 4 of the lethals on chromosome 3 being allelic in 93 flies tested. He, also, tested 20 flies from the July, August, 1962, collection and found 2 lethals in chromosome 2 (10.1'%) and 7 lethals (28.6%) in chromosome 3, none of them alleles. However, he found 4 lethals common to the March and the July, August collections in chromosome 3. The University of Texas Publication much more extensive, Table 3. The pertinent intrapopulation tests of ananassae include: Tutuila light, tests 1, 2, 7, 8; Tutuila dark, tests 10, 11, 16, 17; Palmyra, tests 19, 20, 25, 26; and Rarotonga, tests 28, 29, 30. We haye discussed the fact that some undetermined environmental variable such as a poison or infection caused the Rarotonga egg development values in tests to be low. D. simulans collected at the same time had a normal range of egg development in these tests. It may be that some other tests were low because of some such environmental variable. The con­sistency of results in the normal range in most cases made this improbable. All of the random intrapopulation crosses listed above gave good egg development, Rarotonga excepted. The fertility of pairs was slightly better but still very poor for Tutuila light. The others were relatively fertile, ordinarily ranging from 40 to 60%. DISCUSSION The initial collections for these studies were centered around Christmas Island and the series of atomic and thermonuclear devices tested there in 1962. Although Christmas Island might have received limited fallout, there was no Drosophila population living there. Palmyra and Tutuila received very little fallout from the tests. Their populations showed the usual relation between population size and distribution of viability classes on inbreeding. In fact, D. nasuta on Palmyra with a fairly small population, a few thousand at most at its peak, showed less inbreeding depressions on egg development than most populations so far tested. Marsden et al. (1958, 1965) and Fieldes et al. (1960) reported that some soils on the island of Niue had an unusually high radiation due to radium and its decay products. H. J. Yeabsley and A. C. Stevenson provided a report to the Health Department on Niue dated 18 February 1963, "Data Relation to Population Gam­ma Exposure." They reported 6 to 20 micro-rads per hour on soils where villages are located close to the shore (e.g., a roadside on a hill of Alofi= 10 tc 20 micro­rads per hour) . The good soil in spots in the interior where gardening is often done ranged from 100 to 140 micro-rads per hour in one area, Mr. Harry Ne­maia's farm. They concluded, since only a small portion of the population spend as much as one working day per week in such areas, that the total mean increase of exposure to gamma radiation from background could not be double that pro­duced by the average world ·back ground radiation. Sir Ernest Marsden, using a counter that measured alpha particles, obtained readings up to 350 micro-rads per hour on Mr. Nemaia's farm; but our readings using a counter sensitive only to beta and gamma radiations ranged from 100 to 160 micro-rads per hour. The alpha particles are not serious except when ingested. The taro grown in this rich soil such as that on Mr. Nemaia's farm accumulates the radium of the soil but many fruits and vegetables do not accumulate radium to the same degree. In 1963 Futch (this Bulletin) found two pericentric inversions and a transloca­tion (in one larva) in ananassae collected on Mr. Nemaia's farm on Niue. He checked the chromosomes in four or more larvae from each of the 84 lines estn b­lished from single wild females in the lab at Austin. This is a surprisingly high number of chromosomal abnormalities even though one pericentrlc i:1version wns Stone et al.: Pacific Populations of Drosophila quite short and so subject to minor adverse selection due to crossingover within the inversion. The translocation seems to have occurred in one germ cell. In col­lecting on Niue (Mr. Nemaia's farm) in 1964, Futch (this Bulletin) isolated 82 ananassae females and established 30 control isofemale lines collected at Alofi, an area of little radiation. No chromosomal abnormality was found in checking the chromosomes of four or more larvae from each of the control flies or those from the area of natural radioactivity. Ingested radium or radon could produce alpha particles in the germ cells which could cause chromosomal abnormalities. The absence of chromosomal abnormalities in 1964 indicates this is rare, whether or not the pericentric inversions and translocation found in 1963 were of this origin. The short life cycle of Drosophila does not allow time for much gamma radiation even at a rate 10 to 20 times the average background. The level of detrimental factors in the Niue ananassae population does not differ from these other Pacific island populations as reported here (Table 8). This contrasts sharply with the radiation damage to ananassae found in populations from the Pacific Proving Ground, especially Bikini (see Figure 3b, c, d for comparisons) . These popula­tions, except the control from Ponape, ordinarily gave lower egg development than the new material presented here (Stone, Wheeler and Wilson, 1962). The number of lethal equivalents per gamete was not often higher in the Northern Marshalls..However, there were detrimentals present so that random matings did not give as good development as that ordinarily found in this material from Palmyra and the South Pacific islands. As these new tests give us little added information on the effect of either fallout or natural radioactivity, except lack of appreciable effect of low level natural radiation on Drosophila, our investigations become population and evolution studies. We can measure the effects of detrimental factors and some of the factors contributing to the isolation between populations or species by comparing intra­population tests with interpopulation crosses and further tests. Our results are from tests within the four collections. We checked crosses between collections from different years only to verify species and chromosome arrangements and to test to some extent allelism of genes leading to enzyme polymorphism (Johnson et al., July, 1966). The initial tests on enzyme polymorphism were done with the ananassae collected on Tutuila, Upolu and Savaii in July-August, 1965. We tried to test all this material within a few months of the time it was brought in to the laboratory to get the most information about the populations as they exist under natural selection. Futch (this Bulletin) and Spieth (this Bulletin) used crosses between many of the ananassae strains from different parts of the World to study the evolution of behavioral and other differences between the island populations. Our tests show no species differences other than the Tutuila light and dark ananassae, especially when the two Tutuila forms are the male parents. Futch and Spieth, also, found that the ananassae from the other islands in this study are the same species as the dark Tutuila form, see their papers for more informa­tion. The average number of eggs laid per day by stocks of the several species col­lected in 1962, 1963 and 1964 usually fell between 20 and 30, with few tests over 30 per day and more under 20. There is some difference between species, between island strains, and between years. None of the species characteristically laid over 30 eggs per day like virilis or hydei. D. nasuta is a larger fly than ananassae, melanogaster or simulans; but these island populations do not differ markedly in egg laying which is, undoubtedly, adjusted by selection. The only serious difference between stocks of dark ananas­sae from Tutuila and other islands and crosses was found in 1963, Tables 4 and 6. The low crosses involved Niue, except in one case (Table 4, number 62). Two crosses (Table 4, numbers 39 and 44) involving Niue were only slightly below some of the stocks. All fourteen other low crosses (Table 4, numbers 43, 45, 46, 48, 51, 54, 56 through 61, 63, 64) involved Niue hybrid genotypes of low fe­cundity. However, Niue crosses and hybrids with Rarotonga and Aitutaki from the 1964 collection were of normal fecundity (Table 5) . The data on tests with Palmyra and Tutuila light and dark ananassae collected in July-August, 1962, show the effect of the isolating factors in these species (Tables 3 and 6). The data on fecundity are difficult to analyze. Both Tutuila dark and Palmyra laid about 20.4 eggs per day, so many of the stock tests had values below 20.0 (Table 3, numbers 1-30). The initial crosses (numbers 50­61 ) averaged 21.8 eggs per day and the two three-way crosses averaged 24.5 (numbers 62, 63). The Fi random matings from crosses between Palmyra, Tutuila light and Tutuila dark ananassae averaged 22.0 eggs per day (numbers 70-75), while the F2 random matings from these crosses averaged 20.0 eggs per day (numbers 76-81) for an average of these two groups of 21.0. When these same Fi hybrids were backcrossed, the number of eggs laid per day was low. The backcrosses involving only Tutuila light and dark ananassae averaged 15.3 (num­bers 64-69), those between either Tutuila light or dark and Palmyra averaged 17.7 (numbers 82-93), with average of backcrosses equal to 16.9 eggs per day. These were tested at the same time so it is not probable that the laboratory condi­tions can have been responsible. These suggest that the sperm (or seminal fluid) and the female genital apparatus interact to reduce laying when one parent is of these strains and the other is of hybrid origin. In view of the higher rate of laying in crosses between strains and in crosses between hybrids and the three-way crosses, the tests suggest the occurrence of hybrid substances which interact against the parent strains and reduce backcrosses. Such hybrid reactions would be interesting mechanisms that reduce the effectiveness of hybridization. These may be related to the mechanisms involved in the Niue hybrid fecundity reductions in the 1963 tests. The results for Palmyra and the South Pacific islands differ from the situation found earlier in the northern Marshall Islands (Stone et al., 1962). In the Mar­shalls, we had four or five years of tests involving stocks and crosses. We found these effects in crosses between the ananassae stocks: crosses between Rongelap females (stock 19.7 eggs per day) and Bikini males (stock 26.7 eggs per day) averaged 34.3 eggs per day; their Fi X Fi averaged 26.1, between the average of the stocks but nearer Bikini. This was stimulation of the female reproductive apparatus by alien sperm or seminal fluid to lay more eggs than the parent strains. In contrast, Majuro females (stock 21.4 eggs per day) mated to Ponape males (stock 24.6 eggs per day) averaged only 16.1 eggs per day, while Fi X Fi aver­ Stone et al.: Pacific Populations of Drosophila aged 22.2 eggs per day, falling between the parent types. This is a case of inhibition or lack of stimulation of laying by the alien semen. It represents a special type of isolating mechanism. In the other four crosses (Bikini females by Ponape males, Majuro females by Bikini males, Rongelap females by Majuro males, Ponape females by Rongelap males), the average number of eggs laid per day fell between the parent averages. In the first of these (Bikini X Ponape), the Fi X Fi gave a very slightly lower egg development than the parents or the cross. This is similar but not identical to Niue hybrids in 1963. In the other three cases the Fi X Fi were heterotic, producing on the average 1.5, 3.8 and 1.3 more eggs per day than the higher parent. These averages for large tests over four or five years seem convincing. Summary Table 7 shows the fertility (fertile pairs/ all pairs) of stocks and crosses for each collection. In checking fertility for March, 1962, and July­August, 1962, we see the difference in the Palmyra and the Tutuila light and dark ananassae. The July-August data show the effect of the different species, Tutuila light, especially well. The data for the material collected in July-August, 1963, show a regular in­crease in fertility each generation in the lab for both stock tests and crosses. We do not know what detrimental influence the ananassae were recovering from. Neither egg laying nor development so obviously improved in the lab. However, the average eggs per day laid in crosses involving Niue was decidedly below the average of stocks, which may be a related phenomenon. The fertility of the strains collected in 1964 was good in both the strains and crosses. The Niue strains were only slightly more fertile that year than the F1and F2 strain tests of 1963. In all tests the P1 were mated in pairs. Some of the F1were sib mated; others crossed at random (offspring from different P 1pairs were crossed) . The F2 from each procedure (sibs or random) were random mated in pairs. Although the same lots of F1 flies were used in both random and sib matings, those F2 from sibs were more often sterile. We conclude that there are recessive sterility factors that be­come homozygous on inbreeding which reduce the fertility of flies whose parents were sibs. These sterility factors were present in several island strains tested. Similar results have been obtained for Mainland populations, for example in Drosophila pseudoobscura, Dobzhansky et al. (1963). Viability (development from egg to adult) is one of the more important vari­ables under natural selection. Viability varies with the amount of inbreeding and past inbreeding with fewer unaccounted fluctuations due to changes in food or other laboratory conditions than fecundity or fertility. Information about the genetic structure of a population can be obtained by comparing the viabilities of eggs from tests involving different degrees of inbreeding. Isofemale lines are necessary in measuring viability and useful in studying isozyme variation. Table 8 summarizes the average viabilities with different degrees and histories of inbreed­ing. Tables 2 through 5 give the details. Figure 3 makes some comparisons of these several populations with those from the Pacific Proving Ground. Table 9 shows the lethal equivalents per gamete (lethals plus summed detrimentals) calculated by the method of Morton, Crow and Muller (1956) and Greenberg and Crow (1960). Tables 8 and 9 (and the original data) show clearly that the greater the degree of inbreeding of the parents, the greater the reduction in egg development. This is modified by population size and other factors such as radiation from fallout. These tests involve populations of different sizes from areas of limited to rich ecological divergence. The Drosophila pseudoobscura and Drosophila novamex­icana are Mainland populations of different sorts. The island populations as col­lected from 1962 through 1964 had a relatively high level of egg development when crossbred. The Pacific Proving Ground populations from Bikini, Rongelap and Rongerik did not give a good development even when crossbred. The inbreed­ing depression, or lethal equivalent per gamete which is a related measure, is not often larger than those of the South Pacific island populations and Palmyra, or than those of Ponape in the eastern Carolines (Tables 8 and 9). The Marshall Islands populations have, in addition, a set of mutations with a dominant com­ponent (judged by crosses between population) that reduce the viability (egg de­velopment of the heterozygote) even when outcrossed. In fact, in crosses between populations only Rongelap x Majuro gave a (slightly) higher egg development than random matings within strains. In crosses between Bikini and Ponape, and Ponape and Rongelap, the egg development fell between the values for random matings within the parent strains. Three crosses, Bikini x Rongelap, Majuro X Bikini and Majuro x Ponape, gave egg development values appreciably below either of their parent strains. The structure of a population is determined both by the recessive detrimental factors and those which can have marked detrimental effects in heterozygotes or homozygotes, as in these Pacific Proving Ground popu­lations (see Figure 3 b, c and d for comparisons, using pictographs). Figure 3 a .Drosophila nasula 0 100 Morch 1962 July, August 1962 Morch 1962 July, August 1962 ~ Palmyra Tutuila f:-~~~~·-~~ ~ g70 -50.5 ~ _J~ -64.7 rs! Random _J] Brothers X . fl unrelated '?'?_ll First l!J Cousins --1l Fu11 Sibs _J Perility ~ 1~ -~ ~ 60'--~~~~_,_~~~~__J ~~~-~5=6=. 5'--~~~~--' FIG. 3. Tests of fitness with Drosophila ananassae and Drosophila nasuta. The pictographs diagram certain important components of reproductive fitness, fertility and egg development, in tests involving different degrees of inbreeding or crossbreeding. The histo­~ams show the frequency classes of females that had 0% eggs develop (sterility, plotted down m black), 1-19%, 20-39%, 40-59%, 60-79%, 80-100% (plotted from left to right above the line as percent of fertile females in that class range). A glance at such a histogram shows the :eprodu.ctive effectiveness of the several types of females. Different shading shows the degree of mbreedmg of the cross, see symbol explanation by 3a. 15 Stone et al.: Pacific Populations of Drosophila Figure 3 (continued) b Drosophila ananassae 100 l_M_or_ch_l9_62___.:..;Ju"-'ly.!..';.;.Au:;_,g.:.us.:..;t....;1.:.9.:.62=----, _ - I'° ~ "~ ~Polm~y" ,! 0 -399 ~ 855 861~ ~ -417 o 80 -451 -55 3 -56 o~ t~ _;:~~ 70 -450 172.s w _J -58.6 a: w I-60 en ~ Drosophila ananassae July, August 1963 c Niue 0 100 w Q_ 0 _J w ~ 90 0 ~~.J -53.8 -56.5 -40.0 t~ ""w _J 70 a: (7'-2 w 1­ (f) -56.5 ~ 60 d Pon ape (P) Suva ~' -47.2 ~ -59.I ~To,"' _,:'.'~,J. -49.6 83.6 - -34.2 f! -45.4 Light D. ananassae Dark D. ananassae Morch 19 62 July, August 1962 Morch 1962 July, August 1962 Tutuilo Tutuilo -82.3 ~ rfp; 894 88 2 -501 -453 ~ -5:~1~ -44.2 I -55.1 -78.2 Drosophila ananassae Niue 6 -20.8 -46.5 ~J ~ -32.6 ~9 -43.I July, August 1964 Ailutaki J9J3 -22.3 -32.2 ~ -29.4 ~ -38.6 Drosophffa onanassae JUNE, JULY, AUGUST 1955-59 Majuro(M) Rongelap (A) Rongerik (K) Rorolanga .J,.J J -30.3 82.0 82.1 -46.1 -28.8 ~3 -35.5 Bikini (B) IOO 56 57 58 59 fJ~j~ ~ 80 rs 80.0 ~ -44.6 ~ "¢ti! ~ -32.2 -43.5 ~-40. 60 -33.1 t w a: _J w t; 50 55 56 57 58 59 ~~ .,D . 74.9 _ _J~ Fi-22.9 ~4~ -488 -:~:t! -451 63.7 . 61.0 ~6 -53.7 -36.6 -46.5 ~3 55 56 57 58 59 55 56 57 58 59 55 56 57 58 sl raJl n.J ~-33.3 rll ~-10.6 -1i.~.4 ~ -no .JJ~ "' ~~ -42:.a r4 5....~-435 E2-!6 r ,. F 1 60311111111 -40.6 -4271612 ':::r! 604 -236 -340(57.7 -44.5 -27.9 55 2-63 7 -50.6 Jl!lil. "!­ 146:"4 -52.6 ~ -27.2 40'--.L----'--'----'--.....__ _,__'--~-~1~;:4__.___..____.___.___.__.____,___.'--__.___._.....___,_~ In addition, over the four or five years tested, the average number of pairs fer­tile in initial crosses between members of populations from different islands was appreciably below the fertility of the parent strains in four tests (Majuro X Ponape, Ponape X Rongelap, Rongelap X Majuro and Bikini X Ponape) and in­termediate in the other two tests (Majuro X Bikini and Bikini X Rongelap). These physically semi-isolated populations have developed some behavioral iso­lating mechanisms between them so that they are reluctant to cross-mate, even without choice of potential mates. There remain several interesting analyses. Population size influences the num­ber of lethal equivalents per gamete markedly. Professors Th. Dobzhansky and William Heed collected the pseudoobscura reported in Table 9. Three of the popu­lations were large but Madera Canyon was small; Dr. Heed was able to collect only 200 pseudoobscura in two days because of the dry weather. Dr. Heed col­lected the novamexicana, also, in 1953. Unlike pseudoobscura, this species occurs only in very small populations along desert streams. Both these populations have few lethal equivalents per gamete (0.30 and 0.22). The other three pseudoobscura populations were large and have about 0.51 lethal equivalents per gamete. Many of the larger island populations have a much greater frequency of lethal equiv­alents per gamete, reaching a peak of 1. 40 for Bikini in 195 7. The small Pacific island populations, especially Rongerik ananassae and Pal­myra nasuta, show the effect of small population size in the reduced number of lethal equivalents (0.27, 0.28 and 0.31). Nevertheless, Rongerik ananassae and Palmyra nasuta are very different populations. The Rongerik population had an egg development in 1959 of only 64.6% from random matings. The random mat­ings in the well adapted small populations, using viability as the measure, are: Madera Canyon pseudoobscura, 88.5 % ; Patagonia novamexicana, 91.7% ; Pal­myra nasuta, 87.6 and 85.6%; Tutuila melanogaster, 86.9%. These are in the same ranges as the larger continental populations of the very successful pseudo­obscura. The larger South Pacific island populations are, also, in this range, as is Ponape in the eastern Carolines. However, none of the Marshall Islands popula­tions of ananassae are well adapted on the basis of egg ·development from random matings. For the four years 1956 through 1959, these ranged from 64.6% to 83.6% with three values in the sixties, five values in the seventies and three values in the low eighties (Table 9). The range of lethal equivalents per gamete is about the same as the much better adapted populations from Palmyra and South Pacific islands. Clearly lethal equivalents do not by themselves characterize the adaptive values of populations, nor even describe the total genetic load. The crosses may show heterosis such as the crosses between the pseudoobscura populations (Stone, Wilson and Gerstenberg, 1963) or, as is often the case with the Pacific island populations, show some degree of isolation, culminating in (nearly) complete isolation between the Tutuila light and dark ananassae. The reduction in egg development encountered in crosses may be developmental lethals or sometimes low sperm transfer. The isolation between Tutuila light and dark ananassae is sexual. They cross with reluctance even without choice of mate. When males and females of both types are together in a vial, they do not cross (Futch, this Bulletin) due to sexual (behavioral) isolation (Spieth, this Bulletin). Stone et al.: Pacific Populations of Drosophila Once the initial cross had been made between the two Tutuila species or either of them and Palmyra, or any of the crosses any year involving other island popu­lations, the F,, F2, etc. proved normal. There are a few examples where recombi­nations in F2 or later may not have been as good as the parent strains, but these are too few and scattered to be significant. Nevertheless, some combinations and recombinations with the Tutuila light ananassae and other forms were aberrant, Tables 2 and 3. As they live together and are isolated sexually, they represent an example of Dobzhansky's (1951, and earlier) hypothesis that with contacts fur­ther isolating factors may have developed and been selected since they came to­gether. Limited sperm transfer may be one of them. Selection for sexual isolation from Tutuila light may have incidentally led to the rather considerable isolation between Tutuila dark and Palmyra ananassae although these are different island strains of the same species. The two species of the ananassae complex, Tutuila light and Tutuila dark, are siblings separated almost entirely by sexual isolation. This case resembles the sibling species found by Carson ( 1954) in the willistoni group of Drosophila. These are Drosophila bocainensis and Drosophila parabocainensis. These cannot be separated morphologically but, like these Tutuila species, differ consistently cytologically. Carson could not demonstrate sexual isolation as the critical isolat­ing mechanism and proposed some subtle ecological preference to explain their isolation in nature. The hybrid males with bocainensis mothers are of reduced fertility. These cases resemble each other in many respects. In comparing crosses to intrapopulation tests made in 1963 and 1964, no serious isolation or heterosis was detected in egg development tests. The P1 crosses gave egg development like the random matings within stocks, and so did the random matings of F1, F2 , etc. of crosses. The egg development from sib matings of F1 from crosses was similar to the intrapopulation sib matings. A few low tests were found but they were so scattered and different from other related tests (for ex­ample, the results of the sibs crosses in other ways) that these must be the result of variation in laboratory conditions. Johnson et al. (this Bulletin) tested for the presence of enzyme variation in a number of species of Drosophila. These are extensive and we found, as did Lew­ontin and Hubby (1966), that 40 to 50 percent of the enzymes we have identified are polymorphic in several species. There are some decided differences between subgenera and even between species groups in the enzyme patterns and varia­tion. In a study of the ananassae populations collected on Tutuila, Upolu and Savaii, Johnson et al. (1966a) found about half the identified enzymes show poly­morphism implying that half the genes determining these enzymes (for our purposes 1 gene : 1 enzyme is a reasonable estimate) have two or more alleles. A few have alleles which produce no enzyme, but most are represented by alleles that produce slightly different enzymes, separable by starch electrophoresis, stain­ing and inhibition studies. The light and dark ananassae found on Tutuila and U polu are most informative. Both light and dark forms have several esterase alleles, including a fast, medium and slow moving enzyme. The relative frequen­cies of these three enzyme forms in the light ananassae are alike on Tutuila, Upolu and Savaii. The dark have similar frequencies on Tutuila and Upolu (we have none from Savaii). The dark and light differ markedly on all islands: aver­age values for enzyme types in the light = F, 0.14; M, 0.80; S, 0.06 and dark species averages = F, 0.76; M, 0.23; S, 0.01 (see Johnson et al. for details). The other enzymes differ in some cases, also. The enzyme variation should prove most useful in studying mutation pressure, drift and selection effects in these natural laboratories. Multiple factors, selection for modifiers of dominance, polygenes, and the mag­nitude of mutation variability have been subjects for bitter debates over many years. Here we have revealed the extensive genetic variability that can be very useful for rapid selection. These isozymes differ slightly in their chemical proper­ties as revealed by their separate identification, including crosses that show the allelism or similarity of these genes in the light and dark ananassae. No system proposed for the effects called polygenes has been reasonable. Here we have the genetic variability represented by different isoalleles in various combinations that can account for polygenic (multiple factor) variation and the effectiveness and rapidity of selection in many cases. The extensive genetic divergence of these island populations illustrates well the effectiveness of the system. SUMMARY 1. The 1962 test of atomic and thermonuclear devices on Christmas Island did not drop appreciable fallout on Palmyra Island, the island close to Christmas which had Drosophila populations. Therefore, we have no new data on the genetic effects of fallout. 2. Niue has a high background radiation due to radium and its decay products, reaching 0.350 micro-rads per hour using a counter sensitive to alpha particles in the rich soil where gardening is done. Drosophila ananassae collected in the area of high radiation did not show an appreciably greater inbreeding depres­sion than other island populations in that area. However, in 1963 Futch (this Bulletin) found two pericentric inversions and a translocation (the latter in one fly of a number checked from an isolated female's progeny). These might have been caused by alpha particle radiation in the gonads, but if so this is rare for none were found in a like number of lines (82) from the locality with high background in 1964. 3. Palmyra and these South Pacific islands prove quite interesting in the study of isolated populations of different sizes, living in various ecological conditions. 4. The number of eggs laid per day by young mature fertilized females in these island Drosophila populations usually vary from 20 to 30, although the species studied vary markedly in size. 5. The fertility in pair matings is fairly erratic. However, it was possible to show that inbreeding made recessive sterility factors homozygous which reduced the fertility about 10% in offspring of sib matings. 6. Variability, measured as percent of eggs that developed into adults, varies in­versely with inbreeding. As inbreeding increases, the percent of eggs that develop decreases. If flies whose parents were sibs are mated at random so the progeny developing are heterozygous, egg development is reduced, showing the Stone et al.: Pacific Populations of Drosophila material and/ or paternal effect of the homozygous detrimental factors on their offsprings' development. 7. There is a decided effect of population size on the effect of inbreeding. Small populations have a much smaller inbreeding depression than larger ones, for example Palmyra nasuta and Rongerik ananassae, Figure 3 and Table 9. 8. The summed genetic load of a population is the sum of the effect of recessive factors which reduce egg development on inbreeding and their dominant com­ponents (these can be of minor effect on a population) plus the homozygous recessive and the dominant factors which depress the viability of a population in outcrosses as well as on inbreeding. The latter is the more serious component in reducing the adaptation of a species, see, for example, Rongerik ananassae. This component depresses the reproductive effectiveness of even such effective populations as pseudoobscura, Table 9. 9. These island populations as well as those of the Marshall Islands and Ponape, exhibit varying degrees of genetic divergence and partial genetic isolation. This culminates in the remarkable sibling species, Tutuila dark and Tutuila light ananassae. These live together but very seldom mate in nature for they possess distinctive chromosome arrangements (Futch, this Bulletin) and isozyme poly­morphism (Johnson et al., 1966a). If given no choice of mate, they cross and the F, and F2 show no, or minor, residual isolation. This complete initial isola­tion is, therefore, sexual (behavioral) and dependent on no other isolating factors for its success. LITERATURE CITED Carson, Hampton L. 1954. Interfertile sibling species in the willistcini group of Drosophila. Evolution 8: 148-165. Dobzhansky, Theodosius. 1951. Genetics and the Origin of Species. New York: Columbia University Press. Dobzhansky, Th., B. Spassky, and T. Tidwell. 1963. Genetics of natural populations. XXXII. Inbreeding and the mutational and balanced genetic loads in natural populations of Dro­sophila pseudoobscura. Genetics 48: 361-373. Fieldes. M., G. Bealing, C. G. Claridge, N. Wells, and N. H. Taylor. 1960. Mineralogy and radioactivity of Niue Island soils. New Zealand Journal of Science 3: 658-675. Futch, David G. 1966. A study of speciation in South Pacific island populations of Drosophila ananassae. This Bulletin. Greenberg, R., and J. F. Crow. 1960. A comparison of the effect of lethal and detrimental chromosomes from Drosophila populations. Genetics 45: 1153-1168. Johnson, F. M ., Carmen G. Kanapi, R. H. Richardson, M. R. Wheeler, and W. S. Stone. 1966a. An analysis of polymorphisms among isozyme loci in dark and light Drosophila ananassae strains from American and Western Samoa. Proc. Nat. Acad. Sci. 56: 119~125. Johnson, F. M., Carmen G. Kanapi, R.H. Richardson, M. R. Wheeler and W. S. Stone. 1966b. An operational classification of Drosophila esterases for species comparisons. This Bulletin. Lewontin, R. C., and J. L. Hubby. 1966. A molecular approach to the study of genie heterozy­gosity in natural populations. II. Analysis of geographical populations of Drosophila pseudo­obscura. Genetics (in press). Marsden, E. 1964. Radioactivity of some rocks, soils, plants and bones. In The Natural Radia­tion Environment. Edited by J. A. S. Adams and W. M. Lowder. Chicago: The University of Chicago Press. Marsden, E., G. J. Fergusson, and M. Fields. 1958. Notes on the radioactivity of soils with application to Niue Island. Proc. 2nd Int. Conf. Peaceful Uses of Atomic Energy, Geneva, 18: 514. Morton, N. E., J. F. Crow, and H.J. Muller. 1956. An estimate of the mutational damage in man from data on consanguineous marriages. Proc. Nat. Acad. Sci. 42: 855-863. Spieth, Herman T. 1966. Drosophilid mating behavior. The behavior of D. ananassae and D. ananassae-like flies from the Pacific. This Bulletin. Stone, Wilson S., Marshall R. Wheeler, and Florence D. Wilson. 1962. I. Genetic studies of irradiated natural populations of Drosophi.la. V. Summary and discussion of tests of popula­tions collected in the Pacific Proving Ground from 1955 through 1959. Univ. of Texas Puhl. No. 6205: 1-54. Stone, Wilson S., Florence D. Wilson, and Virginia L. Gerstenberg. 1963. Genetic studies of natural populations of Drosophila: Drosophila pseudoobscura, a large dominant population. Genetics 48: 1089-1106. Tables 2 through 5, general comments These tables give the results within and between populations of the effects of various degrees of inbreeding and crossbreeding on three important components of fitness under natural selection: fertility in pair matings, fecundity of young, mature fertilized females, and viability measured as the percent of eggs that develop laid by these fertilized females. Data are given on the egg development patterns (the important component of fitness which yielded the best quantitative data) ; this is expressed as the percent of fertile females that laid eggs that devel­oped in classes recorded as 1-19%, 20-39%, 40-59%, 60-79% and 80-100%. Taken with the percent sterility, these give a profile of the net productive efficiency of the populations, as shown in Figure 3. TABLE 2 Tests with Drosophila ananassae and D. nasuta from Palmyra, D. ananassae light, D. ananassae dark, D. melanogaster and D. nasuta from Tutuila, collected in March, 1962 Fertility Fccunnirs Percent Days eggs Eggs/ Total Percent tested fertile counted 2. brothers X unrelated 'i? 'i? 3. sibs F 1 tests 4. cousins 5. random from P1 sibs 203 280 116 163 31.1 21.8 37.9 25.8 112 124 85 66 28.8 22.8 20.5 16.5 3222 2826 1741 1087 88.9 62.5 78.9 74.1 4.4 4.0 6.7 8.0 10.0 2.4 37.8 15.0 24.4 24.4 20.0 15.0 73.2 26.7 68.0 60.0 41 45 25 20 C::! ;:s-.<::: ('=> .... "'-. ... ~ 6. sibs from P 1 brothers X unrelated 'i? 'i? 170 26.5 115 22.6 2601 68.0 5.4 27.0 48.6 18.9 37 .Q_ 7. random from P 1 brothers X unrelated 'i? 'i? 138 38.4 99 28.6 2830 89.4 11.4 88.6 35 "l ('=> H ~ F 2 tests 8. random from F 1 random "l:l $::: from P 1 sibs 9. random from F 1 sibs from P 1 brothers X unrelated 'i? 'i? D. ananassae dark (Tutuila) 143 141 48.3 45.4 120 137 38.3 25.2 4591 3454 85.5 86.3 3.8 1.9 10.9 19.2 13.0 75 .0 76.1 52 46 <;J-<..._-.2... o· ;:s P 1 tests 10. random 371 49.9 402 18.6 7450 89.4 2.3 10.1 87.6 129 11. brothers X unrelated 'i? 'i? 243 54.7 287 16.2 4678 88.2 2.2 12.9 84.9 93 12. sibs 267 55.8 352 22.1 7785 74.9 2.5 19.2 25.0 53.3 120 F1 tests 13. cousins 156 42.9 114 1 t.O 1252 80.1 2.7 5.4 1.'3.5 29.7 48.9 .'37 14. random from P, sihs 146 54.1 125 22.3 2793 86.2 9.1 18.2 72.7 44 15. sihs from P, brothers X unrelated 'i? 'i? 232 46.1 170 20.2 3437 63.0 5.3 24.6 43.9 26.3 57 16. random from P1 brothers X unrelated tests F2 17. random from F 1 random from P1 sibs 18. random from F1 sibs from P1 brothers X unrelated D . anana.ssae (Palmyra) tests P1 19. random 20. brothers X unrelated 21. sibs tests F1 22. cousins 23. random from P 1 sibs 24. sibs from P 1 brothers X unrelated 25 . random from P 1 brothers X unrelated F2 tests 26. random from F 1 random from P1 sibs 27. random from F 1 sibs from P 1 brothers X unrelated D. ananassae (Rarotonga) 28. P1 random 29. sibs F1 30. random from F1 sibs F2 D. simulans (Rarotonga) 31. P1 random 32. F1 sibs 33. random from F1 sibs F2 161 141 135 389 270 280 150 159 180 174 141 138 358 181 145 129 156 120 36.6 66.6 45.9 44.7 44.0 41.4 70.6 54.7 65.0 58 .6 52.4 41.3 67.3 61.9 49.0 40.3 34.6 60.8 145 193 133 403 274 220 179 82 171 206 188 131 433 189 145 136 69 202 23.5 26.1 22.7 20.6 19.9 22.0 12.9 23.2 17.3 19.7 21.0 27.3 24.5 24.5 16.2 17.1 25.0 17.6 3411 5033 3017 8296 5451 4848 2314 1901 2951 4062 3944 3572 10,608 4638 2353 2321 1723 3554 84.5 90.2 76.4 86.1 87.9 72.8 76.0 81.7 70.5 87.1 87.1 76.5 65.9 56.4 79.7 87.6 67.3 84.7 2.0 0.8 2.6 6.3 1.4 1.6 4.7 5.9 4.1 1.6 2.1 14.3 0.8 1.1 2.6 6.2 2.8 15.4 18.8 2.5 12.0 1.6 4.2 8.2 6.1 4.5 26.3 16.9 21.1 2.6 1.6 18.4 18.3 29.4 10.2 5.0 20.0 6.3 22.9 12.7 14.3 15.9 17.0 27.6 24.6 21.9 36.6 18.4 13.1 24.5 23.7 22.4 18.4 15.0 52.0 22.2 70.8 87.3 61.2 76.5 77.3 40.8 52.3 71.9 38.0 78.9 83.6 57.1 37.9 23.5 67.3 77.5 16.0 68.3 48 71 49 132 88 76 65 32 71 76 61 49 169 85 49 40 25 63 [/.) ..... § ~ ~ ..... i::i :­ ~ ..... ';::t) (") ..g ~ i::: ~ ..... §" "'c - ~ g .g ;J-­ ..... ~ to w to TABLE 3-Continued -i;:. Fertility Fccun X TdarkJ J 168 51.2 72. P '? '? X T light J J 132 41.7 73. T light'? '? X P J J 120 47.5 74. T dark'? '? X P J J 120 45.0 75. P '? X T dark c1 c1 121 62.8 Fecundity Days eggs Eggs/ counted day/'f (3) (4) 186 24.5 23 3 22.6 114 23.7 128 22.8 153 16.8 197 21.5 Development Total Percent eggs developed (5) (6) 4553 88.4 5255 84.8 2705 83.4 2919 81.8 2573 86.5 4239 88.6 -­1-19 (7) Percent in each class range of egg development ------· 20-39 40-59 60-79 80-100% T otal '?I' (8) (9) ( 10) ( II ) (12) 3.6 21.4 75.0 56 7.5 26.9 65.7 67 7.9 28.9 63.2 38 2.3 6.8 22.7 68 .2 44 18.4 81.6 49 6.2 9.2 84.6 65 "-3 ;:i-­(1) <:::!;:s-· a5 ;;::-­..... ~ .Q.. ~ F 2 random from random from P1 two-way crosses 76. T dark'?'? X T lightJ J 127 50.3 77. T Lght '? '? X T dark J J 11 9 53.7 78. P '? '? X T light J J 142 40.8 79. T light'? '? X P J J 138 48.6 80. P '? X T dark J J 127 55.9 81. T dark'? '? X P J c1 94 31.9 154 172 89 149 196 72 20.8 20.0 23.6 15.6 20.5 19.7 3199 3431 2078 2318 4011 1420 80.7 79.4 69.5 78.8 87.1 85.4 1.9 3.8 9.7 1.9 4.5 9.6 13.2 19.4 5.8 9.8 21.2 26.4 48 .4 42.3 14.8 31.8 67.3 56.6 22.6 50.0 75.4 63.6 52 53 31 52 61 22 ~ "i:I !::: ~ .......2· .....5· ;:s F1 backcrosses 82. Tdark '? X F1 (Tdark '? X Palmyra J J ) J 83. F1 (T dark X P ) '? '? X P J J 135 151 57.0 39.7 180 81 15.0 20.7 2695 1675 85.5 72.4 1.6 6.1 6.1 1.6 9.1 13.1 45 .5 83.6 33.3 61 33 84. T dark~ ~ X F 1 (PX Tdark) t t 118 85. (PX T dark)~ ~ X F1 T darkt t 116 86. P~ ~ X F1 (PX T dark) t t 125 F1 87. (PX T dark)~ ~ X Pt t 126 88. T light~ ~ X F 1 (T light X P) t t 122 F1 89. (Tlight X P) ~ ~ X Pt t 132 90. T light~ ~ X F 1 (P X T light) t t 117 91. (PX Tlight) ~ ~ X F1 T lightt t 109 92. P~ ~ X F1 (PX Tlight) t t 134 F1 93. (PX Tlight) ~ ~ X Pt t 126 64.4 50.0 63.2 69.0 17.2 50.0 44.4 49.5 59.0 44.4 169 114 170 248 17 74 56 118 159 149 12.2 14.9 15.1 17.0 14.6 23 .8 19.1 21.9 21.0 17.8 2065 1699 2575 4227 248 1764 1070 2579 3335 2654 54 90.5 16.7 83.3 91.1 8.1 5.4 86.5 87.0 3.4 15.3 81.4 91.1 13.4 86.6 85 .4 20.0 80.0 83.6 25.0 75.0 85.6 4.5 18.2 77.3 82.1 7.1 28.6 64.3 87.2 9.8 5.9 84.3 80.9 4.4 2.2 4.4 26.7 62.2 37 59 82 5 36 22 42 51 45 (/.) .... 0 ;:s (I) ~ !::) ;­" ;p (") ..... ~ (") ~ "1::1 !::: ~ .... c;· ~ 0 - t::l-; g ~ ~ ..... S" to ~ TABLE 4 N) co Tests with Drosophila ananassae collected on Niue, Suva (Fiji) and Tonga in July, August, 1963 Speril~s. islands and nrntings Fertility ----­P<1irs Percent tested fertile c:olumn : ( ! ) (2) Fecundity Days eggs Eggs/ counted day/'i' (3) (4) Development Tota l Percent eggs developed (5) (6) 1-19 ( 7) Percent in each class range of egg development 20-39 40-59 G0-79 80-100 % T otal '?'? (8) (9 ) ( 10) (11 ) ( 12) D . ananassae Niue(N) , Suva(S), Tonga(T) P 1 tests '"-j ;:::;... ~ 1. N random 424 46.2 423 27.7 11716 89.3 0.7 2.1 2.1 10.6 86.5 141 ~ 2. T random 3. S random 219 195 42.9 52.8 217 287 31.3 27.7 6791 7963 93.9 90.4 3.3 8.3 14.4 91.7 82.2 72 90 ;::!-.\::: ~ 4. N brothers X unrelated~ ~ 418 43.5 467 27.3 12771 88.1 3.8 16.6 79.6 157 ;:;-· 5. T brothers X unrelated~ ~ 262 50.4 341 29.9 10199 89.5 0.8 2.5 2.5 13.1 81.1 122 ..... ~ 6. N sibs 7. T sibs 8. S sibs 434 282 93 43.5 54.6 40.9 426 338 120 29.9 33.5 24.2 12744 11388 2970 70.2 70.4 63.9 0.7 2.6 7.5 6.9 10.8 18.5 20.7 32.4 39.0 37.1 27.0 34.2 32.8 29.7 146 116 37 .Q.. '"-j ~ F 1 tests 9. N random from P 1 random 10. T random from P 1 random 11. S random from P 1 random 279 188 149 61.6 76.1 69.8 333 348 253 24.1 19.5 31.7 8031 6794 7996 91.1 87.0 88.8 1.2 1.0 0.8 2.9 1.7 10.6 16.7 15.7 85.6 80.8 83.1 104 120 83 "ti i:::c:r­-Fl . 12. N sibs from P1 random 13 . T sibs from P 1 random 233 197 57.5 70.0 306 284 28.8 30.0 8801 8522 67.8 70.0 1.0 1.8 6.0 7.1 25.0 21.4 27.0 33.0 41.0 36.6 100 112 ~ -· 0 ;::! 14. S sibs from P 1 random 139 61.9 22Z Z6.3 5836 67.7 10.1 29.1 32.9 Z7 .8 79 15. N random from P 1 sibs 94 43.6 76 25.2 1914 82.3 6.7 13.3 6.7 73.3 30 16. T random from P 1 sibs 76 28.9 40 20.0 803 87.1 11.8 17.6 70.6 17 17. S random from P 1 sibs 138 49.3 195 31.1 6070 82.1 3.0 3.0 7.6 16.7 69.7 66 18. N sibs from P 1 sibs 349 51 .6 399 28.0 11173 64.9 Z.1 15.1 21.Z 36.3 Z5.3 146 19. T sibs from P 1 sibs 183 51.4 230 24.8 5698 61.4 2.5 17.5 23.8 33.8 2Z.5 80 20. S sibs from P 1 sibs 71 54.9 109 ZS.9 2823 6Z.7 2.6 13.2 18.4 47.4 18.4 38 21. N cousins from P 1 brothers X unrelated~ ~ Z20 60.0 262 27.7 7259 88.0 3.5 22.1 74.4 86 22. T cousins from P 1 brothers X unrelated~ ~ 23. N random from P 1 brothers X unrelated ~ ~ 24. T random from P 1 brothers X unrelated~ ~ 25. N sibs from P 1 brothers X unrelated~ ~ 26. T sibs from P 1 brothers X unrelated~ ~ F2 tests 27. N random from F 1 random from P 1 random 28. T random from F1 random from P 1 random 29. S random from F 1 random from P 1 random 30. N sibs from F 1 random from P 1 random 31. T sibs from F1 random from P 1 random 32. S sibs from F 1 random from P 1 random 33. N random from F1 ·. sibs from P 1 sibs 34. N sibs from F1 sibs from P 1 sibs 35. N random from F1 cousins from P 1 brothers X unrelated~ ~ 36. N sibs from F 1 cousins from P 1 brothers X unrelated~ ~ 155 223 193 310 203 111 142 186 201 191 178 158 163 115 202 65.8 65.9 68.4 64.8 71.4 73.9 76.1 68.3 69.2 56.5 69.7 43.7 53.4 61.7 71.3 228 328 284 399 320 217 256 318 337 253 255 127 152 197 327 20.0 23.6 25.3 26.5 27.7 27.9 26.6 20.8 27.2 21.3 21.2 21.4 23.8 26.1 26.4 5769 7733 7188 10569 8875 6054 6814 6568 9167 5393 5406 2720 3623 5143 8623 83.6 86.2 84.1 73.3 69.6 91.0 89.1 85.0 65.7 82.7 66.1 76.5 60.9 90.9 73.4 2.9 1.0 2.0 0.8 1.0 0.9 2.1 2.2 2.3 10.7 1.1 1.9 1.0 2.0 4.7 5.7 0.9 1.9 6.4 10.4 13.0 4.7 7.1 1.1 5.7 5.4 4.9 2.0 16.9 24.6 7.8 4.5 3.9 31.2 16.7 20.7 14.0 30.4 1.1 14.2 27.0 13.6 24.5 33.1 38.5 12.1 5.7 16.5 34.9 32.3 29.3 25.6 25.0 11.4 34.9 67.6 77.7 70.4 43.2 30.3 79.3 89.8 76.7 26.6 38.5 34.8 53.5 26.8 85.2 43.4 74 103 98 148 122 116 88 103 109 96 92 43 56 88 106 (/.) ..... 0 ;::i ~ ~ ..... !:) ~ "ti ~ $ () "ti ~ a i::: o" ~ c - ti.., ~ ;::i-. B" -· l'O c.o TABLE 4--Continued w Fertility Fecundity Development ----· -----·-----Percent in each class ra~ge of egg development Pairs Percent Days eggs Eggs/ Total Percent tested fertile counted day /'i' eggs developed 1-19 Z0-39 40-59 60-79 80-100% Total 'i''i' Species, islands and matings Column: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) ( 12) P 1 two-way crosses 37. S 9 9 X T O' O' 204 34.3 174 25.9 4502 89.9 3.6 3.6 92.7 55 38. T 9 9 X N O' O' 188 72.9 342 29.0 9923 93.1 0.9 0.9 4.3 93.9 115 39. N 9 9 X SO' O' F1 tests from two-way crosses 210 71.9 445 18.8 8358 90.9 0.7 6.6 92.7 137 '"-3;:r.. (';) 40. (S X T) 9 9 (S X T ) O' O' random 41. (T X N) 9 9 X (TX N ) O' O' random 42. (S X T) 9 9 X 188 136 56.9 77.9 242 303 26.6 22.0 6444 6665 89.5 85.9 1.0 1.0 2.6 9.3 9.0 12.4 88.5 76.3 78 97 C:! ;::i-.<::: (';)..., "'-...... ~ ( S X T ) O' O' sibs 163 55.2 234 26.8 6264 69.2 1.3 5.1 36.7 35.4 21 .5 79 .Q, 43. (TX N ) 9 9 X (T X N ) O' O' sibs F1 three-way crosses 156 77 .6 342 16.3 5563 70.6 1.9 8.3 13.9 34.3 41.7 108 i 44. (S X TH 9 X N O' O' 45. (T X N ) 9 9 X SO' O' 46. (N X S) 9 9 X T O' O' 195 138 156 48.7 65.2 73.7 138 263 318 18.6 17.1 12.2 2565 4502 3865 91.3 91.2 87.6 1.1 2.1 1.3 2. 1 9.1 7.6 15.8 90.9 91.1 78.9 44 19 95 "tl i::: ~ ........-· () F 2 tests from three-way crosses 47. random from F1 ~ ....-c . ;::i (S X T) 9 9 X N O' O' 153 73.9 281 24.7 6932 78.9 1.1 17.4 25.0 56.5 92 48. random from F 1 (T X N ) 9 9 X SO' O' 117 61.5 33 15 .3 505 76.8 8.3 33.3 58.3 12 49. random from F1 (N X S) 9 9 X T O' O' 84 94.0 245 24.9 6089 93.7 1.3 1.3 3.9 93.5 77 50. sibs from F1 (S X T ) 9 9 X N O' O' 106 68.9 194 22.6 4393 68 .6 3.3 33.3 36.7 26.7 60 51. sibs from F1 (T X N) 9 9 X SO' O' 1+8 90.5 258 16.8 4340 59.6 4.5 14.6 32.6 28.1 20.2 89 52. sibs from F 1 (N X S) X T t t F1 backcrosses 53. S X (S X T) t t 54. (T X N) X T t t 55. T X (T X N) t t 56. (T X N) X N t t 57. N X (T X N) t t 58. (N X S) X N t t 59. N X (N X S) t t 60. (N X S) X St t 61. S X (N X S) t t F 2 tests from F 1 backcrosses 62. random from F 1 ( S X ST) 63. random from F 1 (TN X N) 64. random from F 1 (N X TN) 65. random from F1 (NS X N) 66. random from F1 (N X NS) 67. random from F1 (NS X S) 68. random from F1 (S X NS) 134 99 143 144 138 122 146 141 140 146 144 136 150 133 141 123 64 81.3 54.5 76.9 79.1 54.3 79.5 70.5 71.6 57.9 30.8 72.2 72.1 77.3 85.7 94.3 91.1 82.8 270 108 241 337 133 125 286 255 187 74 235 68 263 358 444 371 118 21.6 19.3 17.0 21.0 16.6 13.1 -15.3 8.1 14.4 10.7 15.4 11.4 12.7 23.3 21.1 20.9 24.9 5827 2080 4091 7062 2209 1603 4388 2067 2691 789 3619 775 3347 8328 9385 7764 2938 70.0 86.2 84.9 89.5 90.9 91.8 89.5 86.7 89.7 90.0 83.0 80.9 86.5 89.0 89.0 91.8 93.7 25.0 1.2 9.5 1.2 2.4 1.2 0.9 3.8 1.0 2.8 7.3 3.0 2.1 3.5 4.1 4.8 6.0 11.1 3.6 0.9 3.0 3.9 2.5 31.0 27.8 11.0 10.0 11.6 8.5 10.5 17.8 17.5 16_0 26.2 29.6 20.2 15.5 7.6 3-9 7.5 33.3 69.4 78.0 87.0 88.4 89.4 84.9 78.1 77.8 84.0 67.9 59.3 76.2 82.7 85.6 91.3 90.0 84 36 82 100 43 47 86 73 63 25 84 27 84 110 132 103 40 ~ ..... 0 ;::l (I) (I) ..... i::i ~ ~ <"') ..... ~ <"') "tl c "'tl !:: ~ ..... o· ;::l -c"' t:i '-; .g ~ ;:::­ ~ w - w to TABLE 5 Tests with Drosophila ananassae collected on Niue, Rarotonga and Aitutaki during July, August, 1964 Species, islands and matings Column: Ferti lity Pairs Percent tested fertile (1) (2) Fecundity Days egJs Eggs/ counte day/I'(3) ( 4) Development T otal Percent eggs developed (5) (6) 1-19 (7) Percent in each class range of egg development 20-39 40-59 60-79 80-100% T otal l'I' (8) (9) ( 10) ( 11 ) ( 12) "'-j D. ananassae ~ Niue (N ), Aitutaki (A), Rarotonga (R) Stock tests 1. N P1 random 2. N P 1 brothers X 387 53.5 524 28.3 14850 83 .6 1.9 8.1 15.0 75.0 160 C::! ;:i-· (::! (':> ;;i-...... unrelated 'i? 'i? 218 79.2 351 25.7 9022 86.1 0.9 4.4 18.4 76.3 114 ~ 3. N P 1 sibs 225 56.9 276 28.7 7914 63.9 1Z.2 30.0 30.0 27.8 90 .Q.. 4. NF1 cousins Z88 67.4 435 30.3 13178 72.6 1.4 4.3 18.4 32.6 43.3 141 "'-j 5. N F1 random ~ from P 1 random 150 50.7 153 31.0 4738 81.6 0.9 9.2 20.2 69.7 109 ~ 6. N F0 random 't:l from F 1 random 7. A P 1 random 8. AP1 brothers X unrelated 'i? 'i? 9. A P 1 sibs Z82 399 283 236 60.3 76.7 67.8 61.4 370 858 496 302 30.0 28. 1 Z3.2 25 .2 11097 24078 11519 7601 86.0 84.9 85.3 61.3 0.4 1.2 4.6 0.8 1.2 0.6 18.3 4.1 3.5 6.1 23.9 18.9 21.8 19.4 34.9 76.2 73.2 72.7 18.3 1Z'2 257 165 109 >::: ~ .......-(") · i::i ..... s· ;:i 10. A F1 cousins 272 70.6 446 24.8 11040 72.1 1.9 3.8 24.4 31.3 38.8 160 11 . A F 1 random from P 1 random 311 73.3 538 Z3.6 12670 86.0 1.8 1.2 3.7 14.0 79.3 164 tZ. A F2 random from F1 random Z89 55.4 Z88 24.7 7100 85.9 1.9 7.5 17.0 73.6 106 13. R P 1 random 14. R P 1 brothers X unrelated 15. RP1 sibs 16. R F1 cousins 17. R F 1 random from P 1 random 18. R F 2 random from F 1 random P 1 crosses 19. AX N 20. N X R 21. RX A tests F1 22. F1 random from A X N 23. F 1 random from N X R 24. random from RX A F1 25. F1 sibs from AX N 26. F1 sibs from N X R 27. sibs from RX A F1 F 2 three-way crosses 28. R 'i? X F1 (AX N) J J 29. F 1 (N X R)'i?'i?XAJJ 30. (RX A) 'i? 'i? X NJ J F1 379 298 290 292 279 293 393 382 373 279 282 254 280 289 310 281 296 291 69.7 58.4 64.5 71.2 60.9 70.6 61.8 55.2 77.2 74.6 74.5 75.6 71.1 74.0 86.1 64.8 77.0 69.8 682 375 431 501 373 500 540 532 767 523 505 522 505 596 564 376 543 438 30.0 23.8 28.7 30.8 30.7 30.9 24.6 28.5 28.9 28.3 26.2 23.8 27.6 27.5 22.1 26.9 23.4 25.0 20494 8916 12358 15429 11 465 15453 13297 15204 22154 14799 13244 12442 13913 16410 12487 10111 13217 10930 85.0 82.0 63.3 82.1 84.6 84.2 85.3 87.1 88.0 86.1 82.8 84.8 60.5 63.5 69.1 88.2 87.1 85.9 .. 2.2 0.6 1.7 0.6 0.4 0.6 0.6 1.2 3.1 1.6 0.5 3.7 13.0 1.3 3.1 1.2 0.6 0.6 0.8 1.2 1.2 2.3 16.8 10.7 7.8 0.7 2.8 1.4 5.4 7.4 29.0 8.6 5.5 4.7 3.9 1.7 2.5 2.4 9.9 4.1 26.7 32.1 27.5 3.8 4.5 1.4 19.5 22.8 34.8 25.7 15.0 21.1 18.0 15.2 14.8 18.2 16.8 18.7 32.9 36.4 34.7 13.1 11.2 15.9 75.1 66.2 21.0 64.5 76.4 72.5 75.8 82.0 81.6 77.6 71.4 73.7 20.5 19.3 29.5 82.3 81.5 81.4 205 136 138 152 127 171 178 178 244 165 161 171 161 187 193 130 178 145 C'/.l ...... c ~ (1) ...... !:) ;.... '1::i !:) ~ s; ~ '1::i c '1:::l $::: ..._ - c i::i ·...... ~ c - tl 2 .g ~ -. ~ (.).) (.).) The University of Texas Publication TABLE 6 Summary of number of eggs laid per day by young mature inseminated females, both within strains and in crosses March, 1962 July, August, 1962 July, August, 1 963 July, August, 1964 Species Eggs/day/'i' Species Eggs/day /I' Species Eggs/day/'i' Species Eggs/day/'i' D. ananassae D. ananassae D. ananassae D. ananassae Tutuila (T)light 22.8 Tutuila (T) light 25.4 Niue 26.4 Niue 29.0 T dark 19.2 T dark 20.3 Tonga 25.8 Aitutaki 24.9 Palmyra (P) 25.5 Palmyra(P) 20.4 Suva (Fiji) 26.1 Rarotonga 29.2 Rarotonga ( R) 21.7 average, stocks 22.5 average, stocks 26.1 average, stocks 27.7 average, crosses 22.6 average, stocks 22.0 average, crosses 18.8 average, crosses 26.1 average, crosses 19.5 D. melanogaster D. simulans Tutuila 25.5 Rarotonga 19.9 D. nasuta D. nasuta Tutuila 18.8 Tutuila 21.1 Palmyra 29.5 Palmyra 19.7 average, stocks 24.2 average, stocks 20.4 average, crosses 20.8 average, crosses 23.4 T ABLE 7 Summary of fertility of young mature pairs of Drosophila in stocks and crosses from all tests March, 1962 July, August, 1962 July, August, 1963 July, August, 1964 Species Percent fertile Species Percent fertile Species Percent fertile Species Percent fertile D. ananassae D. ananassae D. ananassae D. ananassae Tutuila (T) light 18.3 Tutuila (T)light 33.7 Niue P1 44.4 Niue 61.3 Tdark 44.5 Tdark 50.3 F1 57.8 Aitutaki 67.5 Palmyra(P) 57.1 Palmyra (P ) 52.5 F" 57.2 Rarotonga 65.8 T 1 'i' x Tctc; 40.5 Rarotonga ( R) 59.4 Tct X T 1 CJ 80.4 average, Ave., Niue 55.1 average, crJsses 71.8 T 1 'i' X PC) 57.4 all crosses 44.8* P'i' x T 1 c; 12.5 Tonga P1 49.3 Tct X Pc; 59.1 D. nasuta F1 61.7 P'i' X Tct c; 9.4 Tutuila 59.9 F2 66.3 Palmyra 58.4 D. melanogaster P'i'x T c; 76.5 Ave., Tonga 59.3 Tutuila 92.1 T'i'XP CJ 83.3 Suva, Fiji P1 46.9 D. nasuta D. simulans Fi 58.9 Tutuila 46.5 Rarotonga 45.2 F2 69.0 Palmyra 59.5 • 6 back crosses T'i'X Pc; 69.0 between Tutuila Ave., Suva 58.5 P'i' X T c; 33.1 light and dark, average 29.9 Crosses P 1 59.7 Other crosses F1 64.2 omitting T light, F2 75.6 average 52.5 average, crosses 68.7 Stone et al.: Pacific Populations of Drosophila TABLE 8 Summary of egg development from young mature inseminated females. Percent of egg development from these matings Rand01n Random Sibs Sibs Differences Drosophila from random from sibs Cousins from random from sibs ( 1)-(4)Year Island species Column (1) (2) (3) (4) (5) (6) March, 1962 Tutuila ananassae light 87.0 74.5 12.5 ananassae dark 85.2 62.5 22.7 Palmyra ananassae 88.1 85.5 74.6 13.5 Tutuila melanogaster 86.9 78.8 8.1 nasuta 81.4 67.4 14.0 Palmyra nasuta 85.6 83.5 79.8 5.8 July, August, Tutuila ananassae light 88.3 80.2 78.9 65.3 23.0 1962 ananassae dark 87.4 81.3 80.1 69.0 18.4 Palmyra ananassae 87.0 79.1 76.0 71.7 15.3 Rarotonga ananassae 65.9 79.7 56.4 9.5 simulans 87.6 84.7 67.3 20.3 Tutuila nasuta 85.5 82.9 81.9 70.0 15.5 Palmyra nasuta 87.6 84.7 80.0 81.1 6.5 July, August, Niue ananassae 89.4 79.4 88.1 70.1 64.9 19.3 1963 Tonga ananassae 88.7 87.1 83.6 70.7 61.4 18.0 Suva, Fiji ananassae 88.1 82.1 65.9 62.7 22.2 July, August, Niue ananassae 84.3 72.6 63.9 20.4 1964 Aitutaki ananassae 85.5 72.1 61.3 24.2 Rarotonga ananassae 84.4 82.1 63.3 20.7 all ananassae 85.3* 81.3 79.9 66.9t 63 .0 18.4t all nasuta 85.0 83.8 81.8 74.6 10.4 • 86.9 without Rarotonga, 1962 t 67.7 without Rarotonga, 1962 t 19.2 without Rarotonga, 1962 TABLE 9 An estimate of lethal equivalents per gamete from different types of populations Percent egg development Lethal Year Locality Drosophila species and strains Random from random Sibs from random equivalents per gamete* March, 1962 Tutuila ananassae light 87.0 74.5 0.62 ananassae dark 85.2 62.5 1.24 Palmyra ananassae 88.1 74.6 0.67 Tutuila melanogaster 86.9 78.8 0.39 nasuta 81.4 67.4 0.75 Palmyra nasuta: 85.6 79.8 0.28 July, August, 1962 Tutuila ananassae light 88.3 65.3 1.21 ananassae dark 87.4 69.0 0.95 Palmyra ananassae 87.0 71.7 0.77 Rarotonga ananassae 65.9 56.4 0.62 simulans 87.6 67.3 1.05 Tutuila nasuta 85.5 70.0 0.80 Palmyra nasuta 87.6 81.1 0.31 July, August, 1963 Niue ananassae 89.4 70.1 0.97 Tonga ananassae 88.7 70.7 0.91 Suva, Fiji ananassae 88.1 65.9 1.16 The University of Texas Publication TABLE 9-Continued July, August, 1964 Niue ananassae 84.3 63.9 1.11 Aitutaki ananassae 85.5 61.3 1.33 Rarotonga ananassae 84.0 63.3 1.13 Average ananassae 85.3 66.9 0.97 Average without 1962 Rarotonga anrinassae 86.9 67.7 1.00 Average nasuta 85.0 74.6 0.52 1956t Ponape ananassae 88.3 67.6 1.07 Majuro ananassae 80.8 66.4 0.79 Bikini ananassae 75.4 60.3 0.89 1957t Ponape ananassae 87.8 68.8 0.98 Majuro ananassae 74.9 63.7 0.65 Rongelap ananassae 83.6 64.5 1.04 Bikini ananassae 81.8 57.7 1.40 1958t Ponape anrinassae 81.5 63.4 1.10 Majuro ananassae 69.8 50.3 1.31 Rongelap ananassae 73.5 61.2 0.73 Bikini ananassae 74.7 65.1 0.55 1959t Ponape ananassae 80.0 67.6 0.67 Majuro ananassae 74.8 61.0 0.82 Rongelap ananassae 69.1 62.8 0.38 Rongerik ananassae 64.6 60.4 0.27 1953:j: Arizona novamezzcana Patagonia 91.7 86.7 0.22 1962§ Arizona pseudoobscura Rustler's Park 90.4 80.2 0.48 Cave Creek 92.3 81.5 0.50 Madera Canyon 88.5 82.1 0.30 Mount Lemon 89.8 78.3 0.55 • These estimates of lethal equivalents (lethals plus sum of detrimentals) are calculated by the method used by Morton, Crow and Muller (1956) and Greenberg and Crow (1960). The lethal equivalents per gamete are 4 (log• percent egg development from random matings-loge percent development from sib matings), as only Y+ of the loci become homozy­gous in sib matings. +Data from Stone, Wheeler and Wilson, 1962. t Data from Stone, Alexander and Clayton, 1954. § Data from Stone, Wilson and Gerstenberg, 1963. II. The Nearctic and Neotropical Species of Scaptomyza Hardy (Diptera; Drosophilidae) .1 2 MARSHALL R. WHEELER AND HARUO TAKADA The genus Scaptomyza was established by Hardy (1849, Proceedings of the Berwickshire Naturalists Club, 2: 361) for the two nominal species, Drosophila graminum Fallen and Scaptomyza apicalis Hardy. The first of these, graminum Fallen, was chosen as the type species by Coquillett (1910). The current litera­ture now lists 82 species of Scaptomyza for the world but this figure does not include the mass of new species being described from the Hawaiian Islands by Prof. D. E. Hardy (personal communication) . Dr. Walter Hackman (1959), summarizing the species of the world, reported on 65 species; 17 other species were known to him only from the literature. His very excellent account has been the stimulus for this study, providing the necessary background for a comprehen­sive systematic study of the American species. During the last century the taxonomic treatment of the group has been quite varied. Scaptomyza has sometimes been considered a subgenus of Drosophila; it has also been divided into several subgenera, new related genera have been added, and it has been treated as an entity consisting of several species groups or complexes. In his most recent account, Hackman (1959) reorganized the genus into nine subgenera; four of these occur in the Neotropical and Nearctic areas: Scaptomyza s.s., Parascaptomyza, Mesoscaptomyza, and Hemiscaptomyza. We are rearranging somewhat the species to be included in Parascaptomyza and M esoscaptomyza, and we are including a fifth subgenus, Dentiscaptomyza, which contains four South American species. The apparent absence of the subgenus Trogloscaptomyza in the Americas is surprising; there are numerous species of this subgenus in Hawaii (Prof. Hardy, personal communication), although the type species came from the island of Tristan de Cunha in the South Atlantic. Hence, a connection across South Amer­ica would seem logical. We have checked the male genitalia of a number of species from Central and South America which grossly resemble species of Trogloscaptomyza (of an appropriate size, and lacking ventral branches on the arista). Without exception, so far, these have turned out to be co-generic with Cladochaeta nebulosa (but which run through the available keys to Clastoptero­myia Malloch= Diathoneura Duda) and do not seem to be closely related to Scaptomyza. However, there are still several hundred undescribed drosophilids from the Neotropical region and the possibility remains that among them some species referable to Trogloscaptomyza will be found. Much of the past taxonomic work on the genus has been more confusing than helpful, due in large part to the color variability found in many of the species. 1 This investigation was supported, in part, by Public Health Service Research Grant No. GM-11609 from the National Institutes of Health. 2 Present address: Kushiro Women's College, Midoriga-oka, Kushiro, Hokkaido, Japan. The University of Texas Publication Thus, the thorough studies of Hackman (1955, 1959), utilizing male genitalia! features, represent a most important landmark in Scaptomyza taxonomy. We are now convinced that accurate identification of the species of this genus can only be accomplished with the use of such genitalial characteristics; in the pres­ent account, therefore, we are figuring the male genitalia for 54 American species (including 18 new species and 3 unnamed forms). Several of these species had been illustrated earlier by Hackman (1955, 1959) or by Brncic (1955, 1957a), but we believe that it is more practical to redraw them at this time so that all figures of American species would be directly comparable. For two named species, which we have not seen, we have copied the original author's figures. The genitalia dissections and original figures were made by the junior author; Mrs. Linda Kuich assisted in preparing them for publication. Although this study is based primarily on material collected by members of the Genetics Foundation of the University of Texas1 much additional useful ma­terial was loaned to us from the following collections: U.S. National Museum (loan arranged by Dr. Willis Wirth); California Academy of Sciences (arranged by Dr. Paul Arnaud); Centro Nacional de lnvestigaciones Agricolas, Tibaitata, Bogota, Colombia (arranged by Miss Isabel Sanabria); the lnstituto Oswaldo Cruz, Brazil (collection of Prof. Hugo Souza Lopes, arranged by Dr. 0 . Frota­Pessoa); collection of Dr. Danko Brncic, Universidad de Chile, Santiago. In all, about 1400 specimens have been available for this study. Holotypes of the new species are being placed in the Drosophila Type and Ref­erence Collection, Genetics Foundation, University of Texas, Austin, Texas; some paratypes of each species are being deposited in the U.S. National Museum. Vari­ations from this general plan are stated under the individual descriptions. TYPES oF MALE GENITALIA IN ScAPTOMYzA In Figures 1-5 we have illustrated diagrammatically the major features of male genitalia in the genus. The apparent complexity seen in a number of species is due primarily to the development of one or more "extra" lobes bearing either bristles or stout teeth. Our interpretation of the origin of the more common types of extra lobes is shown in Figure 1, which is arranged in a logical sequence which may represent the evolutionary pattern. The simplest form, Type A, is repre­sented by a single species, nigricosta; the primary clasper is plain, and there are no specializations of either the anal plates or the lower genital arch. Although there are a few other species which are nearly this simple, most members of the genus possess some modification of the anal plate. Types B-H of Figure 1 show the varying degrees of development of secondary claspers (paralobes of Frey and others) through specialization of the lower part of the anal plates. In the sub­genus Scaptomyza (Type B) the differentiation is slight; in Type C (Hemi­scaptomyza and some species of Mesoscaptomyza) the specialization is quite obvious; while in Type D (many species of Mesoscaptomyza) and Type E (Para­scaptomyza) we find the most extreme development, with one to several promi­nent spines adorning the tip. t Field work supported by National Science Foundation grants G-1653 and G-4999. Frn. 1. Types of external male genitalia in Scaptomyza. A:-nigricosta-type; B:-montana­type; C:-terminalis-type; D:-vittata-type; E:-pallida-type; F:-paradusta-type; G:­wheeleri-type; H :-denticauda-type. Three different appearances of the secondary clasper are also shown in Types F-H, but in these there is also shown the development of a specialized portion of the lower part ("toe") of the genital arch. In Type F (paradusta type) and Type G (some species of Mesoscaptomyza) this specialization is only moderate, but in species of the subgenus Dentiscaptomyza. (Type H) the lower part of the arch becomes a tertiary clasper, forming a large and often spectacular part of the external genital apparatus. Still other lobes, with bristles and/ or differentiated teeth, may occur in particular species, for example S. Parascaptomyza clavifera. Figure 2 illustrates the three basic hypandrium types in the genus; the arrows indicate that lateral processes of the hypandrium may develop at these points in varying degree. The major types of such processes are shown in Figure 3, the arrows indicating the four major types (B-D) which seem to have been derived from the basic Type A. The anterior gonopophyses exhibit different relationships to the hypandrium; our interpretation of the major types is illustrated in Figure 4, from a front view and in a hypothetical sectional view. One can represent, symbolically, the general features of the male genitalia for any species of Scaptomyza. by referring to the specific "types" from each of these first five figures; for example, paradusta can be represented as "FBBDG", while intermedia is "HCDCD." Certain combinations occur in several species, and some species are difficult to assign precisely. However, we had hoped that certain The University of Texas Publication :r: w 0 0 m C\J QI .... ::J o> LL.. FIG. 2. Types of hypandrium in Scaptomyza. A:-nigricosta-type; B:-terminalis-type; C:­parandina-type. FIG. 3. Types of lateral processes of the hypandrium in Scaptomyza. A:-montana-type; B: -paravittata-type; B1 -nigricosta-type; :-pallida-type; B3 :--clavifera-type; C:-teinop­ B2 tP.ra-type; D:-denticauda-type; E:-macroptera-type. (Continued on next page) TABLE 1 The morphological types of male genitalia in the subgenera of Scaptomyza, based upon the traits shown in Figures 1-5 Fig.1 Fig. 2 Fig. 3 Fig. 4 Fig. 5 sg. Scaptomyza B ABC ABB1C CD CDEF sg. Mesoscaptomyza nigricosta-type A A B1 B c paravittata-type c B BB1 CD ABC vittata-type D B BB1 CD ABC wheeleri-type G B BB1 A A sg. Hemiscaptomyza c B B1 DE c sg. Parascaptomyza pallida-type E B B1B2B3 D CDH paradusta-type sg. Dentiscaptomyza F H B BC DE ABD DE CDE AG l D clusters of characters would serve to differentiate between the various subgenera; Table I shows that this is not true, although this ideal is nearly reached by the assignment of species on the basis of the type of external genital apparatus (Fig­ure 1). Key to American Subgenera of Scaptomyza ( 1) Acrostichal hairs distinctly 2-rowed, both between the dorsocentral rows and anteriorly ------------------------------·--···--·---------·-·-···-----------------------------------·-·· 2 Acrostichal hair 4-rowed anteriorly, with 2 or more irregular rows posteriorly between dorsocentrals ----·-·--··----·-····-----------···---···-··--·-·-···--·----4 (2) Palpi dark brown; posterior scutellar bristles rather short, and turned upright; presutural dorsocentral bristles often more or less evident ------------------------------------------------------------------------------·· Mesoscaptomyza Palpi pale; apical scutellars normally long and directed posteriorly; presutural bristles not common --------------···--····--·--·----------·-··-···--·-·------------3 (3) Two s•rong humeral bristles; usually dark species; male genitalia of type B (Fig. 1) ; 9 ovipositor plates stout and bearing coarse teeth ------------------------------·-·····--···-··--··-··----·-··-··--··-----------Scaptomyza (part) One strong humeral; genitalia of Type E; of S. Parascaptomyza spinipalpis (.8-.9) ; wing apex of S. Hemiscaptomyza apicipuncta (.10), redrawn from Malloch; wing pattern of Drosophila dila­cerata Becker, redrawn from Becker ( .11). Types and Distribution: Holotype male, allotype and 9 paratypes, Cerro Monte Cristo, El Salvador, 7000 feet, Feb. 1954, W. B. Heed; 3 paratypes, Volcan Boqueron, El Salvador, 4500 feet, Feb. 1954, W. B. Heed; in addition there is one female which probably belongs with this species: Boquete, Chiriqui Prov., Panama, June-July 1959, W. B. Heed and H. L. Carson. 41. S. (Mesoscaptomyza) samurai, new species i1l . Front tan, becoming pale on orbits; antennae light brown; face brownish, becoming pale near eyes; cheeks pale; 1st oral bristle stout, 2nd thin and weak; arista 4/ 2. Mesonotum with three brownish longitudinal stripes, not well defined; median stripe continued broadly onto scutellum; pleural stripe dark brown to black, very prominent; haltere base brown, the knob tan. Legs pale; fore coxae adorned with a series of stout bristles apically (Fig. 19.6). Abdomen almost uni­formly dull brown. Wings (Fig. 19.7) with a striking pattern of dark areas; spur The University of Texas Publication veins occur in the large dark area near the apex of the 2nd longitudinal veirt, and in the dark spot apically on the 4th veirt. Female unknown. Male genitalia: Fig. 19.1-.5, drawn from a paratype. Types and Distribution: Holotype male and two male paratypes, from S.E. of Bogota, Colombia, 10,000 feet, July 1960, W. B. Heed and H. L. Carson. 42. S. (Mesoscaptornyza) setosa, new species o, c;>. Front dull tan, the orbits dark brown and well differentiated; ocellar area dark; 2nd antennal joint dark tan to brown, 3rd joint lighter; face tan, often darker on carina; cheeks rather broad, pale; palpi dark; one prominent oral bristle; arista 3/2 or 4/2. Head and thoracic bristles longer and more conspicuous than usual. Mesonotum dark brown to black, subshining, some stripirtg only faintly indicated on some irtdividuals as paler stripes just irtside dorsocentral lilies; scutellum darker in midline, paler on sides. Upper part of pleura about as dark as mesonotum, becoming pale tan below; legs all pale; halteres pale; wirtgs clear. Abdomen mostly dark brown to black, subshirtirtg, but with the series of broad pale median areas united to form a broad pale midline stripe. Male genitalia: Fig. 18.6-.10, drawn from a paratype. Types and Distribution: Holotype male, allotype, and 13 paratypes, Chilicay, Chimborazo, Ecuador, July 1955, Levi Castillo, collector; all types are the prop­erty of the U.S. National Museum and are beirtg returned to that collection. 43. S. (Mesoscaptornyza) striaticeps, new species o, c;> • Front pale yellow to tan, with a broad blackish stripe from ocelli to anterior margirt; 2nd antennal joint dark brown, 3rd joint very pale; face and cheeks pale; arista 4/2 or 5/2; 2nd oral bristle weak. Mesonotum with three prominent dark longitudirtal stripes; the central stripe extends to the end of the scutellum and the lateral stripes extend onto the sides of the scutellum. A sirtgle dark pleural stripe, large and prominent, contirtued to postscutellum. Base of haltere dark, the knob white. Legs pale; wirtgs clear, the 3C black bristles on the basal 1/5 or 1/6. Abdominal tergites shiny black with a dorsal pattern of pale yellow areas except on the last segment. Male genitalia: Fig. 18.11-.15, drawn from a paratype from Bogota. Types and Distribution: Holotype male, allotype and two paratypes, Tequen­dama Falls area near Bogota, Colombia, Feb. 1958, M. R. Wheeler; 7 paratypes from the same locality, July 1960, W. B. Heed and H. L. Carson; 2 paratypes, Bogota vicinity, 8700 feet, Nov. 1955, W. B. Heed; one paratype, Fusagasuga, S.W. of Bogota, Colombia, Nov. 1955, W. B. Heed. There is irt addition one specimen, badly damaged, which may belong here; it is labelled "Argentina; Jose C. Paz; Ogloblin; 2-X-39" (from Inst. Oswaldo Cruz collection). 44. S. (Mesoscaptornyza) subvittata Hackman Scaptomyza subvittata Hackman 1959: 50. Type from San Jose, Cost Rica. Male genitalia: Fig. 16.19, redrawn from Hackman's original figure of the genital arch and clasper. We have not seen this species; it is said to be very similar in appearance to paravittata. Distribution: Costa Rica. 45. S. (Mesoscaptomyza) vittata (Coquillett) Drosophila vittata Coquillett 1895: 318: Type locality: Charlotte Harbor, Florida. Male genitalia: Fig. 15.1-.5, drawn from a specimen from Coroico, Bolivia, and checked with specimens from Colombia, Ecuador, and Louisiana. Distribution: A widely distributed species; Florida to Louisiana; Cuba to Ja­maica and Puerto Rico; Mexico to Parrama to Colombia to Bolivia and Peru. 46. S. (Mesoscaptomyza) wheeleri Hackman Scaptomyza wheeleri Hackman 1959: 49. Types from Falls Church, Virginia. Male genitalia: Fig. 16.6-.10, drawn from a specimen from Medellin, Colom­bia, and checked with males from El Salvador, and from Great Smoky Mountain National Park, Tennessee. For comparison with the similar coquilletti, the female ovipositor is shown in Fig. 16.17. As was stated under coquilletti, it and wheeleri are not readily separated except by the use of the male genitalia. There is also some overlap in the respective distri­butions but, in general, wheeleri ranges widely on the mainland (Virginia to perhaps Peru) while coquilletti is primarily found on the islands of the Caribbean and in some localities in Central America. Distribution: Tennessee and Virginia south to Mexico to Colombia and Ecua­dor. Probably occurring also in Peru and Bolivia (Hackman 1959). 47. S. (Mesoscaptomyza) species b This undescribed species is known to us by a single male from Cerro de la Muerte, Cost Rica. The male genitalia are illustrated in Fig. 18.16-.20. This specimen shows almost no midfrontal stripe, the face is whitish, the vibrissa is single and strong, the mesonotal striping is moderately pronounced, the pleural stripe is typical and well-developed, the haltere knob seems to be somewhat dis­colored, and the legs and wings are pale. 48. S. (Mesoscaptomyza) species c This undescribed species is knawn from a single male from Pena de Gato, Puebla, Mexico. The genital arch was illustrated by Hsu (1949:90) under the misidentification, vittata. The species has not been recaptured, so we are redraw­ing Hsu's original figure as Fig. 16.20. Key to species of Mesoscaptomyza 1. Mesonotum with three rather prominent dark brown longitudinal stripes ............................................................................................... ,........ Mesonotum dark brown to blackish, without stripes or with faintly indicated paler stripes ................................................. ,,,.......................... 3 2 The University of Texas Publication 2. Pleura all dark, or at least darker below than it is above; wing base_ a~d costal margin black; abdomen black ............................................ nzgrzcosta Pleura dark above, pale below; wings pale; abdomen with some light areas mid-dorsally ···············-····--·-··-····-····-··············-·····-····-···-············· setosa 3. Wings without distinctive apical clouds ··---·-····························-················ ~ Wings with a series of dark spots apically ······-·-·····-···-·-····-·---·········· samuraz 4. Abdominal tergites with a pattern of light and dark areas ····---···-···-·····-··· 6 Abdomen uniformly dark or nearly so -··-····················---·····-·--·········-··-·····-· 5 5. Wings clear; abdomen dull dark brown ··········--···--··-·-····-···--················ dankoi Wing base and costal cell dark, marginal cell slightly darkened; abdo~en . semishining reddish brown ····················-·--···········-········-····-·-·-·-fuscznervzs 6. Vibrissa single and strong, the Q.nd oral bristle thin and usually not half as long as 1st; pleura usually with a single dark brown stripe above.... 9 Two relatively thin and weak oral bristles on each side; pleural stripe indistinct or tending to form two stripes ---································-········-····· 7 7. Male face pale, often whitish ···----·······························································-· 8 Male face with a dark median stripe; pleural stripe tending to bifurcate into two stripes posteriorly ····················································--·-········ bogotae 8. Pleura of male with two distinct stripes .................................... pleurolineata Pleural stripes rather indistinct .................................... vittata; pseudovittata 9. Male face with a dark median stripe ······--············-···-·-··-···········--····· personata Male face uniformly pale, often whitish ·······-··---------·-·-··--·--·------·--·······-·--· 10 10. Front pale tan to orange in middle·····--···--··--wheeleri; coquilletti; subvittata Front usually dark brown medianly, sometimes conspicuously so········---­·-···---···········-····-·········-········· paravittata; salvadorae; striaticeps; nigripalpis Subgenus PARASCAPTOMYZA Duda Type species: Scaptomym pallida (Zetterstedt). 49. S. (Parascaptomyza) adusta (Loew) Drosophila adusta Loew 1862: 231. Type locality: Washington, D.C. Male genitalia: Fig. 13.5-.9, drawn from specimens from Austin, Texas, and compared with males from Mexico, Costa Rica, El Salvador, Colombia and Argentina. Distribution: Very widely distributed in North America from British Colum­bia and Washington to Ontario and Maine, south to Florida and Arizona (Chiricahua Mts.); Mexico to Colombia to Argentina. Hackman (1959) reported specimens from Bermuda; this is the only authenticated record from the Carib­bean islands since the earlier report of this species on Puerto Rico (Townsend and Wheeler 1955 ) cannot now be confirmed due to a lack of specimens. 50. S. (Parascaptomyza) clavifera, new species i, ~.Frontal orbits and ocellar triangle dark brown, subshining; rest of front dull brown becoming tan to orange anteriorly; antennae yellow to tan; face pale, almost white, on males but with brownish carina on females; cheeks and palpi pale; two nearly equal oral bristles on each side; arista usually 4/2. Mesonotum dark brown, evenly pollinose and usually with no visible evidence of longitud­inal stripes; scutellum all dark. Pleura entirely dark brown, subshining; haltere base dark, the knob whitish; legs pale, wings clear. Abdomen dark brown, rather shining posteriorly. Male genitalia: Fig. 14.1-.4, drawn from a paratype. Types and Distribution: Holotype male, allotype, and 5 paratypes, Cuzco, Peru, February-March 1956, DankoBrncic, collector. 51. S. (Parascaptomyza) macroptera, new species This species seems to occur in a dark form and a light form; the holotype comes from the light form. Dark form ( !i? !i? only). Front gray pollinose behind, tan anteriorly, the large ocellar triangle becoming black anteriorly; antennae pale; arista with two dorsal, no ventral, branches; face tan, brownish in midline; cheeks tan; clypeus dark; palpi pale; 1st oral stout, 2nd small. Mesonotum dull grayish brown with three poorly defined longitudinal stripes; pleura dark brown; legs pale, with the termi­nal two tarsal joints darker. Abdomen dull, dark brown, subshining on last seg­ment. Wings slightly brownish, posterior crossvein darkened. Light form ( 5, !i?). Front tan, lightly browned in midline and in ocellar area; face entirely pale ( 5 5 ) or with a narrow dark line in middle ( !i? ) ; palpi, cheeks and clypeus pale; pleura mostly pale, a darker dorsal stripe only indicated; prosternum blackish near posterior edge, variable. Mesonotum tan, only a median longitudinal stripe evident, this continued onto scutellum. Wings clear. Other characteristics about as in the dark form. Male genitalia: Fig. 14.5-.9, drawn from the holotype. Types and Distribution: All specimens but one have the same collection data: Cerro de la Muerte, Costa Rica, July-August, 1956, W. B. Heed, H. L. Carson, and M. Wasserman. Holotype male, allotype, and one paratype of the light form; 4 paratype females of the dark form. One female paratype, dark form, Volcan Irazu, Costa Rica, 11,000 feet, "off mushroom," July 1956; collectors as above. The arista, with two dorsal branches and none ventrally, is unique for Ameri­can species, and suggests the subgenus Trogloscaptomyza (see Hackman 1959: 10, Fig. 11 for S. T. striatifrons Hackman). 52. S. {Parascaptomyza) pallida (Zetterstedt) Drosophila pallida Zetterstedt 1847:2571. Type from Europe. Male genitalia: Fig. 13.1-.4, drawn from specimens from Chichi Jima, Bonin Islands. This is the common American species which has been referred to by most American authors as "graminum Fallen" or "disticha Duda." Distribution: Worldwide and rather common except in the colder northern and southern extremes; not known from Alaska nor from the southern parts of South America. The University of Texas Publication 53. S. (Parascaptomyza) paradusta Wheeler Scaptomyza paradusta Wheeler 1952: 198. Type locality: Mt. San Jacinto, California. Male genitalia: Fig. 13.10-.14, drawn from a male from Oregon, and compared with a paratype. Distribution: California (numerous localities); Oregon (Saddleback Mt., Lin­coln Co., J. C. Dirks-Edmunds coll.); Washington (Mt. Constitution); Arizona (Oak Creek Canyon) . 54. S. (Parascaptomyza) picifemorata Hackman Scaptomyza picifemorata Hackman 1959: 45. Type from Ecuador. Male genitalia: Fig. 14.10-.13, drawn from specimens from Bogota, Colombia and checked with males from Volcan Irazu, Costa Rica, and Cuzco, Peru. Distribution: Costa Rica (Vo lean Irazu) ; Colombia (various localities, some up to 10,000 feet); Ecuador (in Hackman 1959); Venezuela (pass on Pacific­Atlantic divide, 8-9000 feet, Estado Trujillo) ; Peru (Cuzco). The dark femora, mentioned by Hackman, are not characteristic; of about 90 specimens which we have seen, only 14 show the trait. 55. S. (Parascaptomyza) spinipalpis Seguy Scaptomyza spinipalpis Seguy 1934: 11. Type, male, from La Plata, Argentina. Male genitalia: Fig. 13.15-.19, drawn from a specimen from Coroico, Bolivia and compared with a specimen from Ogloblin, Argentina. This species has apparently not been reported since its description. We have seen two males from what appears to be the type locality, loaned by Dr. Blanch­ard of Buenos Aires; in addition we have three other specimens from Argen­tina, four from Bolivia and eight from Colombia. The markings of the wings are distinctive, as is shown in Fig. 19.9 (male) and 19.8 (female). The unusually large apical wing spot of the male is produced by a fusion, more or less complete, of the large spot apically on the 3rd vein with the smaller one apically on the 4th vein. On one male from Bolivia, the spot of the 4th vein is lacking, and this is essentially so also on a male from Colombia. Females have only the spot on the 3rd vein. We have also noted that the speci­mens from Bolivia are decidedly darker than those from elsewhere, but a check of the male genitalia shows that the same species is involved. Distribution: Argentina, Bolivia, Colombia. Our list of specimens examined is as follows: two, La Plata, Argentina, V-VI-1935, R. Costa (loaned by Dr. Blanch­ard); three, Argentina, Ogloblin, Jose C. Paz, 24-VIII-39 and 10-IX-39 (loaned by Dr. Souza Lopes, Inst. Oswaldo Cruz, Brazil); four, Coroico, Bolivia, April 1958, M. Wasserman; eight, Bello, Ant., Colombia, C. Rios, 27-VI-55, VI-16-55, and 21-VIII-56, all but one labelled "Frijol" indicating that they were reared from or captured on bean plants (collection of CNIA, Tibaitata, Bogota, the loan arranged by Miss Isabel Sanabria). . OTHER SPECIES We have seen a single female of an undescribed species belonging to the "andina complex." The mesonotal pattern is strong, consisting of prominent brown and bluish-gray stripes; the mid-facial stripe is broad and very black; the palpi are black; and the fore coxae and all femora are black. This specimen, like others of the andina complex, came from near Bogota, Colombia, in the paramos above the city, and was collected by W. B. Heed and H. L. Carson. Hackman (1959:66, Fig. 28) mentions an unidentified species, probably be­longing to the subgenus Scaptomyza, from Cerra Punta, Panama, (U.S.N.M. collection) . We have not recognized the species among our material. The two individuals studied by Hackman were said to be in very poor condition, and the external features could not he given with any certainty. The single character which sounds highly significant is an unusual one for this genus: dark fore tibiae. Drosophila dilacerata Becker (1919), from Ecuador, has not been reported since its description. Duda (1927:217) expressed the opinion that it was most likely a Scaptomyza. In our opinion, this is most unlikely. Becker (op. cit.) made a point of commenting on the reality of separating Scaptomyza from Drosophila on the presence of only two or four acrostichal rows in the former; thus it seems safe to infer that dilacerata must have more than four rows of acrostichal hairs, and this is not true for any known species of American Scaptomyza (although it is common among Hawaiian species, for example) . The wing pattern figured for dilacerata seems unique; we are reproducing it as Fig. 19.11. REFERENCES Basden, E. B. 1961. Type collections of Drosophilidae (Diptera). 1. The Strobl collection. Beitr. z. Ent. 11: 160-224. Becker, T . 1919. Mission du serv. Geogr. de l'Annee pour la mesure d'un Arc de Meridien Equat. en Emerique du Sud. 10(2), Diptera, Brachyceres: 163-215; Pl. XVII. Brncic, D. 1955. The Chilean species of Scaptomyza Hardy (Diptera, Drosophilidae). Rev. Chilena de Ent. 4: 237-250. ----. 1957a. Las especies Chilenas de Drosophilidae. Monograf. Biol. 8, Univ. Chile, Santiago. pp. 136. ----. 1957b. Los insectos de las Islas Juan Fernandez. 31. Drosophilidae (Diptera). Rev. Chilena de Ent. 5: 391-397. and T. Dobzhansky. 1957. The southernmost Drosophilidae. Amer. Nat. 91 (857): 127-128. Collin, J. E. 1953. On the British species of Scaptomyza Hardy and Parascaptomy za Duda (Dipt., Drosophilidae). The Entomologist 86: 148-151. Coquillett, D. W . 1910. The type-species of the North American genera of Diptera. Proc. U.S.N.M. 37(1719): 499-647. Duda, 0 . 1921 . Kritische Bemerkungen zur Gattung Scaptomyza Hardy (Dipteren). Vereins fiir schlesische Insektenkunde, Breslau, 13: 57-69 (pp. 1-13 in reprint version). ----. 1924. Beitrag zur Systematik der Drosophiliden unter besonderer Beriicksichti­gung der paliiarktischen u. orientalischen Arten (Dipteren). Arch. f. Naturg. 90A3: 172-234. ----. 1927. Die sudamerikanischen Drosophiliden unter Beriicksichtigung auch der anderen neotropischen sowie der nearktischen Arten. Arch. f. Naturg. 91A11-12(1925) : 1-228. The University of Texas Publication Hackman, W. 1955. On the genera Scaptomyza Hardy and Parascaptomyza Duda (Dipt., Drosophilidae). Notul. Ent. 35: 74-91. 1959. On the genus Scaptomyza Hardy (Dipt., Drosophilidae). Acta Zool. Fen­nica 97: 1-73. Hsu, T. C. 1949. The external genital apparatus of male Drosophilidae in relation to system­atics. Univ. Texas Puhl. 4920: 80-142. Malloch, J. R. 1924. Descriptions of neotropical two-winged flies of the family Drosophilidae. Proc. U.S.N.M. 66 (3): 1-11. Segtiy, E. 1934. Etude sur quelques Muscides de l'Amerique Latine. Rev. Soc. Ent. Argentina 6: 9-16. Slosson, A. T. 1902. Additional list of insects taken in alpine region of Mt. Washington. Ent. News 13: 4-8. Takada, H. 1965. Differentiation of the external male genitalia in the Drosophilidae. Kushiro Women's College Puhl. 1: 39-50 (in Japanese; English summary pp. 131-132). Townsend, J. I., and M. R. Wheeler. 1955. Notes on Puerto Rican Drosophilidae including descriptions of two new species of Drosophila. Jciur. Agr. Univ. P. Rico 39: 57-64. Wheeler, M. R. 1949. Taxonomic studies on the Drosophilidae. Univ. Texas Puhl. 4920: 157-195. 1952. The Drosophilidae of the nearctic region, exclusive of the genus Drosophila. Univ. Texas Puhl. 5204: 162-218. ----, and L. H. Throckmorton. 1960. Notes on Alaskan Drosophilidae (Diptera), with the description of a new species. Bull. Brooklyn Ent. Soc. 55: 134-143. III. A Study of Speciation in South Pacific Populations of Drosophila ananassae1 DAVID G. FUTCH INTRODUCTION Students of organic evolution have long recognized the significance that small isolated populations may have with respect to the process of speciation, The con­siderable phenotypic diversity which is often found when peripheral populations of an otherwise relatively uniform species are compared is testimony to why this is so. Carson (1959) has described how such diversity might be expected even­tually to lead to the formation of new species. Such conditions are frequently found among populations of organisms resident on scattered oceanic islands, isolated from each other by miles of water (see Heed, 1962). A situation such as this suggests a very intriguing question. What will happen if two differentiating island populations of this kind renew geographic contact? Will they maintain their isolation, or will they interbreed freely and merge to form a single popula­tion? This question has been answered for one specific case by a pair of sympatric, interfertile, and yet, apparently, reproductively isolated island races of Droso­phila ananassae. Drosophila ananassae Doleschall, a member of the melanogaster species group of the subgenus Sophophora of the genus Drosophila, has a geographic distribu­tion which has been called circum-tropical (Stone, Wheeler, Spencer, Wilson, Neuenschwander, Gregg, Seecof, and Ward, 1957), occurring in the tropical and subtropical parts of the world, and sometimes ranging into slightly more temper­ate climates. In many areas it is one of the most common of the Drosophila species present, especially in and around places of human habitation where it is attracted to fermenting fruit, decaying vegetable matter, and, according to Sturte­vant (1916), sometimes excrement. Over much of this range the species seems to be fairly continuous, even to the extent that populations separated by major geographic barriers, such as oceans and mountains, may experience some gene exchange. Conceivably this could be accomplished by utilizing the close association between the species and man as a means of dispersal. Indeed, as Dobzhansky and Dreyfus (1943) have pointed out, D. ananassae probably originated in the vicinity of southeastern Asia and has depended on man for its present widespread distribution. They cite as evidence for this belief the fact that all but three other members of the melanogaster species group are restricted to this part of the world plus the common occurrence of the same three distinct chromosomal rearrangements in significant frequencies in D. ananassae populations at various places throughout its range. Of the three 1 This investigation was supported, in part, by Public Health Service Research Grant GM-11609 and by Public Health Service Training Grant 5 Tt-GM-337 from the National Institutes of Health. The University of Texas Publication other species not restricted to Asia and the outlying Pacific Islands, two, D. melanogaster and D. simulans, are cosmopolitan and the third, D. kikkawai, is likewise a widely distributed species, being found in Asia, the Pacific Islands, Australia, and Central and South America. Thus, over much of its range, D. ananassae exists in large, stable, heterozygous, frequently domestic populations. There are, however, areas in which the species exists in isolated populations of various sizes. Among the tropical islands of the Pacific Ocean, such isolated populations may be very small and very unstable, being subjected to the pressures of a sometimes harsh and drastically changing environment. Such populations occur on small coral atolls such as those in the northern Marshall Islands of eastern Micronesia (Stone, et al., 195 7; Stone, Wheeler, and Wilson, 1962). On these particular islands, ananassae is the only Drosophila species known to occur. On many of the larger Pacific Islands, how­ever, where environmental conditions are more favorable, a number of other species may also be found, although ananassae is frequently the dominant or, at least, one of the more numerous of those present. This is especially true of the fertile high volcanic islands of some of the major island groups in the Pacific Ocean. Islands such as these commonly host, in addition to the widespread species like ananassae, a few to many endemic species of Drosophila. In the spring and summer of 1962, Dr. M. R. Wheeler and Dr. W. S. Stone visited certain Polynesian Islands of the central equatorial and South Pacific Ocean for the purpose of collecting Drosophila. As might be expected, ananassae was found to be a common and fairly numerous inhabitant. An interesting and obvious characteristic was immediately noted concerning these flies which indi­cated a difference between them and individuals of the same species from other areas, in particular the afore-mentioned populations of the Marshall Islands. Whereas aruznassae is normally thought of as being relatively light in color, the Polynesian flies were found to be quite dark. This darkness is especially apparent as it is expressed in the pigmentation of the dorsal part of the abdomen, which is dull brownish-black in color. One of the places visited during these collecting trips was the rather small high island of Tutuila in American or eastern Samoa. A peculiar thing with respect to the anaruzssae population on this island was discovered. Not only was the dark Polynesian form present in large numbers, but, in addition, a light, yellow form also fitting the description of the species was collected. In a number of instances both dark and light individuals were collected at the same time from the same places. Interestingly no intermediate types were detected in these col­lections as might be expected in view of what has been found to be the case in otherwise similarly polymorphic populations of animals. Initial laboratory tests indicated that this was probably not merely a case of genetic polymorphism, for, although none was found in the natural populations on the island, numerous fertile intermediate individuals could be produced in the laboratory by crossing light and dark flies. Moreover, there were consistent struc­tural differences detectable in the salivary gland chromosomes distinguishing the two forms from each other. Thus, it seemed that the dark and light forms on Tutuila might represent geographically sympatric, interfertile, but genetically Futch: Drosophila ananassae Populations isolated populations which, though phenotypically different, were nevertheless both identifiable as Drosophila ananassae. These seemingly paradoxical observations have led to the work which is to be discussed in the following pages. Cytological analyses, crossing experiments, and mating discrimination tests involving the two varieties from Tutuila, plus, in some instances, stocks selected to represent other geographical populations of ananassae were made in an attempt to determine whether the two forms are really isolated from each other and, if they are, how this isolation is maintained. RELATIONSHIP OF COLOR TYPES AND GEOGRAPHIC DISTRIBUTION As a general background against which to compare the two Samoan:· color types, a survey of all of the ananassae stocks maintained at the University of Texas Genetics Foundation was made to determine how much phenotypic variation of abdominal pigmentation there may be and how this may be related to the geo­graphic distribution of the species. Although limited by a lack of material from certain areas--e.g., Japan, China, and certain other parts of Asia-this survey did, nevertheless, reveal that a considerable amount of color variation does occur within the species, ranging from very pale yellow to nearly black. The interesting thing is that these color types can be seen to exist as geographic races. Thus, the stocks which originated from Asia and the western Pacific-i.e., southern India, New Guinea, and the Truk Islands of the central Caroline Islands of Micronesia-are dull yellow. Flies from the Marshall Islands to the northeast, while also yellow, are somewhat lighter and paler. With the exception of the light stock from Tutuila, all the ananassae stocks from islands to the southeast of New Guinea beginning with the Fijian Islands-i.e., Viti Levu (Fijian Islands), Tongatapu (Tongan Islands), Niue Island, Tutuila (Samoan Islands), and Rarotonga and Aitutaki (southern Cook Islands)-are very dark. The stocks from Oahu (Hawaiian Islands) and Palmyra, one of the Line Islands located south of Hawaii and slightly north of the Equator, are dark gray. All of the stocks from South and Central America and the Caribbean--e.g., Brazil, Panama, Mexico, and Cuba-are brownish yellow. Except for being somewhat lighter, the American flies are very similar to those from Hawaii and Palmyra. The variation in degree of pigmentation is accompanied by variation in the pattern of abdomin·al pigmentation. An indistinct, somewhat darker transverse band of pigment marks the posterior margin of each tergite of the flies, especially females, from all of the stocks. The light Samoan stock differs in this respect. Among its females these bands are quite distinct and contrast strongly with the yellow background. Females from the dark stocks of Polynesia have a noticeable tendency to grow darker with age. This darkening can be seen to be concentrated along the dorsal midline of each tergite. The darker the fly the more "filled" with dark pigment the lateral areas of the tergites become. In females of the darkest types (i.e., Samoa and Tonga), tergites are often almost completely pigmented. EXPERIMENTAL STOCKS AND GENERAL METHODS Table 1 lists all stocks used experimentally, collection localities, University of The University of Texas Publication TABLE 1 List of the stocks used in this investigation, their geographic origin, abdominal pigmentation, University of Texas stock number, and the stock symbols used in the tests Locality Pigmentation Stock No. Symbol Pago Pago, Tutuila, American Samoa Light (yellow) 3038.1 L Pago Pago, Tutuila, American Samoa Dark (black) 3038.2 D Niue Island Dark (black) Niue Merida, Yucatan, Mexico Intermediate H63.1 Mexico (brown) Majuro, Marshall Islands Very light 2370.9 Maj (yellow) Truk, Caroline Islands Light (yellow) 2210.3 Truk Brown River, Port Moresby, Papua, New Guinea Light (yellow) 3020.8 NG-1 (21cc) Popondetta, Papua, New Guinea Light (yellow) 3021.2 NG-2 (22jj) Rarotonga, Cook Islands Dark (black) 3036 Rarotonga Palmyra Ishnd Dark (grey) 3034.1 Palmyra Honolulu, Oahu, Hawaii Dark (grey) 2370.11 Hawaii Santiago de Cuba, Cuba Intermediate 2387.4 Cuba (brown) Texas stock numbers, and symbols used in tests. In addition, the abdominal pig­mentation type of each stock is given. The dark Samoan stock was established by mixing the offspring of thirty pair-matings of dark flies. The pairs which produced this pooled stock were from crosses involving non-inbred third generation flies from the original collection. The light Samoan stock was established in the same manner from thirty fertile pair-matings of light flies. The culture medium used both for maintaining stocks and making crosses consisted of cornmeal, agar, malt, yeast, and water with propionic acid added as a mold inhibitor. Prior to its use the medium was inoculated with Wagner's Y-2 strain yeast cells (Wagner, 1944). Crosses were usually set up as groups of 120 or more individual pair-matings in small shell vials. Transfers to fresh food were made at seven day intervals. Only virgin adults, males and females, aged for five days following eclosion, were used for crosses and behavior tests. In some experiments females were checked for insemination. This was done by dissecting out the spermathecae and ventral receptacle and examining them microscopically for the presence or absence of sperm. Salivary gland chromosomes were prepared for examination by passing the glands quickly through ( 1) forty-five percent (glacial) acetic acid and (2) dis­tilled water. The chromosomes were stained and spread in lacto-aceto-orcein stain consisting of two percent orcein in equal parts of glacial acetic acid and lactic acid. The technique for preparing larval neuroblast chromosomes involved (1) a Futch: Drosophila ananassae Populations ten minute pretreatment of the anterior dorsal and ventral ganglia in hypotonic sodium citrate solution ( 0.35 gms sodium citrate : 50 ml distilled water) in order to produce chromosome swelling, (2) a five minute staining period in alcohol­aceto-orcein stain (three percent orcein in seventy percent glacial acetic acid and thirty percent ethyl alcohol), and ( 3) a second five minute period prior to squash­ing in this same stain supplemented by an equal amount of the latco-aceto-orcein stain used for salivary gland chromosomes. CYTOLOGY Table 2 lists each of the twenty-one paracentric inversions, three pericentric inversions, and one translocation found during the course of this study. The break-points for each rearrangement, as determined with respect to the map (Figure 1), and the symbol chosen to designate the rearrangement are also given in this table. Throughout this paper, these symbols will be used in discussions involving rearrangements. No differences could be detected between Samoan light and dark flies with respect to the neuroblast chromosomes. The metaphase karyotypes of both are the same as the karyotype of ananassae as it as been described by Kaufmann (1937) and Kikkawa (1938). There are two pairs of large V-shaped metacentric auto­somes, a pair of small V-shaped metacentric autosomes (corresponding to the dot chromosome in most other species of Drosophila), and a pair of medium size V-shaped metacentric sex chromosomes in the female. In the male, one of the V-shaped X-chromosomes is replaced by a I-shaped Y-chromosome (see Fig­ure 1). Salivary gland chromosome slides were prepared from larvae produced by the thirty pairs of light and thirty pairs of dark flies, from which the pooled stocks were derived. The practice was to make slides of at least four female larvae from each pair. Basically, the two forms, light and dark, possess the same salivary gland karyotype of six chromosome strands radiating from a darkly stained chro­mocenter. This is the ananassae arrangement as reported by Kaufmann (1937), Kikkawa (1938), and Seecof (Stone, et al., 1957). As Seecof's salivary gland chromosome map for the species indicates (Figure 1), there are two short strands representing the left and right arms of the paired X-chromosomes and four much longer strands representing the left and right arms of the two pairs of large auto­somes, chromosomes 2 and 3. The two small autosomes which are largely hetero­chromatic (Kaufmann, 1937; Kikkawa, 1938) are buried in the chromocenter. As these two authors have shown in ananassae, this chromosome, chromosome 4, is the result of a translocation of a part of the heterochromatic region of the X­chromosome to the small dot chromosome. The cytological similarity between the two forms ends with these rudimentary observations. No significant amount of inversion polymorphism was detected in the dark form while the light form was frequently found to be heterozygous' for a set of five paracentric inversions in chromosome 2. Two of these five inversions form a pair of medium length subterminal overlapping inversions in the right arm of 2, (2RA;2RB); two of the other three form a pair of subterminal over­ ~ TABLE2 List of rearrangements by chromosome, including break-points and alternate designations Chromosome XL XH 2L 2R Paracentric Inversions 1. XLA (5.7-16.0) XHA (2.3-8.5 ) 2LA [CIIL]1 (3.1-30.5) 2. XLB (16.7-20.3) 2LB [CIIL-0] 1 (30.3-34.4) 3. 2LC [h]2 (2.6-12.8) 4. 2LD ( 10.4-17.0) 5. 2LE ( 14.8-31.6) 6. 2LF ( 16.8-34.8) 7. 2LG (2.3-9.2) 2RA [d]2 (3.4-14.0) 2RB (8.9-22.5) 2RC (1.7-7.4) 2RD (19.3-25.6) Pericentric Inversions 1. 2. (2L-2R)A (35.2-28.9) Translocation · (XL-2R)A 1. 11.7 26.0 Symbols c11doscd in brackets are nuntcs given by olher authors to these some inversions. ' C:lll., Clff..(), CIIIL, nnd CUii\: Kikknwa (1938). 2 h nnd d: Seecof in Stone, el al. ( 195 i l. 3L 3LA[CIIIL] 1 (1.0-19.5) 3LB (5.9-12.4) 3LC (2.3-15.4) 3H lo..j 3RA[CIIIR]1 (22.0-27.8) ~ 3RB ~ (3.0-22.9) 3RC ~­(1.9-7.5) ;:;: 3RD ( 16.4-29.0) ~ .Q.. ~ "1:j i::: (3L-3R)A \:)"' ....... 18.3-8.0 §" (3L-3R)B ..... 5· 5.0-5.5 ;::i I -·-------1 _r:: -------.i. ..-----==-___; ~ ID ~ x ...J .._____ ,_ -= I .. ~· --~ - -· ! ­ -=---~ n -~ ~--~ -=-ct a: --N-x ---==. = ct ...J --~ x L--" -- ==-­L =-~ ~ N - ~ CD ~ ~ a:: ~~ .. II x II x ~ )) < r<> w .., ::: NG-1 XLA None XRA None (2LE;2LB) 2LF 2RD (2RA;2RC) 3LC None (3RB;3RC) None None ...... (") 2LC ~-. l::i .... NG-2 XLA None XRA None (2LE;2LB) 2LF (2RA;2RC) None 3LC None None None None -. 0 XLB 2LC 2LG 2RD ;:i Rarotonga Std. None Std. None Std. None Std. None Std. 3LA Std. 3RA None Palmyra Std. None Std. None Std. 2LA Std. None Std. 3LA Std. 3RA None Hawaii Std. None Std. None Std. 2LA Std. None Std. 3LA Std. 3RA None Cuba Std. None Std. None Std. 2LA Std. None 3LA None Std. 3RA None 1 Standard and inverted arrangements have similar frequencies. T ABLE 4 Geographic distribution of paracentric inversions of Drosophila ananassae as recorded in this and other studies. Chromosome Inversion XL XR 2L 2R 3L 3R A SamoaL New Guinea . . . . . . . . . . . ... . ... . . . . .. . . . . . . . . . . .. . New Guinea .... .. .. . . ... .. .. . .. . . . . . . . . . . . . . .. . ...... .... ..... ... .. Japan2 China2 Hawaii Palmyra Brazil3 Mexico Cuba Alabama2 Saipan2 Ponape1 SamoaL New Guinea Majuro1 Japan2 Formosa2 Majuro1 Hawaii Palmyra Rarotonga Brazil3 Mexico Cuba Alabama2 Saipan2 Japan2 China2 Formosa2 Hawaii Palmyra Rarotonga Brazil3 Mexico Cuba Saipan2 B New Guinea ........ . Formosa2 Bikinit SamoaL . ......... Samoa L Samoa L New Guinea . .. . .. .. . . ... .. . .. .. ....... ... ... ... .. . . Majuro1 Ponape1 Samoa L . . . . . . .. . . .. .... .. . . . ... . . . . . . .. .. . .. .. ... ... . .. . New Guinea Saipan2 ...... .... c . ...... ... . .. ... .. . . .. . .. ..... . .... . . .. . .... . .. . . . .. .. .... Ponape Samoa L New Guinea New Guinea . . .... . ... New Guinea New Guinea D . . . . . . . . . . ... .... .. . Samoa L New Guinea . ... . .. .. . Samoa D E New Guinea .... . ... .. F ..... .... New Guinea G New Guinea ....... .. . 1 Seecof, in Stone, et al. (1957) . 'Kikkawa (1938, 1939). • Dobzhansky and Dreyfus (1943). 2LB in strains from Bikini, Majuro, and Ponape Islands. The former two are in the Marshall Islands while the latter is one of the eastern Caroline Islands. At the time he stated that this short inversion was probably identical to one found in ananassae from Formosa and Saipan, an island in the Mariana group, by Kik­kawa ( 1938, 1939) and designated by him as CIIL-0. Moreover, of eight males from Ponape crossed by Seecof to "standard" females, four produced larvae heterozygous for medium length subterminal paracentric inversions in both the left arm and right arm of chromosome 2. The break-points reported by him for these two inversions agree very well with the break-points determined in this study for inversions 2LC and 2RA. Thus it would seem that one-half of each of the two pairs of overlapping inversions in the light Samoan chromosome 2 also exist in this chromosome in the ananassae of Ponape. This could not actually be The University of Texas Publication confirmed at this time because the original stock of ananassae from Ponape could not be located. The salivary gland chromosomes of all of the stocks listed in Table 1 were examined both as they existed in the stock cultures and as they appeared in larvae produced by crossing the stock to the standard Majuro strain. Table 3 lists the cytological characteristics of each of these stocks. Two of them, Majuro and Truk, were homozygous for the standard gene sequence. Five carried at least two of the three widely distributed and frequently reported inversions (Kauf­mann, 193 7; Kikkawa, 1938; Dobzhansky and Dreyfus, 1943; Seecof in Stone, et al., 195 7; and Freire-Maia, 1961) of ananassae. Stocks from Palmyra, Hawaii, Cuba, and Mexico had inversions 2LA, 3LA, and 3RAcorrespondingtoKikkawa's CUL, CIIIL, and CIIIR. The stock from Cuba was homozygous for the terminal inversion 3LA. The stock from Rarotonga was found to have both 3LA and 3RA. These inversions are indicated on the map in Figure 1. The flies from Niue were of special interest since they had been recently col­lected from an area which was known to have an exceptionally high amount of naturally occurring radiation. At least four of the larvae produced by each fly were examined. Of eight-four flies examined in this manner, only three were found to have chromosomal aberrations. Two of these were pericentric inversions and the third was a translocation. This last rearrangement, an exchange of parts between XL and 2R, was only found in one larva, although many others from this culture were examined. The two pericentric inversions, on the other hand, were seen in a number of the F1 larvae and could still be found frequently in larvae from these two strains as many as five generations later. (2L-2R)A was relatively short and was found homozygous several times, while (3L-3R)B was found only when heterozygous. These three chromosomal rearrangements are indicated on the map in Figure 1. The two stocks from Papua, New Guinea, provided the most interest with respect to this attempt to analyze the two Samoan forms. Larvae from both of these stocks were cytologically polymorphic to some extent, being heterozygous for several previously unrecorded inversions. New Guinea stock 21cc was hetero­zygous for inversions 2LF and a pair of overlapping inversions in the right arm of chromosome 2, 2RC, and 2RA. New Guinea stock 22jj was heterozygous for 2LF and 2LG. These inversions are indicated on the maps for 2L (Figure 4) and 2R (Figure 6). Larvae produced by crossing either of these two stocks to the Majuro strain and the Samoan light stock were heterozygous for even more para­centric inversions than were larvae from crosses between the light and dark Samoan flies. Some of these inversions were found to be identical to some of those which had been found in the Samoan light by Samoan dark crosses. Inversions XLA, XRA, 2LB, 2LE, 2RD and 3LC were all homozygous in both New Guinea stocks. In addition, inversion 3RC, which overlaps 3RB, was homozygous in one of the stocks. The (2RA;2RC) complex, heterozygous in 21cc, was homozygous in 22jj and inversion XLB, which was not found in 21cc, was also homozygous in 22jj. Inversions 2LC, 2LB, 2LE, and 2LF, are shown in Figures 3 and 4. In­versions 2RA, 2RC, and 2RD are in Figures 5 and 6; inversions 3RB and 3RC are in Figures 7 and 8; inversion XRA is in Figures 1 and 9; and inversions 3LC and XLB are in Figure 1. B ! b FIG. 5. Photomicrographs of right arm of chromosome 2 (2R) from larvae heterozygous for various complex gene arrangements. A.-Pairing when heterozygous for standard gene sequence and overlapping inversions (2RA;2RB). (Standard X Tutuila light) B.-Pairing when heterozy­gous for standard gene sequence and overlapping inversions (2RA;2RC) and inversion 2RD (arrow c) . Arrows a and b indicate locations of tips of each chromosome. (Standard X New Guinea) C.-Pairing when heterozygous for overlapping inversions 2RB, 2RC, and 2RD. Inver­sion 2RA is homozygous. Arrows a and b indicate chromosome tips. (Tutuila light X New Guinea) The University of Texas Publication CROSSES Table 5 summarizes the outcome of crosses between the light and dark Samoan flies. Included are homogamic control crosses, F1 crosses, and F1 backcrosses. It is clear from the results of all of these crosses that the two forms are fully capable of interbreeding, for in most instances when insemination occurred, fertile off­spring resulted. Moreover, the percentages of fertile heterogamic crosses were not very much lower than the percentage of fertile homogamic crosses between light flies. It was noticed during the examination of the offspring of heterogamic crosses that at least part of the genetic control of abdominal pigmentation must be sex­linked. Females of both crosses had the same intermediate phenotype. The appear­ance of these flies included a darkening and widening of the transverse bands, plus the presence of an indistinct dark spot in the center of each tergite. The males resulting from the two crosses were not exactly alike. Males from hetero­gamic crosses involving dark females were substantially darker than males pro­duced by light females in the reciprocal cross. The crosses listed in Table 5 were made approximately one year after the flies reached The University of Texas laboratory. Some of these same crosses were done earlier, shortly after the flies first arrived, by Mrs. Florence Wilson and Mrs. Virginia Gerstenberg. A comparison of the results of the two groups of tests conducted one year apart is made in Table 6. These data indicate that the fre­quency of crossing fertility had increased during the year that the flies were kept TABLE5 Summary of results of homogamic and heterogamic pair-matings of Samoan light and dark flies, including F1 crosses and backcrosses. Females were dissected 14 days after matings were begun. Temperature approximately 22°C No. Percent inseminated inseminated Percent Percent 'i? 'i? from · 'i? 'i? from all 'i? 'i? Nature of cross No. pairs No. fertile fertile sterile pairs sterile pairs inseminated D'i' X D e! 108 94 87.0 1 7.1 88.0 L'i' X L e! 105 69 65.7 0 0 65.7 D'i'XL .=::;,. co = Q'.) N N N N .....-­-...;;--­ ------= ... ,._ ~ --~ ~----N ~-~ ~ ~~ ~-.N "' N"' "' N"' N N "' N " :~~ -~~ ~~ Cl ~ N~ N er: .. f\J N N NL "'= .... ~ ~-7, - - ~ "' ---~ ........... Ill 0:: N --~~­ = 0 "' <( 0:: (.) ....... -­ NL~' =_ _...... 0:: = N N .....:....;..-N '~ ,,,.,·;,;;~;:-en "' -= ~ -~. ­ """'--­ ::;!J:.-­ .:.. !ii~-.­~-­ ~ ¥ = ~-­ ~-­ L Cl FIG. 6. Chromosome maps of 2R. A.-Standard gene sequence. B.-Ponape: Break-points of 2RA are indicated and the segment is shown inverted. C.-Tutuila light: Break-points of 2RB are indicated. 2RA is inverted and 2RB, which overlaps it, is also shown inverted. D.-New Guinea: Break-points of 2RC and 2RD are indicated. 2RA is inverted; 2RC, which overlaps 2RA, and 2RD are shown inverted. in the laboratory. This increase was especially significant for backcrosses of F1 females, rising approximately fifty-two percent. The progeny of F1 male backcrosses were tested by crossing them back to each of the two parental types. In this particular instance, each hybrid individual was The University of Texas Publication TABLE6 Comparison of frequencies of fertile pairs from matings of Samoan light and Samoan dark flies including F 1 crosses and backcrosses shortly after they were first collected and again one year later August, 1962 August, 1963 Temperature: 24°C Temperature: 22°C Natu re of cross No. pairs No. fertile Percent fertile No pairs No. fertile Percent fertile LQ X D o 142 44 30.9 123 62 50.4 D9 X Lo 128 73 57.0 108 57 52.8 F1 (L9 X D o) 9 X F1 (LQ X Do) o 168 86 51.2 138 100 72.5 F1 (D9 X L o) 9 X F1 (D9 X L o)o 140 69 49.3 136 109 80.1 L 9 X F 1 (L 9 X D o ) o 131 83 63.4 D9 X F1 (L9 X D o)o 79 39 49.4 145 108 74.5 L9xF1 (DQ X Lo)o 95 41 43.2 133 77 57.9 D9 x F, (D9 X L o)o 101 56 55.4 120 87 72.5 F1 (L9 X D o)9X Lo 50 4 8.0 139 73 52.5 F,(L9XDo)9XDo 96 7 3.0 133 94 70.7 F1 (D9XLo)9X Lo 118 13 11.0 121 70 57.8 F1 (D 9 X L o) 9 X Do 111 19 17.1 124 81 65.3 tested against three non-hybrid individuals of the opposite sex; e.g., one hybrid male was mated with three non-hybrid females, etc. This was done in order to increase the probability of hybrid flies becoming involved in at least one success­ful copulation. The results from these crosses are given in Table 7. These data show a difference in the frequency of fertility which can be directly related to the frequencies of "dark" and "light" chromosomes in each type of hybrid. The type of X-chromosome present seems to have been of special importance. Thus TABLE 7 Fertility observed in F2 backcrosses involving 6 and 9 progeny obtained from backcrosses of F1 hybrid (light X dark) o o. Each individual to be tested was mated to three individuals of the opposite sex from one of the two parent stocks. Temperature: 25°C Initial cross: A. Light'f X Darko Initial cross: B. Dark 'f x Lighto No. Percent No. Percent 3'f 'f x 10 00 Cros..c;es crosses tested fe1~ile fertile 3'f 'f x 10 O O Crosses crosses tested fertile fertile 1. D99X(L9XF1 o)o 109 54 49.5 1. D 9 9 X (L 9 X F1 o) o 103 44 42.7 2. L 9 9 X (L 9 X F1 o) o 120 113 94.7 2. L99 X(L9 X F 1 0)0 118 104 88.1 3. DQ 9 X(DQ X F1 o) o 118 109 92.4 3. D 9 9 X (D 9 X F1 o) o 118 97 82.2 4. L99X (D9 X F 1 o) o 120 90 75.0 4. L9 9 X(D9 X F 1 0) 0 110 87 79.1 1'f X3oo 'f '?Tested 1'fX3oo 'f '?Tested 5. (L 9 X F1 o) 9 X D o o 11 5 98 85.2 5. (L9 x F, oHx D oo 105 78 74.3 6. (L9XF,o)9XLoo 122 109 89.3 6. (L9xF,o )9xLoo 114 89 78.1 7. (D9xF,o)9XDoo 116 107 91.5 7. (D 9 x F1 o) 9 X Do o 105 100 95.2 8. (D9XF1 0)9XL00 122 68 55.7 8. (DQ x F1 0)9 X Loo 117 61 52.1 A FIG. 7. Photomicrographs of the right ann of chromosome 3 (3R) from larvae heterozygous for the following: A.-Standard gene arrangement and inversion 3RB. (Standard X Tutuila light) B.-Tutuila light arrangement and inversion 3RC. (3RB is homozygous.) (Tutuila light X New Guinea) C.-Standard gene arrangement and overlapping inversions (3RB; 3RC). (Standard X New Guinea) The University of Texas Publication A zz j 'z •oz/ '' • e1 I ''I" ls1 I"j t1 J z1 I "I 01 j ' I • ! Lj •J s I • l c •11luJ1 1il1111· 1 n u11 :: nfill' :11t f11 i 111111 tni'!w rn vn ( 117r 111 ·um~·litrlJ II r2 I n 24 1ZS I2, I2:-r:r:, B .---------------3RC---------------. I ZJ '•/sj•icj IJ!j lllil Ill Ill JllJII)llUIJ ·r: II u: "r Jlllltl trti:id.: 1\vn 1::a1 ~i Ulll ·u~•~iltll• I , ' I • ' I10 rII I 12 1.. 1 •• 1.5 " I11 I18 l .. I20 I21 I22 1n I2. IZS I2, 2;~, c FIG. 8. Chromosome maps of 3R. A.-Standard gene sequence. B.-Tutuila light: Break­points of 3RB are indicated and the segment is inverted. C.-New Guinea: Break-points of 3RC are indicated. 3RB is inverted and 3RC, which overlaps it, is also shown inverted. hybrid males, sons of light mothers, bearing a "light" X-chromosome, were found to cross successfully with dark females in one test (cross A1) 49.5% of the time and in the other test (cross B 1) 42. 7 % of the time. These same two classes of hybrid males crossed successfully with light females in test A2 94.7% of the time, and in test B2 88.1 % of the time. The differences shown by crosses involv­ing hybrid "dark" X-chromosome males were not as large as those found in the four crosses already described. However, considerable difference could be seen with respect to the success of different classes of hybrid males when they were crossed to dark females. "Dark" X-chromosome males in tests A3 and B3 were much more successful than "light" X-chromosome males in test A1 and BL There was a similar but smaller difference when light females were involved; i.e., "light" X-chromosome males were more successful at mating with light females than were "dark" X-chromosome males. Similar trends were noted in crosses involving hybrid females from these Fi male backcrosses. Wherever dark males were concerned the differences were rather small. Light males, however, produced results comparable to those pro­duced by dark females in crosses to hybrid males. When the hybrid female carried a homozygous pair of "light" X-chromosomes, as in test A6, 89.3% of the matings FIG. 9. Photomicrographs of X-chromosome from larvae heterozygous for the following: A.­Standard gene arrangement and inversion XLA. (Standard X Tutuila light) B.-Standard gene arrangement and inversions XLA and XRA. (Standard X New Guinea) The University of Texas Publication with light males were fertile. The females in test B6 and test A8 were hetero­zygous for "light" and "dark" X-chromosomes. Thus the difference in the fre­quency of fertility as seen in the two crosses-B6 was 78.1 % and A8 only 55.7% -might be due to differences in the average frequencies of "light" and "dark" autosomes. Females from A8 should have averaged more "dark" than "light" and females from B6 more "light" than "dark." When, as in test B8, the female had a homozygous pair of "dark" X-chromosomes, the lowest percentage of fer­tile crosses was seen. There is a possibility that these differences in hybrid crossing behavior might have been due to some cytoplasmic maternal effect passed on by the parent type females. Thus hybrid offspring of dark females might be expected to cross more frequently with dark flies than with light flies, and vice versa. However, the data given in Tables 5 and 6 for F1 backcrosses give no indication of such a system operating in these instances. TABLES Crossing relationships of Samoan light and Samoan dark flies to five different geographic strains of Drosophila ananassae as measured by the frequency of fertility in pair-matings. These matings were terminated 28 days after initiation. Temperature: 25°C Nature of cross Na. pairs Na. fertile Percent fertile Maj'? x DJ 122 116 95.1 Maj'? X LJ 114 74 64.9 D<;? X MajJ 130 124 95.4 L<;? X MajJ 121 47 38.8 Maj<;? X MajJ 119 110 92.4 Trukc;> X DJ 117 99 84.6 Truk<;? X LJ 119 45 37.8 D<;? X TrukJ 104 94 90.4 L<;? X TrukJ 108 39 36.1 Truk c;> X Truk J 130 116 89.2 NG-1 <;? X DJ 191 2 1.05 NG-1 <;? X LJ 197 79 40.1 D<;? X NG-1 J 188 26 13.8 Le;> X NG-1 J 195 97 49.7 NG-1 c;> X NG-1 J 132 124 93.9 Mexc;> X DJ 122 99 81.1 Mex'? X LJ 123 41 33.3 Dc;>x MexJ 132 129 97.7 L<;? X MexJ 125 47 37.6 Mex<;? X MexJ 144 138 95.8 Niuec;> X DJ 131 119 90.8 Niuec;> X LJ 113 52 46.0 D<;? X NiueJ 110 107 97.3 L<;? X NiueJ 109 66 60.6 Niue'? X Niue J 134 125 93.3 Frn. ·to. Chromosome ZR. Photograph, drawing, and schematic diagram of overlapping inver­sions (ZRA;ZRB) showing pairing relationships and break-points. One factor which may have had an effect on the data derived from the crosses discussed thus far has not been mentioned yet. All of the crosses included in Table 5 were kept in a room where the temperature was somewhat variable, although it usually remained at about 22°C. All of the other crosses, except one The University of Texas Publication of the multiple-choice experiments and the crosses recorded in Table 6, were kept in another room where the temperature remained fairly constant at 25°C. This temperature difference may have had an important influence on the mating behavior of the flies. Tables 8 and 9 present data indicating the crossing behavior of both Samoan forms and several different geographic races of ananassae. The results in Table 8 are fertility percentages from long duration matings, while those in Table 9 are insemination frequencies from shorter term matings. The data from either table show quite clearly that, with the exception of the two stocks from New Guinea, the dark Samoan stock crossed as well with any of these widely scattered races TABLE9 Relationships between Samoan light and Samoan dark flies and five different geographic strains of Drosophila ananassae as measured in pair-matings by the frequency of insemination following seven days exposure. Temperature: 25°C Nature of cross D2X Do D2 x Lo L2x Lo L2x Do Mai2 XDo Mai2 x Lo D2 x Maio L2 x Maio Maj2 x Maio NG-12 X Do NG-12XL o D2x NG-to L2 X NG-1 o NG-1 2 X NG-1 o NG-22 x Do NG-22 x Lo D2 x NG-2o L2 X NG-2o NG-2 2 x NG-2 o Mex2 x Do Mex2 x Lo D2 x Mexo L2 x Mexo Mex2 x Mexo Niue2 x Do Niue2 x Lo D2 x Niueo L2 x Niueo Niue2 x Niueo No. pairs 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 ::\"o. fertile 25 1 23 2 25 7 23 2 25 0 17 0 12 24 0 16 3 9 23 24 1 25 2 22 24 1 23 3 22 No. of others inseminated 2 1 0 0 0 0 0 0 0 2 0 2 0 0 1 2 0 0 2 2 2 Percent total matings inseminated 100.0 12.0 96.0 12.0 100.0 32.0 92.0 12.0 100.0 0.0 68.0 0.0 48.0 96.0 0.0 72.0 12.0 40.0 100.0 96.0 4.0 100.0 12.0 96.0 96.0 4.0 100.0 20.0 96.0 as flies from any of them did intraracially in control homogamic crosses. The percentage of fertile crosses involving light Samoan flies, however, was usually much lower. The exceptional results obtained from crosses with either of the New Guinea stocks were quite radical in this sense. Because of this additional matings were made. Crosses between New Guinea ananassae and dark Samoan flies were very unsuccessful, while fertility, although lower than homogamic crosses, was reasonably high in crosses between New Guinea and light Samoan flies. F1 crosses of the offspring produced by all of these crosses were usually fertile; no backcrosses were attempted. These two experiments indicated that time was an important element in these laboratory instigated crosses. Apparently if males were allowed to court long enough, the discriminatory resistance of some of the females was reduced so that a successful copulation could be achieved. The results of all of the mating tests discussed up to this point indicate that if a successful copulation occurs between a dark Samoan fly and a light Samoan fly, fertile offspring will more than likely be produced. Isolating mechanisms such as zygote sterility and insemination reaction were seen not to exist, while gamete and zygote mortality, if they occurred, were not effective enough to be readily apparent. MULTIPLE-CHOICE MATING TESTS The data did indicate in the case of the crosses between the Samoan light and dark forms and other geographic races of ananassae that sexual isolation could occur in the form of discrimination by means of preferential mating behavior. To test this hypothesis two multiple-choice experiments were conducted. The first was only partly successful. This test utilized the technique employed by Patterson, McDanald, and Stone (1947) for testing members of the virilis group for mate discrimination by males. Five light females and five dark females were mated with either a single dark male or a single light male in a vial. At the end of twenty-four hours the females were removed and dissected in order to determine which and how many had been inseminated. Other similar matings were checked at the end of forty-eight hours, still others at seventy-two hours, and so on until an average of more than five females per vial was found to be inseminated. As Table 10 shows, this average was reached in matings involving dark males by the end of forty-eight hours. Isolation indices were calculated for both forty-eight hour matings and seventy-two hour matings by means of the %homogamic ­formula derived by Stalker (1942) (isolation index = % homogamic + % heterogamic . %h . ) . The data suggest that dark males exercise rather strong prefer­ eterogarmc ence by selecting conspecific mates most of the time. Actually, the behavior of the females probably had considerable influence on the behavior of the males. Light females were seen frequently attempting to repulse the male, often shaking the male off if he tried to mount. The other part of the experiment failed to produce results of comparable significance. The average of more than five inseminated females per vial was not reached by the end of 168 hours. The experiment was The University of Texas Publication TABLE 10 Results of three-way multiple-choice experiments in which five of each type g were placed in a vial with a single (I; • g g were examined for insemination at prescribed intervals of time. Temperature: 22°C Homogamic Heterogamic 'i''i' 0 Interval of time Number dissected Number Percent insemi­insemi­nated natecl Nwnber Number insemi­dissected nated Percent insemi­nated Isolation Index L+D Light 24hours 48 hours 25 25 5 4 20.0 16.0 25 25 0 0.0 4.0 72 hours 25 4 16.0 25 0 0.0 96 hours 25 13 52.0 25 6 24.0 120 hours 25 6 24.0 25 6 24.0 144hours 50 23 46.0 50 12 24.0 0.31 168 hours 50 17 34.0 50 9 18.0 0.31 L+D Dark 24 hours 25 9 36.0 25 0 0.0 48 hours 50 39 78.0 50 3 6.0 0.86 72hours 50 41 82.0 50 2 4.0 0.91 terminated then because the supply of matings had been exhausted. Isolation indices were calculated from results of vials from the 144 hour group and the 168 hour group. However, since so few females had been inseminated, these calculations are of no great value as measurements of sexual isolation. One thing should be emphasized with respect to this experiment. This, like the pair-matings between light and dark Samoan flies recorded in Table 5, was conducted at 22°C. During these multiple-choice tests it was noticed that the light males were much less active in their courtship behavior than were the dark males. Looking again at the data from Table 5, it can be seen that the homogamic light cross was considerably less fertile than the homogamic dark cross. However, the small number of pair-matings of light-to-light and dark-to-dark examined in Table 9 indicate that the difference seen in Table 5 may be misleading. At 25°C, light males were observed to be much more active. The distinctive appearance of hybrid adults, especially females, and the cyto­logical differences seen in hybrid larvae, made possible the second of the two multiple-choice experiments. This experiment was designed to allow females as well as males to exercise selection with respect to the choice of mates. Each mat­ing consisted of five light females, five dark females, two light males, and two dark males per vial. At twenty-four hour intervals, one hundred females of each type were isolated individually, allowed to lay their eggs, and then removed. When off-spring appeared they were transferred and allowed to age for five days so that their pigmentation could develop and their types be more easily recog­nized. Larvae produced by flies identified as possibly hybrids were checked cyto­logically in order to verify by the presence of heterozygous inversions XLA and 3RB that these flies had resulted from heterogamic matings. Three instances of suspected heterogamic matings involving light females were found to be homo­gamic cytologically; all instances involving dark females which were checked cytologically proved to be heterogamic. As Table 11 shows, heterogamic crosses were very infrequent. This experiment was conducted at 25°C and as the data clearly indicate, insemination was accomplished very quickly, most of it within the first twenty-four hours. The obviously considerable change in the behavioral activity of light males at 25°C compared to 22°C is also apparent. DISCUSSION There can be little doubt that in sexually reproducing organisms the wide­spread polytypic species present the greatest opportunity for the formation of new species. Numerous authors, for instance, Mayr (1942, 1963), Dobzhansky (1951), Patterson and Stone (1952), Carson (1959), Stone (1962), and Wasser­man (1963), have discussed this idea in theory and illustrated it by citing ex­amples of its occurrence as seen both in nature and in the laboratory. Much of the laboratory evidence has been collected through the efforts of geneticists study­ing groups of closely related species of Drosophila; e.g., Wasserman's extensive study of the numerous species of the repleta species group (Wasserman, 1962a, b,c, and d). The existence of such a closely related, frequently almost identical, cluster of species is usually assumed to indicate a common genetically polytypic ancestor. Although this model for speciation is known from genetic and cytological evidence to occur in Drosophila, the opportunity to observe it genetically and cytologically as it happens, or at least while still a recent event, has at best been rare. Drosophila ananassae certainly appears to qualify as a polytypic species. Its widespread circum-tropical distribution, especially through the scattered island groups of the Pacific Ocean, has permitted recognizable genetic differences be­tween parts of the species population to become well developed so that geographic races can be distinguished. The light and dark color types of the Pacific Islands are certainly deserving of racial distinction. If these two semi-isolated populations can be said to exist as separate races, then the fluid cosmopolitan population spanning the continents and man's major trade routes must be considered as a third distinct population, also a :race. Although there is some color variation over the range of the cosmopolitan race, from the dull brownish-yellow of Asia to the gray of Hawaii and Palmyra to the yellowish-brown of the Americas, it is considered, nevertheless, as one race be­ cause of the occurrence of three particular and characteristic inversions through­ out this range of the species. Wherever the species flourishes in China, Japan, Hawaii and the Americas, inversions 2LA, 3LA, and 3RA, are found (Kaufmann, 1937; Kikkawa, 1938, 1939; Dobzhansky and Dreyfus, 1943). (See Table 4 for a synopsis of the geographic distribution of inversions in the ananassae complex.) That each of these inversions could have occurred independently so many times is inconceivable. It is, therefore, concluded that through them one can trace the spread of the species through the agency of human travel. By this same means it is expected that the parts of this population, however remote they may seem, maintain genetic contact with one another. The ananassae population on Palmyra Island is cited as evidence for this hy­ pothesis. This very small island is one of four small islands which extend in a line north from the equator. They are well isolated and can be presumed to have The University of Texas Publication had a very limited native flora and fauna as they were devoid of human inhabi­tants until very recently. During World War II, Palmyra was used extensively as an air transport point in support of military operations in the Pacific area. In this connection, it is interesting to note that the ananassae population on this island is nearly identical to that in Hawaii, a part of the cosmopolitan race and undoubtedly a recent arrival on the island. The widespread occurrence of the three "cosmopolitan" inversions implies their importance to the success of the species. It has been found in many species of Drosophila that heterotic buffering is associated with chromosomal polymorph­ism. Such would be expected to be the case in large domestic populations of ananassae, and there is evidence that at least for 2LA and 3LA this is true. Heterosis has been found to be associated with these two inversions when hetero­zygous by Moriwaki and his co-workers (Moriwaki, Ohnishi, and Nakajima, 1956; Tobari, 1962; and Moriwaki and Tobari, 1963) . A fourth inversion, 2LB, is found frequently distributed in a different direc­tion. It occurs on Formosa and on Saipan, one of the Mariana Islands, with 2LA, 3LA, and 3RA (Kikkawa, 1938, 1939). However, further to the south, it is found in conjunction with the light ananassae race of the Western Pacific Islands. Thus it has been reported from Ponape of the eastern Carolines and Majuro and Bikini of the Marshalls ( Seecof in Stone, et al., 195 7). It is interesting to note that of the three "cosmopolitan" chromosomes, only 3LA was found and then only from one of forty-seven flies examined from Majuro. During the course of the present study, 2LB has been found to be homozygous in two stocks of light flies from southeastern New Guinea and heterozygous among the light forms from Samoa. Thus there is a cytological relationship among the light flies not shared by all of the cosmopolitan race nor, for that matter, the dark Pacific race. The dark flies of the eastern Pacific islands seem to be relatively free of inver­sions. Of the four populations examined cytologically, only the one from Raro­tonga was found to be significantly polymorphic in this respect. Both 3LA and 3RA were found frequently in these flies, indicating a genetic connection between this island's ananassae population and the cosmopolitan race. It is possible that flies carrying the "cosmopolitan" inversions have been introduced recently by man and have mixed with the dark indigenous population. An interesting observation in connection with the dark race of Pacific flies was the frequency of naturally occurring pericentric inversions. The three discovered at this time increase the number of such inversions reported from natural popu­lations of ananassae to at least eight. The previous count of five (Freire-Maia, 1961) reported from Brazilian populations of the species was said to exceed the number of known pericentric inversions from all other Drosophila species taken together. In this particular case, the highly asymmetrical (3L-3R) A from Samoa was never recovered again, while the two synunetrical pericentric inversions from Niue, (3L-3R)B and (2L-2R)A, were found frequently, the latter even homozygous, as many as five generations later. It is among the Pacific island populations of ananassae that the right conditions for speciation appear most prominently. The two phenotypically distinct light and dark races have seemingly developed independently within the confines of two Futch: Drosophila ananassae Populations general areas of geographically isolated island groups, the islands of Micronesia to the north and the islands of Polynesia to the east of New Guinea. As Mayr (1942, 1963) and Carson (1959) have emphasized, it is from the marginal isolated populations of a polytypic species that new races and in time, perhaps, new species can be expected to evolve. Following the generally accepted model of such an isolate, it would be expected that in contrast to the hetero­geneous genetic constitution of the main population, the individuals in this popu­lation would be decidedly homogeneous. Several reasons can be given for this homogeneity. If the isolate represents the establishment of the species in a new habitat-e.g., a formerly uninhabited island-it is logical to assume that the colonizers were few in number, perhaps only a single fertilized female, and there­fore representative of only a very minute fraction of the total genetic population. If, on the other hand, the population is a small remnant of the species separated from the main body by some change, geographical, climatic, or otherwise, the surviving individuals could be expected to be of highly selected genotypes and therefore lacking variability. In any event, the resulting population would be inbred and homogeneous. Two such populations derived from a polytypic species and established independently of one another might be expected to differ from one another significantly at the outset. Another reason for homogeneity in the small isolate is that it cannot afford the luxury of heterotic buffering provided by such systems as balanced chromo­somal polymorphism. In order to respond to the stress of what is usually a rigor­ous and changeable environment the way must be left clear for frequent experi­mentation through recombinations of genes and the effective dissemination of beneficial mutations. Strickberger (1963) has essentially shown this to be true experimentally. He tested chromosomally monomorphic and polymorphic cage populations of D. pseudoobscura and found that in general the monomorphic forms showed more rapid accumulation of genetic changes improving their rela­tive fitness for the cage environment than did the polymorphic forms. Thus it can be readily seen that it is possible for initially different forms from two separate isolated populations of a polytypic species to diverge considerably from one Another in a stepwise manner by virtue of independent evolutions as they adapt to their own particular environments. The distinctive color difference between the two Pacific Island races of ananassae has probably arisen in such a way. Kalmus (1941a, 1941b) has shown that there is physiological significance associated with the degree of cuticle pig­mentation in insects. He has reported that dark pigmented cuticle provides more protection against ultraviolet light, absorbs more heat, resists desiccation, and imparts more mechanical strength to the exoskeleton by hardening the chitin. In a series of experiments utilizing both mutant and wild-type strains of several species of Drosophila he found that yellow mutants died earlier and lost more weight than wild-type flies of the same species while being starved in dry air at 25°C. Under the same conditions ebony mutants of melanogaster outlived wild­type and lost less weight. These differences were found to decrease with an increase of relative humidity and to disappear at 100% moisture. Another inter­esting discovery was that yellow mutant females of all of the four species that The University of Texas Publication he tested survived longer in moist conditions when provided with drinking water than did wild-type females of the same species. He could offer no explanation for this last phenomenon and added that no such difference was noted in identi­cal tests with yellow and wild-typemales of the same species. The evolutionary significance of pigmentation difference as implied by these experiments is of considerable interest in view of the possible differences, other than merely color, which may exist between the light and dark races. It seems entirely feasible that each form has adapted to a somewhat different way of life and probably has different ecological requirements. What could be expected to happen if individuals of each race were to colonize the same island? As experiments conducted in this study have shown, members of the dark race from Samoa are as fertile when mated with members of the light race from Majuro and Truk as any of the three are in homogamic matings with their own kind. However, interbreeding might not occur in the same manner in nature. There are a number of possible results of such renewed contact. If hy­bridization occurs frequently and there is not much ecological separation, the two races might be expected to integrate into a single population. If hybridization is restricted by divergent ecological habitation the two races could remain essen­tially separate with only a small rate of interracial gene flow. This situation could result in eventual species differentiation. If the two races fail to hybridize and are able to survive, they may be considered different species. If, on the other hand, the two do not hybridize and are unable to adapt to different niches, competition would be expected to develop and one of the two forms should be eliminated from the island. When hybridization does occur between two ecologically differentiated races which have come to occupy overlapping zones, the hybrids may be unable to compete successfully in any of the niches available. These hybrids may be inferior in fitness in a number of ways, including reduced fertility, lower via­bility, etc. In any event, hybridization will be selected against through the failure of hybridizing individuals to contribute successfully to the genetic store of either population. Thus genetic constitutions favorable for interbreeding will be slowly eliminated. Selection occurs in the setting provided by the natural circumstances sur­rounding the specific habitats of the two forms. The factors which suppress inter­breeding in nature may break down in the laboratory. Two examples of sympatric sibling species of Drosophila capable of producing fertile hybrid offspring should illustrate this point. The frequently studied pair of species, D. pseudoobscura and D. persimilis, are two of these. Although capable of producing fertile hybrid females but sterile males when crossed in the laboratory, there are only three instances reported of interbreeding in nature (Spieth, 1958; Mayr, 1963). This is true in spite of the fact that over part of their ranges the two species are sym­patric, that many thousands of both species have been collected through the years, and that some interbreeding occurs in multiple-choice tests involving the two species. Carson ( 1954) studied three sibling species of the willistoni species group occurring in Brazil. Two of these species, D. bocainensis and D. parabocainensis, are sympatric over part of their ranges. Laboratory crosses between the two pro­ Futch: Drosophila ananassae Populations duced numerous fertile hybrid males and females. In spite of this, no instance of hybridization was found among flies collected in the wild. The light race of ananassae from Micronesia and the dark race from Polynesia apparently have not diverged to this point yet. Although very different in color, they are very much alike cytologically and seem to have acquired no appreciable amount of incipient sexual isolation. The situation surrounding the two forms on Tutuila, however, seems to be quite different. Not only are there consistent differences in the gene arrangements of the two, but there are also signs of fairly strong sexual isolation maintained by preferential mating behavior. Hybrid female larvae produced by crosses between the two Samoan forms were always heterozygous for inversions XLA and 3RB. These inversions were never found heterozygous in cultures of only light or dark flies. Subsequent examina­tion showed that the light form differs from the standard gene sequence of both XL and 3R by these two inversions. The light form exhibits a high degree of chromosomal polymorphism with respect to chromosome 2. The five rearrangements found in this chromosome are apparently associated with a balanced heterotic system in this form, although no tests were made to verify this. None of these inversions was ever found among the offspring of dark flies. Thus from all appearances-i.e., the sharp color differences and the consistent cytological difference-these two forms of Drosophila are genetically isolated from one another. This seems to be true in spite of their apparent sympatry, their morphological similarity, and their ability to cross with one another to produce fertile offspring. Although the two forms were collected in the same general area, it is possible that they may be partly ecologically allopatric. The topography and climate of high volcanic Pacific islands like Tutuila ought to afford a number of potential ecological niches for small insects such as these (Bryan, 1951 ) . Some of the features of Tutuila are its elevation (2141 feet), its relatively small area (54 square miles), its abundant annual rainfall (200 inches in some places), and its mild and only slightly variable temperatures (26°C average). Such an environ­ment can support a luxuriant flora which in turn can provide numerous micro­environments for other organisms. During the latter part of July, 1964, the author was fortunate enough to be able to spend ten days on the island of Tutuila. Through the generous coopera­tion of Mr. Neil Spencer, Staff Entomologist of the Government of American Samoa, collections were made at a number of different places on the island. Large numbers of dark ananassae, but no light ananassae, were encountered in and around the small seaside native villages. As a rule, these flies were found con­centrated in the vicinity of pig pens and around the small outdoor ovens used by the Samoans for most of their cooking. These places are usually littered with scraps of cooked or dried coconut and breadfruit meat and are generally fairly well shaded. The light form, but not the dark form, was found, ·although in small numbers, on small fallen fruit on the extremely well shaded floor of a stand of very old, very tall native trees, the only such stand of any size still found on the island. This area, which is situated in a flat inland area of fairly low elevation, The University of Texas Publication also supports small populations of several of the species of Drosophilidae which are restricted to the Samoan Islands. In certain other areas both light and dark ananassae were collected at the same time from.the same places. The grounds of the Experimental Fann proved to be very interesting in this respect. For several days Drosophila were collected over some fallen papaya on an open grassy area. Dark ananassae, melanogaster, and bryani were the predominant species; no light ananassae were found here. At another spot, less than one hundred yards away from the papaya, however, dark ananassae, light ananassae, and bryani, but no melanogaster were collected in fairly large numbers. These flies were taken by sweeping over a rotting breadfruit which had fallen into a place which was fairly well protected by shade provided by the breadfruit tree and several other neighboring trees. Larvae produced by eight dark females and eight light females isolated after having been caught at this spot had chromosomal arrange­ments peculiar to the color type of the known parent (i.e., the female) showing that no interbreeding had occurred at least as far as these individuals were concerned. The evidence, although a more thorough population check is needed to make it conclusive, indicates that the light and dark ananassae forms on Tutuila are not conspecific. They are geographically sympatric and yet they do not exist as a single freely interbreeding population. They are able to interbreed, but do not do so to any appreciable degree as their different chromosomal arrangements testify. They appear to be ecologically at least partly separated. Their different colors certainly suggest physiological differences which could be expected to influence habitat preference and behavior patterns. The relationships between these two forms and other populations of ananassae point out a possible explanation for this situation. Both forms crossed with flies from all of the stocks used in mating tests. Moreover, fertile offspring resulted from each successful cross. However, the dark form was considerably more successful than the light form in all crosses, no matter what color phenotype the other flies had, except those involving either of two stocks from Papua, New Guinea. In every instance crosses between dark flies and New Guinea flies exhibited very low percentages in fertility. Light flies, on the other hand, were frequently successful in crossing with New Guinea flies. Thus, as measured by crossing frequencies, the light form is genetically at least as closely related to the Papuan populations as to any of the others, while the dark form is certainly very much more closely linked in this way to other ananassae populations than to either the Papuan or Samoan light forms. Cytological relationships agree very closely with these crossing observations. The dark form shares a common basic gene arrangement with all of the flies except those from Papua. The light fonn and the Papuan flies have some marked cytological differences but they do have two very significant similarities; they are both homozygous for inversion XLA and one is homozygous for 3RB, two in­versions never reported before from any other population of ananassae. These two aberrant gene sequences, shared in common as homozygous arrange­ments by both the light form and Papuan populations, reinforce indications of a close genetic relationship between the two and suggest that these flies have resulted from a line of evolution different from that which produced the wide­ Futch: Drosophila ananassae Populations spread populations of ananassae. Further evidence is supplied by five sets of overlapping inversions which are found by comparing their chromosomes with the standard arrangement of ananassae. Dobzhansky ( 1944, 1951) has discussed the significance of overlapping inver­sions in Drosophila salivary gland chromosomes for determining evolutionary sequences. A pair of overlapping inversions, by virtue of their break-points, must have happened sequentially in the following manner. First, a single inver­sion occurred in a chromosome. Later, a second inversion occurred which has one of its break-points within and one of them outside the limits of the first inversion. Thus, if the three gene arrangements (ABCDEFGH, AEDCBFGH, and AEDGFBCH) are found, something can be determined about their evolu­tionary relationship to one another. Starting from ABCDEFGH, AEDCBFGH would have had to exist before AEDGFBCH muld have occurred. Likewise, ABCDEFGH could not have evolved directly from AEDGFBCH. Both ABCDE­FGH and AEDGFBCH could have evolved independently from AEDCBFGH. Overlapping inversions are pictured in several ways in Figure 10. The overlapping inversions which are found in the light flies of Tutuila and the Papuan flies are illustrated diagrammatically by the chromosome maps in Figures 4, 6 and 8. The inversions in chromosome 2, both left and right arms, indicate that this chromosome has evolved in each instance from a chromosome Z having the three inversions ZLC, ZLB, and ZRA. In the light form, ZLD has occurred overlapping ZLC, and ZRB overlapping ZRA. In the Papuan form mean­while, ZLE has occurred outside the limits of ZLC but overlapping the Samoan ZLD when the two chromosomes are compared. Also in the Papuan form, 2RC has occurred overlapping ZRA, and ZRD has occurred outside of the limits of either of these two, but overlapping the limits of the Samoan ZRB. Thus, it can be seen that ZLC, i2LB, and ZRA exist in both populations in association with several inversions unique to each population. Furthermore, these population specific inversions occur in overlapping complexes with these three arrangements. The interesting thing is that the speculated predecessor, chromosome Z, with only the three inversions ZLC, ZLB, and ZRA, apparently exists in the Micro­nesian Islands, on Ponape at least. ZLB is found rather frequently in ananassae from other Micronesian Islands and also from the Marianas and Formosa. The situation in the right arm of chromosome 3 is much less complex. Inver­sion 3RB occurred in the evolutionary line of both populations. 3RC occurred after this during the evolution of the Papuan flies. Figure 11 diagrams the possible evolution of the three major chromosomes of ananassae and related forms in a simplified stepwise manner. This diagram, although it obviously lumps a number of steps together, seems to illustrate the final independent evolution of the light Samoan and the Papuan forms from an arrangement which may have evolved from something very similar to the standard ananassae arrangement. Although it is possible for the standard to have evolved from this ancestral arrangement, it seems more likely that it happened the other way, or perhaps from one of the intermediate steps such as the "For­mosa" arrangement. These crossing and chromosomal relationships indicate that light and dark forms on Tutuila have been established as part of the island's fauna at different times. Moreover, they probably came from different sources The University of Texas Publication X CHROMOSOME CHROMOSOME 2 CHROMOSOME 3 FIG. 11 . Diagram of a simple scheme for the possible evolution of chromosomes of Drosophila ananassae and of Tutuila light and New Guinea (Papua) flies. Inversions are indicated by capital letters at their approximate loci. Overlapping inversions are indicated by connecting lines be­tween letters. Standard = the standard gene arrangement of Drosophila ananassae as expressed by the chromosome map in Fig. 1. Ancestral = the gene arrangement possessed by the common ancestor of both the Tutuila light and the Papuan forms. although some form of ananassae must have been the common ancestor of both. It seems likely that the light form was the first to reach the island. The apparent uniqueness of certain inversions which occur frequently in this population (2LD and 2RB) indicates a fairly long residence accompanied by specific adaptive evolution to fit this particular island environment. This form probably was derived from a population of ananassae or ananassae-like forms which had most of the Samoan light inversions, lacking only 2LD and 2RB, and which were also ancestral to the present forms on New Guinea. The presence of inversions XLA, 3RB, 2LC, and 2RA testifies to this relationship. The dark form is apparently a more recent arrival. It seems to have come to this general area as a subunit of the expanding ananassae population. Every recent collection of Drosophila from islands in this part of the Pacific Ocean has yielded dark forms of ananassae. In every case so far examined cytologically, the left arm of the X-chromosome and the right arm of chromosome 3 have been homozygous for the standard ananassae gene sequence. The only exception to this was the frequent occurrence of the "cosmopolitan" ananassae inversions 3LA and 3RA in the Rarotonga population. The light form apparently colonized the island of Tutuila before the dark form and is probably fairly well specialized to fit into the natural environment of the island. The dark form, on the other hand, is probably more domestic, which is in keeping with an important characteristic of ananassae. As is the case with the rest of the species, these flies have probably relied heavily on the travels of man for their far-flung distribution. They are also probably less specialized and there­ Futch: Drosophila ananassae Populations fore much more flexible. Their chromosomal monomorphism would allow for more rapid acquisition of genetic changes through recombination, etc., thus facilitating faster adaptation to environmental changes. Also, if the presence of more dark pigment signifies physiological ·differences such as those found by Kalmus ( 1941 b) in laboratory strains of Drosophila, the dark flies should be expected to be able to withstand environmental changes more easily. In most cases in which closely related sympatric species have been studied, there is no evidence that they are able to interbreed to produce numerous fertile offspring of both sexes as the Samoan light and dark Drosophila can. Although they are apparently able to interbreed successfully, they do not seem to attempt to do so. These observations prodm:e an interesting problem as to why the flies are isolated and how this isolation is maintained. The evidence gathered during the present study does not answer the question as to why isolation is maintained. It is possible to discuss theoretical reasons, however, both in the light of speculation and in the light of studies made on other species of Drosophila. It seems likely that when the dark forms first began to colonize Tutuila and contact was made with the light forms, some hybridization might have occurred. However, if any did occur, there apparently were reasons for such interbreeding to be "discouraged" by natural selection, and isolating mechanisms were rein­forced. This process of natural selection and reinforcement of already existing isolating mechanisms has been observed in mixed experimental laboratory popu­lations of pseudoobscura and persimilis by Koopman ( 1950). Table 6 of the present study indicates that some reproductive isolation may be lost when two formerly sympatric forms are isolated from each other physically. The differences noted in this table for mating behavior with respect to dark and light forms may be due in part (there were also temperature differences) to the fact that the first crosses were made soon after the flies were captured and the others were made after one year of total isolation in the laboratory. During this year, isolating mechanisms, as such, probably would have been selectively neutral and could have been partly lost through random drift, etc. A resumption of contact between breeding populations of the two forms should result in re­newed selection and reinforcement of behavioral isolating mechanisms. Selection for isolation between the two forms may have occurred for the follow­ing reasons: (1 ) Hybrids were inferior in ability to compete economically with the parent types. They, therefore, had less chance to survive and reproduce. (2) Hybrids were not as fertile as non-hybrids. Evidence for the latter is suggested by several investigators working with other species of Drosophi.la. Cooper, Zim­mering, and Krivshenko ( 1955) found, while working with melanogaster, that dominant lethals are produced when two or more pairs of major chromosomes are heterozygous with respect to inversions. Furthermore, they found that chro­mosomal heterozygosity of both major autosomes produced a large increase in dominant lethals without increasing the rate of primary non-disjunction of homo­zygous X-chromosomes. On the other hand, the rate of non-disjunction of hetero­zygous X-chromosomes was greatly increased when either major autosome was also heterozygous. Terzaghi and Knapp (1960) found a similar condition in pseudoobscura when they were able to show a considerable reduction in the num­ The University of Texas Publication her of viable eggs produced by females heterozygous for inversions in two or more chromosome pairs. The results of experiments discussed in these two papers could be of significance in view of the fact that female hybrids of the light and dark forms of Tutuila always have at least two major chromosomes, XL and 3R, heterozygous for inversions, with chromosome 2 also frequently heterozygous in either or both of its arms. The interchromosomal effect of heterozygous inversions on the frequency of crossing over in non-homologous pairs of chromosomes has been a frequently studied phenomenon. It has been found that in most cases a reduction in crossing over in one pair of chromosomes, because of inversion heterozygosity, will result in an increase in crossing over in other pairs of chromosomes in that cell. Carson (1952) has shown in D. robusta that inversions present in each arm of a meta­centric chromosome produce a strong intrachromosomal effect suppressing cross­ing over in the portion of the chromosome between the two inversions. However, the presence of heterozygous inversions in other chromosome pairs may alter this effect and intensify crossing over in this portion. If crossing over can be increased outside of inversions such as these, perhaps it can also be increased within the regions of the inversions. Such an effect would have no significance in the case of single paracentric inversions. However, as Sturtevant (1938) has pointed out, crossing over within a pair of overlapping inversions would produce the same results that crossing over within a pericentric inversion does. Recombi­nant chromosomes would have duplications and deficiencies of genetic regions and yet still be included in egg nuclei. Eggs with these duplication-deficiency chromosomeswould probably produce inviable zygotes. The foregoing has all been conjecture which has yet to be tested thoroughly in laboratory experiments. Such speculation has been prompted by observations made in the current study which indicated very little hybrid sterility but rather strong genetic behavioral isolation as seen in some of the preferential mating tests (see Tables 10 and 11 ). TABLE 11 Results of 4-way multiple-choice experiments in which five of each type were placed in single vials with two of each type ~ . At prescribed intervals of time were isolated and allowed to lay eggs. The resultant offspring were examined and the frequency of heterogamic matings measured by the absence or presence of hybrids. Temperature: 25°C Percent No. with fertile 'i' 'i' Type 'i' 'i' Interval of time No. 'i' 'i' No. fertile hybrid F1 producing hybrids Light 24hours 100 92 0 0.0 48 hours 100 96 1 1.04 72 hours 100 95 0 0.0 96 hours 100 96 3 3.12 Dark 24hours 100 85 0 0.0 48 hours 100 86 0 0.0 72hours 100 93 1 1.08 96 hours 100 93 2 2.15 Futch: Drosophila ananassae Populations Although the majority of the behavioral isolation mechanisms probably devel­oped while the two forms were geographically isolated, it seems reasonable to assume that some have developed, or at least have become reinforced, in response to the pressures of natural selection following the initiation of geographic sym­patry. There is some evidence in support of the idea that behavioral isolation between the two Samoan forms has been strengthened by contact between the two populations. Dr. Takashi Narise (1965 and this bulletin) has observed a peculiar behavioral phenomenon which occurred when both light and dark indi­viduals were placed together in connected migration tubes. There seemed to be a repulsive reaction, for the light flies were seen to migrate away from the center of the complex of tubes, while the dark flies became concentrated in the center. When each type was placed in the same kind of system with individuals from allopatric populations of ananassae no reaction of such severity was noted. The apparent importance of temperature to the effective maintenance of sexual isolation is also indicative. The temperature range on Tutuila is from 23.2°C to 29.2°C. In this study, sexual isolation with respect to light males was found to be weak in temperatures of 22°C and lower. However, at 25°C both forms were found to exercise considerable choice in the selection of mates. This response to temperature differences is similar to that found by Mayr and Dobzhansky ( 1945) with respect to persimilis males in multiple-choice experiments with members of that species and pseudoobscura. It was noted, moreover, that at the lower temperature light flies, especially males, were much less active. The degree of activity probably has considerable influence on the ability to exercise mating preference. Mayr (1946) has listed degree of activity as one of three factors controlling what he calls the "ratio of preference." Another of these three factors is species recognition and attraction. The back­crosses of progeny of F1 male backcrosses indicate that recognition and stimulus characteristics are acquired genetically and that some, but not all, may be sex­linked. ~uch a situation was found to exist (Tan, 1946) with respect to the inheritance of sexual preference by pseudoobscura and persimilis. The possibility of color difference playing a part in mating preference has not been tested yet. However, if it does, it would seem that this recognition character would only pertain to the light form, since the dark flies have been seen to mate quickly and frequently with light ananassae from Majuro and Truk. The evidence gathered thus far from observing and experimentally testing these two very similar forms of Drosophi.la from Tutuila indicates that the two represent valid biological species. The fact that they do not seem to cross in nature provides the best evidence for this opinion. The dark form is undoubtedly a geo­graphic race of the polytypic ananassae. The high frequency of fertile crosses between this form and other ananassae plus the gene arrangements shown by the salivary gland chromosomes indicate that this is true. The light form, on the other hand, since it is chromosomally different and reproductively isolated from ananassae can logically be considered to be another species. If the iight form is another species, there appears to be an available name already in existence. In 1933, Malloch described a fulvous yellow fly from the Marquesas Islands of eastern Polynesia which he named Drosophila errans. In The University of Texas Publication 1934, he reported that he had seen the same fly from the Samoan Islands, and stated that it might be the same as Drosophila ananassrze. Since that time, the name errans has been considered as a synonym of ananassrze. Thus, the situation on Tutuila seems to be resolved by the recognition of two very closely related species which inhabit this island. Still to be answered is the question of the relationships existing between these flies and the New Guinea forms used in these tests. This question can probably be cleared up by noting that the New Guinea population also possesses characteristics suggesting that it is a species which is closely related to both of the two Samoan forms, but at the same time is quite different from either of them; e.g., the indication of genetic isolation as reflected by gene arrangements in the salivary gland chromosomes. These three species might be thought of as forming some sort of a nucleus for the study of a probable complex of ananassae-like forms from Australia, New Guinea, and the numerous outlying islands comprising the Bismarck Archipelago, the Solomon Islands, etc. SUMMARY 1. Light and dark pigmented flies, both apparently forms of Drosophila ananassae, have been found in the same collections from the South Pacific Island of Tutuila, American Samoa. Initial tests showed that fertile males and females of intermediate phenotype could be produced by crossing the two forms. How­ever, the lack of such intermediate forms in the collections, plus certain consistent cytological dissimilarities apparent in the salivary gland chromosomes of hybrid larvae, indicated that such interbreeding does not occur in nature. 2. An examination of stocks representing various geographical populations of ananassae indicates that the species is polytypic with respect to abdominal pig­mentation. Pigmentation types range from flies which are pale yellow to others which were almost totally black. A division of the species into races is proposed partly on the basis of the geographic distribution of these pigmentation types. ~ The very light forms of the islands of Micronesia and the very dark forms of Polynesia constitute two of these races. The third race is composed of populations of ananassrze from all parts of the heavily populated tropical and sub-tropical world. This is the '"cosmopolitan" race of the species, variable to some extent in abdominal pigmentation, but usually in possession of three characteristic para­centric inversions (2LA, 3LA, and 3RA). 3. An examination of mito'tic figures in larval neuroblasts revealed that light and dark forms from Tutuila have the same general karyotypes. In each case X-chromosomes are medium size metacentric, two pairs of autosomes are large metacentric, and one pair of autosomes is small and metacentric. The Y-chromo­some is J-shaped. 4. The salivary gland chromosomes from larvae of twenty-eight of thirty pair­matings of the dark flies from Tutuila were homozygous for the standard gene arrangement of ananassrze. One pair-mating of dark flies produced larvae hetero­zygous for a basal inversion in 3R and one pair produced one larva heterozygous for an asymmetric pericentric inversion in chromosome 3. On the other hand, larvae produced by twenty-six of thirty pair-matings of light flies from Tutuila were heterozygous for a pair of overlapping inversions in 2R (·2.RA and 2RB) , Futch: Drosophila arumassae Populations larvae from five of the thirty matings were heterozygous for a pair of overlapping inversions in 2L ( 2LC and 2LD) and from three of thirty for a small single inversion in 2L (2LB). These inversions were found frequently heterozygous in hybrid larvae produced by crosses between dark and light flies. In addition, these hybrid larvae were always heterozygous for an inversion in 3R (3RB) and, if female, for an inversion in XL (XLA). Both of these inversions were found to be homozygous in light flies. 5. Larvae produced by crosses between the two forms from Tutuila and flies from other stocks of ananassae showed that all except those from Papua, New Guinea, have the same basic gene arrangement as the dark form. Two stocks from Papua were used in these crosses and both were found to be homozygous for XLA and one of the two homozygous for 3RB. In addition, the chromosome 2 inversions-2RA, 2LB, and 2LC-which were found in the Tutuila light form were also present in these flies. This, plus the presence of various overlapping in­versions in each form, indicate an evolutionary relationship between the light Tutuila flies and the Papuan flies, probably through a common ancestor not shared by the other ananassae investigated. The dark form is obviously cyto­logically much more similar to other ananassae than to flies of these two popu­lations. 6. The results of pair-matings between the two Tutuila forms revealed that fertility is generally obtained if insemination is accomplished. Similar results were obtained for pair-matings between each of these two forms and flies from other geographic populations. Is was easily seen that the dark flies had far greater success with respect to the frequency of insemination than the light flies except when the matings involved Papuan flies. Here again it was indicated that the dark Tutuila flies are more similar to other ananassae and that the light Tutuila flies and the Papuan flies are somehow related in a special way. 7. The two multiple-choice preferential mating tests indicated that at temper­atures similar to those occurring on Tutuila, strong preferential mating behavior acts to help maintain sexual isolation. However, isolation breaks down at lower temperatures, especially in matings when light males are involved. 8. Backcrosses of the progeny of F1 light-dark hybrid male back-crosses indi­cated that the ability to recognize and to be recognized as either light or dark is hereditary, much of it sex-linked. The mechanisms involved also seemed to be partly influenced by the ratio of dark and light autosomes present in the hybrid. It is also possible some of this ability is determined by the egg cytoplasm. 9. The relationships suggested by the results of these investigations are that the dark form on Tutuila is part of the dark Polynesian race of ananassae and that the light form is a second species which is very closely related to ananassae. A hypothesis is constructed which attempts to account for the evolution of the light species from a part of a geographically isolated segment of ananassae. Since the light species and flies from Papua apparently have had evolutionary experi­ences in common with one another but unique with respect to the rest of ananassae, it is supposed that this light species had a particular ancestor in common with the Papuan population of ananassae-like flies. It is suggested that the light form probably inhabited Tutuila much earlier than the dark form. When the dark form first arrived, it is supposed that some The University of Texas Publication interbreeding may have occurred initially. However, if hybridization did occur, offspring produced by these crosses must have been inferior in some way un­detected so far in the laboratory, for mechanisms for selective mating behavior apparently were selected for and reinforced. 10. It is pointed out that there appears to be an available name already in existence for the light species from Tutuila. Drosophila errans Malloch, pre­viously considered as a synonym of Drosophila ananassae, may be the taxonomi­cally correct name for this fly. In addition, the Papuan forms apparently repre­sent a third species which has yet to be named. The suggestion is made that these three species may represent a part of a complex of ananassae-like species which have evolved in the Australia-New Guinea area. ACKNOWLEDGMENTS The author wishes to express his gratitude to Dr. Marshall R. Wheeler and Dr. Wilson S. Stone for supervising the majority of this work as a doctoral dis­sertation and for arranging the collecting trip to Hawaii, Samoa, and other Polynesian Islands. He is also indebted to Dr. Harrison D. Stalker for his advice concerning techniques for photographing salivary gland chromosomes, to Dr. Costas D. Kastritsis for his cooperation in developing these techniques, to Mrs. Linda Kuich for the diagrammatic illustrations, and to Mrs. Florence D. Wilson and Mrs. Virginia Gerstenberg for assistance with technical problems. REFERENCES Bryan, E. H . 1951. Central and western Polynesia. In 0. W. Freeman, ed., Geography of the Pacific. John Wiley and Sons, Inc., New York. 3~22. Carson, H. L. 1952. The effects of inversions on crossing over in Drosophila robusta. Genetics, 38: 168-186. ----. 1954. lnterfertile sibling species in the willistoni group of Drosophila. EYolution, 8: 148-165. --­-. 1959. Genetic conditions which promote or retard the formation of species. Cold Spring Harbor Symposia on Quantitative Biology, Volume XXIV: 87-105. Cooper, K. W ., S. Zimmering, and J. Krivshenko. 1955. lnterchromosomal effects and segrega­tion. Proc. Nat. Acad. Sci., 41: 911-914. Dobzhansky, Th. 1944. Chromosomal races in Drosophila pseudoobscura and Drosophila persimilis. In Dobzhansky and Epling, Carn. Inst .Wash. Pub. 554: 47-144. ----. 1951. Genetics and the Origin of Species. 3rd edition, revised. Columbia Univ. Press, New York. Dobzhansky, Th., and A. Dreyfus. 1943. Chromosomal aberrations in Brazilian Drosophila ananassae. Proc. Nat. Acad. Sci., 29: 301-305. Harrison, R. A. 1954. Some notes on and additions to the Drosophilidae of Samoa and Fiji. Trans. R. Ent. Soc. London, 105(6): 97-116. H eed, W. B. 1962. Genetic characteristics of island populations. Univ. of Texas Puhl. 6205: 173-206. Freire-Maia, N. 1961. Peculiar gene arrangements in Brazilian natural populations of Dro­sophila ananassae. Evolution, 15: 486-495. Kalmus, H. 1941a. Physiology and ecology of cuticle colour in insects. Nature, 148: +28-431. Futch: Drosophila ananassae Populations 1941b. Resistance to desiccation of Drosophila mutants affecting body colour. Proc. Roy. Soc. London (B), 130: 185-201. Kaufmann, B. P. 1937. Morphology of the chromosomes of Drosophila ananassae. Cytologia, Fujii Jubilee Vol., pp. 1043-1055. Kikkawa, H. 1938. Studies on the genetics and cytology of Drosophila ananassae. Genetica, 20: 458-516. 1939. Further studies on the genetics and cytology of Drosophila ananassae. Cytologia, 9: 452--459. Koopman. K. F. 1950. Natural selection for reproductive isolation between Drosophila pseudo­obscura and Drosophila persimilis. Evolution, 4: 135-148. Malloch, J. R. 1933. Some acalyptrate Diptera from the Marquesas Islands. Bull. Bishop Mus., 114: 3-31. 1934. Insects of Samoa and other Samoan terrestrial Arthropoda (Diptera). British Mus. (Nat. Hist.), Pt. 6, Fasc. 8: 267-328. Mayr, E. 1942. Systematics and the Origin of Species. Columbia Univ. Press, New York. 1946. Experiments on sexual isolation. VIL The nature of isolating mechanisms between Drosophila pseudoobscura and Drosophila persimilis. Proc. Nat. Acad. Sci., 32: 128-137. 1963. Animal Species and Evolution. Harvard Univ. Press, Cambridge. Mayr, E., and Th. Dobzhansky. 1945. Experiments on sexual isolation in Drosophila. IV. Modification of the degree of isolation between D. pseudoobscura and D. persimilis and sexual preference in D. prosaltans. Proc. Nat. Acad. Sci., 31: 75-81. Moriwaki, D., M. Ohnishi (Shirai), and Y. Nakajima. 1956. Analysis of heterosis in popula­tions of Drosophila ananassae. Proc. Intern. Genet. Symp., Cytologia Suppl. Vol.: 370--379. Moriwaki, D., and Y. N. Tobari. 1963. Maternal effects and heterosis in Drosophila ananassae. Genetics, 48: 171-176. Narise, T. .1965. The migration of D. ananassae under competitive conditions. Drosophila Info. Service, 40: 84. Narise, T. The mode of migration of Drosophila ananassae under competitive conditions. This Bulletin. Patterson, J. T., L. W . McDanald, and W. S. Stone. 1947. Sexual isolation between members of the. virilis group of species. Univ. of Texas Puhl. 4720: 7-31. Patterson, J. T., and W . S. Stone. 1952. Evolution in the Genus Drosophila. Macmillan Com­pany, New York. Spieth, H. T. 1958. Behavior and isolating mechanisms. In A. Roe and G. G. Simpson, eds., Behavior and Evolution. Yale Univ. Press, New Haven. 363-389. Stalker, H. D. 1942. Sexual isolation studies in the species complex of Drosophila virilis. Genetics, 27: 238-257. Stone, W. S. 1962. The dominance of natural selection and the reality of superspecies (species groups) in the evolution of Drosophila. Univ. of Texas Puhl. 6205 : 507-537. Stone, W. S., M. R. Wheeler, W. P. Spencer, F. D. Wilson, J. T. Neuenschwander, T. G. Gregg, R. L. Seecof, and C. L. Ward. 1957. Genetics studies of irradiated natural populations of Drosophila. Univ. of Texas. Puhl. 5721: 260-316. Stone, W. S., M. R. Wheeler, and F. D. Wilson. 1962. Genetic studies of irradiated natural populations of Drosophila. V. Summary and discussion of tests of populations collected in the Pacific Proving Ground from 1955 through 1959. Univ. of Texas Puhl. 6205: 1-54. Strickberger, M. W. 1963. Evolution of fitness in experimental populations of Drosophila pseudoobscura. Evolution, 17: 40--55. The University of Texas Publication Sturtevant, A. H. 1916. Notes on North American Drosophilidae with descriptions of twenty­three new species. Ann. Ent. Soc. Amer., IX: 323-343. 1938. Essays on evolution. III. On the origin of interspecific sterility. Quart. Rev. Biol., 13: 333-335. Tan, C. C. 1946. Genetics of sexual isolation between Drosophila pseudoobscura and Dro­sophila persimilis. Genetics, 31: 558-573. Terzaghi, E., and D. Knapp. 1960. Pattern of chromosome variability in Drosophila pseudo­obscura. Evolution, 14: 347-350. Tobari, Y. N. 1962. Heterosis relating to a terminal inversion in artificial population of Dro­sophila ananassae. lap. Jour. Genet., 37: 302-309. Wagner, R. P. 1944. The nutrition of Drosophila mulleri and D. aldrichi. Growth of the larvae on a cactus extract and the microorganisms found in cactus. Univ. of Texas Puhl. 4445: 104-128. Wasserman, M. 1962a. Cytological studies of the repleta group of the genus Drosophila. III. The mercatorum subgroup. Univ. of Texas Puhl. 6205: 63-71. 1962b. Cytological studies of the repleta group of the genus Drosophila. IV. The hydei subgroup. Univ. of Texas Puhl. 6205: 73-83. 1962c. Cytological studies of the repleta group of the genus Drosophila. V. The mulleri subgroup. Univ. of Texas Puhl. 6205: 85-117. 1962d. Cytological studies of the repleta group of the genus Drosophila. VI. The fasciola subgroup. Univ. of Texas Puhl. 6205: 119-134. 1963. Cytology and phylogeny of Drosophila. American Naturalist, 97: 333-352. IV. The Mode of Migration of Drosophila ananassae Under Competitive Conditions.1 2 TAKASHI NARISE INTRODUCTION According to Patterson and Stone ( 1952), Drosophila ananassae is distributed widely in tropical and sub-tropical areas. In such a widely distributed species, it is expected that there is an abundance of genetic types in competition in popu­lations with some flow of migrants. Under competitive conditions, what is the pattern of this migration? Although there are many problems in relation to migration, one of the more important is the interaction among genotypes in the species under competitive conditions. With regard to migration of Drosophila, Sakai et al. (1958) and Narise (1962) found that the migratory activity of Dro­sophila melanogaster is under genetic control. However, the experiments were made using single isolated strains, not mixed strains. In a polymorphic popula­tion, the mode of migration may be quite different from the migration in single strains, even if ea(:h strain in the mixture has its own genetic migratory activity. Narise (in preparation) has conducted competition experiments with two geno­types of D. melanogaster in an open and a closed population. The experimental results showed that the migratory activity of a weaker competitor was stimulated by the strong competitor and the stimulation of migratory activity prevented the elimination of the weaker competitor in the open population. The problem arises, then, whether such stimulation of migratory activity occurs in other species under competitive conditions. To attack this problem, an experiment was conducted with four strains of Drosophila ananassae, collected on islands in the South Pacific. MATERIALS AND METHODS Four strains of Drosophila ananassae collected by Dr. Wilson S. Stone and Dr. Marshall R. Wheeler on some islands of the South Pacific were used in this experiment. Two of them were so-called "light ananassae" which has yellow body color, and were collected in Pago Pago, Tutuila, American Samoa, and Majuro, Marshall Islands. The other two strains were "dark ananassae" having black body color, and were collected in Pago Pago, Tutuila, American Samoa, and Rarotonga, Cook Islands. 1) Migratory activity of two kinds of ananassae in a mixed population. One hundred flies, consisting of light and dark ananassae in different frequen­cies were introduced into a migration-tube and the tube (called "original tube") was kept 2.4 hours before three fresh tubes were connected with it. Flies which migrated to the three new tubes were wunted after 6 hours. The relative fre­ 1 This investigation was supported by Public Health Service Research Grants No. GM-06492 and GM-11609 from the National Institutes of Health. 2 Present address: National Institute of Genetics, Mishima, Shizuoka-ken, Japan. The University of Texas Publication quency of the dark fonn in different experiments was 0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0. 7, 0.8, 0.9, and 1.0, and each relative frequency was replicated 10 times. The migratory activity for a strain at each relative frequency presented is calcu­ lated as the per cent of that strain in the new tubes to that strain in the original mixture. 2) The mode of migration for dark and light ananassae in mixed populations. Seventy-five pairs of light and seventy-five pairs of dark ananassae were intro­duced into the original tube and another fifteen fresh tubes were connected with the original tube as shown schematically in Figure 1, after 24 hours. Two tubes which were connected directly with the original tube were called the first-step­tubes, and the four tubes which were connected with the two first step-tubes were called the second-step-tubes, and so on. Only one tube was used for the fourth step-tube. The migrant flies were counted 8 hours after the connection, and the number of both strains which migrated to each step-tube was counted. The experi­ment was replicated five times for each combination. FIG. 1. The arrangement of 16 migration tubes. All experiments were conducted at 25° ± 1°C in a dark room. The following abbreviations were used for the four strains:* Dark ananassae from Pago Pago D-pp Dark ananassae from Rarotonga D-rar Light ananassae from Pago Pago L-pp Light ananassae from Majuro L-maj EXPERIMENTAL RESULTS 1) The migratory activity of dark and light ananassae in mixed population. Table 1 and Figure 2 show the migratory activity of dark and light ana.nassae at each relative frequency, and Table 2 shows the results of the analysis of vari­ance on the migratory activities in the four different combinations. a) Migratory activity of L-pp and D-pp strains in a mixed population. The migratory activity of the original D-pp strain is 33.48% and that of L-pp is 34.45%. Much difference regarding the migratory activity between the two strains under the single condition is not detected. However, the activity of D-pp TABLE 1 The migratory activity of dark and light ananassae under competitive condition Relative frequency of dark ananassae Combination 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 L-pp and 34.45 39.11 39.81 41.43 41.42 41.60 42.00 44.50 38.Z5 44.00 D-pp 24.50 24.75 25.33 22.38 30.05 26.67 25.21 28.69 31.44 33.48 L-ppand 34.45 32.83 36.75 37.29 36.50 36.15 32.84 39.50 33.00 41.50 D-rar 24.50 30.00 25.83 26.13 27.60 Z5.17 30.71 29.99 34.44 33.10 L-maj and 36.78 41.48 42.88 32.93 32.75 40.40 37.38 36.83 34.00 35.50 D-rar 35.50 32.00 28.83 32.50 34.40 39.42 40.29 31.38 32.78 33.10 L-maj and 36.78 37.11 42.94 30.80 37.67 37.40 35.38 33.42 41.25 39.50 D-pp Z5.00 29.25 26.77 28.63 29.30 23.00 24.29 23.08 24.22 33.48 TABLE 2 Analysis of variance on migratory activity of dark and light ananassae under competitive condition M .S. Source of deviation d.f. D-pp and L-pp D·pp and L-maj D-rar and L-pp D-rar and L·maj Between strains 1 7243.67** 4575.03** Z599.71 ** 368.03** Between frequency within strain 18 113.72 88.95 220.16** 126.54** Between single and mixed 2 508.45** 162.15* 121.82 65.02 Between frequency within mixed 16 64.38 79.80 232.45** 134.42** Error 380 87.88 83.98 65.96 59.02 • Significant at 5% level. •• Significant at 1 % level. • Editor's note: Futch (This Bulletin) shows that L-pp is probably a sibling species of D. ananassae; L-maj, D-pp, and D-rar are true ananassae. The University of Texas Publication 40 ..... L-pp ond D-pp ......... ...... .... /4., ..... 30 I ' ..... .­ ..... I '•-..... / -.. e------e..., I .........., 40 L-pp ond D-rar >­ I­ ;20L--~-'--~--'-~-'-~--'~~-'--~--'-~-'-~----L~~'"--~~ light I-30 ............. / / L-pp ~ L-maj > D-pp Mass-migratory activity L-pp = L-maj > D-rar > D-pp It is probable that the migration to the first step-tubes from original tube occurs due to mass-migration, and random-migration takes place in the outer tubes with reduced population density. Then, the migration of light ananassae in the D-pp/ light ananassae combination, takes place due to the stimulation in the mass-and random-migratory activity, with the result that light ananassae strain is superior in the number of migrant flies in all tubes. In D-rar and L-pp as well as the D-rar/ L-maj combination, a smaller number of flies of the D-rar strain migrates to the first step-tubes because of lower mass-migratory activity under competitive con­ The University of Texas Publication dition. However, the strain has higher random-migratory act1v1ty than light ananassae strain, resulting in the excess of that strain in the surrounding tubes. From this point of view, the relation between mass-and random-migratory activ­ity should be another important factor in migration under competitive condition. Needless to say, there are many factors controlling migration aside from strain interaction. However, the interaction among strains which exist in the population seems to be one important factor. As pointed out by Lidicker (1962), the immi­grant spreads his genetic materials more widely and there would be more oppor­tunity for advantageous recombination of genetic materials. In such a situation, it is expected that such new recombinations of genes may bring about new stimu­lating effects on migratory activity. However, in a genetically heterogeneous and open population, the mode of migration is complicated so that the interaction among strains becomes important. From this standpoint, migration should be determined not only by genetic migratory activities of strains in isolation, but also by the genetic structure of the population. SUMMARY 1. Four strains of D. ananassae collected on islands of the South Pacific were used in this experiment. Two of them were light ananassae and two were dark ananassae. Z. The migratory activity of light ananassae is stimulated by dark ananassae, but dark ananassae is reduced in activity when coexisting with light forms. The rate of stimulation or diminution of activity depends on the particular strains that coexist in the population. The migratory activity of the strain is also affected by the relative frequencies of strains as shown in the D-rar/ L-pp combination as well as the D-rar/L-maj combination. However, no relation between migratory activity and relative frequency was detected in the combination between D-pp/ L-pp or D-pp/ L-maj. 3. The light ananassae dominates in all tubes when mixed with D-pp. How­ever, dark ananassae dominates in the surrounding tubes in the combination between D-rar and light ananassae strains. This is probably due to high random­migratory activity in the D-rar strain under competitive conditions. 4. Migration under competitive conditions is dependent on the strains which coexist in the population, and the interaction of the strains is a very important factor in migration. ACKNOWLEDGMENT The author is very grateful to Professor Wilson S. Stone for his helpful sug­gestions and comments, and critical reading of this manuscript. The author is also grateful to Dr. Richard C. Lewontin, Department of Zoology, University of Chicago, for his valuable comments. I wish to thank Mr. David E. Briles, Mr. Jay P. Mumma, and Mr. Stephen Goldfarb for their help throughout the experiments. LITERATURE CITED Lidicker, W . Z ., Jr. 1962. Emigration as a possible mechanism permitting the regulation of popu­lation density below carrying capacity. Amer. Nat. 96(886): 29-33. Narise, T. 1962. Studies on competition in plants and animals. X. Genetic variability of migratory activity in natural population of Drosophila melanogaster. Jap. J. of Genet. 37(6): 451-461. Narise, T. (m preparation). The competition between wild and vestigial flies of Drosophila melanogaster in an open and a closed population. Patterson, J. T., and W . S. Stone. 1952. Evolution in Genus Drosophila. The Macmillan Company. p. 51. Sakai, K. I., T . Narise, Y. Hiraizumi, and S. lyama. 1958. Studies on competition in plants and animals. IX. Experimental studies on migration in Drosophila melanogaster. Evolution 12(1): 93-101. V. Mating Behavior of D. ananassae and ananassae-Iike Flies from the Pacific1 2 HERMAN T, SPIETH INTRODUCTION As a part of their extensive investigations of the drosophilids of the Pacific Basin, Stone and Wheeler and their colleagues (see Stone et al., 1957; Stone et al., 1962; Stone et al., this Bulletin; and Futch, this Bulletin) have collected and sub­sequently established at the University of Texas Genetics Foundation a large number of stocks of D. ananassae Doleschall from various localities in the Pacific area. These stocks show considerable variation, especially with respect to over-all coloration and sexual isolation. Particularly interesting is the fact that two dis­tinct populations of flies, a light (yellow) one and a dark (black) one, occur on Tutuila, American Samoa (see Stone et al., and Futch, this Bulletin) . Fifteen Pacific Basin stocks (Table 1) were selected for study with the hope that it might be possible to find visible differences between their courtship patterns. TECHNIQUE Subcultures of each of the stocks were established and from them known vir- TABLE 1 Locality of stocks utilized Strain Uni.v. of Texas General symbol Collection locality stock no. coloration AUS Northern Queensland, Australia 2372.11 Very light (yellow) P01 Popondetta, Papua, New Guinea 3021.3 Very light (yellow) BR1 Brown River, Port Moresby, Papua, New Guinea 3020.9 Very light (yellow) P02 Popondetta, Papua, New Guinea 3021.2 Light (yellow) BRZ Brown River, Port Moresby, Papua, New Guinea 3020.8 Light (yellow) PP1 Pago Pago, Tutuila, American Samoa 3038.1 Light (yellow) PP2 Pago Pago, Tutuila, American Samoa 3038.2 Dark (black) NIU Niue Island, Friendly Islands 3039.2 Dark (black) SUV Suva, Fiji Islands 3039.3 Dark (black) RAR Rarotonga, Cook Islands 3036.2 Dark (black) TON Tonga, Friendly Islands 3039.4 Dark (black) PA1 Palmyra Island 3034.1 Dark (grey) PAZ Palmyra Island 3034.1 Dark (grey) MAJ Majuro Atoll, Marshall Islands 2370.9 Light (grey) RON Rongerik Atoll, Marshall Islands 2401.11 Light (grey) 1 Supported (in part) by a PHS research grant (GM-11609) from the National Institutes of Health, Public Health Service, and Grant (NSF) GB-711 from the National Science Foundation. 2 Guest Investigator, The University of Texas, December 1964 to June 1965; present address: Department of Zoology, The University of California, Davis, California. The University of Texas Publication ginal adults were collected, etherized, and the sexes separated. The separated flies of each sex were then placed into 8-dram isolation vials and supplied with the same type of cornmeal food media on which they had been reared. After they had reached full sexual maturity, the males and females were introduced into an observation cell without etherization and were observed with the aid of a zoom­type dissecting microscope at magnifications of 7X to 30X . The observation cell, constructed of clear plastic, was rectangular in shape. The sides, top and bottom of the cell had inside dimensions of approximately 25 X 25 X 63 mm ( 1 X 1 X 2.5 in.). Both ends of the cell were covered with lightly moistened 25 x 25 mm. pieces of cellulose sponge. One of these sponges was removable and not only served as the "stopper" for the cell but also could be moved back and forth, thus easily changing the internal volume of the cell. When small numbers of individuals (5 to 10) were utilized, the internal dimensions of the cell were adjusted to approximately 25 X 25 X 10 mm., while with larger numbers (20 to 30 individuals) the dimensions were usually about 25 X 25 X 25 mm. During all observations, the flies had available, irrespective of total volume, two vertical 25 X 25 mm. walls of cellulose sponge on which they preferentially adhered and moved about easily. Since the flies evinced apparent irritation when subjected to illumination of the intensity which was produced by a microscope light, accessory lighting was not used and only the normal room illumination from overhead neon lights was employed during the observation periods. All phases of the research were performed in the same laboratory room in which the parent stocks are kept. This laboratory is maintained at approximately 70°F. The relatively low temperature slowed the maturation of the adults and therefore most observations were made upon individuals that were 8 to 20 days old. Typically all individuals were discarded at the end of an observation period, but occasionally specimens were retained for further use at a later time. In such cases the individuals were etherized and returned to isolation vials. Two types of observational procedures were used. The most common technique was to introduce a number of males and females into the observation cell, observe and record the behavior of the flies for a period of a half hour. The males and females could, of course, either be derived from the same stock or one sex might come from one stock and the other from another stock. The second type of observation consisted of introducing males of stock A with females of stock B into the observation cell and observing them under magnifica­tion and recording their behavior for a period of 15 minutes. At the end of this period, the flies (A o o and B 'i' 'i') were shaken into 35 x 100 mm. glass shell vials, partially lined with lightly moistened blotting paper. The vial was then plugged with cotton and the insects retained on the desk where they could easily be observed, but not on the microscope stage. The reciprocal cross was then intro­duced into the observation cell (stock A 'i' 'i' X stock B o o ) . These flies were ob­served for 15 minutes and at the end of the period they were added to the A o o X B 'i' 'i' in the 35 X 100 mm. glass vial. The combined individuals of both stocks were then observed for 30 minutes under magnification. Since many of the stocks are distinctively pigmented, it was possible to differentiate not only the in­dividuals but also the sexes of two such stocks when mixed together and thus com­ Spieth: Mating behavior of D. ananassae pare the behavior of the different males and females not only with respect to the individuals of the opposite sex of their own stock but also simultaneously the in­dividuals of another stock. Because of the smallness and delicacy of many of the flies and because of the low light intensity under which the observations were conducted, it was not deemed feasible at this time to mark the individual flies by wing-cutting or other marks so that they could be separated with certainty in the case of those stocks which can not be distinguished by normal differences in pigmentation. CouRTSHIP OF FLIES The basic courtship pattern of D. ananassae was described by Spieth (1952) utilizing University of Texas stock No. 1836.3, from Monterrey, Mexico, and is as follows: The male taps, moves to the rear of the female, orienting himself facing close to the tip of her abdomen. He then postures by curling the tip of his abdomen under, slightly crouches, usually twists his body slightly on its longitudinal axis and intermittently engages in pulses of wing vibrations. Periodically, the postur­ing male lunges forward and upward onto the female, grasping her abdomen with his fore tarsi, pushing against her under wing surfaces with his head and at the same time curling his abdomen still farther forward and attempting to bring his genitalia into contact with those of the female. Pulses of wing movement may occur independently of a lunge, but the lunge is always accompanied by the characteristic wing action. Receptive females spread their wings as the male achieves genitalic intromis­sion and the male then immediately pulls himself forward between the female's wings by means of his fore legs and grasps the dorsal surface of the basal portion of her wings with his fore tarsal claws. During copula, the pair is quiet until about 1 to 1 ¥2 minutes before termination when the male releases his fore tarsal grasp, pushes his body backwards so that his wing tips are against the substrate and his body arches upward and somewhat posteriorly from the point of genitalic union. At the same time the female places the tip of her metathoracic tibiae against the point of genitalic union and pushes vigorously. The male turns 180° and with­draws at the end of copula, or infrequently withdraws and then turns to dis­mount. Non-receptive females repel the lunging males by kicking, fluttering the closed wings in rapid but small amplitude, sometimes extruding, sometimes curling the tip of the abdomen downward and decamping. Males may circle about non-receptive females. They also court one another, attempt to copulate with each other, and countersignal by fluttering their wings, depressing the abdomen and kicking. In lunging upon the non-receptive female, the male will sometimes achieve full contact between his genitalia and those of the female. In a small fraction of such cases the male apparently is able to achieve partial intromission of the geni­talia. The non-receptive female immediately reacts violently by kicking, running about, and buzzing both wings at a very rapid rate. As a result, she eventually The University of Texas Publication dislodges the male who is never able to achieve the normal copulatory position. In one pair of stocks, AUS and P01, abnormal copulations of this type did assume normal character after a period of time and were completed. Unlike their close relatives, for example D. melanogaster, D. simulfms, and D. auraria, whose females when virginal and sexually mature will often accept the first mounting overture of a male, none of the females of any of these 15 ananassae stocks accepted a male the first time he lunged. Two individuals were observed to accept the male's overtures at the time of his third successive lunge. Typically, the female would not accept until a male or successive males had re­peatedly lunged, and had made several to many genitalic contacts. Often a female would extrude so as to prevent genitalic contact for some time and then suddenly allow the male to. achieve copulation. Furthermore, in none of the observations that were made did all of the females or even a majority accept the males' over­tures. The typical pattern for those experiments during which copulations did occur was that shortly after the flies had been introduced into the observation cell, two or three copulations would occur, especially in the intra-stock experiments. After these, an occasional copulation would occur later during the observation period, usually 10 to 20 minutes after the start of observations. The 15 stocks studied all conform to the basic courtship pattern described above but for some stocks there are discrete, qualitative differences of male wing move­ments and circling actions that serve as identifying characteristics. Thus the 15 stocks can be divided into groups on the basis of the wing action of the males during courtship. The commonest pattern is for the posturing male to spread both wings laterally from the resting position about 5° to 7° and then vibrate both of them very rapidly in a vertical amplitude of less than 1 mm. If the male, while posturing, also rotates his body slightly along the longitudinal axis, then the "upper" wing appears to vibrate through a greater arc of amplitude than does its opposite member. This type of wing vibration was displayed by the following stocks: RON,MAJ,PA1 , PA2, TON,RAR,SUV,NIU,andPP2. Another type of male wing movement consists of the posturing male extending one wing 50° to 90° away from the resting position and then vibrating the vane up and down in much the same pattern as is displayed by the males of melano­gaster and auraria. The opposite wing remains immobile in the resting position. If the male is not directly behind the female, the wing nearest the female's head is invariably the one that is vibrated. This type of wing display is exhibited by stocks AUS, P01, BR1, andPP1. Finally, stocks BR2 and P02 display a male wing pattern that is intermediate between the two types described above. One wing is extended 20°-30° from the mid-line resting position and both wings are vibrated at the same time although the extended winq appears to move through an arc of greater amplitude than does the non-extended wing. The pattern that the males employ in circling about the females while court­ing shows two distinct types. The most common type is for the male to move with a crab-like motion about the female, constantly facing her and occasionally flick­ Spieth: Mating behavior of D. ananassae ing one or both wings outward. When he is in front of the female, he may arc back and forth for a time before returning to the rear to posture. This is the com­mon circling pattern shown by all the stocks studied with the exception of AUS, P01 and BR1. These three stocks have males that display quite unique circling movements in that as the male circles he stops repeatedly, seemingly attaches his feet firmly to the substrate and then sharply bobs his body upward and forward, extending all of his legs to their greatest possible length. Between bobs he returns to a normal stance, moves laterally a short distance and simultaneously flicks both wings out to 90° and then back to the resting position, following which he bobs up and forward toward the female. On the basis of the duration of copulation, as determined by intrastock matings, the flies can be separated into two discrete groups. Stocks AUS, P01 and BR1 average 11 to 12 minutes (see Table 2) while the remaining stocks, excepting the NIU stock, average 4 to 5 minutes per copulation. The NIU stock displayed an average copulatory time of 6'02" based on four timed observations. Other ob­served copulations of this stock were not timed exactly because either the start or completion of the copulation was missed by several seconds, but the records do indicate that they would fall within the range of 3'50" to 7'5 7". In summary, three of the stocks (AUS, P01, BR1; see Table 1) all pale yellow in color and originally collected from Australia or New Guinea, have males which extend and vibrate one wing and bob as they circle about the females. Nine of the stocks do not bob as they circle, and they vibrate both wings in very small amplitude. These 9 include all the gray or blackish colored stocks. The remaining 3 stocks which are yellowish in color (BR2, P02, and PP1) are unique in that (1) the PP1 males do not bob but they clearly extend a single wing for vibration, while (2) the BR2 and P02 stocks do not bob and have a type of wing action in­termediate between that of the grey and black forms and that of the yellow AUS, P01, BR1 and PPI stocks. Thus those pairs of stocks that were collected at the same time and place (i.e., PP2 and PP1 from Samoa, P02 and P01 from Popondetta, New Guinea, and BR2 and BR1 from Brown River, New Guinea) do display differences in their mating behavior that will allow the observer to separate the males from one an­other. Observations on inter-stock crosses In addition to studying the mating behavior of each of the 15 stocks by intra­ stock observations, i.e., utilizing males and females of the same stock, a number of inter-stock observations were made. Of the possible total of 210 different inter­ stock observations, 65 were actually studied. The greatest emphasis was devoted to the PP1 and PP2 stocks (see Table 3) . The behavior of the various stocks considered individually is given below. Table 3 indicates the number of observations conducted (0), the number of specimens utilized (N), the degree of intensity of the male courting actions (D), and the number of copulations resulting from the inter-stock crosses (C). The first number under symbol N represents male specimens and the second number represents 'female specimens. Under symbol D zero indicates that the male tapped The University of Texas Publication TABLE 2 Copulatory time for inter-and intra-stock crosses No. copulations No. timed lVIinimum Mean ,;\Iaximum AUSo X 'i' 14 5 11'00" 11'36" 12'36" P01 o X 'i' 13 8 9'54" 12'30" 14'54" BR1 o X 'i' 3 2 11'41" 12'45" 13'49" P02o X 'i' 4 3 5'16" 5'45" 6'19" BR2o X 'i' 2 2 4'52" 5'20" 5'47" PP1 o X 'i' 7 4 3'27" 4'15" 5'32" PP2o x 'i' 6 3 3'52" 5'13" 6'04" SUV ox 'i' 19 9 3'54" 5'07" 6'12" NIU o X 'i' 7 4 3'50" 6'02" 7'57" RARo X 'i' 5 1 4'50" TONo X 'i' 14 5 4'31" 5'20" 6'43" PA1 o X 'i' 8 6 3'00" 4'49" 6'18" PA2o X 'i' 8 5 4'26" 4'57" 5'28" MAJox'i' 6 6 4'51" 5'43" 6'42" RONo X 'i' 5 2 4'20" 4'35" 4'50" AUSct; X P01 'i' 2 9'27" P01 o X AUS'i' 7 PP1 o X SUV'i' 5 5 3'32" 4'01" 4'25" PP1 o X TON'i' 1 PP1 ox PA1'i' 2 2 2'48" 4'49" 6'49" PP1 o X PA2'i' 2 4'39" PP1 o X MAJ'i' 2 PPZO X PP1 'i' 2 PP2ct; X NIU'i' 1 1 3'01 " PP2o X SUV'i' 3 2 3'56" 4'03" 4'56" PP2o X RAR 'i' PP2o X TON'i' 4 PP2o X MAJ'i' 6 3 5'13" 5'38 6'26" PP2oX RON'i' 4 2 4'00" 4'45" 5'29" NIU o X PP2'i' 4 3 5'11" 5'45" 6'35" SUVo X PP2'i' 1 4'38" TONo X PP2'i' 2 RONo X PP1 'i' 1 3'53" PA1 ct; X PP2'i' 3 PA1 o X TON'i' 4'57" MAJct; X PA1'i' RON o X PP1 'i' 3'35" RON o X PA2'i' 2'23" TABLE 3 l nterstock crosses '? AUS POI BR! P02 BR2 PP! PP2 NIU SUV RAR TON PAI PAZ MAJ HON AUSo 0 2 1 2 2 2 3 1 1 N 17,24 5,6 20,24 15,20 8,12 30,26 18,18 9,9 D 0/3 0 0 0/1 0/1 0 0 0/1 c 2 0 0 0 0 0 0 0 P01o 0 2 2 N 22,27 13,7 18,25 9,3 11,8 D 3 0 2­ 2­ 0 Spieth: Mating behavior of D. ananassae c 7 0 0 0 0 BR1 3 0 2 N 4,19 9,12 11 ,8 21 ,13 17,18 11 ,5 D 3 2-­ 1/ 2 3/ 2 1 2-­ c 0 0 0 0 0 0 P02 3 0 1 2 2 N 9,17 5,10 7,13 6,10 6,7 4,4 D 0 0 2/ 3 2-­ 2 3 c 0 0 0 0 0 0 BR2 3 0 2 N 8,20 7,10 13,17 3,4 D 0 0 0/ 2 0 c 0 0 0 0 PP1 3 0 2 2 1 2 2 2 1 2 2 21,28 14,21 10,8 20,14 14,14 10,8 18,24 8,5 6,1011 ,22 14,8 12,20 11 ,18 D 1/ 2 2 2/ 3 0 2-­ 1/ 2 2 2 2 2 1 3 3 c 0 0 0 0 0 0 5 0 1 2 2 2 0 PP2 3 0 3 2 2 2 2 1 2 2 1 N 33,24 23,21 9,2810,22 7,13 24,32 3,2 12, 12 13,8 14,28 17,29 22,20 16,11 7,15 D 1/ 2 2/ 3 2/ 3 2-­ 2-­ 3 3 3 3 3 0/ 1 0/ 1 3 3 c 0 0 0 0 0 2 3 4 0 0 6 4 NIU 3 0 2 N 11 ,16 7,7 D 0/ 3 3 c 0 4 SUV 3 0 2 N 16,27 11 ,8 7,7 D 3 3 3 c 0 0 2 RAR 3 0 D c TON 3 0 2 2 1 1 N 13,27 10,111 3,14 8,10 15,3 3,4 D 2-­ 2 2--/ 2+ 3 2 0 c 0 0 0 0 2 0 PA1 3 0 1 N 6,8 7,8 6,6 4,6 6,8 7,9 D 2/ 3 2 2 2/ 1 3 2-­ c 0 0 0 0 3 1 PA2 3 0 2 N 14,30 10,7 D 3 2+ c 0 0 MAJ 3 0 2 2 2 N 8,30 13,3117,12 D 2-­ 2 2-­ c 0 1· 0 RON 3 0 2 2 7,7 20,23 35,15 3,3 D 0 0/ 2-­ 0 3 c 0 1 0 O=no. of observations; N==no. of specimens used ( ~,0); D=degree of cow1:ship intensity; C=no. of copulations observed; further details in text. . • Th_is _was an abnormal copulation, with the O on his back during the entire period; he apparently made a force-type mtrom1Ss10n and the female did not kick him off. The University of Texas Publication and turned away, while 1 means that the male taps, orients behind the female and postures but does not lunge; 2 means that the male taps, orients, postures and lunges; if the 2 is followed by a + it indicates that the male's courting activities increased during the period of observation, while -indicates that they decreased during the period of observation. Absence of the+ or -indicates neither increase nor decrease during observation. A 3 for symbol D indicates that the male during this period courted with the same intensity as he normally does with his own fe­males. Since some males in any single observation are more aggressive than others, the double figures, with a slash between them, indicate such observed dif­ferences. PP2MALES When presented with a no-choice situation of AUS, BR2 or P02 females, the PP2 males did, during the first part of any observation period, tap, orient, posture and lunge at the females. Before the end of a half-hour period, however, most if not all of them had ceased to lunge and had broken the courtship pattern at the tapping or orienting stage. With P01 and BR1 females, the complete repertoire of courtship persisted during the observation periods, but not at a high level of sexual excitement. No copulations occurred with any of the females of these five stocks. The PP2 males courted the PP1 , NIU, SUV, RAR, MAJ, and RON fe­males almost as vigorously as they did their own females, and further copulations did occur between the PP2 males and females of each of these 6 stocks. In com­parison, the PP2 males only rarely went farther than the tapping phase with PA1 and PA2 females. The aversion of the PP2 males to these females is clear and striking. When given a choice between their own PP2 females and MAJ and RON females, they courted the MAJ females almost as freely as they did their own, and actually appeared to prefer the RON females to their own. PP1 MALES The PP1 males, in comparison to the PP2 males, varied somewhat in the in­ tensity with which they courted the females of AUS, P01 , BR2 and P02. In gen­ eral the PP1 males were less inclined to court the females of those stocks, and this applied especially to the BR2 females. The females of these four stocks refused to copulate with PP1 males and thus paralleled the behavior of the similar females with the PP2 males. The PP1 males courted the PP2 and the NIU females but at a decreasing in­ tensity as the observation period proceeded and no copulations occurred. The PP1 males courted the SUV, RAR, TON, PA1 and PA2 females at a low level of intensity but consistently throughout the observations. Copulations resulted with the SUV, TON and both PA1 and PA2 females. Surprisingly, the PP1 males courted the MAJ and RON females very vigorously and steadily, but none of the females would accept their overtures. Thus considerable difference existed be­ tween the behavior of the PP1 and PP2 males toward the females of the various stocks studied; also, the responses of the PA1 and PA2, MAJ and RON females were quite different toward the two types of males. Spieth: Mating behavior of D. ananassae AUS, P01 AND BR1 MALES The males of these 3 stocks showed high sexual isolation with respect to the females of other stocks, either due to the lack of courtship on the part of the males themselves or because the females consistently refused the males' overtures when they did occur (see Table 3). The only exceptions were the crosses of AUS X P01 where males courted the females and a number of copulations resulted. When AUS males were paired with P01 females, two copulations occurred and both began in abnormal fashion but in both cases the males were able eventually to assume the normal copulatory position and then each continued in a normal manner to termination. In one case the female attempted to dislodge the male for 8 minutes before the she allowed him to assume the normal copulatory stance. In the second copulation the male achieved normal posture one minute after the ab­normal start. With the reciprocal cross, P01 males and AUS females, seven copulations occurred; three of them started normally, but the other four began abnormally. It can be surmised that in a normal environment, individuals of these two stocks, if they should come into contact, would probably display com­plete sexual isolation. BR2 AND P02 MALES P02 males were crossed with the females of 5 other stocks: AUS, P01, PP1, PP2 and PAL These males went no farther than the tapping stage with the AUS and P01 females, but courted the other females in varying degrees of intensity. The females of these stocks, however, refused the overtures of the P02 males. It is significant that the P02 and P01 stocks were both collected at the same locality and at the same time in New Guinea by the same collector. In both crosses the males refused to court the "foreign" females, even under no-choice conditions. ' The BR2 males were tested against the PP1, PP2, TON and PA2 females. They failed to court any of these females except those from Tonga (see Table 3). NIU, SUV AND TON MALES The dark, black pigmented males from Niue, Suva and Tonga readily courted the dark Tutuila females from Samoa (PP2), who accepted the males' overtures, with copulations resulting during every observation period (see Table 3). But when the males of these three stocks were placed with light Tutuila females (PP1), even though they usually courted at high intensity (i.e., at the 3 level), they were never able to induce a female to engage in copulation. Likewise the SUV males readily courted the AUS females, and the TON males (at a lower pitch of intensity) courted the AUS, BR1 and P02 females, but without succeed­ing in achieving copulation. Finally, the TON males refused to court the PA1 females (see Table 3). PA1, PA2, MAJ AND RON MALES The PA1 and PA2 males courted the PP1 females, although the PA1 males did so at a less intense level than did P A2. Nevertheless, both were unsuccessful in achieving copulation with the females. In comparison, the PP2 females accepted The University of Texas Publication PA1 but not PA2 males. PA1 males courted the AUS, BR2 and P02 females but failed to achieve copulatory acceptance. PA1 males, who courted TON females weakly, did achieve one copulation. With MAJ and RON males, the results are essentially reversed since one PP1 female did copulate with each of these males, whereas the PP2 females were es­sentially unattractive to the MAJ and RON males and no copulation occurred. The copulation between the PP1 female and the MAJ male was quite abnormal in that the male spent the entire period in the "back lying" posture, and was never able to achieve the normal copulatory position. Further, the female did not kick and attempt to dislodge the male, and the copulation was eventually termi­nated by the male rather than as a result of any female action. It can be pre­sumed that this copulation was an abnormal one that started out as a partial in­tromission type, that the female never fully responded, but neither did she at­tempt to dislodge the male as did other individuals during the same observation period. RON males found PA1 females sexually attractive; they courted them vigor­ously, and during a single observation when three individuals of each sex were involved, a copulation did occur, indicating that the PA1 stock is not sexually iso­lated from the RON stock under no-choice conditions. MULTI-CHOICE OBSERVATIONS As indicated above, a small number of multi-choice observations were made using stocks whose males and females were physically distinguishable. In every case the stocks were introduced immediately after the individuals had been used for no-choice observations. The data clearly indicate (Table 4·) that high sexual isolation exists between many of the various stocks that were used for study. All of the males vigorously TABLE 4 Multi-choice crosses Inter-stock, multiple Intra-stock courtships choice courtships Stock I d' I' Stock 2 d' I' Obs. Intensity Copulations Intensity• Copulations AUS 7 15 TON 6 6 high 1 nil low 0 AUS 9 7 RON 7 9 high 5 nil nil 0 P01 11 8 PA1 6 8 1 high 9 nil nil 0 BR1 11 21 MAJ 4 5 1 high 2 nil nil 0 BR1 21 27 SUV 16 13 2 high 11 mod. mod. 0 BR2 7 13 PP2 7 10 1 high 0 low low 0 BR2 13 19 TON 12 17 2 high 3 nil low 0 BR2 3 6 PA1 6 4 high 0 nil nil 0 PP1 14 8 PA2 6 8 high 5 mod. mod. 0 PP2 16 8 MAJ 6 11 high 0 mod. low 0 PP2 7 10 RON 24 15 1 high high low 0 PP2 12 12 PP1 10 11 3 high high low 0 'Lefr , Stork I d' o X Stock 2 I' I'; RighL Stock 2 d' d' X Stock 11' I'. Spieth: Mating behavior of D. ananassae courted their own females and the courtships resulted in 38 intra-stock copula­tions; whereas, by comparison, there were no interstock copulations. Further, the behavior of the males toward the "foreign" females was considerably different in that in 10 out of 24 of the combinations, males never courted "foreign" females (nil intensity); in 7 of the combinations the males courted occasionally but with low intensity; in 5 the males courted with some regularity but infrequently (moderate intensity) when compared to their behavior with their own females; and in only two instances (PP2 o o X RON ~ ~ ; PP2 o o X PP1 ~ ~ ) the males actually courted the foreign females as intensely as they did their own females (high intensity). The females, however, never accepted the courtship of any of the foreign males and no inter-stock copulations occurred under the multi­choice conditions. It should be noted, however, that even under no-choice condi­tions, only four of these same crosses did result in copulations, i.e., PP1 o x PA2~, PP2 o x PP1 ~, PP2 o x MAJ ~ and PP2 o x RON~. DISCUSSION Limited observations upon 15 stocks of D. ananassae-like flies from the Pacific Basin show that they can be divided into three groups on the basis of the male mating behavior: (1) stocks AUS, P01 and BR1, all small pale yellowish flies from northern Queensland, Australia, and eastern New Guinea; (2) stocks BR2, P02 and PP1 from eastern New Guinea and American Samoa, which are also yellow pigmented flies with P02 and BR2 being sympatric with P01 and BR1, respectively; (3) nine other stocks collected from the Marshall, Caroline and Polynesian Islands. Although these 9 stocks display identical mating be­ havior, they can be divided into two groups on the basis of color, i.e., four light northern stocks (PA1, PA2, MAJ, and RON) and five darker southern stocks (PP2, NIU, RAR, SUV and TON) . While these dark stocks are all allopatric with each other, one of them, PP2, is sympatric with the PP1 (light) stock. Interest­ ingly, the PP2 males court the PP1 females as readily as they do their own fe­ males under both no-choice and multi-choice conditions. In comparison the PP1 males scarcely court the PP2 females, either under no-choice or multi-choice conditions. The observed sexual isolation between these stocks varies from weak isolation, such as that between the PP2 stock and the SUV and TON stocks (see Table 3) to extremely marked isolation as between the AUS, P01 and BR1 stocks and all other stocks. The mating behavior of any given species or interbreeding popula­ tion is literally evolutionarily tailored to the specific environment in which the population lives. Where the individuals of the population are forced to live under considerably different environmental conditions, particularly under the dis­ ruptive, abnormal conditions of the experimental laboratory, the isolation mechanisms that operate effectively in nature often break down and give un­ expected and abnormal results. Experience indicates that laboratory existence and no-choice experiments invariably result in lowered sexual isolation between closely related species or populations of Drosophila (Mayr and Dobzhansky, 1945; Mayr, 1963; Spieth, 1958). It is believed that any interpretation of the The University of Texas Publication present data should take this factor into consideration. It is presumed that any of these stocks which exhibit complete sexual isolation under no-choice conditions in the laboratory would be isolated under normal field conditions if and when they chance to come into contact. The males and females of any interbreeding population invariably display a balance between the courting behaviors of the two sexes which insures the fecun­dation of the females without the wastage of gametes either by the male or the female. With essentially cosmopolitan species such as D. ananassae, the males typically exhibit high sexual drive and the females relatively high discrimi­nation. This is particularly true of the D. ananassae females, and thus despite the general sexual aggressiveness and courtship persistence of the males, none of the intra-stock observations resulted in all of the females at any given time accepting the male's overtures. In comparison to other wide-spread species such as D. mela­nogaster, D. simulans, D. hydei or D. virilis, the females of D. ananassae exhibit considerable reluctance to accept male courting overtures. This does not mean, however, that the females of all stocks show identical levels of discriminatory re­sponses. Thus the MAJ and RON females were both vigorously courted by the PP2 and PP1 males, but they accepted the overtures of the PP2 males much more readily than they did those of the PP1 males. In comparison, the MAJ and RON males found the PP2 females unattractive. On the basis of the evidence derived from both intra-and inter-stock crosses of PP2 and RON, and PP2 and MAJ, the conclusion was reached that the sexually aggressive PP2 males found the RON and MAJ females equally, if in fact not more, attractive than their own PP2 females. The RON and MAJ females on their part also found the PP2 males more stimulating than were their own males. In comparison the reciprocal crosses of RON and MAJ males to the PP2 females resulted in no copulations, since the RON and MAJ males not only were rela­tively lethargic but also they found the PP2 females to be of low stimulatory ability. Therefore, as might be expected, no copulations resulted. Thus the RON and MAJ females seem to be relatively easily aroused to sexual receptivity and as a balance to this, the males of these stocks have a relatively low sexual drive. In comparison, the PP2 males have relatively high sexual drive while the PP2 fe­males are more resistant to arousal. Courtship balance of this type would seem to involve quantitative changes in the stimulus-response pattern of the two sexes rather than qualitative differences. That qualitative differences may arise, however, between stocks and on an unpredictable basis can be demonstrated by a number of the inter-stock crosses, of which the PA1 X PP2 and PA2 x PP2 crosses most clearly show this type of difference. Thus PP2 males simply fail to court PA1 and PA2 females, never go­ing beyond the posturing stage, and courtship is usually terminated by the male at the tapping stage. It is not a question of the males being lethargic but rather they seem simply to find the PA1 and PA2 females completely unattractive or even repulsive. The reciprocal crosses, PP2 'i' 'i' X PA1 or PA2 ii, however, re­sult in a high level of courtship by the males and copulatory acceptance by the PP2 females of the PA1 males. In comparison, the PP1 males courted the PA1 and PA2 females at a rather low level of intensity but the courting stimulus was Spieth: Mating behavior of D. ananassae such that the females did accept the PP1 males. To the observer it seemed that the PP1 males displayed a super-stimulus which even though infrequently displayed was extremely effective when it did occur. Significantly, as Futch (this Bulletin) has shown on the basis of a variety of evidence such as cytogenetics, morphology, color, etc., the PA1, PA2, MAJ, and RON stocks must all be relatively closely related. The differences in responses (1) of the PP2 males to the different females, and (2) the PA1, PA2, MAJ and RON males to the PP2 females, clearly supports the thesis that geographical sep­aration can and does result in the origin of novel and varied sexual isolating me­chanisms. LITERATURE CITED Futch, David G. A study of speciation in South Pacific populations of Drosophila ananassae. This Bulletin. Mayr. E., and Th. Dobzhansky. 1945. Experiments on sexual isolation in Drosophila. IV. Modification of .the degree of isolation between D. pseudoobscura and D. persimilis and sexual preference in D. prosaltans. Proc. Nat. Acad. Sci. 31: 7581. Mayr, E. 1963. Animal species and evolution. Harvard Univ. Press, Cambridge. Spieth, Herman T. 1952. Mating behavior within the genus Drosophila (Diptera). Bull. Am. Mus. Nat. Hist. 99(7): 395-474. ----. 1958. Behavior and isolating mechanisms. In A. Roe and G. G. Simpson, eds., Behavior and evolution. Yale Univ. Press, New Haven, pp. 363-389. Stone, Wilson S., Marshall R. Wheeler, Warren P. Spencer, Florence D. Wilson, Jun~ T . Neuen­ schwander, Thomas G. Gregg, Robert L. Seecof, and Calvin L. Ward. 1957. Genetic studies of irradiated natural populations of Drosophila. Univ. Tex. Pub., No. 5721: 260-316. Stone, Wilson S., Marshall R. Wheeler, and Florence D. Wilson. 1962. Genetics studies of irradiated natural populations of Drosophila. V. Summary and discussion of tests of popula­tions collected in the Pacific proving ground from 1955 through 1959. Univ. Tex. Pub., No. 6205: 1-54. Stone, Wilson S., Marshall R. Wheeler, Florence D. Wilson, Virginia L. Gerstenberg, and Hei Yang. Genetic studies of natural populations of Drosophila: IL Pacifiic island populations. This Bulletin. VI. The Influence of Light on the Mating Behavior of Drosophila' JOSEPH GROSSFIELD2 INTRODUCTION Several authors (Patterson and Stone, 1952; Ehrman, Spassky, Pavlovsky, and Dobzhansky, 1965) have pointed to the relative paucity of information concern­ing the effects that environmental variables might have on sexual activity and sexual selection in Drosophila. Since two species of Drosophila, D. subobscura (Philip, Rendel, Spurway, and Haldane, 1944) and D. auraria (Spieth and Hsu, 1950) are known to require light in order to mate, and Spieth and Hsu (1950) have shown that the mating of species in the melanogaster group other than D. auraria are affected by light, this environmental variable appeared to offer an opportunity to further explore the interrelations between organism and environ­ment. To accomplish a meaningful survey of possible light dependence it was decided that, in addition to choosing a species representative of a group, a rela­tively large number of species in one group, as well as a number of strains of a particular species would have to be studied. Since Wheeler's observations ( 1947) of the quinaria group indicated a certain light response and the species in the group have been shown to be quite variable in several different aspects of their biology, it seemed to provide a favorable point of departure. In addition, several pairs of species in the group have been shown (Sears, 1947) to be quite fertile in interspecific crosses. This coupled with the fact that a difference exists between the species with regard to mating ability in dark­ness seemed to provide ample opportunity to investigate the genetic basis of any light response that might be evidenced. METHODS AND MATERIALS A single strain of flies for each of the species has been used, with the exception of D. guttifera and D. subpalustris, for each of which two strains were employed, and D. occidentalis for which five strains were used. The University of Texas collection numbers, if any, and geographical origin of all of the strains used are indicated in Table 1. All stocks were reared and maintained on a medium consisting of banana­agar-malt-yeast-karo, and all experiments were performed at 21° ± 2°C in a room illuminated for periods of 12 to 18 hours each day. Virgins were collected twice daily from all strains, lightly etherized, separated according to sex, and placed in groups of 10 (j> <;> or 10 5 5 in small vials (8dram; 25mm X 95mm) containing cornmeal food. 1 This work was supported in part by a Public Health Service research grant, GM-11609, and training grants, 5-T1-GM-337 and 2-T1-GM-337-06, from the National Institutes of Health, Public Health Service. 2 Present address: Department of Life Sciences, University of California, Riverside. The University of Texas Publication TABLE 1 List of species and strains used, together with their University of Texas collection numbers and the localities where they were collected Collection Species number Locality Subgenus: Drosophila virilis-repleta section virilis group virilis Sturtevant 1801.1 Texmelucan, Puebla, Mexico melanica group melanica Sturtevant 1720.3 Cliff, New Mexico mesophragmatica group pavani Brncic H347.6 Santiago, Chile repleta group mulleri subgroup mulleri Sturtevant 1815.2 Brownsville, Texas hydei subgroup hydei Sturtevant 2375.7 Chile quinaria section immigrans group immigrans Sturtevant 2321.9 Cross Anchor, South Carolina funebris group funebris Fabricius 1732.3 Ain Arubia, Lebanon quinaria group innubila Spencer 2074.6 New Mexico falleni Wheeler 1062.6 Gt. Smoky Mt. N.F., Tennessee phalerata Meigen 1915.1 Beirut, Lebanon occidentalis Spencer 1945.1 Crater Lake N.P., Oregon 1974.5 Dark Canyon, California 2175.3 Mt. San Jacinto, California 2575.3 Hollinger Canyon, Santa Fe N.P., New Mexico 2581.2 Shoshone N.F., Wyoming munda 929.8 Cave Creek, Arizona subpalustris Spencer 1877.9 Georgetown, South Carolina 3012.2 South Carolina palustris Spencer 1757.13 Bemidji, Minnesota guttifera Walker Fayetteville, Arkansas 2086.3 Austin, Texas calloptera group ornatipennis Williston 2378.2 San Vicente, Cuba guarani group guarani subgroup subbadia Patterson and Mainland 2262.24 Vera Cruz, Mexico cardini group acutilabella Stalker H351.2a Windsor, Jamaica tripunctata group tripunctata Loew 2539.2 Austin, Texas Buescher St. P., Texas 1. Species Tests: Flies used for testing the ability of the various species to mate in darkness were aged for various periods of time, as indicated in Tables 2-20. The length of time Grossfield: Light and Mating Behavior in Drosophila the females and males were exposed to each other, after having been aged, under conditions of either LD (light) or DD (darkness) varied from 6-15 days for the quinaria group, to a series of exposure times ranging from 1-7 days for the ma­jority or the other species. Different densities of flies, as well as different exposure times were used. All single pair matings were accomplished in small vials (25mm X 95mm), while all matings utilizing larger numbers were made up in large vials ( 12 dram; 100 mm X 33mm). The matings to be accomplished under conditions of diurnal illumination, hereafter referred to as LD or light matings, were set up by introducing a definite number of males into large vials of the fe­males on banana medium. Those matings that were to be tested in darkness were treated in like fashion but for the fact that males were shaken into the vials containing females in a photographic darkroom in the absence of any light. The experimental vials for DD experiments were then placed into light-tight cardboard boxes or into a wooden light-tight one. The boxes were then sealed with tape, the lights turned on, and the boxes placed on a laboratory shelf. By linear arrangement of the vials, and with the use of cardboard cut into various shapes, particular vials could be identified in darkness by touch alone. In this way, vials containing flies com­parable in age and density to those dissected from the LD (or control) series of vials, could be identified and withdrawn for dissection after the appropriate ex­posure time. All females recovered were dissected in physiological saline and their spermatheca and ventral receptacles checked for the presence of sperm. All storage and experimental vials were checked for larvae before being discarded. The danger that mating will occur before the flies can be etherized may be minimized by removing from the dark, in darkness, only that number of vials containing a number of flies which can be etherized in a time period less than the minimal copulation time for the species. When flies from a particular vial were etherized for dissection, the number of males and females were noted, including the number of each sex which had become mired in the food. The presence of an extra male, or of one or more males less than the counted number intro­duced, was considered sufficient grounds on which to discard the data from a vial. This procedure also allowed the calculation of the percentage of males and fe­males recovered from experiments. In the earliest series of experiments the number of males remaining was not determined, consequently no data is avail­able for the recovery of males for certain strains. Despite all precautions taken, extra females were present in a few vials. This may be seen, for example in Table 3 for the number of females dissected for 1-day LD conditions. T ABLE 2 Insemination data on Drosophila virilis Light Dark Type Expos. Age at mating period start No.'? '? No.'?'? Per cent No.'?'? No. '?'? Per cent '? '? rl 'i' dissected Per cent insem. No.'i''i' intro. Dark No.'i''i' dissected Per cent insem. 10 10 10 5 5 5 4 7 17-27 12--20 12--21 21 31 100 21 30 96 19.0 20.0 33.3 40 30 150 40 30 140 0.0 0.0 0.0 152 147 28.6 220 210 0.0 20 20 20 10 10 10 4 7 13,22 17-22 13,17 20 20 20 60 20 20 20 60 0.0 10.0 25.0 10.0 20 20 20 60 20 19 20 59 0.0 0.0 0.0 0.0 30 15 7 23,25 30 -242 29 236 10.3 21.6 30 310 26 295 0.0 0.0 The use of different densities and lengths of exposure was undertaken to as­certain, if possible, the conditions in which various species would or would not mate in darkness. As a final check, all those species which appeared to be com­pletely light dependent were tested in another fashion. Stock bottles were cleared four to eight hours before virgins were removed. The male and female virgins from a strain were then pooled and, as equally as possible without ether­izing and counting them, divided into two groups, in bottles containing banana food. One of these bottles was left on a laboratory shelf while the other was im­mediately placed in a light-tight box. In this way any effects of etherization and/ or over-aging would be revealed. These bottles were checked for larvae two weeks after they had been set up, this time period well overlapping the aging of the flies used in the earlier experiments. The bottles all contained about 60 flies, and no contraindication of other experimental results was seen. All LD and DD matings for a particular experiment were set up simultan­eously, with the exception of some early experiments with D. guttifera, D. palus­tris, D. subpalustris, and D. phalerata. In these cases some matings were set up in DD, for which no LD control matings were made. In the case of D. guttifera the number of females in mass matings was not determined; hence the discrep­ancy in Tables 16A and 16B between the number of females introduced and number of females dissected in DD. A procedure used for estimating the reliability of such matings, made without a parallel series of LD controls, was to simply place several vials from the DD series on a laboratory shelf, without transferring the flies to fresh food, for a period of a week and then checking the vials for larvae. This post-DD check for mating was carried out with at least one vial from each DD series which had failed to show either insemination or larvae. As a check on sterility, all single pair matings that had failed to show larvae in either LD or DD were transferred to fresh food in LD and rechecked after a week to determine whether the failure of matings had been due to sterility or to experimental conditions. Gross field: Light and Mating Behavior in Drosophila All data for each quinaria group species were gathered through a series of ex­periments, rather than from a single, relatively large scale experiment as was the case with all other species tested. This distribution of investigations over pe­riods ranging from one to seven months effected a check on the repeatability of results, as did the use of unetherized virgins in bottles and the single pair matings made with D. funebris and D. tripunctata. This latter species was also used to compare a laboratory population with one immediately derived from natural population. The progeny derived from eggs laid on laboratory food by wild­caught females, inseminated by males of like origin, were placed as a series of small mass matings under an experimental regime similar to that used for the laboratory population of D. tripunctata. An alternate correlation would have been to test wild-caught flies directly, but this was not attempted owing to the high percentage (92.8) of wild females already inseminated (Patterson, Stone, McDanald, 1947). 2. D. palustris-D. subpalustris 1877.9 Hybridization. Virgin males and females which had been aged for 7 to 10 days were used to make up interspecific single pair matings. Vials which had been kept under LD conditions were checked for larvae after 11 days and at that time all surviving pairs not showing larvae were transferred to fresh food. These were checked after another 16 days and the total number proving fertile was recorded. Those vials kept in DD were checked after an exposure period of 14 days but no transfers were made. A similar procedure was followed in establishing interspecific crosses utilizing 10 males or females of each species in large vials. These matings were kept under either LD or DD conditions for a period of 12 days, at which time dissections were made and the number of females carrying sperm under each light regime determined. The results of this series of experiments constitute Table 23. Two series of LDmatings utilizing 10 females and males of each species were set up to obtain an F1 • The F1 progeny were counted for 12 successive days after the first eclosion for the first series, and for 16 successive days for the second. The P1 flies were transferred to fresh food weekly, but all counts of progeny were made from only the original vials. In all, 34 vials containing 10 males and 10 alien females each were set up for the P ~ X S iS cross and 26 vials for the recip­rocal cross. The F1 progeny resulting from these matings was used to make up F1 inbred and backcross matings, the flies having been aged for 9-12 days (as a matter of convenience) in the case of the F1 inbred and 15-23 days for the back­crosses. Ten pairs of flies per vial were used in establishing all of these crosses. The same counting and transferring procedure was followed in counting the progeny of the F1 inbred and backcrosses, but for the fact that progeny were counted for 18 successive days. These progeny counts are presented in Table 24. Half of the crosses involving (PS)F1 flies were placed in DD for the first week, checked for larvae, and then transferred to fresh food in LD. This was done to ascertain whether any of the actual backcrosses would go in darkness, and was not attempted with those crosses involving (SP)F1 flies owing to the fact that there were not enough flies to risk testing in this fashion. The University of Texas Publication The Fi inbred and backcross offspring, after having been aged, were set up in large vials of 5 males and 10 females each. These progeny tests were placed under appropriate conditions of LD or DD and dissected at the end of a week to 9 days, the results being shown in Table 25. 3. Observations. Small mass matings (2-100 o, 3-10'i' 'i') of various Fi and backcross flies were observed. All observations were carried out under 15x binocular magnifi­ cation in small vials without food, lined for % of their internal circumference with moistened filter paper. RESULTS The insemination frequencies during the various lengths of exposure in LD and DD of the various species tested are shown in Tables 2-20. The numbers at the bottom of each column in these tables represent the summation of all the data, irrespective of the age of the flies, or the period of exposure. The data for different types of matings is pooled, in like fashion, at appropriate places in each table, while the results from single pair matings, if any, are presented at the bottom of each table. For all species, a chi square value was calculated only from the data pooled for various days, ages and types of matings, excluding single pair matings. By handling the data in this fashion, an overall value, characterizing the species, is obtained. This procedure takes into account several of the controll­able experimental variables. An evaluation of the component parts of the data culminating in the overall chi square value is presented below. A general sum­mary of all species tested together with the percent recovery of females and males from experimental conditions is given in Table 21. The various classes of species response shown, with respect to illumination, are presented in Table 22. Also included are some cases from data in the literature. D. virilis (Table 2) mates as well in darkness as it does in light, even if females and males are together for only 24 hours. In fact, the percent of insemination showed no increase over the 1-day values in the sample taken 5 days later. A similar lack of dependence upon illumination for insemination is shown by the 9-day sample of D. funebris (Table 8). This independence is also demonstrated by the results of the single pair matings. With this species, a 24-hour test was made of the effect of density on insemination. As can be seen from Table 8, 98% of the females are inseminated after 24 hours in light. In contrast to this, about 84% were inseminated in darkness. This difference yields a x2 value of 8 (P = .01 < P < .005) with one degree of freedom. Although this difference is signifi­cant, it is doubtful whether it would be maintained after several days of ex­posure. Since the volume of a bottle is three times that of a large vial, an indication is present that, to some small extent, individuals of D. funebris rely on vision for either location of, or courtship of, other individuals. The latter possi­bility seems less likely. D. hydei (Table 6), in the 6-day sample showed equal insemination in LD and DD, the percent of insemination being 100.0 in both cases. The 1-day sample, however, showed more inseminations in DD than in LD, a difference strong Grossfield: Light and Mating Behavior in Drosophila TABLE 4 Insemination data on Drosophila pavani Light Dark Type Expos. Age at mating period start No.'i' 'i' No.'i' 'i' Per cent No.'i' 'i' No.'i' 'i' Per cent 'i''i' 00 (days) (days) intro. dissected insem. intro. dissected insem. 10 5 1 11-16 4-0 39 15.4 4-0 37 0.0 10 5 4 10-17 30 30 20.0 30 28 0.0 10 5 7 11-17 60 57 17.5 70 61 0.0 130 1Z6 17.5 140 126 0.0 20 10 11,14 20 20 0.0 20 13 0.0 20 10 4 15 20 20 5.3 20 18 0.0 20 10 7 10-12 60 55 12.5 80 64 0.0 100 96 7.3 120 95 0.0 30 15 7 11,12 30 30 50.0 30 27 0.0 260 252 15.5 290 248 0.0 TABLE 5 Insemination data on Drosophila mulleri Type mating 'i''i' 00 Expos. period (days) Age at start (days) No.<;' 'i' intro. Light No.'?'? dissected Per cent insem. No.'i''i' intro. Dark No.'i''i' dissected Per cent insem. 10 10 10 5 5 5 1 4 7 14--18 10-19 10-19 40 30 90 4-0 30 90 57.5 56.7 70.0 40 30 110 40 30 108 12.5 40.0 13.0 160 160 64.4 180 178 17.4 20 20 20 10 10 10 4 7 17 15 11 20 20 20 6o 20 20 19 59 45.0 70.0 94.7 69.5 20 20 20 60 20 20 20 60 15.0 11.1 50.0 25.0 30 15 7 16 30 30 83.3 30 28 25 .0 250 249 67.9 270 Z66 19.9 TABLE 6 Insemination data on Drosophila hydei Light Dark Type Expos. Age at mating period start No. 'i' 'i' No.'i' 'i' Per cent No.<;' 'i' No. <;' 'i' Per cent 'i''i' 0 0 (days) (days) intro. dissected insem. intro. dissected lnsen1. 10 5 1 6--14 100 100 4-0.0 100 97 66.0 10 5 6 6--14 110 103 100.0 110 98 100.0 210 203 68.1 210 195 83.1 The University of Texas Publication enough to appear even in the pooled data for both samples. Although a significant difference, it is not derived from sufficient data to designate D. hydei as anything other than a light-independent species, as is done in Table 22. It may be that for brief exposure periods, this species will inseminate more females in darkness, a possibility indicated by a"?" in Table 22. The several species referred to above are all considered to be light independent in their behavior. A consistent pattern, appearing irrespective of density (type mating) or ex­posure period, or light dependence is presented by D. mulleri (Table 5). The dif­ferences in overall inseminations are 68% in LD and 20% in DD, (x2 y = 119; p >p = .001) . Inspection of Table 5 reveals that insemination in LD and DD shows a rela­tively constant difference, but that an increase in insemination results from longer exposure and higher density. Comparison with the 7-day data for the larger mass matings, both in LD and DD, would seem to indicate that, above a certain density, a decrease in insemination ensues. The 40% value of the 4-day DD sample seems to show an erratic peak, perhaps owing to uncontrolled var­iation in experimental conditions, or to sampling error. Such internal inconsist­encies in the data, for this and other species, remain for the most part unexplained, but for the factors mentioned. The 7-day 10:5 data from D. immigrans (Table 7) would seem to show no difference between LD and DD conditions. However, the results from 290 other dissected females as well as the highly significant (x2,. = 14; P > P = .001) differ­ence shown in the overall data reveals a distinct reduction in insemination in darkness. Whether this reduction is a real one can be revealed by future investi­gations. For the present, this species is placed in Class II of Table 22. The data for D. ornatipennis (Table 17) points to the fact that an increase in density clearly obliterates any difference in insemination frequency between light and darkness. In spite of this, the 10: 5 matings, regardless of exposure time, TABLE 7 Insemination data on Drosophila immigrans Light Dark Type mating 'i''i' 00 Expos. period (days) Age at start (days) No.<;' 'i' intro. No.<;' 'i' dissected Per cent insem. No.<;'<;' intro. No.<;' 'i' dissected Per cent insem. 10 5 12-26 30 30 63.3 40 40 30.0 10 5 4 13-27 30 29 58.6 30 30 26.7 10 5 7 13-25 110 95 69.5 120 98 69.4 170 154 66.2 190 168 52.4 20 10 tZ,16 20 20 60.0 20 20 50.0 20 10 4 14 20 20 85.0 zo 19 52.6 20 10 7 12,16 20 20 75.0 20 1Z 33.3 60 60 73.3 60 51 40.0 30 15 7 15 30 11 81.8 30 13 53.9 260 225 68.9 280 232 51.3 Grossfield: Light and Mating Behavior in Drosophila show a marked decrease in DD insemination. The highly significant overall x~ value of 29 (P > P = .001), reflects this fact. Even though the sample sizes show­ing no difference, and a marked difference, in ill and DD inseminations are comparable, the marked reduction that is shown is used as a basis for placing this species in Class II. Another species placed in this group is D. falleni. Inspection of Table 10 shows a uniformly high (about 97%) insemination frequency in LD and a somewhat erratic, but consistently low (about 11 % ) insemination in DD. These data result in a highly significant (P > P = .001 ) x2 y value of 363. D. occidentalis, of which five strains were tested, showed an overall LD in­semination of 95 % and an insemination frequency of 1 % in darkness. This species then, may be classed as distinctly light dependent, as is done in Table 21. An examination of Table 13, however, reveals that the strains of D. occidentalis really fall into two groups; one completely dependent on light for mating, and one that does show some insemination in darkness. Strains 1974.5 and 2175.3 which show a low level of DD insemination, were both collected from the same region but three years intervened between the two collections. These strains show a lower general activity level than the other three strains which were collected at widely divergent points in the range of the species. There is a general agree­ment between the percent insemination shown in DD by mass (2% ) and by single pair (3%) matings. The data from the strains showing some insemination in DD are pooled as Group I in Table 22, and those showing no insemination in DD are pooled in Group II. TABLE 8 Insemination data on Drosophila funebris Light Dark Type mating: '? '? X Canton-S strain ~ (for dominant lethal). At two-day intervals, the males were remated to a new group of virgin females. Females from the previous mating were separated from each other. All in all, seven two-day mating periods, A to G, were allowed before the males were discarded. The sex-linked recessive lethals were tested by the Muller-5 technique (Spencer and Stern, 1948). Only the vials containing 0 or 1 wild type F2 male were scored as lethals. Lethals were checked through the F3 . Translocations were determined by tests using bw and st markers on the second and the third chromosomes, respectively. F1 males (heterozygous for the treated chromosomes and mutant markers) were backcrossed individually to bw, st virgin females. The presence of induced 2-3, Y-2, Y-3, or Y-2-3 translocations in the F2 was then scored according to the absence of certain segregation groups. For 2--3 translocations, flies were checked through F3 by mating wild-type F 2 males to three marker virgins in order to make sure that the translocations came from the treated (P1 ) male. All the doubtfuls were backcrossed in the same way for verification. Dominant lethal frequencies were determined by the number of eclosions vs. Elequin: Modification of Genetic Damage in Drosophila the number of eggs laid. Females from each mating period were allowed to lay eggs for 4 or 5 days before they were discarded. During the days allowed for egg laying the females were transferred to fresh food and the eggs were counted daily. Basing mating periods on Auerbach's tests in D. melanogaster (1954), A repre­sents mature sperm; B, sperm bundles; C and D, spermatids; E, meiosis; F, sper­matocytes; and G, spermatogonia. In determining the 95% limits of expectation, the method of Stevens (1942) was used for sex-linked recessive lethals and translocations. For dominant lethals, the conventional method for estimating standard errors (for binomial distribu­tion) was used. The experimental design is shown in the following table: Experiment Pre-treatJnent number (10 min.) X-rayed 340r/ min. Interval (20 min. ) X-rayed 340r/min. Post-treatment (20min. ) 0 2 0 2; 500r 0 2 0 2; 500r 02 2 02 0 2; 500r A 0 2; 500r 02 3 02 0 2; Or A 0 2; Or 02 4 A A; 500r A A ; 500r A 5 A A ; Or A A; Or A 6 A A; 500r 0 2 A; 500r A 7 A A ; Or 0 2 A ; Or A 8 A A ; 1500r A A ; 1500r A 9 A A ; Or A A; Or A 10 A A; 1500r 0 2 A ; 1500r A 11 A A; Or 02 A;Or A RESULTS AND DISCUSSION The results are summarized in Tables 1 through 4 and Figures 1 through 3. Sex-linked recessive lethals In tests for sex-linked recessive lethals (Table 1, Fig. 1) the in-oxygen X-rayed experiments (1 and 2) showed a significant difference only in period C (sper­matids). There was less damage when the flies were treated with oxygen (Experi­ment 1) as opposed to argon (Experiment 2) between the two X-ray fractions. The in-argon X-rayed experiments showed no significant difference between Experiments 4 and 6, or 8 and 10 in any of the mating periods. There are several possible explanations for the above results. First, it is possible that the period between the two doses is not long enough to show the effect of the intervening gas treatments. Sobels ( 1965) reported that "mutation fixation" takes place sometime between 25 and 50 minutes post-irradiation when the flies were X-rayed in nitrogen, whereas the "fixation" takes place earlier in flies X-rayed in air. This would imply that mutation induction in oxygen is different from that in nitrogen. If the same holds for argon, then no significant difference would be expected between Experiments 4 and 6, and 8 and 10, since the oxygen treatments in Experiments 6 and 10 were administered before the "mutation fixation." Because of the shorter "fixation" in flies irradiated in air (or oxygen) The University of Texas Publication TABLE 1 The frequencies of induced sex-linked recessive lethals Experiment A B c D E F G a 1054 962 656 699 196 415 02-02-02 b 35 20 47 67 13 1 (500 r) (500 r) c 3.3 2.1 7.2 9.6 6.6 0.3 d 2.3 1.3 5.3 7.5 3.6 .01 4.6 3.2 9.4 12.0 11.1 1.3 2 a 452 679 745 573 651 482 497 0 2-A-02 b 13 27 42 53 47 29 12 (500 r) (500 r) c 2.9 4.0 5.6 9.2 7.2 6.0 2.4 d 1.4 2.6 4.1 7.0 5.4 4.1 1.3 4.9 5.7 7.6 11.9 9.5 8.5 4.2 3 a 234 308 291 291 303 282 86 02-A-02 b 0 0 0 1 0 0 (0 r) (0 r) c 0.0 0.0 0.0 0.3 0.0 0.4 0.0 d 0.0 0.0 0.0 .01 0.0 .01 0.0 1.6 1.2 1.3 1.9 1.2 2.0 4.3 4 a 213 452 577 679 875 867 618 A-A-A b 4 6 10 14 17 4 7 (500 r) (500 r) c 1.9 1.3 1.7 2.1 1.9 0.5 1.1 d 0.5 0.5 0.8 1.1 1.2 0.1 0.5 4.8 2.9 3.2 3.4 3.1 1.2 2.3 5 a 164 287 168 235 209 191 11 A-A-A b 0 0 0 0 0 0 0 (0 r) (0 r) c 0.0 0.0 0.0 0.0 0.0 0.0 0.0 d 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.3 1.3 2.2 1.6 1.8 1.9 33.5 6 a 1133 1028 1006 1055 996 1007 1021 A-02-A b 22 21 18 28 13 5 8 (500 r) (500 r) c 1.9 2.0 1.8 2.7 1.3 0.5 0.8 d 1.2 1.3 1.1 1.8 0.7 0.2 0.3 2.9 3.1 2.8 3.8 2.2 1.2 1.5 7 a 331 371 365 334 349 318 370 A-02-A b 1 0 0 0 0 0 0 (0 r) (0 r) c 0.3 0.0 0.0 0.0 0.0 0.0 0.0 d .01 0.0 0.0 0.0 0.0 0.0 0.0 1.7 1.0 1.0 1.1 1.1 1.2 1.0 8 a 1297 1290 1014 57 471 935 1119 A-A-A b 97 115 168 9 11 7 3 (1500 r) (1500 r) c 7.5 8.9 16.6 15.8 2.3 0.7 0.3 d 6.1 7.4 14.4 7.4 1.2 0.3 .05 9.1 10.6 19.0 28.6 4.1 1.5 0.8 10 a 1111 1120 1051 129 1165 1222 1369 A-02-A b 98 137 160 20 24 15 8 (1500r) (1500 r) c 8.8 12.2 15.2 15.5 2.1 1.2 0.6 d 7.2 10.4 13.1 9.9 1.3 0.7 0.3 10.7 14.3 17.6 23.1 3.1 2.0 1.2 a=Total chromosomes tested. b==Number of recessive lethals. c=:Per cent of recessive lethals. d=Lower and upper limits of 95% expectation. Elequin: Modification of Genetic Damage in Drosophila Fm. 1. The frequencies of sex-linked recessive lethals induced in the spermatogenic cells of Drosophila melanogaster. Mating periods after Auerbach (1954): A= mature sperm, B =sperm bundles, C and D = spermatids, E =meiosis, F =spermatocytes, G = spermatogonia. Num­ber. in brackets correspond to the experiment numbers referred to in the text. the effect of oxygen treatment between fractions can be seen, as in spermatids (C) in Experiment 1. Second, the frequency of recessive lethals induced in argon was initially very low and any recovery which might be due to the oxygen treatment between the doses would not have made a difference big enough to be detected. Therefore, the sample size ought to be larger in order to detect any difference in experi­ments where the flies were X-rayed in argon. Third, in addition to the possible differences in the behavior of breaks produced in oxygen and nitrogen (or argon), a portion (30-50%) of sex-linked recessive lethals involves chromosome rearrangements (Demerec, '1937; Slizynski, 1938; Slizynska and Slizynski, 1947). If the treatment given between doses can only affect the recovery of that portion of lethals which resulted from rearrangement, then the difference would have been even harder to detect without a larger sample. Dominant lethals The results for dominant lethals are summarized in Table 2 (Fig. 2) . For the in-oxygen X-rayed experiments (Experiments 1 and 2), a significant increase 17 (f) 16 _J 15 II ~ 10 w (.) 9 w a:: 8 0 w 7 z ~ 6 _J 5 I x 4 w (f) 3 ~ 2 0 0 [1] o--o Oz-02-02 (500r) (500r) [2] ----.... Oz-A -02 (500r) (500r) / [4]0--0 A-A-A / / (500r) (500r) // [6] 0----0 A -02-A ,,JI(' (500r) (500r) / [s] A----0 A -A -A / / (1500r) (1500f) / / / / [1 0] A---• A-02-A / / (1500r) (1500r) •' ' ' ' ' ',,, ' PERIODS The University of Texas Publication TABLE 2 The frequencies of induced dominant lethals Experiment A B c D E F G 02-02-02 (5 00 r) (500 r) a b c d 11410 15977 71.4 0.7 5527 8118 68.0 1.0 3209 7305 43.9 1.1 1414 6100 23.2 1.1 973 3124 31.1 1.6 1784 2349 76.0 1.7 2 a 843 1061 267 75 586 225 622 02 -A-02 (500r) (500r) b c d 1151 73.2 2.6 1419 74.8 2.3 548 48.7 4.2 290 25.9 5.0 2517 23.3 1.7 689 32.7 3.5 1075 57.9 3.0 3 a 311 92 274 542 718 1028 1855 0 2 ­(Or) A-02 (Or) b c d 322 96.6 2.0 95 96.8 3.5 279 98.2 1.6 570 95.1 1.8 761 94.3 1.7 1111 92.5 1.5 2047 90.6 1.3 4 A-A-A (500 r) (500 r ) a b c d 336 399 84.2 3.6 685 766 89.4 2.2 714 837 85.3 2.2 224 267 83 .9 4.4 1342 1773 75.7 2.0 970 1360 71.3 2.4 1022 1451 70.4 2.3 5 A-A-A (0 r ) (0 r) a b c d 449 473 94.9 2.0 46 47 97.9 4.1 119 124 96.0 3.4 246 264 93.2 3.0 148 150 98.7 1.8 415 449 92.4 2.5 454 515 88.2 2.8 6 A-02-A (500 r) (500 r) a b c d 3338 3816 87 .5 1.0 8501 10006 85.0 0.7 1199 1514 76.7 2.1 3850 5543 69.5 1.2 2234 3423 65.3 1.6 2477 2779 89.1 1.2 2204 2512 87.7 1.3 7 A-02-A (0 r) (0 r ) a b c d 1111 1143 97.2 1.0 4610 4841 95.2 0.6 1568 1703 93.5 1.2 2170 2245 96.7 0.7 2771 2985 92.8 0.9 1793 1968 94.2 1.0 603 645 94.0 1.8 8 A-A-A ( 1500 r) (1500 r) a b c d 7854 16156 48.6 0.8 6163 23841 25.9 0.6 1901 171 36 11.1 0.5 93 1268 7.3 1.4 2365 5165 45.8 1.4 2234 2910 76.8 1.6 3021 3338 90.5 1.0 9 A-A-A (0 r) (0 r) a b c d 5348 5638 94.9 0.6 991 1045 94.8 1.4 6568 6875 95.5 0.5 4901 5236 93.6 0.7 4868 5305 91.8 0.8 3058 3304 92.6 0.9 3504 3682 95.2 0.7 10 A-02-A (1500 r ) ( 1500 r) a b c d 9931 23449 42.4 0.6 3839 18542 20.7 0.6 1389 14191 9.8 0.6 172 2913 5.9 0.9 2598 7054 36.8 1.2 5546 10080 55.0 1.0 6214 8865 70.1 1.0 11 A-02-A (0 r) (0 r) a b c d 5680 6176 92.0 0.7 4950 5374 92.1 0.7 7557 7942 95.2 0.5 6180 6618 93.4 0.6 4936 5223 94.5 0.6 6178 6576 93.9 0.6 4563 4875 93 .6 0.7 a::=Number of eclosion. h=Number of eggs. c==Percent of eclosion. d=2 S.E. I 1i I I I T I 1•i T • i 1 • w (j) 0 0 .. a::: " w ' 0 (L z ~ - I­ <( (_) ~ / / / / / m / """ / / / •../ / I'­ ID l{) '¢ 10 C\J (S883 /NOIS01J3) HJl'itH % FIG. 2. The frequencies of dominant lethals induced in the spermatogenic cells of Drosophila melanogaster males. Mating periods (A through G) same as Fig. 1. Numbers in brackets cor­respond to the experiment numbers referred to in the text. The University of Texas Publication in damage was observed in periods C, D, and G when argon was used between the two doses, as compared to the damage when oxygen was used between the doses. This seems to concur with the hypothesis and suggestion that energy is needed for repair processes (Wolff and Atwood, 1954; Wolff and Luippold, 1955, 1956; Beatty and Beatty, 1959, 1960; Sobels and Tates, 1961). No significant difference was found in B, E, and F between the two experiments. The results for the in-argon X-rayed experiments (Experiments 4, 6, 8, and 10) , however, were rather inconsistent. On the basis of the above-mentioned hypothesis, it is expected that the damage should be higher with anoxia obtained by using argon during the interval between two fractions of X-rays. For Experi­ments 4 and 6, the opposite of expectation, however, was obtained for periods B, C and D, and E, which correspond to sperm bundle, spermatids, and meiotic stages, respectively. That is, for these periods, the damages were less when argon treatment was given between two doses of X-rays in argon. The reverse was found for the spermatocyte (F) and spermatogonia (G). The situation observed in periods B through E might have been due to an incomplete elimination of oxygen before the second dose in argon was given. However, this explanation does not seem likely because no similar results were seen for the periods in ques­tion in the two other tests (sex-linked lethals and translocations). It is also pos­sible that since the flies were kept in an anoxic state up to the first X-ray treat­ment in argon, this was a manifestation of an effect of oxygen before the second dose in argon was given, but was not long enough to affect also the two earlier stages (F and G). No significant difference was observed in mature sperm. This seems to agree with the work of Alexander and Bergendahl (1962) on the mature sperm of D. virilis where treatment with gamma rays or X-rays showed no indication of recovery in induced genetic damage as determined by dominant lethal or translocation tests in oxygen or anoxic atmosphere. By varying the dose rate from 50 to 2,000 r/ min, or the time interval between doses from 15 to 60 minutes, they observed no effect of these changes on the genetic damage. For Experiments 8 and 10, damages were higher (or hatch lower) in all mating periods except D when the flies were treated in oxygen between doses than when they were treated in argon. These inconsistencies and peculiarities can perhaps be explained by the possible heterogeneity of the origin of the induced dominant lethality resulting from various phenomena involved in chromosome breakage as found by von Borstel and Rekemeyer (1959) and van Borstel ( 1963). Von Borstel ( 1963) listed a variety of induced chromosomal breakage phenomena which result in domi­nant lethality. Another peculiarity observed in the dominant lethal tests was the lower than expected percentage of egg-hatch obtained in spermatogonia (G) in both Experi­ments 2 and 4 (Table 2, Fig. 2). It should be mentioned that these two experi­ments were done at the same period of time. This is not considered to be an indication of the greater sensitivity of the spermatogonia to X-ray treatment but is more likely due to the dryness of the food used, which to some degree may have interfered with the development of the flies. It should also be kept in mind that some of the eggs deposited by the females might not have actually been fertilized. If this should happen, this would cer­ Elequin: Modification of Genetic Damage in Drosophila tainly be one of the contributing factors for the dominant lethality results. How­ever, we have no way of knowing it with any certainty, for no eggs laid by the females were checked for the absence of sperm. Translocations In the translocation tests (Table 3, Fig. 3) there was no significant difference in the number of translocations induced in the in-argon X-rayed experiments between Experiments 4 and 6 or between Experiments 8 and 10. There was a significant difference between the two in-oxygen X-rayed experiments (1and2) only in the spennatid stages ( C and D) . The number of translocations induced was significantly lower when oxygen was used between the two X-ray doses (Experiment 1) than when argon was used between the two doses (Experi­ment 2). No difference was found between the four comparable in-argon X-rayed experiments (Table 3, Fig. 3, Experiments 4 and 6, 8 and 10). These results parallel those of recessive lethals. If the av·ailability of oxygen (which restores oxidative metabolism after the first dose) enhances the healing of a chromosome Frn. 3. The frequencies of translocations induced in the spennatogenic cells of Drosophila melanogaster males. Mating periods same as Fig. 1. Numbers in brackets correspond to the experiment numbers referred to in the text. 24 22 20 18 CJ) 16 z 0 f-14 <( u 0 12 _J CJ) z <( 10 0:: f­ 8 ~ 0 6 4 2 0 ~---­ I [1) o----0 Oz -Oz-Oz I (500r) (500r) I I I [2) ....__.. Oz-A-Oz I (50on (500rl I I I [4] 0--0 A-A-A I (500r) (5001) I I I (6) D---0 A -Oz-A I ,.. \ (500/) (500r) I I \ I ... I \ I [s] tz----0 A-A-A I I / \ l (150Drl (1500r) I I I I I [10] A---6 A-Oz-A I I (1500r) (1500/) I I I I I I I I I I I I I I I I ,. I I I I I I I I I i. I I I I I I I I I I I I I ·--------· A c D E F MATING PERIODS The University of Texas Publication TABLE 3 The frequencies of induced translocations Experiment A B c D E F G 02-02-02 (500 r ) (500 r) a b c d 104-0 14 1.3 0.7 2.2 1005 19 1.9 1.2 2.9 587 44 7.5 5.5 9.9 500 57 11.4 8.8 14.5 132 3 2.3 0.5 6.5 370 1 0.3 .01 1.5 2 0 2-A-02 (500 r) (500 r) a b c d 907 23 2.5 1.6 3.8 947 28 3.0 2.0 4.2 795 75 9.4 7.5 11.7 151 25 16.6 10.6 22.5 481 48 10.0 7.5 13.0 179 5 2.8 0.9 6.4 447 1 0.2 .01 1.2 3 a 252 194 304 286 306 288 296 02 ­(Or) A­ 0 2 (Or) b c d 0 0.0 0.0 1.5 0 0.0 0.0 1.9 0 0.0 0.0 1.2 0 0.0 0.0 1.3 0 0.0 0.0 1.2 0 0.0 0.0 1.3 0 0.0 0.0 1.2 4 A-A­A (500 r) (500 r) a b c d 761 4 0.5 0.1 924 5 0.5 0.2 772 11 1.4 0.7 802 6 0.7 0.3 791 6 0.7 0.3 815 3 0.4 0.1 794 0 0.0 0.0 1.3 1.3 2.5 1.6 1.7 1.1 0.5 5 a 229 253 214 192 179 167 64 A-A-A b 0 0 0 0 0 0 0 (0 r) (0 r) c d 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.6 1.5 1.7 1.9 2.0 2.2 5.8 6 a 1199 1064 1032 1034 1025 1072 1045 A-02 ­A (500 r) (500 r ) b c d 6 0.5 0.2 1.1 13 1.2 0.7 2.1 17 1.6 1.0 2.6 11 1.1 0.5 1.9 0 0.0 0.0 0.4 0 0.0 0.0 0.3 0 0.0 0.0 0.4 7 a 346 388 371 354 348 341 372 A-02-A (0 r) (0 r) b c rl 0 0.0 0.0 0 0.0 0.0 0 0.0 0.0 0 0.0 0.0 0 0.0 0.0 0 0.0 0.0 0 0.0 0.0 1.1 1.0 1.0 1.0 1.1 1.1 1.0 8 A-A-A (1500r) (1500r) a b c d 1385 71 5.1 4.0 989 134 13.5 11.5 975 203 20.8 18.4 68 16 23.5 14.8 1247 6 0.5 0.2 1273 0 0.0 0.0 1402 0 0.0 0.0 6.4 15.9 23.5 36.6 1.1 0.3 0.3 10 A-02-A (1500r) (1500r) a b c d 1426 102 7.2 5.9 1066 169 15.9 13.7 872 198 22.7 20.0 162 36 22.2 16.3 1098 3 0.3 .05 1255 0 0.0 0.0 1243 0 0.0 0.0 8.6 18.0 25.7 29.7 0.8 0.3 0.3 a=Number of sperm tested. b== Number of translocations. c= Per cent of translocations. d= Lo\ver and upper limits of 95% expectation. more than does argon, then the amount of biological damage observed should be lower with oxygen than with argon between the doses. It seems possible that with a 20-minute period between the doses some of the recovery that may have occurred was disturbed. Liining (1958) suggested that a 30'--40 minute period between doses will prevent the disturbing of any healing which may occur under · anoxic conditions. Using 40 minutes between the fractionated doses, Alexander and Bergendahl (1964) obtained greater amounts of biological damage in argon than with a continuous dose. If healing had occurred in anoxia as suggested by Luning, then lower values would have been obtained. When oxygen was present, however, the damage induced with the fractionated dose was less than the damage induced with continuous dose. This seems to indicate that oxygen enhanced the healing of chromosome breaks and that the 40-minute period used by Alexander and Bergendahl ( 1964) had allowed more healing to occur than the 20-minute period used in this investigation. Again, as in the sex-linked lethals, the "fixation time" may be a factor ( Sobels, 1965). These results would lead one to consider the possibility that broken ends induced in oxygen "behave" differently from those induced in argon. One may even suggest that the breaks induced by X-rays in argon are not as amenable to repair as those induced in oxygen or air. The question may be raised as to what portion of the damage is due to the influence of the pressure and gas alone without the X-ray treatments. It can be seen from Tables 1-3, Figs. 1-3 (Experiments 3, 5, 7, 9, and 11) that it is very negligible indeed. Not a single trans'iocation was found, and the average rate of sex-linked recessive lethals is only 0.05%. The percentages of egg-hatch range from 90--96%, which is as good as one can get without any gas or pressure treatment. Response of different spermatogenic cells .The response to radiation of cells in spermatogenesis varies according to the s1:!3.ge of development of the cells. This has been shown repeatedly by other in­vestigators. The data show that for recessive lethals and translocations in flies X-rayed in oxygen, early spermatids (D, Experiment 2) or meiotic stage (E, Experiment 1) were most sensitive. In Experiment 2 late spermatids (C) and meiotic stage (E) were the next most sensitive, while in Experiment 1 early spermatids (D) and spermatocytes (F) were next. This discrepancy was probably due to possible greater heterogeneity of the germ cell population at the tim~ of treatment in Experiment 1 owing to the age difference (four hours) among the males. Sper­matogonia (G) were the least sensitive. For dominant lethals meiotic cells (E) were the most heavily damaged. It should be noted, however, that with the in­argon X-rayed flies (Experiments 4 and 6) the differential response is much less obvious. This agrees well with Chang's ( 1962) observation. Oster ( 195 7) reported a greater reduction of translocation frequencies from "in air" to "in nitrogen" treatments in spermatids than in mature sperm of D. melanogaster. A greater increase, on the other hand, was observed for sperm than for spermatids from "in air" to "in oxygen" treatments. Oster attributed this to the higher oxygen tension present in spermati.ds than in mature sperm. The University of Texas Publication TABLE 4 Type of translocations and distributions of breaks involved Translocations Minimum number breaks Number Number spem1 Experiment T tested 2 ,3 y,z y,3 y,2,3 y 2 1A B 14 1040 12 0 2 0 2 12 14 c 19 1005 13 2 3 6 16 17 D 44 587 36 2 5 8 39 42 E 57 500 42 7 4 4 15 53 50 F 3 132 1 1 0 2 2 2 G 370 0 0 0 0 Total 138 3634 105 12 15 6 33 123 126 Per cent 76.1 8.7 10.9 4.3 11.7 43.6 44.7 2A 23 907 18 4 1 0 5 22 19 B 28 947 21 1 5 1 7 23 27 c 75 795 53 6 14 2 22 61 69 D 25 151 19 4 2 0 6 23 21 E 48 481 27 12 9 0 21 39 36 F 5 179 3 1 1 0 2 4 4 G 447 0 0 0 0 Total 205 3907 142 28 32 3 63 173 177 Per cent 69.3 13.7 15.6 1.5 15.3 41.9 42.8 4A 4 761 2 1 0 2 3 3 B 5 924 4 0 0 1 4 5 c 11 772 6 4 0 5 10 7 D 6 802 4 0 2 5 5 E 6 791 4 1 0 2 5 5 F 3 815 3 0 0 0 0 3 3 G 0 794 0 0 0 0 0 0 0 Total 35 5659 23 7 5 0 12 30 28 Per cent 65.7 20.0 14.3 0.0 17.1 42.9 40.0 6A 6 1199 3 2 1 0 3 5 4 B 13 1056 10 2 1 0 3 12 11 c 17 1032 14 0 3 0 3 14 17 D 11 1033 8 2 1 0 3 10 9 E 0 1025 0 0 0 0 0 0 0 F 0 1072 0 0 0 0 0 0 0 G 0 1044 0 0 0 0 0 0 0 Total 47 7461 35 6 6 0 12 41 41 Per cent 74.5 12.8 12.8 0.0 12.8 43.6 43.6 BA 71 1385 51 7 13 0 20 58 64 B 134 989 99 15 12 8 35 122 119 c 203 975 137 24 26 16 66 177 179 D 16 68 13 1 3 15 15 E 6 1247 3 1 2 0 3 4 5 F 0 1273 0 0 0 0 0 0 0 G 0 1402 0 0 0 0 0 0 0 Total 430 7339 303 48 54 25 127 376 382 Per cent 70.5 11.2 12.6 5.8 14.4 42.5 43.1 10A 102 1426 72 12 16 2 30 86 90 B 169 1066 128 14 17 10 41 152 155 Elequin: Modification of Genetic Damage in Drosophila c 198 872. 12.8 2.4 2.8 18 70 170 174 D 36 162. 2.0 4 8 4 16 2.8 32. E 3 1098 z 0 0 1 z 3 F 0 12.55 0 0 0 0 0 0 0 G 0 12.43 0 0 0 0 0 0 0 Total 508 712.2. 350 54 70 34 158 438 454 Percent 68.9 10.6 13.8 6.7 15.1 41.7 43.2. Grand Total 1363 3512.2. 958 155 182. 68 405 1181 12.08 PercE>nt 70.3 11.4 13.4 4.9 14.4 42..3 43.3 Sobels (1965) reported that nitrogen favored the repair in mature sperm while oxygen favored it in spermatids. This would explain, at least partially, the more or less even responses of different cell types observed here when the flies were irradiated in argon, if it is assumed that anoxia, not nitrogen per se, favors the repair in the mature sperm. Sensitivity differences in recovery of Y-A( autosome) translocations and chromosomes involved The types of translocations and the chromosomes involved in all tests are summarized in Table 4. In all tests, a total of 1363 translocations was recovered. Of these, 958 (70.3%) involved chromosomes 2 and 3, 155 (11.4%) involved chromosomes Y and 2, 183 (13.4%) involved Y and 3, and 68 (4.9%) were complex translocations involving Y, 2, and 3. In the translocations involving the Y chromosome there was no observable preference on the part of Y chromosome for rejoining with either one of the two major autosomes. This is in agreement with the observations made by Falk (1962). No attempt was made here to analyze further the 68 Y-2-3 translocations. Since the Y-2-3 translocations were not ana­lyzed further and comprise a small portion of the total translocation, plus the observation that no preference was shown by Y chromosome for either chromo­some 2 or 3, from here on all the translocations involving the Y will be referred toasY-A (Aforautosome). From Table 4 one can see that the frequency of 2-3 translocations is 2.4 times as great as the Y-A translocations. Falk ( 1962) contended that differences in recovery of Y-A and 2-3 translocations are perhaps an indication of differences in rejoining characteristic of heterochromatic and euchromatic breaks. He found that prolonged nitrogen treatment between two doses of irradiation in air 14 hours apart delayed the rejoining of most breaks involved in 2-3 translocations, while the rejoining of those involved in Y-A translocations was not delayed to a large extent. No such delay was found by him with similar air treatment between the two irradiations. He also found that with an extended nitrogen treatment of 8 to 9 hours prior to irradiation of pupae, thre was a significant increase in the frequency of 2-3 translocations, but not the Y-A translocations. He suggested that this indicates a lower sensitivity of the "induced (effective) breakability" of the heterochromatin than of the euchromatin in the stage (spermatid) tested. The results of this investigation, however, do not concur with Falk's conten­tions. If it were so, then the significant increase in the frequency of 2-3 transloca­tions in the spermatid (C and D) observed in experiments with the two irradia­ The University of Texas Publication tions given in oxygen and with argon treatment in between (Experiment 1) over that with oxygen in between (Experiment 2) should also be reflected here. One can see from Table 4 that such is not the case. It would be interesting to find out what effect oxygen, argon, and other gases would have using the same pro­cedures that Falk used. In D. virilis, Baker ( 1949) found that interchange involving the Y chromosome is less than half as great (0.4) as one involving any one of the four major auto­somes, which are about equally likely to participate in any interchange. Clayton (1962) reported a slightly higher figure; the Y was involved in the interchange about 0.6 as frequently as the autosomes. On the other hand, earlier investigators stated that the breaks located in Dro­sophila are distributed along the chromosomes in proportion to their length (Bauer et al., 1938; Bauer, 1939, cited by Kaufmann, 1946). Kaufmann and Demerec (cited by Kaufmann, 1954) found in a study of salivary gland chromo­somes of male larvae that breakage in the Y chromosome and in the arms of autosomes occur with approximately equal frequencies. Kaufmann (1946, 1954) reported a close correspondence between the frequencies of breaks in the largely heterochromatic Y and the largely euchromatic autosomes, indicating that breaks, regardless of the nature (euchromatic or heterochromatic) of chromosomes, are distributed at random in proportion to the length of the chromosome at the time of irradiation. Haas et al. (1954) found that in D. virilis the second chromo­some, which is slightly longer than the other autosomes, was more frequently involved in interchanges (23%) than the other autosomes tested (21 %). However, the results show (Table 4) that the Y chromosome is involved in 14.4% of the breaks, or about% as frequently as the two autosomes tested. This is similar to the findings of Baker ( 1949) and Clayton ( 1962). Chromosomes 2 and 3 are about equaily likely to participate in an interchange. Nevertheless, it should be noted that in this investigation only the Y and autosomes 2 and 3 are tested. The method does not allow the testing of chromosome 4 and the X chromo­some. This, plus the fact that no cytological analysis of breaks was made, may bias the distribution of breaks being discussed. SUMMARY Drosophila melanogaster (Oregon-R) males, 17-24 hours old, were X-rayed with two equal doses of 500 or 1,500 r, separated by an interval of 20 minutes. The males were X-rayed in either oxygen or argon at 1 atmosphere of pressure. During the 20 minutes between the two X-ray fractions, flies were kept in oxygen or argon. The flies were pre-treated for 10 minutes before the first fraction and post-treated for 20 minutes after the second fraction. Each male was mated individually and simultaneously within an hour of treatment to: (1) 2 o::-3 Muller-5 females; (2) 2 or 3 bw, st females; (3) 1 heterozygous female (Fi of Austin strain~ X Canton-S strain~ ) . Every two days, each male was remated to the same number of three kinds of virgin females. All together, seven such con­secutive, 2-day mating periods were obtained for each series of experiment. The offspring (F1 ) from such mating periods carried chromosomes which were X-irra­diated in different stages of spermatogenesis in the males (P1 ) . Therefore, the Elequin: Modification of Genetic Damage in Drosophila pattern of X-ray damage could be correlated to the different stages of sperma­togenesis. Chromosome damage was measured by tests for sex-linked recessive lethals, dominant lethals, and translocations. When the X-rays were given in oxygen, the damage was significantly greater in the spermatid stage if the flies were exposed to argon (as opposed to oxygen) between the two X-ray fractions. With the X-rays given in argon, no significant difference was found in the damage measured by sex-linked recessive lethals and translocations between the series with argon and the series with oxygen between the two doses. For the dominant lethal tests, the results were the opposite of what might be expected from the above results, and this is discussed. For the in-oxygen X-ray treatments peak sensitivities occurred at the early spermatid and meiotic stages. In the absence of oxygen during X-ray treatments, however, the differential response to X-rays was very much less obvious. Of the total translocations recovered, the Y-chromosome was involved in less than half and showed no preference for either one of the two major autosomes (2 and 3). The frequency of the translocations involving chromosomes 2 and 3 was 2.3 times as great as the Y-A translocations. Chromosomes 2 and 3 were about equally affected by the X-ray treatment. ACKNOWLEDGMENTS The writer wishes to express her sincere gratitude for the encouragement and guidance of Prof. Wilson S. Stone during this investigation. My special apprecia­tion is also due Mrs. Hei Yung Yang and Mrs. Donya Conine for their technical assistance. REFERENCES Abrahamson, S. 1962. Further studies on the influence of oxygen on X-ray induced rearrange­ ment in Drosophila oocytes. Int. J. Rad. Biol., 4: 113-125. Alexander, M . L., and J. Bergendahl. 1962. Biological damage in the mature sperm of Droso­ phila virilis in oxygen and nitrogen with different dose intensities of gamma rays. Genetics, 47: 71-84. ----. 1964. Dose rate effects in the developing germ cells of male Drosophila. Genetics, 49: 1-16. Auerbach, C. 1954. Sensitivity of the Drosophila testis to the mutagenic action of X-rays. Z. indukt. Abstamm. und Vererb., 86: 113-125. Bacq, Z. M., and P. Alexander. 1961. Fundamentals of Radiobiology. New York: Pergamon Press. Baker, W. K. 1949. The production of chromosome interchanges in Drosophila virilis. Genetics, 34: 167-193. Bauer, H., M. Demerec, and B. P. Kaufmann. 1938. X-ray induced chromosome alterations in Drosophila melanogaster. Genetics, 23: 610-630. Beatty, A. V., and J. W. Beatty. 1959. Metabolic inhibitors and chromosome rejoining. Amer. J. Botany, 46: 317-323. ----. 1960. Postirradiation effects on chromosomal aberrations in Tradescantia micro­spores. Genetics, 45: 331-344. The University of Texas Publication Chang, Tsueng-Hsing. 1962. Further observations on the relation between gas pressure and the X-ray damage in Drosophila melanogaster. Univ. of Texas Publication No. 6205, pp. 385-393. Clayton, F. E. 1962. Effects of X-ray irradiation in Drosophila virilis at different stages of spermatogenesis. Univ. of Texas Publication No. 6205, pp. 345-373. Demerec, M. 1937. Relationship between various chromosome changes in Drosophila melano­gaster. Cytologia, Fujii Jubilee Volume: 1125-1132. Falk, R. 1959. Delay in joining of X-ray induced breaks by anoxia in D. melanogaster. Genetics, 44: 509. -----. 1962. Nitrogen-treatment effects of rearrangement-induction patterns in Droso­phila melanogaster. Int. J . Rad. Biol., 4: 437-455. Gray, L. H. 1959. Cellular radiobiology. Rad. Res., Suppl. 1, pp. 73-101. Haas, F. L., E. Dudgeon, F. E. Clayton, and W. S. Stone. 1954. Measurement and control of some direct and indirect effects of X-radiation. Genetics, 39: 453-471. Kaufmann, B. P. 1946. Organization of the chromosome. I. Break distribution and chromosome recombination in Drosophila melanogaster. J. Exptl. Zool., 102: 293-320. -----. 1954. Chromosome aberrations induced in animal cells by ionizing radiations. Chap. 9, Radiation Biology. A. Hollaender, ed. New York: McGraw-Hill, pp. 627-711. Kihlman, B. 1955a. Chromosome breakage in Allium by 8-ethoxycaffeine and X-rays. Exptl. Cell Res., 8: 345-368. -----. 1955b. Studies on the effect of oxygen on chromosome breakage induced by 8­ethoxycaffeine. Exptl. Cell Res., 8: 404-407. LaChance, Leo E. 1961. Post-irradiative effects of nitrogen and carbon monoxide on hatchability of Habrobracon eggs treated in first meiotic metaphase. Int. J. Rad. Biol., 4: 15-20. Liining, K. G. 1958. Blocking of the recovery of chromosome breaks induced in Drosophila melanogaster sperm. Proc. Int. Conf. on Peaceful Uses of Atomic Energy, Geneva, Vol. 22, pp. 333-335. Oster, I. I. 1957. Suggested mechanism underlying the differential radiosensitivity of cells having condensed chromosomes. Genetics, 42: 387. Powers, E. L., R. B. Webb, and B. F. Kaleta. 1960. 0 2 and NO as modifiers of radiation injury in spores of B. megatherium. Proc. Natl. Acad. Sci., U. S. A., 46: 984-995. Slizynska, H., and B. M. Slizynski. 1947. Genetical and cytological studies of lethals induced by chemical treatment in Drosophila melanogaster. Proc. Royal Soc. Edinburgh, B, 62: 234-242. Slizynski, B. M . 1938. Salivary chromosome studies of lethals in Drosophila melanogaster. Genetics, 23: 283-290. Sobels, F. H . 1965. Radiosensitivity and repair in diffeernt germ cell stages of Drosophila. Proc. XI Internat. Congress of Genetics, The Hague, The Netherlands, Sept., 1963, Vol. 2: 235-255. Sobels, F. H ., and A. D. Tates. 1961. Recovery from pre-mutational damage of X-irradiation in Drosophila spermatogenesis. J. Cell. Comp. Physiol., 58(Suppl.): 189-196. Somner, N. F., M. Creasy, R. J. Romani, and E. C. Maxie. 1964. An oxygen-dependent post­irradiation restoration of Rhizopus stolonifer spnrangiospores. Rad. Res., 22: 21-28. Spen cer, W. P., and C. Stern. 1948. Experiments to test the validity of the linear r-dose/muta­tion frequency relation in Drosophila at low dosage. Genetics, 33: 43-74. Elequin: Modification of Genetic Damage in Drosophila Stevens, W. L. 1942. Accuracy of mutation rates. J. Genet., 43: 301-307. von Borstel, R. C. 1963. Effects of radiation on germ cells of insects: dominant lethals, gamete inactivation and gonial-cell killing. Radiation and Radioisotopes Applied to Insects of Agri­cultural Importance, pp. 367-383. International Atomic Energy Agency, Vienna, 1963. von Borstel, R. C., and M. L. Rekemeyer. 1959. Radiation induced and genetically contrived dominant lethality to Habrobracon and Drosophila. Genetics, 44: 1053-1074. Wolff, S., and K. C. Atwood. 1954. Independent X-ray effects on chromosome breakage and reunion. Proc. Natl. Acad. Sci., U. S. A., 40: 187-192. Wolff, S., and H. E. Luippold. 1955. Metabolism and chromosome-break rejoining. Science, 122: 231-232. ----. 1956. The biochemical aspects of chromosome rejoining. Progress in Radiobiology. J. S. Mitchell, B. E. Holmes, and C. L. Smith, eds. Edinburgh: Oliver and Boyd, pp. 217­ 221. ----. 1958. Modification of chromosomal aberration yield by post-irradiation treatment. Genetics, 43: 493-501. VIII. Descriptions and Notes on Hawaiian Drosophilidae (Diptera) .1 2 D. ELMO HARDY Volume 12 of the Insects of Hawaii (Hardy, 1965) deals with 400 species of Drosophilidae from Hawaii. 18 of these are immigrant species and 382 are con­sidered as endemic to the Hawaiian Islands. Since this book went to press inten­sive field work has been done in connection with a project on evolution and genetics of Hawaiian Drosophilidae which is now under way. This project has involved a number of field collectors and a wide assortment of collecting technics and during the summers of 1964 and 1965 approximately 100 additional new species have been collected. The species described here are those for which names are needed for the studies being reported on by some of the collaborators on the project. I am also giving notes on some apparent species aberrations; it is possible that these are actually sibling species and it appears likely that many siblings will be demonstrated when we have more detailed knowledge concerning the native fauna. It is now obvious that at least 600 species of Drosophilidae occur in the Hawaiian Islands. The work which is planned over the next five or more years will greatly increase our knowledge of this remarkable fauna. I am indebted to the following colleagues for their valuable assistance in the field work: Drs. H. Carson, F. Clayton, W. B. Heed, H. Stalker, H. T. Spieth, L. H. Throckmorton, M. R. Wheeler, and Messrs. D. Gubler and K. Y. Kaneshiro. The illustrations have been prepared by Misses Noreen Naughton and Aileen Matsuyama. I am very appreciative of this valuable assistance. The work was made possible by National Institute of Health Grant GM 10640. Drosophila achlya, new species (Figs. 1a-d) This species runs near D. iki Bryan and superficially resembles this species. It differs distinctly by having yellow antennae, a large brown basal wing spot, no brown marking on them crossvein, and no rim on the labellum, as well as in other details. MALE. Head: Mostly dark colored; the front is velvety black with a faint gray sheen as seen in some lights. The eye orbits are distinctly gray. The occiput is brown to black in ground color and covered with gray pollen. The face is brown,. tinged with yellow to rufous, especially in the furrows. The median portion of the face is slightly raised. The lower margin of each gena, the clypeus and the palpi are black. The mentum is dark brown to black. The labella are pale yellow and fleshy, with no apparent ornamentation. A number of rather prominent black hairs are present along the outside margin of each palpus and no apical bristles are present. The palpus is shaped as in Figure ta. The anterior reclinate bristles 1 Published with the approval of the Director of the University of Hawaii Agricultural Experiment Station as Technical Paper No. 759. 2 Department of Entomology, University of Hawaii, Honolulu. The University of Texas Publication are about equal to the proclinates and are situated slightly above the latter. The antennae are pale yellow, contrasting from the remainder of the head. The aristae each have six dorsal and two ventral rays in addition to the apical fork. Two rather prominent bristles are situated at the upper portion of each vibrissal row. Thorax: Predominantly dark brown, covered with gray-brown pollen, and tinged with yellow to rufous in the ground color of the anterior portion of the mesono­tum. The anterior dorsocentral bristles are two-thirds to three-fourths as long as the posteriors and situated about opposite the second pair of supraalars. Two strong humeral bristles are present. The sternopleural bristles are well developed, the anterior bristle is about three-fourths as long as the posterior, and a rather prominent black seta is situated half-way between the sternopleural bristles. The halteres are pale yellow. Legs: Mostly yellow; the coxae are yellow-brown, the middle femora are predominantly dark brown to black, tinged faintly with yellow to rufous; the bases of the middle tibiae are also brown to black and the hind 0 .225 mm FIG. 1. Drosophila achlra n. sp. a. palpus; b. wing; c. male genitalia, lateral; d. female genitalia, lateral. Hardy: New Species of Hawaiian Drosophilidae femora are yellow, tinged lightly with brown; the apical segments of the tarsi are dark brown. The front legs are not ornate. The tarsi lack ciliation; the basitar­sus is rather slender, slightly more than half as long as the tibia, and two times longer than the second tarsal segment. Wings: Hyaline except for the brown bases and apices and except for a very faint tinge of brown on the m crossvein; the apical one-fourth of the wing is covered by a large brown marking (Fig. 1 b). The third costal section is five times longer than the fourth and the costal fringe extends about two-fifths the distance between the apices of veins R2+3 and R4+5 . The last section of vein M 1+2 is 1.25 times longer than the penultimate section. The ninth tergum is greatly narrowed over the dorsal portion and shaped as in Figure 1c. Length: Body, 2.75-3.0 mm.; Wings, 3.2-3.7 mm. FEMALE. Similar to the male execpt that the apical brown mark on the wing is not quite so extensive, the third antennal segment is tinged with brown, and the bristles of the vibrissal row are stronger; the uppermost bristle is approx­imately equal in size to the ocellars. The ovipositer blades extend slightly beyond the apices of the anal plates, are subacutely pointed and are armed with teeth around the apices (Fig. 1d). Length: Body, 3.2mm.; Wings, 3.9 mm. Holotype male and allotype female from Waikamoi, Maui, 4,000', Oct. 1, 1964 (H. T. Spieth). Twenty-one paratypes, 1!2. males and 9 females, same locality as type, collected July, August and October 1958 and 1964 (H. T. Spieth, D. E. Hardy, L. H. Throckmorton). Type, allotype and some paratypes in the B. P. Bishop Museum. The remainder of the paratypes in the collections of the U.S. National Museum, British Museum (Natural History), and the University of Hawaii. Drosophila amplilobus new species (Figs. 2a-f) While doing a comparative study of the male genitalia of some of the Hawaiian Drosophila in the summer of 1964, Dr. Haruo Takada of Kushiro Women's Col­lege, Kushiro, Japan, discovered that the male genitalia of specimens determined as D. crassifemur Grimshaw from Kauai were distinctly different from those of this species which he had studied from other islands. I have now studied the geni­talia of specimens from five of the main islands and it is obvious that a complex of species is present which fit the previous description of crassifemur. The species from Kauai is being successfully raised in culture and a name is needed for it at this time. Dr. Takada is publishing his detailed studies of the genitalia of the crassifemur complex elsewhere in this Bulletin. Specimens from Maui, Molokai, and Hawaii may possibly represent the same species (typical crassifemur) al­though some differences have been noted in the development of the structures surrounding the aedeagus, and other details of the genitalia. The significance of these differences is not yet understood and it will be necessary that these popu­lations be established in cultures so that they can be more thoroughly studied. FIG. 2. Drosophila amplilobus n. sp. a. front leg of male, lateral; b. thorax dorsal view; c. female genitalia, lateral; d. male genitalia, lateral; e. lamella of ninth tergum of male; £. male genitalia, ventral. D. crassifemur Grimshaw. g. male genitalia, ventral. Hardy: New Species of Hawaiian Drosophilidae Specimens from Oahu represent a distinct species. This is not being described until further collections can be made and until the species can be established in culture. As pointed out in the papers being presented by Drs. Throckmorton and Spieth (this Bulletin), it appears evident that crassifemur and related species should be placed in the genus Scaptomyza; on the basis of internal morphology, egg char­acters, and mating behavior these seem to be Scaptomyza rather than Drosophila. On the basis of external morphology, they fit the present concept of the genus Drosophila. It is obvious that a revision of the generic concepts is needed but I feel that it is premature to make these rather drastic changes until more complete information is obtained concerning all of the species involved. The genitalia are unlike anything which I have seen in Drosophila or Scaptomyza. This complex of species is another striking example of the strange speciation trends exhibited by many of the Hawaiian drosophilids. Externally, this species looks like typical crassifemur. I am unable to find any reliable characters which will separate these except those of the male genitalia. In situ the males of amplilobus can usually be differentiated by the greater size of the membranous lobe which extends from each margin of the ninth tergum; this almost completely obscures the clasper due to its large size. In relaxed or dissected specimens, marked differences can be seen in the claspers, the aedeagus, and surrounding structures. The most striking feature is the presence of a long, slender lobe arising from the base of each clasper and the lack of pointed pro­jections on the parameres (compare Figs. 2d, 2f, and 2g). MALE. Fitting the description of crassifemur so nearly that it would be repiti­tious to describe the general features. The specimens are generally dark in color. The femora are black except for the yellow apices, and the yellow ventral por­tions of the first pair; the venter of each front femur is densely yellow pubescent as in crassifemur. The mesonotum typically has five black vittae extending the full length, except for the interruption of the lateral vittae just beyond the suture; these are set off by yellow-gray pollinose areas as in Figure 2b. The disc of the scutellum is dark brown to black, the margin is narrowly yellow. The humeri are clear yellow. The arista has five or six dorsal and one or two ventral rays in ad­dition to the apical fork. Each palpus has three rather prominent bristles on the outside margin. The front leg is as in Figure 2a. The basitarsus is one-fourth as long as the tibia. The brown to black bands on the middle and hind tibiae are very prominent. The wings are pale brown. The third costal section is four times longer than the fourth and the costal fringe extends about three-fourths the dis­tance between the apices of veins R2+3 and R4+5. The last section of vein M1+2 is about .5 longer than the penultimate section. The abdomen is typically dark brown to black with narrow gray margins on the apices of the terga. The genitalia are dark brown to black, except for the white lamellae arising from the ventral margins of the ninth tergum. These lamellae are sclerotized and brown to black along their anterior borders and the posterior and apical portions are ex­panded ventrally into a conspicuous paper-thin lobe which almost completely covers the clasper; this lobe is microscopically reticulated, when seen under high power (Fig. 2e). The other details of the genitalia are as in Figures 2d and 2f; The University of Texas Publication compare with figure of typical crassifemur, 2g. The aedeagus is large and fleshy and has a pointed projection on the dorsal surface (Fig. 2f). Length: Body, 3.5-3.7 mm.; Wings, 3.9-4.2 mm. FEMALE. The body coloring and markings are as in the male. The ventral portion of the front femur, however, is brown to black in ground color, densely covered with gray pubescence. The ovipositor blades are short and pointed, ex­tending scarcely beyond the apices of the anal plates and shaped as in Figure 2c. Length: Body, 4.0 mm.; Wings, 4.2 mm. Field collected specimens of this species show considerable range in size, from the typical given above to approximately 2.5 mm. for the body and 2.7 mm. for the wings. Holotype male, allotype female and 17 paratypes, 10 males, 7 females from laboratory culture WH47.1, March, 1965 (K. Kaneshiro). The original culture was collected at Halemanu Valley, Kauai, by Dr. Frances Clayton. Also about 50 paratypes, predominantly males, from the following localities in the Kokee area of Kauai: Kokee, 3600' elevation, July, 1952, in banana-bait trap (D. E. Hardy); Mt. Waialeale Trail, 4500', August, 1953 (D. E. Hardy); Halemanu Swamp, Aug., 1953 (D. E. Hardy); Halemanu Swamp, Aug., 1953 (D. E. Hardy); Kainamanu, 3800', July, 195'2 (D. E. Hardy); Nualolo Valley, July, 1952, 3400' (D. E. Hardy); Poomau Valley, 3000', July, 1952 (D. E. Hardy); and Waia­koali Stream, July 14, 1937, South Fork, 3500' at light (E. C. Zimmerman). Type, allotype and a series of paratypes in the B. P. Bishop Museum. Remain­der of paratypes distributed among the following collections; U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. Drosophila atroscutellata new species (Figs. 3a-h) Because of the short, broad second segment of the front tarsus of the male and the dark markings in the apex of the wing, this species would run near D. con­formis Hardy, from Hawaii. The two are apparently not related, however, and atroscutellata is readily differentiated by the predominantly yellow coloring of the thorax, with the brown contrasting scutellum; by the black apices of the mid­dle tibiae; and by the differences in the front tarsi (Fig. 3c) and the wings (Fig. 3b). Superficially it closely resembles kokeensis n. sp., from Kauai. The leg characteristics and the wing venation, however, are very different in these two species (Figs. 3b, 3c, 7 d and 7f). MALE. Head: Predominantly yellow, except for the reddish brown compound eyes and black ocellar triangle. The vertex is brown to black, tinged with yellow, the upper eye orbits are brown to black with a faint yellowish tinge. The occiput is brown with yellow in the ground color. The mouthparts are black; the clypeus is light brown. A thin line of brown to black coloring extends down the inner margin of each gena. The antennae are pale yellow; each arista has six dorsal and two ventral rays in addition to the small apical fork. The anterior reclinate bristles are subequal to the proclinates and situated opposite the latter. The up­ Hardy: New Species of Hawaiian Drosophilidae FIG. 3. Drosophila atroscutellata n. sp. a. mouthparts; b. wing; c. front tarsus of male, dorsal; d. front tarsus, lateral; e. middle tibia of male, lateral; f. male genitalia, ventral; g. male geni­talia, lateral; h. female genitalia, lateral. The University of Texas Publication permost bristle of the vibrissal row is approximately equal in size to the anterior reclinates. Two or three rather prominent black setae are also developed in the vibrissal row. The face is white, tinged faintly with yellow. The median portion is slightly raised. The palpi are rather short and thick, only about two times longer than wide, lacking bristles but with black setae around the apical margin (Fig. 3a). The mentum is rectangular, about two times longer than wide. The labella are fleshy and prominent. In relaxed and cleared specimens the mouth­parts are yellow-brown; in situ they appear black. Thorax: Entirely pale yellow except for the dark brown to black scutellum, a brown spot on the extreme anteromedian portion of the mesonotum, a tinge of brown behind each humerus, and the brownish yellow metanotum. A faint tinge of brown is also present on each humerus. Two prominent humeral bristles are present. Both sternopleural bristles are moderately strong; the anterior bristle is equal in size to the anterior dorsocentral bristle. The anterior dorsocentrals are approximately three-fourths as long as the posterior bristles and are situated slightly in front of a line drawn between the second pair of supraalars. Legs: Entirely pale yellow except for the conspicuously blackened apices of the middle tibiae (Fig. 3e); the middle tarsi are tinged with brown. The front tibiae are not ciliated. The front basitarsus is about two-fifths as long as the tibia, terminates in a prominent point at the dorso­apex, and has a row of about seven prominent, black, anterior setae extending from the apex two-thirds the distance to the base of the segment (Fig. 3c) . The second tarsal segment is short and broad, flattened laterally, about as wide as long. Wings: Hyaline except for a prominent brown mark at the apex, filling the apical portions of cells R3 and R5 and the upper apical portion of cell 2nd M2 (Fig. 3b). Third costal section nearly five times longer than the fourth and the costal fringe extending two-fifths the distance between the apices of veins R2+3 and R4+5. The last section of vein MJ+2 is 1.76 longer than the penultimate section. Abdomen: First segment entirely yellow, venter yellow; remainder of terga yellow on the side, brown down the median portions. The genitalia are yellow. The ninth tergum is greatly narrowed above, reduced to just a narrow line over the dorsal portion (Fig. 3g). The anal plates are acutely pointed ven­trally. The claspers are plainly visible from a lateral view (Fig. 3g). The ventral aspects of the genitalia are as in Figure 3f; the claspers are densely setose on their inner margins. Length: Body, 2.15 mm.; Wings, 2.5 mm. FEMALE. Lacking the prominent brown marking in the wing, with the middle tibiae only faintly marked with brown at the apices, the mesonotum predomi­nantly rufous, and the front and antennae brown. The uppermost bristle of the vibrissal row is almost equ·al in size to the upper reclinate bristles of the front. The first tergum and the bases of terga two to four are yellow to rufous; the apices of two to four are black. Terga five and six are yellow to rufous. The ovi­positor blades are moderately slender, sharp pointed at the apex and with promi­nent teeth on the margin (Fig. 3h) . Holotype male and allotype female from Halemanu Valley, Kauai, 4,000', August 28, 1964 (H. T. Spieth and L. H. Throckmorton). Fifty-six paratypes, Hardy: New Species of Hawaiian Drosophilidae 35 males and 21 females, same locality as type, March 5, 1964, to August 28, 1964 (F. Clayton, M. R. Wheeler, D. E. Hardy, and H . T. Spieth) and Kokee, Kauai, 3600', March 5 to June 22, 1964 (H. L. Carson, F. Clayton, and M. R. Wheeler). Type, allotype and a series of paratypes in the collection of the B. P. Bishop Museum. The remainder of the paratypes are in the collections of the U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. Drosophila ceratostoma new species (Figs. 4a, c-f) In my key to the Hawaiian Drosophila, this runs to couplet 68 near asketostoma Hardy and it appears to show some relationship to this species. D. ceratostoma differs by lacking vittae on the mesonotum; lacking long ciliation on the front tarsi; by the very different mouthparts (Fig. 4a); as well as by the smaller size and other details. MALE. Head: The front is black except for the yellow anterior margin; the yellow marking extends almost to the proclinate bristles. The area between the ocellar triangle and the region occupied by the frontal bristles is velvety black. The orbits and ocellar triangle are black in ground color, covered with brownish­gray pollen. The face is yellow, except for a faint tinge of brown on the epistoma. A distinct carina extends down the median portion of the face. The upper two­thirds of the occiput is brown to black and densely covered with gray pollen; the lower portion is yellow. The clypeus is yellow-brown and the mouthparts are predominantly yellow. The labella are strangely ornate (Fig. 4a). The mouth­parts seem to show some resemblance to asketostoma but the development of the spines and processes is quite different (refer to Fig. 4b of asketostoma). Each palpus has a number of short, black setae and one small bristle at the apex. The anterior reclinate bristles are about two-thirds to three-fourths as long as the proclinates and are situated opposite the latter. The antennae are yellow-brown; each arisia has four dorsal and two ventral rays in addition to the apical fork. The oral vibrissae are poorly developed, consisting of short, black bristles which are slightly smaller than the bristles of the occipital row. The compound eyes are higher than long. The face is rather strongly narrowed above the epistoma; at the narrowest point it is scarcely over half as wide as at the upper portion of the face. Toorax: Predominantly brown to black in ground color covered with brown­ish gray pollen on the mesonotum, gray on the sides. The humeri are yellow, with a faint tinge of brown. The scutellum is brown to black on the disc, yellow on the margin and on the ventor. The pleura are brown to black, tinged with yellow; the margins of the sclerites are often yellow. Only one humeral bristle is present. The anterior sternopleural bristle is about two-thirds as long as the posterior bristle. The anterior dorsocentral bristles are about three-fifths as long as the posteriors and are situated almost opposite the second pair of supraalars. Legs: Yellow except for a tinge of brown on the apices of the tarsi. The legs are not ornate. The front tarsus has short, inconspicuous hairs along the dorsal sur­face (Fig. 4c). Wings: Faintly infuscated, with very pale brown marking at the apices of veins R2+3, R4+5 and M1+2, and over them crossvein. The third The University of Texas Publication FIG. 4. Drosophila ceratostoma n. sp. a. head; c. front tarsus of male; d. wing; e. female genitalia; f. male genitalia, lateral. D. asketostoma Hardy, b. mouthparts of male. Hardy: New S-pecies of Hawaiian Drosophilidae costal section is about five times longer than the fourth and the costal fringe extends about one-third the distance between the apices of veins R2+3 and R4+5 (Fig. 4d). Abdomen: Dark brown to black on the dorsum, largely yellow on the ventor. The terga are thickly covered with short black setae. The cerci are dark brown to black, semi-circular in shape. The ninth tergum is almost parallel sided, blunt on the ventral margins (Fig. 4f). Length: Body and Wings, 2.7-2.85 mm. FEMALE. Fitting the description of the male except for sexual differences. The ovipositor blades are short and blunt, shaped as in Figure 4e. Length: Body, 3.2mm.; Wings, 3.6 mm. Holotype male, allotype female and three paratypes, two males, one female, reared from bracket fungus at Kipuka Ki, Kilauea, Hawaii National Park, Hawaii, Sept. 12, 1964 (H. T. Spieth) . Type and allotype in the B. P. Bishop Museum. Paratypes in the collection of the U.S. National Museum and the University of Hawaii. Drosophila conjectura Hardy, aberration or closely related species? (Fig. 5a) Drosophila conjectura Hardy, 1965, Insects of Hawaii 12:223. Seven specimens are on hand from Bird Park, Kilauea, Hawaii, December 5, 1963, (M. R. Wheeler) which appear to be typical conjectura. This is a new island record. Another series of sixteen specimens from the forest above Paauilo, Hawaii, 3000', August 29, 1963, (L. H. Throckmorton, D. Gubler, and D. E. Hardy) seem to fit conjectura in all respects except that the pleura are brown rather than all yellow and the mesonotum and scutellum are darker brawn to black. I see no structural differences in these. The mouthparts (Fig. 5a), orna­mentation of the front tarsi, and other details appear to be identical. It will prob­ably be necessary to do crossing experiments and to have more information con­cerning the biology and habits before the true position of this population can be decided. Drosophila demipolita Hardy, aberration? Drosophila demipolita Hardy, 1965, Insects of Hawaii 12:239. A series of six specimens, one male, five females, reared from gill-type fungus at Kipuka Ki, Hawaii, Sept. 9-11, 1964 (H. T. Spieth) appears to fit demipolita except for the following differences found in the females: the apices of the wings and the m crossvein are distinctly tinged with brown, the sternopleura are all black, and the genae are comparatively broad, equal in width or slightly broader than the palpi. The male specimen which was bred with the series of females seems to fit demipolita but the wings are evenly tinged with brown. On the basis of the wing markings and the brown to black third antenna! segment the females would fit near haleakalae Grimshaw, but differ by having two strong bristles present on each humerus (these are approximately equal in size) rather than having the secondary bristle tiny, and poorly developed as in haleakalae. They The University of Texas Publication ]l Fw. 5. Drosophila coniectura Hardy. a. mouthparts of male. D. dolichotarsis n. sp. b. head, front view; c. head, lateral; d. wing; e. front leg, lateral; f. male genitalia, lateral; g. male genitalia, ventral. also differ by having the costal fringe extended almost one-half the distance be­tween the apices of veins R2+3 and R4+5, rather than scarcely over one-fourth this distance. The comparative lengths of the third and fourth costal sections and the last two sections of vein M 1+2 also differ in these two species. The specimens on hand may represent a new species but this cannot be decided until a larger series of specimens is studied. Drosophila dolichotarsis new species (Figs. 5b-g) This remarkable species would appear to fit near the haleakalae-polita complex of species because of the thickened black rim around the apex of each labellum and the yellow-white apex of the scutellum. It is strikingly different, however, from any Drosophila which has been described to date. On the basis of the associ­ations we have seen of the species which have the thickened rim on the labellum, Hardy: New S-pecies of Hawaiian Drosophilidae it is probable that this may be a breeder in bracket-type or gill-type fungi. D. dolichotarsis fits near D. venusta Hardy, from Molokai, and is apparently related to that species. It is readily differentiated by the extremely long slender wings and differences in wing markings (Fig. 5d); by all yellow pleura, humeri and notopleural areas of the mesonotum; by the more elongate front basitarsus, equal in length to the tibia, as well as other details as brought out in the description below. D. dolichotarsis also shows close relationship to stenoptera Hardy, from Paliku, Haleakala, Maui. It is differentiated by the all yellow pleura, legs, and front, by the more slender front basitarsus (Fig. 5e) and wings (Fig. 5d) as well as differences in wing markings and other details. MALE. Head: The eyes are oval, .54 higher than long and not noticeably narrowed ventrally (Fig. 5c). The genae are broad, measured from the vibrissal row to the eye margin; each is approximately equal in width to nine rows of eye facets. The occiput is also broad; at its widest point it is approximately one-half as wide as the compound eye. The head is predominantly yellow. The upper median portion of the occiput is polished black in ground color, covered with gray pubescence or microscopic scales. The ocellar triangle is rather narrowly marked in black as shown in Figure 5b. The sides and lower portions of the occi­put as well as the remainder of the head, including the mouthparts, are yellow except for a faint tinge of brown in the ground color of the front, the clypeus, and on the lower margin of each gena. The front is about two times broader than the eyes as seen in direct dorsal view and is densely golden pubescent. The face is just slightly raised down the median portion. Each palpus has a moder­ately long apical bristle and numerous black setae around the posterior margin (Fig. 5c). Each la bell um has a prominent black rim around the margin. Two prominent black bristles are present at the upper edge of each vibrissal row; these are approximately equal in size to the genal bristles. The anterior reclinate bristles are rather small, one-half to three-fifths as long as the proclinates and situated well above the latter, approximately one-half the distance between the proclinates and upper reclinates (Fig. 5c). The antennae are entirely yellow. Each arista has six dorsal and two ventral rays in addition to the apical fork. Thorax: Highly polished black over the mesonotum except for the yellow noto­pleural areas and a yellow margin just above each humerus. The scutellum is black in ground color with a prominent yellow-white apex and is rather densely covered with gray pollen. The humeri are yellow with a faint tinge of brown near the anterior margins. The pleura are pale yellow. The metanotum is polished black covered with gray pollen. Only one humeral bristle is present. Also, only one bristle is present on each sternopleuron; I see no evidence of the second sternopleural bristle. Legs: Lacking ornamentation but with the tarsi very long and slender. The front basitarsus is subequal to the tibia; the tibia is .1 longer than the basitarsus (Fig. 5e). The middle and hind tibiae are approximately .35 longer than the basitarsi of those segments. Wings: Approximately 4.3 times longer than wide, almost straight-sided. A dark brown spot extends over the apex of the wing and a broad brown spot extends over the m crossvein into the basal portion of cell R5 and extends as a slightly less intense marking along the pos­terior margin of the wing to wing base. The r-m crossvein has a faint tinge of The University of Texas Publication brown. The entire wing is lightly tinged with brown. The basal cells are yellow. The third costal section is about seven times longer than the fourth and the costal fringe extends about half the distance between the apices of veins R2+3 and R4+5 . The last section of vein M1+2 is .25 longer than the penultimate section. The last section of vein M3+4 is about equal in length to the m crossvein. Abdomen: Predominantly polished black, gray pollinose over the first tergum, over the dorsal median portion of the second tergum and over the narrow basal portions of the remaining terga. The extreme lateral margins of the terga are light gray pollinose. The genitalia are dark brown to black. The cerci are subacutely pointed ventrally. The ninth tergum is almost straight sided, just slightly nar­rowed over the dorsal portion. The ventral margins of the tergum have short setae around the apices (Fig. 5f) . The claspers are not visible from lateral view. These are very short, inconspicuous, and hidden beneath the lobes of the tergum. As seen from ventral view, the ventral margins of the cerci are densely setose. The claspers are two-three times longer than wide and each has a comb of black teeth along the hind margin (Fig. 5g). The aedeagus is flattened at the apex. Length: Body, 6.0-6.2 mm.; Wings, 7.2 mm. Holotype male and two male paratypes from Waikamoi, Maui, July 11-15, 1964 (H. L. Carson). Type in the B. P. Bishop Museum, paratypes in the collections of the U.S National Museum and the University of Hawaii. Drosophila gubleri new species (Figs. 6a-d) A complex of small brownish yellow species which look superficially alike appear to live in similar or closely related habitats in the Koolau mountains of Oahu. The species at hand would run to couplet 116 in my key to the Drosophila and because of the brown marking in the anteroapical portion of the wing would run to joycei Hardy. These two are not related and the differences in wing mark­ings, body coloration, development of the front tarsi and other characteristics will readily separate them. D. gubleri seems to fit closer to some of the species which lack wing markings. Except for the wings it would resemble fastigata Hardy but the anterior reclinate bristles are not strongly developed and the orna­mentation of the front basitarsus is very different (Fig. 6d; cf. with Fig. 93a, Hardy. 1965: 272). On the basis of the front tarsus it is similar to unicula Hardy, from Hawaii, but the basitarsus has two setae at the apex, not one, and the body coloring is very different. MALE. Head: Yellow except for the reddish eyes and for the brown upper half of the occiput, the vertex, and upper portion of the front. The ocellar triangle is black. The median portion of the front is yellow to the ocellar triangle and the upper lateral portions have a golden brown to bronze sheen. A faint line of brown extends down each vibrissal row. The mouthparts, including the clypeus and the palpi, are yellow, faintly tinged with brown. The antennae are yellow with a slight tinge of brown on the third segments and over the dorsal surface of the second. The arista has five-seven dorsal rays and two ventral rays in addition to the apical fork. The palpi have several black setae around the apices but lack Hardy: New Species of Hawaiian Drosophilidae FIG. 6. Drosophila gubleri n. sp. a. wing; b. apex of front basitarsus of male, anterior view; c. male genitalia, lateral; d. front tarsus of male, lateral. apical bristles. The anterior reclinate bristles are slightly smaller than the pro­clinates and are situated just above the latter. The uppermost bristle of each vibrissal row is almost equal in size to the proclinate bristles of the front. The face is slightly raised down the median portion. Thorax: Predominantly yellow, tinged with brown over the posterior portion of the mesonotum, and with the scutellum brown, faintly tinged with yellow on the disc, pale yellow below. Both pairs of humeral and sternopleural bristles are well-developed. The anterior dorsocentral bristles are approximately two-thirds as long as the posteriors and situated about half-way between the first and second supraalars. The pleura are pale yellow with a slight tinge of brown along the upper portions. The halteres are yellow. Legs: Yellow, tinged with brown at the apices of the tarsi. The front legs are not ornate except for the development of the basitarsi. The basitarsus is rather slender, about two-thirds as long as the tibia, slightly enlarged on the upper apical portion and bearing two anteriorally directed black setae on this prominence (Fig. 6b). A few short erect anterior hairs are present on the apical half of the basitarsus (Fig. 6d). Wing: Subhyaline, with a pale brown marking extending over the anteroapical portion of the wing. This mark extends through the apices of cells Rt and R3 and through the upper portion of cell R5 (Fig. 6a) . The third section of the costa is 4.6 times longer than the fourth and the costal fringe extends about half the distance between the apices of veins R2+3 and R4+5. The last section of vein M1+2 is about .5 longer than the penultimate section. The last section of M3+4 is slightly longer than the m crossvein. Abdo­men: The first and fifth terga are yellow except for a faint tinge of brown on the apex of the fifth. The sides and anterolateral portions of the other terga are yellow, the median and posterior portions of terga two-four are brown. The sixth tergum is short, scarcely visible from dorsal view, and pale yellow in color. The genitalia are yellow. The anal plates ( cerci) are subacutely pointed ventrally, the ninth tergum is strongly narrowed over the dorsal portion, and the claspers are plainly visible from a lateral view as shown in Figure 6c. Length: Body and Wings, Z.3 mm. FEMALE. The female has not been definitely associated with the male. A complex of species occur in this area in which the females are all apparently very similar and an allotype has not been designated. One specimen on hand which seems most nearly to fit with the male differs by lacking the brown marking in the wings and having the front more evenly discolored with brawn, also the abdomen is all brown. The ovipositor plates are short, subacutely pointed. Holotype Male, Pupukea Trail, Oahu; July 17, 1963, (D. Gubler). Four male paratypes collected in the Pupukea area, August, 1963, (D. Gubler, L. H. Throck­morton and D. E. Hardy). Type in the B. P. Bishop Museum; paratypes in the collections of the U.S. National Museum, British Museum (Natural History), and the University of Hawaii. It is a pleasure to name this species after Mr. Duane Gubler who has played an important part in the field studies of Hawaiian Drosophila. Drosophila hirtitarsus Hardy, aberrations or possibly new species (Figs. 7a-b) Two populations from the island of Hawaii would seem to be hirtitarsus except that the mouthparts appear to differ. The specimens from Maui and Molokai, typical hirtitarsus, have the labella distinctly narrowed, with the fleshy portion scarcely visible (Fig. 7a) . The specimens from the island of Hawaii have the labella normal in development with the fleshy portion readily visible (Fig. 7b). These differences are easily seen in situ. I see no other differences which will separate these populations. These may represent sibling species or this may be just an aberration of the mouthparts. It will be necessary to have information concerning the biology and habits of these, and also do cross-mating experiments before a firm decision can be made. The series of about forty specimens from the Frn. 7. Drosophila hirtitarsus Hardy. a. labellum of specimen from Maui; h. labellum of specimen from Hawaii. D. kokeensis n. sp. c. head; d. wing; e. front tarsus of male, posterolateral view; f. front tarsus of male, dorsal; g. male genitalia, lateral. The University of Texas Publication forest above Honokaa, Hawaii, 2000', July 27, 1963 (D. Gubler and D. E. Hardy) differs from a series of sixty-seventy specimens from Kipuka Ki, Hawaii, some reared from gill-type fungi, others collected on gill-fungi, Sept. 9-12, 1964, (H. T. Spieth) by having the wings entirely clear, or evenly but faintly infuscated; the series from Kipuka Ki have the m crossvein faintly marked with brown. Drosophila kokeensis new species (Figs. 7c-g) This species very closely resembles D. atroscutellata n. sp. but the front tarsi are strikingly different in the two (Figs. 3d and 7e). These are apparently not re­lated. In my key to the Hawaiian Drosophila this would run to couplet 112, but does not fit near any of the included species. The predominantly yellow color, with the contrasting dark brown scutellum, the development of the front basitar­sus (Fig. 7e) and the wing markings (Fig. 7d) will differentiate this from any known Drosophila. MALE. Head: The genae and lower portion of the occiput are broad compared to most Drosophila and the eyes are rather strongly narrowed on the lower portion (Fig. 7c) . The head is predominantly yellow except for the reddish brown compound eyes, the black ocellar triangle and the brown tinged upper occiput, upper eye orbits, and lower lateral portions of the face. The antennae are pale yellow, the arista has six to eight dorsal rays and three ventral rays in addition to the apical fork. The anterior reclinate bristle is approximately equal to the proclinate and is situated opposite the latter. Two moderately strong bristles are present in each vibrissal row. The uppermost bristle is about equal in size to the proclinates. The mouthparts are yellow, tinged faintly with brown. The palpi lack conspicuous bristles but have black setae around the apices. Thorax: Yellow except for the brown scutellum, and for a brown tinge along the extreme anterior margin extending to the sides of the mesonotum above the humeri. Two prominent humeral bristles are present and both sternopleurals are well devel­oped. The anterior dorsocentral bristle is two-thirds as long as the posterior and is situated in line with the second pair of supraalars. Legs: Entirely pale yellow. The front tibiae lack ornamentation. The preapical dorsal bristle is small, rather inconspicuous. Only four segments are visible in the front tarsus. The basitarsus is about one-half as long as the tibia, is produced into a short appendage on the apex, which bears two long apical cilia equal in length to the entire basitarsus. A closely set row of dorsal cilia extends over the apical half of the basitarsus. Wings: Hyaline except for a large brown spot over the anteroapical portion (Fig. 7d) ; this extends through the apex of cell R5 anteriorly through the apical por­tion of cell Rt. The third costal section is three times longer than the fourth and the costal fringe extends approximately one-third the distance between the apices of veins R2+3 and R4+5. The last section of vein M1+2 is about .5 times longer than the penultimate section. Abdomen: The first tergum is yellow; the other terga are yellow to rufous at their bases, brown to black, tinged with yellow apically. The genitalia are yellow. The ninth tergum is slender, shaped as in Figure 7g. The anal plates are approximately as high as long. The claspers are plainly visible from a lateral view. Hardy: New Species of Hawaiian Drosophilidae Length: Body, 2.3 mm.; Wings, 2.7 mm. FEMALE. Unknown; Holotype male and four paratypes from Kokee, Kauai, 3600'. Type collected July, 1963 (L. H. Throckmorton); Paratypes collected July, 1963 and November 8, 1963, (L. H. Throckmorton and M. R. Wheeler). Type in the B. P. Bishop Museum; Paratypes in the collections of the U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. Drosophila scitula new species (Figs. 8a-d) This species runs near melanosoma Grimshaw in couplet 72 of my key to the Drosophila because of the shortened front basitarsi and the two are obviously closely related. D. scitula differs from melanosoma by having the apical two­fifths of the wing covered with a dark brown mark, by having abdominal terga one, five, and six yellow rather than black; by having the pleura nearly all yellow, rather than the upper half dark brown to black; and by being smaller and more slender bodied. MALE. Head: The lower half of the front, the face, genae, mouthparts, lower one-third of the occiput, and the antennae are pale yellow. The upper front, vertex and occiput are brown. The ocellar triangle is dark brown to black. The lower half of each compound eye is densely and conspicuously yellow-white pubescent. The compound eyes are oblong to oval in shape, about one-third higher than long. The genae are rather narrow through the median portion; measured from the vibrissal row to the eye margin each gena would be equal in width to about three rows of eye facets. The anterior portion of the gena is produced into a subacute point bearing two rather prominent vibrissae. The other vibrissae are short, inconspicuous and pale. Each palpus has a long, slender, apical bristle equal in length to the proclinate bristles of the front. Each labellum has a promi­nent black rim (Fig. 8a). The anterior reclinate bristles are about three-fifths as long as the proclinates and situated well above the latter, approximately one­third the distance to the upper reclinate bristles (Fig. 8a). Each arista has seven dorsal and three ventral rays in addition to the apical fork. Thorax: The meso­notum is brown to black covered with gray-brown pollen. The scutellum is pre­dominantly brown and the apex is yellow-white. The pleura are pale yellow except for a brown mark at the upper posterior corner of each mesopleuron and in the upper portion of each hypopleuron. The metanotum is yellow, tinged with brown. The humeri are yellow except for a brawn tinge on the upper portions. The anterior dorsocentral bristles are two-thirds to three-fourths as long as the posteriors and are situated about opposite the first pair of supraalars. Two rather well developed humeral bristles are present but only one sternopleural bristle is present. The anterior sternopleural bristle is lacking or represented only by a minute seta. The posterior bristle is well-developed, equal in size to the posterior dorsocentral bristles. The halteres are pale yellow. Legs: Yellow except for a tinge of brown on the apices of the middle and hind femora and on the apical segments of the tarsi. The legs are not ornate and lack conspicuous ciliation. The front basitarsus is short and the second tarsal segment is one-half longer than the first The University of Texas Publication FIG. 8. Drosophila scitula n. sp. a. head; b. female genitalia, lateral; c. front tarsus of male; d. male genitalia, lateral. (Fig. 8c). The preapical dorsal bristle on the front tibia is well developed. Wing: The apical two-fifths is dark brown; this marking extends almost to a level with the m crossvein. The third costal section is 3.65 times longer than the fourth and the costal fringe extends slightly beyond the middle of the distance between the apices of veins R2+3 and R4+5. The last section of vein M1+2 is .4 times longer than the penultimate section. The basal two-thirds of the wing is subhyaline. Abdomen: The first, fifth and sixth terga are yellow. The second tergum is yellow to rufous at the base, brown over the apical one-half to three-fourths. Terga three and four are dark brown. The genitalia are brown, tinged with yellow and shaped as in Figure 8d. The ninth tergum is rather broad over the dorsal portion. narrowed slightly on the sides, and with a prominent preapical bristle on each lateral lobe and a few short setae around the apex. Hardy: New Species of Hawaiian Drosophilid(le Length: Body, 2.5 mm.; Wings, 2.7 mm. FEMALE. The female has not been definitely associated with the male; only one specimen is on hand which appears to fit here. It differs by having the wings clear, however, lacking the brown markings; also the upper half of each pleuron is dark brown to black and the palpi, clypeus, epistoma and third antennal seg­ments are dark brown to black. Each palpus has a prominent apical bristle. The basitarsi of the front legs are short as in the male and the preapical dorsal bristle of the tibia is strong, about equal in length to the basitarsus. The brown apices of the middle and hind femora are more pronounced than in the male. The ovipositor and cerci are dark brown to black; the blades of the ovipositor are ·short, tri­angular, reaching scarcely to the apices of the cerci (Fig. 8b). The females differ from those of melanosoma by having the lower half of each pleuron pale yellow; by the brown apices of the middle and hind femora; and by having the lower reclinate bristles situated distinctly above the proclinates. The female is not being designated as an allotype. Holotype male, Mohihi Stream, Kokee, Kauai; 3700', July, 1963, (L. H. Throckmorton). Eleven male paratypes all from the Kokee area of Kauai, some with same data as type, June and July, 1963 and 1964; (M. R. Wheeler and D. E. Hardy); three from Kokee, 3600', June 22, 1964, and Nov. 8, 1963; two from Halemanu Valley, 4,000', June 25, 1964, (L. H. Throckmorton); and one taken near W aiakoali Stream, 3 700', June 26, 1964 (D. E. Hardy). The female specimen reported above was taken in Alakai Swamp, Kauai, 4000', July 28, 1963 (D. E. Hardy). Type and some of the paratypes in the B. P. Bishop Museum. The remainder of the paratypes are in the collections of the following: U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. Drosophila spectabilis Hardy (Figs. 9a-g) Drosophila spectabilis Hardy, 1965, Insects of Hawaii, Vol. 12: 470. Description of male and new island records. This species was described from a unique female from Puu Kolekole, Molokai. Ten additional specimens are now on hand, five males and five females; five same locality as type, June 10 and July 23, 1964 (D. E. Hardy and M. S. Carson); also five specimens from Waikamoi, Maui, August 14 and Oct. 1, 1964 (H. T. Spieth). The latter is a new island record. Typically the front is darker in color than on the type; on most specimens it is predominantly dark brown to black with a faint tinge of rufous in the ground color of the median portion. The lower one-third of the front is dark brown to black. The face is often discolored with brown to black especially through the median portion. In some specimens a pair of submedian, brownish yellow vittae extend down the mesonotum from the anterior margin to the anterior dorsocentral bristles. It was not mentioned in the original, but the vibrissae are arranged in two distinct rows (Fig. 9a). The male differs from the females by having the front basitarsus flattened laterally (Fig. The University of Texas Publication FrG. 9. Drosophila spectabilis Hardy. a. oral margin showing vibrissal row; b. wing; c. front basitarsus of male, dorsal; d. front tibia and basitarsus of male, lateral; e. male genitalia, ventral; f. female genitalia, lateral; g. male genitalia, lateral. 9d) similar to adiastola, densely covered with moderately long downcurved hairs on the posterior surface and with a row of about six curved anterior cilia extend­ing almost the entire length the segment (Fig. 9c). The front basitarsus is three­fifths to two-thirds as long as the tibia and is nearly three times longer than the second tarsal segment. The front tibia is not ornate except for a rather prominent preapical ventral bristle (Fig. 9d). The wing markings are similar to those of the female except that the hyaline spots are smaller and the anterior margin of the wing is completely brown (Fig. 9b). The lateral aspects of the male genitalia are as in Figure 9g and the ventral aspects are as in Figure 9e. The size of the male is the same as that recorded for the female. The ovipositor plates are shown in Figure 9f. Hardy: New Species of Hawaiian Drosophilidae Drosophila spiethi new species (Figs. 1 Oa-e) This belongs in the species group which is characterized by having a prominent appendage at the apex of the front basitarsus and by having only four tarsal segments developed. It appears to be related to D. fundita Hardy from Maui and Molokai, but differs in many respects. The front basitarsus of spiethi has a prominent, black, dorsal bristle at the base of the segment (Fig. 1 Oe) ; this is ab­sent in fundita; the appendage on the front basitarsus of spiethi is short, thick, and straight-sided (Fig. 1Oe) , rather than slender, pointed, and curved as in fundita; the pleura are black in spiethi, all yellow in fundita; the lower half of the front is yellow in spiethi, the front is all brown in fundita; also the meso­notum of spiethi is black covered with gray pollen, rather than brown, tinged with rufous. This species is also rather closely related to propiofacies Hardy but differs by lacking the long, slender anterior reclinate bristles which are char­acteristic of that species; by the presence of the strong basal bristle on the basitar­sus and by having the lower portion of the front yellow; as well as in other details. MALE. Head: Higher than long, shaped much as in species of Titanochaeta (Fig. 10a) . The front is predominantly yellow, the orbits are brown to black, covered with gray pollen to a level with the proclinate bristles, and the upper portion of the front is brown to black above a level with the lower point of the ocellar triangle, and is distinctly pollinose. The vertex is black in ground color, gray pollinose. The upper two-thirds of the occiput is brown, tinged with rufous in the ground color. The lower portion is yellow. The face, genae, and mouthparts are yellow, including the clypeus. Each palpus has several black setae around the apex but no prominent bristles. The mouthparts are not ornate. The antennae are pale yellow, the aristae each have six dorsal and two ventral rays in addition to the apical fork. The anterior reclinate bristles are approximately equal to the proclinates and are situated opposite the latter. One moderately strong bristle is present in each vibrissal row. This is approximately equal in size to the genal bristles. The face is only slightly raised down the median portion and the sides are slightly convergent. Thorax: Entirely dark colored except for the yellow apex on the scutellum and a tinge of yellow in the ground color of each humerus. The mesonotum is shining black covered with gray pollen and the pleura are dark brown, tinged faintly with rufous in the ground color. The halteres are pale yellow. The anterior dorsocentral bristles are small, poorly developed, scarcely over two times longer than the setae over the mesonotum and approximately one-third to two-fifths as long as the posterior dorsocentrals. Two moderately strong bristles are present on each humerus and both sternopleural bristles are well developed. Legs: Entirely yellow except for the brown apices of the tarsi. The legs are not ornate except for the modifications of the front tarsi. Only four segments are present in the front tarsus. The basitarsus has a prominent append­age at the apex; this is rather broad, not quite equal in length to the second tarsal segment, and is densely covered with black setae along the anterodorsal surface (Fig. 1 Oe). A rather prominent, black, dorsal bristle is present near the base of the basitarsus and several black hairs extend in a line along the anterodorsal surface continuous with the line of hairs over the appendage (Fig. 10e). Wings: Almost The University of Texas Publication Fm. 10. Drosophila spiethi n. sp. a. head ; h. female genitalia, lateral; c. male genitalia, ventral d. male genitalia, lateral; e. front tarsus of male, lateral. hyaline, very faintly tinged with brown. The third costal section is 3.75 times longer than the fourth and the costal fringe extends slightly beyond one-third the distance between the apices of veins R2+3 and R4+5. The last section of vein Mt+2 is .7 times longer than the penultimate section. Abdomen: Brown covered with gray-brown pollen except for a tinge of yellow in the ground color of the basal segment and with segment six largely yellow, tinged with brown. The visible genitalia are yellow. The anal plates ( cerci) are subacutely pointed ven­trally. The ninth tergum is narrow, distinctly lobate ventrally and densely setose on the ventral margin (Fig. 10d). The claspers are hidden from a direct lateral view. The aedeagus is short and blunt; a prominent seta is present on each inside surface of the ninth sternum near the tip of the aedeagus (Fig. 10c). The para­meres are short, just scarely visible from ventral view. Length: Body, 2.3 mm., Wings, 2.5 mm. FEMALE. Fitting the description of the male in most details. The arista has five dorsal and two ventral rays in addition to the apical fork. The uppen;nost bristle of the vibrissal row is stronger than in the male, approximately equal or slightly larger than the proclinate bristles of the front. The third seta of the vibrissal row is also rather well developed. The anterior dorsocentral bristles are strong, almost equal in length to the posteriors. The abdomen is entirely dark brown to black except for a tinge of yellow on the first tergum. The ovipositor blades are rather slender and pointed and are armed with teeth on both the dorsal and ventral surfaces just before the apex (Fig. 1 Ob). Length: Body, 2.5 mm.; Wings, 3.0 mm. Holotype male, allotype female and 41 paratypes, 7 males, 34 females from Bird Park, Kilauea, Hawaii National Park, Hawaii, July 12, 1964 (H. T. Spieth). Also 2 males and 6 females collected same locality as type, July 17, 1964 (L. H . Throckmorton) . Type, allotype and some of the paratypes in the B. P. Bishop Museum. Remainder of the paratypes are being deposited in the collections of the U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. It is a pleasure to dedicate this species to Dr. H. T. Spieth who is making intensive studies of the mating behavior of Hawaiian Drosophilidae. ldiomyia clav.'selae new species (Figs. 11a-f) This remarkable species is distinctly different from any known ldiomyia and I am unable to relate it to any of the species which I have seen to date. The wing markings would somewhat resemble those of I. grimshawi Bryan but the two do not seem to be related. /. clavisetae is differentiated from other species in this genus by the capitate, or clavate, hairs on the posterior portion of the male abdomen and by having the extra crossvein in cell R5 situated about half-way between the r-m and m crossveins (Fig. 11c). MALE. Head: Predominantly yellow; the ocellar triangle and the postero­median portion of the occiput are brown, the vertex is tinged with brown and a brown tinge extends along the orbits in the area occupied by the frontal bristles. The front is largely golden yellow, lightly tinged with brown or rufous. The The University of Texas Publication FIG. 11. Idiomyia clavisetae n. sp. a. head; b. front tibia and tarsus of male; c. wing; d. male genitalia, lateral; e. apex of male abdomen, ventral; f. female genitalia, lateral. Hardy: New Species of Hawaiian Drosophilidae front is comparatively broad, two times wider than the compound eyes as seen from direct dorsal view. The interfrontal area is sparsely covered with short, black setae on the lower half. A line of short setae also extends down each side of the front just outside the frontal bristles. The anterior reclinate bristle is sub­equal to the proclinate and is situated distinctly above the latter, approximately one-third the distance from the proclinate to the upper reclinate. A dense patch of short, black setae occurs on each side of the mid-line of the posterior portion of the occiput. The antennae are entirely yellow; each arista has six dorsal and four ventral rays in addition to the apical fork (Fig. 11a). The face is clear yellow and has a prominent gibbosity in the median portion. The oral vibrissae are moderately developed; the three upper bristles are the strongest and are approxi­mately two-thirds as long as the genal bristle. The mouthparts are entirely yellow except for a tinge of brown at apices of labella. The labella apparently lack teeth but are conspicuously covered with yellow to brown setae. The mentum is well developed and covered with prominent setae. The palpi lack prominent bristles but are covered with black setae on their apices. Thorax: Predominantly yellow, with brown markings on the upper half of each pleuron and on the mesonotum and scutellum. On the type the mesonotum has a broad brown median vitta which fades out slightly at the anterior end of the segment. On the two paratype males present, this vitta is lacking on the anterior third to one-half of the mesonotum. A narrow brown Yitta extends on each side of the mesonotum from near the posterior border to about opposite the anterior supraalars; the area behind each humerus is also discolored with brown. The disc of the scutellum is pale brown, the sides are yellow. The upper two-thirds of each mesopleuron and the portion of each pteropleuron beneath the wing base are marked with brown. The anterior dorsocentral bristles are approximately two-third to three-fourths as long as the posterior pair, and are situated opposite the second supraalar bristles. Two strong bristles are present on each humerus. The sternopleural bristles are strong, the anterior bristle is three-fourths to four-fifths as long as the posterior. A line of fine black setae extends vertically over the median portion of each sternopleuron. The halteres are entirely yellow. Legs: Predominantly yellow, tinged faintly with brown on the apices of the femora and the tibiae. The tarsi are brown. The posteroventral and posterodorsal bristles are well developed on the femora. The front tibiae and tarsi are not ornate, the basitarsus is about three fifths as long as the tibia (Fig. 11b). Wings: Mostly infuscated with dark brown markings over the anterior portion and gray-brown over the posterior portion, leaving rather large hyaline spots in the cells (Fig. 11c). The wing base is yellow-brown to a level with the forking of the radial sector. The apical half of the costal cell is oc­cupied by a brown spot which extends into the basal portion of cell Rt. A large brown spot occupies the median portion of cell Rt and extends through cell R3 to vein R4+5 . A brown spot is also present in the apex of cell Rt extending trans­versely across the wing to cells R3 and R5 and blending with a pale brown mark­ing at the apex of cell 2nd M2. The third costal section is approximately five times longer than the fourth and the costal fringe extends about to the middle of the distance from the apices of vein R2+3 and R4+5. The extra crossvein in cell R5 is more basally placed than in any other known species of ldiomyia; it is sit­ The University of Texas Publication uated halfway between the r-m and them crossveins (Fig. 11c). Abdomen: Pre­dominantly yellow; black on the sides of the first four terga and tinged with brown on the ventral margins of the fifth tergum. The posterior margins of terga two to four are narrowly brown. Except for one row on the postero-lateral mar­gins of the fifth tergum, the peculiar capitate, or clavate, setae are confined to the sixth tergum and sternum (Fig. 11e). The cerci are prominent, semicircular in shape, and with rather conspicuous black setae around the apical margins. From lateral view the genitalia are shaped as in Figure 11 d. The ninth tergum is not noticeably narrowed over the dorsal portion. The claspers are almost hidden from lateral view. Length: Body and Wings, 5.7-6.0 mm. FEMALE. Fitting the description of the male in most details. The third anten­na! segment in brown and the apices of the palpi are often tinged with brown. Each palpus has a small apical bristle. Also the front is rather uniformly brown­ish tinged in ground color. The abdomen is dark brown to black except for the yellow markings on the sides of the first segment. The ovipositor blades are rather short, extending about one-third longer than the cerci, and blunt at apices (Fig. 11f). Length: Body 5. 7 mm., Wings 6.0 mm. Holotype male from Waikamoi, Maui, 4,000'; Oct. 1, 1964 (H. T. Spieth). Allotype female, same locality and collector as type, dated Oct. 1-7, 1964. Seven paratypes, two males, five females, same locality as type, collected Oct. 1-7, 1964; July 8, 1964 (H. L. Carson) and one specimen collected on the Flume Trail at Waikamoi under Clermontia arborescens trees, Aug. 14, 1964 (D. E. Hardy). Type and allotype in the B. P. Bishop Museum. Paratypes in the collections of the U.S. National Museum, British Museum (Natural History), and the Uni­ versity of Hawaii. Idiomyia melanocephala new species (Figs. 12a-e) Because of the strongly arched costa, the wing markings, and the nature of the antennae, this fits near I. perkinsi Grimshaw and the specimens at hand were taken in the same area as specimens of perkinsi. I. melanocephala differs from perkinsi by being predominantly black: the mesonotum all black, the scutellum black except for a narrow yellow apex and the pleura and legs nearly all black. In perkinsi the anterior portion and sides of the mesonotum, the disc of the scutel­ lum and the posterior half of each pleuron are yellow; also the legs are almost completely yellow. I. melanocephala also differs by having the anterior reclinate bristle stronger, more conspicuous than in perkinsi and by having the m cross­ vein gently concave, not sinuate. MALE. Head: Nearly quadrate as seen in direct lateral view; the front is not strongly produced and the face is gently concave as seen in lateral view. The eyes are slightly narrowed ventrally (Fig. 12a). The occiput is rather strongly swollen Hardy: New Species of Hawaiian Drosophilidae FIG. 12. ldiomyia melanocephala n. sp. a. head; b. wing; c. front tibia and tarsus of male; d. female genitalia, lateral; e. male genitalia, lateral. and the median portion is approximately one-half as wide as the eye. The genae are also broad; measured from the eye margin to the vibrissal row each gena is equal to about six rows of eye facets. The head and appendages are predomi­nantly black. The sides of the occiput are yellow, tinged faintly with brown and the front is mostly yellow, black along the orbits to just below the proclinate bristles. The eye orbits, ocellar triangle, vertex, and upper portion of occiput are densely gray-brown pollinose. The anterior reclinate bristle is moderately devel­oped, about one-half to three-fifths as long as the proclinate and situated well above the latter, approximately two-fifths the distance between the proclinate and upper reclinate bristles (Fig. 12a). The oral vibrissae are arranged in two irregular rows and no unusually strong bristles are present in the vibrissal row; the upper three bristles are approximately equal in size to those on the genae. A The University of Texas Publication moderate bristle is present at the apex of each palpus; this is about equal in size to the anterior reclinate bristle. The mouthparts are not ornate. The aristae are rather short, scarcely more than one-fourth longer than the antennae. Each arista has eleven or twelve dorsal rays and five or six ventral rays in addition to the apical fork. The inner surface of the arista is densely covered with short villi. The second antenna! segment has a dense patch of black setae on the dorsal surface. The antennae are rather widely spaced and a prominent carina extends between them; this extends onto the upper median portion of the face. Thorax: The mesonotum and scutellum are dark brown to black in ground color covered with gray-brown pollen. The area above each wing base is yellow; the yellow spot occurs on each anterior margin just inside the humerus and a tinge of yellow occurs in the area behind the humerus and in the notopleural area. The propleura, mesopleura and sternopleura are polished black. The pteropleura are black except for a yellow spot in the middle and beneath the wing base. The metapleura are yellow, tinged faintly with brown. The metanotum is dark brown to black, tinged with rufous in the ground color. The humeri are shining black with a faint tinge of yellow on the upper margins. The scutellum has a small yellow spot at the apex between the apical bristles. The lower humeral bristle is small, scarcely one-third as long as the upper. Two strong sternopleural bristles are present. The anterior dorsocentral bristles are about two-thirds as long as the posteriors and are situated approximately opposite the second pair of supraalars. Legs: Almost all black; yellow at the extreme apices of the femora and tibiae, and at the bases of the tibiae. The tarsi are mostly brown, faintly tinged with yellow. Each front tibia has a series of dorsal hairs at the base but no other conspicuous ciliation. The front basitarsus is approximately two-thirds as long as the tibia and has about four erect, anterodorsal hairs situated just before the apex of the segment. About four of these anterodorsal hairs are also found on the second tarsal seg­ment (Fig. 12c). Wings: The costal margin is strongly arched, so that cell R1 is very broad in the basal portion and narrowed apically. A dark brown spot ex­tends over the apex of the wing. Brown markings are also present on the cross­veins, along vein M1+2 just before the crossvein, at the apices of the first and second costal cells, and near the base of cell Rf just below the break in the costa (Fig. 12b). The extra crossvein in cell R5 is situated just before them crossvein. The m crossvein is gently concave and in some specimens has a small appendix on the outer edge as in Figure 12b. Vein M3+4 is short, scarcely one-fifth as long as the m crossvein. The third costal section is 6.5 times longer than the fourth and the costal fringe extends three-fourths the distance between the apices of veins R2+3 and R4+ 5. Abdomen: Polished black in ground color, covered 'With brown pollen. The genitalia are brown to black; the cerci are tinged with yellow and almost oval in shape. The ninth tergum is just slightly narrowed over the dorsal portion and is lo bate ventrally (Fig. 12e). Only the upper edge of each clasper is visible from a dorsal view; the inner surface is armed with strong teeth. The aedeagus has a dorsal hook before the apex. Length: Body, 6.4 mm.; Wings, 7.0 mm. FEMALE. Fitting the description of the male in most respects. The median Hardy: New Species of Hawaiian Drosophilidae portion of the face, however, is yellow, tinged faintly with brown. The ovipositor blades are rather well developed, extending beyond the apices of the anal plates; they are slightly attenuated apically, and armed with short setae around the margins (Fig. 12d). Length: As in male. Holotype male, allotype female and one male paratype from W aikamoi, Maui, July 10, 1964 (W. B. Heed). Type and allotype in the B. P. Bishop Museum; paratype in the University of Hawaii collection. Idiomyia planitibia new species (Figs. 13a-e) This species is related to/. hemipeza Hardy, from Oahu, but the wing mark­ings, body coloration and other details are quite different. The band on the anterior margin of the wing is less distinct, not so broad as in hemipeza, occupying only cell Rt; the apical brown marking extends into the upper apex of cell 2nd M2 rather than ending at vein Mt+2; the pleura are predominantly polished black, rather than having the lower portion brown and the upper portion yellow; the wings lack the transverse brown band through the middle which extends across the crossveins in hemipeza; and the male genitalia differ by having the an­teriormarginofthe ninth tergumexpanded (Fig. 13b). MALE. Head: The head is distinctly pointed anteriorly, about one-third longer than wide with the face sharply slanting (Fig. 13a). The front is about as wide as ·long and approximately two times wider than the eyes. The front is almost entirely yellow except for a brown to black stripe down each side in the area occupied by the frontal bristles and for-a sharp point extended part way down the middle from the ocellar triangle. The front is covered with golden pubescence. The ocellar triangle and the eye orbits are gray pubescent. The upper portion of the occiput is brown to black in ground color covered with gray pubescence. The lower portion of the occiput, the genae, except for a brown discoloration on the anterior portion, the lower margin of the face, epistoma, the labella and mentum are yellow. The upper four-fifths of the face, the anterior comers of the front, and the antennae are opaque black. The palpi are dark brown to black. Each palpus has a moderately strong apical bristle. The anterior reclinate bristles are small, scarcely over two times longer than the setae found on the lower part of the front and are situated almost half the distance between the proclinate and the upper reclinate bristles. The proclinate and upper reclinate bristles are approximately equal in size. The inner vertical bristles are slightly smaller than the outer and are about equal in size to the ocellar bristles. The compound eyes are almost bare with only scattered microscopic setae present. Thorax: The .mesonotum is predominantly yellow covered with yellow pollen. A large black spot is present on each side just before the suture; this blends with a dark brown to black vitta which extends down each side almost the entire length of the mesonotum. A median brown to black vitta is also present, extending from the posterior margin approximately opposite the presutural bristles. The lower half The University of Texas Publication FIG. 13. ldiomyia planitibia n. sp. a. head; b. male genitalia, lateral; c. wing; d. front tibia and tarsus of male, lateral; e. front tibia of male, dorsal. of each humerus is shining black, the upper portion pale yellow. The mesopleura, sternopleura, hypopleura and lower halves of the pteropleura are polished black; also a black spot is present in the upper portion of each pteropleuron just beneath the wing base. The metapleura are entirely yellow. The scutellum is brown on the disc, yellow on the margins and below. The metanotum is polished black covered with gray pubescence. The halteres are pale yellow. The anterior dorsocentral bristles are two-third to three-fourths as long as the posteriors and are situated slightly in front of a line drawn between the second supraalars. Two humeral bristles are present; the lower is one-half to three-fifths as long as the upper bris­tle. The anterior sternopleural bristle is two-thirds as long as the posterior bristle. Legs: Yellow except for the brown to black middle and hind coxae, the brown to black apices of the middle and hind femora and tibiae, a tinge of brown at the bases of the hind tibiae and at the apices of the front pair. Also the apical seg­ments of the tarsi are brown to black. The legs are slender, the anterior femora lack posteroventral bristles except for two or three situated just before the apex of the segment. The front coxae are unusually long, approximately two-thirds as long as the femora. The front tibia is rather strongly flattened dorsally and the entire dorsal surface is smooth, free of setae except for a tiny poorly developed preapical dorsal seta. Six or eight rather long, curved hairs are situated on the dorsal surface at the extreme base of each tibia and a row of erect posterodorsal and a row of anterodorsal hairs extends down each side of the bare area (Figs. 13d-e) Rather long, erect hairs extend down the anterodorsal and posterodorsal surfaces of the front basitarsus. The front basitarsus is about three-fifths as long as the tibia. Wings: 3.2 times longer than wide, rather intensely tinged with yellow, with a dark brown spot covering the wing apex and a pale brown border extend­ing along the anterior margin. A small brown spot is present near the upper edge of the extra crossvein in cell R5 and another brown spot is present on vein M1+2 just before the extra crossvein. The r-m crossvein is situated near the basal one-third of cell first M2 and the extra crossvein is situated about its own length below them crossvein (Fig. 13c). The last section of vein M3+4 is short, scarcely over half as long as the m crossvein. The basal cells of the wing are in­tense yellow. Abdomen: The first tergum is entirely yellow. The other terga are black down the median portions, along the narrow posterior borders and on the sides, leaving a pair of large lateral spots on each tergum. The genitalia are brown to black, tinged with yellow. The cerci are broad, semicircular in shape, not pro­duced ventrally. The ninth tergum is slightly narrowed over the dorsal portion and is expanded on the anterior median margin. The ventral portion of the tergum is lobate and bears strong bristles at the apex (Fig. 13b). The claspers are only partially visible from lateral view. The row of teeth is borne on the inner margin and is not visible except from ventral view. Length: Body, 7.2 mm.; Wings, 8.3 mm. This species has a wing spread of approximately 18.6 mm. FEMALE. Unknown. Holotype male and four male paratypes from Waikamoi, Maui, July 11-15, 1964 (H. L. Carson, No. C 104. 45, and W. B. Heed). The type is in the B. P. Bishop Museum, the paratypes in the collections of the U.S. National Museum, British Museum (Natural History), and the University of Hawaii. Nudidrosophila aenicta new species (Figs. 14a, c, e, g-h) This species runs near eximia Hardy, from Maui, but differs by having the pleura yellow, not dark brown, and by having the front legs of the male very differently ornamented. Except at the extreme base of the front tibia, the long hairs arranged down the segment are anterior and anteroventral in position rather than being dorsal and anterodorsal in position as in eximia; also the long hairs of the front basitarsus are arranged in two irregular rows on the anterior and antero­dorsal surfaces (Fig. 14e) rather than being in a single row down the dorsal surface as in eximia. The female ovipositor of aenicta is more developed than in eximi.a, nearly twice as long (Fig. 14g). The females of this species show remark­able resemblance to Drosophila hirtitibia Hardy and based upon the females these would appear to be very closely related. In most characteristics, the two appear to be identical and the two species probably occupy the same or similar habitats. Most of the specimens of aenicta on hand were collected in the same habitat with a fairly large series of Drosophila hirtitibia. Besides the overall similarity of characters, the most striking resemblance is in the development of the ovipositors; both have the same type of elongate blades which are equal to or longer than the combined lengths of abdominal segments five and six and which are directed vertically when in resting position (Fig. 14f). I believe it is evident that Nudi­drosophila should not be retained as a genus but probably should be sunk as a direct synonym of Drosophila even though on the basis of the males there appear to be very striking differences and even though the male characters depart radi­cally from the present concept of Drosophila; the characters they exhibit may not prove to be of generic importance in this case. I feel, however, that it is prema­ture to set up this synonymy until this complex can be studied in more detail, and more information can be obtained concerning association with hirtitibia and possibly other Drosophila. Except for the head characters both sexes of aenicta show close resemblance to D. hirtitibia. The ciliation of the front tibiae and tarsi is also similar in the two, even the presence of several long, curved, dorsal hairs at the extreme base of each tibia. In the male of D. hirtitibia the head has the full complement of bristles, the front is not puffed on the sides, the eyes are broader in relation to the occiput (Fig. 14b) and the palpi lack strong bristles. The head of the male Nudidrosophila is as in Figure 14a. The females of D. hirtitibia differ from those of aenicta by having a row of prominent setae extending along each eye orbit from the anterior margin about half the length of the front and by having scattered interfrontal setae over the lower half of the front; these setae are well developed, conspicuous, approximately equal in size to the setae of the occipital row, and the sides of the front are not swollen (Fig. 14d). In aeni~ta the anterior lateral areas of the front are densely covered with minute setae and these areas are slightly puffed (Fig. 14c). Also the anterior reclinate bristles in hirti­tibia are situated much closer to the eye margin than the proclinates and are fairly close to the latter. In aenicta the reclinates are situated almost in line with the proclinates and about one-third the distance to the upper reclinates. The eye is more oval in shape in hirtitibia, only slightly narrowed ventrally; the lower portion of the eye is nearly two times wider than the occiput, rather than being shaped as in Figure 14d and as described below. The wing venation is the same in the two although the costal fringe in hirtitibia is consistently shorter than in aenicta, extending about two-fifths the distance between the apices of veins R2+3 and R4+5, rather than distinctly half or beyond. Also the postocellar bristles are much smaller in hirtitibia, being slightly smaller than the proclinate bristles. In aenicta these bristles are well developed, distinctly stronger and longer than the proclinates and nearly equal in size to the upper reclinate bristles. MALE. Head: Almost quadrate in shape as seen from direct lateral view, with each eye rather strongly narrowed ventrally as in Figure 14a. The front is broad. almost two times wider than each compound eye as seen from direct dorsal view, dark brown to black on the upper two-thirds, yellow below, tinged with brown through the median portion. The anterolateral corners of the front are very densely covered with short, brownish yellow pubescence; these pubescent areas are slightly swollen. Yellow to yellow-brown setae are rather sparsely scat­tered in the median portion of the front and extend posterolaterally above the densely pubescent areas where they become longer, more prominent and about equal in size and development to the setae in the middle of the ocellar triangle. Long yellow laterally projected hairs are developed around the sides of the The University of Texas Publication ocellar triangle. The first two antennal segments are yellow, tinged with brown dorsally. The third segment is brown, tinged faintly with yellow to rufous in the ground color. Each arista has about six dorsal and two ventral rays in addition to the apical fork and has about six short rays along the anterior margin. The upper two-thirds of the occiput is brown, the lower portion yellaw. The genae are yellow except for a line of brown along each vibrissal row. The oral vibrissae are represented by short setae; no bristles are present. The face is yellow, tinged faintly with brown on the lower portion; the median portion is slightly raised. The clypeus is yellow, tinged with brown. The mouthparts including the palpi are yellow. Each palpus has a broad, brown apical spine (Fig. 14a). The labella are not ornate. Thorax: Brown on the dorsum, yellow on the sides; the dorsal surface is lightly gray-brawn pollinose. The lower margins of the humeri are yellow. Two strong humerals and two strong sternopleural bristles are present. The anterior dorsocentral bristles are approximately two-thirds as long as the posteriors and are situated almost in line with the second pair of supraalars. A moderately strong seta (small bristle) is situated on each side in line with the dorsocentrals and almost opposite the suture; this is approximately two times larger than the surrounding setae of the mesonotum. Legs: Entirely pale yellow. Each front tibia has about four long, black, curved, dorsal cilia situated at the base of the segment, also a row of about five or six slender, brown, anterior cilia extending most of the length of the segment and a row of five or six long, brown to black. anteroventral cilia extending along the segment (Fig. 14e). The preapical dorsal bristle is completely lacking on the tibia. The front tibia is 2.4 times longer than the basitarsus. The basitarsus has two prominent, black, anterior cilia, one located near the basal one-fourth and one located near the apical one-third; it also has three anterodorsal brown hairs rather evenly spaced along the segment (Fig. 14e). The second tarsal segment has one strong, black, anterior hair and one brown anterodorsal hair. Wings: Subhyaline, very faintly tinged with brown. The third costal section is 3.3 times longer than the fourth and the costal fringe extends slightly over halfway between the apices of veins R2+3 and R4+5. The last section of vein M1+2 is .7 times longer than the penultimate section. Abdo­men: Predominantly brown, tinged with yellow to rufous in the ground color, especially on the first and on the posterior segments. The genitalia are yellow. The ninth tergum is slightly narrowed over the dorsal portion, expanded ven­trally as in Figure 14h. Only the upper portion of each clasper is clearly visible from a lateral view. The aedeagus has a strong dorsal hook near the base (Fig. 14h). Length: Body, 1.75 mm.; Wings, 2.0 mm. FEMALE. The female differs from the male by having the normal head bristles developed, by having a strong bristle present at the upper portion of each vibrissal row; this is approximately equal in size to the upper reclinate bristles -0f the front; the apical bristle on each palpus is not so broad and distinctly flat­tened as in the male and the front legs are not ciliated. The other characteristics of the female are as pointed out in the introduction above. The ovipositor is as jn Figure 14g. Length: Body, 2.2. mm.; Wings, 2.5 mm. Holotype, male and allotype, female. Kawainui, Oahu; June 20, 1964 (M.R. Wheeler). Twenty paratypes: 3 males, 10 females, same data as type; 3 males, 4 females from Drum Drive, Oahu, June 20, 1964 (M. R. Wheeler) . Type and allotype in the B. P. Bishop Museum. Paratypes in the U.S. National Museum and the University of Hawaii. Scaptomyza (Bunostoma) palmae Hardy, aberration? or new species Specimens of dull black Bunostoma from the island of Oahu appear to agree in all respects with specimens of palmae Hardy, from Hawaii, except that the secondary rows of acrostichal setae are poorly developed. These are represented by one to four setae (usually two) situated on the posterior half of the mesonotum behind the suture. Typical palmae has the secondary rows of acrostichals well developed (four distinct rows present) and represented by 8-10 setae which ex­tend anteriorly to a level about opposite the hind margins of the humeri. The male genitalia of these appear to be identical. The Bunostoma are readily established in laboratory cultures, using the modi­fied, high protein Drosophila medium and genetic studies have been made on the Hawaiian species by Drs. F. Clayton and H. Stalker. The two populations in question apparently behave as species in cultures although cross mating experi­ments have not been carried out. The cultures of palmae aberration? from Oahu have been designated by Dr. Stalker as "Dull type no. I" and I had previously identified this as "n. sp.? rel. to anomala." Subsequent studies indicate that mor­phologically this seems to be the same as palmae except for the more sparsely developed acrostichal setae. This alone seems too trivial to be used as a specific character. It is highly probable that this does represent a sibling species. I feel that this supposition should be supported by more detailed biosystematic data, however, before it is described as new. Dr. F. Clayton has shown (this Bulletin) that the chromosome configurations are different in the population from Oahu than in the other species of Bunostoma. Scaptomyza (Exalloscaptomyza) Hardy Scaptomyza (Exalloscaptomyza) Hardy, 1965, Insects of Hawaii, Vol. 12: 604. Drosophila mauiensis Grimshaw was erected as the type of Scaptomyza (Exal­loscaptomyza) and at the time the manuscript for this book went to the printer, this was the only known species in the subgenus. When the field work for the project on.evolution and genetics of the Hawaiian Drosophilidae began the sum­mer of 1963, it was discovered quite early by Drs. H. L. Carson, L. H. Throck­morton, M. R. Wheeler, and others, that the Exalloscaptomyza are intimately associated with morning-glory flowers and that this actually represents a complex of species; the indications were that each major island in Hawaii had its distinc­tive species of Exalloscaptomyza. The subsequent comparative studies of the male genitalia of populations from the different islands have confirmed this. Two species apparently occur on the island of Hawaii and one each are found on The University of Texas Publication Kauai, Oahu, Molokai, a:id Maui. To date the island of Lanai has not been sampled for Exalloscaptomyza. The development of the genitalia of both sexes is unusual. The ovipositor plates of the female are completely fused (Figs. 16c and 16£) and the homologies of the male structures are not yet thoroughly understood. I find no evidence of true claspers being developed and these structures may possibly be represented by the lobes which are developed from the inner margins of the ninth tergum. The aedeagus of the male is enlarged, and flattened at the apex (Fig. 16a) in all species except molokaiensis. Posterior parameres are developed on the hind margin of the ninth sternum and anterior parameres arise at the sides of the aedeagus. These structures vary considerably in the different species (Figs. 15a, 16d, and 17a). All members of this subgenus breed in the flowers of morning-glories, in the semi-wet areas of the islands, usually at elevations of 1,000 to 3,000 ft. The species will breed in the laboratory in artificial media but the colonies are difficult to maintain; the females lay comparatively few eggs and to date it has not been possible to build up strong enough colonies to begin doing crossing experiments or genetic studies on these flies. Dr. L. H. Throckmorton has studied the internal morphology of this group (refer to paper by Throckmorton in this Bulletin). Key to Species of Scaptomyza (Exalloscaptomyza) 1. Abdomen of male predominantly yellow, terga 4-6 entirely yellow. Legs yellow, tinged with brown; genitalia of both sexes as in Figures 15d, 15e, 16a and 16c .................................................................................................... 2 Abdomen entirely subshining black. Legs predominantly black .................... 3 2. Ninth tergum of male with a prominent, sharp pointed lobe on the antero­ventral margin; this bears two strong setae at the apex (Fig. 15d). North Kona, Hawaii ............................................................................deludens n. sp. Lacking such a lobe on the ninth tergum. Male genitalia as in Figure 16a. Maui ............................................................................ mauiensis (Grimshaw) 3. Sixth sternum and the sides of the eighth tergum of male densely setose. Ninth tergum with a heavily sclerotized lobe developed from each inner margin; this is densely setose at its base (Fig. 16d). Female ovipositor elongate (Fig. 16e), expanded at apex and at base, hour-glass shaped (Fig. 16£). Molokai ........ .............................................................molokaiensis n. sp. Not as above. Inner lobe on each side of tergum weakly sclerotized, not seto0 e (Figs. 15a, 17a, and 17d). Female ovipositor wider than long to scarcely longer than wide (Figs. 15c, 17c, and 17f) .................................... 4 4. Posterior parameres, the lobes from the hind margin of the ninth sternum, sharp pointed and convergent at apices. Anterior parameres well-devel­oped, pointed at apices (Fig. 15a). Female ovipositor slightly longer than wide as seen in ventral view (Fig. 15c). Hawaii ..................caliginosa n. sp. Posterior parameres rounded at tips, usually not convergent. Anterior para­meres poorly developed, represented only by short, rounded lobes at sides of aedeagus (Fig. 17d). Female ovipositor as wide as long, shaped as in Figures 17c and 17f as seen in ventral view ................................................ 5 5. Posterior parameres slender, evenly tapered on the outside margins and Hardy: New Species of Hawaiian Drosophilidae separated by a very narrow space (Fig. 17d). Female ovipositor developed as a distinct, conspicuous lobe as seen in lateral view (Fig. 17 e) and not tapered at apex, as seen from ventral view (Fig. 17f). Kauai ·--·-·-----·-------­-·----------------------------------------------------------------------------------·-------throckmortoni n. sp. Posterior parameres rather broad, straight-sided and with a distinctly broader space between (Fig. 17a). Female ovipositor scarcely protruded as seen from a lateral view (Fig. 17b), and tapered at the apex as seen from a ventral view (Fig. 17 c). Oahu ----·--------------------·--------oahuensis n. sp. Scaptomyza (Exalloscaptomyza) caliginosa new species (Figs. 15a-c) The genitalia of both sexes show close relationship to typical mauiensis but caliginosa differs by having the abdomen and legs black except for the yellow tarsi and except for a tinge of yellow at the bases of the tibiae. The wings are faintly tinged with brown, whereas the wings of mauiensis are hyaline. The dif­ferences in the genitalia of both sexes are as in Figures 15a, 15b and 15c and as discussed under mauiensis. MALE. An almost entirely black species with the body rather densely covered with brownish gray pollen. Fitting the description of the typical species except for the details of body and leg coloration, the faint tinge of brown in the wings, and the rather slight differences in the genitalia. The front is entirely black with a tinge of yellow on the anterolateral margins. The extension of the inner margin of each side of the ninth tergum of the male is developed into a prominent point The University of Texas Publication on the posterior margin and extends as a narrow piece which runs parallel to the edge of the ninth tergum. The posterior parameres, the lobes on the posterior margin of the ninth sterum, are pointed, converging at their apices (Fig. 15a). The anterior parameres are short, rather blunt at their apices. The other aspects of the genitalia are as in Figure 15a. The aedeagus is broad and flat at the apex. The female ovipositor is as in Figures 15b and 15c. The female agrees with the description of the male except for genital characters. Length: Body and Wings, 1.7-2.0 mm. This species is common in the flowers of morning-glory over most of the island of Hawaii at elevations of 1000 to 3000 ft. Holotype male and allotype female, collected near Honokaa, Hamakua Coast, Hawaii, 1,000' elevation, August 29, 1963, in morning-glory flowers (D. E. Hardy). About 250 paratypes from the following localities on Hawaii and col­lected in flowers of morning-glory: a large series same data as type; forest above Paauilo, University of Hawaii Branch Experiment Station, June 19, 1964, (D. E. Hardy); Mud Lane, Hamakua Forest, June 18, 1964, (D. E. Hardy); South Kona, 1,000', June 17, 1964, (D. E. Hardy); and Kawaihae uka, June 26, 1963, (D. E. Hardy); Kamuela, Hawaii, June, 1963, (L. H. Throckmorton). Type, allotype and a series of paratypes in the B. P. Bishop Museum. Paratypes are being deposited in the following collections: U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation, and the University of Hawaii. Scaptomyza (Exalloscaptomyza) deludens new species (Figs. 15d-e) This species is colored exactly like mauiensis and the two are obviously closely related. The series of specimens on hand from north Kona, Hawaii, differ from specimens of mauiensis by having a prominent setose lobe developed on the anteroventral margin of each ninth tergum (Fig. 15d). Fitting the description of mauiensis; the first three abdominal segments are brownish yellow and the posterior segments are entirely yellow. The coxae and femora are brownish yellow. The tibiae and tarsi are yellow. The wings are sub­hyaline, very faintly infuscated (the wings seem slightly more tinged than in mauiensis but I doubt that this is of any significance). The genitalia are as in Figure 15d; the lobe on the anteroventral portion of the ninth tergum is pointed and bears two prominent setae at the apex (Fig. 15d) . Size: as for mauiensis and caliginosa. FEMALE. Fitting the description of mauiensis, with a slight tinge of yellow on the abdomen especially on the posterior portion and with the legs distinctly tinged with yellow: the front and middle tibiae and all of the tarsi are yellow. The ovipositor plate is very similar to that of mauiensis and is shown in Figure 15e. Holotype male, allotype female, and 14 paratypes, 5 males, 9 females from Holualoa, North Kona, Hawaii, April 22, 1944, (N. L. H. Krauss) . E E Q) 0 0 Fm. 16. Scaptomyza (Exalloscaptomyza) mauiensis (Grimshaw). a. male genitalia, ventral; b. female ovipositor plate, lateral; c. female ovipositor plate, ventral. S. (E.) molokaiensis n. sp. d. male genitalia, ventral; e. female ovipositor plate, lateral; f. female ovipositor plate, dorsal. Type and allotype in the B. P. Bishop Museum; paratypes in the collections of the U.S. National Museum, British Museum (Natural History) and the Univer­sity of Hawaii. Scaptomyza (Exalloscaptomyza) mauiensis (Grimshaw) (Figs. 16a-c) Drosophila mauiensis Grimshaw, 1901, Fauna Hawaiiensis 3 (1): 67. Scaptomyza (Exalloscaptomyza) mauiensis (Grimshaw), Hardy, 1965, Insects of Hawaii 12: 604. The typical species is readily recognized by having the posterior portion of the male abdomen pale yellow and the remainder of the abdomen brownish yellow; also by having the coxae of the legs predominantly yellow and the tibiae and femora brownish yellow. The genitalia are distinctive: the structures of both sexes are as in Figures 16a, 16b, and 16c. On the basis of the genital characters this species shows close relationship to caliginosa n. sp., from Hawaii, and is differentiated by the yellow abdomen and yellow markings on the legs of the male; and by having the wings hyaline, not tinged with brown. The inner ex­tensions on the sides of the ninth tergum of the male are rather blunt on their posterior margins compared to those of caliginosa and the anterior parameres also are shorter (compare Figs. 15a and 16a). The female ovipositor plates also show some differences as shown in Figures 15c and 16c. Refer to Hardy ( 1965: 604) for a detailed description of this species. The University of Texas Publication This species appears to be restricted to the semi-wet areas of Maui: the records on hand are from Iao Valley, West Maui; from the slopes of Mt. Haleakala, Makawao Forest Reserve; 3500', and from Keanae, Maui. I have not been able to find it in the morning-glory flowers on the road to Haleakala. I have sampled flowers many times without success, from 1000 to 4000 ft. elevation and this slope may normally be too dry for these flies. The flies may be seasonal in this area, however, since one specimen on hand is labelled ex. morning-glory, 2,000' eleva­tion, Haleakala Road, March 17, 1964, (W. C. Mitchell). Scaptomyza (Exalloscaptomyza) molokaiensis new species (Figs. 16d-f) This species is readily differentiated from other known species of Exalloscap­tomyza by the unusual genitalia of both sexes; refer to Figures 16d and 16£. MALE and FEMALE. Predominantly shining black in ground color, covered with gray-brown pollen on the dorsum of the thorax, gray on the pleura and with gray pollen over the front. The antennae are brown to black, tinged with yellow on the second and on the base of the third segments. The palpi and mouthparts are yellow-brown. Fitting the description of mauiensis except that the abdomen is entirely shining black in ground color and the legs are predominantly black, also the genitalia of both sexes are strikingly different. In situ this species can be readily recognized by the densely setose ventral aspects of the posterior portion of the abdomen of the male, and by the comparatively elongate conspicuous ovipositor of the female (Figs. 16e and 16£). The sixth sternum, the ventral por­tions of the eighth tergum, and the ninth tergum of the male are densely setose. As seen from ventral view, the ninth tergum has a prominent lobe developed on each inner margin. This is rather heavily sclerotized, densely setose at its base, and curved downward at its apex as in Figure 16d. The aedeagus is short, not enlarged or flattened at apex as in other species of this complex. The other aspects of the male genitalia are as in Figure 16d. The female ovipositor is elongate compared to other species and is very differently developed, expanded at the apex and at the base, hour-glass shaped. Length: Male. Body, 1.6 mm.; Wings, 1.8 mm. Female. Body, 1.8-2.0 mm.; Wings, 2.0-2.2 mm. Holotype male, allotype female and ten paratypes, six males, four females from Kaunakakai Ridge, Molokai, 2,000', in morning-glory flowers, July, 1963, (D. E. Hardy and L. H. Throckmorton) . Type and allotype in the B. P. Bishop Museum. Paratypes in the collections of the U.S. National Museum, British Museum (Natural History) and the Uni­versity of Hawaii. Scaptomyza (Exalloscaptomyza) oahuensis new species (Figs. 17a-c) A dark-bodied, clear-winged species which shows close relationship to S. throck­mortoni n. sp., from Kauai, but is differentiated by having the posterior para­meres, the lobes on the hind portion of the ninth sternum, comparatively broad, straight-sided and with a rather wide space separating the parameres as in Figure Hardy: New Species of Hawaiian Drosophilidae FIG. 17. Scaptomyza (Exalloscaptomyza) oahuensis n. sp. a. male genitalia, ventral; b. female ovipositor plate, lateral; c. ovipositor plate, ventral. S. (E.) throckmortoni n. sp. d. male genitalia, ventral; e. ovipositor plate, lateral; £. ovipositor plate, ventral. 17 a. Also, the female ovipositor is just barely protruded as seen in direct lateral view, whereas in throckmortoni it is comparatively prominent (Fig. 17b), and in ventral or dorsal view the ovipositor of oahuensis is distinctly narrowed at the apical portion (compare Figs. 17 c and 17f). Length: Same as for mauiensis. Holotype male and allotype female, Pupukea, Oahu, August, 1963, in morning­glory flowers (L. H. Throckmorton). Approximately 100 paratypes from the following localities on Oahu, taken mostly in morning-glory flowers: same as type; Honouliuli, March 12, 1964, (J. Y. Kim); Pali Highway, 500', July, 1963, (L. H. Throckmorton); Mt. Tantalus, 1800', May 9, 1965 (D. E. Hardy-M. Delfinado); also 3 specimens on hand collected in Kaimuki, March 6, 1910, (Lewis). This species is found in the flowers at elevations from approximately 800' to 2000'. Type, allotype and some paratypes in the B. P. Bishop Museum. Other para­types all being deposited in the collections of the U.S. National Museum, British Museum (Natural History), University of Texas Genetics Foundation and the University of Hawaii. Scaptomyza (Exalloscaptomyza) throckmortoni new species (Figs. 17 d-f) A predominantly dark-colored species appearing to be identical in external characteristics with the other species of this group which have the body and legs predominantly shining black covered with gray pollen. It shows close relationship The University of Texas Publication to oahuensis but is differentiated by having the posterior parameres slender, evently tapered on the outside margins and separated by a very narrow space (Fig. 17d). Also the female ovipositor is distinctly protruded as seen in lateral view and is not tapered at apex as seen in ventral or dorsal view (Fig. 17f). The other aspects of the genitalia are as shown in Figures 17 d and 1 7 e. The inner ex­tensions from the ninth tergum are weakly sclerotized and form a rather broad lobe on each side. The anterior parameres are short, inconspicuous, consisting of small rounded lobes on each side of the aedeagus. The aedeagus is expanded and flat at the apex. Length: same as for mauiensis. This species occurs in morning-glory flowers on the Kokee Road, Kauai, at elevations from 2900 to 3600 ft. To date, it has not been taken at lower eleva­tions. Holotype male, allotype female, Kokee, Kauai, 3600', July 29, 1963, (D. E. Hardy). Thirty-five paratypes, sexes about evenly distributed, same data as type, 2900-3600 ft. (D. E. Hardy and L. H . Throckmorton). Type, allotype and some paratypes in the B. P. Bishop Museum; remainder of paratypes in the collections of the U.S. National Museum, British Museum (Natural History), and the University of Hawaii. It is a pleasure to name this species after Dr. L. H . Throckmorton of the Uni­versity of Chicago who was responsible for first pointing out the morphological differences in the Exalloscaptomyza populations. Scaptomyza (Trogloscaptomyza) platyrhina new species (Figs. 18a-f) Because of the peculiar development of the anal plates ( cerci) of the male this species would somewhat resemble diaphrocerca Hardy, from Molokai, but the genitalia are very differently developed in these and the two species are obviously not related. S. platyrhina will best run in couplet 30 of my key to the Scaptomyza and should be keyed out at this point by having the femora of all legs enlarged, swollen (Fig. 18e) and the anal plates (cerci) developed into long, flattened ven­tral lobes (Fig. 18£). This species is readily differentiated from other known Scaptomyza in Hawaii by the enlarged femora and the distinctive genitalia. MALE. Head: The head is almost quadrate as seen in direct lateral view (Fig. 18a). The front is yellow, tinged lightly with brown except for the brown to black orbits and the black ocellar triangle. The eye orbits, ocellar triangle and the occiput are densely gray pollinose. The ocellar triangle extends forward on the front to a point in line with the anterior reclinate bristles. The anterior reclinates are small, about two-thirds as long as the proclinates and situated distinctly above the iatter. The oral vibrissae are strong, consisting of several prominent bristles. The face, genae, and mouthparts are yellow. The face is tinged with brown in the furrows and has a distinct carina down the middle. The antennae are yellow, tinged with brown. The arista has two dorsal rays in addition to the large apical fork (Fig. 18a). Each palpus has three short bristles at or near the apex and sev­eral short black setae along the outside surface. Thorax: Predominantly yellow, densely yellow pollinose. The mesonotum is discolored with brown on the postero­ Hardy: New Species of Hawaiian Drosophilidae Fm. 18. Scaptomyza (Trogloscaptomyza) platyrhina n. sp. a. head, lateral; b. hind leg of male; c. female genitalia, lateral: d. male genitalia, lateral; e. front leg of male, lateral; f. anal plates of male, end view. median portion, extending anteriorly almost opposite the suture and a slight tinge of brown is present in the ground color on each side in front of the suture. The scutellum is brown on the disc, yellow on the sides. The pleura are yellow, somewhat mottled with brown discolorations on the sternopleura, and with irreg­ular markings of pale brown on the propleura, mesopleura and the upper portions of the pteropleura. The humeri and halteres are yellow. One strong humeral bristle is present. The anterior dorsocentral bristles are situated about opposite the first pair of supraalars. Six strong raws of acrostichal setae are present. The The University of Texas Publication anterior sternopleural bristle is approximately three-fourths as long as the pos­terior bristle. Legs: Entirely yellow except for the brown apical segments of the tarsi. The femora of all legs are rather strongly swollen, three times broader than the tibiae. The front legs are as in Figure 18e; the basitarsus is rather short, scarce­ly one-fourth as long as the tibia. The tibiae of the hind legs are slightly curved (Fig. 18b). Wings: Subhyaline, with the third costal section almost four times longer than the fourth and the costal fringe extending approximately half the distance between the apices of veins R2+ 3 and R4+5. The last section of vein M1 +2 is .6 longer than the penultimate. Abdomen: Predominantly brown to black in the ground color, tinged with yellow on the basal portion and with the first four terga and the median portion of the fifth tergum dull gray pollinose. The lateral margins of the terga are marked with yellow. The sides of the fifth and the entire sixth tergum are polished black. The genitalia are black. The ninth tergum is polished and has a tumescence developed on each side of the dorsum; this is readily visible in situ. Each anal plate is developed into a long, flat lobe on the ventral part as in Figure 18£. The other aspects of the male genitalia are as in Figures 18d and 18f. The claspers are developed into two prominent lobes. Length: Body and Wings, 2.5-2. 7 mm. FEMALE. Similar to the male in most respects; the femora are not strongly swollen however. The ovipositor is entirely membranous, covered with a few short setae as in Figure 18c. Length: As in male. Holotype male and allotype female, Alakai Swamp, Kauai, 4,000', July 28, 1963, (H. L. Carson, No. C72.11). Twenty-seven paratypes, 17 males and 10 females, all same locality as type, some specimens reared from Clermontia, July 22, 1964 (W. H. Heed). Type, allotype, some paratypes in the B. P. Bishop Museum. Other paratypes in the collections of the U.S. National Museum, British Museum (Natural His­tory), and the University of Hawaii. Titanochaeta contestata new species (Figs. 19a-g) This species appears to fit in Titanochaeta although the arista has a ventral ray (Fig. 19a) and the female ovipositor blades are blunt, rounded at apices (Fig. 19d). Because of the ventral ray on the arista, this would run to Drosophila in the generic key; it obviously does not belong here, however, since it is definitely Scaptomyza-like in general facies as well as in genital characters of the male. The arista is also too sparsely plumed for Drosophila, having only two dorsal and one ventral ray in addition to the apical fork (Fig. 19a). This species shows very close resemblance to T . chauliodon Hardy, from Maui, and exhibits all of the Titanochaeta characters except the ones mentioned above. It should be noted however that several specimens of chauliodon and specimens of T. bryani Wirth are on hand which do have a ventral ray on the arista, (it has been noted that the branching of the arista is variable in the group) ; the ovipositor, however, is Hardy: New Species of Hawaiian Drosophilidae Frn. 19. Titanochaeta contestata n. sp. a. head, lateral; b. front tarsus of male; c. ovipositor plates, ventral; d. female genitalia, lateral; e. wing; f. male genitalia, latt>ral; g. male genitalia, ventral. The University of Texas Publication needle-shaped and typical of Titanochaeta. The exact status of Titanochaeta as well as the exact definition of this group cannot be clearly defined until further information is obtained concerning the habits and biologies of the species in­volved. The male genitalia of the presently included species are apparently of several different types and I find no characters in the genitalia which will typify the group. It may prove more logical to treat Titanochaeta as a subgenus of Scaptomyza but on the basis of the distinctive habits (predation on spider eggs) and the structural features which were discussed in my treatment of this group, I prefer to continue treating this as a genus until further information is available. Both sexes are very similar to chauliodon and are best differentiated by the genital characters: the male genitalia as in Fig. 19f, having strong teeth along the inner edge of each clasper but lacking the distal spine and strong apical lobe which are characteristic of chauliodon; the female differing by having the ovipositor blades blunt and rounded rather than sharply pointed, needle-like. It should be noted that one paratype female from Mount Tantalus, Oahu, was recorded under the original description of chauliodon; this should be T. contestata. MALE. Head: Distinctly higher than long, with the eye narrowed ventrally but the occiput not usually expanded. The sides of the face are not visible from direct lateral view and the gena are rather broad (Fig. 19a). Each gena, measured from the vibrissal row to the eye margin, is about equal in width to about four rows of eye facets. The head bristles are strongly developed. The anterior recli­nate bristles are subequal to the proclinates and are situated slightly above the latter. The inner vertical bristles are equal in length to the height of the head. One strong bristle is present at the upper edge of each vibrissal row; this is approx­imately equal in size to the ocellar bristles. The front is about equal in width to about one eye and is predominantly brown, tinged with yellow in the ground color and covered with a golden-brown sheen in the areas between the ocellar triangle and the eye orbits. The ocellar triangle is shining black and a black vitta extends down the median portion of the front to the anterior margin. The anterior margin of the front is dark brown, the orbits are tinged with brown to black. The upper third of the occiput is brown to black on the sides, yellow, tinged lightly with brown in the median portions; the lower part is pale yellow as are the genae, face and mouthparts including the clypeus and the palpi. The face is slightly raised down the median portion. The palpi are large and con­spicuous and lack apical bristles or setae. A series of prominent, black, dorsal setae are present on each palpus and a few short, black setae are also scattered over the median portion of the palpus and a few are located on the ventor at the base (Fig. 19a). The labella are slender, and covered with scattered black setae. The antennae are yellow, tinged with brown on the dorsal surfaces. Each arista has two or three dorsal rays and one ventral ray in addition to the apical fork. Thorax: The dorsum is laq~ely brown; the humeri and lateral margins of the mesonotum, as well as the scutellum, are yellow. A rather indistinct brown-to­black vitta extends longitudinally down the middle portion of the mesonotum; the area on each side of this marking to the dorsocentral row is rufous, tinged lightly with brown. The median portion of the scutellum is brown. The pleura are entirely yellow. The metanotum is yellow-brown. The halteres are pale yellow. The anterior dorsocentral bristles are strong, two-thirds to three-fourths as long as the posteriors and situated almost opposite the first pair of supraalars. Each humerus has two strong bristles plus one moderately developed bristle; the latter is two-thirds to three-fourths as long as the lower humeral bristle. Two strong sternopleural bristles are present and several prominent black setae are located in the area between the two bristles. Legs: Entirely yellow; the segments are densely black setose. The front tibia lacks long ciliation but several of the posteroventral setae on the apical third of the segment are equal in length to the preapical dorsal bristle. The front basitarsus is about one-fourth as long as the tibia. The entire tarsus is very thickly setose. The ventral portion of the basi­tarsus has thickened bristle-like setae (Fig. 19b). The tarsal claws are large and conspicuous, longer than the last tarsal segment; they are yellow to rufous on their bases, black apically. Wings: Subhyaline, very faintly tinged with brown. The third costal section is 3.65 times longer than the fourth and the costal fringe extends about one-third the distance between the apices of veins R2+3 and R4+5. Cell R5 is slightly expanded in the middle, directly above them crossvein. The penultimate section of vein Mt+2 is subequal to the ultimate section. The last section of vein M3+4 is about equal in length to them crossvein (Fig. 19e). Abdomen: Predominantly brown on the dorsum, yellow on the sides, over the entire first tergum, the base and median portion of the second, and the entire margin of the sixth tergum. The genitalia are yellow, tinged faintly with brown. The cerci are pointed ventrally appearing rostrate as seen in direct lateral view. The ninth tergum is broad, not at all narrowed over the dorsal portion and with five or six small black setae near each posteroventral margin. The claspers are narrow, about four or five times longer than wide and strongly toothed along the inner margin, plainly visible from a lateral view (Fig. 19£). As seen from ventral view, each clasper has a clump of setae on the anterior end and six or more moderately long hairs on the inner margin. Each anal plate is developed into a slender, curved lobe visible from ventral view (Fig. 19g). The parameres are well developed and extend beyond the ends of the claspers. Length: Body, 3.5-4.0 mm.; Wings, 3.5-4.2 mm. FEMALE. Fitting the description of the male in most respects. The front basitarsi are not shortened, however, and the segments are not so conspicuously setose. The wing (Fig. 19e) is drawn from the allotype specimen. The mesonotum is more generally brown, the pale areas are much less distinct than in the male. The ovipositor blades are blunt, rounded at the apices, shaped as in Figs. 19c and 19d. The length is the same as for the male. The internal morphology of this species is being discussed in the paper by Dr. Throckmorton. Holotype male, Mt. Tantalus, Oahu; May 19, 1963 (M. R. Wheeler). Allotype female, same locality as type; May, 1953 (D. E. Hardy). Five para­types, 2 males and 3 females, from the following localities: same as type, Nov. 9, 1963, (M. R. Wheeler); Opaeula, Oahu, June 16, 1964, (M. R. Wheeler) ; Drum Drive, Oahu, June 20, 1964 (M. R. Wheeler and H. Carson); and Pupukea Trail, Oahu, June 14, 1964 ( M. R. Wheeler). The University of Texas Publication Type and allotype in the B. P. Bishop Museum; paratypes in the collections of the U.S. National Museum and the University of Hawaii. CHANGES OF NAMES OF HAWAIIAN DROSOPHILA Dr. Marshall Wheeler has pointed out to me that Drosophila fungicola Hardy, 1965, Insects of Hawaii 12: 282, is preoccupied by fungicola Villeneuve, 1921, Ann. Soc. Ent. Belgique 61: 158 and that Drosophila intermedia Hardy op cit.: 328 is preoccupied by Drosophila (Scaptomyza) adusta var. intermedia Duda, 1927, Arch. f. Naturgesch. 91A(11-12) : 125 and 151. I am changing fungicola Hardy to fungiperda, and intermedia to medialis. IX. Courtship Behavior of Endemic Hawaiian Drosophila1 HERMAN T. SPIETH2 TABLE OF CONTENTS Introduction and Technique 246 Summary for Comb Tarsi Explanation of Elements in Male Species Group . 277 Courting Behavior as Charted Bristle Tarsi Species Group-in Table 1 248 Introduction 278 Descriptions of Mating Behavior of D. basimacula Hardy . . . . . . . . . . 278 Individual Species D. fusticula Hardy . 279 Genus Antopocerus Hardy-D. perissopoda Hardy . 279 Introduction .... . ..... . . . 254 D. ( Trichotobregma) A. aduncus Hardy ... ..... . 254 petalopeza Hardy . 280 A. longiseta (Grimshaw) 255 Summary for Bristle Tarsi A. orthopterus Hardy 256 Species Group 281 A. tanythrix Hardy . 257 Modified Mouthpart Species Group- A . villosus Hardy . 257 Introduction 281 Summary for Antopocerus . 257 D. aquila-like . . . . . . . . . . . . . . 282 Genus Drosophila Fallen D. comatifemora Hardy 282 Picture Wing Species Group-D. eurypeza Hardy . . . . . . . . . 283 Introduction 259 D. hirtitarsus-like . 284 D. adiastola Hardy ........ . 259 D. ischnotrix-like 285 D. crucigera Grimshaw 261 D. mimica Hardy . 285 D. engyochracea Hardy 263 D. tendomentum Hardy 286 D. grimshawi Oldenberg . 264 Summary for Modified Mouthpart D. picticornis Grimshaw 267 Species Group . 288 D. pilimana Grimshaw 267 White-tip Scutellum Species D. pilimana-like 268 Group--lntroduction 288 D. punalua Bryan 269 D. fungicola Hardy 288 D. spectabilis Hardy . 269 Scaptomyza-like Species-Idiomyia clavisetae Hardy 270 Introduction 290 Summary for Picture Wing D. crassifemur Grimshaw . 290 Species Group . 271 D. nasalis Grimshaw 291 Spoon Tarsi Species Group--D. parva Grimshaw . 291 Introduction . . ..... . ... . 272 Summary for Scaptomyza-like D. disticha Hardy . 273 Species 292 D. dasycnemia Hardy 272 Unplaced Species-Introduction 292 D. sordidapex Grimshaw . 274 D. achlya Hardy 292 Summary for Spoon Tarsi D. atroscutellata Hardy 293 Species Group 275 D. imparisetae Hardy 294 Comb Tarsi Species Group--D. incompleta Hardy . 296 Introduction 275 Field Observations and Studies 296 D. pectinitarsus Hardy 275 Nature of Mating Behavior . 304 D. proceriseta Hardy . 277 Discussion 309 D. spiethi Hardy . 277 Literature Cited 312 1 This investigation was supported by Public Health Service Research Grants GM-10640 and GM-11609 from the National Institutes of Health, and by Research Grant GB-711 from the National Science Foundation. 2 Visiting Colleague, University of Hawaii, July-December, 1964; Guest Investigator, Genetics Foundation, University of Texas, December, 1964--June, 1965. Present address: Department of Zoology, University of California, Davis, California. The University of Texas Publication INTRODUCTION The upland and mountain rain forests of the six major Hawaiian Islands (Kauai, Oahu, Molokai, Lanai, Maui and Hawaii) support a diversified, ex­tremely rich and unique fauna of Drosophila. The pioneer work of Grimshaw ( 1901 ) first made known its existence, and the subsequent studies of Bryan (1934), Zimmerman (1958) and others, plus the collecting efforts of F. X. Wil­liams and G. B. Mainland, gave further indications of its richness and complexity. Not, however, until Hardy (1965) had systematically collected, described, and delimited the many species was it possible effectively to study their courtship patterns. The present report on courtship behavior is a part of an on-going "group" investigation of the biology, genetics, cytogenetics, systematics, evolution and phylogeny of the Hawaiian drosophilid fauna.3 Over 400 species of endemic Hawaiian drosophilids are no1w known, but since each successive field trip has usually resulted in the capture of hitherto unknown species, it is surmised that between 500 and 600 endemic species live on these islands. The ones now known form species groups or constellations of closely related species for many of which we have only a few specimens, but there are certain species in most, if not all, of the species groups which have been collected in abundance. The present study is based primarily upon these abundant. cur­rently readily collectable species. Thus, despite the fact that the courtships of only 40 species have been involved in the study to date, it has been possible to include several representatives of most of the common species groups. The data here presented are necessarily qualitative in nature. The shortness of the periods of investigation (i.e., the month of July, 1963, and the months of July through November, 1964), coupled with the serious problems that had to be solved in merely keeping the field-collected adults healthy and responsive under laboratory conditions, precluded quantitative studies. Adults of only a limited number of these species can as yet be readily and successfully reared in numbers under laboratory conditions. Intensive study of the sexual biology of the species, plus refinement of current rearing techniques and the development of new ones, will be necessary before quantitative investigations of sexual isolation and its role in the evolution and maintenance of the species can be conducted. The data accumulated by laboratory observations of the 40 species. supple­mented by additional field observations, have made possible, it is believed, a basic understanding of the peculiar and unique pattern of mating behavior dis­played by the Hawaiian drosophilids, and sufficient evidence has also been ac­quired to provide insights into certain aspects of the phylogeny and the complex evolutionary history of this fauna. Indeed, it seems clear that the findings to date add new dimensions to our knowledge and understanding of the sexual biol­ "It is a pleasure to record grateful thanks to other members of the research group for their generous assistance: especially to Elmo Hardy who identified all of the specimens studied, and to Malcolm Brown, Hampton Carson, Frances Clayton, William Heed, Harry Stalker, Wilson Stone, Lynn Throckmorton, and Marshall Wheeler for study specimens, advice, and other as­sistance. I am also indebted to Duane Gubler for assistance in the field and to student assistants Alex Farm, Francis Kamiya, Linda Hiranaga, Kenneth Kaneshiro, 'Janice Nakama, Sharon Pyun, Rae Tanabe, and Fumiko Yamasato at the University of Hawaii as well as to Kathleen Resch and John Murphy at the University of Texas. ogy and its evolutionary history not only of the Hawaiian drosophilids but also of all Drosophila. TECHNIQUE Adult flies, either laboratory-reared or field-collected, were used for observa­tional purposes. The sexes were separated by etherization and maintained in different containers for a period of days before being introduced into an observa­tion cell. During July, 1963, the flies were supplied with a variety of different experi­mental foods as well as with the "regular" formula foods. For the period of July through August, 1964, a special food devised by Wheeler and Clayton ( 1965) was utilized. During both periods it was noted that the flies appeared sluggish and non-responsive. Often prolonged periods of observation regardless of the time of day, size of observation cell, intensity of light, etc., would produce no courtships even when large numbers of individuals (i.e., up to 100) were used. During the last three months of the study (September through November, 1964), a food prepared from agar (15 gms.), karo syrup (50 ml.) and distilled water ( 1000 ml.) was used to maintain the adults. These adults were much more responsive than they had been when given the previous foods, and courtships could usually be observed whenever a number of flies were placed in an appro­priately sized observation cell. Most of the data here reported were acquired during this three month period. A variety of observation cells were experimented with, especially during the early part of the studies when difficulty was being encountered with the flies not engaging in 'Courtship actions. During the last several months, the following four types were used, the particular selection for any observation being determined by the physical size of the flies and the number of individuals available: ( 1) glass shell vial, 25 mm. X 90 mm.; (2) glass shell vial, 35 mm. X 100 mm.; (3) glass cell, 6.3 cm. X 8.8 cm. X 11.4 cm., and ( 4) a cage 23 cm. X 23 cm. X 23 cm. The glass vials were always partially lined with moist blotting paper to maintain a relatively high humidity within the cell and also to provide the flies with a surface to which they could easily adhere. The presence or absence of food within these cells seemed to have no effect upon courting activities and food was therefore usually omitted. The rectangular glass cell, consisting of plate glass walls, was fitted with a 6.3 cm. X 8.8 cm. piece of cellulose sponge which, when moistened, snugly filled and formed one end wall. By moving this sponge in or out, the functional size of the cell could be modified at the observer's will. The moisture in the sponge main­tained a high level of humidity within the cell, and the sponge provided a vertical surface upon which the flies could rest. The cage has metal bottom and back sides, a top and two lateral sides of insect­proof screen, and a glass front. Food and wet sand or water-soaked cellulose sponges, or both, were always placed in the cage before the flies were introduced through a closeable opening in the metal back. When using the cage, observations were conducted with the aid of a Beebe loupe or without magnification. When using the other types of containers, ob­servations were made under a binocular dissecting microscope equipped with a The University of Texas Publication zoom lens that enabled the observer to change magnification from 7X to 30X with rapidity and without visual interruption. It was early determined that high light intensity was detrimental to the court­ing behavior of the flies. Observations were therefore conducted under "normal" room lighting, and usually an opaque shield was set up to prevent indirect window light from directly falling upon the cell. The size (volume) of the observation cell was critical for some of the larger species, e.g., all of the Antopocerus spp., D. crucigera and grimshawi. These species would only infrequently and usually fleetingly court in either of the two shell vials, but were much more apt to engage in prolonged sexual activities in the large glass cell or a comparable sized container, or in the cage. Over and above size of observation cell and intensity of light, there are unknowns that operate and sometimes result in a given group of individuals merely sitting quietly in a cell while at the same time another group, whose history is identical with that of the first group, will engage actively and persistently in courtships. EXPLANATION OF ELEMENTS IN MALE COURTING BEHAVIOR AS CHARTED IN TABLE I A plus ( +) sign indicates the presence and ... indicate the absence of the particular action listed for a column. An asterisk (*) following the name of a species indicates that there exists a good possibility that the entire male repertoire of display has not been observed to date. An asterisk ( *) accompanying a postural element symbol signifies that the uniqueness of the action necessitates reading the description given in the text. PRIME POSTURING POSITION. The numbers 1, 2, or 3 indicate the order of sequence during a courtship in which the following positions are assumed: Front, Side, or Rear. Position assumed by the male with respect to the body of the female after he has tapped and as he commences the courting display. The male, facing the female, stands at such a distance from her that there is no physical contact or overlapping of the bodies of the two insects. Head-under-Wing. Position the male takes directly behind the female and so close to the tip of her abdomen that his head and a portion of his thorax are under her wing tips. ANTENNAL MOVEMENTS. Erect. The third antennal segment is swung forward and up so that its longi­tudinal axis is parallel to the substrate, with the distal end directed forward and the arista pointed directly upward. Arista Contact. When head-under-wing posture is assumed, the tip of the arista is brought into contact with the under surface of the vanes of the female's wings. This means that the antennae are erect at this time. Third Segment Contact. If a male with erect antennae elevates his head suffi­ciently while in head-under-wing posture, not only the arista but also the dorsal (anterior) surface of the third antennal segment rest against the under surface of the female's wings. PROBOSCIDIAL MOVEMENTS Display Position. The male of some species extends (extrudes) his proboscis toward the body of the female. The display may be static or may involve repeated movements (Dynamic Display). F = Proboscidial display in front of the female S = Proboscidial display at side of female R = Proboscidial display atrear of female A minus symbol ( -) represents Static Display A plus symbol ( +) represents Dynamic Display Licks Intermittently. The + symbol indicates that during head-under-wing posturing the male repeatedly licks the genital area of the female with the labellar surfaces of the proboscis. Grasps ~ Genitalia. The + symbol indicates that during head-under-wing posturing, the male grasps all or a portion of the genital area of the female with the labellum or the labellar armament of his proboscis. Grasps Other. The + symbol indicates some other part of body of the female is grasped by the labellum of the male. WING ACTION S symbol indicates that a single wing vane is involved in display. B symbol indicates that both wing vanes are involved in display. Numbers 1, 2, or 3. The males of most Hawaiian species exhibit more than one type of wing action during a normal courtship of the female. These actions are given numbers corresponding to their serial appearance in the courtship. Almost without exception these wing actions belong to different categories of display, e.g., stationary display and vibration. In comparison, a few species may exhibit two unique and distinct types of a single category, e.g., two quite different types of stationary display. Stationary. Male wing vane or vanes are extended and held stationary for a period of time. The extension movement typically involves both outward and upward elements, and may include rotation of the entire vane or vanes. Vibration. The wing vane or vanes are extended laterally from the resting position and then vibrated up and down. The degree of displacement later­ally varies according to the species; the vanes may be held horizontal to the substrate or rotated by some species up to 90° so that the vane is perpendic­ular to the substrate and the vibratory movements are then forward and backward. Vibration usually occurs in bursts or pulses of movement, each lasting for a fraction of a second to a few seconds, but with some species each pulse may be of many minutes' duration. Flicking. The wing vane is quickly displaced away from and returned without pause toward the median line (plane) of the fly's body. Flicking may consist of a single continuous out and in motion of the wing but in some species may be a pulse of repeated movements. Typically the action starts from and returns to the resting position, but in some species a continued burst of flick­ing may occur from an extended position, e.g., the tip of vane moves from 20° to 80° from the median line during the flicking motion. The flicking motion may involve a vertical (upward) as well as a horizontal (outward) com­ponent. Rippling. A slow fluttering vibration. The term fluttering has already been preempted to indicate a wing movement displayed by non-receptive females (Spieth, 1952). Therefore to avoid confusion, the slow fluttering type of wing Prime Antenna! Probosc is Leg Leg :\hdc1111ina l i\Iale to TABLE t po!'ture 'Ving ~ction mo,·en1c:lts n10Yen1e11ts display 1110\"ClllCillS lllO\"ClllClllS male position b " u b ~ .: ~ & ""O "' c ;: "' c .':' ~ 3 C-lale "' " ro111·tsl1 ip -~ 8 c " "' ·-"O i; ·o,, -~ "3 w u cl('n1e11ts • ;: ~ ""' ~ "' " 2 u "' " ;; .2" " ~ "O ;: ~ i5. ~ "' '5" "§" ·u " E 0 " 0 -..i: ~ ·;;; " " § ~ w >, ;: °'" " "'~ ~ " "' e ;!: " 0 ·- "0 " .S "' .: "' c. 0 " "' -g v ~ "' -" §: ~ "Vj ..C -~ " ~ -" 0 c. "' " " .J: ~ .g" :g ~ ~ " i5. -~ ~ " ~ ~ 5 "O " ~ ~ -~ ~ 0 111 1 .2 E 5: 2 ~ 0 !" E c -"' .J:" '• 5 0 ~ :;) 0:: ~ w < ~ c ~ a Vl o:l v;"' > G:: "' "' (J) " c: "-?. :;: 0 " " u > 0 u <" u "' "' ANTOPOCERUS A. aduncus A. longiseta • A. orthopterus• A. tanythrix A. vi!losus • 3 2 2= 2 1 1 1 1 1 + + + + + + + + + + + + + B B B B B t 3 2 1 1 1 2 1 + + D D Ds Vv ? ? D1 Dt D1 + + + ? + ? D ? c ? c + ? ? ? ? DROSOPHILA Picture Wings D. adiastola D. crucigera D. engyochracea D. grimshawi D. picticornis D. pilimana D. pilimana-like D. punalua D. spectabilis I. clavisetae t = t=1 2 t t=t 1=1 2 1 t 1 2 1 1 1 2 + + + + + + F-S-R­R+ R+ F+ S+ R+ F-S­ + + + + + + + B B B B B B B B 1 1 t 3 2 2 t 1 t 1 1 + + + + + + + + + + • Vs Ds Ds + + + + + + + · + + + + D3 Vt Vt V3 Dt D3 + + + + + + DCS cs cs cs cs cs cs ? ? ? + + + + + ? ? ? Spoon Tarsi D. disticha 1 2 F-R­ B,S 1 2 Vs + I2 + c + D. dasycnemia D. sordidapex Comb Tarsi D. pectinitarsus D. proceriseta D. spiethi Bristle Tarsi D. basimacula* D. fusticula* D. perissopoda D. (T.) petalopeza• Modified Mouthparts D. aquila-like D. comatifemora D. eurypeza D. hirtitarsus-like D. ischnotrix-like D. mimica D. tendomentum White-tip Scutellum D. fungicola Scaptomyza-Iike D. crassifemur D. nasalis D. parva Unplaced Species D. achlya D. atroscutellata D. imparisetae D. incompleta • i=i=i 1 2 2 R­ B B 2 i i i B i i i i 2 2 2 R­ B B B s i i=i i i 3 2 i i 2 2 i 2 .. . . R+F+ + + + + + + + B B B B B B B + B ++ ++ + .. . . .. B i 2 i i 2 i 2 .. F­ + + + B B B B i i ,2 2 + + s + + 01* Li,3 + + + c ? +? i .. .. Vv Vv Vv . . .. + i i i 2• i i + w Vs Vs Vs v + V3 L3 .. + + c c C* • * + i* i i i i 2 i i 2 + + + T w T w Ss Vv Vv Vv Di L3 V3 + . . .. .. .. .. . . .. + + + • .C • c + +• i + .. D,C + i . . . . . . Vi . . . . + + i* i 2 1* i* i + + + + T T T s w Ss Vs L3 V3 +· + + + + c,s• c,s • + + • Seep. 248 for explanation of symbols. The University of Texas Publication vibratory movement displayed by the males of some Hawaiian species is termed rippling. Semaphoring. A special type of flicking display which involves both wing vanes of the male. In semaphoring one vane is moved (extended) away from the median line of the insect's body, while simultaneously the other vane is moved (adducted) toward the median line. PIGMENTATION IN VANE. Many species display unique, distinct patterns of melan­istic pigmentation in the wing vanes. These seem to be invariably correlated with the type of wing action displayed by the males in the courtship pattern. LEG DISPLAY. These display movements do not involve contact with the female's body. D = The tarsal segments are drummed, i.e., vibrated, against the substrate. S = The leg is extended away from the body, usually raised upward and then sharply brought down, i.e., struck, against the substrate. T =The foot, i.e., the tip of the tarsus, is tapped against the substrate. V = The entire leg is vibrated, typically extended away from the body and may or may not come into contact with the substrate at any time. W = The entire leg is extended and slowly vibrated (waved). LEG MOVEMENTS. Leg movements that occur while male is engaged in head-under­wing posture. D =Male fore legs strike or vibrate against dorsum of female's abdomen. L =Male fore legs strike or vibrate against side of female's abdomen. V = Male fore legs strike or vibrate against venter of female's abdomen. Subscripts = Male's fore legs slash or strike. Subscript v = Male's fore legs vibrate. Fold. The male raises his fore leg and folds the tibia against the femur. This movement also involves the male positioning the folded fore femur and tibia alongside the head, with the femoral-tibial hinge pointing forward and up­ward. The tarsi hang vertically from the tibia. Gra.sps. The male grasps the tip of the female's abdomen with both of his fore legs, typically the tarsi serving as the grasping agent. ABDOMINAL MOVEMENTS Bending Movements. The male bends (i.e., curls) his abdomen. V = Abdomen bent ventrally. D =Abdomen bent dorsally. L = Abdomen bent laterally. Subscript 1 = Posterior end of abdomen bent 5°to 15 °from longitudinal axis. Subscript 2 = Posterior end of abdomen bent 16° to 90° from longitudinal axis. Subscript 3 = Posterior end of abdomen bent 91 ° to 180° from longitudinal axis. Compress Longitudinally. The male abdomen is compressed by shortening the longitudinal axis. Typically, the amount of shortening of the length is slight and is accompanied by a short tubular extrusion and/ or a droplet of liquid from the anus. Vibration. The abdomen is vibrated rapidly. Depending upon the species, this may be a vertical movement or a horizontal movement. Other. Unique adbominal movements. See text descriptions. CIRCLING. The male circles about the female. This may be from a position in front of the female, to her rear, or from the rear to a position in front of her. AGGRESSIVE ACTION. Males of many species show definite aggressive actions toward each other, and toward females at certain times. Two major complex ritualized aggressive behavioral elements are commonly observed: Curling (Symbol C). The fly raises both wings 15° to 20° and at the same time curls the abdomen laterally 20° to 40°, with the concave body surface facing toward the other fly. The aggressor positions himself so that the longitudinal axis of his body is parallel to the other individual. Typically the two flies face in opposite directions. The aggressor then rushes sidewise with a crab­like motion, and strikes the other fly sharply. Often two males will fight thus, both curling and assaulting the other. Often the impact of the aggressor will knock the other fly "off his feet" and occasionally toss him over his, the aggressor's, body. The defeated fly always flees. Slashing (Symbol S). The fly flicks both wings horizontally 10° to 30° from the resting position and then raises and fully extends both fore legs. At the same time his body is quickly tilted backward, i.e., the head and thorax upward, the abdomen downward. The fly then slashes downward with the rigid, straightened fore legs striking the opponent. The downward movement is accompanied by sharp flicking of both wings outward to 90° and back. Ap­parently the wing action aided by the abdominal movement creates the force necessary to accomplish the dawnward slash of the fore legs. Other Aggressive Actions (Symbol D ) . Some species utilize their legs, especi­ally the fore ones, to spar and thrust at other individuals. These motions may be accompanied by varied quick wing snapping and flicking. COURTING OTHER MALES. Males of many species court each other freely once they have been sexually aroused. Unlike many species of drosophilids from other parts of the world, the males of Hawaiian species do not seem to possess counter­signaling actions other than aggressive actions. DESCRIPTIONS OF MATING BEHAVIOR OF INDIVIDUAL SPECIES The species discussed in this section are segregated into groups. Except for a group of four unplaced species, the other groups represent assemblages of species that are believed to be closely related, not only by their sexual behavior, but also by their morphology. Since species groups have not been delimited for the Ha­waiian drosophilids, vernacular names are used to designate the entities except for the species belonging to the genus Antorxx:erus. It should be further noted that Drosophila (Trichotobregma) petalopeza Hardy is placed with the bristle tarsi group and ldiomyia clavisetae Hardy with the picture wings. The descriptions of the mating behaviors represent the typical pattern displayed by each species. In most cases numerous courtships and in some cases copulations were observed; these formed the basis from which the typical pattern was deter­mined. As was to be expected, individual variation occurred, and, whenever such deviations were repeated consistently and often, they are noted in the descriptions. The University of Texas Publication Genus Antopocerus Hardy Introduction Hardy (1965) established the genus Antopocerus for those Hawaiian drosoph­ilid species which possess a long and large first antenna! segment and also have the arista densely covered on its dorsal surface with short setae (hairs) instead of the usual sparse long aristal setae. At present 11 species are assigned to the genus. The courtship behavior of 5 of these has been observed. Since none of the species can as yet be effectively reared in the laboratory, all observations were made on field-captured adults. Antopocerus aduncus Hardy Courtship. The male taps and erects his antennae so that they are pomtmg directly forward; this action results in the aristae being directed straight upward. At the same time he spreads both wings 20° out and up, goes to the rear of the female and assumes the head-under-wing posture with the tip of the arista in direct contact with the under surface of her wing vanes. If she does not decamp, he may stand in this position for a period as long as 2 minutes, intermittently extending his fore legs forward and parallel, and drumming the tarsi against the substrate. Often, however, the fore tarsal drumming is omitted and, as soon as his antennae are positioned against the wings of the female, he flicks both wings out and upward with a rotating motion so that the vane of each wing is almost vertical, its anal margin directed downward. The amplitude of this flicking motion is such that the tip of the male's wing reaches a point at least 10° to 20° anterior to the costal wing root, i.e., 100° to 110° from the resting position. The wing flick is repeated two or three times and then the male extends his fore legs forward between the vanes of the female's wings and the dorsum of her abdomen in what appears to be a grasping motion. At the same time he rears upward as if to mount, but does not curl under the tip of his abdomen. Two other courting actions of the male were also observed: (1) Standing 1 to 2 ems. away from the female, usually at her side but sometimes diagonally in front of her, the male will flick both wings up and out with the same rotational movement of the vanes as described above, but the forward motion does not exceed 80° from the mid-line of the abdomen, whereas in the posterior position the movement exceeds 100 °. ( 2) Occasionally the typical wing flick, with the male in rear position, also involves a rapid burst of vibration of a fraction of a second's duration just as the wing reaches the limit of its anterior motion. Non-receptive females kick, clean their wings with their metathoracic legs and decamp. Sexually aroused males approach each other, tap and court, but always with reduced intensity. The approached individual and the one which taps typically move apart without any aggressive behavior. Copulations. None observed. Aggressive Behavior. The males infrequently joust vigorously with their fore legs. Study Specimens. Five males and 3 females collected at Waikamoi, Maui, 4000', 1-7.X.64. Field Observations. No courtships or copulations were observed in the field. This species, like all other Antopocerus species, is shy and secretive in the field. Speci­mens were caught by sweeping under bracken fem fronds and around fallen logs. Comments. These relatively large flies, when observed in 25 X 90 mm. vials, only occasionally would engage in courtship, always of short duration, fleeting in nature, apparently due to the females' constant decamping and their general restlessness. When placed in the glass cell (see p. 247) the females were more placid in their behavior and the males much more persistent in their courting activities. The courting actions of the male (e.g., the wing display) are different from those of the clear-winged Antopocerus species: A. orthopterus, A. tanythrix and A. villosus. Unlike these three species which have clear wings, A. aduncus has a smokey colored area in the anterodistal wing portion. Further A. aduncus males do not use abdominal movements as do the other three species listed above. The possible significance of these differences is discussed on p. 305. Antopocerus longiseta (Grimshaw) Courtship. The male taps, erects his antennae, circles to rear and assumes head­under-wing posture. The male's body is depressed close to the substrate, the head and thorax more strikingly depressed than the abdomen; the male fore legs are directed forward with the femora and tibia partially folded but the distal end of the tibia and the tarsi are against the substrate. The male tarsi are then drummed against the substrate and simultaneously the abdomen is bobbed up and down. Periodically, the drumming action is interrupted, and the dorsal and inner (anterior) surfaces of the long basal tarsal segments are drawn across the labellar lobes in a cleaning-type motion. A few seconds after the initiation of the leg action, the male flicks both wings upward 45 ° and forward 100° so that the wing tips reach anteriorly to the base of the wing roots. At the extremity of the wing flick, both wings are vibrated rapidly in small amplitude for a fraction of a second, following which they are returned to the resting position. Two or three bursts of such wing action follow closely together and then the male lunges forward and upward, forcing his head between the female's wings and abdomen but without curling the tip of his abdomen forward. At the same time he extends his fore legs under the female's abdomen and uses the fore tarsi to drum rapidly against the venter of her abdomen. The long setae on the male's first and second tarsal seg­ments, especially the tuft that is found near the distal part of the basitarsus, strike against the female's genital area. Simultaneously with the lunging action, the male's wings are spread to 90° horizontally and vibrated rapidly, the abdomen vibrating in unison with the wings. This complicated pattern of actions continues for approximately one second, and then the male drops back to the head-und.er­wing stance and repeats the entire sequence, i.e., tarsal drumming and wing flicking followed by lunging. The male is extremely sensitive to any kicking action by the female and will not engage in wing flicking and lunging unless the female is completely immobile. The male periodically circles the female, keeping his head directed toward her The University of Texas Publication body at all times and about 1 to 2 cm. away from her body. He interrupts his circling to pause briefly and flick both wings out and upward circa 45°. After circling, he returns to the head-under-wing posture. Non-receptive females decamp, kick, and also often depress their wing tips against the substrate, thus preventing the male from achieving even the head­under-wing posture. Whenever this occurs, the male invariably circles about the female and then returns to the rear and attempts to achieve the typical posture. Copulations. None observed. Aggressive Behavior. The sample size was inadequate for determining the pres­ence or absence of aggression. Study Specimens. One male and 2 females, collected at Puu Kolekole, Molokai, 17.XI.64. Field Observations. A. longiseta, like other species of the genus, are shy, secretive flies and have not been observed directly in the field. The adults can be collected by sweeping around logs and ferns in the vicinity of Cheirodendron trees. Comments. The courting behavior of the male is somewhat similar to that of aduncus. Both species have dark clouds on the apex of the wings and both employ the wings and legs in a manner quite different from that shown by orthopterus, tanythrix and villosus. Furthermore, the dorsad curling of the abdomen is not displayed by longiseta and aduncus. Antopocerus orthopterus Hardy Courtship. The male taps, circles to rear and assumes head-under-wing posture, with his head and the third segment of his antennae erected and pressed against the female's wings. The wings are then extended to goo horizontally and rotated so that the vane stands perpendicularly to the substrate with the costal margin directed upward. From this position the wings are vibrated in pulses of varying speed and duration through an arc of circa goo, i.e., 45° anteriorly and 45° poster­iorly to a transverse axis through the wing bases. At times the vibrations are so rapid that they appear as a moving blur to the observer. At the same time, the abdomen is curled upward so that the dorsal surface appears concave and the male's anal papilla is directed upward. Non-receptive females decamp and kick rearward. Males tap each other, but do not proceed further with the courting sequence. Copulations. None observed. Aggressive Behavior. Males curl and occasionally, when courting a non-receptive female, change quickly from the courting sequence to the curling posture and attack the female. Study Specimens. Numerous specimens, collected at Paliku, Haleakala Crater, 6300', Maui, 27.VII.64. Comments. The fantastically rapid intermittent pulses of wing vibration and the concave posture of the abdomen with the anal papilla turned upward seem char­acteristic of the "clear" winged species of Antopocerus (i.e., wings that lack any pigmentation in membrane of the vanes). The presence of specialized long hairs ( setae) on the tibiae of the males suggests that the complete repertoire of the male's display was not observed. Invariably, when such sexually dimorphic characters are present in the Hawaiian species of Drosophila, they have been found to play a role in the courting behavior. Antopocerus tanythrix Hardy Courtship. The male taps, circles to the rear and assumes head-under-wing posture. His antennae are erected and in contact with the under surface of the female's wings. He then periodically extends his wings to 90° with the vanes parallel to the substrate. The two wings are vibrated horizontally to the sub­strate at a speed which makes them appear as a blur to the observer. The arc of movement of each wing is approximately 60°, i.e., 30° anteriorly and 30° poster­iorly to a transverse axis through the wing bases. Simultaneously, the male curls the tip of his abdomen upward so that the anal papilla points directly up, while the dorsum of the abdomen assumes a concave "saddle" shape. A droplet of clear fluid is partially extruded from the anal opening as the abdomen curls upward and remains visible as long as the curled position is maintained. Non-receptive females decamp and kick. Copulations. None observed. Study Specimens. Three males and 4 females collected at Kilauea, Hawaii Na­tional Park, 4000', Hawaii, ?.VII.63. Comments. The courtship observations were made during July, 1963, under en­vironmental conditions in the laboratory that were not entirely suitable for the wellbeing of the flies. It is believed that the courtship pattern of the male, as described, is incomplete, especially in view of the presence on the male's fore tarsi of numerous long slender setae. Antopocerus villosus Hardy Courtship. The male courtship of this species is similar to that displayed by A. orthopterus except that the males were observed to circle in front of decamping females in an effort to prevent their escape. Non-receptive females decamp and kick. As with all Antopocerus species, the kicking action of the female is highly effective for repulsing the male. Copulations. None observed. Aggressive Behavior. Males occasionally curl, especially when another fly ap­proaches them. Study Specimens. Numerous specimens collected at Puu Kolekole, Molokai, ?.VII.64. Comments. This species displays the courtship pattern that seems basically simi­lar for all of the clear winged Antopocerus species. In all probability the full repertoire of male courting actions has not been observed since the male fore basitarsus has a dense clump of long black hairs near the base and the fore tibiae are densely villose. Summary The basic courtship pattern of the males of the species of Antopocerus includes (1) head-under-wing as the prime posture position, (2) involvement of the modi­fied antennae in the courting sequence, and ( 3) wing display. The University of Texas Publication The five species observed divide into two groups: the clear wing and the pig­mented wing species. Of the former, all ( 1) extend both wings laterally to 90° and from this position vibrate them at extremely high speed through a large arc of movement; (2) at the same time, they curl the abdomen dorsally so that the anus points directly upward; and (3) the males do not make forward lunging motions when courting nonreceptive females. The pigmented wing species ( 1 ) utilize the wings for flicking motions that involve both lateral and vertical components and the vibrations that occur are secondary and of small amplitude, (2) do not curl the abdomen, and (3) do make lunging movements as part of the regular courting sequence, regardless of the receptivity of the female. It is also to be noted that the male of aduncus, which strikes at the dorsal surface of the female with a grasping motion of his fore tarsi as he lunges forward, lacks long specialized setae on the fore tarsi. In comparison longiseta, which drums (vibrates) the fore tarsi against the ventral surface of the female's abdomen as he lunges forward, does have long specialized setae on the fore tarsal segments. It is of interest that those structures which are used for the definition of the genus are also structures which play a specialized role in the courting behavior of the male. It can be theorized that sexual selection pressure resulted in the evolution of the characters which now differentiate these species from other drosophilids. If one seeks to determine the closest relative of Antopocerus among other drosophilids, the most striking evidence derived from the mating behavior is the similarity of the courtship pattern with that of the picture wings D. pilimana, D. pilimana-like, and D. punalua. These three species (1) all have "normal" appearing drosophilid antennae but, unlike any other Hawaiian drosophilids (except the Antopocerus species) the males, while courting, erect the antennae so that the aristae reach a vertical position and the aristal tips are then brought into contact with the wing vanes of the females; (2) the males all extend the wings away from the body before engaging them in movement; (3) in the case of pilimana, the abdomen is compressed longitudinally (shortened), but most interestingly the contraction is greatest on the dorsal surface so that the anus of the male is pulled upward and forward; further, as the pilimana male courts, a small bubble of liquid is repeatedly extruded and retracted from the anal open­ing; (4) D. punalua actually curls its abdomen dorsally, just as do the clear winged Antopocerus species. It is suggested that the mating behavior indicates that the genus Antopocerus has been evolved from a picture wing ancestor related to the stock from which pilimana and its relatives have been derived. The evolution of the most primitive Antopocerus species from a picture wing ancestor would appear to have involved the loss of pigmentation from the wings, the modification of the antennae, and the acquisition of the rapid vibration of both wings of the male during courtship. From this primitive stock the specialized courtships of aduncus and longiseta have evolved by ( 1) the re-acquisition of pigment in the distal antero-apical part of the wing vanes, a kind of pigmentation quite different from that of the picture wings, (2) a different kind of male wing display that is correlated with the pres­ence of the wing pigmentation, ( 3) the reduction of wing vibration, and (4) the loss of abdominal curling. Further, since, on the basis of morphological evidence, aduncus and longiseta belong to different sections of the genus Antopocerus (see Hardy, 1965), it is sug­gested that their courtships, although somewhat similar, have evolved separately, each from a more primitive clear winged ancestor. Significantly, the apparent reacquisition of pigmentation in the wing vanes and the correlated changed function of the wing during the courtship behavior are to be found in other species groups such as the bristle tarsi and the spoon tarsi groups. Genus Drosophila Fallen Picture Wing Species Group Introduction The picture wing species all have wings extensively marked with melanistic pigmentation, including markings in the cells, across the middle of the wings, or at least a well defined spot of pigmentation near the middle of vein R2 +a· The fore tarsal segments are normal with respect to shape and number. The individ­uals are moderately large, i.e., 3 to 5 mm. in body length. Several of the species, unli:ke many of the other Hawaiian drosophilids, are attracted to fermenting materials such as rotting fruit, the exuding sap of cut or broken lobeliads, tree ferns, etc. Adults have been bred from a variety of substances and a number of species can be quite successfully reared in the laboratory. The adults can be col­lected in the field both by baiting and sweeping. Several picture wing species are found at lower elevations than are most other species. Drosophila adiastola Hardy Courtship. The male taps, then, standing at the rear, side or front of the female with his head directed toward her, he spreads both wings outward 20° to 30° and upward 10°, the wing vanes forming a shallow V. Simultaneously, he raises his body by extending his legs slightly, extrudes his proboscis forward toward the female with the tip pointing slightly downward, compresses, straightens and slightly elongates his entire abdomen so that the anal tip is slightly elevated and pointing directly backwards. He then moves to the rear of the female, holding wings, proboscis and abdomen in the position described, but now he depresses his body toward the substrate and assumes the head-under-wing posture with the tip of the extended proboscis close to the tip of the female's abdomen. The male never attempts to touch the female with his proboscis but occasionally ex­tends both fore legs forward and strikes upward against the ventral or latero­ventral side of the female's abdomen. In doing so the fore tarsus is so rotated that hairs of the first tarsal segment strike against the area of the female's genital plates. Sexually aroused males, especially after unsuccessful courting for a time and when standing in front or near the front of the female, will display additional courtship actions. The most common action is a forward and upward rotating flicking of both wings from the shallow V position. More rarely, but repeatedly The University of Texas Publication observed, the male flicks both wings forward and upward, so that the wings extend outward at 90° from the longitudinal axis of his body, and then both vanes are vibrated very rapidly for a period of 1 to 2 seconds. At the same time the abdomen is curled upward and forward so that the abdominal tip is directly above the posterior margin of his thorax and directed forward. If the burst of wing vibration is short, the male's abdominal tip may not reach the extreme position just described but may point directly upward. At the end of the vibra­tory period, the wings are returned to the shallow V position and the abdomen is uncurled. A non-receptive female either decamps or keeps her abdomen slightly curled so that the tip is directed downward and responds to the male fore leg move­ments by kicking rearward, usually with her mid legs. Copulations. A single copulation was observed. Two observers had watched a single male court a female for 10-plus minutes, but the acceptance response of the female to the male's overtures was so rapid and obscure and the male's mounting reaction so quick that neither observer clearly saw the female's accept­ance response. Presumably she elevated the tip of her abdomen slightly; the male probably licked and mounted as a coordinated movement. In the copulatory posi­tion, the female's wings were spread outward by the body of the male, his fore tarsal claws grasped the anterolateral surface of the female mesothorax. The male mesothoracic legs rested on the sides of the female's abdomen and his meta­thoracic legs were on the substrate. His head rested on the female's body at the junction of her thorax and abdomen. The female's body was close to the substrate and her retracted proboscis actually rested on the substrate. During the entire copulatory period the mating pair was completely immobile. The copulatory period lasted for 11 minutes and then the female, without the slightest warning action of any type, literally jumped out from under the male. He dropped to a standing position and immediately moved about in the observation cell, but made no attempt to court any of the other five females. Aggressive Behavior. Males approach and tap another male, and immediately face each other and joust vigorously with their fore legs before breaking apart. Males also occasionally terminate an unsuccessful courtship by treating the fe­male with the same aggressive actions that a male would have received. Further, both males and females occasionally display curling and snapping (seep. 253) of the type observed in other species of picture wings. Study Specimens. Numerous specimens collected at Waikamoi, Maui, 4000', 1-7.X.64 (18 females and 12 males) and 21.X.64 (6 females and 1 male). Field Observations. A series of observations on the behavior of this species were made at Waikamoi, Maui, 1-7.X.64, where it, along with other drosophilids, was observed feeding upon the freshly cut stumps of tree ferns (Cibotium sp.) and lo beliads ( C lermontia sp.) . On the feeding site the individuals were extremely quiet, wary and alert, so much so that it was essentially impossible to approach closely to the insects. With the aid of a Questar telescope it was possible, however, to observe their activities. As many as 12 individuals were seen feeding on one fern stump where, at no time, did any one of them show any indication of aggressive action toward another individual. Like all other species of Hawaiian drosophilids that were ob­served in the field, the behavior of adiastola on the food site is decidedly cryptic. No feeding individuals were seen to engage in courting actions of any sort, not even tapping. Since this species displays obvious sexual dimorphism, it was pos­sible to identify the males in the dense vegetation surrounding the stumps where they were observed to be engaging in typical "fighting" actions with each other, and one male was observed on the side of a Clermontia sp. stump to be engaged in courting actions. Because of the nature of the dense vegetation, it was impos­sible to see if a female was on the hidden side of the stump. As with other species observed in the field, males apparently spend little time on the feeding sites and mostly are to be found perched singly throughout the surrounding vegetation close to feeding sites. If two males come into close proximity under such circum­stances, they engage in aggressive actions which end when one of the two even­tually flees to another twig, limb, or leaf. When sweeping for adults in the field, it was repeatedly observed that the areas under and close by Clermontia sp. and Cyanea sp. shrubs yielded the major­ity of the specimens of adiastola that were captured. Adults have been reared from both rotting leaves and rotting fruits of Clermontia sp. and from the rotting fruits of Cyanea sp. Drosophila crucigera Grimshaw Courtship. The male taps, goes to the rear and assumes the head-under-wing posture. He then extends and rotates his proboscis so that the ventral labellar surfaces face forward. The large, modified anterolateral section of each labellar lobe is then pulsated rapidly with an in-and-out action. If the female does not decamp or kick vigorously, the male will fold and orient his front legs along the side of his head, with the femora directed forward and upward alongside the head, the tips reaching beyond the anterior margin of his eyes. The tibia is ap­pressed against the ventral surface of the femur and the tarsi hang downward. From this position the male will ( 1) extend the proboscis forward and down and firmly grasp the female's anal papilla between the large anterolateral spe­cialized lobes of the proboscis, (2) lunge upward and slightly forward, and (3) place the fore legs upon the female so that the femur and tibia are resting upon the posterior tergites of the female's abdomen. His tarsi encircle the venter of her abdomen in the region of the vaginal plates with the tarsal tips crossing under her abdomen. The lunging movement of the male results in his proboscis being stretched, elongated and actually directed slightly posteriorly. It also exerts upward pressure on the tip of the female's abdomen. The fore legs are then vibrated rapidly, resulting in complex movements that cause the tarsi to move both anteriorly and posteriorly, dorsally and ventrally, while the folded femur and tibia are moved diagonally against the sides of the female's abdomen. At the same time the male's abdomen is slightly curled ventrally. To achieve copulation the male releases his hold on the female's anal papilla, curls his abdomen under and thrusts himself still further'forward onto the female, grasping the dorsum of her abdomen with his fore tarsal claws. The female's wings may merely be pushed upward by the male's body or slightly spread. All of the male's legs may be off the ground and the rear tarsi used to rub the vaginal plate area of the The University of Texas Publication female. Sometimes the two rear legs of the male are on the substrate; then, the caressing motions are absent. Invariably during the last minute or minute and a half of the copulatory period, the male releases all hold on the female with his legs, folds them loosely against his awn thorax and assumes a trance-like state. At the end of the copulatory period, the female kicks with her hind legs and causes the break of the genitalic union. The male falls or slides off and typically lies immobile for 10 to 45 seconds in what appears to be a trance on the substrate. At the end of that period, he regains his feet and will immediately court. A non-receptive female repels the male's overtures by keeping the tip of her abdomen curled downward close to the substrate and by vigorous kicking with the hind legs. Males, when sexually aroused, will court each other and will also attempt to copulate with other males. They have been observed to form courting chains of individuals, up to three in number, behind a courting pair. Males have been seen to curl the tip of the abdomen downward against the substrate and at the same time to extrude from the end of the anal papilla the intra-anal lobe (see Throckmorton, this Bulletin). The male then walks back and forth across the surface with the intra-anal lobe in contact with the substrate. Such actions seem always associated with sexual activity and apparently the male must be laying down a scent. The following observation is pertinent: 9 females and 14 males, all virgins aged 47 days, were introduced into the glass cell, which had been washed with a cleansing agent following its precedent use. Within one minute after introduction of the flies, a male began abdomen-dragging on the floor of the cell. He crisscrossed the central area moving 1 to 2 inches, then turned and assumed a new direction. These locomotory movements brought him into contact with five different individuals. Each time he assumed the courting posi­tion. With the first two encounters he quickly changed from courting to aggres­sion; with the next two he desisted after a few seconds of proboscis movements; the fifth meeting involved a true courtship. What is more significant is that other flies from various parts of the cell converged upon the area where the male was dragging his abdomen and the other males immediately became sexually active, courting each other and the females. It was almost as if a trigger mechanism had been activated. There was no doubt that the male's action had attracted the other flies and had engendered rapid and vigorous courting activity on the part of the other males. Copulations. Sixteen copulations were observed. Nine of these, which occurred in the cage, were not observed in their entirety (see comments below) .1 Data on five timed copulations indicate that the average copulatory period is 3'21". Aggressiveness. Individuals of both sexes display considerable pugnacity toward each other, especially when confined in a relatively small space. In addition to curling and snapping, they often engage in jousting with their pro-and meso­ thoracic legs. Study Specimens. Large numbers of individuals were used for observation in the 1 I am indebted to Mr. Kenneth Kaneshiro for observations on the actual copulatory behavior in the large cage. His diligent and careful observations resulted in the recording of nine copula­tions. laboratory. The number observed at any one time varied from single pairs to 100 or more individuals. Observations were made with the specimens in con­tainers of various sizes, ranging from 8 dram shell vials to the cage, and were made at various times ranging from 0600 to 2400. A number of stocks of D. crucigera have been established in the laboratory. It was possible to utilize virgin individuals as well as field-caught adults from various localities. Field Observations. No courtships were observed in the field. The species will come to baits, i.e., rotting fruits and rotting laboratory food, but no courting and/or copulating pair has ever been observed on the food sites. A five-hour observation period of a large bait mass at Halemanu Valley, Kokee, Kauai on 30.VIIl.64 clearly indicated that when on a feeding site these flies (like all other Hawaiian drosophilids) are quiet, wary, and secretive in their actions, and re­strict their activities at such times to feeding and possibly ovipositing. Comments. In the laboratory the flies display totally unpredictable behavior with respect to courting and mating, regardless of the previous history of the individ­uals. Most often the specimens simply sit; at other times they constantly engage in aggressive activities, and frequently only a single male out of 3 to 10 males being observed will be sexually active. Rarely will most of the males become sexually aroused. The only fact that is clear is that courtships are much more likely to occur where the experiments are conducted in "large" containers or cells, e.g., pint milk bottles, the large glass cell, or the cage. The females are still more unpredictable than the males. Often they will not even tolerate the males' attempts to court, and will constantly flee the presence of the males. At other times, the males will court intensively for long periods of time, up to one hour, without any acceptance responses on the part of the females. Most of the copula­tions. were observed in the cage, which would indicate that the area factor must play a role in the acceptance response of the females. Nevertheless, stocks of these flies can and are being maintained in the laboratory in 8 dram vials, and 2 copu­lations were observed in such vials. Clearly there is much that is still unknown about the sexual biology of this species. Drosophila engyochracea Hardy Courtship. The male taps, goes to the rear of the female and assumes the head­under-wing posture. He then extends his long fore legs along the dorsolateral side of the female's abdomen and attempts to grasp it. Simultaneously he licks her genital area with quick stabbing motions of his proboscis. Males may also engage in a signaling behavior. Apparently the male fixates upon another individual which may be sitting some distance away, e.g., up to 5 to 8 ems., and signals by raising and spreading his wings out and upward 60° to 80°. At the peak of the wing motion, the wings vibrate very slightly for a fraction of a second and then drop back into the resting position. Simultaneously with the wing movements, the abdomen is raised and lowered with a bobbing type motion. A male will display the signaling movement several times, then advance several steps toward the other individual, repeating the signaling, and walking until he is close to the second fly. If a non-aggressive female has been approached, the male will then proceed with courtship. Typically, however, the approached individual, regardless The University of Texas Publication of sex, will assume an aggressive posture and either slash or curl against the suitor. Females repel the males by decamping, kicking and depressing the tip of the abdomen. Males apparently do not court each other. Aggressive Behavior. These long-legged, slender flies are aggressive toward each other. When kept in large glass cylinders with food and adequate moisture, they will "fight" with each other to such an extent that they break their wing vanes; after a number of weeks, they have sometimes reduced the wings to short stubs. The curling type of aggressive behavior is often violent and individuals have been observed to be knocked off their feet and literally tossed upward and over the body of the other fly. Study Specimens. Numerous specimens were collected during the summers of 1963 and 1964 in the vicinity of Kilauea, Hawaii. Observations were made using half-pint milk bottles, the large glass cell, large glass cylinders, and cages. Field Observations. No courtships or matings were observed in the field. The species is restricted to the island of Hawaii, where it has been collected only in the vicinity of Kilauea. Specimens can be collected by sweeping, especially along the trunk of moss-covered trees and logs. The flies are attracted to baits of rotting fruits and/ or rotting Drosophila laboratory food. They have also been collected feeding upon two species of decomposing fungi on the dead limbs and trunks of fallen soapberry trees. When feeding in the field, these flies, like all other Hawai­ian drosophilids, are extremely docile toward other individuals. Comments. The courting behavior pattern consists of a restricted number of elements and is quite atypical for a species of the so-called picture wing group. The signaling sequence which is so characteristic and frequently displayed by this species is quite similar to an action that males of D. pilimana display rela­tively infrequently. Field observations on the feeding behavior lead to the con­clusion that mating behavior most probably occurs at sites away from the feeding and/ or ovipositional sites. Further fragmentary data indicate that the signaling activity as well as scent deposition may be involved in bringing the sexes together. The following observation is cited in support of this idea: Four females and two males were introduced into a half-pint glass milk bottle. The bottle was then laid on its side and the mouth plugged with a flat paper (cardboard) cap. One of the males took up station upon this paper surface. Here he engaged not only in the signaling action but also moved back and forth across the surface dragging the tip of his abdomen against the substrate. These actions continued for approxi­mately 20 minutes. Gradually, the other five flies moved toward this area and one female moved onto the paper surface. The male immediately proceeded to court her. She refused his overtures and within a few minutes he changed from the courting pattern to aggressive actions and drove her away. Similar abdomi­nal tip-dragging movements have been repeatedly observed not only by the author but also by Dr. Frances Clayton who for many weeks kept a number of these flies in a large observation cell on her work bench. Drosophila grimshawi Oldenberg Courtship. The male taps and typically goes to the rear of the female, assuming the head-under-wing posture. If the female does not decamp or kick, he will then direct his proboscis forward, rapidly expand and retract (dose) the labellar lobes and at the same time raise his fore femur and orient it alongside his head with the distal end of the femur level with the top of his head. The tibia and tarsi are directed downward, and the long hairs on the dorsal surface of the tarsi are thus pointing forward toward the tip of the female's abdomen. From such a position, the male strikes forward and upward with his fore legs, the motion causing the dorsal (anteriorly directed) long tarsal hairs to come into contact with the nor­mally downward curled abdomen of the female. The upward vector of the male's fore leg movement lifts the tip of the female's abdomen, and the male then extends his proboscis to bring the labellar surface into contact with her genitalia. Oral contact is held for 1 to 2 seconds, during which time the male releases his fore legs' grasp of the terminal part of the female's abdomen and swings them upward, extending them to grasp the dorsum of her abdomen. Then the male's abdomen is curled under, the proboscis contact is broken, and intromission is achieved. During copulation, the female's wings may or may not be spread by the mounting action of the male. If spread, the male's fore tarsal claws grasp the dorsal surface of the vanes near the wing bases. If not spread, the male attaches to the antero-dorsal portion of the female's abdomen. The male's mesothoracic legs lie alongside the female's abdomen and may or may not touch the substrate. His metathoracic legs have the tarsal segments curled under the female's abdomen so that they are in contact with the genitalic area. During the last 1 to 2 minutes of the copulatory period, the male releases his grasp on the female's body, folds all legs loosely against his body and assumes a trance-like state, being attached to the female only by the genitalic union. Often he will fall off the dorsum of the female, and lie on his side on the substrate. At the end of the copulatory period, the female kicks with her hind tarsi at the point of genitalic union, and after a short time effects a break between the two individuals. The male immedi­ately revives, assumes a normal standing position, and cleans the genitalic area with his hind tarsi. Sexually aroused males, if standing at a distance of 0.5 to 1.0 cm. in front or diagonally in front of another individual, will infrequently spread both wings to 90° and then rotate the wing vanes so that the anal edge of the wing is directed downward. This posture is maintained for approximately a second or slightly less and then the male returns his wings to the normal resting position and at once circles to the head-under-wing posture. Courting males have also been ob­served to leave the head-under-wing posture, circle to the front of a female, and engage in wing spreading as related above. Non-receptive females refuse the male's overtures by kicking, decamping, and depressing the tip of the abdomen when the male attempts to lift it upward with his fore legs. Since the long dorsal hairs of the tarsi and the tip of the tibia appear to be the chief agents in effecting the upward pull on the female's abdomen, no great amount of force can be exerted by the male. Males will court each other. Further, sexually aroused males are attracted to courting pairs; chains of males up to three in number have been observed lined up behind a courting pair, with each male courting the male in front of him. When a number of individuals of both sexes (from 10 to 20) are observed in The University of Texas Publication the large observation cell, the males of grimshawi, like those of crucigcra, engage in abdomen dragging, i.e., the male curls the tip of his abdomen downward against the substrate, and at the same time extrudes the intra-anal lobe and drags it across the substrate as he walks. The length of each dragging action is 0.5 to 1.0 cm. and results in the deposition of a thin film of moisture on the substrate. Other individuals are clearly attracted to the area where a male is engaged in such action. If the females are non-receptive, they generally avoid the area and then only the males congregate. These congregated males engage in jostling, curl­ing, and also courting each other. In fact, they court each other much more vigorously than they do the occasional non-receptive female that wanders into the immediate area. Copulations. Using a laboratory stock (PK2), 7 copulations were observed. The copulatory time varied from a maximum of 3'26" to 2'35". The average was 2'56". Aggressive Behavior. Both males and females display aggressive behavior toward each other, especially when crowded. Slashing and curling are both commonly observed. In the case of slashing, an individual may become aware of another specimen that is moving into the vicinity, but is still several centimeters away. The aggressor may then engage in the slashing sequence even though the action may not result in any physical contact with the intruder. Nevertheless, the ag­gressive action, even without physical contact between the flies, will usually cause the intruder to flee. Study Specimens. All observations were conducted with laboratory-reared flies. The copulations occurred between individuals that were 43 to 49 days old. Field Observations. No courting activities were observed in the field. Comments. Like its close relative crucigera, grimshawi specimens display un­predictable sexual behavior under laboratory conditions. The size of the observa­tion cell obviously affects the behavioral pattern. All copulations occurred in the large glass cell. The same specimens had been previously placed in smaller cells without resulting in copulations. That neither space nor age is solely responsible for non-success in smaller observation cells is shown by the following data: On 14.X.64 at 1400, 14 males and 19 females were introduced into the observa­tion cell. These were observed until 1700, i.e., for three hours, during which time many courtships and one copulation occurred. At the end of the period the flies were aspirated from the cell, etherized, separated as to sex and returned to food vials. At 1720, 8 males and 17 females of the same stock were introduced into the same observation cell. These flies were brothers and sisters of the first group and had emerged at the same time, subsequently receiving the same post-eclosion treatment. Immediately courtships and copulations occurred; by 1800 six copu­lations had been observed. These 25 flies were allowed to remain in the cell until 0900 of the following day, at which time the females were sacrificed. Of the 17 females, only 7 were fecundated. Thus from 1800 on 14.X to 0900 on 15.X, a period of 13 hours, only one additional copulation had occurred. Similar speci­mens, including some of those used between 1400 and 1700, were introduced into the same observation cell 7 and 8 days later, but at these times no copulations and only a few courtships occurred. Drosophila picticornis Grimshaw Courtship. The male approaches another individual, typically at the side or front, taps, and then places himself in front of the tapped individual. If the latter at­tempts to escape, the male will move from side to side laterally to prevent the individual from escaping. Once the courted fly comes to a halt, the male lowers his head, extends his proboscis directing the labellar surfaces toward the other individual and taps the end of his proboscis against the substrate. After one or more periods of proboscis-tapping, the male circles to the rear, and assumes the head-under-wing posture. He then folds his fore legs in a manner similar to that observed in crucigera (seep. 261). The male's wings are rapidly and repeatedly flicked out and back to resting position, with the vanes being held flat. Sometimes the courting male omits the forward position and goes directly to the rear of the female. Non-receptive females repulse the male by decamping, kicking, and depressing the abdomen. Males do not attempt to mount non-receptive females. Males court each other in a manner identical to their courtship of the female. Copulations. None observed. Aggressive Behavior. Both males and females engage in aggressive behavior to­ward each other. Such actions include curling and slashing but this species ap­pears to be not as aggressive as crucigera and grimshawi. Study Specimens. Two females and 4 males, all the offspring of a single female collected at Kokee, Kauai, ?.VI.64. Field Observations. No courtships or mating behavior have been observed in the field. This species is restricted to the island of Kauai where it has been collected in the region about Kokee. The adults come to rotting baits rather freely. The five-hour observation period, described on p. 302, resulted in five times as many picticornis specimens coming to the bait as did crucigera individuals. Like cru­cigera, they are shy and wary. There was no sign of courting, mating, or aggres­siveness toward each other or the specimens of any species, including great num­bers of D. immigrans which is a common feral species on all the Hawaiian Islands, while the individuals were feeding. Courtship and copulation activities appear to be behaviorally separated from feeding and/ or egg-laying activities. Comments. The method of folding the fore leg when positioned behind the female is similar to the fore leg position of courting males of crucigera, pilimana and punalua. The extension of the proboscis when standing in front of the female indicates relationship to adiastola, spectabilis and I diomyia clavisetae. Drosophila pilimana Grimshaw Courtship. The male taps, goes to the rear and assumes the head-under-wing posture, with arista upright and the distal tip in contact with the female wing vanes. Both wings are then flicked outward 80° from the resting position, with vanes flat (parallel) to the substrate, and immediately are rapidly moved back and forth in a range of 20° to 80° from the mid-line for approximately 1 second and then returned to the resting position. At the very termination of the flicking period, the speed and amplitude of the wing motion increases and the wing is The University of Texas Publication carried forward to 90°, i.e., at right angles with the body. After a rest of 1 to£ seconds, the burst of flicking motion is repeated and males sometimes continue the courting activity for periods of one-half hour or longer. At the same time that the wing movements are occurring, the male periodically folds his front legs in a manner identical to that seen in crucigera, adiastola, and punalua. From this folded position, the legs are thrust with a forward and upward grasping motion against the tip of the female's abdomen. At the same time his proboscis is extended and an attempt is made to bring the labellar surface in contact with the female's genital area. The male also slightly compresses his abdomen along the longi­tudinal axis, thus shortening the length. The compression also results in the anal papilla pointing upward, and repeatedly a tiny droplet of anal liquid is extruded as the wings are flicked. A non-receptive female repels the male by decamping, kicking, and particu­larly by keeping the abdomen depressed and not allowing the male to lift it up­ward with his fore leg actions. Males court each other and form chains of individuals behind a courting pair. A male engages in signaling actions similar to those described for engyochracea (seep. 263 ) and will gradually approach another fly; if the individual happens to be a female, he will then court in typical fashion. Copulations. Two. See "Comments" below. Aggressive Behavior. This species is much less aggressive than most other species of picture wings. They do, however, occasionally engage in slashing and curling actions. Study Specimens. Both field-caught and laboratory-reared specimens were ob­served in various types of observation cells ranging from 8 dram shell vials to the large cages. Field Observations. None. Comments. This species can be maintained in the laboratory and is being reared in both Hawaii and Texas. The males court readily and persistently; one or more pairs of courting individuals can almost invariably be seen among the laboratory stocks. Nevertheless, despite many extended observations by the author and others, utilizing virginal individuals as well as laboratory stocks, only two copu­lations have been seen, both by Miss Kathleen Resch at the University of Texas laboratory. Mr. Kenneth Kaneshiro of the University of Hawaii spent more time studying this species than he did in studying crucigera and yet was unsuccessful in observing a copulation. Drosophila pilimana-like Courtship. The courtship is identical with that of D. pilimana except for one element of the behavior, i.e., the male when positioned behind the female curls his abdomen underneath his body so that the tip of the abdomen is directed for­ ward and reaches the posterior edge of the metathoracic sternum. Copulations. None observed. Aggressive Behavior. Similar to pilimana. Study Specimens. One male and 2 females collected 1-7.X.64 at Waikamoi, Maui. Field Observations. The adults were observed feeding upon freshly cut stumps Spieth: Courtship Behavior of Hawaiian Drosophilidae 26g of tree ferns and the lobeliad Clermontia sp., along with other species of picture wings, i.e., adiastola and spectabilis. Comments. This species, restricted apparently to the island of Maui, is a close relative of pilimana Grimshaw. In the collections made during the years 1 g53 and 1g54, the field identifications given specimens from Maui were noted as "pili­mana-like." The mating behavior clearly indicates the close relationship of this species to pilimana. The latter has not been collected to date on Maui, but has been taken on Kauai, Oahu, Molokai, and Hawaii. It appears that this species must be the Maui representative of the pilimana "stock." Drosophila punalua Bryan Courtship. The courtship of D. punalua is identical with that of pilimana except for two elements, i.e., wing and abdominal movements of the punalua males. The punalua male, when at the head-under-wing posture, extends both wings to 40° from mid-line with vanes flat and parallel to substrate; from this position, one wing is extended to goo while the opposite wing is retracted almost to the mid-line. This semaphoring of the wings is repeated frequently but not rhyth­mically. The male also elevates and slightly curls upward his abdomen with the anal area pointing almost directly upward. From a lateral view the abdomen, when held in this position, appears "sway" or "saddle" backed. Very rarely a male will extend both wings to goo from the mid-line and vibrate them for a fraction of a second. Copulations. None observed. Aggressive Behavior. As in D. pilimana. Study Specimens. Three males and 1 female, Pupukea, Oahu, and 2 females, Tan­talus, Oahu. Drosophila spectabilis Hardy Courtship. The male taps, then positions himself at a distance of 0.5 to 1.0 cm. from the female, at the side or rear. He slightly depresses his body toward the substrate and engages in a complex wing action which involves a combination of scissoring, fluttering and semaphoring. Each wing vane is extended outward from the mid-line 20° to 40° with the vane held horizontal (parallel) to the substrate. At the same time the lateral wing motion occurs, the vane is rippled slowly, i.e., fluttered up and down. The motions of the two wings are not in phase; rather, as one is being extended away from the mid-line, the opposite member is being returned to the resting (mid-line) position. These wing actions occur in bursts, with a frequency similar to the bursts of wing action displayed by pilimana and punalua. Accompanying the wing action but not coordinated with it, the male proboscis is periodically extended and elongated forward and slightly downward with the ventral labellar surface pointing toward the female. The elongation of the proboscis is relatively great and the tip is extended well beyond the face of the male. Between proboscis extensions, i.e., when the mouthparts are in normal resting position, the male regularly draws the fore tarsi between the labellar lobes and immediately follow­ing this "cleaning" action he extends and straightens the fore legs, pointing them forward with the tips directed toward the female. The University of Texas Publication Non-receptive females either decamp or ignore the courting actions of the males. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. One male and 2 females collected at Waikamoi, Maui, 1-7.X.64. Field Observations. Dr. Elmo Hardy has identified the study specimens as aber­rant spectabilis individuals. D. spectabilis was described from Molokai and it is possible that when specimens from Molokai are observed the courtship pattern may differ from that displayed by the Maui specimens. The individuals observed in the laboratory were collected at the same places where a number of picture wings, including adiastola, pilimana-like, and I. clavisetae were found. The deep black pigmentation of the body and wing veins which also heavily infuses the wing membranes causes these flies to appear as totally black. In the field they appear to be extraordinarily shy and wary. When feeding on the stumps of freshly cut tree ferns, they blend into the dark background to such an extent that is is almost impossible to detect them. Comments. The courting behavior as observed in the laboratory is unique in that, except for tapping at the beginning of the courting sequence, there is no physical contact between the courting male and the female. Clearly the acceptance re­sponse of the female must therefore involve either visual stimuli or air-borne chemical stimuli. Although only one male was available for study, he courted steadily for a two-hour period, alternating between the two females. We thus have reasonable confidence that the courtship pattern described is accurate. Idiomyia clavisetae Hardy Courtship. The male of this species displays several extraordinary sexual dimor­phic characteristics: (1 ) a greatly enlarged proboscis; (2) modified tergites on the first and second abdominal segments which allow this area to serve as a hinge for the dorsal flexing of the abdomen; (3) a transverse double row of extremely long, flaring, clavate, posteriorly-directed setae on the distal edge of the sixth sternite and the lateral and dorsolateral surfaces of the fifth and sixth sternites. The male taps, then positions himself at the side or in front of the female and extends and greatly elongates his proboscis toward her. The distal labellar sur­face is directed toward the female and is expanded laterally to a width that equals the body width of the male, and at the same time it is compressed dorsoventrally so that the proboscis tip appears as a broad band-like structure. Simultaneously the wings are extended to 90° from the mid-line and rotated so that the vane is held perpendicular to the substrate, with the anterior margin up. The wings are not vibrated but held immobile. Meanwhile, the abdomen is curled up and over the thorax and head of the male, with the modified first and second abdominal segments serving as a hinge. The tip of the curled abdomen reaches forward to a point directly above the male's eyes, and the tip and the long double raw of ab­dominal setae are directed toward the female. Once this contorted stance has been achieved, the male vibrates the end of his abdomen rapidly up and down in an amplitude of 1 to 2 mm. During these bursts of vibration of the abdomen, the posterior end of the male rectum is everted to the outside, forming a tube-like papilla of approximately 1 mm. in length. After displaying in this fashion for a variable period of time, the male reverts to the normal resting position, quickly circles to the rear of the female and posi­tions himself directly behind her with his head under her wings and close to the tip of her abdomen. He then at once strikes the dorsum of her abdomen with a tapping-grasping motion of the fore legs and simultaneously attempts to lick the female genital area with his proboscis. The male is aggressive in his behavior and vigorously fends off the kicking hind legs of a non-receptive female with his fore legs. After a variable period of time, if the female is non-receptive, he will circle to the front of her and again assume the "elevated abdomen" type of dis­play described above. Non-receptive females decamp, kick, and curl the tip of their abdomens toward the substrate. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. One male and 4 females collected at Waikamoi, Maui 1-7.X.64. Field Observations. TI1e species was observed feeding upon the cut stumps of lobeliads (Clermontia sp. ). Despite many previous collections having been made in this area by divers collectors, this species had not been previously captured. It occurs in association with adiastola, spectabilis and other rare picture wings which have not been collected in adequate numbers at any one time to enable the investigator to study their courting behavior. Comments. Few, if any, courting behaviors of Hawaiian drosophilids are as bizarre as the behavior of this species. Apparently the long abdominal hairs serve a fan-like function to direct volatile materials derived from the rectal extrusion toward the female's sense receptors. The wings, although involved in courtship, are held in a stationary position and can conceivably serve as a baffle to prevent stray air currents from dispersing volatile substances that are possibly released by both the extended and expanded proboscis and the rectum. Summary The 10 species of picture wings that were studied illustrate well the complexity of courtship patterns that have evolved among related species of Hawaiian dro­sophilids. Most of the 10 species are being maintained successfully as laboratory stocks. The study of several of these species was, therefore, conducted on both field-captured adults and laboratory-reared specimens. No single unique behav­ioral element or constellation of elements is common to all of the 10 species. All 10 species have extensive pigmentation on the wing veins and all but crucigera and engyochracea utilize wing display. The lack of wing display by these two species apparently is correlated with the substitution of movements of other structures, i.e., proboscis and leg action in engyochracea and a startlingly unique proboscis movement in crucigera. Six species (see Table 1) have a unique and distinctive method of folding the fore legs while courting at the head-under-wing position. Three of these 6 (pili­mana, pilimana-like and punalua) also use the arista in an identical manner The University of Texas Publication while courting. Four species (adiastola, picticornis, spectabilis, and I. clavisetae) employ the proboscis for display movements while at the front or side of the females, in contrast to crucigera and grimshawi which display or employ probos­cidial movements only at the rear during head-under-wing posture. D. adiastola displays the greatest number of separate courtship elements and has been ob­served to display three different constellations of courting elements when pos­turing in front of the female. The overall patterns of mating behavior, the geographical distribution of some of the species at relatively low elevations, the attraction of a number of the species to fermenting substances, and the apparent lack of specialized food sources for many of the larvae indicate that the picture wings are among the most primitivr of the Hawaiian drosophilids. Spoon Tarsi Species Group Introduction The spoon tarsi species group is characterized by having the second segment of the male fore tarsal segment short, flattened, and concave on the inner surface, i.e., spoon-like. Members of the group are typically denizens of the wet rain forests where Cheirodendron and tree ferns are found. Species of this group reared to date oviposit in the ventral surface of fallen, decomposing Cheirodendron leaves; the larval period is spent inside the leaf. In areas where Cheirodendron is abundant and the rainfall is heavy and frequent, individuals of the various species can be found in great numbers, e.g., D. disticha in the W aikamoi area of Maui. The adults are collected by sweeping over the leaf litter of the forests, and especially under the dense growths of bracken fems where the adults congregate on the stems of the fronds and the under surfaces of the fern pinnae. Drosophila dasycnemia Hardy Courtship. The male taps and positions himself beside the female, facing toward her. The position assumed may be at any point from almost directly in front of to almost directly behind the female, but typically is along the side. The male then raises both wings out and up about 20° to 30°. From this wing-raised position, both wings are repeatedly flicked outward another 15° and returned. One vane precedes the other slightly, both for extension and retraction. The flicking continues for many minutes if the female does not decamp. At the same time, the abdomen is compressed, shortened slightly along the longitudinal axis, and slightly concaved dorsally. The compression of the abdomen results in the eversion of the posterior end of the male rectum in the form of a small tubular extrusion. After posturing thus, the male, if the female does not decamp or curl the tip of her abdomen ventrally, will circle rapidly to her rear, extrude his proboscis slightly forward and ventrally until it almost touches the substrate. As soon as he assumes the head-under-wing position behind the female with his face close to the tip of her abdomen, he extends his fore legs under her abdomen and then upward in a grasping motion directed toward the tip of the female's abdomen in such a manner that the male fore-tarsal "spoons" grasp the area of Spieth: Courtship Behavior of Hawaiian Drosophilidae Z73 the vaginal plates. This lifting motion pulls the tip of her abdomen against his extended proboscis. Males do not try to mount non-receptive females. Non-receptive females kick, decamp and depress, the last action preventing the male from grasping the female's abdomen with his fore tarsi. Males court each other without displaying any effective counter-signaling behavior. Copulations. None observed. Aggressive Behavior. Males occasionally engage in curling, but, in common with other species of the "spoon tarsi" group, the species is not particularly pug­nacious. Study Specimens. Numerous specimens collected at Kilauea, Hawaii 13.IX.64. Field Observations. D. dasycnemia Hardy lives in the wet rain forests of the island of Hawaii where Cheirodendron trees are present. Sweeping above the forest floor under the trees themselves usually results in the capture of females, whereas in the area immediately surrounding the trees males are more commonly caught. Comments. Although repeated observations resulting in many prolonged court­ships were made upon field-captured specimens, no copulations were observed. The extrusion of the rectum described above can be duplicated by placing gentle pressure upon the abdomens of etherized males. Drosophila disticha Hardy Courtship. The male taps and then places himself Z to 3 mm. away from the female, facing toward her. The typical position assumed is directly or diagonally in front of the female's head. He raises both wings up and out Z0° to 30°; the wing nearest the female's head is then flicked 15° to Z0° in pulses of activity, with pauses between. Simultaneously, the abdomen is compressed and shortened slightly along the longitudinal axis; the tip, displaying a small rectal extrusion siillilar to that seen in D. dasycnemia, is curled slightly toward the female's face. The proboscis is extended downward .and slightly forward with the tip almost touching the substrate. Such posturing may continue for many minutes. The male then circles to the rear and assumes the head-under-wing position directly behind the female, with his face close to the tip of her abdomen. He extends his fore legs under her abdomen and strikes upward against the side of her abdomen with his tarsi, using his legs alternately. Periodically he cleans the fore legs. Eventually both legs are raised simultaneously upward in a grasping motion directed toward the tip of the female's abdomen in such a manner that the male fore tarsal "spoons" grasp the area of the female's vaginal plates. This lifting motion pulls the tip of the female's abdomen against the male's extended pro­boscis. Males do not try to mount non-receptive females. After engaging in head­under-wing courting actions, the male will circle to the front of the female and repeat the wing and proboscis actions. Non-receptive females decamp, kick, and depress. The last prevents the male from grasping the female's abdomen with his fore tarsi. Males occasionally court each other without displaying any effective counter­signaling behavior. The initiation of courtship by a male attracts other nearby males who may try to intrude upon the courting individuals. The University of Texas Publication Copulations. None observed. Aggressive Behavior. Males occasionally engage in curling, but the species displays little pugnacity. Study Specimens. Numerous specimens collected at Puu Kolekole, Molokai 22.VII.64; Kaulalewelewe, Maui 3-4.VIII.64, and Waikamoi, Maui 1-7.X.64. Field Observations. D. disticha Hardy is abundant in the wet rain forests of Molokai and Maui whenever Cheirodendron trees are present. Sweeping under bracken ferns and over the forest floor will sometimes yield hundreds of speci­mens. Comments. This species is a close relative of dasycnemia Hardy. Not only does it have a somewhat similar courting behavior, but, like that species, is reluctant to copulate under laboratory conditions. Despite repeated observations involving hundreds of individuals in various sized observation cells and at various times of day, no copulations were observed. Since the species cannot as yet be successfully bred in the laboratory, known virginal specimens have not been available for study to date. Drosophila sordidapex Grimshaw Courtship. The male taps, moves to a point diagonally behind the female, spreads both wings outward 15 ° to 25° with vanes almost parallel to the substrate, elongates the tip of his abdomen and curls it slightly toward the female's head. After holding this posture for a few seconds, he then returns wings and abdomen to a resting position, and moves immediately to the head-under-wing position, extends his fore legs under her abdomen and strikes upward against her venter. After several such leg movements, both legs are used to grasp the tip of the female's abdomen, with the "tarsal spoons" placed against her genital area. The fore legs then lift the tip of the female's abdomen upward and bring it into contact with the male's proboscis which has not been extended. If she does not kick-or decamp as a result of these actions, but also does not give the acceptance response, the male will typically circle to the front of the female, face her, spread his wings 15° to 25 °, then vibrate his entire body from side to side, and simul­taneously curl his abdomen laterally to form a U with the tip extending toward the female's face and reaching forward to the level of the anterior margin of the male's eye. Once the abdomen has achieved the maximum limit of curl, the body vibrations cease except that the abdominal tip alone vibrates horizontally. At the same time, the mesothoracic legs are raised and extended, the dorsal surface of the entire leg forming a concave arc with the tarsi elevated above the level of the dorsum of the thorax. From this position, the legs are snapped downward against the substrate. After posturing in front of the female for several seconds, the male returns to the head-under-wing position. Non-receptive females decamp, kick, and depress. Copulations. None; males did occasionally attempt to mount non-receptive females but were unsuccessful. Study Specimens. Two males and 8 females, collected at Niaulani Ranger Cabin, Kilauea, Hawaii 13.IX.64. Field Observations. D. sordidapex is an inhabitant of the wet rain forest in the vicinity of Kilauea and Volcano, Hawaii. It can most effectively be captured by sweeping in the vicinity of Cheirodendron trees. Comments. The small number of males (2) yielded no observational data about male to male courtship responses nor information concerning the aggres­siveness or lack of aggressiveness of the individuals toward each other. The males were persistent suitors and were observe~ to court the females continuously for long periods of time, i.e., 6 hours. Summary The males of the spoon tarsi group studied to date all show a common basic courtship pattern in that ( 1) the first prime postural position is at a distance from the female where the male spreads both wings up and out to a stationary display, and at the same time extrudes the terminal end of the rectum; (2) the male eventually assumes a head-under-wing position, then reaches forward with his fore legs, grasps the tip of the female's abdomen with the fore tarsi, the second tarsal segment spoon grasping the genital area of the female, and pulls the tip of the female's abdomen against the labellar surface of his proboscis. This basic courting pattern is over-laid by numerous species specific actions that distinguish the species and result in rather complex courting repertoires. Interestingly, dasycnemia and sordid.apex can be collected at the same time and at the same site. When, under laboratory conditions, males and females of both species are observed together in the same observation cell, the males of both species will tap and go through the initial courtship postures with females of both species. They usually are less persistent with the "foreign" females than with their own. On the basis of the evidence derived from the courtship behavior, it is suggested that the spoon tarsi species group has been derived from a picture wing ancestor of the D. adiastola type. It is also significant that the larvae of adiastola are decaying-leaf feeders and have been reared from lobeliads. Comb Tarsi Species Group Introduction The small, inconspicuous flies (usually 2 to 3 mm. in body length) that belong to this species group are characterized by the peculiar male fore tarsus which consists of only four segments plus a "comb-like" appendage arising at the junc­tion of the basitarsus and the second tarsal segment. The flies are most commonly found in the forested areas which are relatively drier (i.e., 60 to 80 inches of rain per year) than those occupied by the bristle and spoon tarsal forms. The adults can be collected by sweeping among the bracken ferns and over the leaf litter, but also are commonly found on the under surfaces of the leaves of broadleafed trees at a height of 6 to 8 feet above the ground. Drosophila pectinitarsus Hardy Courtship. The male taps, moves to the rear and assumes head-under-wing posture. The fore legs are then lifted, swung inward to a closely aligned parallel The University of Texas Publication position, with the femur and tibia forming an inverted V, and the tarsi extending forward under the female's abdomen and parallel to the substrate. In the court­ing stance, the comb-like structure of the male's fore tarsi (which originates at the anteroventral junction of the first and second tarsal segments and which normally points downward and forward) is extended at right angles to the tarsi, with the tip of the comb pointing toward the mid-line parallel to the substrate and directly under the female's genital area. The fore tarsi are then vibrated rapidly up and down. The amplitude of these movements is such that the tarsi strike or almost strike the venter of the female, and the tarsal combs seem always to strike her genitalia. The tarsal vibrations occur in pulses or bursts of move­ment. In between the pulses, the male retracts both legs simultaneously and draws the inner surface of the basitarsal segments and the tarsal combs between the labellar lobes of the proboscis in a typical cleaning motion. Occasionally, the male, after tapping, postures in front of the female and spreads both wing vanes out and upward, holding the spread-wing posture for a short period before moving to the rear of the female. Males will continue to court females for prolonged periods of time. A non-receptive female kicks, decamps and sometimes depresses the tip of her abdomen, thus preventing the male from extending his fore tarsi under it. When the male is in head-under-wing posture, the female's wings rest upon the dorsum of his thorax. If she depresses the tip of her abdomen, the male will often raise and lower his body rhythmically in what appears to be an attempt to elevate the female's abdomen. Sometimes he will tap and start toward the rear of the female, and then desist without achieving the head-under-wing posture. No visible action of any sort by the female could be detected, but apparently the male had received a repelling stimulus of some kind. Copulations. None observed. Aggressive Behavior. No obvious aggressive behavior was displayed by the flies at any time. All of the "comb tarsi" species observed to date seem to be non­aggressive in their behavior. Study Specimens. Numerous specimens collected at Kilauea (Bird Park), Hawaii 12-18.VII.64 and 7-14.IX.64. Field Observations. The adults can be captured by sweeping under bracken fems. Wherever the species is abundant, often the males but only occasionally the females can be found resting on the under surface of the leaves of broad­leafed species such as Pipturus hawaiiensis (mamaki), Coprosoma rynchocarpa (pilo) and Haimerliodendron bruonianum. Some flies were seen courting on the under surfaces of leaves of the shrub-like trees named above, but could not be positively identified as belonging to pectinitarsus. It is not known where the larval stage of the species is spent, nor have the adults been observed when feeding in the field. Comments. It is presumed that the fore tarsi of the male are the prime agents in transmitting courting stimuli. The repeated licking (moistening) of the comb indicates that either an air-borne or contact chemical stimulus must be trans­mitted to the female. The lack of any wing, abdominal or extruded proboscis movements when the male is posturing behind the female is characteristic of all the comb tarsi species observed to date. The males never show the slightest indi­cation of attempting to mount the non-receptive females. They are easily repelled by the kicking action of a female, and never show any signs of courting other males. They always appear to be "gentle" in their reactions to other individuals, both in the laboratory and in the field. The visual inconspicuousness of the court­ship, the lack of pugnacity, and their small size are correlated with the fact that flies of this species group are more often found in exposed sites in the field than are those of other species groups. These characteristics all seem to be related to the predator pressure that is believed to exist upon the Hawaiian drosophilids (seep. 311). Drosophila proceriseta Hardy Courtship. The courtship pattern of proceriseta appears to be identical to that displayed by pectinitarsus except that the male does not posture in front of the female. Copulations. None observed. Study Specimens. Numerous field-captured adults from Puu Kolekole, Molokai 22.VII.64 (Carson Collection). Comments. This species is a typical "comb tarsi" representative. Drosophila spiethi Hardy Courtship. This species in all respects appears to behave both in the field and in the laboratory similarly to pectinitarsus except that the male does not posture in front of the female. Copulations. None observed. Study Specimens. Numerous individuals collected at Kilauea (Bird Park and Kipuka Ki), Hawaii, 12-18.VII.64. Comments. Despite the sympatric distribution of pectinitarsus and spiethi, their behavior is strikingly similar. However, when both species are placed in the same observation cell, the males invariably court females of their own species and ignore those of the other species. Apparently they achieve this differentiation at the tapping stage; since spiethi is a yellowish-colored fly and pectinitarsus is much darker, the observer is able to distinguish the individuals with certainty. Summary The courtship of the comb tarsi species studied so far is relatively simple. The male taps, moves to the rear of the female, assumes the head-under-wing position and rapidly vibrates both fore tarsi synchronously under her abdomen. When the fore tarsi are at the vibrating position, the tarsal comb stands at right angles to the longitudinal axis of the tarsus, pointing directly toward the median line. Thus, the tips of the two combs point toward each other. Periodically the male interrupts the vibratory process to draw these combs and the adjacent areas of the tarsi between his labellar lobes in a cleaning-moistening type of action. All species of this group which have been studied have shown surprisingly similar behavior not only with respect to the courting sequence but also in their lack of aggressive behavior. The unique nature of the courtship gives no clue to the rela­tionship of this species group to the other species studied to date. The University of Texas Publication Bristle Tarsi Species Group Introduction The males of this group are characterized by having a cluster of heavy bristles at the dorso-apical end of the fore basitarsus. The number and shape of the bristles varies between species, but most typically the cluster consists of (1) four straight, erect bristles whose tips converge distally and (2) a group of short, usually kinked bristles which point distally with respect to the basitarsal segment and which partially surround the straight bristles distally. The kinked bristles would appear to be a protective barrier for the longer erect bristles. The bristle tarsi species are denizens of the rain forests, especially in those areas where trees of the genera Pterotropia, Tetraplasandra, and Cheirodendron are abundant. The greatest concentrations of individuals can be found in the vicinity of Kokee and Mohihi, Kauai, where great numbers of Pterotropia kaua­iensis (Mann) Hbd. are found. Drosophila basimacula Hardy Courtship. The male taps, goes to the rear of the female and assumes a head­under-wing posture. At the same time he spreads his wings outward and upward 10° to 20° from the resting position. The fore legs are then thrust forward under the female's abdomen and raised sharply upward with a striking motion against her abdomen. The male is so positioned that the clump of large black setae (bristles) arising from the anterodistal surface of the male's first fore tarsal seg­ment strikes against the female's genital area. Non-receptive females respond by kicking sharply rearward with their meta­thoracic legs or by decamping. Copulations. None observed. Aggressive Behavior. These flies, especially the males, are aggressive toward any other fly that is within their immediate vicinity except when actually court­ing. The individuals move rapidly about in the observation cell and whenever they come close together (circa 1 cm. or less ) they spread and elevate the wings 10° to 20°, curl the abdomen toward the other fly and rush laterally against the "opponent." Males, when in courting position directly behind the female, never display aggressive behavior but typically, after several seconds of unsuccessful courtship, the male will move laterally and immediately assume the aggressive posture and "attack" the female. Often the force of the male's impact is sufficient to displace the female physically to the extent of 0.5 cm. One male was observed to start aggressive action against a female near him, change to courtship, which was unsuccessful, and then return to aggression. Study Specimens. Numerous specimens collected at Mohihi (Kokee), Kauai, 4000', 21-23.VII.64 and 2.6.VIII.64. Field Observations. The adults can readily be captured by sweeping under bracken ferns in the area around Kokee, Kauai. They appear to be most abundant in places where trees of Pterotropia sp. and Cheirodendron sp. are numerous. Comments. The courtship feature of the male thrusting the fore legs under the female's abdomen and then striking upward against her body, thus bringing the heavy bristles of the first fore tarsal segment against the female's genital area, appears characteristic of all relatives of basimacula, i.e., the so-called bristle tarsi species. The lack of wing, abdominal and circling actions may indicate a rela­tively "simple" courting behavior or possibly indicates that the full panoply of actions has not as yet been observed. These flies do not survive well under labora­tory conditions and perhaps the entire repertoire of the courtship was not dis­played. Drosophila fusticula Hardy Courtship. The male taps, places himself in front of the female, extends his fore legs toward her, elevates his wings 10° to 15° but does not spread them. He then vibrates his body rapidly from side to side, which results in the wings appearing to vibrate rapidly back and forth with a synchronized lateral motion, both vanes moving in the so.me direction simultaneously. The male then circles rapidly to the rear of the female, and assumes head-under-wing position with proboscis extended. Aggressive Behavior. The male is extremely aggressive toward other flies, and often engages in curling actions against any other individual in his immediate vicinity. Comments. These incomplete observations were made upon a single male cap­tured at Waikamoi, Maui. The several females were extremely non-receptive and decamped as soon as the male reached the rear position. The presence of bristles on the male's fore tarsi indicates that the courtship must involve the use of the fore legs in a manner similar to that described for basimacula. The male was aggressive in his courting and when observed in a vial containing other species was seen to court females of species other than his own and also to engage in repeated aggressive (curling) action against other species, both males and females. Drosophila perissopoda Hardy Courtship. The male taps, elevates his wings 5° to 10° and then positions him­self in front of the female, facing her. He raises (semaphores) one wing upward to 85°, typically holding this extended position for several seconds. The wing is then lowered and the opposite one elevated. The alternate raising and lowering of the wings is accompanied by the curling under and forward of the abdomen, with the tip of the abdomen reaching forward to the posterior margin of the male's thorax. If the female attempts to decamp, the male arcs back and forth in ~ront of her in an effort to prevent her escape. Periodically he circles to the rear, assumes head-under-wing posture, extends his fore legs and strikes upward against the female's abdomen, with the tarsal bristles striking against her genital area. Non-receptive females decamp, kick, and depress the tip of the abdomen. The last action prevents the male from achieving head-under-wing position. Males court each other frequently and persistently. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. Numerous field-captured adults from Kokee, Kauai 12-13. VII.64 and 25-30.VIIl.64. The University of Texas Publication Field Observations. D. perissopoda is abundant in the Kokee area of Kauai and can easily be collected by sweeping. It appears to be most abundant in areas where the Hawaiian tree Pterotropia kauaiensis is present. No observations on the sexual behavior were made in the field. Comments. The male posturing in front of the female appears to be correlated with the male's dimorphic wings which possess a black pigmented infuscation in the distal third of the vane. Drosophila (Trichotobregma) petalopeza Hardy Courtship. The male taps, positions himself diagonally behind the female and flicks the wing nearest the female's head outward to 45° with vane horizontal to the substrate. After a short period of wing flicking, he then assumes head­under-wing posture with the vertex of his head in contact with the ventral sur­faces of the female's wings. The male then (1) rhythmicaly raises and lowers his body, thus forcing the female's wings up and down, (2) periodically elevates and extends his mesothoracic legs towards the female's head and waves each leg with a circular motion. The distal four segments of the tarsi of these legs are densely covered with heavy, short, intensely black pigmented setae that create a distinct black brush which is visible even without magnification. (3) He curls his abdomen laterally so that it forms a U with the tip directed forward. This curling involves compression of the abdomen and as a result a white balloon-like expansion is formed by the white articular membrane that surrounds the anal papilla and the genital sclerites. The darkly pigmented anal papilla and the genital area thus appear as black spots on an irregularly-shaped balloon-like terminal enlargement of the abdominal tip. ( 4) Periodically the fore legs are raised, partially folded and swung toward the mid-line and under the tip of the female's abdomen, and then are used to grasp her with the clump of heavy first tarsal bristles in contact with her genital area. Non-receptive females decamp, kick, and depress the tip of the abdomen, the last preventing the male from achieving his grasp with the fore legs. Males have not been observed to court each other. They do tap one another but the mating sequence does not proceed further. Copulations. None observed. A considerable number of persistent male court­ships were observed. The specimens were field-collected, and dissection of a number of females indicated that at least 20% did not have any sperm stored in the spermatheca and ventral receptacle. It is presumed they were virginal. Aggressive Behavior. The males are extremely aggressive. They employ their long legs effectively in jousting with other individuals, and also they display typical curling actions. Often two males when directly facing each other will elevate the anterior end of their bodies to the extent that the longitudinal axis of the body forms a 45 ° angle with the substrate. They then make physical contact with each other by thrusting the anteroventral surface of the thorax against the other fly in such a manner that the two heads are side by side. This position can best be described as a shoulder-to-shoulder posture. The two flies then visibly thrust against each other until one of the pair is defeated, turns and flees. In large observation cells, i.e., a pint milk bottle or larger, males were observed to select a site or station and establish a "territory" of circa 2 to 4 ems. in diameter which they vigorously defended against intrusion by any other fly. Courting males often change from courtship to aggression and drive the female away. Study Specimens. Numerous field-captured adults from Paliku, Haleakala Crater, Maui ?.VII.63 and 27.VIl.64. Field Observations. D. (T.) petalopeza is an extremely abundant, easily col­lected species in the small forest area that is located at Paliku at the eastern end of the vast, arid crater of the dormant volcano Haleakala on the island of Maui. This is the only locality from which the species has ever been collected. Adults have been reared from Cheirodendron leaves collected at this site. Comments. Hardy ( 1965) established the subgenus Trichotobregma in Dro­sophila for the single species petalopeza. The subgenus is distinguished by the presence in the male of 7 or 8 long reclinate hairs, instead of the normal proclinate and reclinate orbital bristle pattern. Interestingly, the distinct and repeated up and down movement of the male's body as he stands at head-under-wing posture, with his head in firm contact with the under surface of the female's wing vane, causes the orbital and surrounding area of his head to slide back and forth against the female's wing surface. Thus the subgeneric characters seem to be correlated with a specific element of the mating behavior. Because of the presence of the fore basitarsal bristles and the general over-all nature of the courtship pattern of petalopeza, I have grouped it with the bristle tarsi species group. Its divergence from its relatives involves coordinated func­tional and structural modifications which are directly involved in the courting behavior of the male. Summary On the basis of the present data, it would seem that the basic courtship pattern of the bristle tarsi species consists of a stationary wing display and the use of the fore tarsi to strike or grasp the female's abdomen, with the basitarsal bristles being brought into contact with her genital area. Modified Mouthpart Species Group Introduction This group is characterized by the modification of the posterior and lateral setae of the male's labellar lobes into sclerotized structures that are employed for grasping the female's genital area during courtship. These modifications range from those shown by D. hirtitarsus-like where only 10 to 12 setae on each labellar lobe are enlarged and elongated and each retains its normal setal shape, to the heavy armament type structures displayed by D. mimica, D. eurypeza, and others. The various species live in a wide gamut of forest conditions, ranging from the relatively dry forests at Bird Park and Kipuka Ki, Kilauea (Hawaii) with a rainfall of 60 to 64 inches annually to the wetter areas such as at W aikamoi (Maui) with 200 or more inches of rain annually. Some species such as hirti­ tarsus-Iike have been bred from fungi; others such as mimica, from the fallen The University of Texas Publication rotting leaves of Haimerliodendron bruonianum. Adults have been observed feeding upon fungi and the surface of rotting leaves. Drosophila aquila-like Courtship. The male taps, then positions himself directly in front of the female, facing her with about 2 mm. of space between them. He then spreads both wings 10° to 15° laterally, with vanes parallel to the substrate. Quickly, he then de­presses his head and elevates his abdomen until the longitudinal axis of his body is essentially perpendicular to the substrate. This "standing on head" position is maintained for 1 to 2 seconds, after which the male drops to a standing position and moves very rapidly to the rear and head-under-wing position. He extends his fore legs and rapidly strikes the dorsolateral sides of the female's abdomen with downwardly directed movements of the tarsi, at the same time making stabbing-grasping movements of his modified proboscis against the female's genital area. Occasionally, he was observed to stand some distance behind the female, i.e., 0.125 to 0.5 cm., repeatedly extending and retracting his proboscis with a downward and forward movement, and simultaneously "pawing" the substrate with one or the other of his fore legs. A non-receptive female decamps, kicks and engages in what appears to be extruding; at least if the male is not positioned directly behind a female, she curls the tip of her abdomen toward his face. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. One male and 3 females collected 21.X.64 at Waikamoi, Maui. Field Observations. All specimens were collected by sweeping under bracken fems and over leaf litter. Comments. This small, dark species has heavily infuscated wings. The extra­ordinary standing-on-head posture apparently serves to bring the entire wing surfaces into the visual field of the female. The male courted readily all three of the females, and the courtship sequence was repeated many times. Drosophila comatifemora Hardy Courtship. The male taps, goes to the rear and assumes a head-under-wing posture. He then spreads both wings outward 10° to 20°, elevating the costal margin so that the vanes form a shallow V when viewed in cross section. At the same time he raises and slightly curls the abdomen upward so that the abdominal tip is above the level of the wing vanes, and extrudes the posterior portion of the rectum to form a small ( 1 mm.) papilla-like structure. Simultaneously he extends his highly modified proboscis and firmly grasps the tip of the female's abdomen, typically the anal papilla. The labellar setae of the males are modified into chisel­shaped structures, which combine to form a single longitudinal knife-like struc­ture (row) on each lobe; these two parallel blades are used to grasp the female. Having achieved a firm grasp, the male then extends both legs forward and under the female's abdomen, and drums rapidly against her abdominal venter. At the base of each femur, on the dorsal surface, there arises a clump of long setae, and the drumming action causes these setae to strike against the female's genital area. If the female attempts to decamp before the male has achieved his labellar grasp, he will circle to the front of her, flick or snap both wings and extend and retract his proboscis toward her face, thus attempting to prevent her escape., A non-receptive female decamps, kicks, and depresses. The last action prevents the male from grasping the anal papilla, but instead he may then grasp the dorsum of the female's abdomen with his labellum, at the same time encircling it with his fore legs. Males court each other freely and often will attempt to copulate with each other. Copulations. None observed. Aggressive Behavior. The males are extremely aggressive toward one another. In addition to general jousting with their legs, they engage in vigorous curling. Study Specimens. Numerous. Collected 1-7.X.64 and 21.X.64, Waikamoi, Maui. Field Observations. Adults can be captured in the W aikamoi region by sweep­ing over the leaf litter and in the low vegetation in areas where lobeliads of the genus Clermontia are abundant. Comments. The males are aggressive suitors. When sexually aroused, they will continue to court for long periods of time (several hours) and direct their atten­tion not only to the females but also to each other. Often a male, standing alone near a courting pair, will engage in wing spreading and abdominal curling. Drosophila eurypeza Hardy . Courtship. The male taps, goes to the rear and assumes the head-under-wing position. He then extends his proboscis downward and forward under the tip of the female's abdomen and grasps her vaginal plates with his heavily sclerotized, spiny, labellar lobes. Having achieved the oral grasp upon the female, the male raises his abdomen slightly, compresses and shortens it along the longitudinal axis and then flick-waves both wings out and upward with one vane commencing the movement slightly ahead of its opposite. These flick-waving motions usually occur in bursts of two's with an intervening period of inaction. Finally the male extends both fore legs forward under the female's abdomen with the tips of the tarsi reaching the abdominal thoracic junction of the female. His legs then rotate so that the anterior surfaces face upward and are vibrated rapidly against the venter of the female's abdomen. If she is receptive, the male releases his oral grasp, curls his abdomen forward and achieves genitalic union. He then pushes himself forward upon the female, forcing her wings upward and apart and grasp­ing her abdomen with his fore and mid legs. At the end of copulation he slides off the female and shows no sign of a trance. Non-receptive females repel the males by decamping and kicking. If the male has achieved the oral grasp, the kicking of the female will cause him to loosen his hold and retract his proboscis, but, as soon as the female stops kicking (although often this is continued for some time by the female in the form of abdomen and wing cleaning motions), the male will regrasp the female. A non-receptive female engages in violent kicking whenever the male engages in vibrating his fore legs The University of Texas Publication against her venter. This invariably causes the male to retract both legs and pro­boscis. Typically non-successful courtships are terminated by the female decamp­ing, although males will cease courting after long unsuccessful efforts. No male-to-male actions were observed. Perhaps they were masked by the normal wing flick-waving that all individuals constantly engage in as they move about in the rearing cage or bottle. Such motions have also been recorded in the field. Copulations. Two. Aggressive Behavior. No distinguishable aggressive actions have been observed. Study Specimens. Laboratory stock W (H) 31.14 derived from a single female captured at Kumuwela, Kauai. Field Observations. D. eurypeza Hardy appears to be a relatively rare species, and is only infrequently captured. All collection records are from the island of Kauai to which the species is apparently restricted. Comments. The peculiar wing actions that the specimens engage in at all times set this species apart from all others. The use of the male proboscis in courtship is similar to that exhibited by other modified mouthpart species. The acceptance response by the female was not visibly observable and probably is transmitted to the male via his proboscis. Drosophila hirtitarsus-like Courtship. The male taps, then standing at the side or front of the female he extends one fore leg up and outward until the tarsal tip is elevated above the level of his head. The extended leg is then waved (sldwly vibrated) up and down, the amplitude of the movement being such that the tarsi will strike or almost strike the substrate at the bottom of the downstroke. Simultaneously, the mesothoracic leg of the opposite side is partially extended and waved up and down but at a slower rate than the fore leg. The leg waving occurs in bursts of 1 to 2 seconds' duration, and then both legs are returned to the resting position. In the interval between leg-waving bursts of action, the male rapidly flicks both wings horizon­tally and simultaneously hops laterally 1 to 2 mm. After a period of leg waving interspersed with wing flicking and body hopping, the male circles to the rear, assumes head-under-wing posture, extends his proboscis, and grasps the anterior end of the female's vaginal plates with the tips of his modified labellar setae. Retaining his grasp, he then extends his fore legs under the female's abdomen and drums her venter with his tarsi, thus bringing the long male fore tarsal setae against her body. Non-receptive females repel the males by decamping and kicking. Males court each other readily and seem unable to distinguish another male from a female. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. Numerous. Adults reared from larvae collected 7-14.IX.64 at Kipuka Ki, Kilauea, Hawaii. Field Observations. The species utilizes a mushroom type of gill fungus as a breeding site. All of the study specimens were bred from such a fungus collected from a dead limb of Sapindus sa1xmaria (seep. 299) . Comments. The species belongs to the modified mouthpart group, but the degree of modification is slight. On the posterolateral surface of each labellar lobe are located two lightly sclerotized longitudinal bands, each of which bears 5 to 6 long, heavy setae. These setae (2 groups on each labellar lobe) form the grasping organ of the proboscis and it was possible to observe clearly that only the distal ends of these setae are actually brought into contact with the female genital plates during the grasping action. Drosophila ischnotrix-like Courtship. The male taps, places himself directly in front of the female and raises both wings upward to almost 90° with vanes rotated so that they are parallel with the anterior margin facing forward. The male, after a short period of dis­play, returns his wings to the resting position, circles to the rear of the female and assumes a head-under-wing position, then extends his proboscis and grasps the female's genital area with his labellum. This grasp is maintained while his abdomen is curled laterally so that the tip is directed forward and reaches the posterior margin of his metathorax. Occasionally, he will also elevate one wing, rotating it so that the anal margin is directed posteroventrally. Periodically he abandons the posterior position, circles to the front of the female and arcs back and forth in front of her head, frequently raising both wings to almost 90° with vanes parallel. Typically these arcing movements are made 1 to 2 mm. away from the female, but occasionally he approaches near her and strikes her head with his fore legs. After posturing for a period in front of the female, he then returns to the rear and engages in the courting pattern as a hove. The non-receptive female appears to ignore the male's courting actions and also avoids him by decamping. Copulations. None observed. Study Specimens. One male and 2 females collected at Pupukea, Oahu, by Carson ?.VIl.63. Comments. Although the male courted vigorously and persistently, the limited number of individuals involved leaves some doubt as to whether the full female non-receptive repertoire was observed. Drosophila mimica Hardy Courtship. The male taps, goes to the rear of the female and assumes head­under-wing posture. The male's wings are raised and rotated slightly so that the costal edges are elevated 10° to 15° while the anal margin is raised only 0° to 5°. In cross sectional view, the wing vanes thus form a shallow V. The male then extends his proboscis and grasps the female's genital region. Non-receptive females decamp, kick and depress the tip of the abdomen close to the substrate. Males tap each other but do not proceed with courting actions. Copulations. None observed. The University of Texas Publication Aggressive Behavior. When a number of individuals are crowded together, the males are often extremely aggressive toward other individuals. The typical ex­pression of aggression is curling. The curling action usually is atypical in that the abdomen is extended and elongated but at the same time it is arched laterally only to a small degree. The aggressive individual places himself so that he is facing away from his adversary and then rams backward, using the tip of the abdomen to butt or strike against the other fly. Often two males will be seen butting each other in such fashion for several seconds until one of the individuals is defeated and flees. Males also will display aggressive behavior toward non­receptive females that they have been unsuccessfully courting. Study Specimens. Numerous individuals, both adults captured in the field in the vicinity of Kilauea, Hawaii, and specimens from laboratory stocks which had originally been founded by flies captured at Kilauea, Hawaii. Field Observations. D. mimica is one of the most abundant species in Bird Park (Kipuka Puaulu), Kipuka Ki, and the surrounding vicinity. The adults can be captured by sweeping over the forest floor, especially in areas where Haimer­liodendron bruonianum is abundant. In many areas H. bruonianum forms dense thickets. In such places the large broad leaves are continuously being abscissed from the trees and thus form a thick carpet on the forest floor. It is in this leaf litter that mimica spends its life cycle. The adults not only feed upon the rotting leaves, but also lay their eggs upon them. The larval life is spent in rotting masses of Haimerliodendron leaves and the pupal stage in the soil of the same area. Unlike the larvae of those species which live inside Cheirodendron leaves, mimica larvae appear to inhabit both the inside and outside of the rotting Psychotria leaves. The adults spend most of their time hidden in the leaf litter, but any disturbance, such as net-sweeping about the litter or physically moving the leaves, will cause the flies to emerge and fly about. They quickly alight on the surface of the leaf litter and after a few minutes disappear by crawling under the leaf surfaces, especially into the moister areas. Comments. Despite many hours of observation, both in the laboratory and in the field, using different types of observation cells, different light intensities and making the observations at various times of the day and night, only a limited number of courtships and no copulations were observed by the author. This species breeds well under laboratory conditions, but to date only one copulating pair has ever been seen (Wheeler, personal communication) in a laboratory stock. Possibly courtship occurs only in the dark, or at least under light intensities that would make it impossible to observe under laboratory conditions or such as would exist in the deeper parts of the leaf litter. .. Drosophila tendomentum Hardy Courtship. The male taps and, if standing at the side of a female, immediately "cleans" his tarsi by drawing them between the labellar lobes of the proboscis. After "cleaning," the male assumes a normal standing stance except that the fore legs are drawn close together and parallel to the median plane. Simultaneously, both fore legs are then stamped (tapped) against the substrate and the male's proboscis is extended and retracted repeatedly in the direction of the female. The fore tarsi are repeatedly "cleaned" (moistened?) during a courtship bout. After a short display, themale moves to the rear, assumes the head-under-wing posture, spreads both wings 15° to 20° from the resting position, and infrequently attempts to lick or grasp the female's genital area with his highly modified mouthparts. Infrequently, the male will either commence the courtship in front of the female or leave the head-under-wing position to circle forward until he is directly in front of her with his face almost touching her face. He then elevates his body, curls his abdomen under and forward so that the tip reaches the level of his mesothorax and extends both wings flat to 40°. With a slight rippling motion, the wings are then extended and rotated to 110° with the vane vertical to the sub­strate. This extended wing position is held for a fraction of a second and returned to the 40° position. If the female attempts to decamp, the male will arc back and forth in front of her, trying to keep her from escaping and keeping himself directly in front of her. From the front posture he will circle to the rear and assume a head-under-wing posture. In doing so, he may posture at the side of the female as described a hove before finally reaching the rearposition. Non-receptive females decamp, kick and depress. The slightest depression of the female's abdomen is adequate to prevent a male from attempting to lick or grasp the female. Males do not court each other. Copulations. None observed. Aggressive Behavior. None observed. Study Specimens. Six males and 8 females reared from larvae collected 7-14.X.64, Kipuka Ki, Kilauea, Hawaii. Field Observations. The 14 specimens were reared from a gill fungus collected at Kipuka Ki. This mushroom type fungus (unidentified) was found growing on a large, dead soapberry tree limb (Sapindus saponaria) which had been broken off by the wind. Originally the fungus had been growing at a point about 50 to 55 feet above the ground. The limb had a number of fruiting individuals of the fungus living on its surface, and at least six species of Drosophila were observed feeding on these; from the fungal material brought into the laboratory several species were reared, including tendomentum. Comments. This species portrays well the difficulties connected with the study of the sexual behavior of the Hawaiian drosophilids. All specimens were sepa­rated as to sex as soon as they had eclosed from the pupal stage during the first two weeks in October of 1964. There was therefore no doubt as to the fact that the females were virginal at the time the courtship observations were conducted. Repeated observations were made on different days with a total observation time of several hours. At the end of each observation period, the sexes were separated and returned to separate food vials. All individuals appeared healthy and vigor­ous. The males courted readily and persistently, but no female ever gave any indication of an acceptance response to their overtures. The species belongs to the modified mouthpart group; the labellar setae are greatly enlarged and heavily sclerotized, and function as a complex grasping mechanism when the proboscis is extended. Unlike most species of this type, the distal portions of the wings of the males are pigmented and interestingly, unlike the clear winged species, the wings are utilized in the courtship pattern. The University of Texas Publication Summary The modified mouthpart species form a highly diverse group insofar as court­ship patterns are concerned. The only element that is constant is the use of the modified mouthparts to grasp or lick the female's genitalia or some other part of her abdomen. Wing display always involves both wings, usually for a station­ary type of display. White-tip Scutellum Species Group Introduction A member of this complex is characterized by its slender body, low body pro­file, white to yellow tip of the scutellum, and the black, heavily sclerotized lateral border on each labellar lobe. D. fungicola is aberrant in that it lacks this labellar border. The larvae are fungivores and the adults can often be found feeding upon fungi. The members of the group are quite distinct from all other Hawaiian species, not only with respect to their anatomy but also with respect to their mating behavior. The male courtship is essentially Scaptomyza-Iike. In addition to fungicola, another undescribed species was studied intensively in the field and both conform to the same behavioral pattern (see p. 299). The eggs of these species have short filaments and are very similar in appearance to typical Scap­tomyza eggs. The eggs c.re only partially thrust into the body of the fungus or laboratory food by the ovipositing females, unlike many Hawaiian Drosophila which imbed their eggs deeply into the food. Drosophila fungicola Hardy Courtship. The male fixates the female and positions himself in front of her, raises both wings up and outward 45 ° to 50°, holds this position for a second or less, then returns his wings to a resting position and circles rapidly to the rear, grabs the female's abdomen with his fore legs, and curls the tip of his abdomen ventrally. These actions are all done rapidly and are similar to those exhibited by Scaptomyza species. Non-receptive females repel the males by the same type of actions as those displayed by Scaptomyza females, i.e., by decamping, kicking, violent shaking of the body, and by wing :movements. Males court each other as readily as they court the females. A male-to-male courtship results in violent aggressive actions between the two individuals which always terminates with one of the individuals fleeing. Copulation. One copulation was observed in the laboratory while studying flies that had been captured in the field at Kipuka Ki, Kilauea, Hawaii. During copu­lation the pair was extremely quiet, despite constant physical contacts with other flies moving about in the observation cell. The male fore tarsi held on to the top of the base of the female's spread wings, with the hind legs on the substrate. Copulation lasted more than 6 minutes and was terminated by the kicking action of the female. The male did not enter into a trance during the copulation, at the end of which he jumped off the female. Aggressive Behavior. The males of these slender, alert appearing, very active flies are aggressive. They curl as well as fend with their fore legs, and engage in vigorous pushing contests, often accompanied by rapid wing vibrations. Study Specimens. In addition to numerous field-captured adults, 40 individuals (20 males and 20 females) reared from larvae collected in the field were studied. These virginal specimens were 18 days old, but engaged in no courtships even though they seemed healthy and active. Field Observations. These flies were observed feeding on and ovipositing in fungi, especially Polyporus sulphurea. On the food masses the individuals are quiet and cryptic. The males appear to spend much less time feeding than do the females; rather, they take station on the vegetation in the vicinity, i.e., 5 to 20 feet from a feeding-ovipositing fungal site. They prefer large, thin leaves of shrubs such as the mamaki (Pipturus hawaiiensis) and the pilo (Coprosoma spp.) at a height of 2 to 8 feet above the ground. Typically, they sit upon the upper surface of the leaves, although the under surfaces may be used. The fly invariably posi­tions itself, usually at the tip of a leaf, so that it can visually scan the area about and below it. It appears that the "viewing value" of a particular site is the critical factor, and this rather than the upper or lower surface of the leaf determines where a male takes station. Because of the aggressiveness of the males toward each other, unless the leaf is large (e.g., 3 or more inches in diameter) only one fly is found per leaf. In an area where the flies are abundant and a fungus is present, most of the leaves, especially those on the tips of the branches facing the fungus, will each have a fly sitting on them. By lying or sitting on the ground and looking upward, one can see the silhouettes of numerous male heads extending just over the edges of the leaves. The fly does not stay long on a single station. Rather he will fixate upon another individual nearby and fly to that leaf. Sometimes he will land directly upon the other fly, but usually he lands 1 to 2 inches away. The fly thus approached will turn to meet the intruder, and the intruding male will then engage in the wing-raising phase of courting and approach the other individual. If the approached fly is a male, there then ensues a fight which invariably results in one of the two fleeing to another leaf. Although no records were kept, it ap­pears that about half of the fights result in the intruder being defeated. Despite a number of courtships having been observed, no copulations were seen in the field. It is suggested that when sexually receptive females leave the feeding-ovipositing sites they are intercepted by the waiting males when they alight on the surrounding vegetation, and that courtship and copulation then take place. The larval life is spent in the fungus and the larvae, when mature, emerge on the fungal surface, skip off and fall to the ground. Mature fungicola larvae are extremely resistant to desiccation. They immediately burrow into the ground and do not pupate for several days (5 to 20) after they have left the fungus. During this period they appear to burrow about in the soil. Such behavior apparently results in scattering the pupal population, thus preventing a great concentration of the pupae in a smaH area in the immediate vicinity of the larval food site. Comments. The courtship and copulation of this species are Scaptomyza-like in all respects. The male action of "taking station" on a leaf is typical of a number The University of Texas Publication of Hawaiian species. For example, D. imparisetae males are sometimes found with fungicola males on adjacent leaves of a single branch of a shrub. Interest­ingly, males of both species appear to sleep at night on the under surfaces of the leaves that they use as visual stations. In comparison, the females of fungicola are always collected either from fungi or by sweeping the vegetation close to the ground, especially under bracken ferns. The aggressive behavior displayed by this species is a mixture 0£ drosophilid and Scaptomyza-like actions. It should be noted that tapping does not seem to be a mandatory part of the courtship pattern of the male. Under laboratory conditions, tapping was observed to occur only occasionally, but in the field it was not observed. This also indicates a Scaptomyza-like type of behavior. Scaptomyza-like Species Introduction Three species, D. crassifemur, D. nasalis and D. parva, display a typically scaptomyzoid pattern of mating behavior. Hardy ( 1965), on the basis of external anatomy, has placed these three species in the genus Drosophila. I have followed his classification and they are therefore included in this study. Note should be taken that the courting pattern of these species is similar to that of D. fungicola which belongs to the white-tip scutellum complex of Drosophila. Drosophila crassifemur Grimshaw Courtship. The male positions himself at the rear or side of the female, taps with a slashing movement of the fore legs, and then immediately attempts to grasp the female with his fore legs. If at the rear he usually but not always reaches under the female's wings. When the female has been thus grasped, he attempts to pull and push himself into copulatory position, i.e., with her wings pushed apart and with his fore legs grasping the anterolateral surface of her abdomen or her thorax and with his meso-and metathoracic legs along the side of her abdomen. The male then curls his abdomen under that of the female so that its tip reaches anteriorly to her genitalia. He then draws his genitalia back­wards across those of the female, attempting to achieve intromission. If he fails, the abdominal sequence is repeated. A non-receptive female extends the tip of her abdomen posteroventrally, at­tempting to prevent the male from achieving genitalic contact. She also extrudes although the extent of such extrusion appears small. Neither of these actions succeeds in preventing the male from achieving genitalic contact, but even so copulation does not occur unless the female accepts the male's overtures. Even­tually he dismounts from non-receptive females and will then court another individual. Males court each other in exactly the same fashion as that used in courting females. There seems to be no definite countersignaling between males. Copulations. One observed but not timed. Aggressive Behavior. None except to fend with mesothoracic legs in typical puli, a species collected from cot­tonwood trees near Anchorage, Alaska (Wheeler and Throckmorton, 1960) . In Hawaii this type has so far been seen only from the Drosophiloids. It is always DROSOPHILOIDS Genus: IDIOMYIA .1 obscuripes .2 perkinsi .3 picta Genus: ANTOPOCERUS .4 aduncus .5 diamphidiopodus .6 longiseta . 7 orthopterus Genus: NUDIDROSOPHILA .8 aenicta Genus: DROSOPHILA Miscellaneous .9 anomalipes .10 caccabata .11 hirtitibia .12 imparisetae .13 truncipenna Picture wings .14 adiastola .15 crucigera .16 engyochracea .17 fasciculisetae .18 grimshawi .19 musaphilia .20 picticornis .21 pilimana .22punalua .23 villosipedis Modified mouthparts .24aquila .25 araiotrichia .26 asketostoma .27 comatifemora .28 conjectura FIG. 1. Spermathecae .29 dissita .30 eurypeza .31 flavibasis .32 freycinetiae .33 hirticoxa .34 involuta .35 ischnotrix .36kauluai .37 mimica .38 mycetophila .39 residua .40 scolostoma Bristle tarsi .41 apodasta .42 basimacula .43 expansa .44 perissopoda .45 (T.) petalopeza .46 prodita .47 redunca .48 seclusa .49 torula .50 trichaetosa Spoon tarsi .51 conformis .52 contorta .53 disticha .54 incognita .55 neutralis .56 polliciforma .57 sordidapex Split tarsi .58 ancyla .59 fundita .60 pectinitarsus The University of Texas Publication i!P ~ ~ pigmented, either dark brown or black, and the spermathecal envelope is thick and uniformly distributed over the capsule. At its opening into the spermatheca the spermathecal duct may flare widely (Figure 1.16) or not (Figure 1.10). This character of the spermathecal duct shows no indication of phylogenetic pattern, although the flared type is more common in the more derivative Hawaiian groups (Idiomyia, etc.) . The second type of spermatheca is subspherical to quadrate in outline, and the introvert extends inward half to three-quarters of the diameter of the capsule. Only one Drosophiloid species (D. anomali-pes, Figure 1.9) has this type of sperm­atheca. It is far more common among the Scaptoids (Figure 2) and is, indeed, the typical Scaptomyza spermatheca. Its color varies from dark brown to pale yellow. The spermathecal envelope is heavy apically but in many species it is thinner toward the base of the capsule. The third major type of spermatheca is prominent among the Scaptomyza, and it is also characteristic of the "white tip scutellum" forms. The spermatheca in this case lacks an introvert, and there is no evidence that this represents a regressed or degenerate form. This is, substantially, the primitive spermatheca of the Drosophilid stem (see Throckmorton, 1962). Gen­erally the capsule of this spermatheca is very weakly sclerotized and not pig- FIG. 2. Spermathecae (Numbers refer to specimens listed in the Appendix) DROSOPHILOIDS Genus: DROSOPHILA White tip scutellum .1 cilifemorata .2 fungicola .3 haleakalae .4 iki .5 melanoloma .6 melanosoma .7 nigra .8 bipolita .9 canipolita .10 demipolita SCAPTOIDS Genus: DROSOPHILA .11 crassifemur .12 nasalis .13 parva Genus: TITANOCHAETA .14 contestata .15 #8 Genus: SCAPTOMyzA Subgenus: Alloscaptomyza .16 longisetosa .17 stramineifrons Subgenus: Bunostoma .18 anomala .19palmae .20 xanthopleura Subgenus: Exalloscaptomyza .21 mauiensis .22 #165 (Hawaii) .23 #166 (Oahu) .24 #168 (Kauai) .25 #167 (Molokai) Subgenus: Parascaptomyza .26 pallida Subgenus: Tantalia .27 varipicta Subgenus: Trogloscaptomyza .28 argentifrons .29 articulata .30<:onnata .31 hackmani .32 inaequalis .33 intricata .34 latitergum .35 levata .36 retusa .37 rostrata Representatives of spermathecae and parovaria .38 Antopocerus longiseta .39 Drosophil.a musaphilia (picture wing) .40 D. imparisetae (misc.) .41 D. hirtitibia (misc.) .42 D. fungicola (white tip scutellum) .43 D. nigra (white tip scutellum) .44 D. crassifemur (Scaptoid) .45 D. nasalis (Scaptoid) .46 Scaptomyza argentifrons .47 S. retusa The University of Texas Publication mented. In appearance there is very little to distinguish these spermathecae from parovaria. Spermathecae of this type are shown in Figures 2.1 to 2.10. The spermathecae shown in Figures 2.14 and 2.'21 to 2.25 are not of this last type although they may superficially resemble it. They represent regressions from a fully introverted type. The introvert remains as a well sclerotized struc­ture, but the capsule itself is only a shrivelled remnant and hardly sclerotized at all. The form shown in Figure 2.25 is probably near the type these were derived from. It is very weakly sclerotized and collapses readily in phenol, where most specimens evert and give the appearance of those shown in Figures 2.21 to 2.24. However, it is just possible to determine in untreated specimens that the form shown in Figure 2.25 is normal in the living individual. Some of the spermathecae of Scaptoid species have another character not shared with the Drosophiloids. Generally the spermathecal duct is regularly annulate from the base of the capsule to a point above where it enters the vagina (Figures 2.38-.43, etc.) In several of the Scaptoids the annulae do not start immediately below the capsule. Instead there is a continuation of the smooth surface of the spermathecal duct from within the introvert. This may extend a distance almost approximating the diameter of the capsule before the annulations start (Figures 2.11, .28, .35, etc.). Both D. crassifemur and D. nasalis have the Scaptomyza spermatheca, and D. crassifemur (Figure 2.11) also has this smooth apical portion of the spermathecal duct. Parovaria--Some general features of the spermathecae and parovaria are shown in Figures 2.38 to 2.47. The structures are shown as they appear under low magnifications of the compound microscope after clearing in phenol. As a rule the ducts of parovaria are undifferentiated. This is true for all of the new world species of Drosophila that I have examined and, to my knowledge, no dif­ ferentiation of this structure is reported elsewhere. However, many of the Hawai- DROSOPHILOIDS Genus: IDIOMYIA .1 obscuripes .2 perkinsi .3 picta Genus: ANTOPOCERUS .4 aduncus .5 diamphidiopodus .6 longiseta .7 orthopterus Genus: NUDIDROSOPHILA .8 aenicta Genus: DROSOPHILA Miscellaneous .9 anomalipes .10 caccabata .11 hirtitibia .12 imparisetae .13 truncipenna Picture wings Fw. 3. Ventral receptacles .14 adiastola .15 crucigera .16 engyochracea .17 fasciculisetae .18 grimshawi .19 musaphilia .20 picticornis .21 pilimana .22punalua .23 villosipedis Modified mouthparts .24aquila .25 araiotrichia .26 asketostoma .27 comatifemora .28 conjectura .29 dissita .30 eurypeza .31 flavibasis .32 freycinetiae The University of Texas Publication ian species have this duct expanded and sclerotized as shown, for example, in Figure 2.38. The more usual condition for Drosophila and its related genera is shown in Figure 2.41. As for so many of the characteristics of the Hawaiian spe­cies, this character is distributed in a rather random fashion. At least some species from all the major groups (which includes all of the informal groupings) of Drosophiloids have it, although its expression among the "white tip scutellum" flies (Figure 2.43) is somewhat atypical. It is doubtful that this differentiation is present in any of the Scaptoids, but some slight development of this region is seen in some species (Figures 2.45 and .46). Ventral receptacles-The ventral receptacles are shown in Figures 3 to 6. In all Hawaiian species the distal end of the ventral receptacle is completely free of the vagina. Generally there is a straight basal section, then a series of coils, and then a much-folded section. In some species (Figure 3.10) the short coiled section is absent. In others (Figures 5.15, .26, etc.) the coiled section is much expanded and forms a major part of the organ. In one species (D. anomalipes, Figure 3.9) the ventral receptacle lacks the folded section entirely. It has the coiled ventral receptacle that characterizes flies from the subgenus Drosophila from elsewhere in the world. The typical ventral receptacle of the Hawaiian species (Figure 3.3, etc.) may be considered as intermediate between the folded type seen in Pholadoris, Sophophora and Dorsilopha and the coiled type seen in the subgenus Drosophi.la. As inspection of Figures 3 to 6 will show, the ventral receptacles from Drosophiloid and Scaptoid species are substantially alike. The major types, and the ranges of variation within them, are approximately the same in both groups. Some of the species of "white tip scutellum" flies have ventral receptacles with exceptionally long straight sections basally (e.g., Figure 5.8), and no Scaptoid species has a true coiled ventral receptacle, although some approximate this (Figures 5.15, .26) . Otherwise the species are very much alike for this character. The figures show the ventral receptacle as seen with the com­pound microscope after clearing in phenol. The ovipositor-As a genus, Drosophila does not show conspicuous versatility in its ovipositors. Eggs are generally inserted just under the surface of the food DROSOPHILOIDS Modified mouthparts (cont.) .1 hirticoxa .2 involuta .3 ischnotrix .4 kauluai .5 mimica .6 mycetophila .7 residua .8 scolostoma Bristle tarsi .9 apodasta .10 basimacula .11 expansa .12 perissopoda Fw. 4. Ventral receptacles .13 (T.) petalopeza .14 prodita .15 redunca .16 seclusa .17 torula .18 trichaetosa Spoon tarsi .19 conformis .20 contorta .21 disticha .22 incognita .23 neutralis .24 polliciforma .25 sordidapex The University of Texas Publication and the ovipositor required for this is not very elaborate. Many of the Hawaiian Drosophilids have departed from this fashion. They may have unusually long and almost tubular ovipositors, or the ovipositors may be of the more standard type but still long, and heavily sclerotized with stout, peg-like teeth. The develop­ment of the ovipositor is associated with some internal modifications that often give to the Drosophiloid female a truncated, or blunt and heavy bodied aspect in lateral view. Figure 7.1 shows in lateral view the internal reproductive organs of the female. The most conspicuous feature is the pronounced development of the vagina and its associated muscles. For clarity the muscles have been omitted from the figure. As seen in figure 7.1 the vagina is very long. It folds anteriorly before turning ventrally to enter the ovipositor. In many species, particularly those having a long membranous ovipositor (figure 7.8), the inner sheath of the vagina moves freely within the muscular coat and has its wall supported by very tough, spiral cords. In the figure this inner sheath is shown pulled out somewhat. In consequence of this great development, the vagina itself is displaced from its usual position more or less parallel with the rectum. The muscles of the system arise from the inner surface of the last abdominal tergite. On each side one mass extends to the anterior part of the vagina and attaches to it in the region of the spermathecae and ventral receptacle. These muscles exert a force back and down, so the dorsal surface of the vagina is held closely in contact with the ventral sur­face of the rectum, and in many cases the nominal dorsal surface of the vagina is directed more posteriorly than dorsally. In consequence both the spermathecae and parovaria are forced to bend laterally around the anterior end of the vagina and the oviduct. The folded vagina also displaces the ventral receptacle, and it curves laterally (to the left) and dorsally. Generally the distal folded section of the ventral receptacle is attached by tracheae to the left spermatheca. A second set of muscles leads from the last abdominal tergite to the base of the ovipositor. In the relaxed state the ovipositor rests with its tip high (as in Figure 7). If this Fm. 5. Ventral receptacles (numbers refer to specimens listed in the Appendix) DROSOPHILOIDS Split tarsi .1 ancyla .2 fundita .3 pectinitarsus White tip scutellum .4 cilifemorata .5 fungicola .6 haleakalae .7 iki .8 melanoloma .9 melanosoma .10 nigra .11 bipolita .12 cani poli ta .13 demipolita SCAPTOIDS Genus: DROSOPHILA .14 crassifemur .15 nasalis .16 parva Genus: TITANOCHAETA .17 contestata .18 #8 Genus: SCAPTOMYZA Subgenus: Alloscaptomyza .19 longisetosa .20 stramineifrons Subgenus: Bunostoma .21 anomala .22palmae .23 xanthopleura Subgenus: Exalloscaptomyza .24 forms from Kauai, Oahu, and Hawaii .25 form from Molokai .26 mauiensis The University of Texas Publication Fm. 6. Ventral receptacles SCAPTOIDS .5 connata Genus: SCAPTOMYZA .6 hackmani Subgenus: Parascaptomyza .7 inaequalis .1 pallida .8 intricata S1.Jbgenus: Tantalia .9 latitergum .2 varipicta .10 levata Subgenus: Trogloscaptomyza .11 retusa .3 argentifrons .12 rostrata .4 articulata second set of muscles contracts the tip of the ovipositor is lowered. In many species (e.g., Figure 8.2) the proximal section of the ovipositor is modified and has a special extension for the attachment of these muscles. Not all of the Hawaiian Drosophilids show internal differentiation and the development of the ovipositor. This is seen to a greater or lesser degree in all Drosophiloids except the "white tip scutellum" species. Itis not seen in the Scap­toids that I have examined to date. Among the Drosophiloids that do show devel­opment there is considerable variation. Those species that would be considered most primitive on other grounds also have this system less conspicuously devel­oped. It is apparent that the evolution of this suite of characters has been directly correlated with adaptation to a variety of food niches and the exploitation of oviposition sites not generally accessible to Drosophila. Figures 7 and 8 show some examples of ovipositors of Hawaiian species. They are shown as they appear after clearing in phenol. Figures 7.6 or 7.10 show ovipositors that must be very near the primitive type from which all other Hawaiian types were derived. Only the left half, or the left valve, of the ovipositor is shown in the figure. In most cases the ovipositor consists of a pair of sclerotized valves joined together antero-ventrally by a narrow sclerotized bridge. Dorsally the valves are connected by membrane, and the posterior extent of this membrane varies from species to species. In many of the Hawaiian Drosophiloids this mem­branous connection extends almost to the tip of the ovipositor. Among the Scap­toids the connection may be virtually absent and the valves almost unattached to each other. Ventrally there is also a membranous connection that generally does not extend as far posteriorly as does the dorsal connection. This is a continuation of the inner lining of the vagina and its inner surface is covered with short, recurved spines. As the ovipositor is extended and an egg deposited this lining everts and in some species may form a tube-like extension almost ~s long as the ovipositor. The spines would then serve to hold this tube in position while eggs are laid. Not infrequently, at least in laboratory cultures, eggs may be laid in clusters of up to half a dozen or so, and the ovipositor is probably kept in position and not re-inserted for each egg. Figure 7.8 shows a part of this ventral lining. The terminus of the dorsal connection is shown in Figure 7.5. Not all ovipositors are distinctly composed of two valves. Sclerotization in many is reduced so that the ovipositor is entirely membranous. If one wished, an almost complete graded series of ovipositors could be arranged. At one extreme would be those with distinct, heavily sclerotized valves. In Figures 7 and 8 these are indicated by a solid line separating the upper and lower halves of the ovi­positor (Figures 7.3, .11; Figures 8.1-.3, etc.). In other cases the ventral margin of a valve may be sclerotized but the region of sclerotization may grade almost im­perceptibly into the membranous dorsal part. These have been indicated by dashed lines separating the dorsal and ventral halves (Figures 7.2, .5, etc.). Then there are those that are almost completely membranous (Figure 7.4; Figure 8.8, etc.) . These are tube-like and quite flexible. With a dissecting needle they can be turned inside out very readily. Most of the types in which the degree of sclerotization grades from a "strong" ventral margin to a weak dorsal margin exhibit this type of flexibility. Obviously, such ovipositors could not be used to insert eggs into a very firm substrate. They seem more suited to pushing eggs into crevices or into soft, porous materials. Among the Drosophiloids three major trends in ovipositor development may be detected. One is toward the elongate, membranous type of ovipositor just men­ The University of Texas Publication tioned. These are found primarily among Idiomyia, Nudidrosophila and the "picture wings." Incipient development in this direction is seen primarily among flies with modified mouthparts. A second line of development is toward a very strongly sclerotized ovipositor of very characteristic shape. This type is shown in Figure 7.3. It is found in some flies with modified mouthparts (Figure 7.11) but it is most characteristic of flies in the genus Antopxerus, and among the flies with tarsal ornaments in the male, particularly within the "bristle tarsi" group. Very often species in a group may have one or the other of two basic types of oviposi­tors. These two general types are shown in Figure 8 (Figures 8.4 and .5). Many times, two species will be collected at a given locality and be almost indistinguish­able for most of their characters. However, one will have the type of ovipositor shown in Figure 8.4, the other that shown in Figure 8.5. Presumably these reflect some niche separation between the two species. Precisely what each type of ovi­ positor is adapted for remains uncertain. The third trend in ovipositor development is seen among the flies with the white tip scutellum. Some of these (Figure 8.7) have the type usual for the Hawai­ian Drosophiloids, and others show evidence of the trend toward elongate, mem­branous ovipositors (Figures 8.6 and 8.8). Many, however, show a trend toward reduction. This trend is evidenced both in the reduction of sclerotization and in the reduction of size and loss of bristles (Figures 8.10 and .11). Some, of course, have their own peculiarities (Figure 8.9) and evidence independent divergence. The trend toward reduction of ovipositors is most pronounced among the Scap­toids. Here there are many peculiar forms, but a more or less graded series could be arranged for some species. Some Scaptomyza (Figure 8.:22) have the usual ovipositor with a pair of sclerotized valves and with a dorsal and a ventral range of bristles. Others have reduced sclerotization, the shape of the valve is changed, and the arrangement of bristles is less typical (Figures 8.14, .15, .22 and .24). The extreme reduction is seen in such forms as that shown in figure 8.25, where the valve is reduced to a small sclerotized plate and the bristles are almost all lost. Species of Scaptomyza from the subgenus Exalloscaptomyza have the most divergent types. Here the two valves seem to have fused together ventrally to form a single median structure. Some of these are shown in Figures 8.16 to 8.21. In each case both ventral and lateral views are shown. The most divergent type, found in the species of Exr.illoscaptomyza from the island of Molokai (Figures 8.20, .21), is Y-shaped, attached only basally, and is freely movable in a vertical plane. Figures 8.12 and 8.13 show the ovipositors of D. crassifemur and D. nasalis. They are obviously distinct from the usual Drosophiloid type. D. crassifemur Frn. 7.1. Internal genital system of the female shown in lateral view. The species figured is Drosophila hirtitibia. a.p.-anal plate; h.g.-hind gut; o.-ovipositor; ov.-ovary; ovd.-oviduct; p.-paragonium; r.-rectum; r.p.-rectal papilla; s.-spermatheca; v.-vagina; v.r.-ventral receptacle. The remainder of the drawings are of the left side of DROSOPHILOID ovipositors . .2) ldiomyia perkinsi, .3) Antopocerus orthopterus, .4) Nudidrosophila aenicta, .5) Drosophila caccabata, .6) D. imparisetae, .7) D. fasciculisetae (picture wing), .8) D. engyochracea (picture wing), .9) D. asketostoma (modified mouthparts), .10) D. comatifemora (modified mouthparts), .11) D. dissita (modified mouthparts). The University of Texas Publication ~.21~ Q .24~ ~ has an ovipositor with two valves (Figure 8.12). These are each slender and elongate and attached only basally. Each is covered with much elongated bristles, among which the typical dorsal and ventral ranges of bristles can be homologized without too much difficulty. D. nasalis has an ovipositor that is much more Scaptoid in conformation. Sclerotization is much reduced and the valve is more or less fleshy. In addition to the few remaining bristles, the apical part of each valve is covered with many short hairs (Figure 8.13). Eggs-Figures 9 and 10 show eggs of Hawaiian species. There is an unusual variety of types. In order to represent the important features most clearly, each egg is shown in lateral and in ventral view. The egg filaments have been omitted from the ventral view (to the left in each figure) . Among the Scaptoids the number of egg filaments varies from zero to four. Among the Drosophiloids the number varies from two to four. The character of the filaments is quite different between these two major groups. Among the Scaptoids the filaments are short and heavy. In many cases they are more like small lobes than like filaments (e.g., Figure 10.25). In some cases the posterior filaments are simply anterior extensions of a pair of heavy white ridges that extend along the ventral surface of the egg (Figures 10.19, .83, .27). Among the Drosophiloid species there are two major types of eggs. One is seen among the "white tip scutellum" species, the other among the remainder of the Hawaiian Drosophiloids. "White tip scutellum" flies have eggs with either two or four very short filaments (Figures 10.1-.8). These filaments are slender, as is typical for Drosophila filaments. In those having only two filaments (Figures 10.6-.8) these are very fine, very short, and set close together near the ventral margin of the opercular area. The remainder of the Drosophiloids have eggs with four long filaments. Some of these (Figure 9.5) appear to be identical to the type seen from non-Hawaiian species of the subgenus Drosophila. Others are quite different. Generally the anterior and posterior filaments differ in length, with the posterior filaments being the longer. In many instances the posterior filaments are very much longer than the anterior (Figure 9.2, .6, .7, .13). In other cases both the anterior and posterior filaments are very long (e.g., Figure 9.9). Species having the long, membranous ovipositor lay the eggs with the exceptionally long fila­ments. Two distinct patterns of egg chorion can generally be distinguished. In one, typical for most Drosophila, the surface is finely and uniformly sculptured. In the other there are distinct longitudinal striations. These two patterns are shown in the figures by stippling for sculpturing, by lines for striae. In some cases the FIG. 8. Ovipositors DROSOPHILOID: bristle tarsi: .1) Drosophila apodasta, .2) D. basimacula, .3) D. torula; spoon tarsi: .4) D. conformis, .5) D. incognita; white tip scutellum: .6) D. fungicola, .7) D. iki, .8) D. melanolorruz, .9) D. nigra, .10) D. canipolita, .11) D. demipolita. SCAPTOID: .12) Drosophila crassifemur, .13) D. nasalis, .14) Titanochneta contestata, .15) Titanochaeta sp. #8, Scaptomyza (Exalloscaptomyza) species, .16) from Hawaii (ventral), .17) from Hawaii (lateral), .18) from Oahu (ventral), .19) from Oahu (lateral), .20) from Molokai (ventral), .21) from Molokai (lateral), .22) S. (Trogloscaptomyza) latitergum, .23) S. (T.) retusa, .24) S. (T.) rostrata, .25) Scaptomyza species. 354 The University of Texas Publication ' ' ' ' .30 .31 .32 .33 .34 .35 .36 ~' two patterns are mixed, i.e., eggs show both striae and sculpturing. In some the striae are ventral, the sculpturing dorsal (Figures 9.Z, .6), in some the reverse (Figure 9. tz) and in others both are intermingled (Figure 9.18). Both Dro­sophiloid and Scaptoid eggs show these features. In some cases the chorion is almost completely devoid of character. It seems that eggs retained for some time in the vagina tend to loose the features of the chorion mentioned above. For example, the species of Exalloscaptomyza lay very large eggs (Figure 10.17) and, occasionally, larvae. Apparently the female can retain the egg for a rather long period until she finds a suitable oviposition site. During this period the egg may hatch and the larva be retained in the vagina. If newly collected females are dissected there is usually an egg or a larva in the vagina, one nearly mature egg in one ovary, and no egg at all in the other ovary. Females taken from culture generally have no egg in the vagina, suggesting that they lay the egg as soon as it is mature when a suitable oviposition site is available. Eggs taken from culture are very weakly sculptured. Those found in the vagina rarely show any evidence of sculpturing. The same seems to be true of D. nasalis (Figure 10.10), although I have dissected only a few of these females. At least some species of Titanochaeta also follow this pattern. The egg of one undescribed form from this genus is shown in Figure 10.1Z. When a female from another species, T. contestata, was dissected she was found with a larva in the vagina. Titanochaeta are parasitic on spider egg cases, but the ovipositor of this species (Figure 8.14) is not well suited for inserting an egg into a spider egg case. The larva in the vagina of the female was perfectly healthy and vigorous. Its mouth hooks were developed to form a single median stylet that was very long and sharp. When freed from the vagina Fm. 9. Characteristics of eggs. For each form a ventral (left) and lateral (right) view is shown. The egg filaments have been omitted from the ventral view. DROSOPHILOID Genus: ANTOPOCERUS .1 diamphidiopodus Genus: DROSOPHILA Miscellaneous .2 anomalipes .3 imparisetae .4 truncipenna Picture wings .5 adiastola .6 crucigera .7 engyochracea .8 fasciculisetae .9 grimshawi .10 picticornis .11 pilimana .12punalua .13 villosipedis Modified mouthparts .14 aquila .15 asketostoma .16 comatifemora .17 conjectura .18 eurypeza .19 freycinetiae .20 hirticoxa .21 infuscata .22 involuta .23 ischnotrix .24mimica .25 mycetophila .26 residua Bristle tarsi .27 basimacula .28 expansa .29 perissopoda .30 (T.) petalopeza .31 torula Spoon tarsi .32 disticha .33 neutralis .34 sordidapex Split tarsi .35 ancyla .36 pectinitarsus The University of Texas Publication 00 OU OG uo ~~OU .14 .8 .9 .10 .II .12 .13 .15 .16 . 17 .18 .19 .20 .21 : ..... ,·::: ·::: ·~ ®~. .28 .22 .23 .24 FIG. 10. Characteristics of eggs. For each form a ventral (left) and lateral (right)view is shown. The egg filaments have been omitted from the ventral view. DROSOPHILOID .14 stramineifrons Genus: DROSOPHILA Subgenus: Bunostoma White tip scutellum .15 anomala .1 fungicola .16 palmae .2 haleakalae Subgenus: Exalloscaptomyza .3 melanoloma .17 forms from all islands .4 melanosoma Subgenus: Parascaptomrza .5 nigra .18 pallida .6 bipolita Subgenus: Tantalia .7 canipolita .19 varipicta .8 demipolita Subgenus: Trogloscaptomrza SCAPTOID .20 argentifrons Genus: DROSOPHILA .21 articulata .9 crassifemur .22 connata .10 nasalis .23 hadunani .11 parva .24 inaequalis Genus: TITANOCHAETA .25 intricata .12 #8 .26 latitergum Genus: SCAPTOMYZA .27 levata Subgenus: Alloscaptomyza .28 rostr a ta .13 longiseta the larva exhibited a behavior pattern in which it extruded the stylet, held it in an extended position, and then pulled, as if it were attempting to tear its way through a fabric. Whether or not this behavior was adapted to gaining entrance into a spider egg case is, of course, not known. At any rate, some Titanochaeta, some Scaptomyza, and D. nasalis all seem to share the ability to lay few and large eggs which may on occasion hatch in the vagina without adverse effect on either the female or the larva. Two species from the subgenus Alloscaptomyza (Figures 10.13, .14) both lay very small eggs. When females of these species were dissected the vagina was found to be very large and packed with eggs. It occupied almost the entire abdomen and the ovaries were degenerating and almost non-existent. Apparently these species are adapted to do just the reverse of what Exalloscaptomyza and Titanocho.eta do. They lay many small eggs and hold them all until a proper site is found, then lay them en masse. These eggs gave no indication of hatching in the vagina. There is one other important feature of the eggs of Hawaiian species. In many forms the thinly sculptured opercular area extends down the ventral surface of the egg almost to the posterior end. In some cases (Scaptoid) this area is bounded by lateral ridges (Figures 10.19, .23, .27). In others (Drosophiloid) these ridges are lacking (Figures 9.4, .7, .9, etc.) . I will refer to this ventral area as a cleft. Sometimes the edges of this cleft are fused to form a single midventral ridge, which I will refer to as a suture. These features are found in various combina­tions. Some eggs lack both cleft and suture (Figure 9.2), which is the usual state for non-Hawaiian Drosophila. Some have only a cleft, and this may be either long (Figure 9.4) or short (Figure 9.33). Some have only a suture (Figure 10.16). Many have a short cleft plus a suture (Figures 9.26; 10.15. etc.). Eggs having the cleft bounded by lateral ridges are found in species elsewhere in the world. One of these is D. populi. Others are in a pair of related genera, Leu­cophenga and Amiota. To my knowledge, eggs having the simple cleft, the suture, or the cleft and suture are found only in the genus Scaptomyza and in the Hawai­ian Drosophiloids. The paragonia and vasa deferentia-Figures 11, 12 and 13 show the internal reproductive tract of the male. Testes are not included in the figures. The para­gonia of the Drosophiloids show a considerable range of types. They also tend to be somewhat variable within species and often are not bilaterally symmetrical (e.g., Figure 11.38). The number of folds in the paragonia varies from somewhat more than one (Figure 11.9) to about five (Figure 11.38). In this respect these species resemble flies from the "virilis-repleta section" of the subgenus Drosophila or some species of Pholadoris. One of the most characteristic features of the paragonia of Hawaiian Drosophiloids is the low first arch (Figures 11 and 12). This is in marked contrast to the condition in the Scaptoids, where the first arch is generally high (Figures 13.7-.34). A few Drosophiloid species have a rather high first arch (Figures 11.31, 12.Q5, .26), but it is never as high as that seen characteristically in the Scaptoids. However, several of the Scaptoids do have the typical Drosophiloid paragonia (Figures 13.9, .33) . At the present time very little association is seen between type of paragonia and species groups. One of the more characteristic types (Figures 11.1, .10, .23, .27, etc.) tends to be found among ldiomyia, Nudidrosophila and "modified mouthparts." Another (Figure 12.10) tends to be found most frequently among flies with tarsal ornaments in the male. It is possible that more regular groupings will be apparent when the species group taxonomy has been worked out. Until then, the major importance of the characters of the paragonia lie in the evidence The University of Texas Pub!" . they provide for the origin of the Hawaiian Drosophilids and in the evidence they provide for differentiating the Drosophiloids from the Scaptoids in Hawaii. Itmay be noted that the "white tip scutellum" flies, whose eggs and spermathecae are Scaptoid, have paragonia that are fully Drosophiloid. It is probable that the paragonia also provide species characters. In Figures 12 and 13 I have included some possible examples of this. In Figures 12.1 and 12.2 are shown the male genital tracts from flies from two localities on Oahu. The first is from Pupukea, the second from Mt. Tantalus, and both are identified by Hardy as D. ischnotrix. Figures 12.25 and 12.26 show the characteristics of D. sordid­apex, one from Kulani Road and one from Kilauea on the island of Hawaii. Figures 12.32 and 12.33 show the characteristics of D. fungicola from Kipuka Puaulu and the Paauilo Experiment Station on the island of Hawaii. In this last case the individuals also differ in testis color between these two localities. In Figures 13.2 and 13.3 are shown the characteristics of D. melanosoma from Kumuwela Ridge and Halemanu Valley in Kokee State Park on the island of Kauai. In the case of D. fungicola, I have dissected flies from the listed localities both in 1963 and 1964. The characteristics shown are true for the two localities in both seasons. While it seems probable that differences of the type noted here are species differences, it is also possible that they represent simple polymor­phisms or local racial differences. This can only be determined by laboratory breeding tests and until these can be made it seems best to leave the question of species status open. The vasa deferentia tend to follow the curvature of the first arch of the para­gonia in both the Drosophiloids and the Scaptoids. The association is generally Fm. 11. Paragonia and vasa deferentia DROSOPHILOIDS Picture wings Genus: IDIOMYIA .18 adiastola .1 obscuripes .19 crucigera .2 perkinsi .20 engyochracea .3 picta .21 fasciculisetae Genus: ANTOPOCERUS .22 picticornis .4 aduncus .23 pilimana .5 diamphidiopodus .24punalua .6 longiseta .25 villosipedis .7 orthopterus Modified mouthparts .8 tanythrix .26 aquila .9 villosus .27 araiotrichia Genus: NUDIDROSOPHILA · .28 asketostoma .10 aenicta .29 chaetopeza .11 lepidobregma .30 comatifemora Genus: ATELEDROSOPHILA .31 conjectura .16 preapicula .32 dissita , Genus: DROSOPHILA .33 eurypeza .; Miscellaneous .34 flavibasis .12 anomalipes .35 freycinetiae .13 caccabata .36 furvifacies .14 hirtitibia .37 hirticoxa .15 imparisetae .38 infuscata .17 quasianomalipes .39 involuta The University of Texas Publication rather loose, but it is strong in some of the Scaptoids. In this respect the Hawaiian Drosophilids are intermediate between the condition characteristic of the more primitive Pholadoris, Sophophora, Chymomyza and Dorsilopha and the more derivitive Drosophila, Phloridosa, Dettopsomyia, Zaprionus and Mycodrosophiln.. Among the Drosophiloids three general types of vasa are seen. One is an extreme coiled and folded type (Figures 11.3, .12, .17, .29, etc.) that is found particularly among ldiomyia, "picture wings" and flies with modified mouth­parts. Elsewhere (Throckmorton, 1962) I commented on the coiling of the vas deferens among species of Drosophila and concluded (following Stern, 1940) that the coiling of the vas represented countercoils produced as a result of the coiling of the testis. This conclusion still holds for the flies dealt with there, but this interpretation can be only partly correct for the species just mentioned. Here the number of coils in the vas is often higher than the number of coils in the testis, and there are several reversals of coiling direction along the vas (e.g., Figure 11.17). While other explanations are possible, it is probable that there is some intrinsic determination of form for these vasa. Hence, these species show a characteristic that may be peculiar to this group of flies. In any case, this charac­teristic seems to link two odd species (D. anomalipes and D. quasianomalipes) with the "picture wings," ldiomyia, and flies with modified mouthparts. This is of some importance, since the female of D. anomalipes (I have not seen the female of quasianomalipes) is the only Drosophiloid to have a true coiled ventral recep­tacle. She also has a type of spermatheca that is different from other Drosophiloid types, although it is seen among the Scaptoids. This might suggest the inde­pendent origin of anomalipes and its relatives, a question that will be discussed later. However, both with respect to characters of the paragonia and to characters of the vasa, anomalipes is an ordinary Hawaiian Drosophiloid. FIG. 12. Paragonia and vasa deferentia DROSOPHILOIDS Genus: DROSOPHILA Modified mouthparts (cont.) .1 ischnotrix (Pupukea, Oahu) .2 ischnotrix (Mt. Tantalus, Oahu) .3 kauluai .4 mimica .5 mycetophila .6 pychnochaetae .7 residua .8 scolostoma Bristle tarsi .9 apodasta .10 basimacula .11 expansa .12 perissopoda .13 (T.) petalopeza .14 prodita .15 redunca .16 seclusa .17 torula .18 trichaetosa Spoon tarsi .19 confonnis .20 contorta .21 disticha .22 incognita .23 neutralis .24 policiformis .25 sordidapex (Kulani Road, Hawaii) .26 sordidapex (Kilauea, Hawaii) Split tarsi .27 ancyla .28 clavata .29 fundita .30 pectinitarsis White tip scutellum .31 cilifemorata .32 fungicola (Kipuka Puaulu, Hawaii) .33 fungicola (Paauilo Expt. Sta., Hawaii) .34 haleakalae .35 iki The University of Texas p . Throckmorton: Relationships of Hawaiian Drosophilidae The second type of vasa is the one that is most common, both in Harwaii and elsewhere. The distal end of the vas is moderately enlarged and generally coiled about three times. In this type, the number of coils in the vas follows the number of coils in the testis and is always less than the number of coils in the testis. This type is shown in Figures 11.4, .5, .6, etc. In the third type the distal section of the vas is generally strongly enlarged, somewhat sausage-shaped and very little coiled. This is seen, for example, among some species of I diomyia (Figure 11.1) but it is most pronounced among the "white tip scutellum" flies. Here the vas is almost completely uncoiled (Figures 12.31-.35, 13.1-.6). This is, essentially, a Scaptoid characteristic (Figures 13.7­.34). Among the Scaptoids the differentiated section of the vas is completely uncoiled, plump, and nearly barrel-shaped at its most extreme development (Figure 13.14). As a general rule the pigmentation of the vas deferens falls into two general types. In the Drosophiloid type the pigmentation extends from the testis to the base of the vas. Among the Scaptoids the pigmentation extends only over the differentiated section. The base is completely unpigmented. An intermediate condition is found in the "white tip scutellum" species. In these the pigmentation generally extends about midway between the differentiated section and the base. In D. crassifemur (Figure 13.7) , which is otherwise fully Scaptoid, pigmentation of the vas extends almost to its base. In D. nasalis and D. parva pigmentation is of the Scaptoid type. The eiaculatorr bulb-Figures 11, 12 and 13 include some of the character- FIG. 13. Paragonia and vasa deferentia DROSOPHILOIDS Genus: DROSOPHILA White tip scutellum (cont.) .1 melanoloma .2 melanosoma (Kumuwela Ridge, Kauai) .3 melanosoma (Halemanu Valley, Kauai) .4 nanella .5 canipolita .6 demipolita SCAPTOIDS Genus: DROSOPHILA .7 crassifemur .8 nasalis .9 parva Genus: TITANOCHAETA .10 contestata Genus: SCAPTOMY"ZA Subgenus: Alloscaptomyza .11 longisetosa .12 stramineifrons Subgenus: Burwstoma .13 anomala .14palmae .15 xanthopleura Subgenus: Exalloscaptomrza .16 from Hawaii .17 from Maui .18 from Molokai .19 from Oahu .20 from Kauai Subgenus: Parascaptomyza .21 pallida Subgenus: Rosenwaldia .22abrupta Subgenus: Tantalia .23 varipicta Subgenus: Trogloscaptomyza .24 argentifrons .25 articulata .26 connata .27 hackmani .28 inaequalis .29 intricata .30 latitergum .31 levata .32 retusa .33 rostrata .34 sil vicola The University of Texas Publication Throckmorton: Relationships of Hawaiian Drosophilidae istics of the ejaculatory bulbs. Here the gross details are shown and the general conformation of the long caecae can be seen. The Drosophiloids and Scaptoids are sharply different with respect to ejaculatory caecae. These are conspicuous in Scaptoids and either absent or inconspicuous in Drosophiloids. Generally the caecae of Hawaiian Drosophiloids are not apparent until after clearing the ejacu­latory bulb in phenol. In Figure 13 the Scaptoid caecae are shown extended. In situ they make up a tangled mass occupying the posterior part of the abdomen. More complete details of the ejaculatory bulb and ejaculatory apodeme are shown in Figures 14 and 15. These diagrams show the bulb after clearing in phenol. Only the bases of the Scaptoid caecae are shown in Figure 15. In the figures, anterior is toward the right, ventral toward the top. In most Drosophiloid species the ejaculatory bulb is a simple sac-like structure. In general shape the type shown in Figure 14.4 is very near the primitive for the genus Drosophila, and this has not been much modified among the Hawaiian forms. Among the Drosophiloids there is only slight development of ejaculatory caecae. Short caecae are found in some ldiomyia (Figures 14.1-.3), in some Fm. 14. Ejaculatory bulb and ejaculatory apodeme DROSOPHILOID Genus: IDIOMYIA .1 obscuripes .2 perkinsi .3 picta Genus: ANTOPOCERUS .4 aduncus .5 diamphidiopodus .6 longiseta . 7 orthopterus .8 tanythrix .9 villosus Genus: NUDIDROSOPHILA .10 aenicta .11 lepidobregma Genus: ATELEDROSOPHILA .16 preapicula Genus: DROSOPHILA Miscellanous .12 anomalipes .13 caccabata .14 hirtit:bia .15 imparisetae .17 quasianomalipes Picture wings .18 adiastola .19 crucigera .20 engyochracea .21 fasciculisetae .22 grimshawi .23 picticornis .24 pilimana .25 punalua .26 villosi pedis Modified mouth parts .27 aquila .28 araiotrichia .29 asketostoma .30 chaetopeza .31 comatifemora .32 conjectura .33 dissita .34 eurypeza .35 flavibasis .36 freycinetiae .3 7 furvifacies .38 hirticoxa .39 infuscata .40 involuta .41 ischnotrix .42kauluai .43 mimica .44 mycetophila .45 pychnochaetae .46 residua .47 scolostoma Bristle tarsi .48 apodasta .49 basimacula .50 expansa .51 perissopoda .52 (T .) petalopeza .53 prodita .55 seculsa .54 redun<:a .56 torula .5 7 trichaetosa The University of Texas Publication "picture wings" (Figures 14.;23, .26) and in some flies with modified mouthparts (Figures 14.40, .43, .45, .47) . In the latter cases it is problematical whether the structures involved are "lateral lobes" or caecae. Since these probably represent variations of the same thing, the distinction is not critical. Lateral lobes are seen mostly among flies with modified mouthparts (Figures 14.29, .34, .39, .44), although some are also seen in the "bristle tarsi" group (Figures 14.49, .50). So far only one species of the "white tip scutellum" flies (Figure 15.12) has been found with the lateral lobes. Among the Drosophiloids the major development of caecae and lateral lobes is thus seen among ldiomyia, "picture wings" and "modi­fied mouthparts." There has been an unusual amount of diversification of the apodeme among Hawaiian Drosophilids and in many species the apodeme is very large and con­spicuous. Among the "picture wings," Nudidrosophila, and ldiomyia there has Throckmorton: Relationships of Hawaiian Drosophilidae been a trend toward the sclerotization of the posterior ejaculatory duct (toward the left in the figures). The basic apodeme, from which that of the Hawaiian species is almost certainly derived, is similar to the one shown in Figure 14.46. The plate is spade-like and roughly triangular in outline. Among some of the Hawaiian Drosophiloids there has been an increase in the sclerotization lateral to the plate so that the sides of the posterior ejaculatory duct are more or less solidly enclosed (Figures 14.3, .21, .40, etc.). In others the sclerotization continues and completely surrounds the posterior duct, but this additional area is not pig­mented (indicated by dashed lines in Figures 14.2, .10, .11, .12, .14, .17, etc.). In still others the duct is heavily sclerotized and pigmented (Figures 14.1, .18, .19, etc.). Again, these trends are most conspicuous in ldiomyia, Nudidrosophila, "picture wings" and the flies with modified mouthparts. D. anomalipes and D. quasianomalipes also fall into this group. An additional, and apparently inde­pendent, trend is also sE*!n among the species with modified mouthparts. In many of these flies the ejaculatory bulb is very wide (not figured), sometimes almost twice as broad as long. The plate of the ejaculatory apodeme is also very broad in these cases. Generally, an apodeme that surrounds the posterior ejaculatory Fm. 15. Ejaculatory bulb and ejaculatory apodeme DROSOPHILOID Genus: DROSOPHILA Spoon tarsi .1 conformis .2 contorta .3 disticha .4 incognita .5 neutralis .6 polliciforma .7 sordidapex Split tarsi .8 ancyla .9 clavata .10 fundita .11 pectinitarsus White tip scutellum .12 cilifemorata .13 fungicola .14 haleakalae .15 iki .16 melanoloma .17 melanosoma .18 nanella .19 canipolita .20 demipolita SCAPTOID Genus: DROSOPHILA .21 crassifemur .22nasalis .23 parva Genus: TITANOCHAETA .24 contestata Genus: SCAPTOMYZA Subgenus: Alloscaptomyza .25 longisetosa .26 stramineifrons Subgenus: Bunostoma .27 anomala .28 palmae .29 xanthopluera Subgenus: Exalloscaptomyza .30 from Hawaii .31 from Maui .32 from Molokai .33 from Oahu .34 from Kauai Subgenus: Parascaptomyza .35 pallida Subgenus: Rosenwaldia .36 abrupta Subgenus: Tantalia .37 varipicta Subgenus: Trogloscaptomyza .38 argentifrons .39 articulata .40 connata .41 hackmani .42 inaequalis .43 intricata .44 latitergum .45 levata .45 retusa . 4 7 rostrata .48 silvicola The University of Texas Publication duct is narrow (Figure 14.47), and one that is very wide does not surround the duct (Figure 14.28). Some, however, do both (Figures 14.37, .42). In many of these species the ejaculatory apodeme has a slightly different shape that seems to be peculiar to Hawaiian Drosophilids. The difference is difficult to describe but it can be seen by comparing the apodeme in Figure 14.46 with that in Figure 14. 7. That seen in Figure 14. 7 is the Hawaiian type. Elements of this shape can be seen even in some of the more derivative types (e.g., Figures 14.1, .14, etc.). This type of apodeme is found most conspicuously among species of Antopocerus (Figures 14.4--.9) and among those species whose males have tarsal ornaments (Figures 14.48-.57, 15.1-.11). In this latter group the apodeme may occasionally be of the standard type (Figure 15.4, etc.). Among the "white tip scutellum" species the apodeme is also generally of one of these two types, with the additional feature that it is often flexed, sometimes strongly so (e.g., Figure 15.19). In some of these species also the anterior (toward the right) flanges of the plate may encircle the anterior ejaculatory duct and form a completely sclerotized ring around it (not shown, but present in some undescribed species). Among the Scaptoids there is considerably less variation in ejaculatory bulbs and ejaculatory apodemes. All bulbs have caecae, and sometimes these are branched (Figure 13). The apodeme is rarely of the standard type (Figure 15.21). More generally it is of the kind very typical of Scaptomyza. In this the plate is very much flattened so that it almost disappears in lateral view. The anterior flanges or angles of the plate are much reduced (Figure 15.23) or rounded off completely (Figure 15.24, etc.). The handle is almost always simple. At most it may have a small blade (Figure 15.31 ), which is quite in contrast to the elaborate developments (e.g., Figure 14.44) of some Drosophiloids. The testes-Table 1 lists the number of testis coils and testis color among Hawaiian species. The range in number of coils is restricted relative to that found among species of Drosophila from elsewhere in the world. In an earlier sample (Throckmorton, 1962) of 195 species, approximately 36 per cent had more than six coils in the testes. Here (Table 1) only 2 per cent of the Drosophiloids and 4 per cent of the Scaptoids have such high numbers. The range in number of testis coils reaches from zero (elliptical testes) to more than twenty among non­Hawaiian species. It reaches from approximately one to nine coils among the Hawaiian Drosophilids. Thus far no species have been found in Hawaii having elliptical testes, although some species among the "white tip scutellum" flies have less than one coil. These are not listed in Table 1, since the species are un­described. The righthand columns in Table 1 indicate the distribution of testis colors among the Hawaiian forms. A preponderance (about 70 per cent) of the Dro­sophiloids have testes that are basically yellow. About 60 per cent of the Scaptoids have testes that are basically orange. There is a great deal of variation in intensity of color, but no attempt has been made to indicate this in the table. Development of pigment is greatly influenced by age, and since most individuals were collected from the wild, age was unknown. Abdominal sternites in the male-As discussed by Wheeler (1960) and Throckmorton (1962) presence or absence of the first and sixth abdominal ster­nites in the male may provide evidence of evolutionary position. In general the Throckmorton: Relationships of Hawaiian Drosophilidae TABLE 1 Number of testis coils and testis color among the Hawaiian Drosophilids Testis color Number of coils Group 1-3 3--6 6-9 yellow yellow-brown yellow-orange orange Drosophiloid ldiomyia 3 2 Antopocerus 6 6 N udidrosophila 2 2 Miscellaneous 5 3 1 picture wings 2 7 6 2 1 modified mouthparts 5 15 11 4 5 bristle tarsi 10 7 2 spoon tarsi 8 7 1 forked tarsi 4 3 1 white tip scutellum 9 3 6 Scaptoid Titanochaeta 1 Alloscaptomyza 2 Bunostoma 2 3 Exalloscaptomyza 4 5 Parascaptomyza Rosenwaldia Tantalia 1 1 Trogloscaptomyza 2 9 7 4 Total Drosophiloid 27 49 2 50 5 16 7 Total Scaptoid 8 16 10 14 males of Sophophoran species have the sixth sternite present as a polished and generally unbristled plate. Among species of the subgenus Pholadoris and Chy­momyz.a, the sixth sternite may be unreduced and fully bristled or it may be partly reduced and present only as a pair of bristled plates. With the exception of D. testacea, species from the su'bgenus Drosophi/,a are not known to possess the sixth sternite. The first sternite is generally not present in Sophophorans (except D. populi and very faint remnants in some species of the saltans group). It is present in some species of Chymomyza and Pholndoris and absent in others. It is generally absent in species of the subgenus Drosophila and in the other genera and sub­genera closely related to it (Zaprionus, Mycodrosophila, Dettopsomyia, Phlori­dosa). Among the Hawaiian Drosophilids I have seen the sixth sternite present and unreduced in ldiomyia picta, D. adiastola and D. villosipedis. The last two are both "picture wings." It is present but reduced in D. picticornis (present as paired plates) , in D. furvifacies (a normal sclerite thinly sclerotized in the midline) and in D. mimica (only remnants seen). It is present as paired patches in the male of D. nasalis collected from Paliku in Haleakala Crater on the island of Maui. It was not seen in D. nasalis males collected from Molokai. It is very faintly evi­dent in males of a species of Exalloscaptomyza collected at Kamuela on the island The University of Texas Publication .23 .24 .25 .26 .27 .28 FIG. 16. Anterior pupal spiracles, intra-anal lobes and Malpighian tubules. DROSOPHILOID spiracles: .1) Drosophila crucigera, .2) D. adiastola (face view), .3) D adiastola (side view) , .4) D. grimshawi, .5) D. pilimana, .6) D. punalua, .7) D. villosipedis, .8) D. eurypeza, .9) D. ischnotrix, .10) D. mimica, .11) D. mycetophila, .12) D. melanosoma. SCAPTOID spiracles: .13) D. crassifemur (Paliku), .14) D. crassifemur (Kauai), .15) D. parva, .16) Scapto~yza palmae (face view), .17) S. palmae (side view), .18) S. mauiensis, .19) S. pallida, .20) S. varipicta, .21) S. argentifrons. Intra-anal lobes: .22 lateral view, .23 ventral view; a.1.-anal lining; a.p.-anal plate; a.s.-anal sclerite (dashed line); i.1.-intra-anal lobe. Malpighian tubules: .24-.28. See text. Throckmorton: Relationships of Hawaiian Drosophilidae of Hawaii. The area of the sixth sternite is sclerotized but not bristled in D. anomalipes and D. engyochracea. The sixth sternite is thus seen among the Drosophiloids from ldiomyia, the "picture wings," and the "modified mouth­parts." It is also seen in D. anomalipes and D. quasianomalipes. From the Scap­toids it is seen in D. nasalis and in Exalloscaptomyza. The first sternite is present in ldiomyia obscuripes, /. picta and/. perkinsi. It is also seen in S. (Bunostoma) palmae from Kamuela, Hawaii, and in a species of Exalloscaptomyza from Molokai. Intra-anal lobes in the male-Figures 16.22 and 16.23 show a peculiar struc­ture that is present in many of the Hawaiian Drosophiloids. This takes several forms and I have made no attempt to note all of them. The figure shows the most extreme development of the characteristic that I have seen. In these cases there exists a pair of lobes covered with a fine dense pile and lying between the anal plates in the male. These lobes can be completely withdrawn between the anal plates or they can be extruded as shown in the figures. The lobes are not present in the female. Associated with the lobes is a special sclerite that lies in the median line between the lobes when they are present (shown by dashed line in Figure 16.22) . It is anchored in the dorsal membrane just posterior to the genital arch and it apparently serves as a support for muscles that act to retract the lobes when they are not in use. This sclerite is seen in many males that do not show evidence of the lobes themselves. Drosophiloid males may thus fall in one of three categories: without anal sclerite or intra-anal lobes, with anal sclerite but with­out intra-anal lobes, and with both anal sclerite and intra-anal lobes. This is probably also an evolutionary sequence from primitive to derivative, with pres­ence of intra-anal lobes being most derivative. These lobes are most highly devel­oped among "picture wings," but at least some species from all major groups of Drosophiloids except the "white tip scutellum" flies show at least the anal sclerite. Spieth (this Bulletin) has described courtship behavior from many species of Drosophiloid males. Two behavior patterns seem relevant in the present context. In one case the male during courtship elevates and bends the abdomen so that the anal region is pointed toward the female. At the same time, probably through abdominal pressure, the lining of the anal passage may be extruded as a distinct cylinder of tissue and a drop of liquid hangs on its end. Presumably this has some function as a sexual stimulant. I have dissected males of some of the species with this behavior pattern and found the anal sclerite present, but not the intra-anal lobes. Apparently the sclerite functions in the withdrawal of the extruded anal lining after this phase of courtship is complete. Since I have never made a detailed micro-examination of the anal lining in these species, it is possible that some less conspicuous form of the intra-anal lobes was present in these flies also. In other instances the male may be seen touching the tip of its abdomen to the substrate or dragging it along the substrate. In the species which show this type of activity, the male does have the intra-anal lobes. Spieth suggests that the male may be laying an odor trail or laying down scent to act as an aggregation point. The anal lobes, with their dense covering of very short hairs, may be especially adapted for this purpose, although it is quite possible that they also function in the pro­duction of a sexual stimulus during courtship. To date, observations on these peculiar structures of the male Drosophiloid are too fragmentary to warrant The University of Texas Publication further discussion. A systematic study of the structural modifications, correlated with a detailed study of behavior, would undoubtedly be rewarding. Pupal spiracles-The major features of the anterior pupal spiracles of the Hawaiian Drosophilids are shO'wn in Figure 16. They are of a very simple type, but I have not seen this form elsewhere. The branches are short and cylindrical. Generally they stand erect. In this respect they resemble the branches of pupal spiracles from species in the subgenus Drosophila. In many respects these spiracles resemble those of species in the virilis group and they are perhaps as close to them as to any other. However, even here the similarities are not striking. Most of the figures show a face view of the spiracle (a latero-ventral view). Figure 16.3 shows a side view of the spiracle of D. adiastola, and Figure 16.17 shows a side view of the spiracle of S. (Bunostoma) palmae. As inspection of Figure 16 will indicate, these spiracles are very much alike. Although the more typical Drosophiloids spiracle (Figure 16.6) is quite distinct from the typical Scaptoid spiracles (Figure 16.21) the two types intergrade so smoothly that it is impossible, in Hawaii, to be certain of the genus from which a given spiracle comes. One of the characteristic features of the Scaptoid spiracle is that the antibasal branches tend to be elongated and fused basally. This feature is also seen in several of the Drosophiloid spiracles (e.g., Figure 16.9). Conversely, spiracles that are very similar to the Drosophiloid type are seen in Scaptoids (compare Figures 16.11 and .14). One of the Scaptoids, D. crassifemur, has rather variable spiracles. Figure 16.14 shows one extreme of the type found from flies collected at Paliku, Maui. Several pupae were examined from this culture and the number of branches varied from twenty-two (shown) to seventeen. Another culture from flies collected at Kokee State Park, Kauai,* showed spiracles with from fifteen to eleven branches. This latter type is shown in Figure 16.14. It is probable that additional cultures from the two localities might have produced pupae that overlapped completely in their characteristic number of branches. Whether or not these intergrade completely, there is a distinct gradation between these two populations from fully Drosophiloid spiracles (Figure 16.13) toward the Scaptoid type. I am not certain that I have seen pupae from Titanochaeta. I am indebted to Mrs. Meredith Carson for bringing to my attention some puparia she found par­tially embedded in a spider egg case. The egg case was small and there were only three puparia. Two were partially within the egg case and the third was almost completely on its surface. These puparia were all alike and had spiracles very similar to the type shown in Figure 16.12. That is, the spiracles were fully Dro­sophiloid.f Malpighian tubules-Malpighian tubules from Hawaiian species are shown in Figures 16.24 to 16.28. All Drosophiloid species have the type shown in Figure 16.'24. This is the type usual for species in the subgenus Drosophila. The stalks are short and the posterior tubules have their tips fused, the lumen continuous. The Malpighian tubules of the Scaptoids are variable. About half of these species have the type shown in Figure 16.24, and about half have the type shown in • Editor's note: The form from Kauai is a different species; see Hardy, this Bulletin. t During the summer of 1965 I saw pupae from Mrs. Carson's recent collection. Titanochaeta had emerged from them in the laboratory. The spiracles were as shown in Figure 16.12. Throckmorton: Relationships of Hawaiian Drosophilidae Figure 16.25 (posterior tips apposed). Several species show both types. Thus, S. hackmani from Molokai has the posterior tips fused, that from Hawaii has the tips apposed. S. palma.e from Mt. Tantalus, Oahu, has the tips apposed, that from Hawaii has the tips fused. Since these came from different collecting localities it is possible that cryptic species are being dealt with instead of geographical strains. These may also represent polymorphisms, and there are at least two cases where this is probably the case. Species of Scaptomyza from the subgenus Exalloscaptomyza are quite variable in this respect. Flies collected from a single locality at Kokee State Park, Kauai, had both types of Malpighian tubule. Some had the posterior tips apposed. Some had the posterior tips fused. The strain of D. crassifemur collected from Molokai also showed this type of variation. It is probable that more species would be found to show this type of variation if this character were investigated more fully, particularly in wild-caught individuals. Several Scaptoid species have more sharply modified Malpighian tubules. In specimens of Exalloscaptomyza collected from Oahu, Molokai, and Maui, the stalk of the posterior tubule was about 2X the usual length and both the anterior and posterior tubules were very short (Figure 16.26). This type of Malpighian tubule was also found in the two species from the subgenus Alloscaptomyza (S. longiseta and S. stramineifrons) . Two species, S. inaequalis and S. articulata, had this same type of tubule, except that the posterior tubules were apposed (Figure 16.27). In one species, S. varipicta, the posterior tubules were fused and the anterior tubules were much shortened (Figure 16.28) . Of the two species from the genus Titanochaeta, one had the posterior tubules fused, the other had them apposed. Thus, for this character also, Titanochaeta is Scaptoid. D1scussroN The interrelationships of the Drosophiloids-For all their diversity in external features (Hardy, 1965), the Drosophiloids are a remarkably compact and co­hesive group of species. The sharpest division within this group, between the "white tips" and the other Drosophiloids, is not presently recognized by the formal taxonomy, mainly because the characteristics that distinguish these are not among those characters customarily used to differentiate groups of Diptera. Conversely, several groups, otherwise undistinguished, are set off formally as genera because of novelty in conventional characters. This tends to give a mis­leading impression of divergence. Thus the genera, Idiomyia, Antopocerus, and Nudidrosophila are based on far more trivial and less extensive complexes of characters than would be a division between the "white tip scutellum" species and the other Drosophiloids. I am not intending here to quarrel with the formal taxonomy of these species as it stands today. It is simply necessary to remind the reader that taxa of equivalent rank are not necessarily of the same biological or evolutionary significance, and this is particularly true of many of the Hawaiian genera and subgenera. Thus, for most of the earlier descriptions in this paper, it has been possible to treat Hawaiian species in two groups, Drosophiloid and Scaptoid, regardless of their formal classification. It should have been evident in these descriptions that, for the characteristics covered herein, group designa­tions are of little consequence. Most of these characteristics are characteristics of The University of Texas Publication Hawaiian Drosophiloids, not of any particular group of Drosophiloids. Even some of the most distinctive features (some ovipositors, some egg characteristics, some ejaculatory bulbs, etc. ) are not confined to single groups or even char­acteristic of any presently recognized group. While this situation is sympto­matic of the type of evolution occurring among Hawaiian species (see later), and is of interest for that reason, it makes the analysis of Drosophiloid relation­ships inter se very difficult. Oftentimes speculation must be based on only a few characters and hence, for the present, phylogenetic relationships are quite tenta­tive. The method of phylogenetic analysis follows that of Throckmorton (1962, 1965 ). As indicated above, the major cleavage among the Hawaiian Drosophiloids is that between the "white tip scutellum" species and the other forms. The "white tips" have a very characteristic external "gestalt," and they can be recognized and separated in the field without the aid of magnification. At least three sub­groups can be discerned within this group. One includes such extreme forms as nigra, and another is the "polita complex" (in this study, bipolita, canipolita and demipolita ). A third group is made up of such species as fungicola or melanoloma. These last forms occupy a more or less central position within the "white tips" but do not constitute a phylogenetic unit as such. They are the forms that are nearest in general morphology to the other Drosophiloids. D. fungicola and im­parisetae, for example, have a superficial resemblance in wing pattern, and in several other features (the light tip of the scutellum, for example) imparisetae diverges toward the "white tips" in its characteristics. I have grouped imparisetae with the "miscellaneous" species in the Appendix and in the figures. It is cer­tainly not a "white tip," but it represents a rather nondescript cluster of species that have retained the more general Drosophila characters rather than diverging in the peculiar ways of other Hawaiian forms. On the basis of external charac­ters, it seems reasonable to infer an origin of the "white tips" from Hawaiian species not too different from imparisetae and its relatives. On the basis of internal characters also, there can be little question that the "white tips" are descended from the same ancestor as the other Hawaiian species. Even a cursory inspection of Figures 1 through 16 will show the extent to which these species share basic Hawaiian characteristics with other forms. There are only two major characters, the egg and the spermatheca, that have states restricted to these species. Other characters are either the standard Hawaiian types or obvious modifications of them. The fundamental unity of the Hawaiian character complexes can best be appreciated if they are contrasted with like complexes from other groups within the genus. Such data have been presented elsewhere (Throck­morton, 1962) and the reader is referred to that paper for more detailed infor­mation on the characters of continental forms. Three of the major characters are of particular importance in giving evidence of the distinctness and unity of the Hawaiian Drosophiloids. These are the ventral receptacle, the paragonia and the pupal spiracle. A summary of these and other characters is given in Figure 17. The ventral receptacles are of a type rarely seen elsewhere in the genus, and one of their most critical features is not particu­ larly evident in the figures. This concerns the relationship of the ventral recep­ Throckmorton: Relationships of Hawaiian Drosophilidae tacle to the vagina and its disposition among the other internal organs. The ven­tral receptacles of Hawaiian Drosophiloids are of three major types (Figure 17.6­.8), with the two types shown in Figure 17.6 and 17.7 predominating. The type shown in Figure 17.6 appears, in the figure, to be very similar to types seen, for example, among Sophophorans. It differs, however, in that it is completely free of the vagina rather than being reflexed against its surface and partially embedded in it. This sets the Hawaiian ventral receptacle apart from all others of its gen­eral conformation. For this character the "white tip scutellum" flies are quite like other Hawaiian species and are, indeed, often more extreme (e.g., Figure 5.8). As inspection of Figures 3 through 6 will show, there are likewise no other char­acters of the ventral receptacles that serve to set the "white tip scutellum" flies apart from other Drosophiloids. Without reference to the figure legends, it is impossible to determine, in the figures, the point of transition from ventral recep­tacles of other Drosophiloids to ventral receptacles of the "white tip scutellum" species. When the basic similarities seen in Figures 3 through 6 of this paper are contrasted with the wide variety of types seen elsewhere in the genus (Figures 34-4,0 in Throckmorton, 1962) it is difficult to avoid the conclusion that, for this character, Hawaiian Drosophiloids are a compact group, and the "white tip scutellum" species are full members of this group. The paragonia (Figures 17.35-.39) are likewise of a distinct Hawaiian type, even though they show considerable variety among themselves. Inspection of Figures 3 through 14 in Throckmorton (1962) amply enforces this conclusion. One of the most characteristic features of the Drosophiloid paragonium is its low first arch. In general, this sets these forms apart from most other species in the genus. Here, as for the ventral receptacles, the "white tip scutellum" species have characteristics that are of the basic Hawaiian type. Also, they share with the other Drosophiloids certain specific types of paragonia that are unique to Hawaii (e.g., Figures 12.10 and 12.35). Thus, the characteristics of the paragonia indi­cate that the "white tip scutellum" species are full Hawaiian Drosophiloids. So far as my experience extends, the anterior pupal spiracles of Hawaiian Dro­sophiloids are unique. On the basis of pupal characters, it is impossible to separate "white tip scutellum" species from other Drosophiloids. Hence, this character also enforces the conclusion that the Hawaiian Drosophiloids, including the "white tip scutellum" species, are a compact and cohesive group. The reader is referred to Figures 42 through 45 of my earlier paper (1962) and to Figure 16 in this one for a comparison of the spiracle types. For most of the remaining characteristics, character states in Hawaiian forms are seen to diverge from states seen elsewhere in the genus, and at least some Hawaiian species show character states common elsewhere in the world. The egg filaments (Figures 17.12-.16) are good examples of this. The four-filament type seen in Figure 17 .14 is the type common to most species in the subgenus Dro­sophila. The "white tip scutellum" species diverge from this type and have either four or two short filaments. The other Drosophiloids diverge from this type toward eggs with four very long filaments. For this character, therefore, we have evidence for two distinct lineages in Hawaii, one showing a reduction in size and number of filaments, the other an increase in length of the filaments. Bearing in The University of Texas Publication Throckmorton: Relationships of Hawaiian Drosophilidae mind the evidence from the paragonia, ventral receptacles and pupal spiracles, these lineages seem more probably to represent divergence within Hawaii than lineages separately introduced to Hawaii. Evidence from the ovipositors almost directly parallels that from the eggs. As discussed earlier, there are two major trends in ovipositor development among Hawaiian forms. One is toward a reduction of the ovipositor, the other toward either a strongly sclerotized type or a membranous, tubular type. One type of ovipositor (Figures 7.10 and 8.7) is common to both the "white tip scutellum" FIG. 17. Summary of the data Drosophiloid spermathecae .1 D. truncipenna .2 D. anomalipes .3 D. cilifemorata Scaptoid spermathecae .4 S. varipicta .5 S. stramineifrons Drosophiloid ventral receptacles .6 D. caccabata .7 D. truncipenna .8 D. anomalipes Scaptoid ventral receptacles .9 S. palmae .10 S. hackmani .11 D. nasalis Drosophiloid egg filaments .12 D. engyochracea .13 D. punalua .14 D. adiastola .15 D. nigra .16 D. bipolita Scaptoid egg filaments .17 S. inaequalis .18 S. pallida .19 S. latitergum .20 S. stramineifrons Drosophiloid ventral surface of egg .21 D. ancyla .22 D. villosipedis .23 D. disticha .24 D. infuscata Scaptoid ventral surface of egg .25 S. intricata .26 S. rostrata .27 S. argentifrons .28 S. connata .29 S. hackmani Drosophiloid ovipositors .30 D. pseudoobscura .31 D. pattersoni .32 S. retusa .33 D. imparisetae .34 D. immigrans Drosophiloid paragonia .35 D. redunca .36 D. demipolita .37 D. imparisetae .38 D. crucigera .39 D. ischnotrix Scaptoid paragonia .40D. parva .41 S. rostrata .42 T . contestata .43 S. retusa .44 S. pallida Drosophiloid vasa diferentia .45 D. anomalipes .46 D. hirtitibia .47 D. fungicola .48 D. demipolita Scaptoid vasa diferentia .49 D. nasalis .50 S. retusa .51 S. anomala Drosophiloid ejaculatory apodeme .52 D. adiastola .53 D. iki .54 /. obscuripes .55 D. picticornis .56 D. residua .57 D. araiotrichia .58 A . orthopterus Scaptoid ejaculatory apodeme .59 S. articulata .60 D. crassifemur .61 D. nasalis .62 S. palmae Drosophiloid spiracles .63 D. punalua .64 D. mycetophila .65 D. ischnotrix Scaptoid spiracles .66 D. crassifemur (Paliku) .67 D. crassifemur (Kauai) .68 S. mauiensis .69 S. varipicta The University of Texas Publication species and the other Drosophiloids. This type is probably also the "primitive" type for Hawaiian forms. Both the "white tip scutellum" species and the other Drosophiloids diverge from this, and hence evidence from the ovipositors is con­sistent with the development of these two lineages in Hawaii. The spermathecae show the "white tip scutellum" forms to be sharply distinct from the other Drosophiloids. The major types are shown in Figures 17.1 to .3. All of these types are seen in species from elsewhere in the world, although neither that shown in Figure 17.1 nor that in 17.3 is very common except in Hawaii. These cannot be used as evidence for the uniqueness of the Hawaiian species or as evidence for the origin of the two major Hawaiian lineages. How­ever, they do indicate the uniformity of the lineages within themselves. The uniformity of spermathecal type among the "other Drosophiloids" is phenomenal (Figure 1). Continental groups of comparable external diversity show far greater variation of the spermathecae than do the Hawaiian forms (see Figures 27-32 of Throckmorton, 1962) . The vasa of the Hawaiian forms show divergence from a relatively common character state. The important types are shown in Figures 17.45 to .48. The general type, and the type seen elsewhere in the genus, is that shown in Figure 17.46. The "white tip scutellum" species diverge from this toward the type seen in Figure 17.48. The other Drosophiloids generally retain the common type but in some groups the extreme forms seen in Figure 17.45 exists. Here, as for the egg filaments and ovipositors, the "white tip scutellum" species and the other Drosophiloids appear to have diverged from a common type. The same is also true for characteristics of the ejaculatory bulb. The "white tip scutellum" species have retained the simple bulb, and the ejaculatory apodeme has diverged from the basic type (Figure 17.56) on an independent line of its own (Figure 17.53). Here the trend has been toward the sclerotization of the base of the anterior ejaculatory duct by extension of the anterior corners or flanges of the apodeme plate. Among the other Drosophiloids there has been a development of bulbs with lateral lobes, or caecae, or peculiar shapes (Figures 14.2, .12, .40, .47, etc.). There has also been a development of some rather bizarre ejaculatory apodemes (Fig­ures 17.52, .54, .55, .57) based primarily on the sclerotization of the fJOSterior ejaculatory duct. In summary, there is ample evidence for two major lineages among the Hawai­ian Drosophiloids. For all character states but those of the spermathecae, these two lineages either intergrade (diverge from a common type) or they share the same (unique or nearly unique) character states. Evidence from external anat­omy indicates much the same thing. Thus, for the Drosophiloids, derivation from a single ancestral type is strongly indicated. There is no reason to assume that these forms were derived from more than one introduced species, and pre­sumably they were derived from a single individual. Figure 18 shows the tentative phylogeny for the Hawaiian forms. Our present discussion is concerned with the relationships within the left branch of the phylogeny shown to the left in this figure (the Drosophiloid branch of the Hawai­ian Drosophilids). The evidence discussed to this point establishes the dichotomy between the "white tip scutellum" forms and the other Drosophilids. There remains the problem of the relationships among these other species. This can­ Throckmorton: Relationships of Hawaiian Drosophilidae not be resolved with any certainty, but some general trends are apparent. As indicated earlier, the groups named in Figure 18 are groups based on external characters. While the groups themselves are probably amenable to further sub­division, they represent for the most part reasonably good phyletic lineages. The major exception to this is the "picture wings" where probably two major lineages are involved. On the basis of our present meager evidence, the phylogenetic posi­tions of these two lineages of "picture wings" seem to be so nearly the same that I have not attempted to differentiate them further. External anatomy indicates three major complexes of species among the "other Drosophiloids." These are the species with tarsal ornaments in the male, the species with modified mouthparts in the male, and the species with neither modi­fied mouthparts nor tarsal ornaments. As might be expected, the latter category is rather heterogeneous. In Figure 18 all of the species with modified mouthparts are lumped together to the top left. While there is much diversity among them, there is no concrete evidence that they are not each other's nearest relatives, and there is considerable evidence that they are. Characteristics of the ejaculatory bulb in particular tend to suggest that these species are a coherent group in spite of considerable external variation in form. The species with tarsal ornaments in the male (Trichotobregma, bristle tarsi, spoon tarsi, forked tarsi) are likewise treated a& a unit. Most of their characteristics, external as well as internal, show a mosaic pattern of distribution. While no character is exclusive to them, the characters that they do have are distributed in such a seemingly random fashion among them that no rational division of these species can be made at the present time. It is probable that more detailed studies of these forms will detect group boundaries among them. It is unlikely, however, that further studies will show them to have closer relatives than each other, but this remains for the future to determine. The other five groups (Nudidrosophila, Ateledrosophila, ldiomyia, Antopo­cerus and picture wings) fall into the negative category of not having modified mouthparts and not having tarsal ornaments. Of these five groups, four show relationships to the species with modified mouthparts; the other (Antopocerus) seems to share most of its characteristics with species having tarsal ornaments in the male. The evidence for this division can be stated briefly. At least some species of Nudidrosophila, Ateledrosophila, Idiomyia and the picture wings share characteristics of the ovipositor, ejaculatory bulb, ejaculatory apodeme, egg fila­ments, paragonia and vasa with species having modified mouthparts (see earlier descriptions of the characters). Species having tarsal ornaments in the male share characteristics of the ovipositor, ejaculatory bulb and ejaculatory apodeme with Antopocerus. In other respects Antopocerus is neutral. It is placed in the phylog­eny of Figure 18 in an equivocal position at a branch point, although the weight of evidence favors placing it with forms having tarsal ornaments. The other groups (ldiomyia, etc.) are shown as more closely related to species with modi­fied mouthparts, and no detailed relationships can be discerned among them. Many of the "miscellaneous" species are omitted from this phylogeny. They could be placed almost anywhere on the basis of present evidence. One group, represented by im/Xlriseta.e, probably is closely related to the forms from which the "white tip scutellum" species arose. Another group, represented by anomal­ The University of Texas Publication ipes, shares characteristics with Idiomyia, Nudidrosophila, etc. This last form deserves some further comment. It is the one in which the female has a true coiled ventral receptacle. The spermathecae are also of a type unique for Hawai­ian Drosophiloids (but not for Hawaiian Drosophilidae). Were it not for the anomalipes female there would be only two instead of three types of ventral receptacles and spermathecae seen among the Drosophiloids (Figures 17.1-.3; .6-.8) . It is peculiar that a form aberrant for one character should also be aberrant for another. This is particularly true when all of the other characters of the species are fully of the Hawaiian type, and many, indeed, are of highly derivative types. The derivative characters (of the ejaculatory bulb, ejaculatory apodeme, ovi­positor, egg filaments, paragonia and vasa) virtually preclude an independent origin (from a separate introduction) for anomalipes (see the characteristics of this species in Figures 1 to 16). In addition, most Hawaiian Drosophiloids have a distinctive "gestalt" that is recognizable by persons experienced with these spe­cies. To my eye, in external morphology anomalipes is an ordinary Hawaiian Drosophiloid, and M. R. Wheeler (personal communication) concurs with me in this judgment. Hence, the great majority of the characteristics of anomalipes seem to be of Hawaiian origin, and this species cannot be interpreted as a member of a separate lineage independently introduced into Hawaii. The problem this species raises is not that of the origin of the species itself. Rather it is the question of the mechanism of origin of a few "continental" character states that seem to have appeared suddenly in a lineage where they are not expressed in related species. It is possible, however, that the present sample is inadequate and that the characteristics of the spermathecae and of the ventral receptacle are more widespread among Hawaiian species than available evidence suggests. The pres­ent sample of 81 Drosophiloids includes less than a third of the known species. It is quite possible that, as this sample is enlarged, more of the "aberrant" forms will be encountered. The phylogenetic position of the Drosophiloids-There are two major phyletic lineages in the genus Drosophila. One of these includes the genus Chymomyza and the subgenus Sophophora. The other includes the subgenera Dorsilopha, Phloridosa and Drosophila. This latter also includes the genera Scaptomyza, Zaprionus, Mycodrosophila and Dettopsomyia. The subgenus Pholadoris seems to have been derived earlier in time than were the others and hence is considered the most primitive subgenus (see Throckmorton, 1962, 1965) . The character­istics of the Hawaiian Drosophiloids indicate that they were derived from near the base of the major branch leading to the subgenus Drosophila. The evidence from the various characters can be summarized as follows. The spermathecae give no evidence of phylogenetic position within the genus. The ventral recep­tacles indicate an intermediate position between forms having a folded ventral receptacle that is strongly appressed to the vagina and forms having a coiled ventral receptacle that is completely free of the vagina. In the Hawaiian forms the ventral receptacle is completely free of the vagina, coiled basally and folded distally. The folded, appressed condition is characteristic of Pholadoris, Chymo­myza and Sophophora. The free and coiled condition is characteristic of the sub­genus Drosophila, of Phloridosa, and Dettopsomyia. The interm2diate position between these groups is on the major branch leading to the subgenus Drosophila. Throckmorton: Relationships of Hawaiian Drosophilidae The egg filaments likewise indicate a position between two major groups. Gen­erally there are four egg filaments in Drosophiloid eggs but in some groups there are two. The four-filament condition is characteristic of species in the subgenus Drosophi.la, although there are many exceptions and a considerable number of species have two filaments, three filaments, or one filament. The two-filament condition is characteristic of Sophophora. Again, the phylogenetic position that reconciles these variables is on the major branch leading to the subgenus Dro­sophila . .The characteristics of the ovipositor are not very useful for determining phylogenetic position, but I have indicated some characteristic forms in Figure 17. The basic Hawaiian type, Figure 17 .33, is more nearly that of species of the subgenus Drosophi.la (Figure 17.34) than of other major groups (Figures 17.30 and .31). The paragonia are, in themselves, not strongly indicative of phylogenetic posi­tion, except that they indicate that several phylogenetic positions are improbable. They are almost certainly not Sophophoran and not from members of the quinaria section of the subgenus Drosophila. In the relatively large number of folds pres­ent in some forms they resemble species in either the virilis-repleta section of the subgenus Drosophila or possibly species in the subgenus Pholadoris. When char­acteristics of the vasa are considered, the possibilities are narrowed considerably. To date, no species other than those related to the subgenus Drosophila are known to have vasa associated with the paragonia. Almost all of the Drosophiloids do have vasa associated with the paragonia, although the association is not so strong as that seen from most species in the subgenus Drosophila. Thus, the character­istics of the paragonia and the vasa combine to indicate a position on the branch leading to the subgenus Drosophila. Neither the color nor the coiling of the testis can give any indication of phy­ logenetic position since these are variable throughout the genus. The generally simple nature of the ejaculatory bulb precludes a highly derivative position in the genus, but it does not, of itself, give positive indication of phylogenetic position. The ejaculatory apodeme likewise is of a simple, general type, excepting the special Hawaiian types that are irrelevant in this context. The evidence from these structures is consistent with a position on the branch leading to the sub­ genus Drosophila. The characters of the anterior pupal spiracles exclude the Hawaiian forms from large portions of the genus. They do not resemble the highly derivative types found in the subgenus Drosophila, and they are neither Sophophoran nor repre­sentative of Pholadoris. The spiracle is simple, but it differs from other simple spiracles (e.g., those of Pholadoris) in having heavy, erect branches instead of slender, recurved branches. It could be considered a simple derivative of the Pholadoris type, or the converse. This type of spiracle may actually be more primitive than the types seen in Pholadoris. Hence, the spiracles suggest an origin of the Hawaiian forms from a lineage prior to the development of complex derivative types. This is consistent with a position toward the base of the branch leading to the subgenus Drosophila, but, of course, it does not require that they be placed there. The Malpighian tubules of the Drosophiloids are of the type found generally The University of Texas Publication in species of the subgenus Drosophila. The tips of the posterior tubules are fused, the lumen continuous. This condition, however, is also found in species of the subgenus Pholadoris, in Chymomyza, etc. It is consistent with a position on the branch leading to the subgenus Drosophila. The abdominal sternites of the male indicate a fairly early derivation within the genus. Most of the species lack the first sternite, but it is present in some forms. When the first sternite is present, its characters are those of species in the subgenus Pholadoris. Itis not the wide, polished plate of the Sophophorans. Again, this suggests a position on the branch leading to the subgenus Drosophila, but quite close to the base and perhaps rather close to Pholadoris. In summary, characteristics of the egg filaments, ventral receptacles, and vasa strongly indicate a position on the branch leading to the subgenus Drosophila. All other characteristics are consistent with this interpretation. The character­istics of the anterior pupal spiracles and of the abdominal sternites in the male suggest a fairly early separation from this branch. The position indicated to the upper right in Figure 18 is therefore a best approximation for the phylogenetic relationships of these species. The origin of the Drosophiloids-It is fairly easy to suggest where the Dro­sophiloids did not come from. They show no close relationships to any of the major new world groups of species that are knovvn to me. Assuming that these are a reasonably representative cross-section of the new-world Drosophila, North, South and Central America can be excluded as probable sources for the Hawaiian Drosophiloids. There is little if any positive evidence for the origin of the Hawai­ian forms from the islands of the south Pacific, but the Drosophila from this part of the world are not well enough known to allow firm conclusions to be drawn. Recently (Wheeler and Takada, 1964) a number of species have been described from Micronesia and the groups present in Micronesia seem not to be good candi­dates for the ancestors of the Hawaiian species. Four subgenera, Drosophila, Hirtodrosophila, Sophophora and Scaptodrosophila, are present in the area from which Wheeler and Takada's material was drawn, and none of these is known to possess characters that would relate it closely with Hawaiian forms. It is de­sirable, however, that more material be collected from these areas, and particu­larly the higher altitudes in the high islands. It will also be necessary to obtain information on the internal anatomy of these species before final conclusions are reached. For the present, origin of the Hawaiian fauna from the islands of the south Pacific seems doubtful, but it is not absolutely precluded. Having more or less excluded the north, east, south and southwest, few direc­tions remain from which the Hawaiian forms might have come. The east coast of Asia, and particularly Japan, seems a reasonable possibility, and there is some positive evidence to suggest a derivation of the Hawaiian Drosophiloid fauna from the Japanese Drosophilid fauna. Okada ( 1956), in his study of the Japanese Drosophilids, included figures of the internal anatomy of many species. Among these species are some that appear to have characteristics similar to those seen among Hawaiian forms. In particular, the ventral receptacle, and perhaps the paragonia, of some Japanese species suggest affinities with the Hawaiian forms. This possibility is currently being investigated jointly by Dr. Okada of Tokyo Metropolitan University, Dr. Takada of Kushiro Women's College, Dr. Wheeler Throckmorton: Relationships of Hawaiian Drosophilidae of the University of Texas, and myself. The results of this study will be published later. The origin and relationships of the Scaptoids-For discussion, the Scaptoids can be divided into three groups: Titanochaeta, Scaptomyza and the three species of Drosophila that are true Scaptoids. The species of the genus Scaptomyza are, for most of the characters treated here, rather uniform. When they are not uniform, the variation is generally of such a nature that little information of phylogenetic value (for relationship inter se) is provided. Of the subgenera avail­able for this study (see Appendix), Tantalia and Alloscaptomyza are the most derivative, at least for some internal characters, or characters of the eggs. The remainder of the subgenera do not differ among themselves in any way that would suggest phylogenetic sequence. There is very little more that can be said about Scaptomyza, except that one need not postulate more than a single intro­duction to account for it, if, indeed, it is not better interpreted as originating in Hawaii. Titanochaeta is so much a Scaptoid that its origin from the basic Scapto­myza line is almost certain and no additional introduction is required to account for it. Each of the three Drosophila species departs from the more usual Scaptomyza pattern in some way. D. parva, for example, has paragonia that are fully Dro­sophiloid. Otherwise, all of its characteristics are completely Scaptoid. D. nasalis is fully Scaptoid, except that coiling of the vasa and the distribution of the pig­ment along the vasa are both Drosophiloid. Also, the ejaculatory apodeme of nasalis is intermediate in shape between the unmodified Drosophila type and that typical of Scaptomyza (Figure 17.61). D. crassifemur is almost completely Scap­toid, except that the pigmentation of the vasa containues almost to the base (Dro­sophiloid), and the ejaculatory apodeme is of the conventional Drosophila type (Figure 17.60). Also, the anterior pupal spiracle of one strain of crassifemur is completely Drosophiloid (Figure 17.66), although there is some variation within the species toward the Scaptomyza type (Figure 17.67). Only a few species of the genus Titanochaeta were available for this study, but these species were fully Scaptoid in all of their characters. (If the pupae mentioned in the earlier descrip­tions were actually pupae of Titanochaeta, then the pupal spiracles of these spe­cies are Drosophiloid) . Hence, the Scaptoids presently not classified as Scapto­myza tend to vary in the direction of Drosophila, and, for several characters (pig­mentation of the vasa, pupal spiracles and paragonia), specifically in the direc­tion of Hawaiian Drosophiloids. Since three of these species are sufficiently equivocal in their external diagnostic features to have been classified as Dro­sophila, we must investigate the possibility that the Scaptoids (Titanochaeta and Scaptomyza) originated in Hawaii from the same ancestor as did the Dro­sophiloids, and that species such as crassifemur, nasalis and parva reflect some characteristics of early transitional populations. The phylogenetic position of Scaptomyza has been discussed earlier elsewhere (Throckmorton, 1962). At that time only a few species were available, but the position indicated then is the same as that indicated by the larger sample from Hawaii. This position (see Figure 53 of Throckmorton, 1962) is toward the base of the major phyletic branch leading to the subgenus Drosophila, which is sub­stantially the position indicated for the Hawaiian Drosophiloids. The evidence The University of Texas Publication Fm. 18. Left. Preliminary phylogeny of the Hawaiian Drosophilids. Right. The phylogenetic position of the Hawaiian Drosophilids within the genus Drosophila from the Hawaiian Scaptoids for this position will not be recapitulated. The reader should note from Figure 17 that all of the considerations cited for the phylogenetic position of the Drosophiloids apply equally well for the Scaptoids, since these two groups have characters that are very much the same in most cases. Hence, it is possible to indicate a single position for the Hawaiian Dro­sophilidae, and this is shown to the upper right in Figure 18. Of itself, the observation that the Scaptoids and Drosophiloids have the same phylogenetic position need be of no great significance. It is quite possible for extant species groups having rather diverse arrays of characters to be derived from nearly the same level in a phylogeny. However, the problem of the origin of the Scaptoids is complicated by the fact that, for most of the characters of phylogenetic significance, the Hawaiian Drosophiloids and Scaptoids overlap broadly. This can be seen very plainly in Figure 17 where their characters are summarized. For only two characters, the ejaculatory bulb (not shown in Figure 17, see Figures 14 and 15) and the paragonia (Figures 17.35-.44) is there a rather sharp difference between the Scaptoids and the Drosophiloids. The para­gonia actually do overlap in characteristics more broadly than a cursory inspec­tion of Figures 11-13 might indicate, since Drosophiloid types are rare among the Scaptoids and Scaptoid types rare among Drosophiloids. The significant thing is that both types exist in each group (Figures 17.35-.44) and their frequency of occurrence is irrelevant. The types seen in Figure 17.36-.37, and in 17.40-.41, are apparently found only in Hawaii, and it is Femarkable that they are also found in both major groups from Hawaii. Much the same situation holds for many of the other characteristics. In some cases there is a simple overlap ( sperma­ Throckmorton: Relationships of Hawaiian Drosophilidae thecae and egg filaments for example). In other cases there is an overlap, and unique features are shared by the Scaptoids and Drosophiloids. The cleft and suture of the ventral egg surface (Figures 17.21-.i29) is an example of the latter. So also are the characteristics of the ventral receptacle. The "white tip scutellum" species have vasa of the Scaptoid type and they share one type of spermatheca with them. (This type of spermatheca is primitive and cannot be taken as evi­dence of close relationship.) The existence of a unique type of pupal spiracle in Hawaii, and the more or less complete intergradation of spiracle types between Drosophiloids and Scaptoids is very suggestive of a close common origin for these forms. The ejaculatory apodeme of the Scaptoids is distinctive, yet some of the Hawaiian Scaptomyza (articulata and inaequalis, Figures 15.39, .42) have nor­mal Drosophila types, and the characteristics of crassifemur and nasalis have already been mentioned in this respect. Also, most Scaptomyza have reduced and "fleshy" ovipositors, but in Hawaii some forms have ovipositors that are un­reduced (Figure 17.32). These also have the normal Drosophiloid complement and arrangement of bristles. In short, for virtually all of the characters treated here there is a broad overlap, and often there is a sharing of characteristics that are, for the present, unique to Hawaiian forms. This does not seem to be a situation that can be dismissed simply as due to convergence. Too many characters are involved, these characters are virtually identical in many cases, and apparently unique types are shared by Drosophiloids and Scaptoids. Under the circumstances, the simplest and most parsimonious con­clusion to be reached from the existing anatomical data is that the Scaptoids originated in Hawaii from the same stock as did the Drosophiloids. Alternately, however, Drosophilids might have been introduced into Hawaii twice (see Figure 18, lower left), and if this second alternative is correct a most remarkable set of coincidences was involved in the origin of the Ha'Waiian Drosophilid fauna. If the first alternative is correct, then both Titanochaeta and Scaptomyza originated in Hawaii, and Scaptomyza subsequently escaped to the mainland, either directly or through the islands of the Pacific, or both. This escape must have taken place rather early, since the genus Scaptomyza has a world-wide distribution with many species endemic to continental areas. If we envision the introduction of two forms into Hawaii, the Drosophiloid species must have been one that was so closely related to Scaptomyza that its descendents in Hawaii possess substantially the same complex of characters as do the Scaptoids. This is no mean coincidence. For the characteristics treated herein, the range of variation among the Hawaiian Drosophilids (Drosophiloids plus Scaptoids) is of the same order of magnitude as that encountered within some species groups from elsewhere in the genus Drosophila. It is considerably less than is encountered within the major subgenera, and some species groups (e.g., the repwta group) seem to be more variable, for more of these character­istics, than are the Hawaiian Drosophilids (see Throckmorton, 1962 for support­ing data) . Elsewhere in the genus, then, such cohesive character complexes are strong evidence of close common descent, and they must be interpreted as indicat­ing the same thing here. Regardless of the geographical point of origin of the Scaptoids, present evidence requires that they share a close common ancestor with the Ha'Waiian Drosophiloids. If the origin of the Scaptoids occurred outside of The University of Texas Publication Hawaii, then two successful trans-Pacific colonizations are needed. Each one of these was, in itself, an improbable event. That a successful introduction be made twice from the same family of Diptera is even less probable, and that the two successful introductions from the same family should involve species so closely related (but presumably already generically distinct) as to produce the existing patterns of variation in Hawaii is less probable still. And if we do not postulate extremely close relationships between the two original colonizers, we must then explain the broad overlaps between the Drosophiloids and Scaptoids as due to convergent evolution. This requires that we explain why the Scaptoids should have diverged in tne direction of Drosophila (while they were competing with them among the vacant niches of .Hawaii), and why, for some character states at least, they have happened to diverge specifically in the direction of the Hawai­ian Drosophiloids. And why do the Hawaiian Drosophiloids, of all the Drosophi­loids in the world, include a group of species ("white tip scutellum" forms) that share characters with the Scaptoids, their nearest (geographical) relatives? Fi­nally, Spieth (this Bulletin) concludes, on the basis of his studies of behavior, that the "white tip scutellum" species are Scaptoid and that the genus Scaptomyza probably originated in Hawaii. For the present, then, Hawaii must be considered to be the only place in the world where the otherwise sharp distinctions between Scaptomyza and Drosophila tend to disappear. The pattern of variation seen among the Hawaiian Drosophilids is so readily interpreted as being the uncomplicated consequence of divergence from a single ancestral colonizer that alternative explanations seem labored and uncalled-for. If we accept this interpretation, a simple and logical sequence of evolution is seen for the species in Hawaii. This last interpretation does require that at least one species of Scaptomyza escaped from Hawaii, but this is not a serious draw­back. Both of the major alternatives require two trans-Pacific colonizations. One requires two separate introductions into Hawaii. The other requires one intro­duction into Hawaii and one introduction from Hawaii to a continent. There is no reason to think that one of these trips is inherently more difficult or more improbable than the other, and such considerations cannot help us establish the relative merits of the two major alternatives. Another consideration, that of time, has yet to be broached. The most recent and informative treatment of the ages of the Hawaiian vol­canoes is that of McDougall ( 1964). He has determined the potassium-argon ages of these as follows: (in millions of years) Kauai, 5.6-3.8; Oahu, 3.4-2.2; Molo­kai, 1.8-1.3; Maui, 1.3-0.8; Hawaii, <1. He interprets these ages as showing the order of extinction. However, the evidence also indicated that the eruption of the presently exposed lavas was very rapid, so the order of extinction may well represent the order of commencement. One of the most interesting observations was that one sample (Mauna Kuwale trachyte) from West Oahu had an age of about eight million years (two readings of 8.26 and 8.46, respectively). This suggests that two volcanoes differing in age by five million years were active in West Oahu, with the earliest known activity dating from the early to middle Pliocene. The activity that produced the presently existing islands, including the Waianae Range of West Oahu, was late Pliocene to Pleistocene. The islands from which these datings were obtained belong to the "windward" group. These are but the southeastern members of a chain that extends for about 1900 miles from Hawaii to a coral atoll off.to the northwest. The northwestern members, the "leeward" islands, are thought to be the remnants of earlier high islands (see Zimmerman, 1948) and these presumably would have been older than the existing islands of the "windward" group. Bearing these factors in mind, an estimate of ten million years as the age of the older members of the archipelago seems justified. It is not unreasonable to assume that the western islands were inhabitable and inhabited at one time, and that some of the evolution of the endemic Hawaiian Drosophilidae occurred on them. If we grant the first five million years for the development of a flora adequate to support the Drosophilids and assume the successful colonization toward the end of this period, then there would be about five million years available for the development of the endemic Drosophilid fauna. If we assume only three generations per year as a probabl~ minimum estimate for the species and climate involved, then perhaps fifteen million generations of evolutionary change were possible. Since the western-most islands may have been much older than the eight million years estimated for West Oahu, and since less than five million years may have been required for the original successful introduction, fifteen million generations is probably a mini­mum estimate. At any rate, it is a conservative estimate, and hopefully it does not err in the wrong direction. If one were to generalize about evolutionary rates, the best that can be said is that they vary widely, both from lineage to lineage and at different times within the same lineage. There is almost no concrete evidence from which we might estimate evolutionary rates in Drosophila (Simpson, 1945). The recent descrip­tion of a fossil Drosophilid about thirty million years old (Wheeler, 1963) gives us an indication of the age of the genus DrosophiUi and of the major phyletic branches within it, but it does not tell us much about rates of evolution. Simpson (1949) suggests 50,000 years or more for the production of species "fully distinct genetically and morphologically," and he indicates that this may be for cases of rather rapid evolution. For lack of a better, we can take this estimate and see where it leads us. If speciation can occur within 50,000 years, and if five million years are avail­able, then 100 speciation events in sequence are "possible." For the sake of arith­metic, if we assume all speciation to have been synchronous and dichotomous, then 2100 species could have been produce:l in the time available. However, we need time only to produce about 210 species (say 600 endemic Drosophilids plus less than 200 Scaptomyza from elsewhere in the world), for which only 500,000 years might have been sufficient. Apparently the rate could be tenfold slower and still produce the requisite number of species. And while speciation would certainly not be synchronous, it most probably would not be strictly dichotomous either. In many instances speciation would almost certainly produce more than two products. Given environmental opportunity, and such opportunity does seem to have existed in Hawaii, a thousand surviving species in five million years may be a very reasonable estimate. In fact, environmental opportunity may be by far the most critical factor regulating rates of speciation. In its absence the rates of change would be low, but in its presence they may be far higher than educated guesses suggest. At any rate, and granting that we are vastly ignorant of specia­ The University of Texas Publication tion rates in Drosophila, time does not provide us with a critical means for dis­tinguishing between the two major alternatives. Certainly considerations of time alone do not force us to reconsider the implications of the anatomical data for the origin of the genus Scaptomyza. The final resolution of this question must depend on very thorough anatomical studies, particularly of the continental species of Scaptomyza. One bit of evidence of this nature tends to cast some doubt on the interpretation of an island origin for Scaptomyza. Wheeler ( 1952, p. 204) reports that the male of Scaptomyza montana ('Western U.S.) has elliptical testes. This seemingly trivial observation is of some consequence for the present discussion. To date no species of Hawaiian Scaptomyza are known to have elliptical testes. This condition is primitive, and if Scaptomyza originated in Hawaii it is surprising that some Hawaiian species do not also have elliptical testes. It is true that some of the "white tip scutellum" species have less than one complete coil in the testis, and these species are Scaptoid in other respects, but still the absence of elliptical testes among Hawaiian forms is puzzling if Scaptomyza originated there. The entire sequence would require that the original colonizer of Hawaii carried genetic variability sufficient to pro­duce either coiled or elliptical testes. Subsequently species were produced that had coiled testes, but no species with elliptical testes survived in Hawaii. How­ever, the colonizer from Hawaii to the continent still carried genetic variability that could produce elliptical testes, and one of its descendents on the mainland does have elliptical testes. In short, the genetic variability for testis form must have survived two bottlenecks if it is to be expressed in mainland forms, and it is not known to have been expressed at the first stopping point in Hawaii. It is possible, of course, to rationalize these observations. In the first place, only a fraction of the Hawaiian Scaptoids have been investigated. Species in which the male has elliptical testes may yet be discovered there. Second it is not abso­lutely necessary that a Hawaiian form with elliptical testes survive to the present. All that is required is that one persisted to the time when the Scaptomyza colo­nizer left the islands. Evolution in Hawaii may well have involved the sequential replacement of one species by several derivative descendent species, and hence we have no definitive evidence regarding the contents of the Hawaiian gene pools at the time the second colonization occurred. Finally, it is not strictly necessary that the reappearance of an ancestral character result from segregation. Atavistic recurrences may well be due to segregation of rare complex genotypes from gene pools in which the ancestral genetic elements still persist at low frequencies. They may also be due to a simple mutational change, say in a regulator gene, that suppresses, or fails to induce, more derivative developmental sequences. Hence, an ancestral trait may reappear through mutation, and its appearance may not be critical evidence of the evolutionary history of the species. Rationalizations are never very satisfactory for deciding issues, and they indicate only that the question must remain open until more concrete evidence is available. For the purposes of discussion, and because the bulk of the existing evidence so indicates, an origin of Scaptomyza in Hawaii will be assumed in the next section. The evolution of the Hawaiian Drosophilids-1£ we assume a single coloniza­tion of Hawaii, the following general outline of events can be adduced. We would expect that the first individual encountered an environment that was not only Throckmorton: Relationships of Hawaiian Drosophilidae hospitable but full of empty niches and empty of competitors. Under the circum­stances, its immediate descendents probably were not under strong selection pressure, and they probably found two major habitats open to them. One of them was the usual Drosophila habitat, woodland with its associated fungi, flowing sap, fruits, rotting vegetation, etc. The other was the open and semi-open grass­lands whose niches elsewhere in the world are filled by species from other families of Diptera (the leaf-mining Agromyzidae, the Chloropidae, etc.) . Under these circumstances, the first major separation may well have involved a dichotomy between the major habitats, leading to two groups, one primarily woodland, the other grassland. The woodland habitat is characteristically that of Drosophila and probably required little in the way of modification to adapt to it. These forms retained the basic Drosophila way of life and also retained the general Drosophila morphology and behavior. They became the Drosophiloids of this paper. The other forms adapted to a new environment (grassland and semi-grassland) and became modified accordingly. In making the transition from one way of life to another, and from one food source to another, some forms switched to become parasitic on spider eggs (Titanochaeta ) and others eventualy became Scapto­myza. This adaptive shift was accompanied by morphological changes and the distinctive Scaptomyza "gestalt" was the result. The occupation of the leaf-mining niche by Scaptomyza has always puzzled me, since other families of Diptera seem to be firmly established in at least super­ficially similar niches. (This assumes that adaptations of families are more ancient than those of genera, which may not be the case in this instance.) How­ever, if Scaptomyza first entered this niche in the comparative seclusion of Hawaii, and if it perfected its adaptation to this niche in the absence of competition from already well-adapted species, then its position may not be surprising. Once estab­lished in the niche, it may have been able to compete successfully with other Dipteran species when it finally came in contact with them, and this may account for its wide distribution on the mainland. Physiologically it would have been quite different from its competitors in other families, and its adaptations could hardly have precisely duplicated those of continental competitors. Hence, when it "re­turned" to the mainland it may have "seen" many niches that may not have been accessible to an earlier Drosophilid. Be that as it may, the next significant event was the separation of the "white tip scutellum" line. The nature of the events that produced the rather pronounced differences of these species is, of course, unknown. Apparently they are fungus feeders (Hardy, 1965), and they may reflect a second major partition of the available food sources. The first would have been between woodland and grass­land; the second, within woodland, between fungus feeders and non-fungus feed­ers. It is probable that a bottleneck of some kind was passed at their origin. At least, the sharp differences in egg filaments and spermathecae implicate some­thing of the sort, and their distinctive external gestalt is perhaps indicative of the same thing. Subsequent to the separation of the "white tip scutellum" forms, the evolution of the Drosophiloids appears to have involved two major patterns, not necessarily sharply distinct from each other. One of these patterns, most evident among the more recent forms, is the orthodox one of differentiation in isolation. This is par­ The University of Texas Publication ticularly clear for the subgenus Exalloscaptomyza, where there may be one species for each of the major islands. There are also a number of species pairs (or triplets, etc.) whose distribution indicates much the same thing. Thus we find, for example, con;ectura from Kilauea (Hawaii), and "conjectura-like" from Mauna Kea on the same island. We also find fungicola and "fungicola-like," mimica and "mimica-like," etc., in the same two localities, as if the same original "community" had been established in each place, with subsequent species diver­gence producing the present faunas. Alternately, this situation could have been produced by independent colonizations from several sources. The Drosophiloids of Hawaii offer unusual opportunities for studying the evolution of community relationships, niche separations and adaptations together with an analysis of speciation events involving forms whose genetic relationships are very close and which may ultimately be determined. The second major pattern follows from the wide, and almost random, distri­bution of many of the character states peculiar to Hawaiian Drosophiloids. It is as if a population originated, became widespread, developed a great store of genetic variability, and then fragmented. The resulting populations, regardless of their mode of origin or the factors keeping them separate, had much of the original genetic variability in common. They could not and did not utilize it all, and different lineages used different combinations. Thus, no strong evolutionary trends are apparent among the more recent Drosophiloids. A pattern of reticulate evolution of this sort might also be produced by the differentiation of populations in semi-isolation (imperfect isolation by distance) , and the Hawaiian environ­ments and the habits of Drosophila are particularly well suited to this type of evolution. One might visualize an original population slowly spreading to occupy the available geographic range of niches open to it. Local populations of consider­able size might build up, but favorable environments might be rather widely separated from each other, as many, if not most of them, are today. Thus, gene flow between the local populations would be slight and sporadic at best. The general situation could be visualized as a net, with each knot a local population, each thread the evidence of gene flow between them. Early in the sequence gene flow might be reasonably constant between the populations, but as time passed, gene flow, as distinct from immigration, might gradually diminish as a conse­quence of the genetic divergence of the populations involved. Alien genes might be selectively eliminated from each gene pool, or alien behavior patterns (re­gional dialects) might reduce the probability of mating even if immigrations occurred at appropriate seasons. Under such circumstances, and even with con­tinued introduction of individuals from other populations, divergence and even­tual speciation might occur. In the net visualized earlier, the knots would persist but the webs between might gradually and sporadically break. Some would break earlier than others, and some might never break. Newly originating genotypes might diffuse readily through some populations, but only poorly or not at all to others. And at various times along the way some populations might themselves begin to spread and establish new local populations of their own (without necessarily displacing the originals) so that new cycles might well be under way long before the original cycle, with well-established reproductive isolation between all the remaining populations, had been completed. The net effect of such Throckmorton: Relationships of Hawaiian Drosophilidae a system would be to produce complexes of species only slightly differentiated from each other, and with the complexes themselves only partially distinct. One group would overlap another for some character states, have certain other states in common with a second group, and have still others in common with a third group, etc., etc. This is substantially what exists in Hawaii today, and such a product seems more likely to have been produced by highly multiplicative split­tings and fragmentations than through conservative dichotomies. SUMMARY Various anatomical features of the Hawaiian Drosophilids are described and discussed. Two major groups are indicated, the Drosophiloids and the Scaptoids. Phylogenetically, both of these groups are very closely related to each other and they are derived from near the base of the major branch leading to the subgenus Drosophila of the genus Drosophila. Evidence is presented indicating that the genus Scaptomyza originated in Hawaii and arguments for and against this interpretation are discussed. If Scaptomyza originated in Hawaii, then the avail­able evidence favors the introduction of only a single individual (basically a Drosophila) as the progenitor of the more than 400 endemic species of Dro­sophilids. At most, two introductions, presumably of a single individual each; are required if the Scaptoids are thought to have originated from an introduction separate from that for the Drosophiloids. Existing evidence indicates that the Drosophiloid introduction was from east Asia, perhaps from Japan, but this problem is still under investigation. Some general patterns of evolution in Hawaii are also discussed. ACKNOWLEDGMENT I have, at various points in the text, acknowledged the contributio;ns of differ­ent individuals to this work. I would like again to recognize my indebtedness to these people. In addition, I would like to make special mention of the extensive cooperation, advice and assistance of Dr. D. Elmo Hardy, both in the conduct of field work and in the organization and operation of laboratory facilities at the University of Hawaii. Without his thoughtful and kind cooperation this project could not have proceeded so expeditiously, and it probably would not have been possible at all. I am much in his debt and grateful for all the assistance he has given. I also wish to thank Mrs. Linda Kuich for her help in preparing the figures for publication. APPENDIX The species dissected for this study are listed below. Their present classifica­tion is indicated, together with informal designations that serve to clarify general groupings. The identification number relates specific individual specimens in my collection to .the notes on their dissection. When specimens from,more than one locality were dissected, the number given is that for the specimens whose charac­teristics are reported in this paper. The University of Texas Publication APPENDIX Identification Species number Locality Genus: IDIOMYIA obscuripes Grimshaw perkinsi Grimshaw picta Grimshaw Genus: ANTOPOCERUS aduncus Hardy diamphidiopodus Hardy diamphidiopodus Hardy longiseta (Grimshaw) orthopterus Hardy orthopterus Hardy tanythrix Hardy villosus Hardy Genus: NUDIDROSOPHILA aenicta Hardy lepidobregma Hardy Genus: ATELEDROSOPHILA preapicula Hardy Genus: DROSOPHILA Subgenus: Drosophila Miscellaneous species anomalipes Grimshaw caccabata Hardy hirtitibia Hardy imparisetae Hardy imparisetae Hardy quasianomalipes Hardy truncipenna Hardy picture wings adiastola Hardy crucigera Grimshaw engyochracea Hardy fasciculisetae Hardy grimshawi Oldenberg musaphilia Hardy picticornis Grimshaw pilimana Grimshaw punalua Bryan villosipedis Hardy modified mouthparts aquila Hardy araiotrichia Hardy asketostoma Hardy chaetopeza Hardy DROSOPHILOIDS 85 53 99 55 54 134 98 73 59 19 94 135 43 89 20 83 30 61 72 136 93 105 2 137 138 139 140 141 76 6 95 24 Paliku, Maui W aikamoi, Maui W aikamoi, Maui Waikamoi, Maui W aikamoi, Maui Puu Kolekole, Molokai Puu Kolekole, Molokai Paliku, Maui W aikamoi, Maui Kilauea, Hawaii Waikamoi, Maui Drum Drive, Oahu Kipuka Ki, H awaii Opaeula Ridge, Oahu Kokee State Park (Mohihi Stream), Kauai Puu Kolekole, Molokai Drnm Drive, Oahu Kilauea, Hawaii Kipuka Puaulu, Hawaii Kokee State Park (Halemanu Valley), Kauai Waikamoi, Maui Waikamoi, Maui Mt. Tantalus, Oahu Kipuka Puaulu, Hawaii Waikamoi, Maui East Molokai Kipuka Puaulu, Hawaii Kokee State Park, Kauai Mt. Tantalus, Oahu Pupukea, Oahu Kokee State Park (Kumuwela Ridge) , Kauai Kilauea, Hawaii Puu Kolekole, Molokai Haleakala Crater, Maui Kipuka Puaulu, H awaii comatifemora Hardy 34 W aikamoi, Maui Throckmorton: Relationships of Hawaiian Drosophilidae Identification Species number Locality coniectura Hardy 22 Kipuka Puaulu, Hawaii dissita Hardy 78 Kilauea, Hawaii eurypeza Hardy 142 Kokee State Park (Alakai Trail), Kauai flavibasis Hardy 143 Kokee State Park, Kauai freycinetiae Hardy 144 Mt. Tantalus, Oahu furvifacies Hardy 38 Kokee State Park (Mohihi Stream), Kauai hirticoza Hardy 113 Paliku, Maui infuscata Grimshaw 23 Mud Lane, Hawaii involuta Hardy 145 Paauilo Experiment Station, Hawaii ischnotriz Hardy 151 Pupukea, Oahu ischnotriz Hardy 1 Mt. Tantalus, Oahu kauluai Bryan 63 Pupukea, Oahu mimica Hardy 146 Kipuka Puaulu, Hawaii mimica Hardy Paauilo Experiment Station, Hawaii mycetophila Hardy 147 Mt. Tantalus, Oahu pychnochaetae Hardy 9 Pupukea, Oahu residua Hardy 103 Kipuka Ki, Hawaii scolostoma Hardy 96 Paliku, Maui bristle tarsi apodasta Hardy 41 Kokee State Park (Alakai Trail), Kauai basimacula Hardy 44 Kokee State Park (Mohihi Stream), Kauai ezpansa Hardy 68 W aikamoi, Maui perissopoda Hardy 40 Kokee State Park (Mohihi Stream), Kauai prodita Hardy 112 Paliku, Maui redunca Hardy 87 Puu Kolekole, Molokai seclusa Hardy 3, 91 Puu Kolekole, Molokai torula Hardy 69 Waikamoi, Maui trichaetosa Hardy 77 Kilauea, Hawaii spoon tarsi conformis Hardy 27 Mud Lane, Hawaii conformis Hardy Kilauea, Hawaii contorta Hardy 66 W aikamoi, Maui disticha Hardy 10 Puu Kolekole, Molokai disticha Hardy Waikamoi, Maui incognita Hardy 82 Kilauea, Hawaii neutralis Hardy 148 Kilauea, Hawaii polliciforma Hardy 28 Mud Lane, Hawaii sordidapez Grimshaw 81 Kilauea, Hawaii sordidapez Grimshaw 17 Kulani Road, Hawaii split tarsi ancyla Hardy 49 Waikamoi, Maui clavata Hardy 26 Pupukea, Oahu fundita Hardy 48 South of Hanalilolilo, Molokai pectinitarsus Hardy 25 Kipuka Puaulu, Hawaii white tip scutellum cilifemorata Hardy 52 Waikamoi, Maui fungicola Hardy 14 Paauilo Experiment Station, Hawaii fungicola Hardy 152 Kipuka Puaulu, Hawaii The University of Texas Publication Identification Species number Locality haleakalae Grimshaw iki Bryan melanoloma Hardy melanoloma Hardy melanosoma Grimshaw melanosoma Grimshaw nanella Hardy nigra Bryan bipolita Hardy canipolita Hardy demipolita Hardy Subgenus: Trichotobregma petalopeza Hardy Genus: DROSOPHILA crassifemur Grimshaw crassifemur Grimshaw crassifemur Grimshaw crassifemur Grimshaw nasalis Grimshaw nasalis Grimshaw parva Grimshaw Genus: TITANOCHAETA contestata Hardy ·species C Genus: SCAPTOMYZA Subgenus: Alloscaptomyza longisetosa Hackman stramineifrons Hackman Subgenus: Bunostoma anomala Hardy palmae Hardy palmae Hardy zantho pleura Rardy Subgenus: Ezalloscaptomyza mauiensis (Grim. ) species? species? speci2s ? species? Subgenus: Parascaptomyza pallida (Zett.) Subgenus: Rosenwaldia abrupta Hackman Subgenus: Tantalia varipicta Hardy 107 65 108 33 46 31 50 149 150 16 97 SCAPTOIDS 79 80 153 154 155 156 157 13 8 158 159 160 161 162 163 164 165 166 167 168 169 120 170 Paliku, Maui W aikamoi, Maui Paliku, Maui Puu Kolekole, Molokai Kokee State Park (Halemanu Valley), Kauai Kokee State Park (Kumuwela Ridge) , Kauai Koke2 State Park (Halemanu Valley) Kauai Waikamoi, Maui Mt. Tantalus, Oahu Pupukea, Oahu Paauilo Experiment Station, Hawaii Paliku, Maui Kilauea, Hawaii Puu Kolekole, Molokai Kokee State Park, Kauai Paliku, Maui Paliku, Maui Puu Kolekole, Molokai Kokee State Park (Mohihi Stream), Kauai Drum Drive, Oahu Pupukea, Oahu Mt. Tantalus, Oahu Mt. Tantalus, Oahu Kokee State Park (Mohihi Stream), Kauai Mt. Tantalus, Oahu Kamuela, Hawaii Mt. Tantalus, Oahu Iao Valley, Maui Kamuela, Hawaii Pali H ighway, Oahu Molokai Puu Ka Pele, Kauai Pohakaloa, Hawaii Waikamoi, Maui Kokee State Park, Kauai Identification Species number Locality Subgenus: Trogloscaptomyza argentifrons Hardy 171 Kokee State Park (Mohihi Stream), Kauai articulata Hardy 172 Pohakaloa, Hawaii connata Hardy 121 Kokee State Park (Mohihi Stream) , Kauai hackmani Hardy 116 Puu Kolekole, Molokai hackmani Hardy Kipuka Puaulu, Hawaii inaequalis (Grim.) 174 Pohakaloa, Hawaii intricata Hardy 125 Waikamoi, Maui intricata Hardy Puu Kolekole, Molokai latitergum Hardy 139 Haleakala, Maui levata Hardy 122 Kokee State Park (Mohihi Stream), Kauai re:usa Hardy 127 W aikamoi, Maui rostrata Hardy 118 Kokee State Park (Kumuwela Ridge), Kauai S:Zvicola Hardy 126 Waikamoi, Maui LITERATURE CITED Hardy, D. E. 1965. Insects of Hawaii. Vol. 12. Drosophilidae. University of Hawaii Press, 814 pp. -----. Descriptions and notes on Hawaiian Drosophilidae (Diptera). This Bulletin. McDougall, Ian. 1964. Potassium-argon ages from lavas of the Hawaiian Islands. Bull. Geol. Soc. Amer. 75: 107-128. Okada, T. 1956. Systematic study of Drosophilidae and allied families of Japan. Gihodo Co. Ltd., Tokyo, Japan. 183 pp. Perkins, R. C. L. 1913. Introduction. Being a review of the Land-Fauna of Hawaii. In Fauna Hawaiiensis (Ed. by David Sharp). Vol. I: xv-ccxxvii. Cambridge University Press. 759 pp. Simpson, G. G. 1945. Evidence from fossils and from the application of evolutionary rate dis­tributions. In, Symposium on age of the distribution pattern of the gene arrangements in Drosophila pseudoobscura. Lloydia 8: 103-108. -----. 1949. Rates of evolution in animals. In, Genetics, Paleontology and Evolution. Edited by G. L. Jepsen, G. G. Simpson and E. Mayr. Princeton Univ. Press. Spieth, H . Courtship behavior of endemic Hawaiian Drosophila. This Bulletin. Stern, C. 1940. Growth in vitro of the testes of Drosophila. Growth, 4: 377. Throckmorton, L. H. 1962. The problem of phylogeny in the genus Drosophila. Studies in Genetics II. Univ. of Texas Puhl. No. 6205: 207-343. 1965. Similarity versus relationship in Drosophila. Systematic Zoology 14: 221­ 236. Wheeler, M. R. 1952. The Drosophilidae of the nearctic region exclusive of the genus Dro­sophila. Studies in the Genetics of Drosophila VII. Univ. of Texas Puhl. 5204: 162-218. 1960. Sternite modification in males of the Drosophilidae (Diptera). Annals of the Ent. Soc. of Amer. 53: 133-137. The University of Texas Publication 1963. A note on some fossil Drosophilidae (Diptera) from the amber of Chiapas, Mexico. Jour. Paleont. 37: 123-124. Wheeler, M. R., and H. Takada. 1964. Insects of Micronesia. Diptera: Drosophilidae. Insects of Micronesia 14: 163-Z4Z. Published by the Bernice P. Bishop Museum, Honolulu, Hawaii. Wheeler, M. R., and L. H. Throckmorton. 1960. Notes on Alaskan Drosophilidae (Diptera), with the description of a new species. Bull. Brooklyn Ent. Soc. 55 (5): 134-143. Zimmerman, E. C. 1948. Insects of Hawaii. Vol. I. Introduction. University of Hawaii Press, Honolulu. Z06 pp. -----. 1958. 300 species of Drosophila in Hawaii?-A challenge to geneticists and evo­lutionists. Evolution 12.: 557-558. XII. Preliminary Report on the Karyotypes of Hawaiian Drosophilidae1 2 FRANCESE. CLAYTON This report includes a list of the karyotypes of 45 species of the Drosophilidae of Hawaii. The metaphase configurations were obtained from larval brain ganglia which had been stained for 15-20 minutse in acetoorcein and squashed in fifty per cent acetic acid. Since, in most cases, only single larvae were available for the cytological analysis, no detailed study has been made by the author on the salivary chromosomes of the various species. The sex of the larvae could not be determined by the usual method, so detailed studies of the sex chromosomes have not been made. RESULTS AND DISCUSSION The cytological results of this investigation are given in Table 1, with meta­;ihase configurations of some species illustrated in Plates I and U. The table includes collection numbers, localities, metaphases, and any information obtained from salivary gland preparations. Within the genus Drosophila, 34 species are listed; of these, twenty-six had the primitive configuration of five rods and one dot. Among the other species in this genus, four had karyotypes consisting of six rods; the remainder had metaphases including one or two V-shaped chromosomes. Two types of metaphases were found among the Scaptomyza species examined; these consisted of three rods, one V, and one dot or one rod, two V's, and one dot. Among the eleven species studied, nine had one pair of V-shaped chromosomes while only two species had two pairs of V's. Some of the modifications within the genus Drosophila are shown in Figures 1-16. The metaphases for D. crucigera and D. grimshawi (Figs. 1,2) are typical of most of the species listed in Table 1. One modification of this primitive con­dition which occurred among karyotypes of Drosophila species was the addition of heterochromatin to one pair of rods or to the dots. Extra-length rods were observed in D. eurypeza, D. freycinetiae, D. basimacula?, and D. kauluai (Figs. 4-7) . Modifications in the size of the dots are illustrated in Figures 8-12; D. pectinitarsus had larger dots than most species and Antopocerus sp., D. asketo­stoma, D. mimica, and D. coniectura had six pairs of rods. The presence of five arms and one dot in salivary gland preparations of mimica and freycinetiae indi­cate that the modifications were probably by the addition of heterochromatin to the rod or the dot element. Metaphases of the four species with one or two pairs of V's are shown in Figures 13-16. The metaphase of D. crassifemur consisted of one pair of V's, three pairs of rods, and one pair of dots. D. rxirva and D. ischno­trix had two pairs of V's, one pair of rods, and one pair of dots, while D. nasalis 1 This investigation was supported, in part, by Public Health Service Research Grant No. GM-1064-0 from the National Institutes of Health, and grant GB-711 from the National Science Foundation. 2 Departm~nt of Zoology, The University of Arkansas, Fayetteville. (.)..> (.0 TABLE 1 (X) Species Collection number I.ocalily l\Ietaphasc Salivary Scaptomy::,a Bunostoma anomala palmae varifrons xanthopleura n. sp. near anomala Unidentified species Exalloscaptomyza spp. C73.1 BSBA BMLHT BSBK BSBP BSBW BDIIK BDIIML BDIIW BDIIC BDIIKM BDIIP CHt.50 T200 WH35.10 C53.7 (BSAT) BSAP BSAM BSAW BDIM BDITa BDITb BDIW C53.29. C54.8 C53.t Kokee Park, Kauai Alakai Swamp, Kauai Mohihi, Kauai Kohala Mts., Hawaii Pohakaloa, Hawaii W aimea, Hawaii Kamuela, Hawaii Moanaloa Rd., Hawaii W aimea, Hawaii Capt. Cook, Hawaii Kohala Mts., Hawaii Pohakaloa, Hawaii Paliku, Maui Tantalus, Oahu Pupukea, Oahu Tantalus, Oahu Poamoho, Oahu Pupukea, Oahu Manoa Falls, Oahu Waimea, Hawaii Manoa Valley, Oahu Tantalus, Oahu (Datura) 'lantalus, Oahu (Koa) \Vaimea, Hawaii Tantalus, Oahu r.Ian:ia Valley, Oahu T;:ntalus, Oahu 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, tD 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, tD !R, 2V, 1D 1R,2V, 1D 1R, 2V, tD 1R,2V, 1D !R, 2V, ID 3R, IV, 1D 3R, IV, 1D ....., ;::s-. ~ ~ ;::i ;:::· ~ ..... -"' - ...... ~ 0 ....., " ~ l::i "' "ti i::: ...._ ""' -. (") l::i 5· ;::i Trogloscaptomyza argentifrons waialealeae n. sp. spp. Tantalia albovittata Drosophila crucigera HS T90 WH34.16 WH31D.18 WH47.13 C72.8 CH72.11 WH35.9 WH34.13 C53.15 C53.11 WH19.9~ X WH16.6~ WH8.8 WH32.11 WH24.1 CH6.9 CHB.Z C63.4 WH34.2 C64.6 WH29A.7 WH31.2 WH29A.7, WHZ9C.7 Kona Plantation, Hawaii Waimea, Hawaii Pupukea, Oahu Kokee, Kauai Kauai Molokai Kumuwela, Kauai Alakai Swamp, Kauai Alakai Swamp, Kauai Pupukea, Oahu Pupukea, Oahu Tantalus, Oahu Tantalus, Oahu Tantalus, Oahu Tantalus, Oahu Tantalus, Oahu Pupukea, Oahu Pupukea, Oahu Pupukea, Oahu Pupukea, Oahu Pupukea, Oahu Sacred Fall, Oahu Kokee, Kauai Halemanu, Kauai Halemanu, Kokee, Kauai 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 1R, 2V, tD 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 3R, IV, 1D 5R,1D 5R,1D 5R,1D 5R, 1D 5R, 1D 5R,1D 5R, 1D 5R, 1D 5R,1D 5R,1D 5R,1D 5R, 1D 5R,1D 5A, ID 5A,1D 5A, 1D (J S" ~ .... c ~ ~ ~ 0 ~ ~ ~ .Q. g= 2 ..... ..... l::i ;::i tJ d .g"' ;::.. ..... - ~ ~ (JV <.O <.O ..... 0 0 Species Collection number Locality i\Jetaphase Salivary punalua CH6,CH8 Pupukea, Oahu 5R,1D LHT Pupukea, Oahu 5R, 1D WH34.1 Pupukea, Oahu 5R,1D C53.17 Tantalus, Oahu 5R, 1D C53.17,C62.4, pilimana C63.11 CS3.3 Oahu Tantalus, Oahu 5R, 1D SR,1D 5A,1D ~ ;.::r­(I) CS3.3,CH6.8 Tantalus-Pupukea, Oahu 5R,1D SA, 1D (unpaired regions) ~ grimshawi C104.1 LHT WH6.8 Waikamoi, Maui Molokai Lanai 5R,1D 5R,1D SR,1D 5A, 1D (heterozygous inversion) ;::i-. <::::! (I) ;;;-· picticornis C73.13 Kokee, Kauai 5R,1D ..... ~ engyochracea WH20.18 Bird Park, Hawaii SR,1D .Q. adiastola C104.7 Waikamoi, Maui 5R,1D ~ Paliku, Maui SR,1D (I) ~ villosipedis C99.3 Kokee, Kauai SR, 1D ~ "' spectabilis C107.2B Puu Kolekole, Molokai 5R,1D ~ eurypeza freycinetiae pychnochaetae C72.4 WH31B.14 WH8.7 C92.6 Alakai Trail, Kauai Kumuwela Rd., l<,auai Tantalus, Oahu Pupukea, Oahu 5R,1D 5R,1D 5R,1D 5R,1D SA, 1D :.::: ""' ........-.(') ~ ..... c;· chaetopeza C68.23 Kilauea, Hawaii 5R, 1D ;::i pectin: tarsus WH26.3 Bird Park, Hawaii 5R,1D mycetophila CH8.7 Pupukea, Oahu SR, tD 5A, 1D WH32.8 Tantalus, Oahu SR,1D flavibasis WH29C.16 Halemanu, Kauai 5R, 1D kauluai WH34.6 Pupukea, Oahu 5R, 1D mimica LHT Bird Park, Hawaii 6R SA, 1D CH70.3 Kipuka Ki, Hawaii 6R C67.10 Bird Park, Hawaii 6R C67.11 Bird Park, Hawaii 6R 5A, tD C69.4 Bird Park, Hawaii 6R WH20.1 Bird Park, Hawaii 6R (""') Honakaa, Hawaii 6R S" con;ectura WH20.2 WH20.2 Bird Park, Hawaii Bird Park, Hawaii 6R 6R ~ ...... 0 ~ asketostoma Haleakala, Maui 6R ~ ischnotrix WH15.2 Tantalus, Oahu 1R,2V,1D 5A, 1D ;:i C53.2 Tantalus, Oahu 1R,2V,1D 5A, 1D ~ 0 WH33.2 Tantalus, Oahu 1R,2V,1D ...... parva crassifemur WH35.7 WH26.7 WH47.14 CH74.7 Pupukea, Oahu Bird Park, Hawaii Kumuwela, Kauai Mohihi, Kauai 1R,2V,1D 1R,2V,1D 1R,2V,1D 3R, IV, tD 5A, tD (heterozygous inversion) ~ ~ .Q.. ::i:: ~ nasal is Ci 11.10 n. sp. near hirtitarsus WH45.2 n. sp. near larifuga complex WH34.15 Hanaula, W . Maui Paauilo, Hawaii Pupukea, Oahu 2R,2V,1D 5R,1D 5R,1D (heterozygous inversion) 2-·iS. ;::s Unidentified species "modified proboscis" WH47.9 C71 "dark scutellum" C99 "large white-tip scutellum" C71.22 Kumuwela, Kauai Kokee, Kauai Kokee, Kauai Kokee, Kauai 5R, 1D 5R,1D 5R,1D 5R,1D tl '1 0 "' .g;:::-­-· ....... "polita complex" CH6.13 Pupukea, Oahu 5R, 1D ~ basimacula? C72.3 Alakai Swamp, Kauai 5R,1D 5A, 1D ~ melanosoma? WH31B.15 Kumuwela Rd., Kauai 5R,1D .(heterozygous inversion) "large tan" C72.12 · Alakai Swamp, Kauai 5R, 1D Antopocerus sp. C107.10 Puu Kolekole, Molokai 6R +. ....... The University of Texas Publication PLATE I --//­ ............ . . ­ ~,, I. D. crucigera /' ~ I \\ ''· ~ 4 . 0. eurypeza ~\\?­~j·,, 7. D. k a ul u a i ,,/ ~, #II~ IO. D. ask et ostoma -.... II'/ ~ ·. 'l\\ 2 . D. grimshawi ,,(1 --;·,,~ 5. D. freycinetiae \\ /// -··­ /// 8. D. pectinitarsus --....\I I__ -)\ ~ // II. D. mi mica ~ ~­ ."It . ,, 3. Hybti d F1 \ II .. ~ ")) 6. 0. basimacula '? ~.. ~ \ .-;/(/~ 9. Antopocerus sp. ~ ~ ::::: ~­ II ~ 12. D. conjectura (Extra _rod) PLATE II Jl­ 13. D. ischnotrix 14.D. parva 15. D. crassitemur -··\\I« ~ 16. D. nasalis 17. S. (T.) albovittata 18. S.(T.) argentifrons 19. S. (TJ walaleale 20. S. (Exallo.) sp . 21. S. (Exallo.l sp. vL // _j\ ~ - 22.S. (B.) palmae 23. S. (B.) n. sp. 24. S. (Sunosfoma) sp. Hybrid ? The University of Texas Publication had two pairs of V's, two pairs of rods, and one pair of dots. The configuration of nasalis is of interest as it is the only species examined among either the Dro­sophila or Scaptomyza in which the modification from the primitive metaphase appears to involve more than the fusion of rods to form V-shaped chromosomes. The presence of five arms and a dot in the salivary preparation of ischnotrix would indicate that, in this species, the formation of V-shaped chromosomes probably resulted from the fusion of rods with the loss of a centromere. During the period of this investigation a cross was made between crucigera and grimshawi; the metaphase configuration of the F1 hybrid from a crucigera male and a grimshawi female is shown in Figure 3. Several larvae were examined and the arrangement of the rods, as shown in the illustration, was consistent. Normal pairing occurred with three of the rods, one set failed to pair, and the other consisted of rods of unequal length. Eleven species of the Scaptomyza were examined and two types of metaphase configurations were observed. The species had either one V, three rods, and one dot or two V's, one rod and one dot. Both of these configurations were found among species of the Trogloscaptomyza (Figs. 18, 19) and the Bunostoma (Figs. 22, 23); Tantalia albovittata (Fig. 17) and the strains of the Exalloscaptomyza (Figs. 20, 21) had a single pair of V-shaped chromosomes. Two different strains of the Exalloscaptomyza are shown since the specimens from Molokai (Fig. 21) consistently showed a pair of V's noticeably smaller than those from the other islands; strains from Oahu, Kauai, and Hawaii (Fig. 20) had the larger V's. During the investigation of the karyotypes, several unusual configurations were observed. Among species of the genus Drosophila, one crucigera was found in which there were only eleven chromosomes, with one rod missing; in three species, D. ischnotrix, D. basimacula?, and D. con;ectura (Fig. 12), one extra rod was found in single larvae. The same types of configurations were found among species of Bunostoma. In Bunostoma palmae an extra rod was observed and in another Bunostoma species one rod was missing. In Figure 24, a metaphase is shown from one Bunostoma larva; this appears to be a hybrid between a species with one pair of V's and the species with two pairs of V's. However, there is an extra rod present if this configuration is the result of hybridization. This investigation of the karyotypes of some species of the Hawaiian Dro­sophilidae has indicated the need for detailed analyses to determine the X-Y chromosome relationships and distribution of heterochromatin, as well as the karyotypes for the many species not yet analyzed. Another aspect which needs investigation is the analysis of rearrangements; evidence for the presence of inversions and other abnormalities among the populations in Hawaii has been noted in this preliminary report. XIII. Chromosomal Races of Drosophila crucigera from the Islands of Oahu and Kauai, State of Hawaii1 2 HAMPTON L. CARSON The daring prediction of Zimmerman ( 1958) that the Hawaiian Islands might harbor as many as 300 species of Drosophila appears not to have been an over­estimation. Thus, the recent monograph of Hardy ( 1965) lists about Z50 species of this genus of which only seven represent recent introductions under the in­fluence of man. When members of other closely allied genera are added, the number of endemic species in the Family Drosophilidae rises above 400. Extra­ordinary morphological diversity has accompanied this speciation. Recent in­tensification of collecting efforts by a joint research team of the University of Hawaii and Texas in 1963 and 1964 have resulted in the obtaining of spec­imens of many more new species. Compilation of these data continues but there is little question that the drosophilid fauna of Hawaii is one of the most remark­able in the world. Zimmerman may indeed have underestimated these remarkable evolutionary events. The question arises as to what extent the population genetics of the Hawaiian species may provide a clue to this "explosive" evolution. Accordingly, the writer, during the summers of 1963 and 1964, set himself the task of selecting one species of the endemic fauna for intensive study of quantitative samples of natural popu­lations. It was hoped that an intraspecific study of chromosomal morphism might provide data on which interpretations of interspecific evolutionary events could be made. Among the most striking and unique groups of species to be found on the Hawaiian Islands are those related to Drosophila crucigera Grimshaw, the subject of the present paper. Briefly, these are large, long-winged, long-legged and slow­walking flies which generally show extensive dark maculations on wings and body. They have been ascribed by Hardy to the subgenus Drosophila. Flies of this description, numbering at least £0 species, are found on all of the major islands of the Hawaiian chain. Drosophila crucigera occurs in accessible locations on two islands, Oahu and Kauai; it can be obtained in reasonable numbers by both general sweeping and the use of baits. Wild-caught specimens survive well in the laboratory and stocks can be established relatively easily. Wild males can be crossed to laboratory virgins; individual wild females may be isolated and will produce progeny singly. All of these features, as well as the great cytological favorability of the salivary gland chromosomes, have led to the selection of this species for this study. The preliminary results seem to support the idea that isolation, inbreeding and small-population effects may have played a key role in the formation of Hawaiian species of Drosophila. 1 This work was supported by grant G-20107 of the National Science Foundation to the au­thor, by grant (NSF) GB-711, and by Public Health Service Research Grant No. 10640 from the National Institutes of Health. 2 Department of Zoology, Washington University, St. Louis, Missouri. The University of Texas Publication MATERIALS AND METHODS Through his own efforts and those of colleagues and assistants (see Acknowl­edgments), the writer obtained 267 specimens of D. crucigera in 1963 and 1964, 140 on Oahu and 166 on Kauai (Table 1). The species apparently occurs only on these two islands (Hardy 1965) . The collecting areas on Oahu are at moderate elevation, between 1200 and 2000 feet above sea level. Like other endemic species, D. crucigera tends to be only locally common, usually being found in deep, wet rainforests. The species can be net-collected by sweeping over low vegetation in stream valleys; frequently specimens appear to be resting near the ground on the under side of leaves. The species is not, like some endemics, restricted to areas of predominantly endemic vegetation. At the area along Drum Drive, in northern Oahu, for example, this species was swept from a dense stand of introduced lehua, Calliandra inaequilatera Rushy. At Halemanu Valley, Kokee State Park, Kauai, at about 3400 feet above sea level, the flies were quite numerous in a wet stream­side forest which supported an extensive undergrowth of the introduced New Zealand Laurel, Corynocarpus laevigatus Forst. It should be stressed, however, that the distribution of the species appears to be spotty, for reasons which are not clear. Even in favorable areas, the species is never found in numbers even remote­ly like those of comparable endemic species on continental areas. After two years of study, the capture of each specimen continues to be an event of note. On June 16, 1963 the writer collected a mass of slime flux from the trunk of a specimen of the endemic tree Acacia koa, along Forest Trail No. 2 on Mt. Tantalus, Oahu. This slime flux contained the characteristic four-filamented eggs (two very long and two moderately long filaments) of an endemic Drosophila species. Although larvae were obtained, no adults were reared from this collec­tion. A year later (June 18), a second egg-filled sample was collected from the same flux on the same tree and from this a specimen of Drosophila crucigera was reared in the laboratory. Wheeler (unpublished) reared this species from frass TABLE 1 Collections of Drosophila crucigera on Oahu and Kauai, 1963 and 1964 Collected by sweeping Collected by baitingLocality Date n 66 l?I? 66 . Mt. Tantalus, Oahu June-July, 1963 0 5 3Z 7 June-July, 1964 1 15 0 Pupukea, Oahu June-July, 1963 10 8 June-July, 1964 5 14 0 Sacred Falls, Oahu July, 1963 4 4 Opaeula, Oahu June, 1964 0 0 0 Drum Drive, Oahu June-July, 1964 Z1 11 TOTAL, OAHU 41 43 49 7 Kokee, Kauai July, 1963 0 1 0 0 June, 1964 0 0 7 0 Halemanu, Kauai June, 1964 105 7 August, 1964 0 0 3 3 TOTAL, KAUAI 0 1 115 10 Grand Total 41 45 164 17 Z67 obtained from rotting tips of the endemic plant Freycinetia, at Pupukea, Oahu as well as from mushrooms from Mt. Tantalus. Accordingly, the species appears to have some latitude of breeding site. In 1963, baits of fermented banana, papaya, or cabbage were found to attract D. crucigera in moderate numbers (see Table 1, Mt. Tantalus 1963). Baiting in such a manner appeared to attract more females than males. In 1964, a new bait mixture, consisting of extremely decayed tomatoes mixed with leftover high­protein fly food (Wheeler and Clayton, 1965) produced excellent results, attract­ing not only D. crucigera but other related "picture-wing" flies. This mixture was used with extraordinary success at Halemanu, Kauai in June 1964, resulting in the capture of 105 females and 7 males. Daily visits were made to the baits for a period of about one week; collecting was continued for about three hours each afternoon. The leftover fly food from the laboratory cultures was autoclaved and then re-contaminated with molds and yeasts associated with D. crucigera. Follow­ing this, it was allowed to stand for several weeks until it had acquired a nauseous and revolting odor. Access by flies in the laboratory was prevented. Just before use, about ten parts of this material was mixed with one part of decayed tomatoes. When treated in this manner, the bait retains its attractive qualities for crucigera­group flies for several weeks. The bait was carried to the collecting area in plastic buckets and distributed in small containers to various parts of the prospective collecting area. It was found that, if a trap was placed in a favorable spot, flies would visit it immediately. Accordingly, the standard practice was to carry along a bucket of bait, expose it and attempt collections immediately. Traps which failed to attract flies in a half­hour or so were moved until they appeared to be in a more favorable spot. In this manner, the suitability of a woods for selective trapping could be fairly quickly surveyed. Almost all of the 1964 flies were collected in the above manner. The prepon­derance of females is strikingly evident (Table 1), a circumstance which is highly favorable to the analysis of natural populations by the isofemale method. One male and one female fly collected at Mt. Tantalus, Oahu, on June 10, 1963 were the progenitors of a laboratory stock, designated C53.11, which has been chosen as a chromosomally Standard stock for the species. In June 1964 a series of twelve salivary gland chromosome smears from this stock were examined. Each was found to be homozygous for gene arrangement and to display five long chromosome arms and a dot. Metaphases from dividing larval ganglion cells showed a basic chromosomal complement of five rods and a dot. These findings confirm those of Clayton (1966). These chromosomes have been designated X, 2, 3, 4, 5, and 6 and preliminary photographic salivary gland chromosome maps have been prepared. Since its original examination, the Standard stock has been used for crosses to other flies and a total of 52 female and 39 male F1 larvae have been examined in the salivary glands. The conditions in all of these larvae are fully consistent with the structural homozygosity of the Standard stock. The aceto-lactic orcein technique has been used throughout. Of the 267 flies collected, 160 wild females were placed individually on culture medium and salivary gland smears were obtained from 114, or 71 %. Forty wild males were crossed to 3-4 virgin Standard females each and salivary gland chro­ The University of Texas Publication mosome smears of F1 larvae obtained from 18, or45% of them. The development of a high-protein culture medium and the employment of a sand-culture method for inducing pupation (Wheeler and Clayton, 1965) has been of great importance for the culture of D. crucigera, and these methods have been employed routinely. Certain female individuals of this species, however, do not readily oviposit, and this group is responsible for most of the above­mentioned failures to obtain Fi smears. Dissection of approximately twenty such females revealed that all contained mature filamented eggs in the ovary and were heavily inseminated. In several instances, females lived for six months in the laboratory without reproducing. An occasional female will lay eggs which do not hatch. Dissection of such females revealed that several, at least, were not insemi­nated, although it should be stressed that such cases are rare and the insemina­tion rate among wild females appears to be very high. Males of D. crucigera are considerably more ephemeral than females and rarely live for more than a month in the laboratory. OBSERVATIONS Examination of Fi larvae from wild flies from both Oahu and Kauai revealed the presence of a small inversion at the base of chromosome 3 (Fig. 1c; Table 2). Flies heterozygous for this inversion carry one gene arrangement which has the Standard gene pattern in this region (gene arrangement 3) and another arrangement which is relatively inverted and which is designated as gene ar­rangement 3-1. The homokaryotypes Standard 3/ 3 (Fig. 1a) and 3-1/ 3-1 (Fig. 1 b) are easily distinguishable cytologically and all smears have been scored as to which of the three karyotypes it shows. In every case possible, seven smears of Fi larvae from each wild female have been scored in this manner, permitting inference, with a 63/ 64 probability of correctness, of the karyotypes of the wild flies in nature (for details of this method, see Carson, 1958). In this manner, 82 TABLE 2 Frequency (in per cent) of X and third chromosome gene arrangements in natural populations in Drosophila crucigera from Oahu and :K!auai X chromosome 3rd chromosome No. of Gene arrangement : No. of Gene arrangement: chromosomes Standard third chromosomes Standard Locality examined x X-1 examined 3 3-1 OAHU* Mt. Tantalus 20 100.0 0.0 26 88.5 11.5 Pupukea 12 100.0 0.0 19 100.0 0.0 Drum Drive 29 100.0 0.0 37 97.3 2.7 Total, Oahu 61 100.0 0.0 82 95.1 4.9 KAUAI Kokee 14 0.0 100.0 18 66.7 33.3 Halemanu 250 0.0 100.0 332 71.1 28.9 Total, Kauai 264 0.0 100.0 350 70.9 29.1 ' Note added in proof October, 1966: In 1966, a sample of D. crucigera was obtained from Palikea in the Waaianae l\fountam range, Oahu. The localities reported in the Table above were all from the Koolau Range. Of nine X chromo­som.e~ observed from Palikea , all have the gene arrangement X-1, that is, resembling populations from Kauai. Of thirty-six add1t10na_l X chromosomes obtained from the Koolau's in 1965, all have the standard X gene arrangement. The specimens from Pahkea were collected by K. Kaneshiro, J. Murphy and K. Resch. la 1b FIG. 1. Proximal end of chromosome 3 of D. crucigera. 1a shows homokaryotype Standard 3/Standard 3; 1b shows homokaryotype 3-1/3-1. The arrows in these two figures show the extent of the 3-1 inversion. Fig. 1c shows the heterokaryotype, 3 /3-1. The University of Texas Publication FIG. 2. An X chromosome of a female D. crucigera larva which is an F1 from a cross between a Standard Oahu female and an X-1 Kauai male. The distal end of the chromosome is to the left on the photograph. Starting at the distal end, the direction of the Standard banding order in the resulting heterokaryotype X/X-1 is indicated by the arrow. wild 3rd chromosomes from Oahu and 350 from Kauai were examined (right­hand side of Table 2). The 3-1 gene arrangement is more frequent on Kauai than on Oahu (x2 = 4.115; p < .05). No other naturally occurring polymorphism has been found. Nevertheless, crosses between flies from Oahu and Kauai universally show, in females, a single large inversion in the central region of chromosome X (Fig. 2) . Inspection of the regions concerned reveals that the two gene arrangements (Standard X, Oahu and X-1, Kauai) can be easily distinguished in the homozygous or hemizygous condition. Accordingly, the banding order of the X chromosomes was read in each test larva and the type of homokaryotype recorded. The results of this analysis are given in Table 2, left-hand side. It will be seen that all of the X homokaryotypes in larvae produced by Oahu flies are Standard X/ X whereas all of the Kauai ones are X-1/ X-1 homokaryotypes. Without exception, all eight crosses made between flies originating from the different islands show the ex­pected gene arrangement in male larvae and show the heterokaryotype in female larvae. The reciprocal crosses appear to succeed equally well. Inthe examination of the slides in the above manner, each of the other chromo­somes, autosomes 2, 4 and 5 were traced throughout their lengths in every smear to confirm that they were homokaryotypes. No inversions were found within any of the populations or in crosses between them. There are thus no differences in gene order between Kauai and Oahu in these autosomes. A total of 1286 of these three autosomes were examined from the two islands without detecting any polymorphism. DISCUSSION Hardy (1965 and personal communication) has noted no morphological differ­ences between the D. crucigera populations of Kauai and Oahu. The writer has also not noted any morphological differences, although no systematic metrical studies have been made. The data given above, however, show that although crosses between flies from the two islands are easy to obtain in the laboratory, there is a permanent and fixed difference in the gene arrangement of the X chromosome. The gene frequencies of the two arrangements in the third chromo­some are also different. The above facts suggest that the populations on the two islands are best referred to as chromosomal races or subspecies. This conclusion must be regarded as tentative, however, until tests of the fertility of F 1 male and female inter-island hybrids have been carried out. Should such tests uncover sterility, this would constitute evidence that D. crucigera populations of Kauai and Oahu are allo­patric sibling species. The fact that D. crucigera from Kauai and Oahu cross easily in the laboratory and give a vigorous F1 generation does not prevent them from possibly being cryptic species. Clayton (unpublished) obtained a large and vigorous F1 between D. crucigera from Oahu and D. grimshawi Oldenberg from Molokai. F1 females were fertile in backcrosses but F1 males were sterile. D. grimshawi has clear morphological differences from D. crucigera and has long been considered a different species on these grounds. Further studies of these relationships are in progress. Kauai and Oahu are separated by eighty miles of open ocean. Accordingly, it seems unlikely that gene flow between the islands, especially for a mountain­dwelling endemic, could be great. On the contrary, the fact that there appears to be a fixed chromosomal difference between the populations of the two islands, indicates that differentiation between small isolated populations may have played an important role in this case. The facts are in accord with the idea that isolation and small-population effects have played a key role in the formation of Hawaiian species of Drosophila. SUMMARY 1. Drosophila crucigera is a large endemic drosophilid found at moderate eleva­ The University of Texas Publication tions on the islands of Oahu and Kauai, State of Hawaii. Several natural breeding sites are described. 2. Females of this species are selectively attracted to a bait consisting of rotting tomatoes and high-protein fly food. The sex ratio of flies collected by sweeping does not differ from 1 : 1. 3. Chromosomal analyses of populations from the two islands show that both are polymorphic for a very small inversion and that the frequencies of the two arrangements are different on the two islands. A sectionally homozygous labora­tory strain derived from a single wild female from Mt. Tantalus, Oahu has been designated as Standard. 4. All of the X chromosomes studied from Oahu are of the Standard gene arrangement. All of the X chromosomes from Kauai differ from the Standard by a single large inversion. 5. The facts suggest the tentative conclusion that the populations of Kauai and Oahu represent disjunct allopatric chromosomal races of a single species. ACKNOWLEDGMENTS I am especially grateful to Drs. M. R. Wheeler, F. T. Clayton and D. E. Hardy for providing specimens, making collections and permitting use of unpublished data. Dr. H. D. Stalker not only helped in collecting, but provided valuable assist­ance with photography. I wish also to thank the following, all of whom aided in the collecting: W. B. Heed, H. T. Spieth, L. H. Throckmorton, D. Gubler, M. S. Carson, W. S. Stone, T. A. Ohta, and V. T. Gammell. A. S. Farm, Jr. rendered valuable assistance by rearing flies and making cytological preparations. LITERATURE CITED Carson, H. L. 1958. The population genetics of Drosophila robusta. Adv. in Genet. 9: 1-40. Clayton, F. T. 1966. Preliminary report on the karyotypes of Hawaiian Drosophilidae. (This publication). Hardy, D. E. 1965. Insects of Hawaii. Vol. 12. Diptera: Cyclorrapha II. Honolulu: Univ. of Hawaii Press, pp. 814. Wheeler, M. R. and F. T. Clayton. 1965. A new Drosophila culture technique. D.l.S. 40: 98. Zimmerman, E. C. 1958. 300 species of Drosophila in Hawaii? A challenge to geneticists and evolutionists. Evol. 12: 557-558. XIV. Cytological Studies on Some Species of the Tripunctata Group of Drosophi"la1 COSTAS D. KASTRITSIS2 INTRODUCTION The increasing number of described species of the Drosophila tripunctata ~oup places this group among the largest in the genus. It was established by Sturtevant (1942), who included only D. tripunctata in it and doubtfully D. histrio. Hsu ( 1949) and Wheeler ( 1949) removed histrio from the group, while Patterson and Mainland (1944) placed D. unipunctata and D. crocina in it. Hsu (1949) suggested the division of the group into two subgroups. Freire-Maia and Pavan (1950) added three more species (D. mediopunctata, D. mediosignata, and D. mediostriata). Frota-Pessoa ( 1954) described 15 more species and incorpo­rated 10 others, bringing the number to 31 species distributed in four subgroups. Heed and Wheeler (1957) described 10 species, and added four more to bring the number of species belonging to the tripunctata species group to 45. Finally, Pipkin and Heed ( 1964) have recently added 9 more species from Panama, which raises the total to 54. It can be predicted that more species will be described in the future. The tripunctata group belongs to the subgenus Drosophila and its species are distributed in the New World with the greater part occurring in the neotropical and a few in the nearctic region. Only one representative (D. tripunctata) is found in the United States, and it has not been collected elsewhere. Patterson (1957) carried out a hybridization study among 11 species of the tripunctata group and made the conclusion that the species are highly isolated from each other with the exceptions of D. mediopunctata and D. unipunctata as well as D. mediostriata and D. paramediostriata. Heed (1957) found some evi­dence of differentiation among strains of D. crocina of different geographic origin and suggested further tests to determine the mechanisms behind this isolation. The results presented by Patterson in his 1957 paper are rather discouraging to an evolutionist who believes that hybridization tests are one of the major methods of determining relationships among different species. It was decided that a cytological examination of the salivary gland chromo­somes of the available species should be carried out with the hope of discovering similarities among the chromosomes of different species as well as the existence of possible polymorphism of gene arrangements within each one of them. With this background future investigations can be made to determine the series of events involved in the evolution of this species group. 1 This investigation was supported in part by Public Health Service Research Grant No. 11609 from the National Institutes of Health. 2 Present address: Department of Biology, Texas Technological University, Lubbock The University of Texas Publication ANIMALS AND METHODS Most of the stocks used in this study were selected from those maintained in the laboratories of The University of Texas. Some of the stocks from Panama were kindly provided by Dr. Sarah Pipkin. Table 1 indicates the stock numbers of the University of Texas, the location in which the flies were collected, the month and the year of the collection, and the name of the collector(s) . The stocks received from Panama are designated with the symbols used by Dr. Pipkin. A total of eleven species were used in this study: one from subgroup I, three from subgroup II, four from subgroup III, and three from subgroup IV. The flies were raised in half-pint milk bottles on the culture medium of banana, agar, malt, yeast, and karo syrup which is currently in use at the University of Texas laboratory. The temperature of the room where the flies were kept was 21° ± 2°C. Inter-and intraspecific crosses were conducted using small mass matings. The interspecific crosses were conducted among the stocks chosen as standards of the different species, while the intraspecific crosses were made between the different geographic stocks with the standards in each case. The crosses found to be in­compatible by Patterson (1957) were not repeated. Some species, not available to Patterson, were available for this investigation. These were crossed among themselves as well as with the species used in the 1957 study. Two species, mediodelta and angustibucca, were not used for crosses because raising the flies in the laboratory for more than one generation was not successful. Slides were prepared for the purpose of studying both larval neuroblast and larval salivary chromosomes. For neuroblast chromosomes, two stains were used. The neuroblast cells were first treated in a stain consisting of 3 % orcein in 70% glacial acetic acid and 30% ethyl alcohol. After five minutes another stain was added. The second stain, which was the same as that used for the salivary gland preparations, consisted of 2% orcein in equal parts of glacial acetic and lactic acids. From 20 to 150 salivary gland smears per geographic stock were prepared. The standard stock for each species was chosen arbitrarily and usually represented the stock with the least number of heterozygous rearrangements. Both photographic and drawn chromosome maps of all species (except D. angustibucca) were constructed. The maps were labeled by letters of the alphabet using distinct bands or other characteristics as points of recognition. It should be mentioned that the letter which indicates a certain region of a chromosome in one species does not necessarily indicate the same region in the same chromosome of another species. The inversions were placed on the map and named by using the abbreviation In and parentheses. Within the parentheses the number designating the chromo­some appears first and then the regions where the break points of the inversion are located. The symbolism "complex" outside the parentheses indicates a com­plex inversion, while the abbreviation "per" is used to indicate possibly peri­centric inversions. Collection date Stock Origin Month Year Collectors SUBGROUP I D. angustibucca 63C13 Panama 1963 Pipkin SUBGROUP II D. mediodelta 1Lewis1 El Volcan, Panama 1963 Pipkin D. mediopunctata 2211.4 Ponta Grossa, Brazil Cordeiro D. unipunctata H4-00.19 Palmar, Costa Rica June 1959 Carson 15B.6 Barro Colorado, Panama Pipkin Heed, Carson and H181.14 2538.2 H4-04.36 Barro Colorado, Panama Cerro Campana, Panama Changuinola, Panama August December July 1956 1959 1959 Wasserman Pipkin Heed, Carson H91.14 Medellin, Colombia November 1955 Heed H192.29 x Heed, Carson and H192.17 Rio Negro, Colombia September 1956 Wasserman H203.12 Caripe, Venezuela November 1956 Wasserman SUBGROUP III D. crocina H131.2 Rio Piedras, Puerto Rico February 1956 Heed H86.12 Balboa, Panama November 1955 Heed H234.12 H29.1B Trinidad El Salvador July November 1957 1953 Heed and Boyes Heed Heed, Carson and H166.9 H136.25 Costa Rica Jamaica August February 1956 1956 Wasserman Heed D. mediostriata H202.23 Cumanacoa, Venezuela October 1956 Wasserman H203.2 Caripe, Venezuela November 1956 Wasserman H4-03.3 H91.17 Chanfuinolai Panama Mede lin, Co ombia July November 1959 1955 Heed Heed H343.8 H234.5 H341.13 Montero, Bolivia Trinidad Vila Atlantica, Brazil April July April 1958 1957 1958 Wasserman Heed and Boyes Wasserman D. paramediostriata 2327.2 Rio Piedras, Puerto Rico June 1953 Townsend H254.12 2378.5 Mayaguez, Puerto Rico St. Vicente, Cuba August January 1957 1956 Heed and Boyes Breuer D. mediopictoides H4-07.32 45C4 2GL-1 SUBGROUP IV Boquete, Panama Cerro Campana, Panama Panama July 1959 1963 1963 Heed and Carson Pipkin Pipkin D. tripunctata 1910.4 2003.3 1878.6 2300.6 2539.2 2512.1 2010.7 2069.16 D. mediodifjusa Dexter, Missouri New Orleans, Louisiana Albemarle, N. Carolina Biloxi, Mississippi Austin, Texas Nacogdoches, Texas Everglades, Florida Crystal Lake, Nebraska August June July June April April June August 1948 1950 1948 1953 1960 1958 1950 1950 Wheeler and Hsu Hsu and Stephens Wheeler and Hsu Wheeler and Heed Throckmorton Wheeler Hsu and Stephens Wheeler and Stephens H130.6 H260.1 Rio Piedras, Puerto Rico Puerto Rico February October 1956 1957 Heed Wasserman H138.16 H136.5 H355.9 H352.4 H135.2 H411.2 H267.4 D. metzii Jamaica Jamaica Jamaica Jamaica Petionville, Haiti Petionville, Haiti Maricao, Puerto Rico February February July July February July October 1956 1956 1958 1958 1956 1959 1957 Heed Heed Heed and Wasserman Heed and Wasserman Heed Heed Wasserman 58B8 Barro Colorado, Panama 1963 Pipkin The University of Texas Publication TABLE z Results of interspecific crosses among species of the tripunctata species group of Drosophila 'i1'i' medio-uni-medio-paramedio-medio-tri-medio­00 punctata punctata crocina striata striata pictoides punctata diffusa metzii mediopunctata 5<;>+5~· • • -• unipunctata -• crocina -• -• -• mediostriata -• +· -• + . paramediostriata -• +· -• + mediopictoides tripunctata -• • larvae, pupae• mediodiffusa . metzii 1 . --• Note: Minus (-) signs indicate incompatible crosses: Plus ( + ) signs indicate compatible crosses; Asterisks (•) indicate results taken from Patterson (1957). In certain cases the kind and number of hybrids are indicated. When a double entry is indicated, then the lower one shows results obtained in this study. RESULTS A. Hybridization tests Table 2 indicates the results of all the interspecific crosses which were con­ducted in this study as well as some of the results of Patterson ( 195 7). In only one case was the cross found to be compatible; crosses between mediostriata and paramediostriata yielded progeny which, when crossed among themselves, gave viable and fertile offspring. The hybrid larvae were used for cytological analysis. These results are presented later in this paper. Other crosses found as partially compatible by Patterson (1957) failed to produce any hybrids when they were repeated in this study. All the intraspecific crosses which were carried out were found to be compat­ible and fully fertile. The F1 larvae of the intraspecific crosses were also used for cytological analysis. B. Cytological observations It was found that the salivary gland chromosomes of the tripunctata group species which were used in this study present a few characteristics that can be identified as similar in each species; these characteristics will be pointed out in the last part of the Discussion. In the following paragraphs the individual features of the chromosomes of each species are described. Subgroup I Drosophila angustibucca Duda, 1925 (1 ) Neuroblast chromosomes were not studied. (2) Salivary gland chromosomes are very large. There are five arms and one dot. (3) Construction of maps was unsuccessful. ( 4) Rearrangements were not found. (5) All of the general features of tripunctata species group chromosomes were very obvious. Subgroup II Drosophila mediodelta Heed and Wheeler, 1957 ( 1) Neuroblast chromosomes were not studied. (2) Salivary gland chromosomes are very large; there are five arms and one dot. The dot chromosome is the largest seen in any of the species examined and is unique in carrying a heterochromatic mass in its proximal end. ( 3) Chromosome maps, photographic and drawn, are shown in Figures 1 and 2. ( 4) The rearrangements seen were one simple and two included inversions in the 2nd chromosome. The inversions always appeared together in a 30.7% frequency. Both standard and inversion homozygote arrangements were observed. Drosophila mediopunctata Dobzhansky and Pavan, 1943 (1) Neuroblast chromosomes are five pairs of rods and one pair of dots. (2) Salivary gland chromosomes are small. There are five arms. The dot chromosome was not found in the preparations. (3) Chromosome maps, photographic and drawn, are shown in Figures 3 and4. ( 4) Rearrangements: there were two included inversions near the distal end of the 2nd chromosome (Figure 5,I); the inversions always appeared together in a 65% frequency. Both standard and inversion homozygote arrangements were observed. Drosophila unipunctata Patterson and Mainland, 1943 (1) Neuroblast chromosomes are two pairs of rods, one pair of V's, one pair of J's and one pair of dots. (2) Salivary gland chromosomes are large. There are five arms and one dot. ( 3) Chromosome maps, photographic and drawn, are sl;10wn in Figure 6 and 7. For purposes of convenience, the ln(J, Hl) per homozygote was used as standard for the X chromosome. (4) Rearrangements in some of the chromosomes were extremely poly­morphic. Table 3 indicates the inversions and their frequencies in the various populations. Photographs and drawings of the inversions are shown in Figures 8, 9, and 10. Tables 4 and 5 indicate the various gene arrangements in the different geographic strains for the 2L and 3R chro­mosomes respectively. Two fusions have been observed between the sec­ond and fifth, and between the third and fourth chromosomes. Subgroup Ill Drosophila mediopictoides Heed and Wheeler, 195 7 (1) Neuroblast chromosomes are five pairs of rods and one pair of V's with unequal arms. (2) Salivary gland chromosomes are of medium size. There are five arms and a dot. The University of Texas Publication x c. E ;g {: i <> 0 ., ~ ~ " E <> u '::: 0 "' ~ ~ "' "' <> (\J Ci i FIG. 1. Photographic salivary gland chromosome map showing the standard gene arrangement of Drosophila mediodelta. Stock 1 Lewis 1 from El Volcan, Panama, was used for the construc­tion of this map. x ln(2,LR);;J _ r--ln(2,MO) ;jEJl 1n1:tnw.; lo~l< ·J~!f'pIE<'Ulf!.\t:· t •;1417L'·1n.T:-:lnl::llll::l~-:-i.rr,l~lrT~l\ ·rJJ). ·"" .t·rr1J J111~r•r1~~1;MJJ((!!n1hiTnJ'[i~. :1111~Jll1t:l}}r.!1Jfltl1 . JM IN lo IP o IR Is 11 Ir ""'"*~n\':<. i91"'1· .:~' l' 3 ~t;:•11i~;.,;~1q;1!r,~~P& r~n~J) 1~ Ir lo ~ ..... -; - 1::'· ~­ ~ ..... 0 ~ ~ "".:i -; -t;· i::: ~ ~ ..... >:) 11:r•-~v1 ..... 4 Is >:) A M ~ ~ i::: Fw. 2. Drawn salivary gland chromosome map indicating the standard gene arrangement of ~ Drosophila mediodelta, and the inversions which were detected in the population examined. ~ ...... ..... ., I:> E ..... ~ 0.. 0 § E ·~ § ~ .t:; () ., ~ ·.c:() ~ Q. ~ ~ II) 0 "' ~ 0 _J .t:; C\I Cs 3­ FIG. 6. Photographic salivary gland chromosome map showing the standard gene arrangement of Drosophila unipuctata. Stock H4-00.19 from Palmar, Costa Rica, was used for the construction of this map. ~ ~-~ ~ -:> - i~ ~ ,;:~ ;~ ­ --w ·~-..... 0 i~ (.) -~.. <( a:: <( ii__fQ a:: C\J ~ l"0 <( _J l"0 Frn. 7. Drawn salivary gland chromosome map showing the standard gene arrangement of Drosophila unipunctata, the inversions which were detected in the populations examined, and the metaphase configuration chromosomes of male individuals. The University of Texas Publication In (2L,BF), In (2L,EK) i<'IG. 8. Photographs of inversions in the X and the left arm of the second chromosome of Drosophila unipunctata. A schematic diagram of ln(2L,BK)complex is shown. Kastritsis: Cytology of Tripunctata Species Group In (2L,AH) complex, ln(2L,HJ) ln(2L,DJ). ln(2L,HJ), ln(2L,IM) In (2L,H J) ln(2L,IM) ln(2L,HJ), In (2L,IM) Fm. 9. Photographs of the inversions in the left arm of the second chromosome of Droso phila unipunctata., and schematic diagrams of the complex rearrangements. The University of Texas Publication In (~R,AB) In (3R,FK), In (3R,H K) In (3R,DG) In (3R, CK) complex Fw. 10. Photographs of the inversions in the right arm of the third chromosome of Drosophila unipunctata and schematic diagram of In (3R, CK) complex. TABLE+ 'Distribution of arrangements found in the left arm of the second chromosome of Drosophila unipunctata Arrangements Standard Standard Palmar, Costa Rica 100% Barro Colorado Panama + Barro Colorado Panama (1956) - Cerro Campana,Panama - Stocks Changuinola, Panama + Medellin, Colombia + Rio Negro,Colombia Caripe, Venezuela ln(2L,HJ) Standard - - - - 67.8% ~ ..... ""!-· In(2L,HJ) In(2L,HJ) ln(2L,EK) ln(2L,EK) ln(2L,BF) ln(2L,BF) In(2L,EK) ln(2L,BF) ln(2L,HJ)ln(2L,IM) Standard In (2L,HJ)In(2L,IM) ln(2L,HJ)ln(2L,IM) In(2L,HJ)In(2L,IM) ln(2L,AB) ln(2L,DJ) In( 2L,AB) In(2L,DJ) ln(2L,AB)ln(2L,DJ) ln(2L,HJ) ln(2L,IM) ln(2L,HJ) In(2L,AH)com.In ( 2L,HJ) Standard ---------- ----42.0% + ---- ----+ 63.8% + -- + ---+ 45% - + --- ---4-0% --- + + 50% ~f?. ~ 0 C" ~ .Q.. '.:;3 ~· !:::: ~ ~ s f ~· ~ d -§ In(2L,AH)com.ln(2L,HJ) ln(2L,AH)com.ln(2L,HJ) ln(2L,AB)ln(2L,BK)com.except In(2L,EK) In(2L,AB) In(2L,EK) In( 2L,AB) In(2L,EK) In(2L,AB)In(2L,EK) In(2L,AB)ln(2L,BK)com.except ln(2L,EK) ln(2L,AB) ln(2L,BK)com.except ln(2L,EK) ---- ---- ---- ---- --- + --- 4-0,7% + + to '° The University of Texa.s Publication C\J Fm. 11. Photographic salivary gland chromosome map showing the standard gene arrange­ment of Drosophila mediopictoides. Stock H407.32 from Boquete, Panama, was used for the con­struction of this map. ination of the hybrids revealed perfect pairing of the chromosomes. Many rearrangements were observed (see Discussion). Photographs of the 2nd, 3rd, 4th, and 5th chromosomes of the species hybrids are shown in Fig­ure 5, IV. ::IE p:·­ u-7...­ ~ ~~ a:,'!;i:-~ «=-­ = r~-­ I Cl ~ ~ -o r: ­ ~­ !m £~ . ~ -~<{ r<> v f: E ~~ II) ~ -~ -~ ~ .~ ~ Cb 0. ~ 0 E ~~ :(: g ~:g II) E ~ 0 C\J cS~ Fm. 12. Drawn salivary gland chromosome map showing the standard gene arrangement of Drosophila mediopictoides, the inversion which was detected in the populations examined, and the metaphase configuration chromosomes of male individuals. ..j>.. w to TABLE 5 Distribution of arrangements found in the right arm of the third chromosome of D. unipunctata Arrangements Standard In(3R,AB) In(3R,DG) In (3R,HK) In( 3R,CD) In( 3R,DG) In (3R,AB)In(3R,CD)In(3R,HK) In(3R,AB)ln(3R,FK)In(3R,HK) ln(3R,AB) In(3R,CD) In (3R,FK) In(3R,HK) In(3R,CD) ln(3R,DG) In (3R,FK) In(3R,HK) ln(3R,AB) In(3R,CD) In(3R,DG) In(3R,FK) In(3R,HK) Pal mar, Costa Rica + - - - 60% - - - - - Barro Colorado Panama + - - - 40.6% - - - - - Barro Colorado Panama ( 195fi) + -26.6% - - - - - Ce1To Campana, Panan1a + -- - 50% ---- - Stocks Clrnugui110IH, Pauama + - - -14% -- - -- Me:) "' "i:j ~ ,_ """' -. ~ >:) ...... 0 · - ;:i FIG. 13. Photographic salivary gland chromosome map showing the standard gene arrange­ment of Drosophila mediostriata. Stock H202.23 from Cumanacoa, Venezuela, was used for the construction of this map. Drosophila paramediostriata Townsend and Wheeler, 1955 (1 ) Neuroblast chromosomes are five pairs of rods and one pair of dots. (2) Salivary gland chromosomes are the same size as those of D. mediostriata. The fact that the 2nd chromosome appears to be longer than that of medio­striata is probably the result of stretching. TABLE6 Frequencies and distribution of arrangements found in the second chromosome of Drosophila mediopictoides Stocks Cerro Compana, 2GL-1 Arrangements Boquete, Panama Panama Panama Standard Standard + + 100% ln(2,JN) 35% 42.8% Standard ln(2,JN) ln(2,JN) + + ..;:.. w TABLE 7 ..;:.. Inversion frequwcies found in different geographic stocks of Drosophila medioslriata 2nd Chromosome 3rd Chromosome Stock In(2,AC) In(2,BC) ln(2,DJ)com In(2,FJ) ln(2,Fl) ln(2,FG) ln(3,AD) ln(3,AC) ln(3,EK)com ln(3,F[) ln(3,IK) Cumanacca, Venezuela Caripe, Venezuela - - - - - - 100% 100% Changuinola, Pan:imJ - 27.5% 27.5% 27.5%. 27.5%. 27.5%. 100% 100% 27.5% 27.5%. 27.5%. "3 ;::i-­ Med2llin, ~ Colombi:i Montero, - 100% - 30% - 45% 45% - - - ~ ;::i-· <:::: Bolivia 100% 100% 100% - 100% - - - 40% - ~ "' Trinidad - - 100% - 35% 35% - 35% - ~­..... Vila Atlantica, Brazil - - - 57% 66.6% - 38% 38% - 33% 33% ~ 0- "3 ~ 4th Chromosome 5th Chromosome ~ Stock ln(4,AC) In(4,AD) ln(4,AB) ln(4,DH) ln(4,FH ) In(5,CF) In(5,El) Iu (5,FH) "' Cumanacoa, Venezuela 7.7% - - - - - - - '"i:1 ~ Caripe, \:)-' .......-· Venezuela 35% - - - - - - - ~ ~ Changuinola, ..... c:; · Panama - - 27.5% 27.5% 27.5% 18% 18% 18% ;::i Med21lin, Colombia - 40% 40% 100% Mont2ro, Bolivia - 25% 100% - - 35% 35% Trinidad 35% - 35% Vila Atlantica, Brazil 14% - - - - - 23.8% The University of Texas Publication In (2, BC) In (2,AC), ln(2,BC) In (2,DJ) complex ln(2,FI) In (2,FJ), ln(2,FI) In (3,AD), In (3,AC) ln(3, EK) complex ln(4,AB) ln(4,AC), ln(4,AB) ln(4,AD), ln(4,AB) ln(4,DH) In (4,0H), ln(4,FH) FIG. 16. ~hotographs of the inversions in the third and fourth chromosome of Drosophila mediostriata. A schematic diagram of ln(3,EK)complez is shown. The University of Texas Publication ln(5,EI), In (5,FH) ln(5,CF), ln(5,EI), ln(5,FH) In (3, AC) ln(3,EH), ln(3,FH) ln(4,IJ) TABLES Distribution of arrangements found in the second chromosome of Drosophila mediostriata Chan­Vila Stocks Cumanacoa, Caripe, guinola, Medellin, Montero, Atlantica, AITangements Venezuela Venezuela Panama Colombia Bolivia Trinidad Brazil Standard 100% 100% + + In(2,FI) 9.5% In(2,FI) In(2,FI) 100% In(2,FJ) In(2,FI) 57% In(2,BC)In(2,FI) In(2,BC) 30% In(2,BC) In(2,DJ) com 27.5% In(2,AC) In(2,BC)In(2,FJ)In(2,FG) In(2,AC) In(2,BC) In (2,FJ)ln(2,FG) 100% In(2,BC) In(2,BC) + In(2,BC)In(2,FI) In(2,BC) In(2,FI) + TABLE9 Distribution of arrangements found in the third chromosome of Drosophila mediostriata Chan­ Vila Arrangements Stocks Cumanacoa, Caripe, guinola, Venezuela Venezuela Panama Medellin, Montero, Colombia Bolivia Trinidad Atlantica, Braz;[ Standard 100% + + + + In(3,FI) 40% 25% In(3,FI) In(3,FI) + + ln(3,FI) In(3,IK) 38% ln(3,AD)In(3,AC) 45% 25% 38% In(3,AD)In(3,AC) ln(3,AD)In(3,AC) 100% + + + ln(3,AD)In(3,AC) ln (3,FI) 10% ln(3,AD)ln(3,AC)In(3,EK)com ln(3,AD)In(3,AC) 27.5% ln(3,AD)ln(3,AC)In(3,EK)com ln(3,AD)ln(3,AC)In(3,EK)com + The University of Texas Publication TABLE 10 Distribution of arrangements found in the fourth chromosome of Drosophila mediostriata Chan· Vila Stocks Cumanacoa , Caripe, guinola, Medellin, Montero, Atlantica, Arrangements Venezuela Venezuela Panama Colombia Bolivia Trinidad Brazil Standard ·standard In(4,AC) Standard In(4,AC) In(4,AC) In (4,DH) In(4,DH) In(4.FH) In(4,FH) In( 4,AC) In ( 4,AB) Standard In (4,AC)In(4,AB) In(4,AC)In(4,AB) In( 4,AB) In( 4,DH) In(4,DH) In(4,AB)In(4,DH) In(4,AB)In(4,DH) In(4,AB)In (4,DH) In(4,AD)In(4,DH) In(4,AB)In(4,DH) In(4,FH) In( 4,AD)In(4,DH) In(4,AD)In(4,DH) + + + + 7.7% 35% 14% + + + + + 35% + 25% + + + 45% 27.5% + (3) Chromosome maps, photographic and drawn, are shown in Figures 18 and 19. X chromosomes of male larvae were used for the construction of the map because the X chromosomes of female larvae, similar to those of D. mediostriata, were very poor cytologically. ( 4) Rearrangements: one simple and two included inversions were found TABLE 11 Distribution of arrangements found in the fifth chromosome of Drosophi"la mediostriata Vila Stocks Medellin, Montero, Atlantica, Arrangements cv::~~~l:'v~~::ia i~:~~ Colombia Bolivia Trinidad Brazil Chan· Standard 100% 100% 100% 100% + + + In(5,EI) 15% 23.8% In(5,FH) 15% ln(5,EI)In(5,FH) 20% In(5,CF)In(5,EI)In(5,FH) 18% Frequencies and distribution of arrangements found in the third and fourth chromosomes of Drosophila paramediostriata Stocks Rio Piedras, Mayaguez, Arrangements Puerto Rico Puerto Rico St. Vicente, Cuba Standard Standard + + + In(3,AC)In(3,EH)In(3,FH) 44% 60% 62%Standard In(3,AC)In(3,EH)In(3,FH) In(3,AC)ln(3,EH)In(3,FH) + + + Standard 100% 100% Standard + In(4,IJ) 32%Standard ln(4,IJ) In(4,IJ) + heterozygous in the 3rd chromosome of all stocks. One of the stocks had one very small, simple inversion near the base of the 4th chromosome. Photographs of the inversions are shown in Figure 17, II. Table 12 shows the distribution and the frequencies of the various gene arrangements. Drosophila crocina Patterson and Mainland, 1944 (1) Neuroblast chromosomes are five pairs of rods and one pair of dots. (Q,) Salivary gland chromosomes are small and are very poor cytological ma­terial. There are five arms and one dot. (3) Chromosome maps, photographic and drawn, are shown in Figures 20 and 21. ( 4) The 2nd, 3rd, and 4th chromosomes had a number of heterozygous inver­sions. Table 13 indicates the inversions found and their frequencies in the various geographic stocks. Photographs of the inversions are shown in Figure 22. Tables 14 and 15 indicate the distribution of gene arrangements in the stocks examined. Subgroup IV Drosophila tripunctata Loew, 1862 (1) Neuroblast chromosomes are five pairs of rods and one pair of dots. (2) Salivary gland chromosomes are large. There are five arms and one dot. (3) Chromosome maps, photographic and drawn, are shown in Figures 23 andQ.4. (4) Rearrangements in all chromosomes (except the dot) of this species were found to be polymorphic. Table 16 shows the inversions and their fre­quencies in the different geographic stocks. Photographs of the inver­sions are shown in Figure 25. Tables 17, 18, 19 and 20 indicate the dis­tributions of gene arrangements in the stocks examined. The University of T exas Publication ~ -a .t> IO c E ~ Cl) Cl> E ·~ c "IS ,,, Cb 0 0 <.O ~ E ... .c ~ ~ " C\I ~ :i:" Q. c ... ~ Clc:s 0 ~ 0 .c cs ..s- Frn. 18. Photographic salivary gland chromosome map showing the standard gene arrange­ment of Drosophila paramediostriata. Stock 2327 .2 from Rio Piedras, Puerto Rico, was used fo1 the com lruction of this map. (tlml:t '.fn-mr,r.~t;rr~wmn\\I, x ~.-rii.:w~~r:m F IG H ~ m}1JrnmJf:1!11J~ll11! f~~~i11,:~":1 )r![)tr~1!O~~Jfllff,~'f!flfj M l ~ C' I. In (3, .AC)l . lln (.3,EH) . ( O' ,_ l I'" 3,FH ~ ]n1111·'"1~·1ilf'rrt~;mn -Q.. u 1.;;«n \h ~~\IJ "1JJ{ll.~I ~ 3 ~~~!!,:~1n1!~1~1i111;riM,nllll!11 ""( ~· lI IJ K L i::: ;::s ~ Drosophila poromediostrioto {j Chromosome map n;lW~m1:nmr:n~~miFtJJJll;rffltlli J A Fm. 19. Drawn salivary gland chromosome map showing the standard gene arrangement of Drosophila paramediostriata, the inversions which were detected in the populations examined, and tthe metaphase configuration chromosomes of male individuals. t w t TABLE 13 ~ Inversion frequencies found in different geographic stocks of Drosophila crocina ~ 2nd Chromosome 3rd Chromosome 4th Chromosome ~ Stock ln(2,AB) ln(2,CG) ln(2,CE) ln(2,GL) ln(2,IM) ln(2,KM) ln(3,DE) ln(3,FJ) ln(4,BE) ln(4,DE) ln(4,FJ) ~­ Rio Piedras, Puerto Rico ;;; ~­ Balboa, Panama 45% - 45% 20% - - 70% 70% .Q._ Trinidad 46.6% 40% - 93% 80% 100% 100% 100% 100% ~ El Salvador Costa Rica Jamaica 40% 40% ----17.6% ---35% 35% 35% 40% -40% 40% Note: When frequency of 100% is indicated, the inversion has been found to be carriecl in the homozygous condition. -35% -23.5% -23.5% 33% 29% ~ \:)"',.._ §" ..... g· TABLE 14 Distribution of arrangements found in the second chromosome of Drosophila crocina Stocks llio Piedras Balboa, El Arrangements P. llico Panama Trinidad Salvador Costa llica Jamaica Standard 100% + + + + + In(2,AB) 5.8% In(2,GL) 20% In(2,AB)In(2,CE) 70% 40% In(2,IM)ln(2,KM) 26.6% 23.5% In(2,AB)ln(2,CE)ln(2,GL) 40% In(2,AB)ln(2,IM)ln(2,KM) 11.7% In(2,CE)ln(2,IM)In(2,KM) 20% In(2,CG) In (2,CE) In(2,IM) 13.5% In(2,CG)In(2,IM)ln(2,KM) 26.6% ln(2,CG)ln(2,CE) ln(2,IM)ln(2,KM) 6.6% Drosophila mediodiffusa Heed and Wheeler, 1957 (1) Neuroblast chromosomes are five pairs of rods 'and one pair of dots. The X is approximately twice as long as the Y. There are satellites on one of the autosomes. (2) Salivary gland chromosomes are large. There are five arms and one dot . . . (3) Chromosome maps, photographic and drawn, are shown in Figures 26 and 27. (4) The only rearrangement seen was a simple inversion in the X chromosome of one of the stocks from Petionville, Haiti. The inversion appeared in a frequency of 33 % ; a photograph of it is shown in Figure 5, III. Drosophila metzii Sturtevant, 1921 (1) Neuroblast chromosomes were not studied. (2) Salivary gland chromosomes are of medium size. There are five arms. The TABLE 15 Distribution of arrangements found in the third and fourth chromosomes of Drosophila crocina Stocks llio Piedras Balboa, El Arrangements P . Rico Panama Trinidad Salvador Costa Rica 1amaica Standard In(3,DE)ln(3,FJ) ln(3,DE)ln(3,FJ) In(3,DE)In(3,FJ) 100% + 70% + 100% 100% + 35% + 100% Standard ln(4,FJ) 100% 100% + 33% + 5.8% 100% In(4,BE)ln(4,DE)ln(4,FJ) ln(4,BE)ln(4,DE) ln(4,BE)ln(4,DE) 100% 23.5 The University of Texas Publication Drosophila crocino (photographic FIG. 20. Photographic salivary gland chromosome map showing the standard gene arrange­ment of Drosophila crocina. Stock H131.2 from Rio Piedras, Puerto Rico, was used for the con­struction of this map. quality of the slides was not good. The dot chromosome was not found, but this might have been because of the condition of the slides. (3) Chromosome maps, photographic and drawn, are shown in Figures 28 and 29. ( 4) No rearrangements were observed. DISCUSSION It has been accepted as fact that gene and chromosomal mutations are major contributing factors to the evolutionary process. However, gene mutations have been found to be reversible events, their results being mimicked by similar mu­ tations at other loci (stated by Stone 1955). Chromosomal aberrations are unique events; the probability of the same aberration occurring twice is very low. Con­ sequently, when found in related forms, the aberration can be considered as having a common origin. Dobzhansky and Epling ( 1944) described chromosomal rearrangements as rare events. They further stated that very few inversions have one breakage point in common while the other is a different one; all chromo­ somal rearrangements are results of two chromosome breaks. Bauer, Demerec, and Kaufmann (1938), and Helfer (1941), using D. melano­ gaster and D. pseudoobscura, respectively, proved that the distribution of inver­ sions on the chromosomes of irradiated flies is random with a possible increase in frequency at the distal ends of the chromosomes. The University of Texas Publication . In (2, AB) \. ~' In (2,AB), In (2, CE) ln(2,CG) In (2,IM) ln(2, IM), In (2,KM) In (3, DE) ln(3,FJ) In (4, BE) , In ( 4,DE) In (4,FJ) FJG. 22. Photographs of inversions in the chromosomes of Drosophila crocina. X Chromosome 2nd Chromosome 3rd Chromosome 4th Chromosome 5th Chromosome Stock In(I,BH)com In(I,KM) In(2,AC) In(2,EI) In(2,LP) ln(2,MO) In(3,EG) In(3,JM) In(4,AD) In(4,EG) In(4,HK) In(5,GK) TABLE 16 Inversion frequencies found in different geographic stocks of Drosophila tripunctata Dexter, Missouri - - 43% - - - - - 47% - 47% New Orleans, Louisiana 100% 100% - 100% 41.8% 41.8% Albemarle, North Carolina 43% 43% 25% 39% - 44% 45% 48% 19% 7% 33% 24% Biloxi, Mississippi 48% 48% 43% 43% - 43% 100% 36% 23% 23% 28% Austin, Texas - 31.7% 40% - - - 19% 25% 31% 26% 23% 23% Nacogdoches, Texas - 100% - 54% - 56% - 45% 69% - 67% Everglades, Florida - - - 100% 49% 49% - 40% 47% 47% Crystal Lake, Nebraska - 100% 50% - - - 47% 47% - - - 41 % When frequency of 100% is indicated, the inversion has been found to be carried in the homozygous condition. The University of Texas Publicat:on Q) ..... \.) E I::; ~ ~ 0 .$ E \... ~ ..... .i:: " ., 1::3 .i:: " ...... 0. ~ ~ ~ c:r> 0 0 Cl "' .i:: 0. x (!) Ci N FIG. 23. Photographic salivary gland chromosome map showing the standard gene arrange­ment of Drosophila tripunctata. Stock 1910.4 from Dexter, Missouri, was used for the con­struction of this map. Stone (1962) mentions that if the basic metaphase configuration of the genus is considered, and if a random distribution of breaks is assumed then the fre­quencies of rearrangements should be: translocations > paracentric inversions > pericentric inversions > fusions. However, observations on natural populations ...I .... ~ ~ ____9.__e s a 0 LL./ 0 "' ~ ~ C5 ~ 0 gw '"":) .. ; = =­ c.. r ~.,; ~ x E LO . CD £-­ ~ ~ ~ v "ti .<:> "ti Q. Q) 0 e: E ., :.$! E 0 ~ g. "' 0 E II) ~ <:> .c ~ (.) ~ .._ U) FIG. 27. Drawn salivary gland chromosome map showing the standard gene arrangement of Drosophi.la mediodifjusa, the inversion detected in one of the populations examined, and the metaphase configuration chromosomes of male individuals. . The University of Texas Publication TABLE 19 Distribution of arrangments found in the third chromosome of Drosophila tripunctata New Albemarle, Nacog-Crystal Stocks Dexter, Orleans, North Biloxi, Austin, doches, Everglades, Lake, Arrangements l'vlissouri Louisiana Carolina Mississippi Texas Texas Florida Nebraska Standard Standard 100% 100% + + + + + ln(3,EG) Standard 8% 1% 5% ln(3,JM) Standard 11 % 7% 45% 40% 5% ln(3,EG) ln(3,EG) + ln(3,JM) In(3,JM) + + ln(3,EG)ln(3,JM) Standard 37% 18% 42% ln(3,EG)ln(3,JM) ln(3,EG) 36% ln(3,EG)ln (3,JM) ln(3,EG)ln(3,JM) + + + + TABLE zo Distribution of arrangements found in the fourth and fifth chromosomes of Drosophila tripunctata New Albemarle, Nacog-Crystal Stocks Dexter, Orleans, North Biloxi, Austin, doches, Everglades, Lake, Arrangements lVlissouri Louisiana Carolina Mississippi Texas Texas Florida Nebraska Standard 100% 100% + + + + + + ln(4,AD) 8% 5% 11% In(4,EG) 2% ln(4,HK) 24% 26% 16% 9% ln(4,AD)ln(4,EG) 2% 21% 19% 47% ln(4,AD)ln(4,HK) 47% 6% 58% In( 4,AD)In( 4,EG) ln(4,HK) 3% 2% 7% Standard 100% 100% 100% 100% 100% Standard + + + ln (5,GK) 24% 23% 41% Standard In (5,GK) ln (5,GK) + + + show that the number of paracentric inversions is much larger than that of trans­ locations and the number of chromosomal fusions is greater than that of peri­ centric inversions. Drosophila species, as many other sexually reproducing species, are found to be polymorphic in several different ways; one is the inversion polymorphism observed in the salivary gland chromosomes. Dobzhanksky (1951) summarizes observations made on the subject in populations of D. pseudoobscura. The same chromosomal rearrangements appear in different frequencies in the various pop­ ulations, while some populations carry rearrangements not found in others. In­ versions in the heterozygous condition serve as cross-over supressors, thus allow­ ing the population to maintain gene sequences which have been selected as ad­ vantageous for a certain set of environmental conditions. In some cases, inversion homozygosity is promoted; thus, recombination can occur, permitting some more experiments of these species toward adaptation in a new set of circumstances as suggested by Carson ( 1965, 1959, 1961) for the case of small marginal popula­ tions. The results obtained from studies in different species of Drosophila fall into three categories: ( 1) species with a rather stable gene sequence in which very little or no chromosomal polymorphism is found, as, for example, the repleta group as discovered by Wharton (1942) and Wasserman (1954), and in D. virilis as reported by Patterson, Stone, and Griffen ( 1940), and Warters ( 1944) ; ·.(2) species in which all the chromosomes show several gene arrangements, for example D. ananassae as found by Kaufmann (1936) , Kikkawa (1938), Dob­zhansky and Dreyfus (1943), D. algonquin as found by Miller (1939), D. robusta as found by Carson and Stalker (1947), D. willistoni (with forty inversions evently distributed among the chromosomes) as found by da Cunha, Burla, and . Dobzhansky (1950), to mention a few of many such examples; and (3) species with most of the chromosomes carrying a rather stable gene sequence with the exception of one chromosome which is extremely polymorphic, for example D. nebulosa as found by Pavan (1946) , species of the melanica group as found by Ward (1962) and Stalker (1960, 1964a,b), and D. pseudoobscura as observed by Dobzhansky and Sturtevant ( 1938). The existence of overlapping inversions in populations of one or more related species usually permits the discovery of the steps which were probably involved in the phylogeny of the species. Both the results of the hybridization tests and the examination of the salivary gland chromosomes performed in this study support the conclusion of Patterson (1957) that there is a high degree of isolation among the various species of the group. In the next paragraphs the inversion polymorphism found in the different species from which several geographic stocks were available will be discussed and an attempt will be made to propose possible schemes concerning the relation­ ships of the various gene arrangements. In D. unipunctata most of the inversion polymorphism was found in two arms in the salivary gland chromosomes, the left arm of the second chromosome and the right arm of the third. The X chromosome was found to exist in three differ­ ent arrangements; most of the stocks were homozygous for two overlapping, The University of Texas Publicat:on x a.. z ...J 0 I{) :;::: -t:! a. x ~~ b CD :-=::; E ~~ ~~ ~~ c::s c3 Fm. 29. Drawn salivary gland chromosome map showing the standard gene arrangement of Drosophila metzii. The University of Texas Publication possibly pericentric, inversions involving a part of the chromocentric hetero­chromatin which has been found to exist in all tripunctata group species. Only one stock from Barro Colorado, Panama, carried those two inversions in the heterozygous condition, as well as both the inversion and the normal homozy­gotes. Since the inversion homozygote is very widespread, ranging from Costa Rica to Venezuela, it is possible that it is the most primitive arrangement. If one considers the geographic locations from which the unipunctata stocks were collected, one observes that the number of heterozygous inversions in the left arm of the second chromosome increases as one goes from Central to South America, the most complex combinations being found in the two stocks from Colombia. The available data indicate certain relationships between the different gene arrangements on 2L which occur in the various geographic stocks. The hypo­thetical scheme, A, below, shows the probable relationships. SCHEME A In(21, HJ)~----In(21,HJ)In(21,IM) ' In 21 In 21 IM / In(21,AB)In 21,DJ) + / In(21,DJ)< ~ In(21,AB)In(21,DJ) t or / 1 In(21,AB) ~ 1 / [complex inversion contains In(21, BF) and In(21,EK) and probably In(21,HilJ 1 I r In(21 .AB)In(21, BK)complex except In(21, EK)~---~ In(21,A};l)com plexin(21, HJ) In(21,AB)In(21, EK) ... \ [Complex inversion probably \ contains In(21,AB) and \ -4 In(21, BF)) In(21, BF) In(21,EK) Note: Dotted lines indicate that there are probably several steps in the s'hown change. This scheme was derived as follows: The stock used as the standard originated in the southwestern part of Costa Rica and carried no heterozygous inversions in the left arm of the second chromosome. The next nearest stock, geographically speaking, came from Changuinola, Panama, in the northeastern part of that country; it presented In(2L, HJ)/Standard as well as the standard and the inver­sion homozygote arrangements. The next stock came from Cerro Campana, Panama, near the Canal Zone, where In(2L, HJ) was found to be homozygous, while In(2L, IM) appeared as a heterozygous inversion with its proximal end near the chromocenter and the distal end within the limits of In(2L, HJ). Both of the other stocks from Panama originated from Barro Colorado Island in the Canal Zone; one of these stocks carried the combination In(2L, HJ) In(2L, IM)/ Standard as well as both the standard and the inversion homozygote arrange­ments. The other stock carried the combination In(2L,HJJin(2L,IM/In(2L,ABJ In(2L, DJ) and also both kinds of inversion homozygotes. The stock from Medel­lin, which is located in the northeastern part of Colombia, was found to carry a complex inversion next to /n(2L, HJ). When the stock was crossed to the stand­ard, the Fi offspring were found to have the chromosomes of either one or the -other of the parents. This would indicate that all the inversions are carried on one chromosome. It is possible that ln(2L, AB) is part of the complex, as is ln(2L, BF). The latter is also retained heterozygous in the stock from Rio Negro, Colombia, in combination with ln(2L, EK) and ln(2L, HJ) as parts of the /n(2L, BKJcomplex. The Rio Negro stock, when crossed to the standard, gave two kinds of offspring: In(2L, AB)In(2L, EK)/Standard and In(2L, AB) ln(2L, BK)complex except ln(2L, EK)/ Standard. Therefore, ln(2L, EK) and ln(2L, BF) are carried in different chromosomes; this condition is retained in the Venezuelan stock which, when crossed to the standard, gave Fi offspring which were all heterozygous for the one or the other inversion. No heterozygous rearrangements were found in the right arm of the second or the left arm of the third chromosome of D. unipunctata. The right arm of the third chromosome of the Costa Rican stock of unipunctata carried In(3R, CD) overlapping with In(3R, DG). This arrangement seems to be the most widespread heterozygous arrangement found in the populations ex­ amined. Three out of four stocks from Panama were found to carry these overlapping inversions, while the fourth (from Changuinola) carried only ln(3R, DG). ln(3R, CD) and ln(3R, DG) were carried in different chromo­ somes as can be seen from the photograph of the overlapping inversions in Figure 10. ln(3R, CD) was retained in both the Colombian and the Venezuelan stocks. The first appearance of ln(3R, AB), ln(3R, FK), and In(3R,HK) was in the stock from Medellin, Colombia. When combined with In(3R, DG), they form /n(3R, CKJcomplex in Rio Negro, Colombia. As can be seen in Table 5, /n(3R, ABJ/n(3R, CKJcomplex is the only heterozygous arrangement observed in the Rio Negro stock; no cross-over types have been observed. This is in contrast to the Venezuelan stock in which some cross-over products were recovered. For example /n(3R, AB) and /n(3R, CDJ/n(3R, DGJ/n(3R, FK)]n(3R, HKJ can result from single crossing over between the complex and the short inversion in the tip. ln(3R, AB)In(3R, FK)ln(3R, HK) is probably the result of double cross­ ing over, although the other class of offspring has not been recovered. One may conclude that the number of heterozygous inversions in 3R increases when going from Costa Rica to Colombia and Venezuela. The following hypo­ thetical scheme (B) is proposed in an effort to explain the way the various gene arrangements have been produced in the right arm of the third chromosome of D. unipunctata. · Wharton (1943) suggested that the J chromosome which appears in the meta­phase configuration of D. unipunctata probably results from a fusion of part of the microchromosome to one of the rods, while the free centromere with some euchromatic material represents the small dot seen in the metaphase plate. Examination of the polytene chromosomes seems to suggest a different explana­tion: it was found in all cases that the arms which represent the second and the fifth chromosome in the rest of the species of the group, as well as the arms which represent the third and the fourth chromosome, appear to be fused to each The University of Texas Publication /In(3R,DG) + >ln(3R,CD)ln(3R,DG) SCHEME B \ln(3R,CD) In(3R ,AB)In(3R, CD)In(3R, DG)In(3R, FK)In(3R, HK) t In(3R,AB)In(3R, CD) 'In(3R, FK)In(3R, HK) / ~ \ In(3R,AB)in(3R, CD)In(3R, HK) or In(3R, CD)In(3R, FK)In(3R, HK)/ t In(3R, CD)In(3R, HK) other, respectively, in unipunctata. This suggests that the V chromosome of the metaphase is the result of a fusion between the third and the fourth chromosomes which are approximately equal in length while the long arm of the J represents the second, and the short arm represents the fifth chromosome. Following this assumption, the dot would represent the obscure short arm in the salivary gland chromosome complement. The last rod of the metaphase could be heterochromatin added to a centromere freed by one of the fusions and is not represented in the salivary gland smears. The data collected from the cytological examination of Drosophila mediostriata stocks do not permit a great number of conclusions. The stocks, as in most of the species examined, were highly inbred. They originated from different numbers of flies, the samples never being large enough to give a clear cytological picture of the species. The two Venezuelan stocks were found to carry no heterozygous rearrange­ments in the second chromosome. The stock from Changuinola, Panama, presents a very different picture from that of the rest of the stocks; it appears to be a species hybrid between D. mediostriata and D. paramediostriata and will be dis­cussed later. The stocks from Colombia, Trinidad, and Brazil were found to carry ln(2, Fl), either heterozygous or homozygous, alone or in combination with other inversions. ln(2, F J) appears only in the Bolivian and Brazilian stocks, while ln(2, BC) is carried homozygous in the stocks from Colombia and Bolivia. Finally, ln(2, AC) and ln(2, FG) appear homozygous only in the Bolivian stock. The existence of ln(2, Fl) alone in some of the individuals from Brazil might be explained as the result of double crossing over, although the other product has not been recovered. No arrangement, other than the arbitrarily chosen standard arrangement, has been found to exist in more than one stock. Scheme C, below, is proposed to indicate the relationships of the arrangements of the second chro­mosome of mediostriata. The standard stock from Cumanacoa, Venezuela, did not carry any heterozy­ SCHEME C ~.,.In(2,FJ) Io(2,FJ)Jn(2,Fp~ In(2,Fl)~J > In(2,BC)In(2,FJ) In(2,BC)In(2,FI)c t / l --Hn(2,BC) In(2 ,AC)In(2, BC)In(2 ,FJ) f In(2 ,AC)In(2, BC)In(2, FJ)In(Z, FG) gous arrangements in the third chromosome. The stock from Caripe, Venezuela, when crossed to the standard, showed that it carries the included inversions ln(3, AD)ln(3, AC) homozygous. The same included inversions were found to exist in the heterozygous condition in the stocks from Trinidad, Colombia, and Brazil. The Trinidad strain also showed an ln(3, Fl) arrangement and an ln(3, AD)ln(3, AC)ln(3, Fl) arrangement. After crossing to standard it was found that ln(3, AD) and In(3, AC) are carried in the same chromosome, while ln(3, Fl) is in a different one. It should be concluded, then, from the results of this cross and the data presented in Table 9, that the structure of the Trinidad stock is In(3,AD)ln(3,AC)/ln(3,Fl). The arrangements ln(3,AD)ln(3,AC) and ln(3,Fl) represent cross-over classes of the following constitution: ln(3,AD) ln(3,AC)ln(3,FI)jln(3,AD)ln(3,AC), ln(3,AD)ln(3,AC)In(3,FI)jln(3,Fl), ln(3,Fl)/+, and In(3,AD)ln(3,AC)/+. ln(3,FI) was the only heterozygous inversion in the third chromosome which has been observed in the Bolivian population. When the Brazilian stock was crossed to the standard, two classes of F1 off­spring appeared, namely ln(3,AD)ln(3,AC)/+, ln(3,Fl)ln(3,IK)j+. It is easily .understood, then, that the two sets of inversions are in different chromosomes; in such a case, a class of individuals heterozygous for all four inversions should . be' expected but such individuals were not found in the actual results. Given that the size of the sample is not very large, the most probable explanation is that the existence of the four arrangements together reduces the viability of such indi­ viduals. · However, this is only speculation with an insufficient amount of sup­ porting evidence. Considering the standard as a starting point, the following scheme (D) is pro­ posed for the inversions of the third chromosome of mediostriata. The standard Venezuelan stock of mediostriata carried In(4,AC) heterozygous in the fourth chromosome in a very low frequency. The other stock from Vene­ zuela showed the same inversion but in a higher frequency. In(4,AC) was also found heterozygous in the Brazilian stock. The stock from Trinidad carried the same inversion but in combination with ln(4,AB) which was included in it. ln(4,AB), on the other hand, was observed heterozygous and in combination with the homozygous ln(4,DH) in the Bolivian stock, a fact which was easily demonstrated by crossing the stock with standard and obtaining two kinds of offspring in the F1 generation: ln(4,AB)ln(4,DH)/+ and ln(4,DH)/+. The Colombian stock presented one more inversion, In(4,AD), which was in a differ­ ent chromosome from that in which ln(4,AB) was found. When the stock was The University of Texas Publication crossed to standard, the F 1 offspring were of the following constitution: In(4,AB) In(4,DH)/+ and In(4,AD) In(4,DH)/+. From the data presented above it can be seen that In(4,AC) seems to be a well­adapted arrangement retained across a long distance in the populations of Vene­zuela, Trinidad, and Brazil. On the other hand, In(4,AB) must also be well­adapted because it, too, appears in the populations of Trinidad, Colombia, and Bolivia. The populations from North and East seem to favor In(4,AC), while the Western populations favor In(4,AB) ; both appear together in an island popu­lation. In(4,DH) was found to be homozygous and usually in association with In(4,AB). If standard is taken as the starting point, then the following scheme (E) could represent the relationships between the inversions on the fourth chromosome of mediostriata. SCHEMED+ +----In(3,FI) In(3,FI)In(3,IK) __,,, ,.., ' ~ 1/ ' >!,. ~ In(3 , AC) In (3 ,AD) In(3 ,AD)In(3 ,AC) ~-~ In(3, FI) \ / In(3 ,AD)In(3 ,AC) , \ / In(3,FI)In(3,IK) '"' ~ In(3 ,AD)In(3 ,AC) / Note: Dotted lines indicate the two alternative ways arrangement In (3 ,AD)In(3 ,AC) could be produced. SCHEME E + ~--~ In(4,AC) ,AB) ""'- In(4,AC)In(41 !~"~ / I In(4,DH) ~ )In(4,AB) In(4,AB)In(4,DH) In(4 ,AD)In(4, DH) I ~ In(4 ,AD}In(4, DH) In(4,AB}In(4,DH) SCH EME F + In (5, EI) In (5, EI) In (5, F H) The standard arrangement was the most widespread sequence found in the fifth chromosome of Drosophila mediostriata. It occurred in all stocks examined and it was the only sequence found in the stocks from Venezuela, Trinidad, and Colombia. The Brazilian stock presented ln(5,EI) which was also found in the Bolivian stock in combination with ln(5,FH) which was included in it. Classes ln(5,El) and ln(5,FH) (see Table 11), found in the Bolivian stock, should be considered as cross-over classes. If the standard arrangement is considered as primitive for the fifth chromo­ some, scheme F, above, could be proposed. All the stocks of Drosophila paramediostriata which were examined presented the same three arrangements in the third chromosome; ln(3,AC) appears to be the same as that found in Drosophila mediostriata, the only difference being that ln(3,AC) of mediostriata has always been found associated with ln(3,AD). The . stock from Mayaguez, Puerto Rico, also carried a very small inversion in the proximal end of the fourth chromosome (see Table 12). As mentioned before, Drosophila mediostriata was found to cross readily with Drosophila paramediostriata, producing viable and fertile hybrids. The cyto­ logical examination of these hybrids provided a considerable amount of infor­ mation concerning the chromosomal differences between the two species. The X chromosome appeared to be the same in both species. This could not be confirmed, however, because, as was mentioned before, the female individuals present an X chromosome of such a condition that a detailed cytological examination was practically impossible. The second chromosomes were found to differ by ln(2,BC) and ln(2,DJ) complex, which was the arrangement found in the stock of medio­ striata from Changuinola, Panama. It would seem that these inversions are carried homozygous in paramediostriata. Jn(2,DJ) complex contained ln(2,Fl), ln(2,FI), and ln(2,FG) as well as at least one more inversion, the distal break point of which is obviously located in region D of the second chromosome. In the third chromosome the hybrid larvae were found to carry In(3,AD) and ln(3,EK) complex. Of course, the study of ten hybrid larvae cannot be considered as a statistically adequate sample. Nevertheless, all ten of them showed the above mentioned pattern. Similar to the above is the situation found in the stock from Changuinola, Panama, with the only exception that ln(3,AD) and ln(3,AC) were found to be homozygous in that stock. It seems that In(3,EK) complex con­ tains ln(3,Fl) and In(3,IK) of mediostriata, but no safe assumptions can be made for the third chromosome due to the fact that the standard stock of paramedio­ striata carried three heterozygous inversions, the involvement of which in ln(3,EK) complex is not certain. It seems that ln(3,AD) is carried homozygous in D. paramediostriata, while ln(3,AC) is still retained in the heterozygous con­ dition. This is in contrast to mediostriata where these two inversions are usually both found heterozygous. The fourth chromosome of the species hybrids appeared to be exactly like the fourth chromosome of mediostriata from the Changuinola stock which was discovered to be heterozygous for In(4,AB),ln(4,DH) , and ln(4,FH). Finally, the fifth chromosome of the hybrid resembled that of the Changuinola stock carrying the arrangement ln(5,CF)In(5,EI)ln(5,FH). Dro­ sophila paramediostriata is a species of island distribution. Since Changuinola is near the coast in Panama where many fruit boats from the islands unload or pick up their cargo, it would be possible to have transportation of an island species to the mainland, an event which could very easily result in the for­ mation of species hybrids. When one further considers the fact that the species The University of Texas Publication greatly resemble each other morphologically and cross fairly readily in the lab­oratory, it can be concluded that the stock from Changuinola, Panama, most probably represents a species hybrid found in nature. The data presented in Tables 13 and 14 on the rearrangements observed in the second chromosome of the populations of Drosophila crocina do not allow a great many conclusions. It is obvious that this chromosome is the most poly­morphic of the species. It is unfortunate that the species is not good for cytological studies. Consequently, the size of the samples taken from each stock of crocina probably did not represent the real picture of the chromosome arrangements. At any rate, the available data suggest scheme G below. SCHEME G In(2,C G)In(2 ,CE)In(2 ,IM)In(2,KM) -------­ \ ln(2,CE) In(2,AB/)In(2,CE)In(2,GL) In(2,CE)In(2,IM),2,KM) ~ In(2 ,AB)In(2, CE) J-' --~ I In(2 ,IM)In(2 ,KM) ,?f + ~In(2,AB) ~ / '-'rn(2 ,KM)>'/ In(2 ,AB)In(2, IM)In(2, KM) Note: Dotted lines indicate the two alternative ways arrangement In(2, IM)in(2, KM) could be produced. It can be seen from Table 13 that ln(2,AB) appears in four of the examined stocks, the same being true for ln(2,CE) . Nevertheless, it is obvious that these two inversions do not always appear together. It could be suggested, then, that each one has occurred independently of the other as a one-step rearrangement from the standard sequence. Something similar could hold true for ln(2,IM) ln(2,KM) as resulting from a two-step rearrangement from the standard. It is not possible to determine which inversion took place first because the inversions are only included. From the sequences mentioned above, the rest of the ar­rangements observed in the populations could be explained either as products of a combination of two or more inversions, or as products of a further rearrange-. ment of the chromosomes where the previously mentioned inversions exist. It is obvious, of course, that the scheme proposed above does not include all the possible ways in which the events could have happened, but it indicates the simplest ex­planation. All other arrangements can be considered as recombinant classes, although in most cases some of the expected combinations are missing. Kastritsis: Cytology of Tripunctata Species Group The third and fourth chromosomes of crocina have not presented many hetero­zygous rearrangements. ln(3,DE)ln(3,FJ) appears heterozygous in the stocks from Panama and Costa Rica, while the Trinidad strain carries the inversions in the homozygous condition. No recombination classes have been observed. ln(4,FJ) appears heterozygous in the stock from El Salvador and represents the only inversion found in the fourth chromosome of that population. The same inversion appears as a recombinant class in the Costa Rican stock, although the other recombinant class has not been observed. In(4,BE) ln( 4,DE)ln(4,FJ) seems to be the dominating heterozygous arrangement. In(4,BE)ln(4,DE) was found homozygous in the stock from Trinidad. Heed (1957), after examining three stocks, one from Puerto Rico (the same stock as used in this study), one from Trinidad, and one from El Salvador, found that there was some morphological differentiation in different stocks of crocina. His work pointed out that the Trinidad strain was highly differentiated from the Puerto Rican strain and less so from the El Salvador strain. Although the stocks from Trinidad and El Salvador used in this study were not the same ones used by Heed in his experiments, they present characteristics similar to those described in the 1957 paper. An observation of the results of the cytological examination raises the question of whether the inversion differences which exist in these · populations (see Table 13) could have any bearing on the above mentioned differentiation. An investigation dealing with this problem is under way. It can be seen from Tables 16 and 17 that /n(J ,KMJ of the X chromosome of Drosophila tripunctata appears with great frequency in the stocks examined in this study. In three stocks the inversion appears homozygous either alone, as in the stocks from Nacogdoches, Texas, and Crystal Lake, Nebraska, or in combina­tion with Jn(J, BHJcomplex as in the stock from New Orleans, Louisia~. Finally, both ln(J, BH)complex and ln(J, KM) appear in the heterozygous condition in the stocks from North Carolina and Mississippi. The arrangements /n ( J, KM)/ standard and ln(J, BH)complex/standard of the North Carolinian stock should be considered as recombination classes. If ln(J, KM), the most widespread rear­ ·. rangement, is considered as more primitive than In( 1, BH)complex, the following diagram (H) can be used to illustrate the events involved in the evolution of the X chromosome of tripunctata. It seems that the rearrangements, which occupy almost half the total length of the X chromosome, are well adapted, creating an arrangement which would not permit recombination, at least within the limits of the inversions. Tables 16 and 18, indicating the frequencies and arrangements of the inver­sions found in the second chromosome of tripunctata, suggest the follawing scheme (J). SCHEME H +~In(l,KM) ~-----------~In(l,BH)complexln(l,KM) Note: Dotted line represents a series of inversions not found as independent in the populations, but being part of In(l.BH)complex. The University of Texas Publication In!2 ,EI)In(2, MO) ~In2,AC) In(2,AC) ; ­ +~I) In(2,EI)In(2,MO) In z"riIn 2 MO \ _____,,.... In(2 ,EI)In 2 ,LP ln(2,El)In(2,LP) ~ Considering "+" as the starting point, ln(2, AC) could arise as a one-step rearrangement from it, a fact which is supported by the existence of the arrange­ments ln(2, AC)/standa.rd and In(2, EJ)ln(2, MO)/ln(2, AC). On the other hand, In(2, El) is involved in two kinds of arrangements: ln(2, EI)ln(2, MO) and ln(2, EJ)ln(2, LP) as is demonstrated when the stock from New Orleans, Louisiana, is crossed to the standard stock, yielding the above mentioned arrange­ments in an approximately 1: 1 frequency. The least complicated way of obtain­ing these combinations would be to consider ln(2, El) as a one-step rearrange­ment from the standard, and then have In(2, MO) and ln(2, LP) taking place independently and producing the combinations ln(2,El ) In(2,MO) and ln(2,El) ln(2,LP). The existence of both kinds of inversion homozygotes seems to support this hypothesis. In(2,AC) seems to be a well-adapted inversion found in all but two of the stocks. ln(2,MO) also presents a rather wide distribution correspond­ing to that of ln(2,El). ln(2,LP) has been observed only in the southern popu­lations of New Orleans, Louisiana, and Everglades, Florida. The existence of the many arrangements in the stock of tripunctata from North Carolina can be ex­plained as a result of recombination. Although the sample taken was a large one, no recombination classes have been found in the stock from Biloxi, Mississippi, which carries the same inversions as that of North Carolina. Table 19 illustrates that ln(3,JM) is widely distributed, found either alone or in combination with ln(3,EG) in all but two of the populations examined. /n(3,EG) is less frequent, while only two stocks were found to carry the "stand­ard" configuration in a 100% frequency. Both Jn(3,EG) and Jn(3,JM) are in the same chromosome and remain inseparable from one another when the stocks which carry them are crossed to the standard. The classes ln(3,EG)/ standard and ln(3 ,JM)/standard, which appear in the populations of tripunctata from Austin, Texas, North Carolina, and Nebraska, are apparently recombination classes resulting from crossing over between the two inversions. If the standard is considered as starting point, then the sequence of events which have occurred could be represented by scheme K below. Finally, from Table 20, which indicates the arrangements of the inversions in the fourth and fifth chromosomes, the following conclusion can be made: ln(4,AD) and Jn(4,HK) are widespread in the populations of Drosophila tri­punctata. It seems that these inversions have high adaptive value and are retained as heterozygous even in highly inbred stocks; on the other hand, In(4,EG) appears only in four of the eight stocks examined. By observing the data pre­sented in Table ·20, one arrives at the conclusion that when Jn(4,EG) is combined with In(4,AD) and In(4,HK), it does not produce a combination of high adaptive value because the inversions are continuously separated by crossing over to pro­duce recombination classes which are found quite frequently in the stocks. From the available data, the following scheme (L) is suggested in an effort to explain the sequence of events involved in the evolution of the fourth chromosome of D. tripunctata. SCHEME K '~ + 'f -)' In(3,IM)~----~ In(3,EG)In(3 ,IM) Note: Dotted lines indicate the aiternative way In(3, EG)In(3, IM) could be produced. SCHEME L In(4,AD) In(4, AD)In(4, HK) •( >fo(4,AD)ln(4,EG) In(4,EG) ~In(4 ,AD)In(4 ,EG)In(4 ,HK) Most of the stocks examined present the standard arrangement in the fifth chromosome which, if considered as primitive, would give In(5,GK) as a one­step rearrangement from it. The stocks of Drosophila mediodifjusa used in this study represented popula­tions from almost the entire area of species distribution (with the exception of Cuba) and exhibited practically no intraspecific chromosomal variation (the only exception being the observation of a heterozygous inversion in the X chromo­some). It seems that the standard arrangement of mediodifjusa is well adapted. Because the populations are small, there is a tendency toward maintaining that homozygosity. The remaining four species were represented by very few stocks (one in three of the species, and three in D. mediopictoides) and these were not sufficient to permit a study of inversion polymorphism. The inversions found in these stocks were mentioned in the "Results" portion of this study. However, the chromo­somes of these species proved very interesting because they presented several characteristics common to the chromosomes of the tripunctata species group. The following paragraphs are an attempt to point out these common characteristics, as well as to mention such observations which might prove to be of interest in further investigations. The X chromosome is readily identifiable in the salivary gland smears because it was found to carry a large heterochromatic mass at its proximal end. The distal end of all the X chromosomes, except those of subgroup IV, is occupied by The University of Texas Publication a puff-like structure which is followed by two dark bands and, in most cases, three subsequent dark bands. In subgroup IV, tripunctata presented a very long X chromosome with an excessively large puff in region C which appeared in all the approximately 1500 slides which were examined; its position was altered in the stocks homozygous for (ln1,BH) complex. On the other hand, the tips as well as the bases, of the X chromosomes of mediodifjusa and metzii are very similar. Regions K-M of mediodiffusa are probably the same as regions K-M of metzii. Region J of metzii is represented in mediodifjusa as part of regions D and E. The remaining regions of the X chromosome are so extensively rearranged that it is virtually impossible to point out similarities among these chromosomes. In all the species with a known metaphase configuration, the X chromosomes are repre­sented by one pair of rods. There are a few features of the second chromosome of the tripunctata group species which make it readily identifiable. It is the longest chromosome of the complement, usually with a small puff-like region at the tip which follows five bands, two or three of which are dotted bands. There is one section found in the second chromosome which appears in all but two of the species (tripunctata and mediodiffusa). This section occupies the first half of region Nin mediodelta, the second half of region K in the 2L arm of unipunctata, the last portion of region F in mediopunctata, the first half of region F in mediopictoides, region I in medio­striata, region H in paramediostriata, the first half of region K in crocina, and the first portion of region F in metzii. Other similarities in banding structure of this chromosome are thought to exist, but, since the chromosomes are highly rearranged, no definite statements can be made. The second chromosome is represented by one pair of rods in five of the seven species with a known meta­phase configuration. One pair of J's represents the fusion product of the second and fifth chromosomes in unipunctata. One pair of V's with unequal arms prob­ably indicates the presence of additional heterochromatin to the second chromo­some of mediopictoides. The right arm of the second chromosome of unipunctata will be considered for purposes of convenience together with the fifth chromo­somes of the other species. The third chromosome of the tripunctata group species was by far the most easily identified chromosome because there are portions of this chromosome which are found in every species examined. Both the proximal and distal ends of this chromosome were found to present similar banding patterns in most cases throughout the different species. The chromosome maps of the various species indicate that essentially identical portions of the third chromosome exist in every species, i.e., region E in mediodelta, region C in mediopunctata, region D and part of E in unipunctata, region G in mediopictoides, region F in mediostriata, region F in paramediostriata, parts of regions F and G in crocina, region E in tripunctata, region E of mediodifjusa, and parts of regions D, E, and F in metzii are alike. Although it is impossible to determine the changes which have occurred during the evolution of this chromosome, it is relatively safe to suggest that it has not undergone quite as much change as have the other chromosomes. In most cases the third chromosomes are represented by one pair of rods in the known metaphase plates. In D. unipunctata, however, they are found fused to the fourth chromosomes. The situation in the fourth and fifth chromosomes of the studied species ap­peared more complicated than that found in the other chromosomes. The process of exclusion was usually applied to identify the fourth chromosome of the various species. In some cases it was possible to see similarities in the tips and bases of the chromosomes, but nothing could be established as a definite rule. The size of the chromosome, being somewhat longer than the fifth, helped in its identifica­tion. In metaphase plates the fourth chromosomes are represented by one pair of rods, except for D. unipunctata in which they are fused to the third chromosomes to form one pair of V's. With some experience one can identify the fifth chromosome of the tripunctata group in the salivary gland smears relatively easily: two gray bands are followed by two dark bands, a non-banded short region, one gray band, and, finally, two more dark bands in the distal end of the chromosome. The fifth was found to be the shortest chromosome of the complement. D. mediodiffusa presented a puff­like structure distally. The base of the chromosome resembles that of tripunctata with some other regions similar to regions of the fifth tripunctata chromosome (i.e., region K is similar to region F of tripunctata and inverted, and regions F, G, and part of Hare similar to regions G and Hof tripunctata and also inverted) . In metaphase plates the fifth chromosomes were usually represented by one pair of short rods which, in the case of unipunctata, were fused to the pair of rods representing the second chromosomes. Both the fourth and fifth chromosomes should be considered as chromosomes which have undergone a great amount of rearrangement during their evolution but appear now as rather stable structures with a successful gene arrangement well adapted in their environment. The salivary gland chromosomes of D. bandeirantorum, as presented by Salzano ( 1963), seem to have some characteristics in common with the chromo­somes of the tripunctata group species examined in this study, but no accurate comparisons can be made by using the maps presented in his paper. It should be pointed out that the data presented in this paper provide a very good case of adaptive polymorphism as exhibited by numerous inbred laboratory stocks which retained, in most case, many chromosomal rearrangements in the heterozygous condition. An interesting and potentially important observation was made: it was observed that the salivary chromosomes of the tripunctata group species are in some ways similar to the chromosomes of D. melanica. The second, third, fourth, and fifth chromosomes of melanica correspond respectively to the second, fourth, fifth, and sixth chromosomes of the tripunctata group. The XR and XL of melanica correspond to the X and third chromosome of the species used in this study. It is felt that further investigations on the subject could be very significant in the study of evolution in the genus Drosophila. SUMMARY 1. Forty-nine geographical strains distributed among eleven different species of the tripunctata group of Drosophila were used for hybridization tests and cyto­logical investigation. Photographic and drawn chromosome maps of ten of these The University of Texas Publication species were constructed and both the homozygous and heterozygous inversions found in the different stocks were recorded. 2. In only three of the species was the X chromosome found to carry any heterozygous rearrangements. One of these three species (D. unipunctata) pre­sented two overlapping, possibly pericentric, inversions; most of the strains examined carried the inversions in the homozygous condition. Two paracentric inversions, one of which was extremely complex, were found in the second species, while the third presented only one simple inversion occurring in only one of the stocks. 3. With the exception of four species, the second chromosome was found to be polymorphic. It was the most polymorphic chromosome of the group. Thirty­three inversions were recorded from the second chromosome, three of which were extremely complex. One fusion between the second and fifth chromosome was observed in one of the species ( D. unipunctata) and probably, in one other case, addition of heterochromatin to the second chromosome has taken place. The distribution of the arrangements found in the second chromosome, as well as their possible relationships, were presented. 4. Four of the species carried polymorphic third chromosomes. Twelve re­arrangements were recorded, one of which was a complex inversion. One of the rearrangements was found to exist as heterozygous in two closely related species. The arrangements, their distribution, and possible relationships were presented. One fusion of the third with the fourth chromosome was recorded (D. unipunc­tata) . The third chromosome should be considered as the one which underwent the least amount of reorganization in this species group. 5. Seventeen inversions, distributed among five of the species, were found in the fourth chromosome. Their combinations and possible phylogenetic relation­ships within each species were discussed. It is thought that a large amount of reorganization occurred during the evolution of this chromosome. One chromo­somal fusion to the third chromosome was observed (see above). 6. Only two species were found to be polymorphic for the fifth chromosome. Subsequently, four inversions were recorded. A considerable amount of reorgan­ization has probably taken place during the evolution of the chromosome. One chromosomal fusion to the second chromosome was apparent in one of the species (see above) . 7. One case of laboratory hybridization between two species was discussed; it is thought that the same hybridization also occurred in nature at Changuinola, Panama. The cytological differences of the two species were presented. 8. With the exception of mediostriata and paramediostriata, the different species examined did not appear to carry any chromosomal rearrangements which could be determined to be identical. 9. There is now cytological evidence giving ground to further investigations concerning the differentiation observed by Heed (1957) in different geographic stocks of D. crocina. 10. It was observed that the banding pattern of the salivary chromosomes of the tripunctata group species is similar to that of Drosophila melanica, a fact which could prove to be important in further investigations. 11. It can be concluded that the species of the tripunctata group which were used in this investigation are highly isolated from each other and that they are polymorphic in their chromosome structure. The second chromosome is the most polymorphic chromosome of the group, but in some cases not the only one. From the available data no interdependence of arrangements in different chromosome arms can be observed. ACKNOWLEDGMENTS The writer wishes to express his deepest gratitude to Dr. W. S. Stone for his supervision and support during the course of this study. Much appreciation is also due to Dr. M. R. Wheeler, Dr. B. H. Judd, and Dr. L. J. Reed for serving as the supervisory committee during the graduate studies of the author; to Dr. Sarah Pipkin for useful suggestions and for providing some of the stocks from Panama; and to the Genetics Foundation group of The University of Texas for all their kindnesses. This work was supported in part by PHS research grants (GM-06492 and GM-11609) and training grant (2G-337) from the National Institutes of Health, Public Health Service. BIBLIOGRAPHY Bauer, H., M. Demerec, and B. P. Kaufmann, 1938. X-ray induced chromosomal alterations in Drosophila melanogaster. Genetics, 23: 610-630. Carson, L. H. 1955. The genetic characteristics of marginal populations of Drosophila. Cold Spring Harbor Symposia on Quantitative Biology, XX: 276-287. ----. 1959. Genetic conditions which promote or retard the formation of species. Cold Spring Harbor Symposia on Quantitative Biology, XXIV: 87-105. ----. 1961. Relative fitness' of genetically open and closed experimental populations of Drosophila robusta. Genetics, 46: 553-567. ----,and H. D. Stalker. 1947. Gene arrangements in natural populations of Drosophila robusta Sturtevant. Evolution, 1: 113-133. Cunha, A. B. Da, H . Burla, and Th. Dobzhansky. 1950. Adaptive chromosomal polymorphism in Drosophila willistoni. Evolution, 4: 212-235. Dobzhansky, Th. 1951. Genetics and the origin of species. 3rd edition, revised. Columbia Univ. Press, New York. ----, and A. H. Sturtevant. 1938. Inversions in the chromosome of Drosophila pseudo­obscura. Genetics, 23: 28-64. ----, and A. Dreyfus. 1943. Chromosomal aberrations in Brazilian Drosophila ananas­sae. Proc. Natl. Acad. Sci., 29: 301-305. ----, and C. Epling. 1944. Contributions to the genetics, taxonomy, and ecology of Drosophila pseudoobscura and its relatives. Carne. Inst. Puhl., Washington, D.C., 554: 1-183. Freire-Maia, N., and C. Pavan. 1950. lntroducao ao estudo da Drosophila. Cultus, 1949, 5: 5-70. Frota-Pessoa, 0. 1954. Revision of the tripunctata group of Drosophila with description of fifteen news species (Drosophilidae, Diptera). Arquiv. do museu paranaense Curitiba, 10: 253-330. Heed, W. B. 1957. Intraspecific relationships of Drosophila crocina Patterson and Mainland from three localities. Univ. Texas Puhl., 5721: 15-16. ----, and M. R. Wheeler. 1957. Thirteen new species in the genus Drosophila from the neotropical region. Univ. Texas Pub., 5721: 17-38. The University of Texas Publication Helfer, R. G. 1941. A comparison of X-ray induced and naturally occurring chromosomal varia­tions in Drosophila pseudoobscura. Genetics, 26: 1-22. Hsu, T. C. 1949. The external genital apparatus of male Drosophilidae in relation to sys­tematics. Univ. Texas Puhl., 4920: 80-142. Kaufmann, B. P. 1936. A terminal inversion in Drosophila ananassae. Proc. Natl. Acad. Sci., 22: 591-594. Kikkawa, H. 1938. Studies on the genetics and cytology of Drosophila ananassae. Genetica, 20: 458-516. Miller, D. D. 1939. Structure and variation of the chromosomes in Drosophila algonquin. Genetics, 24: 699-708. Patterson, J. T. 1957. A study of interspecific hybridization between members of the tripunctata group of Drosophila. Univ. Texas Pub!., 5721: 7-14. -----, W. S. Stone, and A. B. Griffen. 1940b. Evolution of the virilis group of Dro­s:iphila. Univ. Texas Pub!., 4032: 218-250. -----,and G. B. Mainland, in Patterson. 1943. The Drosophilidae of the Southwest. Univ. Texas Pub!., 4313: 7-214. -----,and G. B. Mainland. 1944. The Drosophilidae of Mexico. Univ. Texas Pub!., 4445: 9-101. Pavan, C. 1946. Chromosomal variation in Drosophila nebulosa. Genetics, 31: 546-557. Pipkin, S., and W. B. Heed. 1964. Nine new members of the Drosophila tripunctata species group (Diptera: Drosophilidae). Pacific Insects, vol. 6, no. 2: 256-273. Salzano, F. M. 1963. Chromosome inversion in two species of "Drosophila." Rev. Brasil. Biol., 23(2): 141-145. Stalker, H. D. 1960. Chromosomal polymorphism in Drosophila paramelanica Patterson. Ge­ netics, 45: 95-114. ----_,~/-. 1964a. Chromosomal polymorphism in Drosophila euronotus. Genetics, 49: 669-687. -----. 1964b. The salivary gland chromosomes of Drosophila nigromelanica. Genetics, 49: 883-893. Stone, W . S. 1955. Genetic and chromosomal variability in Drosophila. Cold Spring Harbor Symposia on Quantitative Biology, XX: 256-270. -----. 1962. The dominance of natural selection and the reality of superspecies (species groups) in the evolution of Drosophi1a. Univ. Texas Pub!., 6205: 506-537. Sturtevant, A. H. 1942. The classification of the genus Drosophila, with descriptions of nine new species. Univ. Texas Pub!., 4213: 5-51. Wasserman, M . 1954. Cytological studies of the repleta group. Univ. Texas Pub!., 5422: 130­ 152. Ward, C. L. 1952. Chromosome variation in Drosophila melanica. Univ. Texas Puhl., 5204: 137-157. Warters, M. 1944. Chromosomal aberrations in wild populations of Drosophila. Univ. Texas Pub!., 4445: 129-174. Wharton, L. T. 1942. Analysis of the repleta group of Drosophila. Univ. Texas Pub!., 4228: 23-53. -----. 1943. Analysis of the metaphase and salivary chromosome morphology within the genus Drosophila. Univ. Texas Pub!., 4313: 282-319. Wheeler, M. R. 1949. Taxonomic studies on the Drosophilidae. Univ. Texas Pub!., 4920: 156­ 195. ' XV. A Study of the Royal Jelly Gland Cells of the Honey Bee as Revealed by Electron Microscopy1 2 THEOPHILUS S. PAINTER AND JOHN J. BIESELE For a short time in the life of the adult honey bee, the hypopharyngeal glands in the head secrete the royal jelly, which constitutes, along with honey, the food of queen bees during their development and their adult life. These acinous glands are unique in the great variety of substances they secrete in the royal jelly. They have proved to be especially favorable for a study of fine structure, for it is possi­ ble to obtain all stages of cellular differentiation and to follow the morphological changes made in preparation for and accompanying secretion as well as those associated with cessation of this activity. Over the years there have been many analyses of the chemical composition of the royal jelly and, as better methods have been developed for identifying com­ plex organic compounds, more and more separate components have been found. A recent analysis by Rembold (1964) gives the following composition: water, although somewhat variable, averages about 60 percent and dry matter about 40 percent. The latter consists of various lipids up to 10 percent, dialyzable material including 8 vitamins at about 52 percent, and nondialyzable proteins at about 38 percent. The complexity of the royal jelly is not surprising, for there must be available to the queen all the materials needed for her growth and for the forma­ tion of the large number of eggs laid each day under normal summer conditions. The embryological development of the royal jelly glands as well as the mor­phological changes in secreting cells during the adult life of a worker were de­scribed by Painter ( 1945) from light microscope studies. A brief summary of pertinent facts will be useful for understanding the electron micrographs to be presented in the following pages. Figure 1 is a semidiagrammatic drawing of a gland cell in a state of active secretion, as seen with a light microscope. It will be noted that the cytoplasm is penetrated by an intracellular duct, which runs a tortuous course around the nucleus and passes through secretion reservoirs. The appearance of the nucleus varies greatly; there may be dozens of nucleoli or there may be none. When nucleoli are present one sees many Feulgen-positive granules attached to their surfaces, just as Painter and Taylor (1942) reported for oocytes of the toad. If methyl green-pyronin is used to stain the cells, both the nucleoli and the cyto­plasm stain strongly for RNA. Digestion of such cells with ribonuclease removes the pyronin-staining RNA. At the start of gland cell differentiation in an early pupal stage, the presump­tive gland cell rests on a basal cell, which is at least tetraploid. The basal cell fuses 1 This investigation was supported by USPHS research grant GM-06492 from the National Institute of General Medical Sciences and by USPHS research grant HD-01016 from the National Institute of Child Health and Human Development. 2 Recipient of USPHS Research Career Award K6-CA-18,366 from the National Cancer Institute. The University of Texas Publication ER tubules Fws. 1 and Z. Semi-diagrammatic drawings showing (Fig. 1) the gross morphological struc­tures visible in a light microscope and (Fig. Z) the fine structure and the relations of cell organel­les as revealed in electron photographs. The nucleus is divided into three parts representing dif­ferent phases of an endomitotic division cycle as indicated by the labels. with or is absorbed by the gland cell and there follows a series of endomitotic division cycles in the greatly enlarging gland cell. Most of the gland cell growth takes place after the imago emerges from the comb. During this period the young adult feeds very heavily on bee-bread (pollen) and honey. Active secretion of royal jelly begins about the fifth or sixth day of hive life and continues through the eleventh day or so under normal summer conditions. When the worker has completed its tour of duty as a nurse-maid, it usually stays in the hive as a wax­worker and guard for some ten days; then it becomes a forager bee collecting nectar, pollen, and water. During this period, it eats mainly honey and it eventu­ally dies at about the age of thirty-six days. The cessation of secretory activity is heralded by the disappearance of RNA from nucleoli, then by dissolution of the nucleoli, and by the elimination of royal jelly from the reservoirs. This causes some shrinkage of the gland cells, but the intense basophily of the cytoplasm persists. The purpose of the present study is to follow the above changes in the gland cells at magnifications available in electron micrographs, which should reveal the fine structure of the cell organelles involved in the whole process just out­lined (see Painter and Biesele, 1966). Aside from many details that will be of interest to those concerned with the fine structure of cells, there are two aspects of our findings of more general in­terest. Starting with a relatively simple gland cell in a very young pupa, it has been possible to determine the origin of the ducts of the endoplasmic reticulum and to follow the development of a complex ergastoplasm. What we have found in a very favorable material may presumably hold for protein-secreting gland cells in general. Equally as interesting is the information obtained bearing on the fundamental structure of chromosomes in our material, for during any endomitotic division cycle there is a tendency for the chromosomes to undergo the same general changes seen in normal mitosis. At what corresponds to the interphase of mitosis, nucleoli are seen in the gland cells but the replicating chromosomes are not easily identified. With the onset of the endomitotic prophase, nucleoli tend to disappear l,>y fragmentation and the chromosomes begin to shorten and form bundles of chromonemata, just as Painter and Reindorp (1939) showed in the nurse cells of the Drosophila ovary. MATERIALS AND METHODS Living honey bees (domesticated A pis mellifera L.) in various stages of de­velopment were taken directly from the hive. Under these conditions it is not possible to determine exactly how old a given individual is. To do so would require extensive labeling experiments. Little is to be gained by doing this, for one glance at the cytoplasm of a gland cell establishes the stage differentiation in the given case. The lateral pharyngeal glands were dissected from the severed heads in Ringer's saline solution for insect tissues and quickly fixed by one of several methods, involving primarily either osmium tetroxide or glutaraldehyde. In the first case, fixation was in Palade's 1 percent osmium tetroxide in aqueous solution, either buffered at pH 7.2 or brought to a pH of 6.0 with 0.1 N HCl. The University of Texas Publication Although the lower pH damaged mitochondria somewhat, we consider it prefer­able for fixation of chromosomes (cf. Claude, 1961) . Tissues were fixed for 2 hours at 0° to 4°C. They were then rinsed once in distilled water, dehydrated through a graded series of alcohols, and impregnated in an upgraded series of epoxy plastic and acetone mixtures. Final embedding was completed by heating to 80°C after thorough evacuation. Thin sections were cut on an ultramicrotome with diamond knife and were post-stained with either strontium permanganate or uranyl acetate (Huxley and Zubay, 1961). Fixation with glutaraldehyde took the following pattern: After dissection, the glands were immersed in 2.5 percent glutaraldehyde in 0.05 M Sorensen's buffer, pH 7.3, at room temperature for one hour (Sabatini et al., 1963). Thereafter they were post-fixed in 1% osmium tetroxide solution, likewise buffered with phos­phates at pH 7.2, for one hour. This was followed by immersion in 0.5% uranyl acetate solution for about 20 hours. After dehydration, embedding, and section­ing as above, sections were post-stained with lead citrate (Reynolds, 1963). Sections were examined in a Siemens electron microscope Elmiskop I, and images were recorded on Dupont Ortho-litho S film, Kodak Projector Slide Plates (contrast), or Kodak Contrast Process Ortho film. OBSERVATIONS Entire cross-sections of secreting gland cells are much too large to show details of fine structure (see Figure 3 for a cell showing the complex nature of the ergastoplasm); for the reader's guidance, Figure 2 is presented in order to show diagrammatically the cell organelles involved, their topographical relations to each other, and the different conditions found in nuclei in various phases of an endomitotic division cycle. A conspicuous structure in the cytoplasm in thin sections is the intracellular duct. In cross-sections its walls are seen to contain electron-dense lamellae, or longitudinal staves, which probably serve to keep this duct open. The duct passes through secretion reservoirs, which show openings into the lumens of the endo­plasmic tubules. These tubules bear very conspicuous polyribosomes on their outer surface. Some of the tubules are interconnected and there may be local en­largements, the cisternae. The size of these local enlargements, as well as the size of the secretion reservoirs, may depend on how recently the worker under con­sideration has discharged its royal jelly. Mitochondria lie everywhere but are most numerous in the vicinity of the reservoirs and the intracellular duct. We have not identified Golgi apparatus in secreting mature cells, but many dicty­osomes are found in developmental stages. Several features of the nuclei deserve attention. The nuclear envelope contains thousands of pores. In Figure 2, three typical conditions regularly found in nuclei and related to stages of endomitosis are represented. In the bottom sector, corres­ ponding to the interphase of mitosis, there are numerous nucleoli, the number of which depends on the size (degree of endopolyploidy) of a given nucleus. The nucleoli lie in a felt-like precipitate of nuclear sap, which becomes progressively denser as the cell approaches the period of synthesis and secretion. Ordinarily one cannot identify the chromosomes at this time as their fine filaments are more or less hidden in the nuclear sap. To the left in the diagram, corresponding to an early prophase, the nucleoli are undergoing fragmentation (see Figure 17). The nuclear sap is crowded with nucleolar fragments and the pores in the envelope are prominent. In many electron micrographs, one can identify tangled masses of filaments, as will be shown later. These are bundles of contracting chromonemata. To the right in the diagram, it will be noted that there are no nucleoli and that the chromosomes have become organized into bundles of chromonemata lying in parallel array (see Figs. 20 or 21 ). The nuclear pores are scarcely visible. This stage corresponds to a very late prophase or to metaphase of mitosis. Similar bundles of chromonemata appear during endomitosis in the nurse-cells of Dro­sophila (Painter and Reindorp, 1939). In a worker bee pupa about 2 days before it would have emerged as an adult there is very little endoplasmic reticulum (Figure 4). It will be noted that, aside from sections of the intracellular duct, the most conspicuous structures are the numerous mitochondria. Besides these there are vesiculated areas, labeled Y, which probably represent dictyosomes, and from time to time we have found a structure consisting of concentric lamellae, labeled X. It has been observed twice at this stage and its nature unknown. It will be noted that the endoplasmic reti­culum (ER) is very scant; the tubules are of small diameter, and polyribosomes lie free, widely scattered among the mitochondria and other cytoplasmic organel­les. In young bees taken as they were cutting their way out of the comb (Figures 5 and 6), there has been some increase in the diameter of the endoplasmic tubules, but most of the polyribosomes still lie free in the cytoplasm. Older worker bees contain transition stages that fill in the gap between Figure 4 and the fully dif­ ferentiated ergastoplasm of Figure 3. In Figure 7 the pores of the nuclear envelop are crowded with ribosomes, which emerge as polyribosomes. It will also be noted that the endoplasmic reticulum tubules are larger in diameter and that polyribo­ somes are attached in many areas. Figure 8 is of still more advanced stage in dif­ ferentiation, and the ER tubules are approaching the mature secretory stage. In this figure one finds in the lumen of the ER tubules many local enlargements con­ taining a fibrous precipitate, presumably protein, so it is assumed that the syn­ thesis of proteins has already begun. When a worker bee finishes its tour of duty as a nursemaid to larvae and the queen, it ordinarily leaves the brood chamber and for a period of some ten days acts as a wax-worker and guard for the hive. During this period, it eats less and less bee-bread and, going hand in hand with this, the secretory activity of the royal jelly gland cells wanes. This is accompanied by the loss of RNA from nucleoli and the disappearance of nucleoli from the nucleus. As Figure 11 shows, the ER tubules appear shrunken but otherwise intact. The presence of distended secretion reservoirs in this cell suggests that the individual concerned was prob­ ably doing guard duty when it was taken from the edge of a hive. Figure 12 shows the condition of the nucleus in an old bee. The chromosome bundles have condensed into rounded masses to which most of the remaining nuclear sap is attached. In the light microscope, these balls of chromosomes are seen as large Feulgen-positive granules in the nuclei of old bees. The University of Texas Publication Having given in general outline the course of events from very young pupae to old bees, we now turn to a more detailed description of certain cell organelles and to questions raised by our observations. Origin of the secretory reservoirs. The secretory reservoirs develop in the cy­toplasm around the intracellular duct. The process of their formation seems much involved, in that finger-like extensions of the ground cytoplasm run through them and connect to the wall of the intracellular duct. An early stage in reservoir formation is presented in Figure 13. It will be noted that the reservoir wall is lined externally with ribosomes and that the lumens of the ER tubules merge in turn with the reservoir. On such figures rests the conclusion that the reservoirs are formed by the merging of ER tubules. Golgi apparatus. It is also within the realm of the likely, in view of the current understanding of cellular organelles, that the point of origin of the secretory reservoir is a saccule or vacuole of the Golgi apparatus, ultimately connected to the rough-surfaced endoplasmic reticulum. Massive distension of the reservoir may lead not only to incorporation of the immediately connected portions of the ER into it, but also to the apparent obliteration of the Golgi apparatus as we conventionally recognize it. As mentioned before, more nearly typical dictyo­somes with multiple saccules or cisternae accompanied by a number of vesicles are recognizable in early pupal stages. Microtubules. These are frequently seen in the region of the cytoplasm sur­rounding the intracellular duct. We have not identified any centrioles in these gland cells and, in the absence of evidence for centriolar attachment of the micro­tubules, it may be tentatively proposed that the microtubules in these cells have something to do with a contractile mechanism for the discharge of the royal jelly. Changes in nuclear contour. Throughout this study it has been noted that the contour of the nuclear wall is notably labile, especially in early developmental stages, and we have found blunt processes, as for example in Figures 4 or 7, which we interpret as showing changes in surface tension. When Painter and Taylor (1942) made their study of the toad's egg, similar blunt processes were found to accompany the passage of nucleolar material from the nucleus to the cytopla -m. FIG. 3. Electron micrograph showing the complex nature of the endoplasmic reticulum (ER) in an actively secreting cell. It will be noted that the walls of the ER tubules show polyribosomes PR. The lumen of tubules usually shows a fibrous precipitate which is formed, presumably, by proteins in the royal jelly. M indicates a mitochondrion. Osmic acid fixation with post-staining with strontium permanganate (Os. and Str.). 50,000X. FIG. 4. Electron micrograph made of a gland cell taken from a late pupal stage. Aside from sections of the intracellular duct (D) and the secretory reservoir (SR), mitochondria (M) are the most conspicuous feature of the cytoplasm. The ER system of tubules is not well developed, but one sees tiny tubes lying about the cytoplasm with a few ribosomes attached. There are a number of vesiculate structures, labeled G, which probably represent Golgi complexes. In very young pupae there occurs a structure, labeled X, the nature of which is not known. The fine granular appearance of the cytoplasm is due, in the main, to free polyribosomes (PR) not yet associated with ER tubules. Left of center there is a fold in the nuclear envelope which shows (after uranyl acetate post-staining) the pores (P) in the nuclear envelope as round black dots. There are not intact nucleoli in this cell but there are numerous nucleolar fragments (NF). Os. and Uran. acet. 6,500X. FIGs. 5 and 6. Electron micrographs of gland cells taken from workers as they were cutting their way out of the comb. Fig. 5 shows, at a higher magnification than Fig. 4, the rudiments of the ER tubules. The nucleolus (NCL) is apparent on the left. Os. and Str. perm. 18,760X. Fig. 6 shows the prevalence of free polyribosomes (PR) at this early stage before the ER tubules are well differentiated. Os. and Str. perm. 15,500X. This raises the question of what causes these changes and what forces the nucle­olar material out of the nucleus into the cytoplasm. At present we are unable to explain the forces at work other than to point out that a great deal of material is being synthesized in the nucleus and that internal pressure might therefore be a factor. Origin of the ER tubules. In gland cells of late pupal stages and of newly emerged workers the endoplasmic tubules are very small in diameter, sparse in number, and scattered generally about the cytoplasm even to the edges of the plasma membranes of the cell. Nowhere in micrographs of such stages has there been the least hint of the way in which the tubule anlagen arise. An answer was found in still earlier pupal stages. In pupae that normally would not have emerged as adults for five or six days it was found that the outer membrane of the nuclear envelope showed numerous outpocketings, or blebs, some of which were elab­orately folded (Figures 14, 15 and 16). These blebs in the outer membrane are cut off (Figure 16) and form the anlagen of the ER tubular system. At the earlier developmental stage we have examined, there are already many such anlagen of tubules in the cytoplasm (Figure 14 and 16) from which it is evident that the blebbing process must have begun at earlier stages of development. Evidence of blebbing diminishes in older pupae, and by the time of emergence of the imago, no blebbing is evident in the micrographs. The abundance of ER tubule anlagen near the outer plasma membrane of gland cells raises the question of whether some of the anlagen might have arisen from the cell membrane. The possibility of such an origin certainly lies open. It would be in line with the observation made over a decade ago by Palade ( 1955) that invaginations of the plasma membrane may contribute to and be continuous with smooth-surfaced endoplasmic reticulum, which in turn may be continuous with the protein-synthesizing rough-surfaced endoplasmic reticulum. Our obser­ vations mentioned above, however, suggest that in the case of these gland cells of the bee the chief derivation of the rough-surfaced endoplasmic reticulum is from the outer nuclear membrane. It has long been evident that the perinuclear cavity of the nuclear envelope is one of the largest cisternae of the endoplasmic reticulum (Watson, 1955), and in mitotic cells the virtual interchangeability of pieces of the nuclear envelope and elements of the endoplasmic reticulum was recog­ nized by Porter and Machado ( 1960) . Structure and behavior of nucleoli. Studies of nucleoli with the light micro­scope have shown that in presumed diploid cells of worker bees there are at most 8 nucleoli; hence Painter (1945) concluded that there were 8 organizers in the diploid genome of the bee. In actively secreting cells, Painter counted up to 80 separate nucleoli in a single nucleus. Their variation in size suggested a certain amount of coalescence. The number of nucleolar organizers may therefore have been the next higher power of 2, or 128. Because endomitosis provides no mecha­nism for separating the replicates of a single chromosome, it is frequently ob­served that nucleoli occur in aggregates of a considerable number of separate ele­ments. In electron micrographs, the fine structure of nucleoli is markedly affected by the fixative used initially as well as by post-fixation treatment. It should be re­ The University of Texas Publication membered that Vincent ( 1955) showed that in isolated starfish nucleoli the pro­tein content runs as high as 90 percent of the total weight. Numerous recent studies have shown that ribonucleoproteins constitute only about 33 percent of the proteins, the rest being other types of proteins synthesized in the nucleus. As we have pointed out, with osmic acid fixation at either pH 7.2 or pH 6.0, followed by post-staining with strontium permanganate, the non-ribonucleopro­teins form a relatively less dense felt-like precipitate, and if a ribosome is present in a nucleolar unit (often referred to as granule by other workers), it can be seen. If glutaraldehyde or other aldehyde is used in the initial fixation, the vaguely fibrous or felt-like protein is denser to electrons and ribosomes, if present, are not so readily distinguished. Figure 17 which was made from a very young pupa, as may be judged from the lack of a prominent ER system, shows two typical conditions. In the upper part of this figure typical nucleoli are shown and in the lower part of the nucleus, nucleolar fragments are much in evidence. Intact nucleoli show more or less con­tinuous denser bands, sometimes termed nucleolonemata (Estable and Sotelo, 1951), around the periphery and a more diffuse pars amorpha in the center. In some micrographs these denser areas are not apparent; we think Figure 15 indi­cates the reason. In this figure it is readily seen that the nucleolus is made pre­dominantly of units about 150A-200A in diameter in which dense granules of slightly smaller diameter may be embedded. Not all of the units show these granules, which may be interpreted as ribosomes, possibly still nascent. In a few units we have seen suggestions of the fusion of two granules of different sizes into a single ribosome. In Figure 18 one still sees evidence of a worm-like coiling of the nucleolar units, which suggests that what has been interpreted as a separate component, the nucleolonema, may be an expression of close packing of the units. That the felt-like or vaguely fibrous mass covering the ribosomal granule (if present) in the nucleolar unit consists of protein was shown by Schoefl (1964), for it is removed by pepsin digestion. During the fragmentation of nucleoli small masses of units, or perhaps even single units, break free from the general mass and, as these pass toward or through the nuclear envelope, the fibrous component appears to separate from the ribo­somes; as a result the nuclear sap in secreting cells is quite dense in felt-like material. Figure 7, from material fixed with 1 percent osmium tetroxide, shows that ribosomes are present in the nuclear pores and that they appear to emerge as polyribosomes into the cytoplasm. Figure 8, also after osmium tetroxide fixation, shows that felt-like proteins are emerging from the pores with ribosomes some­what obscured among them. FIGs. 7 and 8. Electron micrographs showing a progressive differentiation of the ER system. In Fig. 7 the nucleus (above) shows many nucleolar fragments (NF ). The nuclear envelope is cut in such a way as to show, unusually well, nuclear pores containing polyribosomes (PR and un­labeled arrows). A fold in the nuclear envelope, on the left, shows many free polyribosomes as well as some which have taken their places on the ER tubules. Microtubules (MT) show well. Os. and Str. perm. 29,400X. Fig. 8 shows the presence of a fibrous precipitate in the cisternae of the ER system, which suggests that protein synthesis has already set in. Material seems to be passing through the nuclear pores (arrows). Os. and Str. perm. 39,100X. The University of Texas Publication Within the past five years the fine structure of nucleoli has received a great deal of attention. From the many publications we have selected the work of three authors, for a consideration of their observations linked with ours points clearly to a better understanding of nucleolar fine structure. The point we wish to make is that in differentiated cells the structure of the nucleolus reflects the needs of that type of cell. Most recent workers agree that the nucleolus is made up of "granules" (we prefer the term, nucleolar units) some 150 to 200A in diameter, some but not all of which carry ribosomes (Schoefl, 1964; Marinozzi, 1964; Stevens, 1965). Among nucleolar constituents, Schoefl ( 1964) listed RNP parti­cles (ribosomes) embedded in a matrix of fibrous proteins, denser areas termed the nucleolonemata, and amorphous material normally contained in depressions in the nucleolonemal masses. When kidney cells of the baboon were subjected to pepsin digestion, the protein matrix of the nucleolar units was dissolved away and the RNP particles stood out with great clarity (see Schoefl' s Figure 2a) . When cells were subjected to both pepsin and ribonuclease digestion, the units disap­peared but the nucleolonema still showed as dense masses (see Schoefl's Figure 2b). When baboon kidney cells were treated with actinomycin D, there was a tendency for the several nucleolar components to sort out (see Schoefl's Figure 4). Marinozzi ( 1964) reported nucleolar units, fibrous masses, and a protein matrix in which the fibrillar masses were embedded. No nucleolonema was seen in pancreatic cells of the rat. Stevens (1965) likewise found granules about 150A in diameter in grasshopper neuroblasts. The nucleolar material was more or less segregated into two areas: the ribosome-carrying units along the edges and the amorphous protein material in the center. In the hypopharyngeal gland cell of the bee, part of the nucleolus shows units carrying ribosomes and part shows nucleolar units apparently lacking them. There is no sharp separation between these structural elements, as the authors cited above found. How can our observations be reconciled with the excellent electron micrographs upon which the three authors cited have based their con­clusions? · First of all, although some cells of the bee gland show dense worm-like areas usually called nucleolonemata, other cells do not show such a nucleolar differ­entiation. We think that our Figure 18 suggests a closer packing of the nucleolar units in some regions and that this is part of the explanation when nucleolone­mata are found. In the second place, we think that the fine structure of the nucleolus may well differ from one differentiated cell to another. After all, much protein appears to be synthesized in the nucleus, but the need for such protein in a baboon's kidney F10. 9. Electron micrograph made from the gland cell of a worker in the process of emerging from the comb. Here the intricate pattern of lamellae of the intracellular duct (D) may be seen. The nucleus shows only nucleolar fragments (NF ). The chromosomes (CHR) show as a tangle of fine threads. This represents a late prophase stage of the endomitotic cycle. Os. and Str. perm. 16,400X. Fw. 10. Electron micrograph taken from a bee somewhat older than conditions shown in Fig. 9. It will be noted that dense material seems to be passing out through the nuclear pores (P). There are no nucleoli in this cell and the condensation of chromosome bundles (CHR) is well advanced. Os. and Str. perm. 39,000X. The University of Texas Publication cell is probably quite different from the demands made by neuroblasts of the grasshopper or the royal jelly gland cells of the bee. Nuclear pores. Although it is hazardous to generalize about the morphology of any cytoplasmic or nuclear structure from a study of a single type of cell, we think that the transitory appearance of the pores in the royal jelly glands of the bee may well explain some of the conflicting views in the literature about the details of pore structure. A fence line with all the gates closed looks quite different from one in which the gates are open with traffic passing through. The traffic in the bee gland consists of ribosome associations, or polyribosomes, and small masses of felt-like protein (see Figures 8, and 10). Of special interest is the central granule interpreted as a ribosome seen in the pores of Figure 7. Many observers have reported the presence of a small granule (Afzelius, 1955; Watson, 1959) in the center of nuclear pores (see arrows in Figure 7); some maintain that there is a thin diaphragm across the pore (Watson, 1955; Merriam, 1961; Fawcett, 1966), as well as other diffuse and indistinct material within the pore channel (Watson, 1959; Merriam, 1961 ). We have found this central gran­ule regularly in cells fixed initially with osmium tetroxide (it is not so readily seen after fixation with glutaraldehyde). The other vague material within the pore, which is better seen with glutaraldehyde fixation, may well be proteins in transit to the cytoplasm, as Watson ( 1955) early suggested. Merriam ( 1961) has contended, however, that the diffuse material in the neighborhood of the problematical diaphragm is a structural part of the nuclear envelope, because it remains with the rest of the fixed envelope when the latter is dissected off of the nuclei of frog eggs. In the bee the central granule in the pore appears to be one end of a ribosome complex (polyribosome) , which, as we interpret our micrographs, is either or­ganized within the pore or stripped of an obscuring protein coat within the pore. The felt-like proteins in the nuclear sap make it difficult to determine whether or not ribosomes may first make their attachment to messenger RNA in the nuclear sap and examination of many micrographs has revealed no firm evidence one way or the other. At any event, when ribosomes emerge from the nuclear pores, they are usually (always?) connected in strings of 12 or 15 or so, by mes­senger RNA. In very early stages they lie free in the cytoplasm of the hypo­pharyngeal gland cell. When ribosomes and other material are not passing through the nuclear pores, as we interpret the situation, the nuclear pores are not obvious. Watson's classical paper (1955) suggested that Palade's ribonucleoprotein FIGs. 11 and 12. Electron micrographs made from bees that had completed their nursz-maid duties. In Fig. 11 the ER tubules appear shrunken but there is still royal jelly in the secretory reservoirs (SR). All nucleoli, except one, have disappeared from the nucleus. Os. and Str. perm. 6,700X. Fig. 12 shows a somewhat later stage in the condensation of chromatid bundles (CHR). The single nucleolus (lower left-hand corner) does not show typical structures. Os. and Str. perm. 7,000X. Fig. 13 is presented to show an early stage in the differentiation of secretory reservoirs (SR). Ribosomes may be seen lying along the edge of the reservoir (left-hand arrow). At the lower right the merging of a tubule with the reservoir is seen (right-hand arrow) . Os. and Str. perm. 36,600X. The University of Texas Publication Painter and Biesele: Royal Jelly Gland Cells of the Bee particles, manufactured in the nucleus, were transported through the nuclear pores into the cytoplasm. Gall (1956) later suggested that the subunits of the annulus, Watson's (1959) "cytoplasmic cuff" surrounding the pore on the outer nuclear membrane, might be ribonucleoprotein particles, and Watson (1959) ;.pointed out that the similarity in diameters of the ribosomal spirals and crescents {polyribosomes) in the cytoplasm and the nuclear pores suggested an association ·between them. We are of the opinion that this association does exist, in spite of ~Watson's (1959) contention that the central granule of the pore does not stain with lead hydroxide as heavily as do cytoplasmic ribonucleoprotein particles, and in spite of Merriam's (1961) finding that ribonuclease affects no element of the pore complex in frog oocytes, although trypsin digestion does eliminate both the ' cytoplasmic cuff of the pore annulus and the diffuse material of the pore dia­phragm. ·· Chromosome structure and behavior. The contraction or coiling of chromo­nemata and their uncoiling, which occur during each endomitotic division cycle, give opportunities for observing the fine structure of chromosomes. There is, how­. ever, one pitfall that must be guarded against. After fixation and post-staining, .th~ nuclear sap, the secretion reservoirs, and the cisternae of the ER tubules all show many branching fibrillae, often with apparent knots of material. Such .. ~tructures doubtless represent precipitated proteins. For this reason, it is not easy . to identify the chromosomes in interphase stages of endomitosis, when nucleoli -~e prominent. After the nucleoli have fragmented, however, masses of very fine fibrillae are observed (Figures 9 and 19), and the paired chromomeric structure of the chromosomes is clearly evident (see arrow, Figure 19). At a somewhat .later stage, bundles of chromonemata appear (Figure 20). ,. -Figures 21to23 show the nature of these chromonema bundles at high magni­fication. It will be noted that the bundles are made up of chromomeric threads in . parallel array. Perhaps the best time to observe their structure is at the end of the secretory . cycle, when the chromosomes begin to contract into ball-like masses (Figure 12) ; two such "old" nuclei are shown in Figures 24 and 25. In Figure 24 the chromo­meric nature of the strands is easily seen. In Figure 25, which is of a somewhat earlier stage, the chromosomes are reminiscent of lampbrush configurations. DISCUSSION The observations recorded above have a bearing on a number of interesting biological problems; in the following pages we shall post a number of questions and attempt to answer some of them. How is it possible for the queen bee to lay so many eggs in a 24-hour period? A queen bee lays about 1200 eggs a day under ordinary summer conditions and a much higher number in some exceptional circumstances. The ovary of the bee is structured much like those of many other insects. It is made up of a number of Fms. 14, 15, and 16. Electron micrographs showing numerous blebs (see arrows) formed by the outer membrane of the nuclear envelope. Glutaraldehyde. Figs. 14 and 15 represent ZB,OOOX and Fig. 16 is 42,200X. The University of Texas Publication :·SR .,,,, .... ,•.. "?*-. ~,:' ­ .. ® ovarioles, or egg strings, each of which contains a series of egg follicles in various stages of egg maturation. The egg follicle is encased in a single layer of follicular cells and there is a number of nurse cells attached to the oocyte. The food of the queen bee consists of royal jelly and honey. The royal jelly contains proteins, lipids, vitamins, and, as will appear below, almost certainly DNA precursors. It is clear that passage of some of these materials undigested and intact into the eggs would be a more efficient process than their digestion by the queen and then their resynthesis by the ovum or other cells of the ovary into large macromolecules. When one of us (Painter and Reindorp, 1939) described the process of endo­mitosis in nurse cells of Drosophila melanogaster, it was assumed that the nurse cells passed on to the oocyte proteins synthesized by them and that the ingestion of nurse cell nuclei supplied the "building blocks" for the assembling of new cleavage chromosomes. Because the egg nucleus is very small during oogenesis, it was concluded that it played a minor role in the formation of the yolk. Since this early period and especially in the last decade, a great deal of experi­mental work has been carried out and, in a recent excellent review, Telfer ( 1965) presented the evidence for the following general conclusions. Most of the yolk of the insect egg is extraovarian in origin. There is much good evidence that, what­ever the site and the mode of ingress may be, some proteins of the blood of insects pass substantially unchanged into oocyte cytoplasm during vitellogenesis. Fur­thermore, while almost any protein injected into the blood of an insect may be recovered in the cytoplasm of the eggs, in general, the process is a highly selective one. In.some cases, there may be twenty times as much of the protein in the egg as in the blood stream, but usually the amount recovered from the egg is very small. While extraovarian sources contribute heavily to the yolk supply of the egg, the components of the egg follicle make contributions also. As has long been known, one function of the follicular sheath is the formation of the chitinous shell that encloses the fully formed egg. In light of the fact that in early stages of egg formation, labeled amino acids are quickly taken up by the nurse cells, the follicular cells, and the ovum itself, all of these different cells must carry on some protein synthesis. Work done in our laboratory by Painter and his students has shown that, even within the order Diptera, there are marked histological differ­ences in the mosquito oocyte as compared to that of the fruit fly. It seems reason­able to conclude that the relative importance of the synthetic activity of nurse cells, follicular cells, and the egg itself may well vary in related taxonomic groups. It is now well established that in Drosophila melanogaster, for example, the cytoplasm has a reservoir of DNA (Schultz, 1956). Where does this DNA come from? Painter concluded years ago that cytoplasmic DNA was derived from in- Fm. 17 . . Electron micrograph made from a late pupal stage. Above, it shows more or less intact nucleoli (NCL) while in the lower part of the nucleus large numbers of nucleolar frag­ments (NF) are present. Note that the cytoplasm contains many free-lying polyribosomes. Glutaraldehyde. 16,BOOX. Fm. 18. Electron micrograph showing details of nucleolar structure. Some of the nucleolar units contain ribosomes (see arrow). Os. and Str. perm. 73,200X. The University of Texas Publication Painter and Biesele: Royal Jelly Gland Cells of the Bee gested nurse cell nuclei. This still is the most probable source in the cytoplasm of the fruit fly egg, but in other forms the origin of the DNA is not so obvious. Recently, Nigon and Gillot ( 1964) attempted to gain insight into the problem by injecting tritiated thymidine into the blood of Drosophila. Radioactivity was found in egg cytoplasm, but cytoplasmic DNA was not removable with deoxy­ribonuclease or by means of other procedures that ordinarily remove DNA from tissues. There is a large amount of evidence that the nurse cells synthesize a great deal of ribonucleoprotein that is passed on to the cytoplasm of the egg. Electron micro­graphs have shown that the cytoplasm of nurse cells has an unusually high con­centration of RNP particles (to judge from our bee gland work, it seems reason­ably certain that these are polysome aggregates), which are passed into the cyto­plasm of the egg. Therefore, it may well be that a primary role of nurse cells is to synthesize ribosomes and polyribosomes needed as the egg begins embryonic de­velopment. Nurse cells undergo endomitosis; while this point has not been veri­fied, it seems probable that the fragmentation of nucleoli in early prophase stages of an ~ndomitotic cycle releases ribosomes and felt-like protein into the nuclear sap arid that these eventually pass through nuclear pores into the cytoplasm, just as inthe royal jelly gland cells described above. · In the light .of the experimental evidence cited above, we may now explain the extremely rapid formation of yolk in the queen bee's eggs. The royal jelly is ·extremely rich in a variety of proteins, lipids, and vitamins, as has been noted. Apparently these materials are not broken down but pass substantially unchanged through the intestinal wall, enter the blood stream and, from the hemolymph, pass again unchanged, presumably by pinocytosis, directly into the cytoplasm of the oocyte. It must be remembered that there is need for precursors of DNA in order that endomitosis can go on in nurse cells and eventually supply deoxyribo-1 nucleotides to the egg cytoplasm. Again, as in the case of proteins, the need for efficiency and the limitations of time dictate that DNA synthesis be shortened by niaking use of salvage pathways and starting with relatively complex precursors closer in structure to the eventual product. This consideration led Painter to give some royal jelly to Dr. E. M. Lansford, .of the Clayton Foundation Biochemical Institute, The University of Texas, who very kindly carried out a bioassay with a thymineless strain of the lactic acid bacillus. As was to be expected, no free thymine or thymidine was found, but when the royal jelly was subjected to a mild alkaline hydrolysis for one hour measurable quantities of thymidine were set free. Presumably this would hold true for the other nitrogenous bases of DNA. From these preliminary assays it was concluded that DNA precursors were present but were bound up or hidden by other compounds; the recent work of Nigon and Gillot (1964) points in this FIG. 19. Electron micrograph showing above, in the nucleus, condensing chromosomes. To the left (see arrow), paired chromomeres are visible for a short distance. The protein precipitate in cistemae may be seen (right-hand arrow) . Os. and Str. perm. 54,000X. · FIG. 20. Electron micrograph showing three chromatid bundles (CUR and unlabeled arrows) one of which appears in cross-section. Os. and Str. perm. 54,000X. The University of Texas Publication same direction. It should be recalled that the eggwhite of hens, amphibia, and per­haps other organisms contains avidin, a protein that may be complexed with DNA. As has been pointed out, when the young worker bee emerges from pupation, the cells of the royal jelly gland show little differentiation of cell organelles, such as the endoplasmic reticular system of tubules. Only after five or six days of hive life is an active secretory condition reached. What is it that controls this dif­ferentiation process? And, at the end of the nurse maid tour of duty, what is it that stops the activity of these glands? The newly emerged worker eats large quantities of bee-bread (stored pollen) and honey; it is a nearly foregone con­clusion that intake of the proper food in appropriate quantities plays a significant role. For many years beekeepers have sought a substitute for stored pollen in order to maintain the strength of a colony in adverse seasons. At the time Painter was working on the bee problem in 1944, there had just appeared an analysis of the substances present in bee-bread and, with the help of Dr. Roger J. Williams, three diets were made up. Diet I consisted of vitamin-free casein, all the vitamins known to occur in bee-bread (lipoic acid has since been discovered), and some crude DNA. Diet II consisted of casein and vitamins. Diet III was casein plus DNA. Young worker bees were taken as they were emerging from the comb and placed in a container with the selected diet and invert sugar. With the "complete" diet I there was a beginning of some differentiation, including the synthesis of some.RNA in gland cells, but it did not go far. With casein and vitamins alone, there was a little RNA formed but no significant change. With casein and DNA there was no change from the initial condition. While these unpublished experi­ments led to no firm conclusions as to the relative importance of the several in­gredients, they did make it clear that the embryonic gland cells would not dif­ferentiate without an extrinsic source of the proper food. It is fair to conclude, then, that food is an important factor for differentiation. What causes the cessation of synthetic activity of the royal jelly glands? Here again, as has been noted, when the worker leaves the brood chamber and assumes duties of working wax and guarding the hive, it eats less and less bee-bread and, when it becomes a forager, its food is mainly honey. The conclusion that glands cease to function when suitable precursors are not available in their nutrient is readily drawn. It is probably oversimplified, however, because the total situation involves not only behavioral changes, including those in feeding habits, but also a temporal programming of development and the senescence that precedes termi­nation. Another question of interest is, when does endomitosis cease in the differen­tiated gland cells? A final answer to this question would involve the careful mark- FIGs. 21, 22, and 23. Electron micrographs showing chromatid bundles (CHR) at a high magnification. In both Figs. 21 and 22 chromomeres are easily seen. Os. and Str. perm. 54,500X. FIGs. 24 and 25. Electron micrographs taken from older bees. Again the chromomeric nature of the chromatids may be seen (arrow). Fig. 25 is a somewhat earlier stage and the chromosomes are seen to resemble lampbrush chromosomes (arrow) . Os. and Str. perm. 54,500X. The University of Texas Publication ing of workers as they emerged from the comb and an electron microscopic study of gland cell nuclei in old workers of precisely known age. Our general impres­sion is that endomitosis, with the accompanying breakdown of nucleoli releasing polysomes to the cytoplasm, continues in any cell that is still synthesizing royal jelly. ACKNOWLEDGMENTS The authors are indebted to Dr. William R. Lee for bees, to Mr. Robert W. Riess for electron microscopy, and to Mr. John Alvermann for photographic work. LITERATURE CITED Afzelius, B. A. 1955. The ultrastructure of the nuclear membrane of the sea urchin oocyte as studied with the electron microscope. Exptl. Cell Research, 8: 147-158. Claude, A. 1961. Problems of fixation for electron microscopy. Results of fixation with osmium tetroxide in acid and alkaline media. Pathologie et Biologie 9: 933-947. Estable, C., and J. R. Sotelo. 1951. Una nueva estructura celular: El nucleolonema. Instituto de Investigaciones de Ciencias Biologicas, Publicaciones 1: 47-68. Fawcett, D. W. 1966. An Atlas of Fine Structure, The Cell, Its Organelles and Inclusions. W. B. Saunders Company, Philadelphia and London. 448 pp. Gall, J. G. 1956. Small granules in the amphibian oocyte nucleus and their relationship to RNA. J. Biophysic. and Biochem. Cytol., 2(4, suppl.): 393-396. Huxley, H. E., and G. Zubay. 1961. Preferential staining of nucleic acid-containing structures for electron microscopy. J. Biophysic. and Biochem. Cytol., 11: 273-295. Marinozzi, V. 1964. Cytochimie ultrastructurale du nucleole-RNA et proteines intranucleo­laires. J. Ultrastruct. Research 10: 433-456. Merriam, R. W. 1961. On the fine structure and composition of the nuclear envelope. J. Biophysic. and Biochem. Cytol., 11 : 559-570. Nigon, V., and S. Gillot. 1964. L'incorporation de la thymidine au cours de l'ovogenese et du development embryonnaire chez la drosophile. Exptl. Cell Research 33: 29-35. Painter, T. S. 1945. Nuclear phenomena Hssociated with secretion in certain gland cells with especial reference to the origin of cytoplasmic nucleic acid. J. Exptl. Zool., 100: 523-547. Painter, T. S., and J. J. Biesele. 1966. The fine structure of the hypopharyngeal gland cell of the honey bee during development and secretion. Proc. Nat. Acad. Sci. 55: 1414-1419. Painter, T. S., and E. C. Reindorp. 1939. Endomitosis in the nurse cells of the ovary of Dro­sophila melanogaster. Chromosoma 1: 276-283. Painter, T. S., and A. N. Taylor. 1942. Nucleic acid storage in the toad's egg. Proc. Nat. Acad. Sci., 28: 311-317. Palade, G. E. 1955. Studies on the endoplasmic reticulum. II. Simple disposition in cells in situ. J. Biophysic. and Biochem. Cytol., 1: 567-582. Porter, K. R., and R. D. Machado. 1960. Studies on the endoplasmic reticulum. IV. Its form and distribution during mitosis in cells of onion root tip. J. Biophysic. and Biochem. Cytol., 7: 167-180. Rembold, H. 1964. Die Kastenentstehung bei der Honigbiene, Apis mellifica L. Die Natur­wissenschaften 51 (3): 49-54. Reynolds, E. S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol., 17: 208-212. Painter and Biesele: Royal Jelly Gland Cells of the Bee Sabatini, D. D., K. Bensch, and R. J. Barrnett. 1963. Cytochemistry and electron microscopy: the preservation of cellular ultrastructure and enzyme activity by aldehyde fixation. J. Cell Biol., 17: 19-57. Schoefl, G. I. 1964. The effects of actinomycin on the fine structure of the nucleolus. J. Ultra­struct. Research 10: 224--243. Schultz, J. 1956. The relation of the heterochromatic chromosome regions to the nucleic acids of the cell. Cold Spring Harbor Symp. Quant. Biol., 21: 307-325. Stevens, B. J. 1965. The fine structure of the nucleolus during mitosis in the grasshopper neuroblast cell. J. Cell Biol., 24: 349-368. Telfer, W . H. 1965. The mechanism and control of yolk formation. Ann. Rev. Entom:il., 10: 161-184. Vincent, W. S. 1955. Structure and chemistry of nucle:ili. Internat. Rev. Cytol., 4: 269-298. Watson, M. L. 1955. The nuclear envelope. Its structure and relation to cytoplasmic mem­branes. J. Biophysic and Biochem. Cytol., 1: 257-270. Watson, M. L. 1959. Further observations on the nuclear envelope of the animal cell. J. Biophysic. and Biochem. Cytol., 6: 147-156. XVI. Sulfoxide Protection of Bacteriophage Against X-irradiation lnactivation1 LLOYDS. LOCKINGEN INTRODUCTION Ashwood-Smith ( 1961) tested several sulfoxides for radioprotective action against a lethal dose of X-rays in mice and found that dimethyl sulfoxide con­ferred a 70 per cent protection. He obtained some protection with other sulfoxides tested but found that diethyl sulfoxide failed to protect the mice. Bridges ( 1962) observed that 1.0 M dimethyl sulfoxide protected Pseudomonas both aerobically and anaerobically against gamma-irradiation, and Dewey (1964) found that dimethyl sulfoxide protected Serratia marcescens against X-irradiation anaero­bically. MATERIALS AND METHODS Strains of microorganisms and the preparation of phage stocks have been described by Cotton, Rinehart, Petrusek and Stone ( 1964); irradiation techniques were those of Cotton and Lockingen ( 1963). Dimethyl sulfoxide and diethyl sulfide were commercial products which were redistilled before use. Diethyl sulfoxide, and dimethyl and diethyl sulfone were synthesized by Dr. J. M. Lagowski. RESULTS In all experiments the bacteriophage were suspended in a buffered M-9 solu­tion containing no carbon except the additives. All test materials were added to this system ten minutes before irradiation. Several different batches of phage were prepared at different times and tested in this system with qualitatively similar, but quantitatively different results. All data reported were obtained from a single batch of phage, and represent two or more independent experiments. From Figure 1 it can be seen that the control curve has two components with different slopes. Since all of the curves are exponential over the region from 40 to 90 Kr, a dose modifying factor due to the additive was determined by com­puting regression lines of the form y =ax+ b from 40 to 90 Kr inclusive and dividing the slope of the curve in the absence of additive by that obtained in the presence of additive (Bridges, 1962). The data in Figure 1 and Table 1 show clearly that diethyl sulfoxide protects the T2r bacteriophage against inactivation by X-irradiation, and that it is as effective as dimethyl sulfoxide, but both are less efficient than diethyl sulfide. The data also show that diethyl sulfone, at the 0.01 M concentration, has no protective effect and at the higher concentration sensitizes the phage to greater damage. 1 This investigation was supported (in part) by Public Health Service Research Grant No. GM 11609. by contract AT-( 40-1) -Z95Z with the Atomic Energy Commission, and by a grant from the Robert A. Welch Foundation. The University of Texas Publication \ \• \ \10-2 \ \ c 0 ~ \ -(.) \ 0 10-3 ~ LL \ .. 0\ c ' '., > 10-4 > ~ ::J Cf) 10-~ '•' ' -., ,.. ' '• 106.._~....__~....._~.......~--~~--~.......~........~.......~--~-­ 10 20 30 40 50 60 70 80 90 Dose in Kr FrG. 1. Survival of bacteriophage after X-irradiation. • Control, 0 0.01 M Diethyl sulfoxide. DISCUSSION It has been demonstrated that metabolizing systems often respond differently from non-metabolizing systems in tests of chemical and physical modifiers of radiation damage. It is not surprising therefore that diethyl sulfoxide protected the phage but did not protect the mice. Neither is it surprising that the phage were protected at concentrations of the sulfoxide down to 0.001 M whereas Serratia marcescens was not protected at concentrations below 0.01 M dimethyl sulfoxide (Dewey, 1964) and Pseudomonas required relatively high concentra­tions for full protection (Bridges, 1962). Lockingen: Sul/oxide protection of Bacteriophage The phage results are in general agreement with the results obtained with mice insofar as diethyl sulfone did not protect the phage and dimethyl sulfone failed to protect the mice (Ashwood-Smith, 1961). In other tests done in this laboratory it has been demonstrated that dimethyl sulfone does not protect the phage. Ashwood-Smith (1961) would not have seen an enhancing effect of the sulfone since he was using a lethal dose of X-rays, but enhancement was demon­strated at high concentrations in the phage experiments. Since both sulfides and sulfoxides confer protection and the sulfones do not, it is tempting to speculate that the electronic configuration about the sulfur atom is critical. The phage experiments are consistent with such a hypothesis, but offer no clues as to the mechanism by which protection is effected. ACKNOWLEDGMENTS The author wishes to thank the Chemical Products Division of the Crown Zellerbach Corp. for generous donations of dimethyl sulfoxide and also to thank Dr. J.M. Lagowski for the synthesis and purification of several of the compounds tested. The following undergraduate students participated in the experiments: Jeffry W. Roberts (presently a National Science Foundation Graduate Fellow at Harvard University), Mildred Considine, Sharon Rae and Mary K. Walker. TABLE 1 Dose modification factors for various additives Molar Slope Dose Additive concentration of curve modification factor None .42 1.0 Dimethyl sulfox:de 1.0 .14 3.0 0.1 .1Z 3.5 0.01 .11 3.8 0.001 .19 2.2 Diethyl sulfoxide 0.01 .1 2 3.5 0.001 .14 3.0 Diethyl sulfone 1.0 .57 0.75 0.01 .42 1.0 Diethy1 sulfide 0.001 .05 7.9 LITERATURE CITED Ashwood-Smith, M . J. 1961. The radioprotective action of dimethyl sulphoxide and various other sulphoxides. Int. J. Rad. Biol. 3: 41-48. Bridges, B. A. 1962. Protection of Pseudomona.s sp. against gamma-radiation by dimethyl sulph­oxide. Int. J. Rad. Biol. 5: 101-104. Cotton, I. M., and L. S. Lockingen. 1963. Inactivation of bacteriophage by chloroform and X-irradiation. Proc. Nat. Acad. Sci. 50: 363-367. Cotton, I. M., R. R. Rinehart, R. Petrusek, and W. S. Stone. 1964. Nitric oxide protection of bacteriophage against X-irradiation damage. Rad. Res. 21: 481-491. Dewey, D. L. 1963. The effect of protectors on Serratia marcescens during anaerobic X-irradia­tion. Int. J. Rad. Biol. 7: 151-154. XVII. Dimethyl Sulf oxide Treatment of Drosophila1 MARY L. ALEXANDER Dimethyl sulfoxide has the unique characteristics of very rapid cell penetra­tion and a versatile solvent action for a number of organic and inorganic com­pounds. It is suitable as a replacement for glycerol in the storage of human and bovine red cells and bovine spermatozoa (Lovelock and Bishop, 1959) . Other types of cells and cell components such as mitochondria and mouse bone marrow retain their activity after freezing in solutions of dimethyl sulfoxide (Anony­mous, 1962). The radioprotective action of dimethyl sulfoxide for survival in mice was reported by Ashwood-Smith (1961a), and Lockingen (this Bulletin) found a protective action of dimethyl sulfoxide against x-radiation inactivation of T~r bacteriophage. The difference in the efficiency of the protective action of dimethyl sulfoxide in metabolizing systems such as observed for the mouse and in non-metabolizing systems such as in bacteriophage makes testing radiopro­tection in the germ cell cycle of Drosophila interesting because of the different cellular types present. The reported accumulation of dimethyl sulfoxide in the testes of the mouse by Ashwood-Smith ( 1961 b) also offered the possibility of selective protection against radiation-induced genetic damage. MATERIALS AND METHODS Adult males of Drosophila melanogaster were treated with dimethyl sulfoxide by dipping, injection or by feeding. In the dipping experiments, the abdomens of young adult males, 7 to 10 hours after emergence from the pupae cases, were dipped in concentrated dimethyl sulfoxide. For the injections, .0004 ml. of a 1 X 10-1 M solution of dimethyl sulfoxide was injected into the abdomen of 7 to 10 hour old adult males. For the feeding experiments adult flies were placed on regular Drosophila food supplemented with 0.5% dimethyl sulfoxide. The eggs deposited on the food were allowed to develop into adults on the supplemented food. Several higher concentrations of dimethyl sulfoxide food supplement were tested but were toxic either to adult flies or to the developing offspring. The X-ray treatments were made with a Westinghouse Quadrocondex Machine operating at 250 KVP, 15 ma and using a filter of Yz mm Cu and 1 mm Al. Dose rates were either 540 or 620 r/ minute. X-ray treatments were made 4-5 hours after chemical treatments by dipping or injections of dimethyl sulfoxide. In the feeding experiment males were treated with 2000 r of X-rays one or two days after emerging. The experiments with dimethyl sulfoxide are listed in Table 1 along with the type of chemical treatments, X-ray dose, and types of germ cells tested for genetic damage. The germ cells of the spermatogenic cycle (mature sperm, spermatids, 1 This investigation was supported in part by Public Health Service Research Grants No. CA-07265 and GM-11609 from the National Institutes of Health, and by contracts AT-(40-1)­2952 and AT-(40-1)-3014 with the U.S. Atomic Energy Commission. The University of Texas Publication spermatocytes, and spermatogonia) are estimates of germ cell types based on the characteristics of genetic damage observed for various germ cell types in previous experiments. Germ cell types of the spermatogenic cycle were sampled by re­mating treated males, one male to one virgin female, every 3 days. The spenn cells, more mature at the time of treatment were sampled first (Period A) and TABLE 1 Dimethyl sulfoxide experiments Type of Experiment chemical treatment DMII Dipping pure chemical DMIII Dipping pure chemical DMVII Dipping pure chemical DMV Injection 1 X 10-1M X-5 Feeding X-52 .5% media supplement X·ray dose 4213r 1674r 2000r 2000r 2000r T,·pe of ge1m cells tested Mature sperm Mature sperm, spermatids, spermatocytes Mature sperm, spermatids, spermatocytes, ( sperma togonia?) Mature sperm, spermatids, spermatocytes, ( sperma togonia?) Mature sperm, spermatids, spermatocytes Genetic damage tested Dominant lethals Sex-linked recessive lethals Sex-linked recessive lethals; translocations Sex-linked recessive lethals Sex-linked recessive lethals Tables and figures Table2 Table3 Figure 1 Table4 Figures2, 3 Table 5 Figure 4 Tables 6, 7, 8 Figure 5 more immature cells were sampled in the following periods: Period C = sperma­tids; Period D = spermatocytes; Period E = spermatogonia. The types of genetic damage tested in each experiment are indicated in the fourth column of Table 1. The Tables and Figures illustrating the data for each experiment are indicated in the last column. Dominant lethals were measured as the per cent development of adults from egg samples counted. Sex-linked recessive lethals were tested by the usual Muller-5 method using the Marker-9 stock (In (1) sc-~1 , S, uf'sc•; Cy/ Bl V). Translocations between the two autosomes and between autosomes and the Y­chromosome were measured by genetic segregation of the recessive marker genes bw and st on the 2nd and 3rd chromosomes. TABLE 2 Dominant lethal damage induced in mature sperm dipped in dimethyl sulfoxide and treated with X-radiation Cont1·0Is Dimethyl sulfoxide X-ray Dimethyl sulfoxide + X·ray Total adults 2137 3959 776 718 Total eggs 2236 4172 4498 4574 Per cent emergence 95.6 94.9 17.3 15.7 Per cent lethals 4.4 5.1 82.7 84.3 Alexander: Dimethyl Sulfoxide Treatment of Drosophila RESULTS The data for dominant lethal damage in mature sperm are presented in Table 2. In the control test a 95.6% emergence was observed. The dimethyl sulfoxide treatment showed almost identical results with the controls, 94.9% emergence. With X-ray treatment a value of only 17.3 % emergence ( 82. 7% dominant lethals) was observed. When dimethyl sulfoxide treatment was used before X-ray treatment, the rate of 15. 7% emergence was similar to the X-ray test. Dominant lethal damage was not modified by dimethyl sulfoxide treatment before X-ray treatment in these tests. In experiments DM III and DM VII, the dipping experiments were extended to include spermatids, spermatocytes and possibly spermatogonia. The data for TABLE 3 Sex-linked recessive lethals induced by dipping in dimethyl sulfoxide and treated with X-radiation Mating period Period days A 1-3 B 3--6 c 6-9 D 9-12 Total Dimethyl sulfoxide Lethals Total sample %lethals 0 1102 0 2 946 0.21 0 684 0 3 1037 0.29 5 3769 0.13 X-ray 1675r Lethals Total sample %lethals 36 1202 2.98 48 1001 4.79 27 217 12.44 11 560 1.96 122 2980 4.09 Dimethyl sulfoxide +X-ray Lethals Total sample %lethals 39 1101 3.54 41 963 4.26 57 675 8.44 22 846 2.60 159 3585 4.43 DM III are given in Table 3 and illustrated in Figure 1. In this experiment there is a slight increase in mutation rate after dimethyl sulfoxide treatment. There was also a reduction in sex-linked recessive lethals in spermatids (period C) when flies were dipped in dimethyl sulfoxide before X-ray treatment. In the other mating periods there were no differences in the sex-linked recessive lethal rates when the X-ray and combined (chemical and X-ray) tests were compared. Although the reduction in spermatid cells was not significant it was impressive and the test was repeated in Experiment VII. Both sex-linked recessive lethals and translocations were tested in mature sperm, spermatids, spermatocytes and possibly spermatogonial cells. The results are given in Table 4 and Figures 2 and 3. In the sex-linked recessive lethal tests, the rate was increased slightly with dimethyl sulfoxide treatment as in the previous test. The reduction in genetic damage in Period C in the combined test was not repeated in this experiment. Almost identical amounts of sex-linked recessive lethals were induced with X-ray treatment alone and in the chemical+ X-ray test. The translocation tests (Figure 3) were similar to the sex-linked recessive lethals in that the pretreatment with dimethyl sulfoxide before X-ray treatment did not modify the results observed The University of Texas Publication 12 CJ) ...J CJ) CJ) 8 w u w er: 0 w 6 z ~ ::::::; I x w CJ) 4 ..... z w u er: w 2 a. oL...~&..-===~======-ai.:::::::::::~::.....J Periods A B C D Doys 1-3 3-6 6-9 9-12 MATING PERIODS FIG. 1. Induction of sex-linked recessive lethals with X-radiation after dipping in dimethyl sulfoxide. with X-ray alone. The effects of dimethyl sulfoxide for inducing genetic damage and modifying X-ray-induced damage were tested by injecting the chemical in Experiment V (Table 1) and the results are given in Table 5 and Figure 4. The rate for sex-linked recessive lethals after dimethyl sulfoxide treatment was slightly higher than the control rate of .04% for the Oregon-R Oak Ridge Stock observed in previous tests (Alexander and Bergendahl, 1964). To test for an increase in mutation rate with dimethyl sulfoxide, a control experiment was carried out using flies taken at the same time as those used in the other tests. The rate for the control test was 0.036%, which is similar to the previous control rate. There was a slight increase in the rate after dimethyl sulfoxide treatment. Al­though the difference was not significant the increase was repeated in two of the dipping experiments and in the injection experiment. In the injection experiment the X-ray and combined tests were similar (the values being within 3% of each other) throughout the germ cell cycle. The com­bined test was lower than X-ray alone in Period C but equally higher in Period D. Alexander: Dimethyl Sul/oxide Treatment of Drosophila TABLE 4 Genetic damage induced by dipping in dimethyl sulfoxide and treated with X-radiation Mating periods Period A B c D E Test days 1-3 ~G 6-9 9-12 12-15 Total SEX-LINKED RECESSIVE LETHALS Dimethyl sulfoxide Lethals 0 2 0 1 0 3 Total sample 555 711 419 586 540 2811 % lethals 0.0 0.28 0.0 0.17 0.0 0.1 1 X-ray 2000r Lethals 29 33 25 16 6 109 Total sample 826 693 205 616 735 3075 % lethals 3.51 4.76 12.19 2.59 0.82 3.54 Dimethyl sulfoxide Lethals 45 50 30 11 7 143 +X-ray Total sample 955 958 260 638 695 3506 % lethals 4.71 5.32 11.5 1.72 1.01 4.08 TRANSLOCATIONS Dimethyl sulfoxide Translocations 0 2 0 0 0 2 Total sample 560 593 567 533 542 2795 %translocations 0.0 0.34 0.0 0.0 0.0 0.07 X-ray 2000r Transloca tions 26 34 54 31 5 150 Total sample 766 707 461 582 673 3189 % translocations 3.39 4.81 11.71 5.33 0.74 4.70 Dimethyl sulfoxide Translocations 30 39 101 40 16 226 +X-ray Total sample 944 942 736 564 839 4025 % translocations 3. 18 4.14 13.72 7.09 1.91 5.61 TABLE 5 Sex-linked recessive lethals induced with injections of 10-1M of dimethyl sulfoxide and X-radiation treatments Mating periods Period A B c D E Test days 1-3 3--0 6-9 9-12 12-15 Total Controls Lethals 0 0 0 0 1 1 Total sample 598 499 525 543 571 2736 % lethals 0.0 0.0 0.0 0.0 .18 .036 Dimethyl sulfoxide Lethals 0 1 0 1 1 3 Total sample 254 550 206 555 536 2101 %lethals 0.0 0.18 0.0 0.18 0.19 0.14 X-ray2000r Lethals 35 32 8 14 10 99 Total sample 573 566 79 433 674 2325 %lethals 6.11 5.65 10.13 3.23 1.48 4.26 Dimethyl sulfoxide Lethals 17 24 7 5 6 59 +X-ray Total sample 381 513 96 94 471 1555 % lethals 4.46 4.69 7.29 5.32 1.27 3.79 The University of Texas Publication 12 en ...J ~10 I- w ....J w > ~8 w u w 0:: 0 w 6 :::c:: z ::J I x w Cf) 4 I- z w u 0:: w a.. 2 oL...--....,::::::::::~===---=:!::llt=-...i&-J Periods A B C D E Days 1-3 3-6 6-9 9-12 12-15 MATING PERIODS FIG. 2. Induction of sex-linked recessive lethals with 2000r of X-radiation after dipping in dimethyl sulfoxide. There does not appear to be a significant modification in X-ray damage with injections of dimethyl sulfoxide. The feeding experiments are given in Tables 6, 7, and 8 and the totals are illustrated in Figure 5. The experiments X-52 and X-5 were separate experi­ments. Males were treated with X-ray one or two days after emergence. The results for induced sex-linked recessive lethals after dimethyl sulfoxide feeding are given in Table 6. There were no recessive lethals induced by feeding dimethyl sulfoxide. The results after X-ray treatment (without dimethyl sulfoxide treat­ment) are given in Table 7 and the combined chemical and X-ray treatment is given in Table 8. With X-ray alone the rate for spermatids (Period C) was unusually low. The sex-linked recessive rates for Period C are usually higher than 6.8% with a 2000r treatment. The X-ray tests in Experiments DM III, VII and V gave rates of 10.13% to 12.44%. The combined treatments (Table 8) gave Alexander: Dimethyl Sulfoxide Treatment of Drosophila 10 X-ray \ z \ I­ 0 \ \ ct \ X-ray+ 0 8 u \ OMS ....J \ (/) z ~ ct I a: \ t­ \ 6 \ I­ \ z \ w \ u a: \ w \ Q. 4 \ \ \ \ \ 2 1 Periods Days A 1-3 B 3-6 C 6-9 D 9-12 E 12-15 MATING PERIODS FIG. 3. oxide. Indu ction of transloca tion with 2000r of X-radiation aft er dipping in dimethyl sulf­ higher values for Period C except Experiment X-52 using 1-day-old o o. The totals for all the experiments gave values of 3.6% for mature sperm which in­creased to 10.1 % in Period C. This value is higher than the X-ray value of 6.8%. The combined treatment is higher but the differences are not significant and the unusually low value for X-ray treatment makes the higher rate for the combined treatnl.~nt of less importance. D1scussroN The various germ cells of the spermatogenic cycle showed no protection from dimethyl sulfoxide when chemical treatments were made by injections or dipping four hours before radiation treatment. Protection was not afforded any type of germ cell stage, neither the non-dividing mature sperm nor metabolically active The University of Texas Publication 12 ~ w (f) (/) 8 w u LLJ a:: c w 6 ~ X-ray+ z 40MS :J I \ x \ (/) \ w 4 \ t­ \ z w \ u \ a:: \ w a.. 2 \ \ oL..~a...-==~==_.......:Y~==~~ Periods A B C D E Days 1-3 3-6 6-9 9-12 12-15 MATING PERIODS FIG. 4. Induction of sex-linked recessive lethals after injection of 1 X 10-1 M solution of dimethyl sulfoxide and treatment with 2000r of X-rays. spermatocytes. The time which elapsed between the chemical treatment and X-radiation treatment may be too long to detect protection if dimethyl sulfoxide metabolizes to another form. The maximum protection in mice was obtained from 5 minutes to one hour after intraperitoneal injection and then decreased in efficiency the next hour (Ashwood-Smith, 1961a). An unexpected reduction in the sulfoxide content of the spleen a short time after injection was first reported by Ashwood-Smith ( 1961 b). The possibility of a lower oxygen tension and there­fore radiation protection in the spleen was studied further by Meer, Valkenburg and Remmelts ( 1963) . Two strains of mice showed a difference in the protective action of dimethyl sulfoxide and this difference corresponded with changes in oxygen tension in the spleen of the two strains. This suggests that dimethyl sulfoxide protects in the mouse by causing hypoxia in the spleen and in other blood forming organs. Alexander: Dimethyl Sulfoxide Treatment of Drosophila TABLE 6 Sex-linked recessive lethal tests after dimethyl sulfoxide feeding Mating periods Periods A c D Experiment days 1-3 6-9 9--12 Average X52 Lethals 0 0 0 0 1-day-old 5 5 Total sample 294 291 289 874 % lethals 0.0 0.0 0.0 0.0 X-5 Lethals 0 0 0 0 1-day-old 5 5 Total sample 351 358 265 974 % lethals 0.0 0.0 0.0 0.0 X52 Lethals 0 0 0 0 2-day-old 5 5 Total sample 345 345 320 1010 %lethals 0.0 0.0 0.0 0.0 X-5 Lethals 0 0 0 0 2-day-old 5 5 Total sample 245 220 353 818 % lethals 0.0 0.0 0.0 0.0 Totals Lethals 0 0 0 0 Total sample 1235 1214 1227 3676 % lethals 0.0 0.0 0.0 0.0 TABLE 7 Sex-linked recessive lethal damage after X-ray treatment Mating periods Periods A c D Experiment days 1-3 6-9 9-1 2 Average X52 Lethals 15 14 13 42 1-day-old 5 5 Total sample 461 177 461 1099 % lethals 3.2 7.9 2.8 3.8 X-5 Lethals 19 4 3 26 1-day-old 5 5 Total sample 482 65 48 595 % lethals 3.9 6.2 6.2 4.4 X52 Lethals 4 2 4 10 2-day-old 5 5 Total sample 89 47 84 220 % lethals 4.5 4.3 4.8 4.5 X-5 Lethals 6 6 7 19 2-day-old 5 5 Total sample 156 96 283 535 % lethals 3.8 6.3 2.5 3.6 Totals Lethals 44 26 27 97 Total sample 1188 385 876 2449 % lethals 3.7 6.8 3. 1 4.0 The University of Texas Publication TABLE 8 Sex-linked recessive lethal damage after dimethyl sulfoxide feeding and X-ray treatment Mating periods Periods A c D Experiment days 1-3 6-9 9-1 2 Average X52 Lethals 7 10 3 20 1-day-old 5 5 Total sample 151 132 153 436 %lethals 4.6 7.6 1.9 4.6 X-5 Lethals 10 8 7 Z5 1-day-old 5 5 Total sample 345 62 152 559 %lethals 2.9 12.9 4.6 4.5 X52 Lethals 17 20 10 47 2-day-old 5 5 Total sample 524 219 504 1247 %lethals 3.2 9.1 2.0 3.8 X-5 Lethals 1Z 10 5 27 2-day-old 5 5 Total sample 262 64 140 466 %lethals 4.6 15.6 3.6 5.8 Totals Lethals 46 48 25 119 Total sample 1282 477 949 2708 %lethals 3.6 10.1 2.6 4.4 If protection should be offered by dimethyl sulfoxide in Drosophila it does not depend upon secondary reactions since there was no protection from radiation when treatment was given 4 hours after chemical treatment. The feeding experi­ments indicated that there were no products from dimethyl sulfoxide feeding built into the germ cells which offered protection against irradiation. There is an accumulation of dimethyl sulfoxide in mouse testes but there was no protection from irradiation when damage was measured as reduction in testicular weight or by histological methods (Ashwood-Smith, 1961 b). The concentration of di­methyl sulfoxide was reported to be high enough to offer protection if free radical inactivation was involved. The work in the mouse indicates that oxygen tension can be effected by di~ methyl sulfoxide and therefore modify radiation damage. There is also a corre­lation of radiation protection with compounds with different oxygen affinities in non-metabolizing systems (see Lockingen, this Bulletin). Dimethyl sulfoxide appears to have an effect upon oxygen tension in some systems and may require shorter periods between chemical and radiation treatment to be detected in Drosophila. SUMMARY The developing germ cells of the spermatogenic cycle of Drosophila were treated with dimethyl sulfoxide by injections, dipping and feeding techniques. Dimethyl sulfoxide treatment increases the sex-linked recessive lethal mutation rate slightly above the control rate in some experiments. The increase is not significant and additional data are required to show whether there is even a slight increase in mutation rate. Alexander: Dimethyl Sul/oxide Treatment of Drosophila 12 en ....I ~ 10 ~ I­ I \ w I \ ....I ·/ \ w I \ I \ > Ci) I \ 8 en I \ w I \ u I \ w /I \ \ a:: X-ray 0 w 6 ~ z ::i I x w en 4 I- z w u a:: w a.. 2 OMS 0'--~-41---~~~~~--~~~'--~19--~---' Periods A c 0 Days 1-3 6-9 9-12 MATING PERIODS FIG. 5. Induction of sex-linked recessive lethals with 2000r of X-radiation after dimethyl sulfoxide feeding. There were no indications of protection against radiation-induced genetic damage when X-ray treatment was given 4 hours after chemical injections or dipping. Feeding experiments also failed to protect germ cells from radiation damage. ACKNOWLEDGMENTS We would like to thank the Crown Zellerbach Corporation, Camas, Washing­ton, for furnishing the research samples of dimethyl sulfoxide. LITERATURE CITED Alexander, Mary L., and Janet Bergendahl. 1964. Dose rate effects in the developing germ cells of Drosophila. Genetics 49: 1-16. Anonymous. 1962. "Dimethyl Sulfoxide Reaction Medium and Reactant." Crown Zellerbach Corporation, Chemical Products Division, Camas, Washington, pp. 25-27. The University of Texas Publication Ashwood-Smith, M. J. 1961a. The radioprotective action of dimethyl sulphoxide and various other sulphoxides. Int. J. Rad. Biol. 3: 41-48. -----. 1961 b. Inability of dimethyl sulphoxide to protect mouse testis against the effect of X-radiation. Int. J. Rad. Biol. 3: 101-103. Lockingen, L. S., this Bulletin. Lovelock, J. E., and M. W. H. Bishop. 1959. Prevention of freezing damage to living cells by dimethyl sulphoxide. Nature 183: 1394-5. Meer, C. Van der, P. W. Valkenburg, and M. Remmelts. 1963. Experiments on the radiopro­tective action of dimethyl sulphoxide. Int. J. Rad. Biol. 6: 151-155. XVIII. An Operational Classification of Drosophila Esterases for Species Comparisons1 F. M. JOHNSON, CARMEN G. KANAPI, R.H. RICHARDSON, M. R. WHEELER, and w. s. STONE INTRODUCTION Biochemical techniques designed to show the relatedness of species on the basis of protein differences were initiated when Nuttall (1904) used immunological methods to compare the serum of humans with that of other primates. Increasing sophistication since then has led to much higher achievement at the protein level in taxonomy-the determination of amino acid sequences in homologous proteins from organisms of different taxonomic categories (Anfinsen et al., 1959; Zucker­kandl et al., 1960). At the genetic level the comparison of nucleotide sequences has been approached by comparing the degree of hybridization of heterologous melted DNA from closely and distantly related organisms (Hoyer et al., 1964). A technically more convenient but rather crude method of comparing molecu­lar differences between species is to measure the electrophoretic mobility of pro­teins in a gel medium such as starch (Smithies, 1955) or acrylamide (Raymond and Weintraub, 1959). Among primate groups certain serum proteins appear to have characteristic migration rates (Beckman, 1963). Protein with enzymatic properties can be compared on the basis of catalytic activity in the presence or absence of inhibitors. In a comparison of species, for example, the diphosphopyridine nucleotidases from ruminant mammals were found to be inhibited by isonicotinic acid hydrazide but the same enzyme from other mammals was not inhibited (Kaplan et al., 1960; Kaplan et al., 1959; Zat­man et al., 1954). A combination of gel electrophoresis and histochemical enzyme detection tech­niques (Hunter and Markert, 1957) makes it possible to combine electrophoretic mobility and catalytic activity comparisons. Substrate affinities and inhibition­activation properties of erythrocyte esterases from prilnates have been charac­terized in starch gels with p-chloromercuribenzoate, diisopropylfluorophosphate, eserine and acetazolamide (Tashian, 1965). The present communication is a similar approach with species groups in several subgenera of the genus Dro­sophila. Previous investigations of Drosophila esterases have been primarily concerned with the inheritance of variants of electrophoretic mobility in D. melanogaster. Thus Wright (1963) and Beckman and Johnson (1964) demonstrated the co­dominant allelic control of electrophoretic variants of two independent esterases 1 The research reported here was supported (in part) by USPHS research grant (GM-11609 to W. S. Stone and M . R. Wheeler ) and training grant (2 Tt-GM-337-06 to R. P. Wagner, et al.) from the National Institutes of Health, U. S. Public Health Service. The work is, in part, research by the senior author to be included in a dissertation to be submitted to The University of Texas in partial fulfillment of the Ph.D. requirements in Zoology. The University of Texas Publication separable by starch gel electrophoresis. Wright's ( 1963) esterase is termed Ester­ase 6 and that of Beckman and Johnson ( 1964) is Esterase C. They are presumed to be controlled by the Est 6 and Est C structural loci respectively. In D. melano­gaster there are at least three additional enzymes which show considerable ester­ase activity, termed Esterase A, Esterase E, and Esterase F, and which are prob­ably also under independent control. MATERIALS AND METHODS Table 1 lists the University of Texas (Genetics Foundation) stock numbers, collection localities and dates, and the data obtained from the species investigated in this study. In all cases only single adult flies from laboratory stock cultures were examined. Culture medium was either standard cornmeal-molasses or banana. Six to· eight individuals from each stock were examined. Full data were obtained on 76 species including 108 different strains and partial data on one additional species and 6 additional strains. Starch gel electrophoresis was carried out horizontally with the discontinuous system of buffers described by Poulik ( 195 7). Some details of the procedure have been published previously (Beckman and Johnson, 1964; Johnson, 1966). Starch concentration in the gels was 13 per cent and it was poured to a depth of 9 mm in 19 cm x 21 cm lucite molds. Each gel contained samples from the species under examination, and, in addition, samples of D. melanogaster for reference. Electro­phoresis was carried out for 3 hours at a voltage gradient of 10 volts per cm. After electrophoresis each gel was cut into four horizontal replicate slices before the enzyme assays were performed. The bottom slice of the gel was 3 mm thick and the other three were 2 mm thick. The top slice was discarded, since it often pro­duced distorted isozyme patterns. A survey of a small number of diverse Drosophila species was first started by utilizing several potential substrates and inhibitors in various concentrations and combinations. Among the substrates which seemed particularly useful in demon­strating specificity differences was a mixture containing the esters a-and {3­napthyl acetate. N-propyl alcohol was found to have both inhibitory and activat­ing characteristics on various esterases. Therefore, a-and [3-naphthyl acetate and alcohol were used to continue the investigation. The staining technique for assaying the esterases using the mixed substrate and alcohol effects was a modification from Beckman and Johnson ( 1964). The a­napthyl acetate. (ANA) and [3-napthyl acetate (BNA) were used as substrates as follows. Stock solutions of ANA and BNA were prepared by dissolving 1 g of the ester in 100 ml of an acetone-water mixture (1: 1 by volume). The solution of dye coupler was prepared by dissolving 75 mg Fast Blue RR salt in 100 ml of 0.1 M sodium phosphate buffer, pH 6.0. Staining solutions were as follows: Bottom slice: 1.5 ml ANA stock 2 ml BNA stock 5 ml n-propanol (absolute) 100 ml dye-buffer Middle slice: 1.5 ml ANA stock 10 ml n-propanol (absolute) 100 ml dye-buffer Johnson et al.: Drosophila Esterases and Taxonomy Upp€r slice: 1.5 ml ANA stock 100 ml dye-buffer All gels were incubated in the staining solutions at 37°C for 2 hours. This stain­ing system allowed the simultaneous determination, on replicate samples, of the esterase phenotyp€ of various Drosophila species compared with patterns observed in D. melanogaster. ANA-BNA specificity is loosely defined in one respect. Usually, if ANA or BNA are used separately as substrates for gels prepared from a single Sp€Cies, similar banding patterns are obtained on the homologous gels and no substrate specificity is detected. The only variation between gels is a color difference; the ,8-napthol Fast Blue RR precipitate is red, but the a-naphthol Fast Blue RR com­plex is blue-black. However, if the ANA and BNA substrates are together in the same staining solution the rates of hydrolysis of the two esters by some esterases are sufficiently different that predominately red or predominately blue-black precipitates are deposited in many instances. Resulting enzyme patterns show some bands which are clearly red and others which are blue-black (Figure 1a) . The esterase specificity referred to in this report concerns this difference. The 1.5 ANA: 2 BNA ratio was selected on the basis of good average differentiation. Tashian ( 1965) has reported a similar specificity difference with primate eryth­rocyte esterases. When 10 ml n-propanol is added to the staining mixture one esterase in D. melanogaster (Esterase 6) is inhibited if only the ANA substrate is present. If both ANA and BNA are present no appreciable inhibition is detected. In this case a smaller amount (5 ml) of n-propanol was found actually to improve the zymo­grams, resulting in greater color differentiation. This accounts for the inclusion of alcohol in the slice stained with a mixture of ANA and BNA. RESULTS The various electrophoretic esterase phenotypes of laboratory strains of D. melanogaster have been examined under the above conditions including replicate samples of the same phenotyp€ but from independent sources. Each independent enzyme has certain properties which distinguish it regardless of electrophoretic variation or source of strain. For example, Esterase A enzymes migrate relatively fast, have low activity which is increased when alcohol is present, and are specific for a-napthyl esters when both a-and (3-forms are present together. Esterase C enzymes are not affected by the presence of alcohol and are specific for a-naph­thol esters when both substrates are present. Esterase 6 enzymes are inhibited by alcohol when only the a substrate is present and are specific for the ,8 ester when both are present. These characteristics apply to all electrophoretic types which have been examined. This does not imply that exceptions are impossible but only that they should be rare compared to the generalized esterase phenotype. It is assumed on this basis that a small sample from any given species will allow the determination of its generalized phenotype; the correctness of this idea should be borne out by success in separating species groups, sections and subgenera accord­ing to the phenotypes of their member species. A wide range of electrophoretic mobility is often encountered for proteins rep­ The University of Texas Publication C\I c .0 u Johnson et al.: Drosophila Esterases and Taxonomy resenting a genetically homologous but polymorphic locus within a species. In some instances several alternate allelic forms may be present in high frequencies, even within highly inbred lines, but in other cases it may be that various alleles are fixed in different lines. Comparison of electrophoretic mobility is of no value in such cases. Efficient interspecific comparisons require that one be able to allo­cate variability to either intraspecific polymorphism or to true interspecific dif­ferences. Assuming that the majority of electrophoretic variants of the same system retain similar catalytic properties generally, it is then possible to compare species on the basis of presence or absence of enzymatically homologous systems without the confusion of taxonomically less meaningful electrophoretic differ­ences. In many cases enzymatic homology probably involves a degree of genetic homology among related species. For comparative purposes the esterase classification scheme was defined as follows : A. Type A Esterase differences. Any band in the upper quarter of the mobility range which was ANA specific and of lesser staining intensity than at least one other esterase zone was defined as an A esterase. Its density was compared with that of D. melanogaster and scored L (=light), if no darker than D. melanogaster Esterase A, or D (=dark) if decidedly denser in staining intensity. Both middle In the list of species and strains used in this investigation, the University of Texas stock num­ber, collection site and collection dates are listed if they were available. The letter following the species name refers to the medium on which the stock is cultured: banana = b, cornmeal = c. Polymorphisms are indicated as belonging to the system analogous to Esterase A of D. melano­gaster (A) or to fi-esterase (/1) or to a-esterase (a) systems. lnterstrain polymorphisms for a given species are indicated with the first strain if there were qualitative differences.* Esterase classification data follow polymorphism data. The text should be consulted for a complete key to the last four columns. A zero in any column is equal to "no" or "none" and a blank space in a column indicates that the information was not obtained or was not available. •A 0 indicates no variation observed among strains, while a-indicates only one strain examined. Fm. 1. Photographs of Drosophila species homogenates after electrophoresis in a gel sliced horizontally into three replicates (a, b, and c) which are differentially stained to reveal the cate­gories of esterase differences. Fig. ta is the bottom slice stained in the presence of both ANA and BNA; 1 b is the middle slice which was stained with ANA and alcohol; and 1c is the upper slice which was stained with ANA but without alcohol. Part t of the gels shows the pattern from two individuals of D. melanogaster which was used for reference purposes. The subgenus Sophophora is represented by parts 2 (D. athabasca of the obscura group), 3 and 4 (D. serrata and D. athabasca of the obscura group) , 3 and 4 (D. serrata and D. simulans of the melanogaster group), 5 (D. austrosaltans of the saltans group), and part 6 contains samples of D. nebulosa from the willistoni group. The subgenus Drosophila is repre­sented by 7 (D. innubila of the quinaria group) and 8 (D . americana of the virilis group) . Direction of mobility is indicated by the arrows. From ta the major esterase specificity pat­terns may be scored a, {1 or a-/1 on the basis of predominately blue-black, red or about equal red and black zones. Alcohol inhibition of /1 esterases may be observed by locating the red bands of slice ta on slices tb and 1c and comparing their activities as the degree of band darkness between the latter two slices. The activating effect of alcohol on A esterase bands may be observed similarly; however, the L, D scoring of A esterases is not reliable from photographs. U\ TABLE 1 to to Esterase A !JroMJflh ila Food Colleclion number L"cal:ty Colleclion date Tnlrastra in polymorphism Interstrain variation +Ale -Ale Maj or csterase Major inhibition Subgenus: Hirtodrosophila duncani b 2311.9 Florida /3 - D D /3 {31 Subgenus: Dorsilopha busckii b H 360.127 Panama 1958 a,/3 - L L a.{3 {31 Subgenus: Sophophora ~ Group: obscura ;:r... pseudr;obscura c 2258.1 Mexico 1952 0 - L L /3 {31 ~ algonquin c 2258.5 Wisconsin 1958 0 - D L /3 {31 ~ ;:i azteca c 1166.3 Mexico 1952 0 - D L /3 {31 ~· narragansetl c 2528.9 Nebraska 1958 a - D L /3 {31 ~ .... athabasca c Manit:>ba 0 a,/3 L L a/3 {31 "'-....... athabasca c 3000.2 Alaska 1960 0 L L a/3 {31 ~ 0- ambigua c 0 - L L a/3 {31 affinis c 2528.6 Nebraska 1958 A a D D /3 {31 ~ affinis c Arkansas D D /3 {31 ~ i::i affinis c 2528.8 N ebraska 1958 "' subobscura c Norway a,{3 - D L a.{3 {31 "i:i i::: Group: m elanogaster ~ ...._-. serrata c 2040.6 Australia 1956 0 0 L L a.{3 {31 ~ i::i serrata c 3007.1 Australia 1960 0 L L a/3 {31 .....-· rufa c 1736.3 China 1947 a,{3 - L L a/3 {31 0 ;:i simulans c H435.4 Columbia 1960 0 a L L a/3 {31 simulans c ..... Rarotonga 1962 a,{3 L L a/3 {31 bipectinata c . . . . . Suva 1963 0 - L L a {31 yakuba c 2371.6 Africa 1956 a - L L af3 {31 seguyi c 2371.4 Africa 1956 a - D D a/3 {31 auraria c 1736.1 China 1947 0 - L L a/3 {31 takahashii c 2363.4 Nepal 1953 A - D D a/3 {31 mayri c 3007.2 N ew Guinea 1960 /3 - L L a {31 szentivanii c 3007.3 N:)w Guinea 1960 /3 0 L L a {3I szentivanii c melanogaster c Group: saltans emarginata c emarginata c saltans c saltans c septentriosaltans c nigrosaltans c austrosal:ans c prosaltans c prosaltans c prosaltans c prosaltans c parasaltans c pseudosaltans c subsaltans c lusaltans c neoelliptica c neocordata c ear lei c Group: willistoni nebulosa c capricorni c sucinea c willistoni c Subgenus: Pholad?ris Groups: various (mixed) latifasciaeformis c latifasciaeformis c pattersoni b cancellata c 3033.11 62.51 2262.8 1911.4 1401.4 H103.21 H360.92 2536.4 H4-00 H436.5 H428.33 H180.40 2536.1 2536.10 2536.2 H411.20 2536.5 2536.7 2374.3 2394.2 H442.22 H57.30 H75.4 3012.1 2372.1 Malaya Salvador Mexico Mexico Mexico Columbia Panama Brazil Costa Rica Columbia Panama Costa Rica Brazil Brazil Brazil Haiti Brazil Brazil Brazil Peru C:ilumbia Nicaragua Costa Rica Costa Rica Australia 1962 1954 1952 1948 1943 1955 1958 1959 1959 1960 1960 1956 1959 1959 1959 1959 1959 1959 1956 1956 1960 1954 1955 1961 1955 a A,a,{3 0 A 0 0 0 0 A 0 0 0 0 /3 0 0 0 0 0 0 A 0 0 0 0 0 a f3 A,a,/3 a 0 --- a,/3 --- - - - - --- - 0 - - L L L L L L L L L L L L L L D L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L L a afJ /3 /3 /3 /3 /3 /3 /3 (J /3 a/3 f3 /3 f3 /3 /3 /3 (3 /3 (3 0 0 0 (3 (3 /31 {31 {31 {JI {31 /31 {JI {31 {31 (JI {31 {JI /31 {31 {31 {31 {31 {31 0 /31 {JI [3I 0 0 0 0 .... 0 ;:::... ;::$ "'0 ~ (I:) .... ~ :.­ ti .... 0 "'0 'l;j ;:::... ~. El" ~ .... (I:) .... (I:) "' ~ "' ~ ;::$ ~ ~ N 0 ::i 0 ~ ~ (}\ to w U"l TABLE I-Continued l'O + Esterase A Collection Collection Intrastrain luterstrain Major Major Drosophila Food number I.ocn li ty dote polymorphism variation +Ale -Ale esterase inhibition Subgenus: Drosophila Section : quinaria Group: calloptera ornatipennis b 2378.2 Cuba 1956 0 - D L a {31 Group : immigrans immigrans b 2321.9 South Carolina 1953 a - D D a {31 "l ~ Group: testacea putrida putrida b b 2067.3 2539.1 Nebraska Texas 1950 1960 0 a 0 L L L L a a {31 {31 (1) c::! ;:i-. ~ Group: tripunctata mediostriata b H341.13 Brazil 1958 a,/3 0 L L a /31 (1) .... "'-...... mediostriata crocina b b H442.5 H1 31.2 Columbia Puerto Rico 1960 1956 a a a L L a 0 ~ 0- croczna b H 86.12 Canal Zone 1955 a L L a 0 "l crocina b H29.1 b El Salvador 1953 0 L L a 0 ~ crocina b no number (dark variety) 0 L L a 0 "' crocina tripunctata unipunctata mediodiffusa Group: cardini b b b b H437.1 0 2539.2 H91.14 H 260.1 Brazil Texas Columbia Puerto Rico 1960 1960 1955 1957 a /3 a a --- L L L L L L a a a /31 al 0 "ti >:: ""' .._-.() $::) .....-.0 ;:i cardini b H336.31 Brazil 1958 0 a,{3 D D a/3 0 cardini b H332.18 Trinidad 1958 a L L a 0 cardinoides cardinoides neomorpha neomorpha dunni b b b b b 2256.2 H1 91.21 H183.4 H80.8 H239.6 Mexico Columbia Panama Canal Zone West Indies 1952 1956 1956 1955 1957 /3 0 /3 /3 0 /3 /3 a L L L L L L L L L L /3 /3 af3 a/3 a /31 {31 /31 {31 0 dunni b H1Z2.1 West Indies 1956 a L L a 0 parthenogenetica b H56.85 Nicaragua 1954 /3 0 L L /3 /31 partherwgenetica acutilabella acutilabella Group: quinaria innubila innubila falleni falleni phalerata occidentalis occidentalis munda subpalustris subpalustris palustris quinaria Group: funebris funebris Section: virilis-repleta Group: virilis littoralis littoralis borealis borealis amerzcana americana amerzcana amerzcana americana novamexicana novamexicana Group : repleta repleta repleta pachea b b b b b b b b b b b b b b b b c c c c c c c c c c c b b b(mod.) H50.28 H351.2a H350.9 2074.6 2056.10 1062.6 1915.1 2575.3 ..... 1877.9 3012.2 1757.13 1753.7 1732.3 2000.3 2096.1 2077.4d 1950.1 45 (stand. ) 1893.10 1760.Si 1901 .5a 1773.42 A21 3 1714.4 3006.4 H435.43 A113.1 Honduras Jamaica Jamaica New Mexico Arizona Tennessee Lebanon New Mexico South Carolina South Carolina Minnesota Minnesota Lebanon Switzerland Switzerland Minnesota Idaho Vermont Montana Michigan N ebraskn Arizona New Mexico Brazil Columbia Mexico 1954 1958 1958 1950 1950 1942 1949 1960 1948 1961 1947 1947 1947 1950 1951 1950 1950 1948 1947 1948 1947 1947 1960 1960 1965 /3 0 0 0 a a a 0 a a A 0 A 0 0 a a 0 0 0 0 /3 a,/3 /3 0 0 0 0 0 A ,a /3 a 0 -0 - /3 - - 0 a,{3 /3 a 0 -L L L L D L L L L D L L D D D L L L L D L L L D L L D D D L L L L L L L L L L D D L L L D D L L D L L L D L L L L L /3 a a a a a a a a a a a a a a/3 a/3 af3 a/3 af3 a(3 a/3 a/3 a{J a{J a(3 a/3 a/3 af3 {JI {JI 0 0 0 0 {JI {JI {JI {3I /3I /3I {3I {3I {3I 0 0 0 0 0 0 0 0 0 0 0 0 0 {3I ....... 0 ~ ~ 0 ;:l ~ ...... ~ .._ : ti ..., 0 0 "' 'c::l ~ .._ ~ ...... "' ~ (1) ..., ~ "'~ "' ~ ;:l ~ ~ f.< 0 ;:l 0 ~ ~ U\ to U\ (.)\ to 0) TABLE 1-Continued '"-3 ~ ('\:) Drosophila Food Collection n umber Loca lity Coll er lion date --- lntrastrain polymorphi sm · Interstrain va riation Eslerase A +Ale -Ale Major C' StE'r:) "' '\:j I::: I:)­.....-(") . >:) ..... ~-. 0 ;::$ Johnson et al.: Drosophila Esterases and Taxonomy and upper gel slices were scored so that staining intensities were determined both with and without the effects of alcohol. B. Specificity in major esterases. Major bands were defined as any band except A esterases and except any zone migrating at an identical rate to Esterase E in D. melanogaster. With these restrictions the esterase pattern was judged as pre­dominantly ANA or BNA specific on the basis of red vs. blue-black color of the bands. Those which showed both red and blue-black bands of approximately equal number and intensity were classified in a third category. Representative patterns are shown in Figure 1a. C. Alcohol inhibition of major esterases. All bands demonstrated in the staining of the bottom slice (ANA + BNA + propanol) were also observed in the staining of the upper slice (ANA); however, they were not differentiated as to ANA ­BNA specificity in the latter stain. Major esterase bands demonstrated in the bot­tom and upper slices but not observed or very weakly stained in the middle slice (ANA + propanol) were designated as alcohol inhibited and judged ANA or BNA specific from their color in the bottom slice (ANA + BNA + propanol). The complete comparison data are presented in Table 1. In a few cases in­volving strains of a species from more than one locality, there was sufficient dif­ference to warrant separate categories for a part of the total esterase phenotype. This is recorded in Tables 2, 3 and 4 by fractional species within categories. The difference was usually due to the absence of a major esterase in one strain. Rela­tively rare polymorphism of this type has been noted in D. melanogaster esterases (Johnson, 1964) and it is an expected source of occasional error here. Electro­phoretic differences between and within strains appeared far more frequently than null variants, but the classification scheme was not affected by this type of difference. In all species a nearly constant band which corresponded in mobility to Este­rase E. of D. melanogaster was found. This band was of much less staining in­tensity in D. nigrosaltans and somewhat less intense in a number of other species. Esterase E from D. melanogaster is inhibited by 10-s M eserine as is the enzyme of similar mobility in many other Drosophila species (Johnson, unpublished data). The fact that such a constant enzyme is found may suggest that some pro­teins continue unaltered in the course of evolution. This may be because they do not follow the pattern of frequently occurring, catalytically neutral amino acid substitutions. In this case it is conceivable that almost any change alters the struc­ture enough to make new catalytic properties incompatible with life in Dro­sophila. A less exciting possibility is that the changes occur but are simply not detectable. Both esterase A classifications, with and without alcohol added to the stain, show considerable variability among species within most groups. A resemblance to D. melanogaster is the common trend as is indicated by the summarized data in Table 2. Here the first letter in the hyphenated Esterase A type refers to the density in the stain containing alcohol while the second letter refers to the density in the ANA stain with no alcohol. A letter L indicates a band density of about the same as an Esterase A in D. melanogaster and a D indicates that the band is appreciably darker. This system is subject to some error from electro­phoretic variants since only one third of the mobility range is considered. The University of Texas Publication TABLE 2 Summary of species differences in subgenera and groups regarding Esterase A types* No. of species with No. of esterase A type: species examined in group L-L L-D D-L D-D Subgenus Sophophora obscura group 9 4 0 4 1 saltans group 13 12 0 1 0 melan:igast2r group 11 9 0 0 2 willistoni group 4 4 0 0 0 Total 37 29 0 5 3 Subgenus Pholadoris (small groups) Total 3 3 0 0 0 Subgenus Drosophila quinaria section tripunctata group 5 5 0 0 0 cardini group 6 6 0 0 0 quinaria group 7 3.5 2.5 0 Total 18 14.5 2.5 0 virilis-repleta section virilis group 4 2.6 1 0 0.4 repleta group 8 0 0 8 0 Total 12 2.6 8 0.4 • The first letter in the hyphenated A·type designations refers to the density of the Esterase A band in an ANA stain containing alcohol and the second num ber refers to the density of the band when no alcohol is present. A letter L indicates a staining intensity similar to the D. melanogaster enzyme and a letter Dan intensity which is appreciably greater. Some exceptions to the trend of similarity to D_ melanogaster are apparent. About one half of the species examined in the obscura group and the quinaria group showed higher activity than D. melanogaster when alcohol was present. In the virilis-repleta section of the subgenus Drosophila all species examined from the repleta group showed the D-L type, while none of the species in the virilis group showed such a pattern. If the sample of species from these groups is rep­resentative, this difference appears to be quite "section specific" in the subgenus Drosophila. The A esterases in D. melanogaster, in addition to being stimulated by alcohol, show differences in activity among larval, pupal and adult stages. The pupal stage shows highest Esterase A activity and much of this activity remains in the empty pupa case when the adult emerges (Johnson, unpublished data). Should metab­olic-developmental control mechanisms be involved, this system should prove interesting in further species comparisons. The ANA-BNA specificity differences are summarized in Table 3. Among subgenera the pooled distribution in the four specificity classes is clearly differ­ent. The subgenus Sophophora contains mostly species with a predominantely f3 pattern. It is contrasted with the quinaria section of the subgenus Drosophila which is high in a class species and with the virilis-repleta section which is high in the a-/3 category. The small sample from the subgenus Pholadoris contained one species with very little esterase activity other than the nearly ubiquitous band in common mobility with Esterase E of D. melanogaster. One species in the Johnson et al.: Drosophila Esterases and Taxonomy TABLE 3 Summary of species differences in subgenera and groups regarding major esterases specificity No. of species No. of with major est erases specificity species examined in group a a/3 {3 no est. Subgenus Sophophora obscura group 9 0 4 5 0 saltans group 13 0 1 12 0 melanogaster group 11 3 8 0 0 willistoni group 4 0 0 3 Total 37 3 13 20 Subgenus Pholadoris (small groups) Total 3 0 0 2 Subgenus Drosophila quinaria section tripunctata group 5 5 0 0 0 cardini group 6 2.5 1.5 2 0 quinaria group 7 7 0 0 0 Total 18 14.5 1.5 2 0 virilis-repleta section virilis group 4 0 4 0 0 repleta group 8 2 4.5 1.5 0 Total 12 2 8.5 1.5 0 willistoni group of the subgenus Sophophora showed a similar lack of activity. All groups have at least one class of esterase pattern unfilled. The inhibition characteristic with alcohol are presented in summary form in Table 4. All species in the subgenus Sophophora except one (nebulosa) show {3­esterase inhibition. In the Pholadoris subgenus none of the three species tested show inhibition. Similarly, in the quinaria section of the subgenus Drosophila several species are lacking in the {3-esterase and consequently show no inhibition. Thus the esterase inhibition is not completely independent from the major ester­ase classification in Table 3. The species in the virilis-repleta section all show some {3-esterase specificity but, with one exception, fail to show appreciable es­terase inhibition. The lack of esterase inhibition combined with major esterase pattern characteristics appears to separate the two sections in the Drosophila subgenus remarkably well. The repleta group contains the exceptional species, D. pachea, which shows a {3-esterase which is alcohol inhibited. This species is atypical of all other Drosophila in its ecology (Heed and Kircher, 1964), being the only Drosophila which breeds in stems of the senita cactus. Furthermore, it bears an anatomical resemblance to the phylogenetically older nannoptera species group (Throckmorton, 1962) and that species was previously assigned to the sub­genus Sophophora (Wheeler, 1949). D. pachea would be an acceptable candidate for the Sophophora subgenus on the basis of its esterase pattern. DISCUSSION In spite of the small number of samples examined from each strain, many The University of Texas Publication TABLE 4 Summary of species differences in subgenera and groups regarding alcohol inhibition of major esterases No. of species examined in group No. of species with inhibition of: a {3 none Subgenus Sophophora obscura group 9 0 9 0 saltans group 13 0 13 0 melanogaster group 11 0 11 0 willistoni group 4 0 3 Total 37 0 36 Subgenus Pholadoris (small groups) Total 3 0 0 3 Subgenus Drosophila quinaria section tripunctata group 5 z z cardini group 6 0 3.5 Z.5 quinaria group 7 0 5 z Total 18 10.5 6.5 virilis-repleta section virilis group 4 0 0 4 repleta group 8 0 7 Total 12 0 11 polymorphisms were found (Table 1) . Both the a-and the ,B-esterases were poly­morphic in the same strain. It is apparent that extensive polymorphism persists in the very old laboratory cultures, some of which were of single female origin. In a number of species there were both intrastrain and interstrain polymorphisms, some occurring in the same esterase system. Consequently, there is likely to be extensive multiple allelism in esterase polymorphisms, which can further con­tribute to possible confusion in interpretation based solely on electrophoretic vari­ation. Despite one's inclination to compare polymorphisms in various groups, it is impossible to attach any taxonomic significance to apparent trends because of the many confounded variables, such as original culture size, numbers of strains per species tested, differences in collections, etc. The electrophoretic variation detected is probably a conservative estimate of total protein heterogeneity. The data in this investigation indicate that 54 out of 114 strains are polymorphic in one or more esterases. For comparisons based only on mo'bility, this would indicate that a rather large number of individuals from more species should be examined in order to obtain a representative sample, greatly increasing the time and expense of the assay. Furthermore, detailed molecular analysis becomes impractical on the proteins of single small organisms (Drosophila-size) so that mass pooling of individuals would be required. After random pooling, however, much of the variability can no longer be detected and variants cannot be properly weighted in the determination of the characteristics of the species. As a result of the prevalence of polymorphisms found in this study, one would expect that mobility comparisons of proteins, with no known homology, would be Johnson et al.: Drosophila Esterases and Taxonomy of little use as a taxonomic trait. When other properties are considered along with mobility, the comparison becomes better since one is another step closer to estab­lishing genetic homology. However, homology is still far from being established and mobility difference in any practical system still has a chance of being am­biguous. Previous works in biochemical Drosophila taxonomy have been carried out under the potential handicaps inherent in the requirement for multiple fly samples and have relied heavily on mobility differences. For the most part these studies measured composite properties of a species. Whether or not known biases have affected the average sufficiently to mislead the reconstruction of Drosophila phylogeny is unknown, but in general the results probably can be trusted. As shown by the extensive comparisons with chromatograms of fluorescing com­pounds (Throckmorton, 1962) and with protein banding patterns in acrylamide gels (Hubby and Throckmorton, 1965), there exists considerable agreement with taxonomic relations proposed on the basis of a host of morphological and genetic data. However, further refinements should improve the agreement, especially when a number of systems are compared. The present communication compared only esterases in Drosophila species groups without relying solely on mobility differences or resorting to mass homo­genates. Thus multiple enzyme forms resulting from polymorphism did not enter directly into consideration of phylogenies, but was considered in establishing a meaningful scoring system. Some bias still remains. Only laboratory cultures of the various species were examined and most of them have been maintained for several years. Many of the cultures were started from a small initial collection and all were subject to changes caused by mutation, selection and drift. This paper was not intended to be a comprehensive survey of species differences, but to introduce one method by which such a survey can be initiated and to examine some of the interpretative complications of using the method. A large number of additional isozyme systems are available for similar study; twenty or more ad­ditional enzyme systems presently are known to be available. This extension is in progress as well as plans for the analysis of individuals taken directly from nature. With an increasing number of systems to consider, a new and expanding spectrum of biochemical characteristics promises to materially supplement the cytological and morphological ones long used and, consequently, improve taxo­nomic resolution. SUMMARY The esterases from several species in the genus Drosophila were studied from the point of view of substrate specificity and alcohol inhibition-activation charac­teristics after separation by starch gel electrophoresis. The observed differences between species were correlated with previously established taxonomic categories. Interesting and suggestive differences differentiating subgenera and even species groups are shown in summary Tables 2, 3 and 4. The advantages of the isozyme comparisons in combination with discriminatory criteria in addition to mobility alone are discussed, as are the advantages of single fly homogenates vs. mass homogenates. Some complications of interpreting isozyme polymorphisms for phylogenetic studies are mentioned. The University of Texas Publication ACKNOWLEDGMENTS The authors acknowledge with thanks the constructive criticism of Dr. H. E. Sutton during the preparation of the manuscript and the technical assistance of Misses Cynthia Greer, Susan Rockwood and Barbara Carroll. LITERATURE CITED Anfinsen, C. B., S. E. G. Aquist, J. P. Cooke, and B. Jonsson. 1959. A comparative study of the structure of bovine and ovine pancreatic ribonuclease. J. Biol. Chem. 234: 1118. Beckman, L. 1963. Serum protein variations in primates. Symp. Zool. Soc. Lond. 10: 251. Beckman, L., and F. M . Johnson. 1964. Esterase variations in Drosophila melanogaster. Hereditas 51: 212. Heed, W. B., and H. W. Kircher. 1964. Unique sterol in the ecology and nutrition of Dro­sophila pachea. Science 149: 758. Hoyer, B. H., B. J. McCarthy, and E. T. Bolton. 1964. A molecular approach in the syste­matics of higher organisms. Science 144: 959. Hubby, J. L., and L. H. Throckmorton. 1965. Protein differences in Drosophila. II. Compara­tive species genetics anphila) carinata x Nearly cosmopolitan 24. D. (D.) polychaeta x Texas, Hawaii, Carolines, Marianas, Central and South America 25. D. (D.) nasuta x x x x Widespread in Pacific 26. D. (D.) nasutoides x x x x Solomons 27. D. (D.?) upoluae x 28. Leucophenga samoaensis x 29. Liodrosophila pallidipennis x x Niue 30. Lissocephala versicolor x x x x 31. Microdrosophila convergens x x x x Fiji, Niue, Carolines, Marshalls 32. M. laticlavia* x (M. suvae)* Fiji 33. Mycodrosophila buxtoni x x 34. M. gratiosa x x x Widespread in Pacific 35. M. nigrithorax x x 36. M. recula* x 37. Neotanygastrella samoana* x 38. Paramycodrosophila bimaculata x 39. P. pictifrons x x 40. Samoaia attenuata* x 41. S. comma x x 42. S. hirta x x 43. S. mallochi• x x 44. S. leonensis* x 45. S.nuda x x x 46. S. ocellaris x x 47. Scaptomyza (Bunostoma) bicolor x 48. Zygothrica samoaensis x x x • Described as new. The University of Texas Publication considerable doubt about the identification since there appears to be a complex of seminigra-like fonns in the southeastern Asian area. 7. Drosophila albifrontata Malloch. Male genitalia: Fig. 4. 1-. 6; 17. 2. Mal­loch reported a single female from Malololelei, Upolu. We have 18 specimens from Malololelei and Afiamalu, Upolu. 8. Drosophila fiavipes (Harrison), NEW COMBINATION, for Liodrosophila fiavipes Harrison 1954: 113. Male genitalia: Fig. 4. 7-. 12; 17. 5-. 6. The name fiavipes has been used in Drosophila but it is not nomenclaturally preoccupied in that genus (see Wheeler, 1959: 190). The genus Liodrosophila is very poorly known, but judging from the seven named species which we have seen previously, fiavipes is much too large, it has well developed postvertical bristles, a high costal index, and is not nearly as shiny as the typical species. In addition, the large frontal triangle is formed differently, being widely separated by dull area from the shiny orbits. Harrison described fiavipes from a male from Malololelei, Upolu. We have two specimens, from the Afiamalu road, Upolv. 9. Drosophila nigrifrons Malloch. Male genitalia: Fig. 5. 1-. 6; 17. 3-. 4. Mal­loch's specimens came from Savaii and Upolui Harrison reported it from Upolu and Manono Island. We have nine specimens from Malololelei and Afiamalu, Upolu. 10. Drosophila pleurovittata Harrison. This species is known only by the holo­type female from Malololelei, U polu. 11. Drosophila stramineipes (Malloch), NEW COMBINATION, for Scapto­myza stramineipes Malloch 1934: 295. Male genitalia: Fig. 5. 7-. 12. Malloch had only a male from Savaii; we have 16 specimens from the Leone area, Tutuila. A study of the male genitalia shows clearly that this species is closely related to albifrontata, fiavipes and nigrifrons. We are therefore placing it with them in Drosophila although some key characters are unusual in that genus. 12. Drosophila nublada new species. Male genitalia: Fig. 6. 1-. 6; 17. 7. ~, !i?. Front tan, rather narrow; orbitals spaced well apart, the middle orbital about half length other two; smaller frontal hairs rather numerous. Antennae tan; arista with three dorsal and two ventral branches basal to the terminal fork. Face tan, narrow; carina prominent, rounded; cheeks pale, very narrow, making the eye look quite large; pile of eyes short. Vibrissa single; palpi tan. Mesonotum dark tan; on some specimens a faint pattern can be seen, usually as a pair of darker stripes near midline and as splotches in front of and behind the transverse suture. Prescutellars strong; basal scutellars divergent and long. Pleura yellowish tan; three strong sternopleurals; three humerals. Legs all pale. Halteres pale. Abdomen pale tan, without a defined pattern. Wings rather dark, especially anteriorly. Costal index about 3. 5; 4th vein index about 1. 6. Third costal sec­ tion with small black bristles on the basal 2/ 3. Body length about 3.0 mm., but most specimens appear to be about 2. 0 mm. due to body bending. Relationship: belongs to the subgenus Scaptodrosophila (= Pholadoris). Holotype male, allotype female, 15 paratypes and 13 additional specimens Wheeler and Kambysellis: Drosophilidae of Samoa from the Leone area, Tutuila Island, Samoa, July 27-August 5, 1962, M. R. Wheeler; one specimen from Malololelei and two from Afiamalu, Upolu Island, Samoa, July 22-23, 1964, M. R. Wheeler and D. G. Futch. 13. Drosophila samoaensis Harrison. Male genitalia: Fig. 6. 7-. 12. Harrison collected four males on Manono Island; we have a single male from the Leone area, Tutuila. The species is closely related to anuda Curran from the Solomon Islands. 14. Drosophila fuscovittata Harrison. Male genitalia: Fig. 7. 1-. 6; 17. 8. The three original specimens came from Manono Island and Vailima, Upolu. We have 39 specimens from Pago Pago and the Leone area, Tutuila, all collected from flowers. In addition we have two specimens from Suva, Fiji Islands. 15. Drosophila marjoryae Harrison. Male genitalia: Fig. 7. 7-. 12; 17. 9-. 10. This was described from four females from Manono Island. On the holotype and a paratype, borrowed from Dr. Harrison for study, the frons was strongly whitish pruinose when viewed from certain oblique angles. This is also true on 12 speci­mens from Niue Island and one from Nukualofa, Tonga Islands, which we be­lieve to be correctly identified as marjoryae. 16. Drosophila bryani Malloch. Male genitalia: Wheeler and Takada, 1964, Fig. 6. This species is widespread in the Pacific region (see Wheeler and Takada, 1964). It is common on all the Samoan Islands. 17. Drosophila convexa Malloch. This species is known only from the original holotype and paratype males from Malololelei, Upolu. 18. Drosophila excepta Malloch. Male genitalia: Fig. 8. 1-. 6. The species was described from a female from Malololelei, Upolu. We have four specimens, from Pago Pago and Leone, Tutuila, and Afiamalu, Upolu. 19. Drosophila ananassae Doleschall. Male genitalia: Wheeler and Takada, 1964, Fig. 2. This widespread, circumtropical species is the dominant species of Drosophila in Samoa, and is known from all islands of the group. However, Futch (this Bulletin) has found that two color phases occur in Samoa, the darker one being con-specific with ananassae from other areas while the lighter form appears to be an undescribed sibling species. 20. Drosophila bipectinata Duda. Male genitalia: Okada, 1954, Pl. 3, Fig. 14. Previous records are Formosa, India, Nepal and Japan. We collected several hundred individuals from bait and garbage at Pago Pago, Taputimu, and Leone area, Tutuila, and at Nafanua, Upolu. 21. Drosophila kikkawai Burla. Male genitalia: Wheeler and Takada, 1964, Fig. 3. This is a widespread, circumtropical species. Malloch reported it (as montium de Meijere) from Apia and Malololelei, Upolu. 22. Drosophila melanogaster Meigen. Male genitalia: Wheeler and Takada, 1964, Fig. 2. Malloch reported specimens from Apia, Upolu. We collected a small number of individuals at garbage and at banana bait at Pago Pago and Taputimu, Tutuila. 23. Drosophila carinata Grimshaw. Male genitalia: Wasserman, 1962, Fig. 2. D. mercatorum Patterson and Wheeler, 1942, is a synonym (see Hardy 1965: 204). We collected 25 to 30 specimens at banana baits near Leone, Tutuila. In the same local area were various grocery items, including fresh vegetables, from The University of Texas Publication Hawaii. Since this species occurs in Hawaii we suspect that the Samoan flies were recently introduced. 24. Drosophila polychaeta Patterson and Wheeler. Male genitalia: Wheeler and Takada, 1964, Fig. 4. We captured a single female at banana bait at Taputi­mu Farm, Tutuila. The species has been found sporadically at many isolated points, e.g., Texas, Hawaii, Caroline and S. Mariana Islands, Central and South America. 25. Drosophila nasuta Lamb. Male genitalia: Wheeler and Takada, 1964, Fig. 5. This is a widespread, circumtropical species. It is found on all of the Sa­moan Islands and frequently forms very large local populations. 26. Drosophila nasutoides Okada (replacement name for hypopygialis Mal­loch, preocc.). Male genitalia: Fig. 8. 7.-10; 17. 11-. 13. This species is rela­tively common in Samoa and is known from all the islands. We also have specimens from Guadalcanal and New Georgia (Solomon Is. ) . 27. Drosophila upoluae Malloch. This species has not been seen since Mal­loch's description which was based on a female from Apia, Upolu. 28. Leucophenga samoaensis Harrison. This species is known only from a single male from Manono Island. 29. Liodrosophila pallidipennis (Harrison), NEW COMBINATION, for Lis­socephala pallidipennis Harrison 1954: 113. Male genitalia: Fig. 9. 1-. 6. As was stated above with respect to Drosophila fiavipes the genus Liodrosophila is still poorly understood, but this species seems to be a fairly typical member of it. Harrison had one specimen from Vailima, Upolu. We have ten specimens, from the Leone area, Tutuila, Malololelei, Upolu, and Niue Island. 30. Lissocephala versicolor Malloch. Male genitalia: Fig. 9. 7-. 12; 17. 14. Malloch reported the species from Upolu and Savaii; Harrison found it also on Manono Island. We collected 46 specimens from Pago Pago and the Leone area, Tutuila, and Apia, Upolu. 31. Microdrosophila convergens (Malloch); = Microdrosophila errator Wheeler and Takada, 1964; 215, NEW SYN. Male genitalia: Wheeler and Ta­kada, 1964, Fig. 19 as errator. Malloch reported five specimens, from Vailima, Upolu and Safune, Savaii, as well as some from Viti Levu, Fiji. Harrison added Manono Island and Malololelei, Upolu. We have collected 44 specimens from Pago Pago and Leone, Tutuila, as well as 25 specimens from Suva, Fiji and 12 from the Island of Niue. As errator, the species is also known from Kusaie and Ponape (Caroline Islands) , and from Majuro, Arno and Namorik atolls (Mar­shall Islanus) . 32. Microdrosophila laticlavia new species. Male genitalia: Fig. 10.1-.5. 3, 'i' . Frons with pale yellow orbits, the large central triangle much browner; proclinate orbital situated well in front of the posterior reclinate and about 2/3 as long as the latter; middle orbital extremely reduced. Second antennal segment brown, third pale except at base of arista; aristal branches usually eight dorsal and three ventral basal to the terminal fork. Face pale yellow and distinctly cari­nate; cheeks pale, very narrow; palpi dark brown to black. Mesonotum yellowish with a broad, brown, median stripe (Fig. 10. 6). Pleura with a broad brownish stripe above, all pale below. Legs all pale. Halteres brown. Wheeler and Kambysellis: Drosophilidae of Samoa Basal scutellars about 1/ 3 length apicals. Abdomen dull dark brown, the genital area pale. Wings hyaline; costal index about 1. 0 to 1. 1; 4th vein index about 4.0. Third costal section with the small black bristles on the basal 9/ 10. Body length about 1.8 mm. Holotype male, allotype female, and 11 paratypes, Leone area, Tutuila Island, American Samoa, July 27-August5, 1962, M. R. Wheeler. There is a very similar appearing species in the Fiji Islands, but it is clearly different as determined by the study of the male genitalia. We are taking this opportunity to name this Fijian form. Microdrosophila suvae new species. Male genitalia: Fig. 10.7-.11; 17.15-.17. ~, ~.Very similar in morphology to M . laticlavia, described above. The front is not so brown centrally; the brown mesonotal stripe is gradually paler anteriorly and is quite noticeably separated into right and left portions anteriorly. The 4th vein index is about 4.5. The sixth sternite of the male is quite unusual and is shown in Figure 10.11. Holotype male, allotype female, and 5 paratypes, Suva, Fiji Islands, July, 1962, M. R. Wheeler. 33. Mycodrosophila buxtoni Malloch. Male genitalia: Fig. 11. 1-. 6. The de­scription was based on a female from Malololelei, Upolu. We have 24 specimens, from Pago Pago and Leone area, Tutuila. The male antennae are much paler than those of the female; the haltere knob is discolored on both sexes. 34. Mycodrosophila gratiosa (de Meijere). Male genitalia: Wheeler and Ta­kada, 1964, Fig. 12. This is widespread throughout Asia and the Pacific region, being previously reported from Indonesia, Fiji, Solomon Islands, and Micronesia from Guam to Kusaie. Malloch had specimens from Upolu and Savaii. We have 27 specimens from Pago Pago and the Leone area, Tutuila. 35. Mycodrosophila nigrithorax Malloch. This species is known only from Malloch's two specimens from Malololelei, Upolu and Salailua, Savaii. 36. Mycodrosophila recula new species. Male genitalia: Fig. 11. 7-. 12. ~, ~. Front of head dull black, appearing whitish anteriorly when seen from cer­tain angles; proclinate and posterior reclinate orbitals subequal, middle orbital minute. Antennae dark brown; arista with four branches above and one below basal to the terminal fork. Face dark brown, cheeks and clypeus darker, almost black. Proboscis pale yellow, palpi dark brown. Mesonotum shiny black; acrostichal hairs in about ten rows; scutellum velvety black; basal scutellars about 1/ 4 length apicals. Pleura dark above, and the meso­pleura with a broad dark stripe which just fails to reach the base of the front coxa; rest of pleura pale. Legs pale, with a little discoloration apically on femora; male fore tarsi without long hairs. Halteres dark. Male genital area and last two tergites pale yellow; preceding four tergites with transverse dark apical bands which are only a little wider in midline; all bands bend anteriorly at extreme lateral margin. Abdominal pattern of female similar to that of the male but there is even less widening of the bands medianly. Wing blade pale; distal costal break deep, forming a large black lappet; base of 6th vein a little darkened. Costal index about 1. 2; 4th vein index about 3. 5. Third costal section with small black bristles on the basal 2/3. The University of Texas Publication Body length about 1. 8 mm. Holotype male and paratype male, Leone area, Tutuila Island, Samoa, July 27-August 5, 1962, M. R. Wheeler. 37. Neotanygastrella samoana, new species. Male genitalia: Fig. 12. 1-. 6. i!J, ~.This new species is quite similar to N. pacifica Wheeler and Takada (1964: 212) from Micronesia (Ponape, Kusaie and Truk). The mesonotum is light brown rather than tan, and the front is also darker brown. The face, cheeks and clypeus are brown (very pale in pacifica) , and the palpi are tan and not espe­cially darker at the apex. The arista has only three dorsal branches. The pleura is darker brown than in pacifica and the legs are all brownish; in pacifica the fore femora and tibiae are black. The wing vein indices are the same as those in pa­cifica but the blade is much darker. This species is similar to, but not the same as N. sp. A of Wheeler and Takada, 1964: 212 from Ponape, Caroline Islands. Holotype male and allotype female, Afiamalu Road, Upolu, Samoa, July 22­23, 1964, M. R. Wheeler and D. G. Futch. 38. Paramycodrosophila bimaculata (Malloch) . This was described in the genus Upolumyia Malloch which is a synonym of Paramycodrosophila (see Wheeler and Takada, 1964: 206). The species has not been re-collected, being known only from the four original specimens from Salailua, Savaii. 39. Paramycodrosophila pictifrons (Malloch). The holotype came from Vail­ima, Upolu, and the allotype and two paratypes were from Salailua, Savaii. This may be the same species as P. pictula (de Meijere), a widespread species known from Indonesia, Taiwan, the Caroline Islands and the Fiji Islands. There are, however, a few discrepancies between Malloch's description and specimens from Micronesia and Fiji. 40. Samoaia attenuata new species. Male genitalia: Fig. 12. 7-. 11. i!J, ~ . This species is extremely similar in external morphology to nuda, and the only fea­tures which definitely distinguish the two are found in the male genitalia. How­ever, most specimens can be sorted correctly on the basis of the color pattern of the mesonotum (Fig. 12. 12); on attenuata the dark stripes are much blacker and more clearly delimited, while on nuda they tend to be a lighter brown and some­what poorly defined. In nuda the scutellum is more regularly pointed apically, while in most specimens of attenuata, there is a fairly distinct point at which the apex becomes more abruptly pointed. Holotype male, allotype female, and 15 paratypes, all descendants of one orig­inal female collected along the Afiamalu Road, Upolu, Western Samoa, July, 1965, W. S. Stone and C. P. Oliver, collectors. The species has been maintained in laboratory culture since its collection by using the special culture medium described by Wheeler and Clayton, 1965 (Dro­sophila Information Service, 40: 98). 41. Samoaia comma Malloch. Male genitalia: Fig. 13. 1-. 5; 18. 1-. 5. This species was described from eight specimens from Malololelei and Vaea, Upolu, and Salailua, Savaii. We have one specimen from Malololelei and one from Afia­malu Road, Upolu. 42. Samoa.ia hirta Malloch. Male genitalia: Fig. 13. 6-. 10. Malloch's speci­ Wheeler and Kambysellis: Drosophilidae of Samoa mens were from Malololelei, Upolu, and Salailua, Savaii. Harrison captured one at Malololelei. We collected 15 individuals from the Afiamalu Road region, Upolu. The sexual dimorphism in pleural coloration is quite prominent on our specimens. 43. Samoaia mallochi new species. Male genitalia: Fig. 14. 1-. 5. In his de­scription of nuda, Malloch mentioned a paratype which was badly "abraded" and which he thought might not belong to nuda. We have found the same form, and are pleased to honor the famous dipterist by naming the species for him. &, 'i?. Frons tan, the ocellar area black and with a brownish mark on the orbit near the posterior reclinate orbital; lunular area brown. Head bristles rather short and stubby (as compared with those of other species of this genus); anter­ior reclinate and proclinate orbitals unusually thin. Antennae tan to brown; arista with five to seven dorsal and two ventral branches in addition to the termi­nal fork. Face with a large brownish carina, paler in foveae but with a dark brown spot near vibrissa; cheeks pale and with two light brown marks: one just below the eye and one more posteriorly. Clypeus as brown as carina; palpi very pale yellow; eyes with coarse pile. Mesonotal bristles relatively short and stubby; anterior dorsocentral much shorter and thinner than posterior; no evident presutural dorsocentral. Acrosti­chal hairs rather short and sparse. The total visual effect of small, stubby bristles and sparse, thin hairs is that the thorax has been badly rubbed-just the impres­sion that Malloch had when viewing his one specimen. Mesonotum dark tan with a pattern of dark brown stripes: the midline is pale; there is a broad stripe on each side which ends abruptly posteriorly, leaving a pale area between the two dorsocentrals of each side; laterally there are broken brown stripes and blotches. Scutellum dark except for a pale stripe in the midline and paler dorso-lateral margins. Pleura and legs pale with a complex pattern of brown spots; the pattern is almost that of nuda but the spots are larger and darker. Abdomen mostly dull brownish black, but each tergite has a pale tan area laterally which is completely bordered by the dark color. Wings somewhat variable, dark brown with irregular paler patches on each side of both crossveins; the 2nd longitudinal vein bends away from the costa before its apex, then turns suddenly back to the costa, form­ing a short spur vein at the angle of the turning. Costal index about 2. O; 4th vein index about 4. 1. Third costal section with the small black bristles on the basal 2/ 3. Body length about 5. 0 mm. Holotype male, allotype female and two paratype females, Malololelei, Upolu, Western Samoa, July 22-23, 1964, M. R. Wheeler and D. G. Futch. 44. Samoaia leonensis new species. Male genitalia: Fig. 14. 6-. 10. & , 'i?. This species is extremely similar in external morphology to ocellaris Malloch; in ad­dition to the clear differences in the male genitalia of the two, there is also a simple difference in the color pattern of the cheek. In ocellaris there is a definite black spot at the vibrissal base, and a large, rather extensive dark brownish area on the cheek below the eye, extending far back to the rear angle of the cheek. In leonensis, the vibrissal spot is about the same, but just below the eye there is only a small, pale brownish spot, the rear cheek angle being wholly pale. Other traits, The University of Texas Publication especially in wing pattern, are probably present, but the pattern is too variable to make this reliable. r ofotype m1le, allotype female, and 15 paratypes, as well as 20 additional specimens, from the Leone area, Tutuila Island, Samoa, July 27-August 5, 1962, M. R. Wheeler. 45. Samoaia nuda Malloch. Male genitalia: Fig. 15. 1-. 6; 18.6 .9. The holo­type ( 'i' ) came from Salailua, Savaii. Malloch mentioned a second specimen from the same locality which he thought might belong to this species but it was con­sidered to be damaged by abrasion. We have described the latter form, above, as S. mallochi. We have collected 18 specimens of nuda, from Leone, Tutuila, and Afiamalu and Malololelei, Upolu. 46. Samoaia ocellaris Malloch. Male genitalia: Fig. 15.7-. 11. Malloch's speci­mens came from Apia and Malololelei, Upolu, and Salailua, Savaii. Harrison collected several individuals at Vailima and Malololelei. We have nine speci­mens, from Malololelei and Afiamalu, Upolu. 47. Scaptomyza bicolor Malloch. Male genitalia: Fig. 16.1-. 6; 18. 10-. 13. This species has not been reported since Malloch's description, based on a male from Malololelei, Upolu. We collected a single male from Afiamalu, Upolu. The general features, and more especially, the features of the male genitalia, show clearly that it belongs to the subgenus Bunostoma, a group which is very well represented in Hawaii. 48. Zygothrica samoaensis Malloch. Male genitalia: Fig. 16. 7-. 12; 18. 14. This was described from 21 specimens from Malololelei, Upolu, and Safune, Savaii. Harrison found one at Vailima, Upolu. We have 14 specimens, from Iliili and Leone, Tutuila, and Afiamalu and Malololelei, Upolu. D1scuss10N It may be premature to discuss the phylogeny of the Samoan species but re­marks concerning some of them seem desirable. Our remarks will be centered around the photographs of portions of the male genitalia shown in Figures 17 and 18. These particular photographs were chosen instead of drawings since they show rather clearly the points which we wish to emphasize. Figure 17.1, Drosophila (Hirtodrosophila) unicolor. The ventral view shows the long extensions of the genital arch (a) which articulate with the hypandrium (b). These structures have been found so far in every Pacific species of the sub­genus Hirtodrosophila (see, for example, Wheeler and Takada, 1964, Fig. 10 a, e, g) , but they have not been seen in Oriental species (judging from Figs. 43-50 in Okada, 1956) nor in any North American species. We cannot say, however, that the Pacific species form a distinct species group. Figure 17.2, Drosophila albifrontata (ventral view); 17.3-.4, D. nigrifrons (dorsal and ventral views); 17.5-.6, D. fiavipes (lateral and ventral views) . These three species, along with stramineipes, form one of the two distinctive en­demic Samoan groups. Judging from the description, D. pleurovittata may also belong here, and, with less reason, D. convexa since Malloch placed it close to the above. Although the male genitalia of the four studied species are very similar Wheeler and Kambysellis: Drosophilidae of Samoa (almost identical, in fact) the species have diverged sufficiently in other char­acters that Malloch included stramineipes in Scaptomyza while Harrison placed flavipes in Liodrosophila! We consider this complex to represent a new species group (the albifrontata species group) of the subgenus Scaptodrosophila, endemic to Samoa, and having the following male genital characteristics: hypandrium elliptical or nearly so, with a pair of small bristles (c) ventrally; dorsal part of hypandrium (d) re­duced to a great extent; anterior gonapophyses ( e) nearly as long as the penis, with small sensilla (f) laterally; posterior gonapophyses (g) smaller than the anterior; penis consisting of two symmetrical halves (h) connected with each other by a membraneous structure; apodeme of the penis (i) rodlike and flattened laterally. Claspers (k) greatly reduced, with a distal row of bristlelike teeth, and with a large prolongation posteriorly (1); bridge (m) with a characteristic shape (see drawings), connected to the claspers by a membraneous structure (n). The articulation points of the external and internal genitalia are located approxi­mately in the middle part of the hypandrium (o) (see also Figures 4, 5) . Figure 17.7, Drosophila nublada. The lateral view shows the small gonapo­physes (p), curved penis (q), and the unusual hypandrial shape (r), character­istic for samoaensis as well as nublada. Figure 17.8, Drosophila fuscovittata. A lateral view of the internal genitalia showing the winglike extension of the penis ( s) . Figure 17.9-.10, Drosophila marjoryae. The arrangement of the teeth on the claspers is shown ( .9 t) and the winglike extension of the penis ( .10 s). Figure 17.11-.13, Drosophila nasutoides. This shows the characteristic mem­ braneous structure (v) and the hooklike apophyses (w) of the penis (dorsal, ven­ tral, and lateral views) . Figure 17.14, Lissocephala versicolor. The shape of the internal genitalia seems to be characteristic for the genus (see also Fig. 9.7-.11 and Wheeler and Takada 1964, Fig. 20). Figure 17.15-.17, Microdrosophila suvae. The male genitalia are modified to such an extent that homology with the typical structures of other genera is quite uncertain. The hypandrium is greatly reduced, the remaining parts probably rep­resenting a growth of the outer border (a), the distal bow (b), and the para­median expansions ( c) with the paramedian spines ( d). The penis is less modi­fied, showing clearly the apodeme (e) and a structure (£) specialized for an ar­ticulation with the distal bow. The anal plates (g) are fused to the genital arch (in most species of the genus) and the claspers (h) are modified to a great ex­tent. Similar structures are seen in M. laticlavia (Fig. 10.1-.6) and in most of the Micronesian species of Microdrosophila (see Wheeler and Takada 1964, Fig. 19). Figure 18.1-.5, Samoaia comma; 18.6-.9, S. nuda. These two species represent very well the characters of this endemic Samoan genus. The internal genitalia of both species remind one of the subgenus Drosophila, repleta group. On the other hand, the external genitalia show an evolution of the genus into two groups: a group containing comma, hirta and attenuata, and another group containing ocellaris, nuda and leonensis. The seventh species, mallochi, has diverged con­siderably more but may be an offshoot of the attenuata group. On the basis of The University of Texas Publication general morphology, nuda and attenuata seem to be sibling species, so that their separation into different species groups on the basis of genitalia is rather sur­pnsmg. The fusion of the bridge and claspers (a), the presence of a membraneous structure (b) connecting the anal plates with the genital arch, and the long hairs on the genital arch (c), all seem to be generic characters, found in both groups. The position of the claspers (d) and the secondary (?) claspers (e) are probably group characters as can be seen in Figures 18.4, .5, .7, .8, and on the appropriate drawings in Figures 12.7to15.11. Figure 18.10-.13, Scaptomyza (Bunostoma) bicolor. This species shows a series of characters typical of the subgenus Bunostoma. There is a long, thin, laterally flattened penis (f) with a distal modification (g); a long extension of the novasternum (h), small elliptical gonapophyses ( i) with a pair of sensilla, and numerous long hairs on the external genitalia (k). Similar structures are present on the Hawaiian species of Bunostoma. Figure 18.14, Zygothrica samoaensis. A study of the male genitalia shows be­yond doubt that this is properly placed in Zygothrica. The most important ge­neric characters, discussed by Burla ( 1956), are the pointed prolongations ventrally from the anal plates (1), the bristles on the claspers (m), and a mem­braneous structure (n) connecting the claspers with the anal plates. TECHNIQUES Due to the great importance of the detailed structure of the male genitalia for taxonomic studies, an attempt was made to develop a series of reliable techniques for preparing illustrations of these structures as realistically as possible. Free­hand drawings must always be greatly influenced by the artistic ability of the investigator. The use of the camera lucida, often employed, has the advtantage of speeding up the drawing time and gives a rather realistic proportional outline of the structure, but the details must still be filled in by the artist. Moreover, with both methods, to re-examine a specimen one must make new observations of the mounted genitalia, and this is often inconvenient. As an alternative, the junior author has spent considerable effort in developing a photographic technique with which realistic pictures can be made, and from them, in turn, accurate drawings of the individual parts. Photography of the male genitalia posed two major problems. The male genital structures are suffi­ ciently thick that depth of focus is often a serious problem. Further, unless sev­ eral specimens are available, one must photograph a structure from several dif­ ferent views using a compound microscope with magnifications of 200 X to 300 X; this requires taking the photographs while the specimen is in liquid and without a cover glass. The method described by Komp (1942) for use in the taxo­ nomic study of mosquitoes has been very useful; in fact, several steps in our technique, especially the staining and mounting of specimens, have been used with scarcely any modification. For completeness, the methods used here to pre­ pare the illustrations are being presented in considerable detail (see Figure 1). l clear dissect acetic ll'I in --"""'•~ in EtOH I phenol creosote "" I./ Gages stain 95%EIOH "' S td JOO%EtOH ·-­ ....~ =~ ? relax specimen photographing slide i camera set-up microneedles dissection apparatus FIG. 1. Diagram of the steps required for photographic reproduction of male genital characters. A. Clearing and staining 1. When dry, pinned specimens were to be used, they were relaxed and soft­ened by placing them in a chamber with hot water for 5-10 minutes (Figure 1); with freshly killed material this step is not necessary. 2. Under the lower power of a dissecting microscope, the tip of the abdomen was cut off using a pair of No. 3 dissecting forceps. 3. The tip of the abdomen was placed in a test tube with 10% KOH; the re­mainder of the fly was returned to the collection, and was given an identifying number to match one used thereafter with the genitalia specimen. 4. The male genitalia were boiled in KOH for 3-5 minutes, the time varrying according to the size and degree of sclerotization of the species. 5. The specimen was removed from the KOH with a pipette and placed in a shell vial. The KOH was removed and replaced with acetic alcohol (3 parts 50% ethanol: 1 part glacial acetic acid). 6. After 15 minutes the acetic alcohol was removed with a clean pipette and replaced with dilute Gage's Stain ( 1 part stain : 5 parts distilled water) ; the The University of Texas Publication specimens were left in the stain for 12-24 hours. Gage's Stain has the following composition: acid fuchsin, 0.5 g; 10% HCl, 25.0 cc; distilled water, 300.0 cc. 7. The stain was removed and replaced with 95% ethanol for five minutes. 8. The alcohol was removed and replaced with absolute ethyl alcohol for about 15 minutes. 9. The absolute alcohol was replaced with glycerin (or the specimen was transferred to phenol; see below). Specimens may be left in glycerin for several weeks, but after 5-6 days the specimen becomes destained. The glyecrin step can be eliminated, and we have used it only when problems of overstaining or of storage were involved. 10. To complete the dissection, the specimen was transferred from the vial to a double-depression slide and placed in a drop of aqueous, liquid phenol. We found that the easiest way to transfer the specimen was to trap it with a small amount of glycerin with a pair of dissecting forceps, remove it from the vial, pick it up with a microneedle (see below) and place it in the phenol. It helps to carry along as little glycerin as possible. The specimen becomes cleared in the phenol in about five minutes. 11. After clearing, the specimen was removed to the other side of the depres­sion slide and placed in a drop of creosote (Beechwood). The specimen may re­main here for a long time if necessary (weeks to months) without altering its structure or color, but occasional drops of creosote must be added periodically to replace that lost by evaporation. B. Dissection All dissections were made in creosote using the low power (30-40 X) of a dis­secting microscope, and using transmitted light. The dissecting apparatus (Fig­ure 1) with a fitted work-top of ~-inch glass, combined with a light source and a pair of easily adjustable mirrors, was especially designed to use transmitted light. The large size gave ample space for resting the hands so that essentially perfect stability during dissection was obtained. A pair of microneedles (Figure 1), using "minuten nadeln" insect pins in­serted into a glass rod or into wood (a match stick for example), was used to sep­arate the genitalia from the rest of the tissues. Before the actual separation of the external and internal portions of the genitalia, a careful examination of the geni­tal complex was made under the higher power of the dissecting microscope ( 120 X) in order to understand the complex as a unit. Several photographs (usually a dorsal, a ventral and a lateral view) of the entire genital complex were taken with the compound microscope (see below); then separation of the ex­ternal from the internal genitalia was made with magnification of 70 X to 100 X. This is the most critical part of the dissection and extreme steadiness of the hands is required. The points of articulation were always located before any attempts at separation were made (see, for example, points a and bin Figures 2.1, 2.5, and 2.9). C. Photography Photographs were taken with the genitalia under the high-dry objective (ca. 450 X) of a compound microscope. To do this a special slide was prepared (Fig­ure 1) as follows: four small pieces of broken coverglass were glued with "Duco" household cement to a clean slide in such a way that a small rectangular cavity, several mm. in length and ± 0.5 mm. in width (always a little wider than the width of the male genitalia) was formed. Due to variations in the size of the genitalia of different species, it was convenient to prepare two or three slides of different widths of cavities. A very small amount of creosote was trapped with forceps and released into the cavity formed by the pieces of glass; the amount should be just enough to fill the cavity. The genitalia were transferred to the creosote with a microneedle and were oriented in the desired position. The slide was then placed, with great care, on the stage of the microscope to which was attached the photographic ap­paratus. In photographing, we used a 10 x objective, a 10 x ocular fitted into the camera adapter, and camera bellows for greater magnification. The camera was the Asahi Pentax, a single lens reflex type, with its lens re­moved. Light from an AO Spencer Ortho-Illuminator with a blue filter gave the best results. The sequences of steps followed in making the photographs is as follows: 1. The iris diaphragm of the substage condenser of the microscope was closed almost maximally. 2. The Field Diaphragm of the Illuminator (the diaphragm nearer the micro­scope) was turned to the right to close the opening until a clear view of the geni­talia, with a good depth of focus, was obtained. 3. We then focused on the level in which were were primarily interested (see below). 4. The Intensity Filter Turret of the Illuminator was placed in position No. 2; the result is a low light intensity-so low that it is difficult for the eye to see the details of the structures. 5. Exposure time was 5 sec. using a fine-grain black-and-white film such as Kodak Panatomic-X (F135) with an ASA rating of 32. Three pictures were usually taken for each structure: a dorsal, ventral, and lateral view. Development of the film was done in a fine-grain developer (for example, Kodak Microdol-X diluted 1: 3 with distilled water, with a development time of 13 min. at 21 °C). Prints were made with an enlarger on Kodak Polycon­trast Rapid Paper (FSW single weight type) using filters No. 2 and No. 3 of the Kodak Polycontrast Filter Kit, Model A. D. Preparation of Illustrations One complete set of photographs was kept as a permanent record, and a second set was used to prepare the drawings (see Figure 2.1-2.8). By comparing the actual specimen under the microscope with the photographs, we outlined the most significant features on the photograph with waterproof drawing ink (Fig­ures 2.5-2.8), using a Rapidograph pen, No. 00. The inked outline was then copied on Dietzen 198 drawing paper (Figures 2.9-2.12). The orientation of the structure being photographed is very important and The University of Texas Publication .3 .II .15 significantly influences the apparent morphology of the specimen. An extreme case (Drosophila ma.nonoensis ) is shown in Figure 2. In this species a part of the hypandrium is rolled between the apodeme of the penis and the rest of the hy­pandrium, as is shown in the lateral view; thus, a wide angle is formed between those two structures. As a result, one gets opposite impressions of the structure of the hypandrium when seen from other views (2.2 and 2.3). In a dorsal view (2.2), when the focus is on the apodeme, the hypandrium is quite out of focus so that in the drawings (2.10) it appears to be very small. On the other hand, in the ventral view (2.3) when we are focused on the hypandrium, the result is just the reverse (2.11). As a rule we have found it best to focus on the apodeme of the penis for the dorsal view, to focus on the hypandrium for the ventral view, and to focus on the penis for the best lateral view. It should be pointed out that ex­treme cases such as that of manonoensis are very rare; generally, the dorsal and ventral views are comparable. E. Permanent mounts When all of the necessary drawings and records had been made, the speci­mens were mounted on slides, and labelled with a number to match the one at­tached to remainder of the pinned specimen. In mounting, small drops of balsam were placed with a microneedle on a clean slide, and in each drop one separate genital structure was placed in an appropriate orientation. The slide was then kept for a week or two in a dustproof box to allow the drops to harden. Then a drop of balsam was added to the slide, filling· in the space between the original drops, and a coverglass was added. REFERENCES Burla, H. 1956. Die Drosophilidengattung Zygothrica und ihre Beziehung zur Drosophila­untergattung Hirtodrosophila. Mitt. Zool. Musewn, Berlin 32(2): 189-321. Hardy, D. E. 1965. Insects of Hawaii, Vol. 12. Diptera Cyclorrhapha II. Family Drosophili­dae. Honolulu, University of Hawaii, pp. 814. Harrison, R. A. 1954. Some notes on and additions to the Drosophilidae (Diptera) of Samoa and Fiji. Trans. R. Ent. Soc. Lond. 105: 97-116. Komp, W. H. W. 1942. A technique for staining, dissecting, and mounting the male termi­ nalia of mosquitoes. Public Health Repts. 57 (36): 1327-1333. Malloch, J. R. 1934. Insects of Samoa, Part VI, Fasc. 8: 267-312. Okada, T. 1954. Comparative morphology of the drosophilid flies. I. Phallic organs of the melanogaster group. Kontyu 22: 36-46. ----. 1956. Systematic study of Drosophilidae and allied families of Japan. Gihodo Co., Ltd., 183 pp. Wasserman, M. 1962. Cytological studies of the repleta group of the genus Drosophila: III. The mercatorum subgroup. Univ. Texas Puhl. 6205: 63-71. Wheeler, M. R. 1959. A nomenclatural study of the genus Drosophila. Univ. Texas Puhl. 5914: 181-205. ----, and H . Takada. 1964. Diptera: Drosophilidae. In Insects of Micronesia 14(6): 163-242. FIG. 2. Drosophila (Hirtodrosophila) manonoensis Harrison, showing the technique used to prepare drawings from photographs of the male genitalia. 2.1-.4 are photographs of the struc­tures: .1, ventral view of the internal and external genitalia; .2, dorsal, .3, ventral, and .4, lateral view of the internal genitalia. 2.5-.8 are the same major structures with their outlines inked in. 2.9-.12 are tracings of the inked outlines. 2.13, ventral view of the external genitalia; 2.14, lateral view of the right half of the external genitalia; 2.15, bridge. The University of Texas Publication 0 0 (.) co l2 c :::J C\J --: cil 0 FrG. 3.1-.6, male genitalia of D. innocua. 3.1, dorsal view, 3.Z, ventral view, and 3.3, lateral view of the internal genitalia. 3.4, ventral view of the external genitalia; 3.5, lateral view of the right half of the external genitalia; 3.6, bridge. The figures of the male genitalia of the remaining species are presented in this same sequence of drawings; 3.7-.1Z, b. unicolor. ~--=: :: ---------­ (j) 0 0-c e cg <.O N -_c 0 cil Fw. 4.1-.6, D. albifrontata; 4.7-.1Z, D. flavipes. The University of Texas Publication "! Fw. 5.1-.6, D. nigrifrons; 5.7-.12, D. stramineipes. gil C! ·a; c E c .... C/) - cil I{) C/) c -e .... C> c ci I "' c: "' Q) 0 0 E 0 "' cil 0 FIG. 6.1-.6, D. nublada; 6.7-.12, D. samoaensis. The University of Texas Publication (Jl > 0 0 I/) :J cil Fw. 7.1-.6, D. fuscovittata; 7.7-.12, D. marjoryae. Wheeler and Kambysellis: Drosophilidae of Samoa rn a.> "O 0 ~ rn 0 c: -: cil 0 -: I'-: Fm. 8.1-.6, D. excepta; 8.7-.12, D. nasutoides. The University of Texas Publication .... 83 C\J ..2 0 u IJ) co .... Q) > _jj _jl FIG. 9.1-.6, Liodrosophila pallidipennis; 9.7-.iZ, Lissocephala versicolor. 0 > 0 u -0 ~I FrG. 10.1-.6, Microdrosophila laticlavia; 10.7-.12, M . suvae. The University of Texas Publication c 0 -x :::i .0 ~I -: FIG. 11.1-.6, Mycodrosophila buxtoni; 11.7-.12, M. recula. r0 J l. ,' ' /.' ', ' l[) v D 0 c: C\J (,!) 0 0 E 0