HARVARD UNIVERSITY Library of the Museum of Comparative Zoology APR 3 0 1937 Bulletin of the Museum of Comparative 2ioblogr AT HARVARD COLLEGE Vol. 81, No. 1 NOTES ON THE ORNITHOLOGY OF TROPICAL EAST AFRICA By Herbert Friedmann and Arthur Loveridge CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM April, 1937 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the Bulletin, Vols. I to LXV, LXVI, No. 1 & 2, LXVII to LXXIX No. 1, 2, 3 & 4, and LXXX, No. 1, of the Memoirs, Vol. I to LIV No. 1, 2 & 3. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. 81, No. 1 NOTES ON THE ORNITHOLOGY OF TROPICAL EAST AFRICA By Herbert Friedmann and Arthur Loveridge CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM April, 1937 A\ C -z. - L. No. 1. — Notes on the Ornithology of Tropical East Africa By Herbert Friedma.nn and Arthur Loveridge INTRODUCTION The collection of birds reported on in this paper was made by A. Loveridge when residing in East Africa from October, 1915, until May, 1923. About 4,000 skins of some 800 species and subspecies were preserved. The preponderating majority of these specimens came from the northern half of Tanganyika Territory, but two natives, especially trained for the purpose, made collections in Bukoba, British Ruanda, Uganda and Kenya Colony, while further material was obtained dur- ing the closing stages of the East African Campaign in the vicinity of Lumbo, Mozambique. As many of the localities are scarcely known outside the countries in which they occur, a geographical list of them is attached for the convenience of investigators using this paper. Four years after Arthur Loveridge joined the staff of the Museum of Comparative Zoology, the bulk of the collection, comprising at least a pair of each form (where more than one specimen had been obtained) was acquired for the Museum of Comparative Zoology through the generosity of Dr. Thomas Barbour. Several hundred duplicate specimens were purchased by the East Africa and Uganda Natural History Society, the American Museum of Natural History and the Academy of Natural Sciences of Philadelphia; those from the latter institutions have been available when needed in the present study. However, only the specimens in the "main" collection in the Museum of Comparative Zoology have been listed under each species, the others being referred to only by localities. Unfortunately the locality records of some of the earlier duplicates were not retained before the skins were dispersed by donations to various museums in England and East Africa, but the data for some of the birds now in the Coryndon Memorial Museum in Nairobi are here included. In addition, a few birds collected by Loveridge while with the Smithsonian-Chrysler Expedition in 1926, and now in the U. S. National Museum, have been included in this report because they came from localities visited during the work involved in amassing the main collection. The identifications of some of the birds were made years ago at Tring by the late Dr. Ernst Hartert and Mr. Arthur Goodson, but all have been gone over and checked by the principal author (H. Fried- mann), who is alone responsible for the systematic part of this report. In 1923 the collector (A. Loveridge) compiled a systematic list of the 4 bulletin: museum of comparative zoology collection with brief notes, the latter chiefly regarding birds which were subsequently disposed of and which, therefore, were not available to the principal author in his study of the material. This list was freely consulted in writing this report. The principal author began to study this unusually fine and com- plete collection in 1928 at the Museum of Comparative Zoology, and continued to do so there until he left for the United States National Museum in September, 1929, where the work has been brought to a conclusion. He has written the entire body of the paper; the junior author is responsible for the descriptive list of the collecting localities, as well as for the short notes given under each species in quotations followed by his initials. His share in the paper is really more extensive than this would indicate, as he has given freely of his extensive knowl- edge of the country and has helped the senior author in understanding many otherwise puzzling distributional facts. The following museums have loaned material for use in the present study — The Academy of Natural Sciences of Philadelphia, the Am- erican Museum of Natural History, the Carnegie Museum, the Cleveland Museum of Natural History, the Field Museum, the Museum of Comparative Zoology, and the United States National Museum. Information concerning certain specimens under their care has been kindly supplied by Dr. Erwin Stresemann, Berlin; M. Jacques Berlioz, Paris; and Dr. V. G. L. van Someren, Nairobi. This report is precisely what its title implies — an attempt to put in usable form and to place on record a great many facts, observations, reflections, and opinions about a large number of kinds of birds found in tropical East Africa. It is not a faunal study — the series of specimens were usually too small to allow for that, but it is hoped that the facts herein recorded may be useful as a partial basis for such a study at some future time. In some respects this paper may be looked upon as a supplement to the work van Someren has done in Kenya Colony and Uganda. His main paper, published at Tring in Novitates Zoologicae, 29, 1922, pp. 1-246, is a mine of information about the dis- tribution of teeming bird life of the countries it deals with. The present collection, probably the finest one ever gathered together in the north- ern half of Tanganyika Territory, extends the limits far to the south of van Someren's localities. The birds collected in Ruanda and Mozambique are relatively few and of decidedly minor importance in this connection. When one considers the amount of collecting that has been done in tropical East Africa the results revealed by the present study are FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 5 really surprising. Thus, out of some 800 forms of birds involved, we are able to extend appreciably the previously known ranges of not less than 104 species and subspecies. Ten forms new to science have been found amongst them and have been described elsewhere; thirty-one birds are new to the avifauna of Tanganyika Territory, one of them, Muscicapa striata tyrrhenica, being recorded for the first time from the Ethiopian region. By "new to Tanganyika Territory" or "extension of range" we mean new with reference to published records. There may be other Tanganyikan specimens in other museums, but unless they have been recorded in print, we have no way of knowing about them. The classification followed for the arrangement of orders and fam- ilies is that formed by Wetmore (Proc. U. S. Nat. Museum, 76, 1930, pp. 1-8). Within the families the sequence of genera, species, and sub- species is patterned after the arrangement in Sclater's "Systema Avium Aethiopicarum." All measurements are in millimeters. Brief Summary of Results For the convenience of students interested in the distribution of East African birds the following brief summary may prove a useful guide to the contents of this paper. 1. New Forms described by the principal author from the Present Collection. Sarothrura e. y. . inguens Rhinoptilus africanus illmtris Eremialcctor decor atus lovcridgei Slreptopelia capicola anceps Rhinopomastus minor extimus Pogoniulus bilineatus conciliator Ncocichla gutturalis anguslus Cossypha heuglini euronota Sheppardia cyornithopsis bangsi Anthreptes orientalis barbouri Of these, one, Cossypha heuglini euronota, is from Mozambique; the rest are from Tanganyika Territory. The following Tanganyikan bird was described from another col- lection in connection with the present report: Bycanistes cristatus brevis 6 bulletin: museum of comparative zoology Birds New to Tanganyika Territory. Ardeola idae Spatula clypeata Melierax metabales metabates Falco ardosiaceus Falco dickinsoni Sarothrura elegans languens Rhinoptilus africanus illustris Eremialector decor atus loveridgei Streptopelia capicola anceps Caprimulgus europaeus unwini Rhinopomastus minor extimus Bycanistes cristaius brevis Tricholaema diadematurn diadematum Pogoniulus bilineatus conciliator Mesopicos goertae centralis Mirafra africana, doherlyi Anthoscopus roccatii taruensis Neocichla gutturalis angustus Pycnonotus tricolor minor Pycnonotus layardi fayi Sheppardia cyornithopsis bangsi Oenanthe oenanthe rostrata Irania gutturalis Hippolais iclerina Muscicapa striata tyrrhenica Bradornis microrhynchus taruensis Anthus nicholsoni neumannianus Anthreptes orientalis barbouri Euplectes taha intercedens Estrilda charmosyna kiwanukae Uraeginlhus bengalus brunneigularis Extensions of Range other than those already Accounted for as New to Tanganyika Territory. Butastur rufipennis Falco tinnunculus tinnunculus FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY Falco tinnunculus carlo Francolinus coqui hubbardi Francolinus shelleyi trothae Francolinus squamalus zappeyi Lissotis hartlaubii Hoplopterus armatus Numenius ar quota linealus Rhinoptilus cinctus emini Turturoena delegorguei sharper Ruivenzorornis johnstoni kivuensis Corythaeola cristala yalensis Crinifer zonurus Centropus senegalensis fasciipygialis Centropus burchellii Agapornis pullaria ugandac Glaucidium capense scheffleri Caprimulgus europaeus meridionalis Colius indicus pallidus Coitus macrourus pulcher Halcyon chelicuti zinjense Aerops boehmi Melittophagus lafresnayii oreobales Melittophagus variegalus loringi Coracias weigalli Bycanistes subcylindricus Lybius zombae albigularis Tricholaema ?nelanocephalum stigmatothorax Tricholaema lacrymosum ruahae Indicator variegatus variegatus Pmdotiscus insignis reichenowi Campethera nubica pallida Dendropicos lafresnayi hartlaubii Mesopicos griseocephalus kilimensis Mirafra africanoides intercedens Hirundo angolensis arctincta Campephaga quiscalina milnzneri Oriolus chlorocephalus 8 bulletin: museum of comparative zoology Corvus rhipidurus Parus rufiventris pallidiventris Parus fringillinus Argya rubiginosa emini Andropadus insularis insularis Cossypha heuglini euronota Neocossyphus rufus rufus Oenanthe isabellina Acrocephalus baeticatus suahelicus Camaroptera brachyura littoralis Alseonax adustus fiilleborni Anthus lineiventris Laniarius funebris degener Laniarius fiilleborni Laniarius lilhderi luhderi Dryoscopus gambensis erwini Pomatorhynchus australis littoralis Sigmodus retzii tricolor Eurocephalus ruppelii bohmi Onychognathus tenuirostris Nectarinia melanogastra melanogastra Pseudonigrita arnaudi emini Sorella eminibey Ploceus aureoflavus Ploceus bojeri Euplectes hordeacea sylvatica Euplectes hordeacea changamwensis Euodice cantans meridionalis Pytilia melba grotei Pytilia 7nelba belli Lagonosticta rubricata haematocephala Lagonosticta senegala kikuyuensis Lagonosticta senegala somaliensis Uraeginlhus cyanocephalus Poliospiza atrogularis reichenowi Linurgus kilimensis kilimensis FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 9 This report was completed and left the senior author's hands in the spring of 1932. In the summer of 1936 he revised it as best he could without actual access to the specimens, and in this way attempted to bring it up to date as a number of important papers dealing in whole or in part with the same region had appeared in the interval. The most important of these are: Bangs, O. and Loveridge, A. 1933. Bull. Mus. Comp. Zool., 75; pp. 143-221. Chapin, J. P. 1932. Bull. Amer. Mus. Nat. Hist., 65. Cheesman, R. E. and Sclater, W. L. 1935-36. Ibis (several parts). Delacour, J. and Edmond-Blanc, F. 1933-34. L'Oiseaux, 3 and 4. Granvik, H. 1934. Rev. Zool. Bot. Afr., 25, pp. 1-190. Lynes, H. 1934. Journ. f. Ornith., 82, Sonderheft, pp. 1-147. Moreau, R. E., and Sclater, W. L. 1936. Proc. Zool. Soc. Lond., 1935, pp. 843-891. Peters, J. L. and Loveridge, A. 1936. Bull. Mus. Comp. Zool., 79, pp. 129-205. Sclater, W. L. and Moreau, R. E. 1932-33. Ibis (several parts). van Someren, V. G. L. 1932. Nov. Zool., 37, pp. 252-380. Vincent, J. 1933-36. Ibis (several parts). In addition to these longer papers, a number of important notes by Grant and Mackworth-Praed appeared in the Ibis and the Bulletin of the British Ornithologists' Club; by Grote and Meise in Ornitholog- isches Monatsberichte, etc. Inasmuch as the specimens were not available during this revision of the manuscript, changes have been reduced to a minimum; it is possible that more would have been effected under other circumstances, but it is felt that no major ones have been left out in this way. 10 bulletin: museum of comparative zoology Alphabetical List of Collecting Localities ANKOLE. A district of the Western Province of Uganda. It lies northwest of Bnkoba and west of Lake Victoria. Chiefly a grass country, it has an altitude of about 5,000 feet. Collections were made here in 1919 by Kamau wa Kiragu. BAGILO. Altitude 5,000 feet. A native village on the northern slopes of the Uluguru Mountains, Tanganyika Territory. It is situated on former forest land and is still adjacent to extensive rain forest. Very different ecological conditions are to be found in the rain forest in contrast to the grass-grown valleys with their swift- flowing streams. For further details see Mem. Mus. Comp. Zool., 1929, 50, p. 96. BAHI. Altitude 2,731 feet. A station on the Central Railway of Tanganyika Territory west of Dodoma. A rather arid, acacia- thorn country during the dry season, but many acres are com- pletely flooded during the rains at which time waterfowl assemble in great numbers. BOGOTI. See Wami River. BUCHOSA. Stated by Salimu bin Asmani to be near the Kagera River and a good day's inarch north of Bukoba on the west shore of Lake Victoria. BUDDU (BUDU). Now a saza, or chieftainship, in the Masaka district of the Buganda Province of Uganda. It lies due west of the Sesse Islands in Lake Victoria. Both forests and swamps are well represented in the district. Kamau wa Kiragu collected here in 1919 while on his way to Ruanda. BUNGU. Mr. R. E. Moreau writes me that, "Bungu is the name of a jumbeate half as big as the entire East Usambara Mountains and running from the Korogwe-Makuyuni railway line to near Bum- buli," Tanganyika Territory. It therefore comprises altitudes of from 2,000 to 4,000 feet, this explains the strange admixture of savanna and rain-forest forms from this locality. Maps show two villages bearing the name of Bungu, situated near each other and about ten miles north of Korogwe Station on the Tanga Railway. Collections were made at Bungu by Kamau wa Kiragu in 1921. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 11 CHANTWARA. Stated by Salimu bin Asmani to be six hour's march (i. e. about eighteen miles) north of Bukoba on the west shore of Lake Victoria. CHANZURU. Altitude circa 1,500 feet. A village ten miles north- east of Kilosa station on the Central Railway of Tanganyika Territory. Ecological conditions essentially those of Kilosa. DAR ES SALAAM. Altitude 42 feet. Capital and chief port of Tanganyika Territory on the east coast. Collecting was carried out along the shore and on the flats (semi-swamped during the rains) which lie due west of the town. Some shooting was done among the baobabs and coconut palms north of the town. Most collecting was accomplished in June, 1918 and on other occasions while awaiting the arrival of transports. DODOMA. Altitude 3,706 feet. Chief town of Ugogo, situated on the Central Railway of Tanganyika Territory 260 miles west of Dar es Salaam. Eor a detailed description see: Loveridge, 1928, Proc. U. S. Nat. Mus., 73, Art. 17, pp. 3-4. EKAGANGO (KAGANGO). Not located on maps. In the Ankole district of the Western Province of Uganda. ELDORET. Altitude 6,875 feet. Chief town of the Uasin Gishu Plateau, Kenya Colony, and a station on the main line of the Kenya-Uganda Railway. I carried out a little collecting in the open grass lands and along a stream a few miles south of the township. FRERE TOWN. Altitude 50 feet. A settlement for freed slaves on the mainland opposite Mombasa Island, Kenya Colony. Typical coastal vegetation with mangroves abundant along the shore, mangoes and coconut palms among the scattered huts. Most of my collecting was done in close proximity to the Church Mission- ary Society's station. GULWE (IGULWE). Altitude 2,512 feet. A station, east of Dodoma, on the Central Railway of Tanganyika Territory. An inhospitable desertlike spot with dense patches of thorny scrub varied by more open sandy areas with scattered baobabs and rocky kopjes. GWAO'S VILLAGE. Altitude circa 4,145 feet. Situated thirty miles south of Singida on the Singida-Udatu road, Tanganyika 12 bulletin: museum of comparative zoology Territory. In dry upland country with dense patches of wind- blown scrub and thorn bush; scattered kopjes and worn boulders of large size are quite a feature of the landscape. Roads are lined with manvara hedges to serve as windbrakes. IKIKUYU (Not KIKUKU in the same province). Altitude circa 2,500 feet. Three days march south of Gulwe Station on the Cen- tral Railway of Tanganyika Territory. Scattered tembes (kraals) and unstumped clearings for growing mtama (millet) occur at intervals in the 'bush' which consists of an almost impenetrable thorny thicket. Kopjes are not uncommon. I LONG A. A village near Kilosa in Central Tanganyika Territory. For conditions see under Kilosa. ISOONA. SeeSuna. ISUNA. SeeSuna. ISIKISIA (IZIKISIA). Altitude 3,900 feet. A village of Tabora district a few hour's march north of Tabora on the Tabora- Mwanza road, Tanganyika Territory. At the time of my visit a region of rather dry miombo bush and orchard forest. KABALE. In Rukiga saza, which see. Presumably many of the birds labeled Kabale, were shot in the vicinity rather than in the town- ship, Tanganyika Territory. KABARE, BUKOBA. Stated by Salimu bin Asmani to be a two and a half hours' march (about seven miles) north of Bukoba on the west shore of Lake Victoria. KABURA. A village in Mawokota saza in Entebbe district, due west of Entebbe, Uganda. One of the halts made by Kamau wa Kiragu where he collected in 1919 when on his way to Ruanda. KAKINDU. (Not KAKINDU in Buso Province.) Stated by Salimu bin Asmani to be five hours' march north of the Kagera River and therefore near the southern border of Uganda. KARUMIYA (KARUM). Stated by Salimu bin Asmani to be nine bours' by boat west of Mwanza on the southern shore of Lake Victoria. KATABASUNGU. A village in Budu saza, Masaka district, Buganda Province, Uganda where Kamau wa Kiragu collected a few birds in 1919 when on his way to Ruanda. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 13 KAYANDA. Presumably a village lying between Maziba and Kabura, Mawokota saza, Entebbe district, Buganda Province, Uganda. KEDONG VALLEY. This well-known valley lies south of Escarp- ment Station on the Kenya-Uganda Railway. Upland acacia- thorn country with rocky escarpments and towering cliffs in the gorge near Lake Naivasha. All specimens from this region are in the Coryndon Memorial Museum, Nairobi. KIBOSI. Apparently a village near Rutaka, in Kigezi district, south- western Uganda. One of the localities in which Kamau collected in 1919. KIDETE STATION. Altitude 2,175 feet. On the Central Railway of Tanganyika Territory between Kilosa and Gulwe stations. Dry- bush country — baobab, thorn, miombo — , ecological conditions not unlike those of Kilosa. KIDIDIMO. Altitude circa 3,000 feet. One or other of two villages of this name within five miles of each other and about fifteen miles southwest of Dodoma, Tanganyika Territory. KILIMATINDE. Altitude 3,591 feet. In Manyoni district ten miles south of Saranda Station, Central Railway, Tanganyika Terri- tory. Dry scrub, open dry forest, scattered baobab trees, man- yara hedges, kopjes. In the flats below a few swamps form during the rains. KILOSA. Altitude 1,040 feet but most of my collecting was accom- plished in the hills behind the Otto (now Kilosa) Plantation, circa. 1,800 feet. Kilosa station is now considered to be 175 miles by rail west of Dar es Salaam. Dry miombo bush covers the hills and parts of the flats. Native gardens of maize and cotton extend over a great area radiating from the station. Uncultivated land gives rise to dense growth of grass and sword grass. Waterbirds were collected along the river and in some stretches of country flooded by the heavy rains. One might characterize the Kilosa district as typically weaver bird country for birds representing many genera of this family are dominant and their nesting activi- ties subsequent to the rains could not escape the notice of the least observant of travellers. KIMAMBA. Altitude 1,502 feet. A station nineteen miles east of Kilosa on the Central Railway of Tanganyika Territory. It is 14 bulletin: museum of comparative zoology one of the principal cotton producing centres of the territory. In general its topographical features are similar to those of Kilosa though extensive open forest occurs in close proximity to the cotton plantations and most of my collecting was done along the edge of this forest. KIMBWABWA'S VILLAGE. See Mkata Plains. KINYAMBWA (NYAMBWA). A village about thirty miles south- west of Dodoma and twenty miles east of Mahaka. Physical features similar to those of Dodoma. KIPERA. Altitude circa 1,600 feet. A village on the Kilosa-Iringa road about nine miles south of Kilosa, Tanganyika Territory, with essentially similar vegetation but differing physiographically by the absence of hills. A few species occur here which were not found at Kilosa. KISANGA. Altitude 2,500 feet. A village on the Kilosa-Iringa road at the foot of Elton's Pass fifty-one miles by road south of Kilosa, Tanganyika Territory. Conditions similar to those obtaining at Kilosa. KITORTU. In the Budu saza, Masaka district, Buganda Province, LTganda. KOME ISLAND. Altitude 889 feet. An island at the south end of Lake Victoria about twenty-five miles west of Mwanza. It was visited by Salimu bin Asmani who collected there in 1922. KONGWA. Altitude circa 3,000 feet. About twenty miles almost due north of Gulwe Station on the Central Railway of Tanganyika Territory. Kongwa is best known by the Church Missionary Society's training college which was my headquarters during the week that I spent there. Open mimosa forest and acacia bush with scattered baobab trees form the principal vegetative features of the landscape; it is a dry country except during the rainy seasons. KONSIGWE (KONZIG WE) is near Kinyambwa, which see. KYADONDO. A saza, or chieftainship, of Mengo district, Buganda Province, Uganda where collecting was done by Kamau wa Kiragu in 1919. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 15 LALAGO. Altitude circa 4,500 feet. A large village with Arab and Indian shops in the Maswa district of Mwanza, Tanganyika Territory. It lies four days march on a footpath from Mkalama to Mwanza. The country-side for miles about the village is very much under cultivation and grazing — rolling downs sparsely covered with thorn bush. At the time of my visit (October 17, 1922) the midday heat and glare were trying though the nights were cold. Crows were especially abundant. Crested Cranes, Sacred Ibis and Secretary Birds were seen in the vicinity. LASICALET. (LESICARET.) A village in Budu saza, Masaka district, Buganda Province, Uganda. LONGIDO. Altitude 4,200 feet. A little collecting was done (Febru- ary, 1916) in the vicinity of the British Camp at the foot of the western slope of Mt. Longido, fifty-three miles from Arusha, on the Arusha-Nairobi road. Open mimosa thorn bush with rocky hills dominates the scene; streams descend from the mountain which is totally different in its vegetation from the surrounding- steppe. LUKAYA see RUKAYA. LUMBO. Altitude 50 feet. On the mainland opposite Mozambique Island, Mozambique. Typical East Coast vegetation — coconut palms flourished in the sandy soil but it supported only a scrubby vegetation and comparatively few trees, of these the baobab was dominant. Mangroves throve in the muddy sheltered inlets and waterfowl were collected here in the vicinity of the British Camp from June to November, 1918. MADAZINI. Altitude 2,500 feet. A few huts on the Kilosa-Iringa road between Kipera and Kisanga, two days march south of Kilosa, Tanganyika Territory. Camp was pitched in the middle of the road in a patch of primary forest bordering a stream. This forest supported a fauna related to that of the rain forests of the Uluguru Mountains. At the same time I saw and captured the rare tsetse — Glossina austeni — the only example of this species which I ever encountered. MAHAKA. Altitude circa 3,000 feet. A village on a scarp twenty- five miles southeast of Kilimatinde, Manyoni district, Tanganyika Territory. Salimu collected here in 1922. 16 bulletin: museum of comparative zoology MASAKA. Altitude 4,200 feet. A township twenty-four miles west of Bukakata port (northwest shore of Lake Victoria) in the Masaka district, Buganda Province of Uganda. A coffee and cot- ton growing region with an average rainfull of thirty-seven inches. Kamau wa Kiragu collected at Masaka in 1919. MASOMUNTU MUKUBWA. Not located, according to the label it is in British Ruanda, i. e. the Kigezi district, Western Province, Uganda. MAZIBA. A village in Mawokota saza, due west of Entebbe, Uganda. Another of Kamau's collecting localities in 1919. MBALA. A village near Kisanga in the southern part of Kilosa district, Tanganyika Territory. A stream flows by the village. At the time of my visit (February 26-28, 1923) bare-throated francolins were common, two species of guineafowl were present in very small numbers. MB ETA. A village in the Ulugurn Mountains, which see. MBONOA. Altitude circa 4,300 feet. A few scattered huts in Singida districts, seven miles north of Itigi Station on the Central Railway of Tanganyika Territory. Around the huts were clearings made by the natives, beyond was open thorn bush and dense scrub; surface water was scarce. MBUGWE. In the Budu saza, Masaka district, Buganda Province, Uganda. MBULU'S VILLAGE (Not Mbulu township, Mbulu district, northern Province). Altitude circa 4,500 feet. Situated two days' march south of Singida on, or near, the Itigi-Singida road, Tanganyika Territory. Plenty of open acacia-thorn rnbugwe as well as dense thickets such as occur in the vicinity of Gwao's village which is only about six miles distant. MDJENGO'S VILLAGE (MADJENGO). Altitude circa 5,000 feet. A village fifteen miles north of Singida on the Singida-Mkalama road, Tanganyika Territory. Dry scrub, both open and in dense thickets, clothes the hillsides; rocky kopjes are a feature of this country which is only moderately well watered in the dry season. MHALALA see Muhalala. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 17 MIKANDO. Presumably in the Kigezi district, Western Province, Uganda. MKALAMA. Altitude 3,953 feet. Due north of Singida, Tanganyika Territory, of which it was formerly a district headquarters. There are several rivers in the vicinity, but the few birds collected were shot in dry, open bush a few miles outside the town. MKATA PLAINS, RIVER and STATION. Altitude 1,314 feet. Mkata Station is two hundred and forty kilometres west of Dar es Salaam, Tanganyika Territory, and in close proximity to the river from which it takes its name. Such collecting as was done was carried out north of the station where the extensive plains are open, or sprinkled with scattered acacia bushes, mimosa and other thorn trees. Thickets occur quite frequently while along the banks of the big river grow huge trees usually associated with primary forest. Kimbwabwa's village was at the northern end of the plain. MKINDO RIVER (MKINDU). Altitude circa 1,400 feet. A small tributary of the Wami River with which it unites almost exactly forty miles due north of Morogoro, Tanganyika Territory. At the time of my visit, rank growth of sedge and grass made it difficult to get about. Erom the luxuriant riot of vegetation arose mimosa trees and in some places they were so numerous as to amount to a forest. The river rises in the neighboring mountain Kirui which was climbed for 2,000 feet; near its summit is a vegetation totally different from that growing at its base. MLEWA'S. A village in the Singida district of Tanganyika Territory. MOROGORO. Altitude 1,628 feet. Situated on the railway a hundred and twenty-six miles west of Dar es Salaam, Tanganyika Terri- tory. It is the capital town of the Morogoro district. Most of my collecting was done in the vicinity of Government House a mile and a half south of the town and situated much higher than the town on the lower slopes of the Uluguru Mountains. In the dry valleys and ravines which scar the mountain side many birds were found which usually occurred only at higher levels. Most of the accipitrines were shot in the orchard forest of the extensive flats which stretch away to the north of the town. Waterfowl were procured on the lower reaches of the river which flows through the town. 18 bulletin: museum of comparative zoology MPINGA. Altitude circa 3,000 feet. A village in Dodoma district, twenty-five miles due west of Dodoma Station and about fifteen miles southeast of Bahi Station on the Central Railway of Tan- ganyika Territory. MSIMBx\. Altitude circa 1,600 feet. A small village near Ilonga and about ten miles northeast of Kilosa, whose ecological conditions it shares. Tanganyika Territory. MTALI'S. The village of Chief Mtali. Situated about twenty miles south of Mkalama, on the Mkalama-Singida road. Since the clos- ing of Mkalama station it comes under Singida district, Tangan- yika Territory. MUHALALA (MHALALA). Altitude circa 2,500 feet. Low-lying country in the Manyoni district between Saranda Station and Kilimatinde, Tanganyika Territory. Salimu stayed for a night on his way through to Mahaka. MUKANDO. SeeMikando. MWADIRA. Altitude circa 4,500 feet. A few scattered huts in Mwanza province, Tanganyika Territory, between the Semo and Simlyu Rivers, a hard march northwest of Lalago. The region is devoted to cattle grazing and maize cultivation. In the adjacent mbugwe, bull's horn acacia is abundant : on these extensive plains, inundated during the rains, such birds as sandgrouse, bustard and ostrich were found. MWANZA. Altitude 3,760 feet. The well-known township on the south or southeastern shore of Lake Victoria between Smith Sound and Speke Gulf. Mwanza is rich in bird life; this is particularly noticeable along the swampy marshes of the lake shore where cormorants, ibis, geese and kingfishers are much in evidence. MYOMBO (MEYOMBO). Altitude 1,600 feet. A village on the Kil- osa-Iringa road where it crosses the Myombo River about nine miles south of Kilosa, Tanganyika Territory. The vegetation is similar to that of Kilosa but open miombo forest containing many large trees covers an extensive area. NAIROBI. Altitude 5,490 feet. The capital city of Kenya Colony three hundred and thirty miles inland from Mombasa. It pro- vides suitable environment for many types of bird life. Forest- FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 19 loving species find sanctuary in the Karura and Ngong Forests as well as in the Parklands Reserve. Without the town the almost limitless plains stretch away towards the horizon; swamp-haunt- ing species occur among the marshes and, during the dry season, along the course of the then stagnant Nairobi River. Other kinds may be found in the extensive banana plantations which fill the valleys while many kinds of weavers frequent the native gardens as well as the plains and forest-edge. NDALA. Altitude circa 4,000 feet. A mission station and village forty miles northwest of Tabora just east of the Tabora-Shin- yanga road, Tanganyika Territory. Faunistically similar to Tabora. NDTEZA. In the Ankole district, Western Province, Uganda. NDUGUYU RIVER. A river about twenty-five miles northwest of Mkalama, Singida district, Central Province, Tanganyika Terri- tory. The river was dry when we arrived (October 14, 1922) and we had to dig for water after a trying march in withering heat over sunbaked and cracked thorn-bush steppe. NGARI MTONI. Altitude circa 4,600 feet. Site of a British camp by the ford across the river at the foot of Mount Meru on the Arusha-Nairobi road, Tanganyika Territory. Conditions much like those at Nairobi : very little collecting was done. NYAMBITA. Altitude circa 4,400 feet. A village about six miles northwest of Sagayo on the Sagayo-Mwanza road and about eighty miles from Mwanza, Tanganyika Territory. The country consisted of rolling grassland with scattered trees; manyara hedges had been planted as windbrakes near many villages in this region. At the time of my visit (November 1922) the flats were inundated with water by the frequent and heavy rainstorms. PARKLANDS. Altitude circa 6,000 feet. A suburb of Nairobi in which occurs the extensive Forest Reserve ; the few birds from this locality were collected on the outskirts of the forest. POOMA. Altitude circa 4,500 feet. Fifteen miles southwest of Singida on the Singida-Udatu road, Tanganyika Territory. For a description of this region see Gwao's Village. RUANDA. British Ruanda in the Kigezi district, southwestern Uganda. 20 bulletin: museum of comparative zoology RUKAYA (LUKAYA). A village in Mawokota saza, due west of En- tebbe, Uganda. Kamau wa Kiragu collected at Rukaya on his way to or from Ruanda. RUKIGA (LUKIGA). A saza, or chieftainship, near Kabale, capital town of the Kigezi district, Western Province, Uganda. The altitude is about 6,000 feet. The region is also known as British Ruanda. RUTAKA. In western Uganda, but not located. SAGAYO. Altitude 4,429 feet. An important village on the Simiyu River about sixty miles in a straight line southwest of Mwanza; it is in the Maswa district of Mwanza Province, Tanganyika Territory. In the immediate vicinity of the village the reddish soil was very dusty (October, 1922), beyond was the open common land with stunted and more or less widely scattered thorn bush. This land furnished the grazing for large flocks of goats and herds of cattle. Nearby were mbugwe, the black soil much fissured, in some cases covered with knee-deep dry grass, in others the grass had been burnt off and fresh green blades were springing up through the charred stubble. In some areas there was fairly open thorn bush with a few widely scattered larger trees, chiefly acacias, growing on the black cotton soil. Rocky granite kopjes with ever- green shrubs growing in the interstices of their rocks occurred here and there. Along the banks of the dry bed of the Simiyu River were great trees with dense thickets at their base, in places these thickets were often as much as thirty yards broad and ex- tended for some distance. At the time of our visit the Simiyu con- sisted of a few scattered pools interrupted by sand banks and stretches of pebbles. SAMUMBA. Altitude circa 4,145 feet. A village in the south of Singida district, close to Gwao's Village (which see for a descrip- tion of the ecological conditions, Tanganyika Territory). Salimu bin Asmani collected here in 1922 when turned back by an out- break of plague at Singida. SANGA. A village in the uplands of Ankole, Uganda. SANG A (MSANGA). Altitude between 4,000 and 5,000 feet. A river just northeast and east of the Semu (Semo) River to the north- west of Mkalama but in the Maswa district of Mwanza Province, Tanganyika Territory. A day's march from Lalago. FKIEDM\NN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 21 SARANDA. Altitude 3,511 feet. A station on the Central Railway of Tanganyika Territory between Dodoma and Tabora. It is in tin Manyoni district of Singida Province. Open flats, which are more or less flooded during the rains, are baked, cracked and dusty at other times. To the north they give place to bull's horn acacia and open mimosa forest. To the south dense thorn scrub. To the west miombo bush along the rift valley escarpment. SHANWA (SHANDWA). Altitude between 4,000 and 5,000 feet. In Maswa district, of Mwanza Province, Tanganyika Territory, 33° 40° W. 3° 17' S. An important village in which there are several Arab stores. It is situated at the foot of an extensive rocky hill which rises sharply from a vast plain. Waders and marsh birds come winging their way from distant swamps to roost in the stunted trees which grow from the precipitous scarps of this hill. SIMBA (SIMBO). Altitude 4,400 feet. Thirty miles northwest of Tabora near the western edge of the Wembere Flats and near the border of the Tabora-Nzega districts of Tabora Province, Tan- ganyika Territory. SIMBITI RIVER. A river about thirty miles northwest of Mkalama, Singida district, Central Province, Tanganyika Territory. It flows into Lake Eyasi. The plains crossed on the way from Mka- lama were very hot (October, 1922) and a terrific east wind was blowing. I have rarely seen such a great variety of game birds as were to be found in the vicinity of this river at this time. Guinea- fowl were present in great numbers, two or three species of franco- lin, two species of sandgrouse, two species of bustard, and doves innumerable, the Cape long-tailed species vying with the Senegal in numerical strength. SINGIDA. This well-known government station lies sixty miles due north of Manyoni on the Central Railway, i.e. midway between Mkalama and Kilimatinde. It is headquarters of the Singida dis- trict, Central Province, Tanganyika Territory. SINGO. A saza, or chieftainship, of Mubende district, Buganda Prov- ince, Uganda. SUNA (ISOONA, ISUNA). Altitude circa 4,500 feet. A village thirty-three miles north of Itigi on the Central Railway of Tanganyika Territory. The region was very arid at the time of my first visit (October, 1921) and only one waterhole occurred 22 bulletin: museum of comparative zoology in the whole trek from Itigi to Suna, the route being through dense thorn bush. In the vicinity of Suna the country is more open and subject to cattle grazing. A few baobabs and bussu palms are almost the only trees but manyara hedges, planted as windbrakes, are a conspicuous feature of the vegetation and give a touch of vivid green when everything else is burnt to shades of brown. Rocky watercourses occur in the lower ground between the rolling hills and water is retained in scattered pools. SUNGWIRI. A very small village near Kilosa (which see), Tangan- yika Territory. TABORA. Altitude 4,000 feet. A station on the Central Railway of Tanganyika Territory. Capital town of the Tabora district, Tabora Province. This well-known and long-established town has an annual average rainfall of thirty-three inches. At various times I collected here for brief periods, sometimes among the kopjes to the west, the more open country, or in proximity to the bush which crowns the hills to the south of the town. TINDIGA. Altitude circa 1,500 feet. A village a few miles outside Kilosa, Tanganyika Territory, in a large cultivated area. There are, however, many acres of sword grass which provide a retreat for lions and elephant. During the rainy season miles of country become inundated and so attract many species of waterfowl which are rarely seen at other times. TUMUTUMU. Altitude circa 5,000 feet. A village between Fort Hall and Nyeri, Kenya Colony. Typical Kikuyu country with quantities of bananas growing in the well-watered small valleys; remnants of virgin forest allowed to remain on the steeper slopes ; numerous patches of cultivation interspersed with old gardens lying fallow and overgrown with thickets. Being the home village of one of my collectors, Kamau wa Kiragu, on one or two occa- sions he did a little collecting there when home on furlough. ULUGU. Altitude circa 4,500 feet. A village on the Ushora River, Singida district, Central Province, Tanganyika Territory. See Ushora. [JLUGURU MOUNTAINS. Altitude where collecting was done 5,000-6,000 feet. Specimens bearing this name as the only locality were collected in 1922 by Salimu bin Asmani when on a visit to FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 23 his home village of Bagilo (which see). A great variety of topo- graphical conditions occur in close proximity to the village. USHORA (USSHORA, USCHORA). Altitude circa 4,500 feet. Collecting was carried on in the neighbourhood of the Ushora River which lies about fifty miles west and slightly north of Singida, Singida district, Central Province, Tanganyika Territory. A region close to the eastern edge of the Wembere Flats, dry country in which mimosa and acacia are dominant; along the sandy river bed (dry at the time of our visit) were many bussu palms. USURWE (USSURE). Altitude 4,691 feet. Perhaps USURE is the more correct spelling of this populous village of scattered kraals situated in the Singida district of the Central Province of Tan- ganyika Territory. It is thirty-five miles due west of Singida township and in precisely similar country except that there are no large bodies of water in the vicinity. It is on the Kulusi (Ikulusi) River. WAMI RIVER. Altitude circa 1,300 feet where collecting was done. This is a really important river which empties into the Indian Ocean at Sadani, north of Bagamoyo, Tanganyika Territory. Birds bearing this locality, however, were collected at a point thirty to forty miles northwest of Morogoro in country essentially similar to the flats below Morogoro township. WEMBERE FLATS (WAEMBERE STEPPE). Altitude circa 3,500 feet. In the Tabora district, Tabora Province, Tanganyika Territory; their southern end about eighty miles due west of Tabora and extending northwards nearly to Lake Eyasi. These plains were crossed between Ushora and Tambali at a time when water was unobtainable necessitating an all-night march. During the rainy season these flats are flooded and game is abundant. ZENGERAGUSU. Altitude circa 4,500 feet. Camp was made in open miombo bush and forest two miles north of the few scattered kraals which are known by the name Zengeragusu, Singida district, Central Province, Tanganyika Territory. Water was obtained from a pool in the otherwise dried-up river bed; the ground round about this pool for a radius of a hundred yards was a maze of lion and leopard tracks. 24 bulletin: museum of comparative zoology Superorder PALAEOGNATHAE Order STRUTHIONIFORMES Family STRUTHIONIDAE. Ostriches Struthio camelus massaicus Neumann Struthio massaicus Neumann, Journ. f. Ornith., 46, 1898, p. 243: Ukamba, Kenya Colony. 1 egg, Kimamba, near Kilosa, Tanganyika Territory, 1921, collected by native. This egg from Kimamba is very different from three of the more northern race molybdophanes. It is more ovate (long diameter relatively shorter) and is less pitted, the pits being small and without brownish pigment while in the other three the pits are large, deep, and brownish in color, giving the eggs a flecked appearance. The Masai ostrich occurs in Tangan}dka Territory and Southern Kenya Colony, north to the Tana River. Struthio camelus molybdophanes Reichenow Struthio molybdophanes Reichenow, Mitt. Orn. Ver. Wien, 1883, p. 202: Somaliland. 3 eggs, Western foothills of Mt. Kenya, Kenya Colony. The subspecific identification of these eggs is based on geographic grounds. This form is the one recorded from the Kenya region.1 The Somali ostrich ranges from the Tana River north to Gallaland and Somaliland. Superorder NEOGNATHAE Order COLYMBIFORMES Family COLYMBIDAE. Grebes Poliocephalus ruficollis capensis (Salvadori) Podiceps capensis Salvadori, Ann. Mus. Civ. Genova (2), 1, 1884, p. 252: Shoa. 1 c? , 1 9 , Kilosa, Tanganyika Territory, 18 April 1921. 1 immature d\ Kilosa, Tanganyika Territory, 23 June 1921. The adult cf is darker and larger than the 9 . The 9 was shot while trying to decoy from the nest and clutch of eggs. Van Someren2 1 cf. van Someren, Nov. Zool., 29, 1922, p. 4 — Archer's Post. 2 Nov. Zool., 29, 1922, p. 4. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 25 writes that in all his adult specimens the underside is pure white, lack- ing the black mottling of the northern birds. The two adults from Kilosa (breeding birds) have the breast, sides, and flanks mottled but the rest of the underparts pure white. The amount of white in the wings varies somewhat in the series in the Museum of Comparative Zoology. The young bird has the sides of the head and neck white with heavy dark grayish brown malar and ocular stripes, the gray brown of the top of the head projecting ventrally as a band which is practically continuous across the lower throat, separating the white of the throat (splotched with brownish) from the immaculate underparts of the body. The African little grebe occurs locally throughout the regions under consideration in this report. Order PELICANIFORMES Family PHALACROCORACIDAE. Cormorants Phalacrocorax carbo lucidus (Lichtenstein) Halieus lucidus Lichtenstein, Verz. Doubl., 1823, p. 86: Cape of Good Hope. 1 d\ Lasicalet, Buddu Uganda, 29 August 1919. The soft parts are recorded as follows: iris yellow; feet black; bill gray, upper bill black. This race is not recorded from Uganda by van Someren1, who even states that it does not occur on the inland lakes, or by Sclater.- The specimen has the white underparts of immaturity. Phalacrocorax carbo lugubris Riippell Phalacrocorax lugubris Ruppell, Syst. Uebers., 1845, p. 134, pi. L: Ethiopia. 1 d", 1 9, Mwanza, Tanganyika Territory, 2 December 1922. The male is immature and has the entire underparts white, most of the feathers tipped with dark brown, giving it a very spotted appear- ance. The female has just come into adult plumage and still has a considerable mixture of black and white on the breast, while the entire underside of the neck is whitish washed and spotted with light brown- ish. i Nov. Zool., 29, 1922, p. 5; 37, 1932, p. 253. 2 Syst. Avium Ethiop., pt. 1, 1924, p. 20. 26 bulletin: museum of comparative zoology Dr. van Someren1 notes that the bare skin of the throat is more extensive in this race than in lucidus, an observation the specimens in the series examined fail to confirm. Phalacrocorax africanus africanus (Gmelin) Pelecanus africanus Gmelin, Syst. Nat., 1, pt. 2, 1789, p. 577: Africa. 1 cf, 1 9 , Dar es Salaam, Tanganyika Territory, 19 November 1918. The female is immature. This bird is found throughout eastern Africa. Family ANHINGIDAE. Snake-birds Anhinga rufa rufa (Lacepede and Daudin) Plotus rufus Lacepede and Daudin, in Buffon's Hist. Nat. (18 mo Didot ed.) Quadr., 14, p. 319; Ois, 17, 1802, p. 81: Senegal (ex Daubenton, PI. Enl. 107). 1 d1, 1 9 , Dar es Salaam, Tanganyika Territory, 7 November 1918. "Fairly common at Dar es Salaam and on Lake Victoria." (A.L.) While not recorded at Mombasa by van Someren, it probably occurs there, and I have seen it at Kisumu, Kenya Colony, although van Someren does not include it in his 1922 list of Kenya birds. In a more recent paper,2 however, he states that it is very common on all the lakes and rivers of Kenya Colony and Uganda. The male is darker on the top of head and neck than the female, but is brownish, not blackish below. Order CIRCONIIFORMES Family ARDEIDAE. Herons, Bitterns Ardea cinerea cinerea Linnaeus Ardea cinerea Linnaeus, Syst. Nat. 10th ed., 1758, p. 143: Europe: restricted type locality, Sweden. 1 cf , Dar es Salaam, Tanganyika Territory, 19 November 1918. 1 d", Kilosa, Tanganyika Territory, 6 July 1921. 1 9 , Sagayo, Tanganyika Territory, 7 November 1922. "A male at Dar es Salaam on 5 July 1918 was the only other specimen shot." (A.L.) i Nov. Zool., 29, 1922, p. 5. = Nov. Zool., 37, 1932, p. 254. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 27 The female is not fully adult and has much gray on the top of the head and black streaks on the lower neck and breast. It was shot at 2 A. M. on a moonlight night while fishing in a little body of water. Widely distributed throughout the areas under discussion in this report. Ardea melanocephala Vigors and Children Ardea melanocephala Vigors and Children, in Denh. and Clapp, Trav. 2, App. xxi, 1826, p. 201 : probably near Lake Chad. 1 cf, Morogoro, Tanganyika Territory, 29 November 1918. 1 9 , Tabora, Tanganyika Territory, 13 December 1918. "Also Dar es Salaam, Kilosa, Sagayo. Probably the commonest species in Tanganyika Territory." (A.L.) The female is slightly smaller than the male and lacks the long occipital plumes. On 13th of December Loveridge found the black-headed heron nesting in a large tree in the middle of a cultivated area near Tabora. Ardea goliath Cretzschmar Ardea goliath Cretzschmar, in Riipp. Atlas, 1826, p. 39, pi. xxxvi: Bahhar Abiad, that is, White Nile. 1 d\ Tindiga, Tanganyika Territory, 15 August 1922. "The above specimen was the only one seen in Kilosa District; the species is decidedly scarce in the Territory except at Bahi after the big rains." (A.L.) The underparts are streaked with rufous, gray and white, although the upperparts look mature, indicating that the bird was not yet adult at the time of its death. Two adults from Kenya Colony have the lower mandible more yellowish, that is, the yellow color more extensive, than the Tindiga bird. Pyrrherodia purpurea purpurea (Linnaeus) Ardea purpurea Linnaeus, Syst. Nat., 12th ed., 1, 1766, p. 236: "in Oriente." 1 immature 9 , Dar es Salaam, Tanganyika Territory, 27 January 1919. 1 adult 9 , Kilosa, Tanganyika Territory, 23 June 1921. 1 adult d\ Kilosa, Tanganyika Territory, 6 July 1921. "Also Tindiga; seen at Morogoro. A common species." (A.L.) 28 bulletin: museum of comparative zoology The immature bird has the underparts of the body pale tawny, and lacks the long black stripes on the neck. The male is larger than the female and (to judge by dried skins) has darker legs and less white on the breast. A male from Akaba, Arabia, is slightly paler below. The young bird is less rufous than one of comparable age from South Africa and has a slightly shorter wing than the latter. Casmerodius albus melanorhynchus (Wagler) Ardea melanorhynchos Wagler, Syst. Av. Additamenta (last page), 1827: Senegambia. 1 d\ Morogoro, Tanganyika Territory, 17 July 1917. 1 d" , Morogoro, Tanganyika Territory, 24 November 1917. This bird and the typical European form which migrates to north Africa are often confused and all records of the species from tropical Africa probably belong to this, the resident form. The two may be easily identified by the color of the bare portion of the tibia — yellow in typical albus, and black like the tarsometatarsus in melanorhynchus. Care must be taken in using this character, as van Someren1 points out that the young of both forms have black legs. The African great white egret occurs in suitable places throughout Eastern Africa. Mesophoyx intermedium brachyrhynchus (Brehm) Herodias (Egretta) brachyrhynchus Brehm, Journ. f. Ornith., 1858, p. 471: Blue Nile. 1 9 , Kilosa, Tanganyika Territory, 28 June 1921. "Also seen at Dar es Salaam." (A.L.) This specimen agrees with the description of this race. It has a shorter bill than any example of typical intermedins from China and Borneo in the Museum of Comparative Zoology, but I have no com- parative African material. It is widely distributed in Eastern Africa. Melanophoyx ardesiaca (Wagler) Ardea ardesiaca Wagler, Syst. Av., Ardea, no. 20, 1827: Senegambia. 1 d* , Dar es Salaam, Tanganyika Territory, 1 July 1918. i Nov. Zool., 37, 1932, p. 250. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 29 Dr. van Someren1 writes of this species, "common on the coast of Tanganyika Territory. A few extending to the Pangani region," and records another specimen taken by Loveridge at Dar es Salaam. Egretta garzetta garzetta (Linnaeus) Ardea garzetta Linnaeus, Syst. Nat., 12th ed., 1, 1766, p. 237: "in Oriente." 1 d\ 1 9 , Dar es Salaam, Tanganyika Territory, 27 June 1918. Both specimens have the plumes on the back well developed. The male also has long plumes around the base of the neck and the long occipital plumes as well. Although specimens were collected only on the coast, the species ranges inland throughout the territory. The little egret is a fairly common bird in eastern Africa. Bubulcus ibis (Linnaeus) Ardea ibis Linnaeus, Syst. Nat., 10th ed., 1758, p. 144: Egypt. 1 d\ 1 9 , Morogoro, Tanganyika Territory, 31 January 1918. 1 cf , 1 9, Dar es Salaam, Tanganyika Territory, 18 November 1918. "Seen at Kilosa, Tabora, and Mwanza. Very common." (A.L.) The specimens from Dar es Salaam are in breeding plumage; the two from Morogoro are not. The buffy feathers on the breast, top of head, and the dorsal plumes are darker in the female than in the male. The buff-backed heron is one of the most abundant and widely dis- tributed members of its familv in Africa. Ardeola ralloides (Scopoli) Ardea ralloides Scopoli, Annus, 1, Hist. Nat., 1769, p. 88: Carniola. 1 d\ 1 9 , Kilosa, Tanganyika Territory, 29 June 1921. Both specimens are not yet adult, having black shafts to the prim- aries. The female has the shortest wing and tarsus of any bird in a series of eleven. Its measurements are as follows — wing 187 (as against 195-207 for other females); tail 69 (as against 70.5 to 74); culmen 58.5 (as against 57-60); tarsus 51 (as against 53-56 mm.). ' Nov. Zool., 29, 1922, p. 9. 30 bulletin: museum of comparative zoology Ardeola idae (Hartlaub) Ardea idae Hartlaub, Journ. f. Ornith., 1860, p. 167: Madagascar. 1 9 , Dar es Salaam, Tanganyika Territory, 15 June 1918. "A second individual was seen among the mangroves on the shore during the same month." (A.L.) As far as I know, this is the first record for this species in Tanganyika Territory. It has been recorded in Kenya Colony (Nairobi and Kijabe) and even as far inland as Mount Elgon (Granvik1). Its occurrence along the east coast of Africa is not surprising but the bird can hardly be either common or even regular there. Butorides striatus atricapillus (Afzelius) Ardea atricapilla Afzelius, Kongl. Vet.-Akad. nya Handl. Stockh., 25, 1804, p. 264: Sierra Leone. 1 o",l 9, Dar es Salaam, Tanganyika Territory, 15 June 1918. 1 d% Lumbo, Mozambique, 31 July 1918. 1 immature cf, Kilosa, Tanganyika Territory, 7 July 1921. The bird from Lumbo, Mozambique, has all the wing coverts much greener than either of the Dar es Salaam specimens, and is also larger than either of them. The bird from Kilosa, an immature specimen in brownish, streaked plumage, has a longer bill than either of the two adults from Dar es Salaam. Nycticorax nycticorax nycticorax (Linnaeus) Ardea nycticorax Linnaeus, Syst. Nat., 10th ed., 1758, p. 142: south Europe. 1 cf, 1 9, Tindiga, Tanganyika Territory, 25 January 1922. "These two were the only birds of this species observed. A large area around Tindiga had been inundated by excep- tionally heavy rains and these and other water birds not usually present, put in their appearance." (A.L.) These two birds have considerably larger, heavier bills and longer tarsi than any of a long series from Africa, Europe, and Asia. They 1 Journ. f. Ornith., 1923, Sonderheft, p. 46. FRIEDMANN AND LOVERIDGE : EAST AFRICAN ORNITHOLOGY 31 may represent an undescribed local race of this heron. Chinese birds average bluer, less greenish, above than European and African ex- amples. It seems as though there may be two forms of nycticorax in Africa, a heavy-billed, long-legged, resident race of which these two are specimens, and a migratory race, the typical European form. How- ever, the present species is not especially noted for its migratory habits and without more African material, I prefer not to separate a local tropical East African race. Ixobrychus minutus minutus (Linnaeus) Ardea minuta Linnaeus, Syst. Nat., 12th ed., 1, 1766, p. 240: Switzerland. 1 9 , Kilosa, Tanganyika Territory, 19 April 1921. 1 Proc. Zool. Soc. Lond., 1923, p. 916. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 51 Family FALCONIDAE. Falcons, Caracaras Falco peregrinus minor Schlegel Falco communis minor Schlegel, Abhandl. Geb. Zool. Vergl. Anat. pt. iii, p. 20, circa 1844: Mountains of the Cape of Good Hope. 1 d>, Morogoro, Tanganyika Territory, 16 January 1918. "The bird had a kingfisher in its claws when shot." (A.L.) The specimen is not quite fully adult, having tawny edges to the feathers of the upperparts, and little white on the cheeks. Falco biarmicus biarmicus Temminck Falco biarmicus Temminck, PI. Col. livr. 55, pi. cccxxiv, 1825: Caffraria and Cape of Good Hope. 1 9, Morogoro, Tanganyika Territory, 11 December 1917. "Also Kilosa and Tabora. "This bird had a chicken in its claws when collected." (A.L.) The specimen has the underparts unusually rufous, possibly due in part to staining. Another Morogoro bird collected by Loveridge is listed by van Someren1 as abyssinicus. Falco subbuteo subbuteo Linnaeus Falco subbuteo Linnaeus, Syst. Nat. 10th ed., 1758, p. 89: Europe; restricted type locality, Sweden apud Hartert. 1 9 , Simba, Tabora, Tanganyika Territory, 14 November 1921. 1 d\ Isikisia, near Tabora, Tanganyika Territory, 16 November 1921. 1 9 , Mahaka, Dodoma, Tanganyika Territory, 21 March 1922. 1 d1, Kome Island, Mwanza, Tanganyika Territory, 24 November 1922. The male from Kome Island and the female from Mahaka have very little rufous on the thighs, lower flanks, and under tail coverts; the latter bird has these parts almost white. The birds from near Tabora are much redder on these areas. i Nov. Zool., 29, 1922, p. 43. 52 bulletin: museum of comparative zoology Falco cuvieri A. Smith Falco cuvieri A. Smith, S. Afr. Quart. Journ. 1st ser., 1830, p. 392: Kei River, Cape Province. 1 cT, Morogoro, Tanganyika Territory, 31 October 1917. 1 9 , Kipera, Kilosa, Tanganyika Territory, 27 September 1922. "The female had grasshoppers in its stomach." (A.L.) The male is darker below than the female. The sexual dimorphism in this species is very small. Falco chiquera ruficollis Swainson Falco ruficollis Swainson, Birds W. Afr., 1, 1837, p. 107: Senegal. 1 9 , Morogoro, Tanganyika Territory, 29 January 1918. 1 9 (?), Isikisia, near Tabora, Tanganyika Territory, 16 November 1921. The bird from Morogoro is not fully adult and has the underparts much washed with cinnamon buff. The other specimen is labeled as a male, but doubtfully. On the basis of size and on comparison with carefully sexed specimens, it seems without much question to be a female. Falco tinnunculus tinnunculus Linnaeus Falco tinnunculus Linnaeus, Syst. Nat. 10th ed., 1758, p. 90: Europe; restricted type locality, Sweden apud Hartert. 1 c? , Morogoro, Tanganyika Territory, 4 November 1917. 1 9 , Morogoro, Tanganyika Territory, 5 February 1918. 1 d\ Morogoro, Tanganyika Territory, 19 March 1918. 1 9 , Dodoma, Tanganyika Territory, 6 December 1918. 1 9, Dodoma, Tanganyika Territory, 29 November 1921. "Also Mbala, Kilosa, Tabora. Very common at Dodoma during the 'winter' months where they find the scattered euphorbias offer them a good vantage point from which to descend upon their prey." (A.L.) The female taken at Dodoma on 29 November 1921 has the tail and rump considerably grayer than any of the others. These specimens constitute the southernmost records for this bird. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 53 Falco tinnunculus carlo (Hartert and Neumann) Cerchneis tinnunculus carlo Hartert and Neumann, Journ. f. Ornith., 1907, p. 592: Bissidimo, near Harrar. 1 d\ Morogoro, Tanganyika Territory, 29 January 1918. This specimen probably belongs to this race, but is somewhat intermediate between carlo and rupicolacformis, but nearer the former. It has the under wing coverts lighter than the former, but not as light as the latter race. This specimen is the southernmost record for carlo as far as I know, with the exception of one collected by Loveridge at Mkarazi, Uluguru Mountains. Recently Bangs and Loveridge1 have recorded five birds from Ujiji, Lake Tanganyika, a most unexpected locality. In their revision of the African races of this Kestrel, Grant and Mackworth-Praed2 consider the present specimen as rupicolacformis and not carlo. This seems to be a matter of opinion rather than of fact as it is really an intermediate example. Falco natjmanni naumanni Fleischer Falco naumanni Fleischer, in Laurop and Fischer's 'Sylvan' for 1817-18, p. 174, 1818: S. Germany. 1 immature cf\ Morogoro, Tanganyika Territory, 12 March 1917. 1 adult d" , 1 9 , Isikisia, Tanganyika Territory, 16 November 1921. "Also Sagayo." (A.L.) The two birds from Isikisia were apparently a mated pair as they were together; in fact, both were killed with one shot. The species is not known to breed in Africa south of the Sahara, however. Falco ardosiacus Bonnaterre and Vieillot Falco ardosiacus Bonnaterre and Vieillot, Encycl. Meth. Orn. pt. 3, 1823, p. 1238: Senegal. 1 cf, 1 9 , Kome Island, Mwanza, Tanganyika Territory, 22 November 1922. These specimens are the first to be recorded from Tanganyika Territory and the southernmost known for the species. They extend i Bull. Mus. Comp. Zool., 75, 1933, p. 149. 2 Bull. Brit. Orn. CI., 64, 1934, p. 79. 54 bulletin: museum of comparative zoology the known range southward by several hundred miles. Since they were collected Bangs and Loveridge1 have recorded a specimen from Ukerewe Island, also in Lake Victoria, June 17, 1930. See also remarks by Sclater and Moreau2 regarding birds seen at Amani, Usambara Mountains recorded under the name of Falco concolor Temminck. Falco dickinsoni Sclater Falco dickinsoni P. L. Sclater, Proc. Zool. Soc. Lond., 1864, p. 248: Chibasa, Shire River, Nyasaland. 1 cf ?, Morogoro, Tanganyika Territory, 18 June 1917. 1 cf, 1 9 , Kilosa, Tanganyika Territory, 6 June 1921. The bird from Morogoro is in brownish, immature plumage. Dickinson's kestrel was not known previously from north of Mozam- bique and Nyasaland. The present specimens constitute a notable extension of range and are the first from Tanganyika Territory. Polihierax semitorquatus major Bowen Polihierax semitorquatus major Bowen, Proc. Acad. Nat. Sci. Phila., 83, 1931, p. 259: Mbuyuni, Taita district, Kenya Colony. 1 9 , Mwadira, Mwanza, Tanganyika Territory, 19 October 1922. This race of the pigmy falcon occurs in northern Tanganyika Territory and southern Kenya Colony. It is larger and somewhat darker above than the typical South African form. It is not known how far south its range extends as relatively few specimens have been taken in Tanganyika Territory. The present specimen was considered typical semitorquatus by Friedmann3 but since then Bowen has considered the Tanganyika birds to be a distinct subspecies. The validity of major is very ques- tionable however, although van Someren4 has recently upheld it. > Bull. Mus. Comp. Zool., 75, 1933, p. 150. 2 Ibis, 1932, p. 498. s Bull. 153, U. S. N. M., 1930, p. 101. * Nov. Zool., 37, 1932, p. 271. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 55 Order GALLIFORMES Family PHASIANIDAE. Pheasants, Peacocks Francolinus coqui hubbardi Ogilvie-Grant Francolinus hubbardi Ogilvie-Grant, Bull. Brit. Orn. CI., 4, 1895, p. xxvii: Nassa, s. e. shore of Victoria Nyanza. 1 c?, 1 9 , Sanga, Mwanza, Tanganyika Territory, 16 October 1922. "Purchased from a native who had snared them for food." (A.L.) This constitutes the southernmost record for this bird, the nearest locality, Nassa, on Speke Gulf, being a short distance to the northeast of Sanga. Francolinus sephaena grantii Hartlaub Francolinus grantii Hartlaub, Proc. Zool. Soc. Lond., for 1865, p. 665, pi. xxxix, fig. 1, 1866: Unyamwezi country, Tanganyika Territory. 1 d1 , Dodoma, Tanganyika Territory, 5 December 1918. 1 9 , Saranda, Tanganyika Territory, 6 October 1921. "It is a common species throughout the Dodoma-Singida thorn-bush steppe." (A.L.) Besides the two specimens listed above, Loveridge shot two males and one female at Saranda, and a female at Mahaka. Van Someren1 lists three specimens from Dodoma. Although he does not say so, these were also collected by Loveridge. Of these three, van Someren writes that they are not identical and can be matched by birds from Kisumu. The Dodoma male is different from all others examined (16 speci- mens) in being practically white, untinged with buffy, on the abdomen. It is certainly the lightest bird seen. The lower back is grayer than the average, and the upper back has more dark margins to the white feathers than do most examples in the series. The female is not fully adult, is lighter, more reddish above, and paler, less marked below, than other birds. While working with the Smithsonian-Chrysler Expedition in 1926, Loveridge collected three more specimens of this bird, as follows : cf , Dodoma, Tanganyika Territory, 25 June 1926. 9 , Dodoma, Tanganyika Territory, 2 July 1926 cf , Saranda, Tanganyika Territory, 15 July 1926. i Nov. Zool., 29, 1922, p. 30. 56 bulletin: museum of comparative zoology This Dodoma male is quite unlike the one described above. It is buffier, less whitish below, less marked with reddish on the base of the throat and sides of the neck, and slightly lighter above, and lacks the subterminal dark marks on the upper tail coverts. The Saranda bird is similar but slightly darker. Francolinus africanus uluensis Ogilvie-Grant Francolinus uluensis Ogilvie-Grant, Ibis, 1892, p. 44: Machakos, Kenya Colony. cf, Nairobi District, Kenya Colony, 21 August 1920. The specimen is young although in adult plumage. The tarsal spurs are merely little yellowish knobs. This individual is lighter, and gen- erally brighter in color than two males from Guaso Nyiro and Fort Hall, and also differs from them in that it has no black spots on the white throat patch. Loveridge collected another male and a female also at Nairobi. The female is now in the Museum of the Academy of Natural Sciences at Philadelphia and the male in the American Museum of Natural History. Francolinus shelleyi trothae Reichenow Francolinus shelleyi var. trothae Reichenow, Vog. Afr. 1, 1901, p. 490: Ugalla, Tanganyika Territory. 2 d", Sagayo, Mwanza, Tanganyika Territory, 9 November 1922. Reichenow described trothae from a single specimen (cf ) in 1901. Since that time no further specimens have been collected with the exception of the present two, which are therefore of unusual interest. The type came from Ugalla (which is somewhere on the river of that name) while Sagayo, whence come these birds collected by Loveridge, is roughly 200 miles farther north. So then these two specimens not only constitute a virtual "rediscovery" of the form, but also a con- siderable extension of its known range. Unfortunately, I have no comparative material of shelleyi with which to compare them, but they differ from the colored figure of typical shelleyi1 in all the characters given by Reichenow for trothae but one, namely, that the rufous flecks on the breast and sides are not darker than those in the plate. i Cat. Birds Brit. Mus., 22, 1893, pi. vi. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 57 However, Mackworth-Praed1 records shelleyi from Southwestern Uganda, which means that the present specimens come from territory just about midway between Ugalla (type locality of trothae) and the range of typical shelleyi. It is consequently not very surprising to find some degree of intergradation in these birds. On the whole, however, they seem best referred to trothae. One of the birds is younger than the other, has the spurs still blunt and small, less rufous on the breast, and the lores and a stripe from thence through the eye, whitish without black markings. Both birds are equal in size, and have wings 168 mm. long. Jackson2 lists 3 females from Mulema, Southeast Uganda (Doggett coll.) which appear, "... to be much closer to F. s. trothae, which is found in Tanganyika Territory." Chapin3 has recently cast considerable doubt on the validity of trothae. Its final status must be left for someone with adequate ma- terial to decide. Francolinus levaillanti kikuyuensis Grant Francolinus kikuyuensis Ogilvie-Grant, Bull. Brit. Orn. CI., 6, 1897, p. xxiii: "Kikuyu distr.", in error, Eldoret apud Mackworth-Praed. 1 d", Eldoret, Kenya Colony, 8 November 1920. This topotypical specimen is slightly smaller than the average, hav- ing a wing length of 165 mm. Van Someren4 gives the wing length as 165-179 mm. Francolinus hildebrandti fischeri Reichenow Francolinus fischeri Reichenow, Journ. f. Ornith., 1887, p. 51: Ussere, Wem- bere Steppes, Tanganyika Territory. 1 9 , Myombo, Tanganyika Territory, 4 December 1920. 1 cf, Kilosa, Tanganyika Territory, 5 January 1921. 1 9 , Kilosa, Tanganyika Territory, 8 August 1922. 1 cT, Dodoma, Tanganyika Territory, 3 July 1926. The last of the four specimens listed was taken by Loveridge while 1 Ibis, 1922, p. 119. 2 Game Birds of Kenya and Uganda, 1926, pp. 28-29. 3 Bull. Amer. Mus. N. H., 65, 1932, p. 700. 4 Journ. E. Air. and Uganda Nat. Hist. Soc, 1926, p. 47. 58 bulletin: museum of comparative zoology with the Smithsonian-Chrysler Expedition and is now in the United States National Museum. The male from Kilosa is a young bird and is much browner than the one from Dodoma. The latter is considerably larger as well. The distribution of Francolinus hildcbrandti in northcentral Tan- ganyika Territory is as follows: the typical form hildcbrandti occurs high in the mountains (Nyange, Uluguru Mountains, 1 specimen), while the bird of the steppes and plains is fischeri and that of the coastal district from the Rufigi River southward grotci. As far as I know, altumi has not been taken in Tanganyika Territory. Recently Lynes1 has recorded johnstoni from Iringa, where it was breeding. Of the first three birds listed Loveridge writes as follows, "These were the only three specimens obtained by my collectors although the birds were very abundant and often heard calling close to the house. The species feeds in thick undergrowth and is difficult to approach on account of its wariness. At night they roosted in the trees of a rubber plantation. The immature male has a spotted brown breast, the adults, which I have compared with hildcbrandti typica in the British Museum collection, are not nearly so strikingly marked as in the typical form." An immature bird (sexed as 9 , but probably ad71) collected at Pooma, Singida, Tanganyika Territory, 5 October 1922, may belong to this race, but it is so young and the feathering so unlike anything else I have seen that I cannot be certain. On June 4, 1921, at Kilosa, a newly hatched chick, possibly two days old, was found wandering in the road. The specimen was not saved, but the observation gives an idea of the breeding season in that region. Francolinus squamatus maranensis Mearns Francolinus schuetti maranensis Mearns, Smiths. Misc. Coll. 56> no. 14, 1910, p. 1 : Maranu, Kilimanjaro, 5,000 feet. 1 adult 9 , Nairobi, Kenya Colony, 13 October 1915. 1 adult d\ 1 immature 9 , Ngong Forest, Nairobi, Kenya Colony, 26 July 1919. "Also Tumutumu, Kenya Colony, 3 d" , 1 9 , October 1920." (A.L.) The adult female is darker and smaller than the male, but has the throat whiter. The immature bird is about two-thirds full size and is darker above than either of the adults, the feathers of the back being 1 Journ. f. Orn., 82, 1934, Sonderheft, p. 47. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 59 very broadly subterminally tipped with blackish. The feathers of the breast and sides have white shafts and white terminal shaft spots; the bill is dark brownish, quite in contrast to the orange-red bill of the adults. The male is remarkable in that it has two well developed spurs on each foot. Conover1 has recently separated the birds of the Usambara Moun- tains under the name usambarae. This race is characterized by the presence of a full black line extending from behind the ear coverts forward to the back of the eye, and a streak of thickly spread dull black spots beginning at the gape and extending below the eye to behind the auriculars. The specimen recorded by me as marancnsis- belongs to Conover's race. Francolinus squamatus zappeyi Mearns i Francolinus schuetti zappeyi Mearns, Smiths. Misc. Coll., 56, no. 20, 1911, p. 4: E. shore of Lake Victoria. 1 adult d1, 2 adult 9, Chantwara, Bukoba, Tanganyika Territory, 23-27 December 1922. These three specimens are interesting in that they are very light for zappeyi and approach the characters given for doivashanus. Van Someren3 has recorded schuetti from not far to the north of Chantwara (South Ankole, Uganda), but whether schuetti and zappeyi are distinct seems questionable. However, I judge the former race solely by description as I have seen no material. One female has the grayish-brown median stripes on the abdominal feathers very narrow, and agrees with the characters of doivashanus, while the other has these stripes much broader and darker, and agrees with zappeyi, to which race I refer all three specimens, as doivashanus has not been recorded west of Lake Victoria. These specimens constitute the first record for this bird in north- western (west of Lake Victoria) Tanganyika Territory, but the race is known from Uganda and the eastern Belgian Congo. Since the above was written Mackworth-Praed and Grant4 has reviewed the forms of this francolin and have decided that zappeyi, i Auk, 1928, p. 356. 2 Ibis, 1928, p. 75. 3 Nov. Zool., 29, 1922, p. 27. * Ibis, 1936, pp. 374-376. 60 bulletin: museum of comparative zoology dowashanus, tetraoninus, and schuetti are all the same, in which case the last name would be the proper one for these birds. Pternistes afer humboldtii (Peters) Francolinus humboldtii Peters, Monatsb. Akad. Berlin, 1854, p. 134: Tete, Mozambique. 1 d\ 1 9, Lumbo, Mozambique, 31 July 1918. "Common in the natives' cultivated plots." (A.L.) The female has more white on the black abdominal feathers than the male. Sclater1 gives the range of this bird as north to southern Tan- ganyika Territory. Grote2 listed a bird from the south part of the coast of Tanganyika Territory but refrained from identifying it as to sub- species. He, however, compared it with Nyasaland specimens, which are leucoparaeus, and naturally found differences. His coastal bird is probably humboldtii. Sclater3 has reviewed the races of this spur-fowl and his conclusions seem correct as far as my material indicates. Bowen4 has examined the present specimens and agrees in the identi- fication. He also records humboldtii from Mkata Plains and the Mah- enge district of Tanganyika Territory. Pternistes afer itigi Bowen Pternistes cranchii itigi Bowen, Proc. Acad. Nat. Sci. Phila., 1930, p. 86: Gwao's, near Itigi, Tanganyika Territory. 1 9 , Saranda, Tanganyika Territory, 8 October 1921. 1 d\ 1 9, Gwao's, near Itigi, Tanganyika Territory, 11 October 1921. "Also Pooma (1 9 , 14 October 1921). A very common species in the Singida thorn-bush steppe. They are to be seen feeding on the roads in the mornings, but seek refuge in the dense thorn thickets which are often quite impenetrable to man, especially between Itigi and Suna. "The series shows considerable variation in coloring." (A.L.) Bowen has reviewed the forms of this spur-fowl in detail5 and I follow him in my disposition of the present material. 1 Syst. Avium Ethiop., pt. 1, 1924, p. 90. 2 Journ. f. Ornith., 1919, p. 298. 3 Bull. Brit. Orn. CI., 41, 1921, p. 134. * Proc. Acad. Nat. Sci. Phila., 82, 1930, pp. 149-164. s Proc. Acad. Nat. Sci. Phila., 82, 1930, pp. 149-164. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 01 This form inhabits the high central plateau of Tanganyika Territory west to the Ukongo region north of Lake Rukwa where it- intergrades with intercedens. At Tabora it merges with bohmi. The birds from Gwao's (type locality) have the shaft stripes of the feathers of the underparts much wider than do specimens of bohmi from Tabora, and in the former the margins of these feathers are con- siderably whiter, less vermiculated than in the latter. The Gwao's male, has all the feathers of the lower breast and middle abdomen with extremely broad black median stripes bordered with wide white ones, and each feather laterally margined with very wide, deep bay-brown stripes, giving the underparts a rich tricolored appearance. It cannot be a young bird as it has the tarsal spurs well developed. The other Gwao's bird is labeled 9 , but has small blunt, developing spurs, and is undoubtedly a male. This specimen is subadult, but does not agree with the characters given by van Someren.1 Reichenow2 writes that a very old bird from Iringa has the forehead, forepart of the superciliaries, and cheeks deep black. This is true of the two birds from Gwao's as well, but not of those from Saranda and from Pooma. It seems to be an indication of age and not a geographical variation. Meise3 has recently described two new forms, tornowi from Mkiri, Lake Nyasa, and tertius from Kitiniko, Ruhuhu, Tanganyika Terri- tory. I have seen no specimens of either, but the descriptions sound convincing. Pternistes afer bohmi Reichenow Pternistes bohmi Reichenow, Journ. f. Ornith., 1885, p. 465: Igonda, Tangan- yika Territory. 1 cf, 1 9, Tabora, Tanganyika Territory, 9 December 1918. This is the race inhabiting the fairly low country from Kigoma to Tabora and Igonda, intergrading with itigi east of Tabora. The male has the shaft streaks and vermiculations of the breast and lower throat black, while in the female they are dusky brown. The feathers of the mantle (interscapulars) are likewise vermiculated in the male and have no mottling in the female. The latter has much darker and broader shaft streaks in the feathers of the back and upper wing coverts than does the male. 1 Journ. E. Afr. and Uganda Nat. Hist. Soc, 1926, pp. 97-99. 2 Vog. Afrikas, 1, 1901, p. 682. 3 Orn. Monatsh, 41, 1933, pp. 142-144. G2 bulletin: museum of comparative zoology Pternistes afer nyanzae Conover Pternistes cranchi nyanzae Conover, Auk, 46, 1929, p. 345: Fort Ternan, Kenya Colony. 1 adult c\ 2 adult 9 , Locaret, Ankole, Uganda, 3 September 1919. 1 juvenal 9 , Singo, Ruanda, Uganda, 25 September 1919. 1 juvenal 9 , Kibosi, Ruanda, Uganda, 4 October 1919. 1 adult cT, Ekagango, Ankole, Uganda, 17 October 1919. 1 adult d\ 1 subadult 9, Mwanza, Tanganyika Territory, 1 December 1922. The subadult female collected at Mwanza is molting into adult plumage, and is in poor condition for study. However, on geographic grounds there can be no question of its subspecific identity. This race occurs from Fort Ternan and the Kisumu district, Kenya Colony, across Uganda to the Congo border, south to Ruanda, and east of Urundi, to as far as the southern end of Lake Victoria. Conover1 records one male from Nyanza, northeastern shore of Lake Tanganyika as belonging to this race, but not typical. This is by far the southern- most record. Dr. von Someren has recently2 described the juvenal plumage of P. c. intcrcedetis, although his birds are what would now be called nyanzae. The two juvenal birds from Ruanda (Singo and Kibosi) do not agree with the description given by van Someren. The Singo bird is only partly grown and is wholly in juvenal dress while the Kibosi specimen is in postjuvenal molt. The former may then be used for the description of the first pennaceous plumage. It is dull dark brown on the head and nape, the feathers of the latter with whitish gray tips; the upper back, scapulars, and inner upper wing coverts sandy tawny brown, each feather with a very broad fuscous-brown center; the feathers of the lower back less grayish, finely marked with wavy blackish-brown lines ; lores, orbital region, chin, and upper throat light grayish with a buffy wash, brownish basally; the forehead similar but each feather with a black median streak; breast, abdomen, flanks, and sides broadly barred dark brown and grayish white, the dark bands becoming paler towards the middle of the abdomen; the thighs light creamy gray banded with pale sandy brown; remiges dull brown, the outer webs lighter, more tawny, and mottled with dark brown; rectrices like the lower back; bill (dried) brown; feet (dried) orange-yellowish brown. 1 loc. cit. 2 Journ. E. Afr. and Uganda Nat. Hist. Soc, 192C, p. 97. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY G3 The Kibosi bird has assumed the subadult plumage on the upper breast, sides, nape, and upper back, and has the upper throat, chin, lores, and orbital space bare as in adults. The subadult plumage resembles the adult stage except in that the general tone of the coloration is paler, the feathers of the breast and upper abdomen with fewer vermiculations, more noticeably whitish. The adult male from Mwanza has broad chestnut margins on some of the feathers of the upper breast as well as on the abdomen. Pternistes rufopictus Reichenow Pternistes rufopictus Reichenow, Journ. f. Ornith., 1887, p. 52: Wembere Steppes, Tanganyika Territory. 1 d", 1 9, Sagayo, Mwanza, Tanganyika Territory, 31 October 1922. The female is much smaller than the male, having a wing length of 202 mm., as against 222 mm., in the latter. Even so, the female is unusually large, as Reichenow1 gives the wing length of females as 180-190 mm. Besides the pair listed, Loveridge collected one male and five more females at Sagayo, during October 1922. Two of these are in the American Museum of Natural History; the rest in the Academy of Natural Sciences at Philadelphia. Pternistes leucoscepus infuscatus Cabanis Pternistes infuscatus Cabanis, Journ. f. Ornith., 1868, p. 413: Lake Jipe, near Kilimanjaro. "Several collected at Tumutumu, Kenya Colony, now in Nairobi Museum." (A.L.) This bird is common throughout Tanganyika Territory from the Pangani River north into Kenya Colony through which it ranges to the Elgon and Baringo districts. i Vog. Afr., 1, 1901, p. 65. 64 bulletin: museum of comparative zoology Coturnix coturnix africana (Temminck and Schlegel) Coturnix vulgaris africana Temminck and Schlegel, Faun. Jap. 1849, p. 103: South Africa. "One shot at West Kenya in November 1915. Now in Nairobi Museum." (A.L.) This is the West Kenya record listed by van Someren.1 Coturnix delegorguei Delegorgue Coturnix delegorguei Delegorgue, Voy. Afr. Austr., 2, 1847, p. 615: Oury, that is, Upper Limpopo River. 19?, Nairobi, Kenya Colony, 2 July 1915. 1 9 , Morogoro, Tanganyika Territory, 18 June 1917. 1 cf , Nairobi, Kenya Colony, 24 August 1920. 1 cf , Samumba, Singida, Tanganyika Territory, 25 February 1922. "The Morogoro bird is probably a female although labeled a male." (A.L.) The male from Samumba is larger than the one from Nairobi and has the black ventral patch more extensive than the latter. The former is browner, the latter, grayer above. Family NUMIDIDAE. Guinea-fowl NUMIDA MELEAGRIS MITRATA Pallas Numida mitrata Pallas, Spic. Zool., 1, fasc. iv, pi. iii, 1767: Madagascar. 1 d\ Morogoro, Tanganyika Territory, 18 October 1917. 1 immature , 1 9, Mahaka, Tanganyika Territory, 28 March 1922. "Three birds probably belonging to this race were seen at Kidete in 1923." (A.L.) The two skins listed above have not been examined by me as they were given to Tring Museum. While with the Smithsonian-Chrysler Expedition in 1926, Loveridge collected two females at Saranda, Dodoma, on 15 July. These birds are now in the United States National Museum. According to Sclater1 granti is only doubtfully distinct. However, the limited material at hand (2 specimens) agrees with the description of the race. Family GRUIDAE. Cranes Balearica pavonina gibbericeps Reichenow Balearica gibbericeps Reichenow, Journ. f. Ornith., 1892, p. 126: Lake Jipe, near Kilimanjaro. 1 d" , Tindiga, Tanganyika Territory, 21 September 1921. 1 d" , Lalago, Tanganyika Territory, 17 October 1922. "Also one at Singida, Tanganyika Territory; they are very numerous at Kilimatinde and throughout the Mwanza district." (A.L.) The Tindiga specimen has the secondaries longer than the primaries while that from Lalago has the primaries slightly longer than the secondaries. The latter has the bare area on the sides of the head very much broader antero-posteriorly than the former and also has a more forward projecting forehead. » Syat. Avium Ethiop., pt. 1, 1924, p. 99. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 07 Family RALLIDAE. Rails, Coots, Gallinules Rallus caerulescens Gmelin Rallus caerulescens Gmelin, Syst. Nat., 1, pt.2, 1789, p. 716: Cape of Good Hope 1 d\ Nairobi, Kenya Colony, 5 November 1915. 1 9 , Morogoro, Tanganyika Territory, 30 November 1918. Sclater1 gives the range of this rail as extending north to Kenya Colony in the eastern part of Africa. However, Erlanger2 records a specimen taken by Hilgert near Adis Ababa, Ethiopia, a considerable northern extension of the range. The two specimens listed above agree with topotypical material from the Cape of Good Hope. Crex crex (Linnaeus) Rallus crex Linnaeus, Syst. Nat. 10th ed., 1758, p. 153: Europe; restricted type locality, Sweden (Hartert). 1 cf, Nairobi, Kenya Colony, 15 April 1919. "With the exception of a corn crake brought into the house at Nairobi by a cat in April 1916, I have met with no other examples of this bird." (A.L.) The specimen is an adult in fine plumage. Limnocorax flavirostra (Swainson) Gallinula flavirostra Swainson, Bds. West Afr., 2, 1837, p. 244, pi. xxviii: Senegal. 1 cf , 1 9 , Dar es Salaam, Tanganyika Territory, 16 November 1918. 1 cf, Kabare, Bukoba, Uganda, 21 August 1919. 1 immature 9, Kilosa, Tanganyika Territory, 8 July 1921. "Also Nairobi, Kenya Colony. This is one of the commonest of water birds in the Territory." (A.L.) West African males are blacker, less brownish above than East African examples of like sex. The latter retain a slight brownish cast to the back even in full breeding plumage. Females are not separable. Young birds have whitish throats. The immature female from Kilosa 1 Syst. Avium Ethiop., pt. 1, 1924, p. 101. 2 Journ. f. Ornith., 1905, p. 87. 68 bulletin: museum of comparative zoology has a light patch on the maxilla surrounding the nares, which in life was probably yellowish. Unfortunately no note was made of the color in life. Porzana porzana (Linnaeus) Rallus porzana Linnaeus, Syst. Nat. 12th ed., 1, 1766, p. 262: Europe; restricted type locality, France (Hartert). 1 d\ 1 9, Kilosa, Tanganyika Territory, 14 April 1921. The male has the entire throat and chin abundantly flecked with whitish, while the female has these parts dark slate gray with only a few lateral spots. The species is extremely variable in the amount of freckling it shows.1 Porzana pusilla obscura Neumann Porzana obscura Neumann, Orn. Monatsb., 1897, p. 191 : Kibaya, Tanganyika Territory. 1 adult d\ 1 adult 9, 1 immature 9, Kilosa, Tanganyika Territory, • 22 April 1921. 1 immature 9, Kilosa, Tanganyika Territory, 22 June 1921. "This is a common species at Kilosa and Tindiga." (A.L.) The adult female is lighter than the male, particularly on the throat, and is, in fact, more like intermedia than like obscura. The throat is lighter than the rest of the underparts. The male is lighter, particularly on the head, than the race interna from South Africa (Moorddrift, Transvaal.). In this connection it is noteworthy that van Someren2 writes of a female from Lake Naivasha, Kenya Colony, that the, "... gray which is appearing on the head and throat is paler than in southern birds. I suspect that the northern birds are distinct." Van Someren's suspicion has been shown to be correct; the northern birds are obscura and the southern ones interna. The type locality of obscura is Kibaya in the Masai country of northern Tanganyika Territory, and inasmuch as van Someren's birds to the north of that locality and Loveridge's to the south of it agree, there can be no question but that the present birds are obscura. The adult female was in breeding condition when shot; large ova were present in the ovary. 1 cf. van Someren, Nov. Zool.. 29, 1922, p. 21. 2 Nov. Zool., 29, 1922, p. 21. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 69 The two immature birds are not similar. One is heavily barred with brown completely across the breast and abdomen, while the other has the bars confined to the sides and flanks. Sarothrura elegans languens Friedmann Sarothrura elegans languens Friedmann, Proc. N. Eng. Zool. CI., 10, 1928, p. 68: near Bagilo, Uluguru Mountains, Tanganyika Territory. 1 9, near Bagilo, Uluguru Mountains, Tanganyika Territory, 19 May 1921. "This was the only example of the genus collected or seen from 1915 to 1925." (A.L.) This specimen is the type and only known example of languens, and constitutes the first record for the species in Tanganyika Territory, serving to connect the ranges of elegans of South Africa and of loringi of Mt. Kenya. S. e. languens is similar to loringi, but lighter above, somewhat more narrowly barred below, and slightly smaller in size. The Somaliland form, buryi, is easily told from languens by the fact that the center of the breast and abdomen is conspicuously white in the former and not so in the latter. PORPHYRIO MADAGASCARIENSIS (Latham) Gallinula madagascariensis Latham, Index Orn. Suppl., 1801, p. lxviii: Mada- gascar. 1 Journ. f. Ornith., 1921, p. 264. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 95 While with the Smithsonian-Chrysler Expedition, Loveridge col- lected a female anceps at Dodoma, Tanganyika Territory, on 23 June 1916. This form is intermediate between tropica and somalica in appearance, resembling the former on the upperparts, and the latter on the underparts, but with the breast somewhat more vinaceous than in somalica. In Tanganyika Territory three races of S. capicola occur as follows : 1. tropica: west of the Rift Valley, extending eastward in the southern part of the Territory. 2. somalica : coastal districts south to the Pangani River. 3. anceps: definitely known from Kilosa and Dodoma, probably ranges from the eastern escarpment of the Rift Valley to the coastal districts where it meets and intergrades with somalica. Stigmatopelia senegalensis aeqtjatorialis (Erlanger) Turlur senegalensis aequatorialis Erlanger, On. Monatsb., 1904, p. 98: Mena- balla, Ethiopia. 1 cf , Morogoro, Tanganyika Territory, 23 August 1917. 1 d\ Lumbo, Mozambique, 15 July 1918. 1 9 , Kilosa, Tanganyika Territory, 30 December 1920. "Also Dodoma and Mpinga, Tanganyika Territory." (A.L.) The eastern form, to which I have referred these specimens, is of somewhat doubtful validitv. However, in the absence of anv material of typical senegalensis, I prefer to follow the conclusions reached by Sclater1 who keeps the two distinct. These three specimens certainly agree with the characters and plate given by Erlanger2 but others from Kenya Colony and Ethiopia are more reddish above and below, thereby suggesting that the color characters are very variable. If aequatorialis should eventually prove to be indistinguishable from senegalensis (as indeed is claimed by several writers), then all the birds of the territory covered by this report, would have to be considered as of the typical race. This pigeon occurs throughout the region under discussion except in forest areas, and not higher than 6,000 feet in the mountains or high- lands. It occurs in cultivated districts as well as in the thorn-bush country. A nest with two eggs was found at Morogoro on 1 August 1917. i Syst. Avium Ethiop., pt. 1, 1924, p. 170. 5 Journ. f. Ornith., 1905, p. 117. 96 bulletin: museum of comparative zoology Oena capensis capensis (Linnaeus) Columba capensis Linnaeus, Syst. Nat. 12th ed., 1, 1766, p. 286: Cape of Good Hope. 1 d" , 1 9 , Dodoma, Tanganyika Territory, 6 December 1918. "Also Nairobi and Kilosa." (A.L.) While with the Smithsonian-Chrysler Expedition, Loveridge col- lected another male at Dodoma on 14 June 1926. It is now in the United States National Museum. This species is distributed throughout the region under discussion. Oena capensis anonyma Oberholser1 is a synonym. Tympanistria tympanistria fraseri Bonaparte Tympanistria fraseri Bonaparte, Comp. Av. 2, 1855, p. 67: Fernando Po. 1 d\ Mbugwe, Buddu, Uganda, 31 August 1919. 1 cf , Morogoro, Tanganyika Territory, 28 February 1918. 1 d\ 1 9, Bungu, Usambara, Tanganyika Territory, September 1921. 2 d\ Kome Island, Mwanza, Tanganyika Territory, 24 November 1922. "Also Tumutumu and Nairobi, Kenya Colony; Uluguru Mountains, and Kilosa, Tanganyika Territory." (A.L.) This dove is widespread throughout the region covered by this paper, but is found only in well-wooded areas, such as tree-lined river banks as well as in true forest. It occurs in the mountains up to as high as 7,000 feet, but is not common at altitudes of more than about 6,000 feet. The metallic wing spots are deep purplish-blue in the males from Morogoro, Bungu, and one from Kome Island, while in the other two (Mbugwe and Kome Island) they are green. In this connection it may be noted that Ogilvie-Grant2 writes of a closely allied species Chalco- pelia afra that the wing spots are purple in freshly molted birds, but become green with wear and exposure. In the present case there is no difference in wear or age of the plumage between the birds with green spots and those with blue ones. In fact, the two birds taken together at Kome Island are exactly alike as regards wear, etc., but differ in the color of the spots. It certainly appears that Ogilvie-Grant was wrong in his statement, and that he was probably misled by a coincidence of wear and spot-color. > Proc. U. S. Nat. Mus., 28, 1905, p. 843. 2 Ruwenzori Exp. Rpts., Aves, Trans. Zool. Soc. Lond., 19, 1910, p. 450. FRIEDMANN AND LOVERIDGE : EAST AFRICAN ORNITHOLOGY 97 Van Someren1 writes that the mature but, " . . . youngish female has a decided band (on the breast) which is tinged with ashy-brown." Judging by this observation, the female taken at Bungu is "youngish" as it has the whole pectoral area dull gray, each feather tipped with pale brownish. Turtur chalcospilos CHALCOSPiLOS (Wagler) Columba chalcospilos Wagler, Syst. Av. Columba, sp. 82, 1827: "Terra Caf- frorum," that is, Eastern Cape Province. 1 c? , 1 9 , Dar es Salaam, Tanganyika Territory, 2 July 1918. 1 9 , Kilosa, Tanganyika Territory, 16 July 1921. 1 cf, Kilosa, Tanganyika Territory, 21 July 1921. "Also Tumutumo, Kenya Colony; Morogoro, Tanganyika Territory; and Lumbo, Mozambique." (A.L.) While with the Smithsonian-Chrysler Expedition, Loveridge col- lected another ( 9 ) at Dodoma, Tanganyika Territory, 13 June 1926. The Lumbo specimens are listed as subsp. caffra ? by van Someren2 who writes that they (2 c?, 9 ) are darker on the underside than East African specimens and the wing speculum is not constantly green, being purplish-blue in one bird, half blue, half green in another, and in the other green. This little dove is widely distributed over the territory covered by the present paper, where it is chiefly found in the dry thornbush country, but also along the coastal areas, while in the uplands of the interior it occurs commonly in the coffee shambas. A great many names have been proposed for this bird, but all the so-called eastern forms are untenable. The birds from Tanganyika Territory and Kenya Colony would have to be considered as acanthina Oberholser if that race were recognized. The birds from Lumbo, questionably refered to caffra by van Someren, would probably have to be considered as zambeziensis Roberts, were the characters of that form constant. Turtur afer kilimensis (Mearns) Chalcopelia afra kilimensis Mearns, Proc. U. S. Nat. Mus., 48, 1915, p. 383: Kilimanjaro. 1 d\ Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 1 Journ. E. Africa and Uganda Nat. Hist. Soc, no. 32, January 1928, p. 162. 2 Nov. Zool., 29, 1922, p. 36. 98 bulletin: museum of comparative zoology Gyldenstolpe1 writes that sclateri Rothschild is distinct from kili- mensis and that the former differs from the latter in having darker brown upperparts and the flanks and abdomen less tinged with cinna- mon, more whitish. The present specimen is intermediate between the two, having the lighter upperparts of kilimensis and the whiter, less cinnamoneus flanks and abdomen of sclateri. However, the char- acters given by Gyldenstolpe hold none too well for West African birds. Specimens from Cameroon and Liberia vary in the color of the ab- domen from practically pure white to much washed with cinnamon, and that of the upperparts likewise varies in shade. For the present, at least, I prefer to consider sclateri only doubtfully distinct from kilimensis and to call all the Tanganyika birds by the latter name. According to Gyldenstolpe, the birds of the western type occur east to Tanganyika Territory. In the eastern part of the region under con- sideration in this paper this species stops at the Tanganyika-Kenya border and occurs in the latter country only in its southwestern part (Kavirondo and Kisii), not being found in the coastal areas or in the highlands. In Tanganyika Territory and Uganda it is widespread in bushy, savanna districts, shunning true forest and dry Acacia country. Aplopelia larvata larvata (Temminck) Columba larvata Temminck, Pig., Colombes, 1810, p. 71, pi. xxxi: Le pays d'Antiniquoi, that is, Knysna, Cape Province. 1 cf, 2 9, Bagilo, Uluguru Mountains, Tanganyika Territory, 13 June 1922. "Also seen at Madazini, south of Kilosa." (A.L.) A. 1. kilimensis Neumann is a synonym. This dove is a bird of the true forests, and, consequently, its dis- tribution in the territory under discussion is very broken and dis- continuous. In Tanganyika Territory it is known only from Kiliman- jaro, Mt. Meru, the Usambara Mountains (Mlalo), and the Uluguru Mountains (Bagilo, Mkarazi, and Nyange). In Kenya Colony it has been taken at Taveta, Nairobi, Kyambu, the Meru forest near Mt. Kenya, and on Mt. Kenya itself, Kakamega, Escarpment, Mau, Cherangani, W. Elgon. It does not appear to be known from farther west than Mt. Elgon. « Kung. Sv. Vet. Akad. Handlngr., 1024, p. 312. FRIEDMANN AND LOVERIDGE : EAST AFRICAN ORNITHOLOGY 99 VlNAGO CALVA SALVADORII Dubois Vinago calva salvadorii Dubois, Proc. Zool. Soc. Lond., 1897, p. 784: eastern and central tropical Africa; restricted type locality, W. shores of Tan- ganyika (Hartert). 1 9 , Kibosi, Ruanda, Uganda, 4 October 1919. 1 d\ 2 9 , Chantwara, Bukoba, Tanganyika Territory, 3 January 1923. One of the Chantwara females is more or less intermediate between salvadorii and brevicera, and the one from Kibosi, Ruanda, is also not quite typical of this race. This form occurs in Tanganyika Territory west of the Rift Valley (more especially west of Lakes Tanganyika in the Congo, and Victoria in Tanganyika Territory), thence north throughout Uganda, east to Mt. Elgon and the Elgeyu escarpment and Mau in Kenya Colony. All the fruit pigeons are great wanderers, their presence in any one spot being largely correlated with the ripening of wild figs and other fruits. Vinago calva brevicera (Hartert and Goodson) Treron calva brevicera Hartert and Goodson, Nov. Zool., 25, 1918, p. 353: Moshi, Kilimanjaro district. 1 cf , Ngong Forest, Nairobi, Kenya Colony, 21 September 1920. This race occurs east of the Rift Valley east as far as the Kilimanjaro- Taveta district. According to van Someren1 this race is present in the Sotik and Kavirondo districts, around the north and east shores of Lake Victoria, but it is quite likely that the birds from the southern Sotik district may prove to be granviki or, at least, intermediate between brevicera and granviki. Vinago calva granviki (Grote) Treron calva granviki Grote, Journ. f. Ornith., 1924, p. 102: Ukerewe Island, Lake Victoria. 1 9 , Sagayo, Mwanza, Tanganyika Territory, 2 November 1922. This specimen is referable to this race. The race, however, is merely an intermediate between brevicera and salvadorii and occupies the territory between their respective ranges, that is, west of the Rift Valley from Mkalama and the Wembere Steppes to the southern end 1 Journ. E. Afr. and Uganda Nat. Hist. Soc. Jan., 1928, p. 177. 100 bulletin: museum of comparative zoology of Victoria Nyanza. It may be recognized as far as the material examined indicates, as it seems to be fairly constant in its characters. The naked basal part of the bill is nearly as extensive as in salvadorii, but the coloration of the plumage is lighter, especially on the under- parts where it is noticeably lighter green, with a yellowish tone on the breast and middle of the abdomen. Vinago delalandii granti van Someren Vinago delalandii granti van Someren, Bull. Brit. Orn. CI., 40, 1919, p. 20: Kilwa, Tanganyika Territory. 1 cf, Morogoro, Tanganyika Territory, 15 August 1917. 1 cf , Dar es Salaam, Tanganyika Territory, 7 November 1918. 1 9 , Kilosa, Tanganyika Territory, 8 July 1921. unsexed, Kilosa, Tanganyika Territory, 13 July 1921. "Also Tindiga, Tanganyika Territory." (A.L.) Grant's fruit pigeon is a coastal and subcoastal bird in the eastern half of Tanganyika Territory (south well into Mozambique) and occurs in the southern portion of the coastal districts of Kenya Colony as well. In southern Mozambique it is replaced by typical delalandii which is larger and less olive, more grayish, on the head and breast, less golden, more olive on the back, wings, and interscapulars, and less yellowish on the abdomen. However, a specimen (M.C.Z. 81103) from Komati River, Transvaal (near the border of Mozambique) is typical granti. I have seen no material of V. d. orientalis Gunning and Roberts, but if it is a form of delalandii and not of wakefieldii (as treated by Sclater,1 it may be that it will prove to be the same as granti. The brief description of orientalis given by Roberts in his South African Check List2 fits the specimen from Komati River, which is also identi- cal with granti. Vinago wakefieldii wakefieldii (Sharpe) Treron wakefieldii Sharpe, Proc. Zool. Soc. Lond., for 1873, p. 715, pi. lviii, fig. 2, 1874: Mombasa. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. i Syst. Avium Ethiop., pt. 1, 1924, p. 176. 2 Ann. Transvaal Mus., 10, 1924, p. 124. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 101 Wakefield's fruit pigeon occurs in the northern half of Tanganyika Territory east of the Rift Valley, merging south of the Pangani River with V. to. oricntalis. Its range extends northward into the coastal area of Kenya Colony, spreading inland along the valleys of the Tana and Juba Rivers. In Tanganyika Territory it occurs away from the coastal districts only in mountainous districts such as the Usambara range, whence it is known from Bungu, Amani, and Lushoto. It appears not to have been taken in the Uluguru range as yet, but quite possibly occurs there. It is, however, probably sporadic in its distribution, as forested areas are few and far between in the inland nyika of the Territory. Bangs and Loveridge1 record it from Ilolo, Rungwe district. Lynes2 has taken a specimen and seen numbers of individuals at Iringa. Order CUCULIFORMES Family MUSOPHAGIDAE. Plantain-eaters Turacus livingstonii cabanisi (Reichenow) Corythaix cabanisi Reichenow, Journ. f. Ornith., 1883, p. 221: Bagamoyo, Tanganyika Territory. 1 d71, Morogoro, Tanganyika Territory, 28 November 1918. 1 9 , Uluguru Mountains, Tanganyika Territory, 13 May 1921. 1 9, Uluguru Mountains, Tanganyika Territory, 19 May 1921. "Also Bagilo, Madazini, Mbala." (A.L.) This race differs from typical livingstonii in having the rump and tail more bluish. Neumann3 has reviewed the remarkable and perplexing color varia- tions of the birds of the livingstonii group and has found it difficult to know what to do with T. reichenowi Fischer. This latter bird is more bluish and occurs side by side with livingstonii cabanisi in Tanganyika Territory. Reichenow described birds intermediate in color between the two under the name hybridus, making three so-called forms living- together and differing only in the amount of greenish or bluish color on the upperparts. I have examined a series of twelve adults from the Uluguru Mountains and find all three types represented, that is, some green birds (cabanisi); some more bluish (hybridus); and some bluish i Bull. Mus. Comp. Zool., 75, 1933, p. 165. 2 Journ. f. Orn., 82, 1934, Sonderheft, p. 53. 3 Nov. Zool.. 15, 1908, pp. 376-377. 102 bulletin: museum of comparative zoology ones {reichenowi) . However, the series is unbroken from one extreme to the other, and it would be not only impossible and arbitrary to divide it into named groups, but also unnatural and misleading. T. livingstonii cabanisi is a variable bird and hybridus and reichenowi are synonyms based on selected individuals. In my report on Loveridge's Uluguru-Usambara collection1 1 referred his series to cabanisi without any comment on the color variations as that paper was rather hurriedly written. However, it should be noted that Reichenow2 calls Uluguru specimens hybridus. Of the three, hybridus is by description the most variable, and it is therefore not impossible, although quite improbable, that I have no true cabanisi or reichenowi in my series at all, but only a variable series of the inter- mediate type. The measurements of the birds are as follows: c? wing 165-177; tail 190-200; oilmen from base 21-23 mm. 9 wing 165-179; tail 187-213; culmen from base 22-25.5 mm. These figures extend the limits of variation, beyond those given by Reichenow3 who writes as follows: wing 170-180; tail 200-215; culmen 23-25 mm. The exact range of cabanisi depends on whether we recognize reichenowi and hybridus as different, or put them all together. I incline to the latter view, in which case the bird is known to inhabit Tangan- yika Territory from the coast at Bagamoyo and Dar es Salaam south to Beira in Mozambique; inland only in Tanganyika Territory to the Uluguru Mountains, Morogoro, Kipunga, Usegua, Usagara, Mawudjo, Mbala, Maruji, Madazini, Ugogo, Mpwapwa, to the Uzungwe Moun- tains and the Mahenge district and the north end of Lake Nyasa where it intergrades with typical livingstonii. Its altitudinal range is there- fore from sea level to about 8,000 feet. The breeding season is probably rather indefinitely spread out, but a female taken at Bagilo, Uluguru Mountains, 22 September, had an ovarian egg 15 mm., in diameter. Turacus fischeri (Reichenow) Corythaixfischeri Reichenow, Orn. Centralbl., 1878, p. 88: Witu, Kenya Colony 2 , Kabale, Ruanda, Uganda, 20 September 1919. 2 adult d\ Chantwara, Bukoba, Tanganyika Territory, 5-9 January 1923. "Also Mombasa, Bungu, Dar es Salaam, Morogoro, Kilosa, and Mahaka." (A.L.) This cuckoo, like the didric, is widely distributed throughout tropical East Africa. It is somewhat more inclined to live in forest areas (not very dense forest) and consequently is more local in its range. How- ever, while occurring higher in the mountains than the didric it is found in the lowlands as well. Granvik2 obtained klaasi at 7,000 feet on Mt. Elgon; Sjostedt3 procured it "high up in the cultivated zone," that is, about 6,000 feet on Kilimanjaro; and Carruthers4 found it at 6,000 feet on Ruwenzori. Not only does klaasi replace caprius in the higher, more wooded areas in East Africa, but it largely takes its place in northeastern Africa, Ethiopia, Shoa, etc. However, the two species occur and breed in the same general localities in the lowlands so the two cannot be considered as ecological races. Furthermore Klaas's cuckoo is as nearly related to the emerald as to the didric cuckoo, as was pointed out by van Someren.5 i cf. Grant, Trans. Zool. Soc. Lond , 19, 1910, pp. 424-425. 2 Journ. f. Ornith., 1923, Sonderheft, p. 83. 3 Kilimandjaro-Meru Exp., 1910, part 3, p. 86. 4 cf. Grant, Trans. Zool. Soc. Lond., 19, 1910, p. 425. 5 Ibis, 1925, pp. 660-662. 118 bulletin: museum of comparative zoology Centropus monachus monachus Ruppell Centropus monachus Ruppell, N. Wirbelth., Vog. 1837, p. 57, pi. xxi, fig. 2: Kulla, n. Ethiopia. 1 c? , Nairobi, Kenya Colony, 6 October 1915. 1 9 , Nairobi, Kenya Colony, 21 October 1915. The male is in juvenal plumage, but is fully grown in size. The female is adult, and has the inner secondaries rather dark olive brown like the western race occidentalis. It also has the tail feathers purplish brown instead of greenish, resembling in this respect the southern cupreicaudus. This species occurs only in the northern part of the region under discussion in this paper — the Kikuyu district of Kenya Colony. The southern cupreicaudus is known from southern Nyasaland and may occur in adjacent parts of Mozambique, but has not been taken there as yet. It is a bird of dense thickets, edges of forests, etc., and, in Kenya Colony, the blue headed coucal is known only from Nairobi (probably the forest at Parklands), Kyambu, and Fort Hall. Centropus senegalensis fasciipygialis Reichenow Centropus fasciipygialis Reichenow, Orn. Monatsb., 6, 1898, p. 23: Quilimane, Mozambique. 1 d1, Morogoro, Tanganyika Territory, 15 January 1918. "Also at Dar es Salaam." (A.L.) Van Someren1 writes that a bird of the, "... senegalensis type is reported from Zanzibar (Kirk)," and a specimen is listed by Reiche- now2 as possibly burchelli or fasciipygialis. Probably because of the indefiniteness of this record, and the lack of others north of southern Tanganyika Territory (Mikindani and Lindi3 and Maliwe, near Kilwa4) Sclater5 states the range of fasciipygialis to be Beira and the valley of the lower Zambesi, north to southern Nyasaland and southern Tan- ganyika Territory. The present specimen therefore constitutes a con- siderable northward extension of the range of the race and renders more i Nov. Zool., 29, 1922, p. 50. 2 Vog. Afr., 2, 1903, p. 60. 3 cf. Grote, Journ. f. Ornith., 1912, p. 521. 1 Kothe, Zool. Ergebn, Exp. Hauptm. a. D. Fromm, 2, Aves, 1911, p. 351. s Syst. Avium Ethiop., pt. 1, 1924, p. 186. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 119 intelligible the Zanzibar record. At present Morogoro is the most northern locality for the bird. Owing to the absence of records other than those mentioned above, it is difficult to state much about the distribution and numerical status of the race, except that it is probably uncommon and local. I have not seen the Dar es Salaam bird so cannot say whether the identification was correct. The Morogoro bird has many of the hackle-like, lanceolate feathers of the sides of the throat and breast margined with blackish as in C. superciliosus. Its measurements are as follows: wing 150; tail 179; culmen 27 mm. Centropus burchellii Swainson Centropus burchellii Swainson, Anim. Menag., 1838, p. 321: South Africa (ex Burchell, Cape Province). 1 d\ 1 9 , Lumbo, Mozambique, 10 July 1918. The species Centropus senegalensis and Centropus burchellii have been curiously confounded by students of African birds. To take two recent works for example, we find the following sorts of contradictory statements. Bannerman1 writes that C. burchellii and C. fasciipygialis are not races of C. senegalensis, but that the first named is a distinct species, of which the second is a race. Both of them have very dis- tinctly barred rumps and upper tail coverts, while in senegalensis, these feathers are unbarred. Sclater2 considers them as forms of senegalensis, listing five races of that species, as follows: 1. C. s. senegalensis: Senegambia to n. Angola, east to Somaliland, the Upper Nile, and northern Uganda, skipping the equatorial forests. 2. C. s. acgyptius: Nubia and Egypt. 3. C. s. flecki: Bechuanaland to the Upper Zambesi valley and Northern Rhodesia, east to Mashonaland, occasionally south to the Transvaal. 4. C. s. burchellii: Cape Province and Natal, north to the Transvaal and to Inhambane, Mozambique. 5. C. s. fasciipygialis: Beira and the Zambesi Valley, north to southern Tanganyika Territory and southern Nyasaland. In connection with the writing of the present paper I have had the opportunity of investigating the taxonomy of these coucals, and find that neither Bannerman nor Sclater has completely arrived at the i Rev. Zool. Africaine, 10, fasc. 2, 1922, pp. 125-126. *Syst. Avium Ethiop., pt. 1, 1924, p. 186. 120 bulletin: museum of comparative zoology correct conclusions. C. fasciipygiaUs is, as the latter decided, a race of C. senegalensis, but C. burchellii is a distinct species, as the former author thought. The solution of the problem resolved itself around the discovery that C. burchellii ranges much farther to the north than hitherto suspected, inhabiting the same general area as C. senegalensis fasciipygalis. The present specimens from Lumbo constitute the northernmost records for C. burchellii yet known. Van Someren1 lists eight specimens of Centropus senegalensis flecki from Lumbo, now in the Nairobi Museum. All of these were collected by Loveridge together with the present two and may well be burchellii. C. burchellii may easily be told from flecki by the barred upper tail coverts and rump, which are greenish, unbarred in flecki. C. senegalen- sis fasciipygialis is very similar to C. burchellii, and as it occurs to- gether with it, may often be confused with the latter. The former is, however, much smaller (wing 145-155 mm.), has the top of the head earth-brown, the feathers with black shafts, has no light loreal spot, and has most of the narrow lanceolate feathers of the sides of the throat and breast edged with black as in C. super ciliosus. It also has the basal barring on the rectrices more extensive distally than in C. burchellii. Centropus superciliosus superciliosus Hemprich and Ehrenberg Centropus superciliosus Hemprich and Ehrenberg, Symb. Phys. fol. R., 1828, pi. xi: S. Arabia. 1 Ibia, 1915, p. 258. 2 Syst. Avium Ethiop., pt. 1, 1924, p. 237. » Journ. f. Ornith., 1912, p. 520. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 131 from the range of suahelica. Furthermore, van Someren1 writes that his series of twelve specimens from Kenya and Uganda show every gradation from bright golden brown to deep blackish brown, and can be matched with specimens in Tring, "... which have been placed as S. w. nuchale, bohndorffi, suahelicum, and nigricantius." It seems quite certain from the series (unfortunately small) that I have studied that nigricantia and suahelica are merely color phases of the same thing. Van Someren suggests this without actually saying so, but uses the name suahelica, which is antedated by one year by nigricantia. Claude Grant2 went even farther in concluding that no valid subspecies can be maintained as the differences seem to be merely individual variations, there being at least three phases, a blackish, a cinnamon, and a grayish one. Van Someren3 states that young birds show color phases that are not geographic; one of his birds is grayish buff, two others are sandy buff, while a third is very dark brown. The dimorphism of this species is further complicated by the fact that darkness of color appears to increase with age in all phases. Syrnium sansibaricum Reichenow4 is also a synonym of nigricantia. October seems to be the breeding season throughout Tanganyika Territory as well as Kenya Colony. Grote5 found eggs on 16 October at Lindi and at Mikindani. This owl occurs throughout the territory under discussion but is restricted to dense bush and forested areas, so that it is rather local, and its distribution discontinuous. Otus senegalensis graueri Chapin Otus senegalensis graueri Chapin, Am. Mus. Novit.; no. 412, 1930, p. 4: Lueba, northwest shore Lake Tanganyika. 1 9 , Morogoro, Tanganyika Territory, 12 December 1917. 1 d% Mkata River, Kimbwabwa's, Tanganyika Territory, 27 August 1921. "The outline and colouring of the male so blended with the bare thorn tree on which it was sitting in bright sunlight, that it appeared to be a broken off branch, so that I had the greatest difficulty in seeing it when pointed out by my col- lector." (A.L.) " 1 Nov. Zool., 29, 1922, p. 45. 2 Ibis, 1915, p. 255. 3 loc. cit. * in Werther's Die Mittl. Hochl. des nordl. Deutsch-Ost-Afr. 1898, p. 272: Zanzibar. « Journ. f. Ornith., 1912, p. 520. 132 bulletin: museum of comparative zoology The female is darker and more rufous than the male. The measurements of the two are as follows: male wing 118, tail 51, culmen from cere 11.5 mm.; female wing 119, tail 52, culmen from cere 11.0 mm. On the whole this owl seems to be less dichromatic than most, and consequently color characters are somewhat more reliable. Chapin1 writes that graueri may be expected to be found in the interior of Tanganyika Territory. "A female collected by Loveridge at Morogoro ... is very similar in color to specimens of graueri from near the type locality, but its wing measures only 119 mm. Other specimens from Tanganyika Territory are grayer." Otus leucotis granti (Kollibay) Pisorhina leucotis granti Kollibay, Orn. Monatsb., 18, 1910, p. 148: South- west Africa. 1 9, Morogoro, Tanganyika Territory, 9 July 1917. 1 cf, 1 9, Mtali's near Mkalama, Tanganyika Territory, 19 October 1921. Besides the above three specimens, Loveridge collected a female at Dodoma, Tanganyika Territory, on 3 July 1926 while with the Smith- sonian-Chrysler Expedition. This specimen (now in the United States National Museum) agrees with the other females but is slightly lighter in color. The range of this form as given by Sclater2 is too restricted. He writes, "... South Africa, north to Angola and Nyasaland," but it really extends north to north-central Tanganyika Territory. Just where it meets or intergrades with typical leucotis is not known. How- ever, the Mtali male is darker, especially on the upper throat and the crown than another from South Africa, and may therefore be con- sidered as somewhat intermediate in nature. In the character of the barring of the primaries it is like the South African bird. The two females vary in color, the Mtali specimen being noticeably darker than that from Morogoro. 1 loc. cit. 2Syst. Avium Ethiop., pt. 1, 1924, p. 242. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 133 Glaucidium perlatum (Vieillot) Strix perlata Vieillot, N. Diet. d'Hist. Nat. 7, 1817, p. 26: Senegal. 1 cf , Dar es Salaam, Tanganyika Territory, 18 November 1918. 1 9 , Mahaka, Dodoma, Tanganyika Territory, 1 April 1922. 1 9 , Kinyambwa, Dodoma, Tanganyika Territory, 7 April 1922. "Also Kilosa." (A.L.) The material of this species examined (12 specimens) is not suf- ficient to enable me to attempt a subspecific study of this owl, par- ticularly inasmuch as Claude Grant1 and Sclater and Mackworth- Praed2 were unable to recognize any races with a series six times as great as mine. Van Someren, on the other hand,3 suggests that there are two if not three races, but I cannot follow his argument. He states that there is a western form and an eastern one, but that East African birds are intermediate ! All three specimens have the crown spotted with white, but I have seen some individuals in which the spotting was confined to the fore- head, apparently an individual variation not correlated with sex, age, season, or geography. This owl is generally distributed throughout our region. The Kinyambwa bird is molting the rectrices. Glaucidium capense scheffleri Neumann Glaucidium capense scheffleri Neumann, Orn. Monatsb., 19, 1911, p. 184: Kibwezi, Kenya Colony. 1 cf, Kilosa, Tanganyika Territory, 18 January 1921. 1 9 , Wami River, Tanganyika Territory, 3 September 1921. "Also a male at Kipera. The call of this owl at sunset or dusk is a very characteristic bush sound in the Kilosa district." (A.L.) ' These specimens appear to be the westernmost ever taken in Tanganyika Territory, as birds from Kakoma, Marunge, Mkigwa, and the Unyamwesi country generally are typical capense. The present race has heretofore been known only from Northeastern Tanganyika and Southern Kenya Colony (the Pangani River to Ukamba). It differs from typical capense in having a deeper and purer > Ibis, 1915, p. 256. 2 ibid., 1919, pp. 681-682. 3 Nov. Zool., 29, 1922, p. 46. 134 bulletin: museum of comparative zoology brown color, in having the light bars on the back absent or only poorly developed, and in having the color of the occiput separated from that of the back by an area of yellowish and dark brown bars. The measurements of the two birds listed above are : c? , wing 137, tail 83, culmen from cere 13 mm. 9 , wing 137, tail 83.5, culmen from cere 12.5 mm. Bubo africanus africanus (Temminck) Strix africana Temminck, PI. Col. livr. 9, 1823, pi. l: Cape of Good Hope. 1 o", 1 9, Morogoro, Tanganyika Territory, 9 June 1917. 1 d", Lumbo, Mozambique, 3 September 1918. "Also one from Kilosa where it was very abundant." (A.L.) The male from Lumbo and the female from Morogoro are of the brownish phase while the male from Morogoro is grayish, but even the latter is more brownish than cinerascens of farther north. Claude Grant1 writes that Asio inaculosus amerimnus (not americ- anus as he misquotes it) of Oberholser appears to be a synonym of Bubo africanus cinerascens, but if the original description of amerimnus be examined, it will be found that the bird in question is africanus and not cinerascens, as the type came from Natal, South Africa. The Kilimanjaro specimen listed by Oberholser is not the type as Claude Grant appears to feel. Furthermore, africanus occurs as far north as Nairobi and Nakuru in Kenya Colony, so the Kilimanjaro birds must belong to this race. This owl is widely distributed throughout the region under dis- cussion. Bubo lacteus (Temminck) Strix lactea Temminck, PI. Col. livr. i, 1824, p. 4: Senegal. 1 9 , Kilosa, Tanganyika Territory, 14 April 1921. "Also a female from Tindiga. Seen at Kilimatinde. It is not uncommon in the Kilosa district." (A.L.) The specimen listed above is an adult in fresh plumage. Verreaux's eagle-owl occurs throughout the region under discussion but appears to be numerous nowhere. It does not stray far from trees, usually fairly large ones with rather dense foliage, in the shade of which it passes most of the daytime. i Ibis, 1915, p. 252. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 135 Order CAPRIMULGIFORMES Family CAPRIMULGIDAE. Goatsuckers Caprimulgus europaeus meridionalis Hartert Caprimulgus europaeus meridionalis Hartert, Ibis, 1896, p. 370: Greece {vide Vog. pal. Fauna, p. 848)'. 1 d\ Kilosa, Tanganyika Territory, 4 February 1921. 1 9 , Suna, Singida, Tanganyika Territory, 27 February 1922. Meinertzhagen has reviewed the subspecies of this nightjar in detail1 and I find his conclusions are supported by the material in the Museum of Comparative Zoology. Meridionalis differs from typical europaeus chiefly in size (the wing length of adult males being 189-204 in europaeus and 174-189 mm. in meridionalis), and while the present male is rather large (wing 188 mm.) yet it is clearly referable to the Mediterranean subspecies. It has no white mark on the outer web of the next to the outermost primary and neither does the female, al- though not infrequently such a mark is present in this form. Both birds are in fresh plumage, but the outermost primaries are only partly grown and still enclosed in their sheaths for at least half their length. This race does not appear to migrate as far to the south in winter as either the typical form or the central Asian one, unwini. According to Sclater,2 this subspecies is not known from south of the Pangani River in Tanganyika Territory, so the present records con- stitute new southern limits. Caprimulgus europaeus unwini Hume Caprimulgus unwini Hume, Ibis, 1871, p. 406: Agrore valley, Hazara district, n. w. India. 1 cf , Morogoro, Tanganyika Territory, 24 September 1917. This specimen appears to be unwini but I am not entirely certain as I have seen almost no comparative material of this race. It fits the description given by Hartert3 and has a wing length of 196 mm. As far as I have been able to ascertain, this is the first time this bird 1 Ibis, 1922, pp. 43-48. 2 Syst. Avium Ethiop., pt. 1, 1924, p. 248. s Vog. pal. Fauna, 7, 1912, p. 849. 136 bulletin: museum of comparative zoology has been taken in Tanganyika Territory, although van Someren1 has listed it from as nearby in Kenya Colony as Tsavo and Mombasa. It has also been taken in Natal. This example is in fairly worn plumage, a fact, which, taken into consideration with the date of its capture, indicates that in this bird the postnuptial molt occurs after the migration and not before it. Caprimulgus fossei fossei Hartlaub Caprimulgus fossei Hartlaub, Orn. Westafr., 1857, p. 23: Gaboon. 1 d\ Lumbo, Mozambique, 15 August 1918. A second specimen collected at Lumbo was referred to C. fossei mosambiquus by van Someren2 but I cannot see that mosambiquus differs constantly in any way from typical fossei. The present example is really more or less intermediate between fossei and clarus, which latter I consider a race of the former, and not a distinct species as does van Someren. In East Africa this race is chiefly coastal in its distribution, its range extending northward even as far as Zanzibar, but usually stopping at about Mikindani. The limits of its range inland and those of clarus are not yet known. Bowen3 considers mosambiquus as distinct from fossei, which latter race he restricts to Gaboon. I am not convinced that the two are really separable. Caprimulgus fossei clarus Reichenow Caprimulgus clarus Reichenow, Journ. f. Ornith., 1892, p. 29: Bukoba, Lake Victoria. 1 9, Morogoro, Tanganyika Territory, 23 August 1917. 1 9 , Mombasa, Kenya Colony, 31 May 1918. 1 d\ 1 9 , Dar es Salaam, Tanganyika Territory, 7 June 1918. "Also Kilosa, Tanganyika Territory." (A.L.) Sclater4 gives the range of clarus as the, "... coastal districts of Kenya Colony inland to the eastern shores of Victoria Nyanza," but the type locality is on the western shore of Lake Victoria and van i Nov. Zool., 29, 1922, p. 84. 2 Nov. Zool., 29, 1922, p. 86. » Proc. Acad. Nat. Sci. Phil., 83, 1931, pp. 40-42. * Syst. Avium Ethiop., pt. 1, 1924, p. 253. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 137 Someren1 records the bird from Jinja, Kampala, and Bugoma, in Uganda. The range should be further extended to include the inland parts of the northern half of Tanganyika Territory (Morogoro, Kilosa, to north end of Lake Tanganyika). Another Morogoro bird was referred to C. fossci mosambiqmis by van Someren2 but is really clarus. Sjostedt3 lists a specimen of apatelius from Mt. Kilimanjaro (based on Neumann's statement4) but this is undoubtedly a mistake and should be referred to clarus. Van Someren5 has come to this conclusion. COSMETORNIS VEXILLARIUS (Gould) Semeiophorus vexillarius Gould, Icon. Av. pt. 2, pi. iii, 1838: Sierra Leone. 1 d\ Kilosa, Tanganyika Territory, 3 January 1921. "While sitting up in a tree one night waiting for a leopard at Mbweni, I enjoyed the interesting spectacle of seeing some three or four of these nightjars flying beneath me. The birds themselves were almost, if not entirely, invisible, but their white pennants were floating and flickering about in the gloom, in quite an uncanny manner." (A.L.) The single specimen collected is an adult male in full breeding plumage with very long "pennants." Chapin6 has discussed the breeding season and the migrations of this remarkable goatsucker, and has found that the species does not nest north of the equator, but migrates as far as the southern Sudan (March to July inclusive). The breeding and non-breeding ranges appear to overlap, or at least to meet, in northern Tanganyika Territory, so it appears likely that this bird is more or less of a permanent resident in that part of our area. The present bird from Kilosa is of interest in that it seems to be the northeasternmost breeding locality record, the nearest one to it indicated on Chapin's map being Iringa, about 100 miles to the south- west of it. 1 Nov. Zool., 29, 1922, p. 86. 2 cit. supra. 3 Kilimanjaro-Meru Exp., 1910, p. 102. « Orn. Montsb., 12, 1904, p. 143. 6 loc. cit. 6 Bull. Amer. Mus. Nat. Hist., 35, 1916, pp. 73-81. 138 bulletin: museum of comparative zoology Order MICROPODIFORMES Family MICROPODIDAE. Swifts Micropus affinis abessynicus (Streubel) Cypselus abessynicus Streubel, Isis, 1848, col. 354: Abyssinia. 1 d\ 1 9, Mombasa, Kenya Colony, 20 May 1918. 1 9 , Kilosa, Tanganyika Territory, 5 May 1922. "Also Morogoro." (A.L.) The female from Kilosa was in breeding condition when collected. It has the white throat patch smaller than in the Mombasa female; similar to the Mombasa male. The Kilosa bird is also smaller than either of those from Mombasa. The former has a wing length of 124.5 mm., while the latter measure 130 (d71) and 128.5 mm. (9), re- spectively. This swift is found throughout the region under discussion, but appears to be more or less uncommon throughout much of its range. Cypsiurus parvus myochrous (Reichenow) Cypselus myochrous Reichenow, Journ. f. Ornith., 1886, p. 116: Karema, Tanganyika Territory. 1 adult 9 , Morogoro, Tanganyika Territory, 17 July 1917. 1 immature 9 , Dar es Salaam, Tanganyika Territory, 10 January 1919. "Also Kilosa." (A.L.) The immature bird has the upper wing coverts, scapulars, and feathers of the back and rump edged with rusty rufous, and has no whitish on the upper throat as in adult birds. The palm swift is widely distributed throughout eastern Africa, but as far as known, does not breed in any but palm trees, so that its breed- ing range is necessarily rather discontinuous or "spotty." Loveridge's notes on the curious nesting habits of the palm swift have been published fully in the Memoirs and Proceedings of the Manchester Literary and Philosophical Society, 61, 1919, part 2. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 139 Order COLIIFORMES Family COLIIDAE. Colies Colius striatus affinis Shelley Colius leucotis affinis Shelley, Ibis, 1885, p. 312: Dar es Salaam. 1 cf, 1 9, Morogoro, Tanganyika Territory, 20 August 1917. 1 cf , 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 20 May 1920. 1 d\ 1 9 , Kilosa, Tanganyika Territory, 2-3 February 1921. "Also Dar es Salaam." (A.L.) The pair from Morogoro are erythristic and at first sight are ex- tremely different from the others. They are also slightly smaller in size. The wing lengths of the birds are as follows: Morogoro, cf 88, 9 87; Kilosa, d" 93.5, 9 91; Bagilo, cf 89, 9 87 mm. The exact western limits of the range of affinis are not known. The form berlepschi, which is nearest to affinis, but darker generally, occurs, in the regions here under discussion, only in Southwestern Tanganyika Territory. Apparently the area of intergradation between the two is very extensive, as I know of no undoubted examples of affinis from any locality west of Kilosa. The form cinerascens Neumann, described from Irangi, is based on an intermediate, but being nearer to affinis than to berlepschi may be better synonymized with the former than the latter. Colius striatus mombassicus van Someren Colius striatus mombassicus van Someren, Bull. Brit. Orn. CI., 40, 1919, p. 26: Changamwe. 1 d", Mombasa, Kenya Colony, 31 May 1918. Sclater1 regards mombassicus as a synonym of affinis, but I find that the characters given by van Someren hold. I have seen three examples of mombassicus and eight of affinis. The former race differs from the latter in being lighter, more grayish, on the head, neck, and mantle, and more white on the cheeks. The wing length in mombassicus ranges from 87-92 mm. ; that of the present specimen is 91 mm. This race inhabits the coastal area of Kenya Colony from Mombasa 1 Syst. Avium Ethiop., pt. 1, 1924, p. 265. 140 bulletin: museum of comparative zoology (inland to Changamwe and Voi) north to Lamu, Kismayu, and southern Somaliland. The single specimen collected was molting the remiges and rectrices, but no ecdysis is visible on the head and body. Colius striatus jebelensis Mearns Colius striatus jebelensis Mearns. Proc. U. S. Nat. Mus., 48, 1915, p. 394: Gondokoro. 1 o", Kabura, Mawokota, Uganda, 27 August 1919. 1 9 , Kabare, Bukoba, Tanganyika Territory, December 1922. Colius s. ugandensis is a synonym. Hartert1 writes that kikuyuensis is really recognizably distinct from ugandensis, and a good series of both supports the supposed characters of the former. This is the coly of Uganda and Northwestern Tanganyika Territory, merging in the Mt. Elgon region with kikuyuensis. Colius striatus kikuy, ensis van Someren Colius striatus kikuyuensis van Someren, Bull. Brit. Orn. CI., 40, 1919, p. 27: Nairobi. 1 immature d\ Nairobi, Kenya Colony, 2 June 1919. 1 d", 1 9 , Ngong, Nairobi, Kenya Colony, 14 July 1919. "Also Tumutumu, Kenya Colony." (A.L.) The immature bird is so very young that it must have been taken from the nest. It is just acquiring the first pennaceous feathering, which resembles that of the adult on the underparts, the remiges and rectrices, and the head. The upper wing coverts are broadly tipped with sandy rufous, and the back is broadly striped sandy rufous buff and fuscous brown. This race is darker and larger than jebelensis. On 7 June 1919 a nest and three eggs were found at Nairobi. Colius indicus pallidus Reichenow Colius indicus pallidus Reichenow, Orn. Monatsb., 1896, p. 4: Kionga, Rovuma River. 1 o",l 9, Lumbo, Mozambique, 15 July 1918. As far as I am able to ascertain, these two specimens constitute the southermost records for the race. They agree with the description of « Nov. Zool., 31, 1924, p. 128. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 141 pallidas which I have not otherwise seen, but may be somewhat inter- mediate between that form and the typical race. Both specimens are adults in fairly fresh plumage. The wing mea- surements are 91.5 mm., in the male, and 89 mm., in the female. Loveridge1 found a dozen or so nests at Lumbo in October. Nearly all of them contained highly incubated eggs or young. The clutch was invariably three. Colius macrourus pulcher Neumann Colius macrourus pulcher Neumann, Journ. f. Ornith., 1900, p. 190: Teita. 1 9 , Morogoro, Tanganyika Territory, 26 November 1918. "Also seen at Dar es Salaam." (A.L.) Sclater2 gives the distribution of this bird as "Baro River and S. W. Abyssinia, south to eastern Uganda and Kenya Colony." This should be extended to include the northeastern parts of Tan- ganyika Territory to Morogoro and Dar es Salaam. It appears to be uncommon in the southern part of its range. Order TROGONIFORMES Family TROGONIDAE. Trogons Apaloderma narina narina (Stephens) Trogon narina Stephens, in Shaw's Genl. Zool., 9, 1815, p. 14: Anteniquoi, that is, Knysna district (ex Levaillant). 1 9 , Nairobi, Kenya Colony, 18 October 1915. 1 9 , Ngong Forest, Kenya Colony, 14 July 1919. 1 d\ Ngong Forest, Kenya Colony, 16 September 1920. "Also Morogoro, Tanganyika Territory." (A.L.) On the back of the label of the last specimen listed, the collector wrote, "Keeps to thick forest and is very rarely seen. This is the third specimen I have met with." This trogon is found throughout our region but as it is purely sylvan in habitat, its range is very discontinuous and patchy. i Proc. Zool. Soc. Lond., 1922, pp. 851-852. 2 Syst. Avium Ethiop., pt, 1, 1924, p. 268. 142 bulletin: museum of comparative zoology The female from the Ngong Forest was molting into adult plumage. It still has a few juvenal inner greater upper wing coverts (with large buffy white spots). The under tail coverts are pale grayish, although Reichenow1 writes that in young birds the belly and under tail coverts are pale red. The male was molting the rectrices, the molt being centrifugal. Recently van Someren2 has described a coastal race from the Sokoke Forest, Kenya Colony, littoralis, said to be smaller than narina. Heterotrogon vittatum vittatum (Shelley) Hapaloderma vittatwn Shelley, Proc. Zool. Soc. Lond., 1882, p. 306: Mamboia, Tanganyika Territory. 1 9, Uluguru Mountains, Tanganyika Territory, 19 May 1921. 1 cf1, Bungu, Usambara Mountains, Tanganyika Territory, September 1921. A series of 17 specimens reveals several hitherto unknown facts about the plumages of this rather uncommon trogon. The juvenal males resemble the females on the underparts and have the belly and under tail coverts reddish pink, the breast and throat tawny brown. Just as the juvenal stage of Apaloderma narina the inner greater upper wing coverts and tertials with large buffy white spots, so the corre- sponding plumage of the present species has these feathers with white spots, while the majority of the upper wing coverts are finally barred with white as in the adults. The postjuvenal molt is complete. The tail molt is centrifugal. This trogon occurs in mountain forests throughout the country represented by the present collection. Van Someren3 attempts to separate Uganda birds as H. v. minus (Chapin), but this name is a synonym of Apaloderma aequatoriale Sharpe, and is therefore not available. In a more recent paper van Someren4 calls these Uganda birds camerunensis Reichenow. H. v. keniensis Bowen, described from the Meru area requires further confirmation before it can be accepted definitely as a valid race. Van Someren's material did not uphold it. i Vog. Afr., 2, 1903, p. 213. 2 Bull. Brit. Orn. CI., 61, 1931, p. 80. a Nov. Zool., 29, 1922, p. 73. * Nov. Zool., 37, 1932, p. 286. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 143 Order CORACIIFORMES Family ALCEDINIDAE. Kingfishers Ceryle rudis rudis (Linnaeus) Alcedo rudis Linnaeus, Syst. Nat. 10th ed., 1758, p. 116: Egypt (ex Hasselquist) 1 cf, 1 9 , Dar es Salaam, Tanganyika Territory, 13 June 1918. "Also Kilosa. Towards evening at Dar es Salaam, king- fishers of this species which, during the day have been scat- tered along the coast engaged in fishing, assemble in large flocks and fly up the rivers to roost. A very common bird." (A.L.) This kingfisher is widely distributed throughout the territory under discussion in this paper, but is not found away from water. It is one of the true fish-eating kingfishers, and unlike the species of Halcyon does not feed on insects at all. The male has the broad pectoral black band continuous across the breast while the female has it interrupted medially by white. The former sex also has a second, more posterior, narrow black pectoral band which the latter lacks. The female is the longest-billed bird of a series of 40 specimens examined, having a culmen length of 63 mm. Megaceryle maxima maxima (Pallas) Alcedo maxima Pallas, Spic. Zool. fasc. vi, 1769, p. 14: Cape of Good Hope. 1 9 , Kilosa, Tanganyika Territory, 21 April 1921. 1 , 1 9 , Dar es Salaam, Tanganyika Territory, 8 November 1918. 1 9 , Rutaka, Ruanda, Uganda, 5 October 1919. 1 cf, Kome Island, Mwanza, Tanganyika Territory, 25 November 1922. 1 9 , Kabare, Bukoba, Tanganyika Territory, 17 January 1923. "Also Tindiga." (A.L.) This series shows considerable variation in the width of the black pectoral band and in the color of the upper tail coverts which are darker and bluer in some birds and lighter and greener in others. The two specimens from Dar es Salaam are in very worn plumage, the others are fresh. The typical, western subspecies of this bee-eater, has a smaller bill than maior, but the other characters ascribed to it do not hold with any degree of constancy. In tropical East Africa this bird occurs south to the vicinity of Dar es Salaam, but not beyond that point. There appear to be no known i Journ. f. Ornith., 1905, pi. ix, fig. 1. 154 bulletin: museum of comparative zoology instances of the species breeding anywhere in Tanganyika Territory, and Sclater1 states that it breeds only in the northern half of its range. Aerops boehmi (Reichenow) Merops (Melittophagus) boehmi Reichenow, Orn. Centralbl., 1882, p. 62; Bumi, Tanganyika Territory. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 1 9 , Tindiga, near Kilosa, Tanganyika Territory, 25 August 1922. This rare species has not hitherto been known from central and northeastern Tanganyika Territory, the previously known range hav- ing been confined to the southwestern portion of the country from Msima, the Katuma River, and Bumi to the Rovuma River, Langen- burg (i. e. Manda), and Nyasaland. The present specimens extend the range northwards and eastwards for about 400 miles. Reichenow2 gives the range of this species as follows: wing 80; tail with middle feathers, 120-130, without middle feathers 70-75; bill 27-30 mm. The present two specimens have slightly shorter wings, longer middle rectrices, and longer bills. They measure as follows: wing 76-77; tail with middle rectrices 130-140, without middle feathers 68-70; bill 31-32 mm. While it may be that the present birds represent a different, unnamed form, I hesitate to describe them in the absence of any comparative material. Melittophagus pusillus meridionalis Sharpe Melittophagus meridionalis Sharpe, Cat. Bds. Brit. Mus., 17, 1892, p. 45, pi. i, fig. 4: type in British Museum from Pinetown, Natal. 1 cf , 1 9 , Dar es Salaam, Tanganyika Territory, 13 June 1918. 1 9 , Masaka, Buganda, Uganda, 30 August 1919. 1 cf1, Kabura, Mawokota, Uganda, 27 August 1919. 1 d71, Singo, Ruanda, Uganda, 25 September 1919. 1 d\ Ilonga, Kilosa, Tanganyika Territory, 22 March 1923. "Also Bagilo, Morogoro, Kilosa, Tanganyika Territory, and Lumbo, Mozambique." (A.L.) This race occurs throughout Tanganyika Territory except in the Kilimanjaro area and east from there to the coast south to the Pan- i Syst. Avium Ethiop., pt. 1, 1924, p. 221. 2 Vog. Afr., 2, 1903, p. 319. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 155 gani River, in which area it is replaced by cyanostictus. The two forms may be told apart by the fact that meridionalis has no blue band on the forehead while cyanostictus has. An inhabitant of the thorn-bush country, this bird is frequently found far from water, but is equally common along streams and the sea coast. Melittophagus pustlltjs cyanostictus (Cabanis) Merops cyanostictus Cabanis, in von der Decken's Reisen, 3, 1869, p. 34: Mombasa. 1 9 , Mombasa, Kenya Colony, 25 May 1918. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. "Intermediates between cyanostictus and meridionalis occur atKilosa." (A.L.) The female from Mombasa is in juvenal plumage. This subspecies occurs from British Somaliland, Shoa, Arussiland, and Gallaland, south through Kenya Colony to the Kilimanjaro district and along the Tanganyikan seacoast to the Pangani River, south of which it is represented by meridionalis. In its ecological habitat this bird is similar to meridionalis. Melittophagus lafresnayii oreobates Sharpe Melittophagus oreobates Sharpe, Ibis, 1892, p. 320: Save, Mt. Elgon. {i.e. Sabei, northern slopes, fide Loveridge, 1936). 1 adult cf, 1 immature cf, 1 adult 9, Ngong, Nairobi, Kenya Colony, 28 July 1919. "Also Tumutumu and Morogoro." (A.L.) This bird occurs in the highland districts from Ruwenzori to Elgon, N. Guaso Nyiro, Nairobi, Kilimanjaro, and the Usambara Mountains. In the western part of its range it extends south to the mountain forests west of Lake Tanganyika where Grauer obtained specimens. Sclater1 seems to have overlooked Sassi's report on Grauer's birds2 as he gives only Ruwenzori as the western limit of the range of this bee-eater. Granvik3 has pointed out that there are two color types of this bird, 1 Syst. Avium Ethiop., pt. 1, 1924, p. 222. 2 Ann. K. K. Naturhist. Hof. 1912, p. 370. * Journ. f. Ornith., 1923, Sonderheft, pp. 107-108. 156 bulletin: museum of comparative zoology a greenish one and a more bluish phase. The series examined by me (14 specimens) corroborates this and shows that there is no geographi- cal or sexual significance to this dimorphism. It appears to be a matter of age as Granvik suggested. He recorded that in the blue type the middle rectrices are blue and all the others have a blue edge. A male from Guaso Nyiro, Kenya Colony (Mus. Comp. Zool. 56464) has the middle rectrices blue and all the others green with no blue edges. The central pair of feathers are worn, the others are new. It appears, there- fore, that in first adult plumage, the blue color is well developed, and that in older birds it is absent. According to van Someren,1 this species breeds in December, February, and July in Kenya Colony where it is very common. He also notes that it is apt to make small, local migrations, a fact which should be kept in mind with reference to the few Tanganyikan records. Morogoro appears to be the southernmost point from which the bird has been reported as yet. Melittophagus variegatus loringi Mearns Melittophagus variegatus loringi Mearns, Proc. U. S. Nat. Mus., 48, 1915, p. 393: Butiaba, Lake Albert, Uganda. 1 9 , Kome Island, Mwanza, Tanganyika Territory, 27 November 1922. This race is not too well differentiated from the typical western form. The differences are only average ones, but the subspecies is recogniz- able. It differs from variegatus in being slightly bluer on the forehead and in having darker blue superciliaries. In his original description of this form Mearns gave the characters as " . . . larger and much paler and purer yellow on the underparts," than in variegatus. He had seen but one adult of the latter race at the time, however. Grant2 correctly showed that the characters given by Mearns did not hold, but that on the basis of other slight differences, such as the color of the forehead and superciliaries, the form loringi could be maintained. This specimen is the second locality record for the race from Tan- ganyika Territory. Emin collected it at Bukoba just on or across the border from Uganda, and so did Grauer many years later. Gyldenstolpe collected several in Ruanda, and it seems that the bird probably occurs throughout northwestern Tanganyika Territory. Mwanza is the southeasternmost point at which it has been taken. i Ibis, 1916, p. 246. 2 Ibis, 1915, pp. 297-298. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 157 Melittophagus bullockoides (Smith) Merops bullockoides A. Smith. S. A. Quart. Journ. 2nd ser., 1834, p. 320: S. Africa. 1 cf , 1 9 , Ngare Mtoni, Arusha, Tanganyika Territory, 17 April 1916. "Also Kedong Valley, Kenya Colony." (A.L.) This species is widely distributed in Tanganyika Territory, Mozam- bique, and southward, but appears to be numerous nowhere. In Kenya Colony van Someren1 records it from Nakuru, Naivasha, KisumUj and the Turkwell River, the latter two records constituting an extension of the range as given by Reichenow.2 Both specimens have newly molted rectrices, some of the tail feathers being not quite fully grown, and still basally enclosed in their sheaths. The male is slightly smaller than the female having a wing length of 111 as against 115 mm. A series of nine birds from Kenya Colony and Tanganyika Territory are somewhat darker above and below than two from South Africa and Damaraland, but the difference is rather slight. Dicrocercus hirundineus hirundineus (Lichtenstein) Merops hirundineus A. A. H. Lichtenstein, Cat. rer. rar., 1793, p. 21: no loc: Orange River (ex Levaillant, Hist. Ois. Paradis, 1806, p. 36). 1 a* , 1 9, Lumbo, Mozambique, 6 August 1918. "Also a pair at Morogoro on 15 June 1917." (A.L.) This bee-eater is scarce in collections and figures relatively little in the literature of East African birds. It seems therefore that it is prob- ably local and uncommon throughout its range. Sclater3 gives Dar es salaam as the northernmost locality for the typical race, but van Someren4 writes that, "... specimens from Vanga undoubtedly belong to the southern race . . . ," which extends the range of hirun- dineus northward for about 250 miles. Both specimens are in fresh, adult plumage. i Nov. Zool., 29, 1922, p. 80. 2 Vog. Afr., 2, 1903, pp. 311-312. 3 Syst. Avium Ethiop., pt. 1, 1924, p. 224. • Nov. Zool., 29, 1922, p. 80. 158 bulletin: museum of comparative zoology Family CORACIIDAE. Rollers Coracias garrulus garrulus Linnaeus Coracius garrulus Linnaeus, Syst. Nat. 10th ed., 1758, p. 107: Europe: re- stricted type locality, s. Sweden. 1 9 , Parklands, Nairobi, Kenya Colony, 29 October 1915. 1 d\ 1 9, Morogoro, Tanganyika Territory, 10 December 1917. "Also Ilonga, Kilosa, and Zengeragusu." (A.L.) The European roller is a common winter visitor from the north throughout all of tropical and Southern Africa except in forested areas. All three specimens are in very worn plumage, while birds collected in Kenya Colony in March by the senior author are in fresh plumage. Van Someren1 writes that in that country the birds are in full fresh plumage in February and March. It follows that the molt takes place in late December, January, and early February. The Nairobi bird is young, having the top of the head greenish, the undersides paler, and the upper parts generally lighter than the adults. It is molting the feathers of the interscapulars and back, the new brown feathers being much darker than the old ones. This species is very noticeable on account of its choice of exposed perches. Along the railway lines it is often seen resting on the telegraph wires; elsewhere it chooses the top, outermost branches of the Acacia trees. It is a bird that every traveler sees during the northern winter and consequently it is widely known throughout Africa. It winters in Kenya Colony, Uganda, and southwards, but appears to be wholly a migrant in the Sudan. Meinertzhagen2 writes that only typical garrulus migrates to East Africa, although semenowi breeds in the valley of the Jordan, not far from the breeding grounds of garrulus (coastal plains of Palestine and the Judea highlands). The migrants arrive in the Sudan early in September and occur throughout October at the end of which month practically all have gone south; in Northern Somaliland in the middle of October and in Novem- ber; in Kenya Colony they arrive in late October and early November, and while most of them go through to more southern areas, many of them winter. Meinertzhagen writes that in Kenya Colony, birds of the year arrive about a fortnight before the adults. « Nov. Zool., 29, 1922, p. 73. « Ibis, 1922, pp. 50-51. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 159 On the way north, the birds from South Africa and Rhodesia pass through Tanganyika Territory and Kenya Colony during March and early April. Practically all have left tropical East Africa by the second week in April. The migration routes followed in spring and autumn are somewhat different. In autumn the main flight is down the valley of the Nile although birds come across from Arabia into southern Somaliland and down the east coast as well. In the spring, however, relatively few birds go north through Egypt; the majority appear to follow the coast and then fly across the Red Sea. Although not enough data are available for a serious study of the migrations of this roller, the following suggestion may be of interest. The heavy rains in the Sudan occur during July, August, and September and consequently vegetation is at its best and insects most numerous at that time. During this season the rollers come through and undoubtedly find an abundance of food. During March and April the food supply is not as rich, the birds are in more of a hurry, and cover longer distances between halts for feeding, and are, consequently, less attracted by the valley of the Nile. Coracias spatulatus Trimen Coracias spatulatus Trimen, Proc. Zool. Soc. London, 1880, p. 31: Leshumo valley, near Victoria Falls. 1 9 , Kipera, Kilosa, Tanganyika Territory, 22 February 1923. The outer rectrices are not fully grown in this specimen, being less than an inch longer than those of the next pair. I have seen three Angolan specimens and cannot find any constant characters on which to uphold the validity of dispar, which I therefore consider a synonym. Sclater1 lists it as doubtfully distinct. The Racquet-tailed Roller is distributed throughout the interior of Tanganyika Territory from the Unyamwesi and Ugogo districts through the Usegua, Ugalla, and Kakoma country south to Nyasaland, Southern Rhodesia, and Mozambique. It also occurs in the Kasai and Katanga districts of the Belgian Congo, and in Northern Rhodesia. Kothe2 records it from the Urungu area where it seems to be com- moner than in the rest of Tanganyika Territory, as many as eight being Been at one time. 1 Syst. Avium Ethiop., pt. 1, 1924, p. 207. 2 Zool. Erg. Exp. Hauptmann a. D. Fromm, 1911, 2, Aves, pp. 354-355. 160 bulletin: museum of comparative zoology Coracias weigalli Dresser Coracias weigalli Dresser, Ann. Mag. N. H., 1890, p. 351: Newala, north of Rovuma River. 1 cf, Kipera, near Kilosa, Tanganyika Territory, 7 August 1922. According to Sclater,1 weigalli is apparently based on a young spatu- latus. However, the present specimen certainly does not look like a young bird. Furthermore, Reichenow2 states that the young of spatulatus has the upperparts paler than the adults, the outer tail feathers not elongated, and the three outermost ones cinnamomeous on the terminal parts of the outer webs. This specimen has the back practically the same shade as an adult spatulatus (somewhat more greenish on the head), has no cinnamon-brown on any of the rectrices and the outermost pair are much elongated, exceeding the next pair by 70 mm. Until convincing proof of the identity of the two species is brought forth, it seems more in keeping with what little we know to regard them as separate. It must be admitted that it is rather curious that two such closely related forms should inhabit the same country. Coracias weigalli seems to be a rare bird in collections. Reichenow3 gives only two localities — Newala and Lumbuti. I know of no other records. Grote4 lists weigalli in his paper on the birds of Southeastern Tanganyika Territory, but refers only to Newala on the authority of Weigall. The present specimen then seems to be the third one known. (Loveridge collected another also at Kipera, which he presented to the Tring Museum, making four in all.) It extends the known range of this bird northwards for about 300 miles. Since the above was written, Bangs and Loveridge5 have concluded that weigalli is not a young Coracias spatulatus, a happy confirmation of my remarks above. Co'racias caudatus caudatus Linnaeus Coracias caudatus Linnaeus, Syst. Nat. 12th ed., 1, 1766, p. 160: Angola. 1 d\ 1 9, Kongwa, Tanganyika Territory, 23 April 1917. 1 d1, Mbala, Kilosa, Tanganyika Territory, 27 February 1923. "Also Morogoro, Kipera, Kinyambwa, and Lalago, Tangan- yika Territory, and Lumbo, Mozambique." (A.L.) 1 Syst. Avium Aethiop., pt. 1, 1924, p. 20C. 2 Vog. Afr., 2, 1903, pp. 222-223. 3 loc. cit. * Journ. f. Ornith., 1912, p. 525. » Bull. Mus. Comp. Zool., 75, 1933, p. 171. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 161 Neumann1 separated East African from typical Angolan and South African birds under the name suahelicus on the basis of darker rump, upper tail coverts and lesser upper wing coverts. I have seen no birds from Angola but three adults from South Africa (which, according to Neumann should be typical caudatus) vary greatly, two of them being exactly like East African birds, while one has a light blue rump. One East African bird also has a light rump, but most of the series examined are dark. Claude Grant2 writes that none of the characters for sua- helicus hold true, and therefore synonymizes it with caudatus, a result with which my observations agree. The Mbala bird is darker than either of the others both above and below. This roller is common and widely distributed over the whole of the territory under discussion in this paper, excepting, of course, forested areas. Loveridge3 found this roller nesting at Bissel, Kenya Colony, on 20 December 1915, and at Lumbo, Mozambique, in October 1918. Coracias naevius naevius Daudin Coracias naevia Daudin, Trait6, 2, 1800, p. 258: Senegal. 1 9 , Sanga, Mwanza, Tanganyika Territory, 16 October 1922. 1 cf, Mwanza, Tanganyika Territory, 14 November 1922. "Also Suna and Sagayo. Not nearly as common in Tangan- yika Territory as C. garrulus and C. caudatus. I have seen one on the telegraph wires near the Central Railway at Tabora, which, with Suna, is the most easterly point at which I have encountered this roller." (A.L.) C. n. sharpei is a synonym. These two specimens are quite unlike, the female being much more brightly colored below with much narrower white shaft stripes on the feathers, and with the crown more rufous-purplish, the back more greenish, less brownish than the male. The female has the crown mixed greenish and brownish in about equal proportions. The collector's notes indicate that this bird is scarce in eastern Tanganyika Territory, but it is not absent, as Kirk obtained a specimen at Dar es Salaam, Bohm at Igonda, and Fischer at Soboro. It does i Journ. f. Ornith., 1907, p. 593. 2 Ibis, 1915, pp. 261-262. »Proc. Zool. Soc. Lond., 1922, p. 849. 162 bulletin: museum of comparative zoology not seem to occur far south, however, the latitude of Dar es Salaam being its southern limit. In Kenya Colony it is fairly widespread, and it is found in northern and northeastern Uganda as well. The measurements of the birds are as follows : male — wing 190, tail 151, culmen 42 mm. female — wing 180, tail 135, culmen 37.5 mm. Eurystomus afer suahelicus Neumann Eurystomus afer suahelicus Neumann, Journ. f. Ornith., 1905, p. 186: Tschara, (i.e. Charra) Tana River. 1 d% Morogoro, Tanganyika Territory, 24 October 1917. 1 9 , Kilosa, Tanganyika Territory, 13 December 1920. 1 immature d% Kilosa, Tanganyika Territory, 21 January 1921. "Very common at Kilosa, far more so than at Morogoro." (A.L.) The young bird agrees quite well with the description given by Reichenow.1 The only noteworthy feature presented by it not men- tioned by Reichenow is that the feathers of the flanks and under tail coverts have dark, blackish shaft streaks. These are also present in two other young birds in the Museum of Comparative Zoology, but are better developed in the present one than in either of the others. Inas- much as one of the others is a male and one a female, this difference cannot be sexual. This broad-billed roller is widely distributed over most of the territory represented by the present collection. In Uganda (east to Mt. Elgon) it is replaced by another race, rufobuccalis, which has the cheeks more rufescent, less violaceous, in color. Sclater2 gives the range of suahelicus as including the Eastern Belgian Congo. However, Sassi3 records rufobuccalis from Uvira, Usumbura, Kissenji, Beni, Irumu, etc. and an intermediate between rufobuccalis and suahelicus from Beni. On the other hand, Gyldenstolpe4 writes that birds from the Semliki valley and the Eastern Belgian Congo are intermediate between rufobuccalis and typical afer! i Vog. Afr., 2, 1903, p. 229. 2 Syst. Avium Ethiop., pt. 1, 1924, p. 209. 3 Ann. K. K. Naturhist. Hofmus. Wien, 26, 1912, pp. 364-365. « Kungl. Sv. Vet. Akad. Handlngr., 1924, pp. 280-281. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 163 Family UPUPIDAE. Hoopoes Upupa africana Bechstein Upupa africana Bechstein, Kurze Uebers., 4, 1811, p. 172: Congo to the Cape. 1 d\ Morogoro, Tanganyika Territory, 27 June 1917. 1 9 , Morogoro, Tanganyika Territory, 13 July 1917. 1 immature d\ 1 immature 9, Kilosa, Tanganyika Territory, 18 January 1921. 1 unsexed, Kilosa, Tanganyika Territory, 12 February 1921. "Also Nairobi, Kenya Colony, and Lumbo, Mozambique." (A.L.) The unsexed bird from Kilosa is probably a female as it is consider- ably paler than the male from Morogoro. Van Someren1 considers africana a race of epops, but, although the two are geographically distinct in their breeding ranges, the difference in the pattern of the primaries is so great as to constitute specific characterization. This hoopoe occurs throughout the territory represented by the present collection. Family PHOENICULIDAE. Wood-hoopoes Phoeniculus ptjrpureus marwitzi (Reichenow) Irrisor erythrorhynchus marwitzi Reichenow, Orn. Monatsb., 16, 1906, p. 171: Mkalama, Kondoa Irangi district, Tanganyika Territory. 1 adult d" ?, 1 immature cf, 1 adult 9, Kilosa, Tanganyika Territory, 21 January, 1921. "Also Bungu, Morogoro, Kongwa, Msimba, and Rukaya." (A.L.) The immature bird resembles the adults but has the feathers of the chin and throat light brownish and the bill is black instead of red as in older birds. The adult cT ? is probably a female as it is rather small, agreeing in size with the female listed above and with others in the series examined. The measurements are as follows: cT ? (= 9 ) adult: wing 136.5; tail 200; culmen 34 mm. 9 adult : wing 134.0; tail 202; culmen 36.5 mm. c? immature : wing 142.0; tail 198; culmen 44.5 mm. i Nov. Zool., 29, 1922, p. 81. 164 bulletin: museum of comparative zoology This race occurs throughout the territory under discussion. In northern Tanganyika Territory and Kenya Colony its range is limited on the west by the Rift Valley according to Sclater,1 but this is not so as it occurs at Kisumu and in Uganda. According to Grant,2 Irrisor erythrorhynchiis brevirostris Gunning and Roberts (described from Boror, Mozambique) is a synonym of marwitzi. Roberts3 however, maintains the name in his check-list of South African birds. Phoeniculus bollei jacksoni (Sharpe) Irrisor jacksoni Sharpe, Ann. Mag. Nat. Hist. (6), 6, 1890, p. 503: Kikuyu country, Kenya Colony. 1 d\ 1 9 , Ngong Forest, near Nairobi, Kenya Colony, 18 July 1919. 1 immature d\ Ngong Forest, near Nairobi, Kenya Colony, 16 September 1920. The immature bird has no white spots on the remiges and rectrices and therefore must belong to this species. Yet it has no whitish on the head, but has that part wholly metallic greenish like the breast and mantle! Claude Grant4 writes that in a young bird examined by him the head and throat were almost pure white with an admixture of black. Granvik5 states that in young specimens the, "... bronze-green patches on the back and under surface are washed with blue, which tint, however, disappears more and more in old birds and is replaced by a more uniform green, sometimes yellowish-green, glossy colour." In the present case there is no bluish wash in the young bird but there is in the two adults. The young bird has a blackish-gray bill, about two-thirds grown. The measurements of the two adults are as follows : male — wing 135; tail 162; culmen from the base 47 mm. female — wing 126; tail 131; culmen from the base 37.5 mm. Van Someren6 suggests that there may be three forms united under the name jacksoni and that Kikuyu birds have the general color of the head and mantle more golden green, while those from the Elgeyu Escarpment are more bluish. This is not confirmed by the material examined. i Syst. Avium Ethiop., pt. 1, 1924, p. 233. 2 Ibis, 1915, p. 285. 3 Ann. Transv. Mus., 10, 1924, p. 149. « Ibis, 1915, p. 287. 5 Journ. f. Ornith., 1923, Sonderheft, p. 113. 6 Nov. Zool., 29, 1922, p. 82. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 165 Rhinopomastus cyanomelas schalowi Neumann Rhinopomastus schalowi Neumann, Journ. f. Ornith., 1900, p. 221: Usandawi, Tanganyika Territory. 1 9 , Morogoro, Tanganyika Territory, 24 May 1917. 1 cf , Morogoro, Tanganyika Territory, 12 July 1917. "Also Dar es Salaam, Kilosa, and Lumbo." (A.L.) The female is molting the wings and tail; the male is in somewhat worn plumage. The latter is more reddish-purple, less bluish-violet on the scapulars and interscapulars than any other comparable specimens examined. The number of primary coverts that are white is variable in this bird. Both the above specimens have only the innermost greater upper primary covert white, but other specimens have several of this color, and one has them all dark blue. This bird is found throughout the territory under discussion and is quite common in most of its range. Rhinopomastus minor extimus Friedmann Rhinopomastus minor extimus Friedmann, Proc. N. Eng. Zool. CI., 11, 1929, p. 29: Dodoma, Tanganyika Territory. 1 d1, Dodoma, Tanganyika Territory, 25 December 1918. 1 9 , Dodoma, Tanganyika Territory, 1 December 1921. "Also seen at Mtali's. Rather a scarce bird, not common like R. c. schalowi." (A.L.) While with the Smithsonian-Chrysler Expedition, Loveridge col- lected another male at Dodoma on 7 July 1926, which agrees with the male listed above but has a somewhat longer bill and shorter wing. The difference, however, is wholly within the limits of ordinary in- dividual variation. The present subspecies, which resembles cabanisi, but is larger (wing 106-112 mm. as against 97.5-102 mm., in the latter), occurs from Mangasini, Usandawi, and the Dodoma district, central Tan- ganyika Territory, north to southern Kenya Colony (Taveta and Teita districts to Southern Ukamba and Southern Kavirondo). The male listed above is the type of extimus. Van Someren,1 while agreeing that southern birds cannot be con- i Nov. Zool., 37, 1932, pp. 290-291. 166 bulletin: museum of comparative zoology sidered cabanisi, suggests that calling them extimus merely provides one more name with many intermediate birds harder to place than before. It seems just the contrary to me; they are more readily allocated as the limits of intergradation are narrowed. Family BUCEROTIDAE. Hornbills Bycanistes buccinator buccinator (Temminck) Buceros buccinator Temminck, PI. Col. livr. 48, pi. cclxxxiv, 1824: Cape of Good Hope. 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 13 June 1922. 1 d\ 2 9 , Tindiga, near Kilosa, Tanganyika Territory, 13-19 August 1922. 1 cf, Chanzuru, near Kilosa, Tanganyika Territory, 22 March 1923. The trumpeter hornbill occurs in forested areas throughout Eastern Africa from the Tana River southwards to the Cape Province. Grote1 has considered fistulator, sharpii, duboisi, and buccinator as races of one species, although Sclater2 considers them as three species {sharpii and duboisi being conspecific). On the whole, Grote's arrange- ment appears more logical and is adopted in this report. There is a tremendous amount of variation in the size of the bill and casque, apparently correlated with age. In comparing specimens of the four races this factor must be kept in mind. Bycanistes subcylindricus (Sclater) Buceros subcylindricus P. L. Sclater, Proc. Zool. Soc. London, 1870, p. 668, pi* xxxix: West Africa. 1 Syst. Avium Ethiop., pt. 1, 1924, p. 295. 2 Nov. Zool., 29, 1922, p. 63. • Journ. f. Ornith., 62, p. 394. * Bull. Brit. Orn. CI., 46, 1925, p. 14. 5 Orn. Monatsb., 1915, p. 26: Usagara. 6 Ann. Trans. Mus., 3, p. 112, 1911: Villa Pereira, Boror. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 191 of the top of the head) is similar to that in C. nubica. The young males have red only on the occiput and nape, the forehead and crown being black with small white spots, and the malar stripe black and white. Cameethera cailliautii cailliautii (Malherbe) Chrysopicos cailliautii Malherbe, Rev. Mag. Zool., 1849, p. 540: Africa; Mombasa, apud C. Grant, Ibis, 1915, p. 454. 1 cf, Dar es Salaam, Tanganyika Territory, 3 July 1918. 1 o\ Kilosa, Tanganyika Territory, 2 February 1921. 1 c? , Uluguru Mountains, Tanganyika Territory, 1 June 1921. "Also Morogoro, Bungu, and Mombasa." (A.L.) Campethera loveridgei Hartert1 is a synonym. Hartert2 writes that loveridgei, "... only differs from C. c. cailliautii, with the designated type locality Mombasa ... in being slightly more greenish above and below and in having the round breast spots larger ..." A specimen from the Tana River, Kenya Colony (the northernmost record for the species, F. R. Wulsin coll.) which, on geographical grounds must be considered cailliautii, exactly reverses these char- acters. Otherwise all the above birds would be loveridgei. Van Someren3 writes that Dar es Salaam birds are more heavily spotted below than typical cailliautii. More material is needed before loveridgei can be accepted as a valid race. Sclater4 gives the range of this woodpecker as "Island of Zanzibar and adjacent mainland from Mombasa to Dar es Salaam." As al- ready shown, the bird occurs north to the Tana River. Campethera cailliautii fulleborni (Neumann) Dendromus malherbei fulleborni Neumann, Journ. f. Ornith., 1900, p. 204: Langenburg, {i.e. Manda) n. e. of Lake Nyasa. Loveridge collected a male at Lumbo, Mozambique, in August, 1918. This specimen, now in the Nairobi Museum, and which I have not examined, was determined by Dr. van Someren and listed by him.5 i Bull. Brit. Orn. CI., 40, 1920, p. 139: Morogoro. 2 Nov. Zool., 32, 1925, p. 149. 3 Nov. Zool., 29, 1922, p. 64. 4 Syst. Avium Ethiop., pt. 1, 1924, p. 295. 6 Nov. Zool., 29, 1922, p. 64. 192 bulletin: museum of comparative zoology Campethera abingoni suahelica (Reichenow) Dendromus chrysurus suahelicus Reichenow, Vog. Afr. 2, 1902, p. 175: Great Arusha, Tanganyika Territory. 1 cf, 1 9, Morogoro, Tanganyika Territory, 20 August 1917. 1 9, Kilosa, Tanganyika Territory, 2 February 1921. 1 d1, Uluguru Mountains, Tanganyika Territory, 13 May 1921. "Also Lumbo, Mozambique." (A.L.) Van Someren1 refers a specimen collected by Loveridge at Lumbo, Mozambique, to suahelica, but according to the most recent reviewer of the races of this species2 the typical, southern abingoni occurs from Natal to the Lake Nyasa region. However, Grote3 records suahelica from Marunga, lowTer Rovuma River, so it appears that the range of this form may extend southwards along the coast in Mozambique while the nominate form occurs in the interior north to Nyasaland. The birds from Kilosa and the Uluguru Mountains are in molt in the wings and tail ; the other two are not, but the Morogoro male is in fresh plumage while the female is worn. Recently van Someren4 obtained topotypical examples of suahelica and found that Lumbo birds, while of this race, are paler throughout and are probably near annectens Neumann. Vincent5 writes that abingoni and annectens intergrade in the Tete Province of Mozam- bique. Campethera abingoni mombassica (Fischer and Reichenow) Picus (Campothera) mombassicus Fischer and Reichenow, Journ. f. Ornith.' 1884, p. 262: Mombasa (vide Journ. f. Ornith., 1878, p. 254). 1 o71 , Mombasa, Kenya Colony, 17 May 1918. This race is characterized by having only very small, rather indis- tinct, light spots on the back and upper wing coverts, and also by the olive-brownish bases of the red feathers on the top of the head. It inhabits the coastal strip of East Africa from Mombasa north to southern Italian Somaliland. ' Nov. Zool., 29, 1922, p. 63. » Neumann, Bull. Brit. Orn. CI., 21, 1908, pp. 95-96. 3 Journ. f. Ornith., 1912, p. 523. * Nov. Zool., 37, 1932, p. 282. » Ibis, 1935, p. 19. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 193 The single specimen collected is molting on the crown and the tail. Its dimensions are as follows — wing 107.0, tail 67.5, culmen — (broken). I have seen no material of C. a. kavirondensis van Someren1 but its supposed characters (wider black streaks on breast and belly, and a fusion of the throat streaks to form a wedge-shaped patch extending from the chin to the upper breast) do not sound very convincing, especially as the four specimens of suahelicus in the present collection vary in the width of the black streaks. Moreau2 has recently recorded mombassica from Lyamungu, Albizzia forest, within 40 miles of the type locality of suahelicus! Dendropicos lafresnayi lepidus (Cabanis and Heine) Ipodonus lepidus Cabanis and Heine, Mus .Hein. 4, pt. 2, 1863, p. 118: Ethiopia. 1 9 , Ndeza, Ankole, Uganda, 7 September 1919. 1 9 , Chantwara, Bukoba, Tanganyika Territory, 29 December 1922. 1 d\ Kabare, Bukoba, Tanganyika Territory, 15 January 1923. This race is smaller and less noticeably barred above than its eastern neighbor hartlaubii. However, the dorsal barring color pattern is more pronounced in young than in old birds within the subspecies. D. I. lepidus ranges from western and southwestern Ethiopia and the Bahr el Ghazal through Uganda from the Congo border to western Kenya Colony and northwestern Tanganyika Territory, east to the Rift Valley. Sclater3 writes that it occurs from Uganda to Kenya Colony, "... east of the Rift Valley," but he doubtlessly meant, "east to the Rift Valley." The Ndeza bird is molting the wings. The two females have wings 81 mm. in length; the male, 85.5 mm. Dendropicos lafresnayi hartlaubii Malherbe Dendropicos hartlaubii Malherbe, Rev. Mag. Zool., 1849, p. 532: Zanzibar. 2 cf , 1 9 , Morogoro, Tanganyika Territory, 27 July-23 August 1917. 1 d", Ilonga, Kilosa, Tanganyika Territory, 23 March 1923. "Also Dodoma, Tanganyika Territory, and Lumbo, Mozam- bique." (A.L.) 1 Bull. Brit. Orn. CI., 47, 1927, p. 70: Lolgorien, S. Kavirondo. 5 Proc. Zool. Soc, London, 1935, p. 873. 3 Syst. Avium Ethiop., pt. 1, 1924, p. 298. 194 bulletin: museum of comparative zoology Van Someren1 erroneously refers birds from Morogoro and Lumbo (collected by Loveridge) to Dendropicos fuscescens centralis, but they really are D. lafresnayi hartlaubii! D. f. centralis is not a recognizable form, but a synonym of massaicus. This misidentification is probably the cause of van Someren's statements about centralis that, "... These birds differ so markedly from D.f. massaicus that I am compelled to recognize them as distinct. They are more distinct from D. f. fuscescens than 'massaicus' is from the typical bird . . . ' In a more recent paper2 he gives further details, which should be consulted. The distribution of hartlaubii is as follows: the coastal districts of East Africa from Lamu and Mombasa to Mozambique, including the island of Zanzibar. It does not appear to occur inland very far in the north, the Morogoro, Kilosa, and Dodoma records being the western- most records for northern Tanganyika, but in the southern part of that country it ranges to Iringa and the Uhehe country to the north end of Lake Nyasa, and even into Nyasaland. Lynes3 gives its range as Tanganyika Territory and Mozambique, but mentions Vincent's Nyasaland birds as being of this race as well. Thripias namaquus intermedius Grant Thripias namaquus intermedius C. Grant, Bull. Brit. Orn. CI., 35, 1915, p. 101: Ugogo, Tanganyika Territory. 1 d" , Morogoro, Tanganyika Territory, 23 August 1917. 1 9 , Morogoro, Tanganyika Territory, 17 September 1917. 1 cf, 1 9, Kilosa, Tanganyika Territory, 22 January 1921. Sclater1 casts some doubt on the validity of intermedius, and the present four birds confirm his suspicions, but for the present I hesitate to put it into the synonymy of namaquus because some of the speci- mens are nearer to schoensis. In general coloration the birds called intermedius by Grant resemble the typical South African race, but they are said to agree with schoensis in having the black auricular patch extending caudally beyond the ear-coverts, and more or less connected with the lateral throat stripes. The male from Morogoro has the auricular and throat stripes connected as in schoensis, the other three specimens lack an actual connection between these marks, but i Nov. Zool.29, , 1922, p. 68. 2 Nov. Zool., 37, 1932, p. 283. 3 Journ. f. Orn., 82, 1934, Sonderheft, pp. 68-69. « Syst. Avium Ethiop., pt. 1, 1924, p. 301. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 195 they have the auricular stripe more posteriorly extended than in South African birds. The two Kilosa birds are in molt. The present race occurs throughout the northern half of Tanganyika Territory, southern Ukamba, Teita, and Kikuyu districts, Kenya Colony, and southwestern Uganda. The island of Zanzibar is inhabited by a grayer, less greenish form, T. n. decipiens. Van Someren1 suggests that decipiens is the form of Southern Kenya Colony inland to Nairobi, but he does so on the assumption that intermedins is not constantly different from it, but the material examined in the present study upholds the distinctness of intermedins . Mesopicos goertae centralis Reichenow Mesopicos goertae centralis Reichenow, Orn. Monatsb., 1900, p. 59: Ndussuma, west of Lake Albert. 1 9 , Kabare, Bukoba, Tanganyika Territory, 15 January 1923. This specimen, the first to be recorded from Tanganyika Territory, is not typical centralis but somewhat intermediate between that form and rhodeogaster. It is, however, nearer to the former race and is therefore referred to centralis. The specimen is subadult, has the flanks, abdomen, and under tail-coverts barred and the whole underparts washed with light yellowish green. The red feathers in the mid ventral area are not well developed and are lighter, more orange, in color than in adults. The upper back is duskier, less golden, than in rhodeogaster, slightly brighter yellow than in typical centralis. The range of this subspecies is Uganda west to the Niam Niam country in the eastern Belgian Congo, and south to Bukoba, on the Uganda-Tanganyika border. Hartert2 writes that centralis and poicephalus should be united, and Sclater3 writes that the former is doubtfully distinct from the latter, but lists both races. I find (with very small series) that the two are separable, the western poicephalus being lighter below than the Central African centralis, just as do Sclater and Praed.4 However, the chances are that longer series would reverse this, as Neumann, Grant, Hartert, i Nov. Zool., 37, 1932, p. 2S3. 2 Nov. Zool., 23, 1921, p. 103. s Syst. Avium Ethiop., pt. 1, 1924, p. 302. * Ibis, 1919, p. 633. 196 bulletin: museum of comparative zoology and others have maintained. Yet, in accordance with my policy of agreeing with the arrangement of the Systema Avium Ethiopicarum in all cases where I cannot definitely decide to the contrary, I accept the two forms. Mesopicos goertae rhodeogaster (Fischer and Reichenow) Picus (Mesopicos) rhodeogaster Fischer and Reichenow, Journ. f. Ornith., 1884, p. 180: Masailand, that is, northern Tanganyika Territory. 1 cf , 1 9 , Nairobi, Kenya Colony, 1 August-28 September 1920. This race is a dark, brightly colored form with the maximum de- velopment of red on the middle of the abdomen (in which character it is equalled only by spodocephalus), the brightest, most golden, yellow on the upper back of all the subspecies, and with the grayish underparts without a greenish wash (as is present in spodocephalus, abessinicus, and centralis). It occurs in southern Kenya Colony (Simba to Kisumu) and in northern and western Tanganyika Territory. Mesopicos griseocephalus kilimensis Neumann Mesopicos griseocephalus kilimensis Neumann, Orn. Monatsb., 34, 1926, p. 80: Mountain forests of Kilimanjaro. 1 9, Uluguru Mountains, Tanganyika Territory, 11 May 1921. The Uluguru Mountains appear to be the southernmost locality from which this woodpecker has been reported. The present specimen is the second one known from there (actually the first one collected), the other being a male taken at Nyingwa, Uluguru Mountains, 19 October 1926, by Arthur Loveridge, and reported on as M. g. griseo- cephalus.1 The total known range is very limited (Kilimanjaro to the Uluguru). Not recorded from the Usambara Mountains. Recently Lynes2 has recorded specimens from Njombe forest, near the north end of Lake Nyasa, as "near kilimensis,'" which would constitute a notable extension of range. However, it may be asked if these birds are not wrongly named. Moreau3 calls Usambara Moun- tain specimens typical griseocephalus, and Bangs and Loveridge4 identify birds from the Uzungwe Mountains as ruwenzori Sharpe. 1 Friedmann, Ibis, 1928, p. 83. 2 Journ. f. Orn., 82, 1934, Sonderheft, p. 70. 3Proc. Zool. Soc. Lond., 1935 (1936), pp. 873-874. < Bull. Mus. Comp. Zool., 75, 1933, p. 182. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 197 Jynx ruficollis coseni Grant Jynx ruficollis coseni Grant, Bull. Brit. Orn. CI., 35, 1915, p. 102: Amala River, Kenya Colony. 2 d\ Nairobi, Kenya Colony, 17 August 1920. Cosen's red-breasted wryneck is similar to the typical South African subspecies, but is larger, wings 94-101 mm., as against 90-95 mm. The present two specimens are small for their race, having wings 95 mm. long. Van Someren1 has pointed out that the nature of the stripes on the belly is variable, some birds being broadly, others narrowly striped. This wryneck occurs in southern Kenya Colony from Simba to the Elgeyu Escarpment and Mt. Elgon. Its altitudinal range appears to be from 3,400-9,000 feet. Order PASSERIFORMES Family EURYLAIMIDAE. Broadbills Smithornis capensis suahelicus Grote Smithornis capensis suahelicus Grote, Orn. Monatsb., 34, 1926, p. 17: Magogoni, Tanganyika Territory. 1 d\ 1 9, Uluguru Mountains, Tanganyika Territory, 13 May 1921. "Also Bagilo and Kilosa. The note of this bird is very like a frog's call. On hearing it for the first time at Kipera I actually searched for a tree frog before locating the bird." (A.L.) The two birds listed above are in fresh plumage; in fact they were just completing their molt when shot, the outermost primaries being only partly grown in both. Their dimensions are as follows — male: wing 70.5, tail 51.5, culmen 17; female: wing 70, tail 49, culmen 17 mm. The races of Smithornis capensis are based on rather fine differences, but the species appears to vary so much more geographically than in- dividually, that all the described forms are valid. The present one is nearest to the Angolan race albigularis but is a little more olive- brownish above, and not so white, more washed with pale buffy, below, and with heavier ventral streaks. It is somewhat smaller than the i Nov. Zool., 29, 1922, p. 62. 198 bulletin: museum of comparative zoology Kenyan form medianus; less grayish above than the southern, typical, capensis; lighter than meinertzhageni of Mt. Elgon and the Kavirondo countries. The nominate form occurs from South Africa north through Nyasaland and Mozambique to southern Tanganyika Territory (Lindi, and the Rovuma River), while suahelicus is known from the Kilimanjaro district (Kibonoto and Kahe), Magogoni, Usaramo, and the Uluguru Mountains. It is not known from the Usambara range although it probably occurs there. The six specimens from the Uluguru Mountains which I referred to medianus1 really belong to the present recently described form, the name of which was overlooked when writing that paper. Following Bannerman's account of the subspecies of Smithornis capensis2 there are now seven named forms, but considering the fact that these birds are rather shy inhabitants of the mountain forests and therefore very local with wide gaps in their distribution, it seems not unlikely that additional subspecies still remain to be discovered. Family ALAUDIDAE. Larks MlRAFRA AFRICANA TROPICALIS Hartert Mirafra africana tropicalis Hartert, Nov. Zool., 7, 1900, p. 45: Bukoba, west shore Lake Victoria. 1 d", Kabale, Ruanda, Uganda, 24 September 1919. 1 9 , Rukaya, Mawokota, Uganda, 3 November 1919. 2 d*, 1 9 , Chantwara, Bukoba, Tanganyika Territory, 27 December 1922 to 3 January 1923. The subspecies of Mirafra africana are based on average characters and are therefore sometimes difficult to identify without large series. Inasmuch as the material available for study is rather inadequate (some 20 birds in all, representing five forms) I can do no better than to follow Sclater.3 The present birds agree with the characters of tropicalis in having the feathers of the back largely blackish, and the lesser, upper primary coverts bright rufous cinnamon with almost no dark brown markings. This form inhabits Uganda, western Kenya Colony (Kavirondo and Sotik districts), the Ikoma and Mwanza areas of Tanganyika i Ibis, 1928, p. 83: Bagilo, Nyange, and Vituri. 2 Ibis, 1923, pp. 718-719. 3 Syst. Avium Aethiop, part ii, 1930, pp. 310-312. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 199 Territory, west to the eastern Belgian Congo (Kivu), Ruanda, and Urundi. On the slopes of Ruwenzori it is replaced by a darker form, ruicenzoria Kinnear. I have seen a specimen of the latter, and find it to be very slightly darker above than tropicalis, but the difference is so slight that I am not too convinced of the validity of the Ruwenzori race. Van Someren1 records the breeding season in western Uganda as being in June. MlRAFRA AFRICANA DOHERTYI Hartert Mirafra africana dohertyi Hartert, Bull. Brit. Orn. CI., 19, 1907, p. 93: Escarpment, 6,500 feet, Kenya Colony. 1 cf, Samumba, Singida, Tanganyika Territory, 25 February 1922. This specimen appears to be the first record for Tanganyika Terri- tory, and the southernmost one for the race. It was compared with a good series at Tring by either Dr. Hartert or Mr. Goodson and identi- fied as dohertyi. I find it agrees with another of this race from Lake Naivasha, Kenya Colony, although with my limited material, I am not unwilling to "lump" dohertyi with athi. It is, however, more rufous on the crown than the Naivasha bird, and may be intermediate between dohertyi and tropicalis. The range of dohertyi is as follows: the highlands of the interior of Kenya Colony on both sides of the Rift Valley, east to Nairobi, and south to Singida, Tanganyika Territory. Mirafra africana athi Hartert Mirafra africana athi Hartert, Nov. Zool., 7, 1900, p. 46: Athi Plains, Kenya Colony. 1 cf, 1 9 , Ngong, near Nairobi, Kenya Colony, 7 July 1919. 1 d% Nairobi district, Kenya Colony, 18 August 1920. Van Someren2 and Sclater3 recognize harterti as distinct from athi, although the latter admits its distinctness as doubtful. I cannot add anything positive to the problem as I have seen no actual harterti, but the specimens of athi examined lead me to wonder if it is not a form ilbis, 1916, p. 434. 2 Nov. Zool. 29, 1922, p. 175. 3 Syst. Avium Aethiop., part ii, 1930, p. 312. 200 bulletin: museum of compaeative zoology with a tendency toward dichromatism, of which the rufous phase may be harterti. Tentatively, I assume this to be the case, in which event the range of athi includes the drier parts of Kenya Colony from Nairobi and the Athi Plains to Magadi, Ukamba, and the Teita country. The male taken near Nairobi on 18 August is somewhat intermediate between athi and dohertyi. It is in fine fresh plumage, while the two July birds are much abraded. Mirafra fischeri kawirondensis van Someren Mirafra fischeri kawirondensis van Someren, Bull. Brit. Orn. CI., 41, 1921: p. 125, Kisumu, Lake Victoria. 1 9 , Lalago, Mwanza, Tanganyika Territory, 18 October 1922. 1 d\ Sagayo, Mwanza, Tanganyika Territory, 24 October 1922. 1 cf, Kome Island, Mwanza, Tanganyika Territory, 23 November 1922. I have no topotypical kawirondensis available for comparison, and have placed these birds in that race somewhat hesitantly. The female from Lalago is considerably paler above and below than either of the males, and approaches zombae, from which it differs only in lacking the slight grayish bloom found on the upperparts of the latter (which difference may be due to wear). Oustalet's form tigrina is known to me only from descriptions, but it may be noted that when the two males listed above were examined at Tring, either Dr. Hartert or Mr. Goodson penciled on the back of the labels," ? tigrina Oust., ? kaviron- densis v. Som." It seems not unlikely that the Mwanza area is the meeting place of zombae, fischeri, and kawirondensis and that in that region a great variety of intergrades occur. The name to be used for them is open to question and would probably be different in different cases, as some specimens are nearer to one form than to another. It might perhaps, be nearer the truth to record the two males as kawiron- densis > fischeri and the female as fischeri (or zombae) Ibis, 1916, p. 464. 2 Orn. Monatsb., 1906, p. 148: Usegua. 222 bulletin: museum of comparative zoology This specimen is in molt; the old feathers are extremely abraded? contrasting markedly with the dark, new ones. The breeding season near Nairobi is from December to March. Pseudoalcippe atriceps (Sharpe) Turdinus atriceps Sharpe, Bull. Brit. Orn. CI., 13, 1902, p. 10: Ruwenzori. 1 cf, 1 9, Singo, Ruanda, Uganda, 25 September 1919. Iris brown; upper mandible black, lower mandible gray; feet gray. The Ruwenzori hill-babbler occurs from the Kivu volcanoes north through Ruanda and the highlands of the eastern Belgian Congo to Ruwenzori and to southern Ankole in southwestern Uganda. It is not very common in Uganda, being more numerous in the Kivu. district. Like so many of the mountain birds of central Africa, it reappears in the highlands of Cameroon (Genderu and Banso Moun- tains). On Ruwenzori, Woosnam1 found it from, "... 6,500 to 9,000 feet, frequenting the forest zone and occasionally the lower part of the bamboo . . . going about in small parties of four or five individuals. They hop along through the ferns and tangled vegetation . . . ' Young birds have the crown and nape dark brown instead of black as in the adults. The present two examples are fully adult and have the following dimensions: male — wing 69, tail 65, culmen 14, tarsus 21 mm. ; female — wing 66, tail 56, culmen 14, tarsus 22.5 mm. Lynes2 collected a male about to breed, in December, at Njombe. Pseudoalcippe abyssinicus abyssinicus (Riippell) Drymophila abyssinica Riippell, N. Wirbelth., Vogel, 1840, p. 108, pi. xl, fig. 2: Simien, Ethiopia. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. "Also Tumutumu, Kenya Colony." (A.L.) Grote.3 has separated the birds of the Usambara Mountains from typical abyssinicus under the name micrus on the basis of supposedly smaller size. He writes that the latter birds have wings 66-68 mm., i Trans. Zool. Soc. Lond., 19, 1910, p. 381. 2 Journ. f. Orn., 82, 1934, Sonderheft, p. 74. 3 Orn. Monatsb., 36, 1928, p. 77. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 223 in length as against 72-76 mm. in Ethiopian birds. Sclater1 considers micrus a synonym of abyssinicus. The present specimen has a wing length of 65 mm., and would therefore appear to corroborate Grote's findings. However, a male abyssinicus from Aletta, Ethiopia, has this dimension only 66 mm., so the difference is nil. I agree with Sclater that micrus is unrecognizable. The same is true of the Kilimanjaro "subspecies" kilimensis Shelley. This little hill-babbler reaches its southern limits in the Usambara Mountains. It is a mountain forest bird, and consequently its range is very discontinuous. Pseudoalcippe stierlingi (Reichenow) Turdinus stierlingi Reichenow, Orn. Monatsb., 6, 1898, p. 82: Iringa, Tangan- yika Territory. 1 9 , Uluguru Mountains, Tanganyika Territory, 23 May 1921. 1 cf , Uluguru Mountains, Tanganyika Territory, 3 June 1921. Hartert2 described the Uluguru birds as a race P. s. uluguru, of which the present pair are topotypes. His material consisted of a single female of uluguru which was compared with the types of stier- lingi in Berlin by Neumann. I have seen 7 specimens from the Uluguru Mountains and find the race uluguru is not valid. The characters by which it was said to differ from the nominate form are those of larger size; wing 70 instead of 64-66 mm. ; bill larger, and upper parts a little brighter rufous. Five males from the Uluguru Mountains have wings of from 61 to 65.5 mm.; two females measure 64 mm., each, in this respect. Pseudoalcippe pyrrhopterus (Reichenow and Neumann) Callene pyrrhoptera Reichenow and Neumann, Orn. Monatsb., 3, 1895, p. 75: Mau, Kenya Colony. 1 d", Ruanda, Uganda, 28 September 1919. Iris brown; bill black; feet gray. Neumann's Kivu race kivuensis is not distinguishable. He compared his type with immature instead of adult birds. Likewise Reichenow's tanganyicae is also a synonym. 1 Syst. Avium Aethiop., part ii, 1930, p. 364. 2 Bull. Brit. Orn. CI., 42, 1922, p. 50. 224 bulletin: museum of comparative zoology This hill-babbler occurs from Ruwenzori south to Ruanda, and the highlands west of Lake Tanganyika, and east to Mt. Elgon, Mau, Aberdare Mountains, Nandi, and the Kaimosi-Kakamega district, in Kenya Colony. Granvik1 has separated the Elgon birds on the basis of larger size and brighter reddish-brown dorsal coloration. I have seen no Elgon specimens, but Sclater2 considers elgonensis a synonym of pyrrhoptirus. The present specimen has a wing length of 72 mm., while Granvik's Elgon males measure 73 and 78 mm., respectively. Argya rubiginosa emini (Reichenow) Argya rubiginosa emini Reichenow, Orn. Monatsb., 15, 1907, p. 30: Scamuja, Unyamwesi district, Tanganyika Territory. 1 cf, 1 9, Nduguyu River, Mkalama, Tanganyika Territory, 14 October 1922. "Also Sanga and Sagayo. The female from Nduguyu River held a perfectly formed, pure white egg." (A.L.) The present race of the rufous chatterer is a very scarce bird in col- lections, being known only from Scamuja and the three localities mentioned above. As Hartert has pointed out,3 emini differs from rubi- ginosa in having a slenderer bill, the forehead and most of the crown grayish with pale tips to the feathers, more grayish lores, and shorter wings. The present male has wings 81 mm. in length, the female 82 mm. Argya rubiginosa heuglini Sharpe Argya heuglini Sharpe, Cat. Bds. Brit. Mus., 7, 1883, p. 391: Zanzibar. "One collected at Dar es Salaam, 4 July 1918, now in Nairobi Museum." (A.L.) Argya aylmeri mentalis Reichenow Argya mentalis Reichenow, Journ. f. Ornith., 1887, p. 75: Soboro, Kondoa Irangi district, Tanganyika Territory. 1 d\ 1 9, Mbonoa, Singida, Tanganyika Territory, 29 September 1922. 1 9 , Dodoma, Tanganyika Territory, 7 July 1921. "Also Kinyambwa. Common at Dodoma." (A.L.) i Journ. f. Ornith., 1923, Sonderheft, p. 256. 2 Syst. Avium Aethiop., part ii, 1930, p. 365. s Bull. Brit. Orn. CI., 43, 1923, p. 134. ! FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 225 This form of the scaly chatterer is darker than the typical race of Somaliland, Arussi-Gallalahd, and northern Kenya Colony. I con- sider keniana Jackson and loveridgei Hartert as synonyms of mentalis. The birds listed above are in somewhat worn plumage. Nothing appears to be known regarding the habits of this babbler, except that it seems to be decidedly uncommon in most parts of its range. It is therefore rather remarkable that Loveridge should have found it numerous at Dodoma. Neocichla gutturalis angustus Friedmann Neocichla gutturalis angustus Friedmann, Journ. Wash. Acad. Sci., 20, 1930, p. 434: Muhalala, Tanganyika Territory. 1 cf, 1 9 , Muhalala, Kilimatinde, Tanganyika Territory, 3 March 1922. This bird is considered a babbler by Sclater1 but it is by no means certain that it is not more nearly related to the starlings. The male is in the plumage described by Biittikofer2 as N. kelleni. This is supposed to be the immature plumage of gutturalis, but Neave3 considers kelleni a distinct species, as all the specimens he saw (a flock of about twenty) were in the spotted kelleni plumage. The present male differs from the figure given by Biittikofer (loc. cii.) in being pale buffy gray on the rump and lower back and on the edges of the feathers of the upper back, instead of bright tawny. Schalow4 in his report on Bohm's East African collections, describes a juvenal gutturalis from Kakoma, Tanganyika Territory, which approximates that of "kelleni," and agrees with the present male, except that he writes that the head is dark brown whereas it is black in the present example, and likewise the neck is black and white, not brown and white as in his description. It seems, then, that kelleni is really the young of gutturalis, but judging from the present male, it appears as though the kelleni type of plumage is worn for at least two years as the present example is in no way juvenal in appearance, and has adult primaries and generally fresh plumage. It also has the tip of the bill black, not wholly yellow, as in Biittikofer's plate. Aside from the evidence afforded by Schalow's notes, the geograph- 1 Syst. Avium Aethiop, part ii, 1930, p. 359. 2 Notes Leyden Museum, 10, note 22, 1S88, p. 229, pi. ix, fig. 1. a Ibis, 1910, p. 137. « Journ. f. Ornith., 1883, pp. 357-358. 226 bulletin: museum of comparative zoology ical evidence would indicate the identity of kelleni and gutturalis. It would be very unusual, to say the least, if the two sole members of a genus were to have precisely the same ranges. The measurements of these two specimens are as follows: male — wing 108.5, tail 82, culmen 23 mm. ; female — wing 109, tail 87, culmen 22.5 mm. The female is molting. Inasmuch as the female is undergoing a complete molt, and the male is in very fresh plumage, it may be inferred that the breeding season had ended no great length of time before the birds were shot (3 March). Bohm1 collected a male with enlarged testes in the beginning of December and found juvenal birds out of the nest in March. Only a few collectors have obtained this species in Tanganyika Territory, the most recent being Schuster2 who met with it in Unyam- wesi, and Carnochan who collected a pair near Tabora (now in the American Museum of Natural History). Schuster saw a flock of about 20 birds at Senga on 20 June, obviously a sign of post-breeding gre- gariousness. The present race differs from the nominate, Angolan, form in having very much narrower white tips to the rectrices. In two Angolan birds the white tips of the outermost rectrices measure 12 to 15 mm. in width (measured along the shaft), while in five Tanganyikan specimens they measure from .5 to 5.0 mm. The nominate race is conspicuously a bird with white tips to the tail feathers; angustus is often practically devoid of tips or, at most, has them in the form of narrow bands. I have seen no material from the Katanga or from Northern Rhode- sia and cannot say to which form the birds of those regions belong. I have seen angustus from the following localities (all in Tanganyika Territory) : Muhulala, Kilimatinde, Manyoni, Kakoma, and Kawewe's, 35 miles south of Tabora. Illadopsis stictigula stictigula (Reichenow) Turdinus stictigula Reichenow, Orn. Monatsb., 14, 1906, p. 10: Mbaramo, Usambara Mountains. 1 9, Uluguru Mountains, Tanganyika Territory, 30 May 1921. 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 15 May 1922. This very distinct species of Illadopsis is restricted, as far as known, 1 Journ. f. Ornith., 1883, pp. 189-190. - Journ. f. Ornith., 1926, p. 738. FRIEDMANN AND LOVERIDGE : EAST AFRICAN ORNITHOLOGY 227 to the Usambara and the Uluguru Mountains, the Uzungwe and Ukinga range, and Mt. Rung we, Tanganyika Territory. Aside from the original small series collected by Roehl, the present race is known only from the present two examples, and two others also taken by Loveridge.1 Bangs and Loveridge2 have described /. s. pressa from the highlands of southwestern Tanganyika Territory on the basis of its smaller size and darker and duller coloration. Suaheliornis kretschmeri kretschmeri (Reichenow and Neumann) Phyllostrephus kretschmeri Reichenow and Neumann, Orn. Monatsb., 3, 1895, p. 75: Kibosho, Kilimanjaro. 1 cf, 1 9, Bungu, Usambara Mountains, Tanganyika Territory, September 1921. This peculiar bird is found, as far as known, only on Mt. Kiliman- jaro and the Usambara and Uluguru Mountains, Tanganyika Terri- tory. In the extreme southeastern part of that country another race replaces it — griseiceps Grote, characterized by its much grayer head and underparts. I have seen no material of the latter form, but its characters appear, from description, to be well marked. Sclater3 considers S. albigula Grote as of uncertain status. It was described from the Usambara Mountains, and while it is rather puzzling to find two species of this genus living together in one small mountain area, a similar condition is to be met with in Upper Guinea where Macrosphenus concolor and M . Icempi occur side by side. Grote4 writes that a specimen from the western slopes of the Usam- bara Range differs from the type (from Kilimanjaro) in having the head slightly more grayish, less washed with greenish, and the longi- tudinal streaks on the underparts greener and more distinct. I have seen no Kilimanjaro birds. The present female has the throat and the ground color of the breast and abdomen grayer than in the male, making the greenish stripes stand out more distinctly. » Ibis, 1928, p. 96. *Proc. N. Eng. Zool. CI., 12, 1931, p. 94. 3 Syst. Avium Aethiop., part ii, 1930, p. 366. 4 Journ. f. Ornith., 1921, p. 136. 228 bulletin: museum of comparative zoology Family PYCNONOTIDAE. Bulbuls Pycnonotus tricolor minor Heuglin Pycnonotus nigricans var. minor Heuglin, Orn. Nordost. Afr., 1, 1869, p. 398: Bahr el Abiad, that is, Upper White Nile. 1 cf , 1 immature 9 , Kome Island, Mwanza, Tanganyika Territory, 23 October 1922. These two specimens constitute the first records for minor for north-central Tanganyika Territory. However, inasmuch as Reiche- now's so-called race tanganjicae is now known to be identical with minor, this record is not remarkable. The form occurs south to Usumbura and to Nyanza on Lake Tanganyika. Both specimens are in very worn plumage. Van Someren1 has recently reviewed the races of the Kenyan and Ugandan yellow-vented bulbuls, but does not include Tanganyika Territory in the range of minor. Pycnonotus tricolor pallidus Roberts Pycnonotus layardi pallidus Roberts, Journ. S. Afr. Ornith. Union, 8, 1912, p. 49: Boror, Mozambique. 1 6", 1 9, Lumbo, Mozambique, 30 August 1918. I agree with van Someren2 that pallidus is a recognizable, although somewhat intermediate form, connecting layardi with micrus. It resembles the former but is slightly paler and has the lower abdomen more washed with yellow posteriorly. Both birds are in worn plumage. Loveridge found this bulbul very abundant at Lumbo. Five nests, each containing two eggs, were found there on 28 October.3 Pycnonotus tricolor micrus Oberholser Pycnonotus layardi micrus Oberholser, Proc. U. S. Nat. Mus., 28, 1905, p. 891: Taveta, Kenya Colony. 1 cf, Morogoro, Tanganyika Territory, 15 June 1917. 1 9 , Mombasa, Kenya Colony, 25 May 1918. "Also Dar es Salaam, Kilosa, Dodoma, and Tabora." (A.L.) ■ Nov. Zool., 37, 1932, p. 347. 2 Nov. Zool., 29, 1922, p. 190. 3 Proc. Zool. Soc. Lond., 1922, p. 845. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 229 This is the race of the coastal plain of Southern Kenya Colony and of Tanganyika Territory south to Dar es Salaam and Zanzibar, and inland to Kilosa, Dodoma, and even to Tabora, where it intergrades with minor. On 7 January 1923, at Kilosa, Loveridge found a nest of this geelgat with three eggs. A week before he had found another nest with two eggs. The yellow-vented bulbul breeds during the greater and lesser rains, if not all the year round. Many nests were found at Moro- goro on the following dates: two young birds in a nest in boughs of a flamboyant, 8 December 1917; two nests, one with two eggs and one with three, in similar sites, on 14 December 1917; nest with one nestling in a banana plant on 16 January 1918; a nest with two eggs on 26 March; and one with one egg on 11 April.1 Pycnontus tricolor fayi Mearns Pycnonotus layardi fayi Mearns, Smiths. Misc. Coll., 56, no. 20, p. 7, 1911: Fay's Farm, Njabini, Kenya Colony. Id"?, Ngong, near Nairobi, Kenya Colony, 21 September 1915. 1 9 , Ngong, near Nairobi, Kenya Colony, 21 July 1919. 1 9, Uluguru Mountains, Tanganyika Territory, 14 May 1921. 1 d", Bungu Usambara Mountains, Tanganyika Territory, September, 1921. "Also Tumutumu." (A.L.) The Uluguru and Usambara Mountain birds are the first records for fayi in Tanganyika Territory and are of interest in that they in- dicate again that altitude is the important factor in the distribution of subspecies in eastern Africa. Quite naturally, a student having specimens from two localities as close together as Morogoro and the Uluguru Mountains and finding them different, while the latter birds agree with others from as distant a place as Mt. Kenya or Mau Escarpment, would suppose the variations were wholly individual and that no races were admissible. However, the thing is perfectly clear and simple when we remember that fayi is a highland race and micrus a lowland form. In the Ibis for 1928 (p. 91) I recorded two birds from Amani and Bumbuli, Usambara Mountains, as micrus, but on reexamination I find them to be fayi. 1 Recorded by Loveridge, Proc. Zool. Soc. Lond., 1922, p. 845. 230 bulletin: museum of comparative zoology The July bird from Ngong is in molt ; the September bird is abraded as is also the September specimen from Bungu. The Uluguru speci- men is in fresh plumage. Atimastillas flavicollis pallidigula (Sharpe) Xenocichla pallidigula Sharpe, Bull. Brit. Orn. CI., 7, 1897, p. 7: Entebbe. 1 9 , Masomuntu Mukubwa, Ruanda, Uganda, 26 September 1919. 1 9 , Kome Island, Mwanza, Tanganyika Territory, 22 October 1922. 1 Journ. f. Ornith., 1926, p. 739. 2 Ibis, 1916, p. 465. 248 bulletin: museum of comparative zoology Saxicola torquata axillaris (Shelley) Pratincola axillaris Shelley, Proc. Zool. Soc. Lond., 1884, p. 556: Mt. Kiliman- jaro, 7,000 feet. 1 d\ Ngong, Nairobi, Kenya Colony, 16 July 1919. 1 9, Nairobi district, Kenya Colony, 11 October 1920. 1 d\ Eldoret, Kenya Colony, 6 November 1920. "Also Tumutumu." (A.L.) The Ngong bird is heavily spotted and marked with black on the breast and abdomen. It is in very worn plumage; the others are with- out any black spots on the breast and abdomen and are in fresh plumage. Saxicola torquata promiscua Hartert Saxicola torquata promiscua Hartert, Bull. Brit. Orn. CI., 42, 1922, p. 51: Uluguru Mountains, Tanganyika Territory. 2 d\ Bungu, Usambara Mountains, Tanganyika Territory, September 1921. The Usambara Mountains appear to be the northern limit of the range of this form. On Kilimanjaro and Meru, not so far away, axillaris replaces it. It would be of interest to know what form occurs in the Pare and Bura Hills. The present form differs from axillaris in having the black area on the throat of the male more extensive, and the brown of the breast less so. The breeding season, in the Uluguru Mountains, appears to be in October.1 Saxicola rubetra rubetra (Linnaeus) Motacilla rubetra Linnaeus, Syst. Nat., 10th ed., 1758, p. 186: Europe; re- stricted type locality, Sweden. 1 0% 1 9 , Kome Island, Mwanza, Tanganyika Territory, 21 November 1922. These two specimens are in winter plumage. The European whinchat occurs abundantly in Kenya Colony and Uganda in winter, and south through the western half of Tanganyika Territory to the north end of Lake Nyasa. It does not seem to occur in the eastern portions of that country. i cf. Ibis, 1928, p. 95. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 249 COSSYPHA HEUGLINI HEUGLINI Hartlaub Cossypha heuglini Hartlaub, Journ. f. Ornith., 1866, p. 36: "Keren," error, Wau, Bahr el Ghazal (Heuglin, Orn. Nordost. Afr., 1, p. 375). 1 d\ Nairobi, Kenya Colony, 16 July 1919. 1 d" , Kilosa, Tanganyika Territory, 25 December 1920. 1 d\ 1 9, Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 2 d\ Bagilo, Uluguru Mountains, Tanganyika Territory, 15-26 May 1922. The birds from the Usambara Mountains are intermediate between heuglini and intermedia, but somewhat nearer the former. Grote1 has recorded Usambara birds as intermedia, so it appears to be the meeting place of the two races, some specimens being nearer to the one, others to the other form. The July bird from Nairobi is in fresh plumage; the September specimens are rather worn. Cossypha heuglini occidentalis Reichenow Cossypha heuglini occidentalis Reichenow, Journ. f. Ornith., 1909, p. 108: Lufuku, west of Lake Tanganyika. 1 6", 1 9, Kabale, Ruanda, Uganda, 21 September 1919. 1 d\ Chantwara, Bukoba, Tanganyika Territory, 7 January 1923. 1 d\ 1 9 , Kabale, Bukoba, Tanganyika Territory, 12 January 1923. This is a more richly colored race than typical heuglini. It occurs from southwestern Uganda (southern Ankole) south through the eastern Ituri district and Ruanda to Lake Kivu and the west shore of Lake Tanganyika. It is a valid form, although not listed as such by Sclater.2 Cossypha heuglini intermedia (Cabanis) Bessornis intermedia Cabanis, Journ. f. Ornith., 1868, p. 412: coastal districts of East Africa. In his manuscript list of the collection, Loveridge records a specimen of this form taken at Mombasa on 23 May 1918. I have not been able to find this bird, but there is no reason for doubting the identification. » Journ. f. Ornith.. 1921, p. 138. 2 Syst. Avium Aethiop., part ii, 1930, p. 470. 250 bulletin: museum of comparative zoology Cossypha heuglini euronota Friedmann Cossypha heuglini euronota Friedmann, Occ. Papers Bost. Soc. Nat. Hist., 5, 1930, p. 327: Lumbo, Mozambique. 1 9 , Lumbo, Mozambique, July 17, 1918. (Type.) This form combines the color characters of heuglini with the size of intermedia. It is definitely known as yet only from Lumbo, but it is quite probable that it occurs north to Mikindani, where Grote1 col- lected this species. At the latter place the breeding season is in Febru- ary and March. It has recently been recorded from Mt. Silinda, in Gazaland, by Bangs.2 Van Someren3 writes that the color characters of this race are purely sexual, the males being slatey gray on the back, and that euronota is a synonym of intermedia. However, Bangs records a male from Mt. Silinda as exactly matching the type (female) in color. Cossypha semirufa intercedens (Cabanis) Bessornis intercedens Cabanis, Journ. f. Ornith., 1878, pp. 205, 219: Kitui, Ukamba district, Kenya Colony. 1 9, Nairobi, Kenya Colony, 20 October 1915. 1 cf, Ngong, near Nairobi, Kenya Colony, 10 July 1919. The forms of Cossypha semirufa are very closely allied to those of C. heuglini, and it may well be that they are all one species. The present form inhabits the high districts of south-central Kenya Colony, from the Ukamba and Kikuyu districts to Mt. Kenya and the Aberdare Mountains. In eastern Ethiopia (Harar and eastern Galla- land) a shorter winged form, donaldsoni, replaces it, while in the rest of Ethiopia two other smaller forms (one northern, the other southern) with olive-green backs are found. This species also occurs on Mt. Kilimanjaro (1 specimen seen), but whether the Kilimanjaro form is intercedens or donaldsoni or distinct from both is not very clear. For the present I refer it to intercedens but more material is needed. The present specimens are in fresh plumage. i Journ. f. Ornith., 1913, p. 141. 2 Proc. N. Eng. Zool. Club, 12, 1931, p. 67. 3 Nov. Zool., 37, 1932, p. 378. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 251 COSSYPHA NATALENSIS NATALENSIS Smith Cossypha natalensis A. Smith, 111. Zool. S. Afr., Aves, 1840, pi. Ix: near Port Natal, that is, Durban. 1 c\ Uluguru Mountains, Tanganyika Territory, 30 May 1921. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 1 d1, Mahaka, Tanganyika Territory, 24 March 1922. 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 17 May 1922. 1 9 , Tindiga, near Kilosa, Tanganyika Territory, 25 August 1922. C. natalensis interna Mearns is a synonym, but, as far as may be judged from a single specimen, C. n. garguensis Mearns appears to be a valid race — paler and smaller. Van Someren1 writes as follows: "Though I have examined a big series, I cannot recognize any constant differences, warranting the separation into geographical forms. I have no Uraguess birds but birds from this locality may very likely be different, because of the peculiar nature of the country." This robinchat is a common bird throughout the regions represented by the present collection. In Uganda it breeds in April and May. Cossypha niveicapilla melanonota (Cabanis) Bessonornis melanonota Cabanis, Journ. f. Ornith., 1875, p. 235: Tschent- schoscho, Portuguese Congo. 1 unsexed, Uganda, 1919. 1 c? , 1 9 , Chantwara, Bukoba, Tanganyika Territory, 26 December 1922. The snowy-headed robinchat is a bird of the forests and occurs from Cameroon to Loango, and east to Uganda, and extreme western Kenya Colony (Kavirondo to Elgon) . It appears to be uncommon on Mt. Elgon, as Granvik did not meet with it during his sojourn there. According to van Someren2 the nesting season in Uganda is in May and November. The unsexed bird is blacker on the upper back than the female. These birds agree with two from Gaboon. i Nov. Zool., 29, 1922, p. 239. 2 Ibis, 1916, p. 472. 252 bulletin: museum of comparative zoology COSSYPHA CAFFRA IOLAEMA Reichenow Cossypha caffra iolaema Reichenow, Orn. Monatsb., 8, 1900, p. 5: Mt. Kili- manjaro, East Africa. 1 immature c\ 1 adult 9, Nairobi, Kenya Colony, 25 September 1920. 1 adult d\ Eldoret, Kenya Colony, 8 November 1920. "Also Tumutumu and the Uluguru Mountains." (A.L.) C. c. mawensis Neumann is a synonym. This robinchat occurs in the highlands from Nyasaland through Tanganyika Territory to Lake Kivu, Ruanda, Urundi, Ankole, and to Kenya Colony (north to Mt. Elgon and Mt. Kenya district). In the east it does not occur north of Kilimanjaro. Schuster1 found this bird in the Usambara, Nguru, Uluguru, north and south Rubeho, and Uzungwe Mountains at altitudes of from 1,000-1,200 meters. Bessonornis albigularis albigularis (Reichenow) Callene albigularis Reichenow, Orn. Monatsb., 3, 1895, pp. 87, 96: Uluguru, that is, Uluguru Mtns., Morogoro district, Tanganyika Territory. 1 d\ 1 9 , Uluguru Mountains, Tanganyika Territory, 20 May 1921. This robinchat is a very rare bird in collections. As far as I know these two, and two females collected by Loveridge at Nyingwa, Uluguru Mountains2 are the only specimens known other than the types collected by Stuhlmann some thirty years before. The figure given by Reichenow3 does not fit these specimens any too well. It is paler than the birds, especially above, and has the auriculars practically surrounded by whitish, while in the specimens examined there is no posterior whitish border to the auriculars. The female has some old upper primary coverts with rufous terminal shaft spots — apparently Juvenal feathers retained into the first adult feathering. The measurements of the present two birds are as follows : male — wing 74.0, tail 65.0, culmen 19.0, tarsus 32.0 mm. female — wing 73.5, tail 62.0, culmen 18.0, tarsus 28.0 mm. 1 Journ. f. Ornith., 1926, pp. 740-741. 2 See Ibis, 1928, p. 94. 3 Vog. Afr. Atlas, 1905, pi. xxviii, fig. 2. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 253 Bangs and Loveridge1 have recently described a race porotoensis from the Poroto Mountains and Mt. Rungwe, in southwestern Tan- ganyika Territory. It is a distinct form and differs from the nominate one in having a smaller and shorter bill, and paler, more yellowish flanks and under tail-coverts. Unfortunately, Sclater, after comparing a paratype with the type of macclounii (Shelley) from the Nyika Plateau, finds that porotensis must be regarded as a synonym of macclounii which he had referred to Callene in the Systema. The name, therefore, is Bessornis albigularis macclounii (Shelley). Sheppardia cyornithopsis bangsi Friedmann Sheppardia cyornithopsis bangsi Friedmann, Occ. Papers Bost. Soc. Nat. Hist., 5, 1930, p. 323: Uluguru Mountains. 1 adult cf, 1 adult 9 , 1 female immature, Uluguru Mountains, Tanganyika Territory, 23 May 1921. The adult male is the type of bangsi. This race, confined, as far as known, to the Uluguru Mountains is somewhat intermediate between sharpei of Nyasaland and aequatori- alis of southwestern Kenya Colony; agrees with the former in having wide, white superciliary stripes and the whole of the abdomen (except the sides and flanks) white; and agrees with aequatorialis in the och- raceous wash of the orange tawny on the throat and breast; differs from both in being slightly more greenish olive, less brownish olive above; tail longer than in sharpei (51 mm. as against 45 mm., in the latter). Measurements: male (type) wing 66, tail 51, culmen 14; female wing 66, tail 51, culmen 13 mm. The immature female is similar to the adult but has tawny orange terminal shaft stripes on the feathers on the forehead and crown. Alethe poliocephalus akeleyae Dearborn Alethe akeleyae Dearborn, Field Mus., Publ. Orn., 1, 1909, p. 170: Mt. Kenya. 1 c? , Karura Forest, Kenya Colony, 9 October 1920. Alethe kikuyuensis Jackson is a synonym. This race of the brown-chested Alethe is larger and more olivaceous, 1 Proc. N. Eng. Zool. CI., 12, 1931, p. 94. 254 bulletin: museum of comparative zoology less reddish above than carruthersi of Uganda and adjacent parts of western Kenya Colony and of the eastern Belgian Congo. The single specimen obtained is in fairly fresh plumage. Its dimen- sions are as follows: wing 96.0, tail 61.5, culmen 19.5, tarsus 24.5 mm. Alethe fulleborni usambarae Reichenow Alethe fulleborni usambarae Reichenow Orn. Monatsb., 13, 1905, p. 182: Mlalo, near Wilhelmstal (i.e. Lushoto), Usambara Mountains. 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, May 1922. This race of the white-chested Alethe appears to be confined to the Usambara and the Uluguru Mountains. The nominate form occurs in the Rungwe country southwest of Uhehe, northeast of Lake Nyasa. The present race is more yellowish brown, less rufous brown, above than the typical subspecies. Inasmuch as very few specimens of this form exist in collections, I append the measurements of this specimen — wing 100.0, tail 77.0, culmen 26.0, tarsus 33.0 mm. In his original description of usambarae Reichenow records a wing length of 103 mm. Grote1 records five specimens (including the type) all collected by Roehl in the western slopes of the Usambara Mountains. As far as I know, only six specimens are known, the original five and the present one. ClCHLADUSA ARQUATA Peters Cichladusa arquata Peters, Monatsb., Akad. Berlin for 1863, p. 134: Sena, on the Zambesi. 1 d\ 1 9, Lumbo, Mozambique, 10 July 1918. "Also Dar es Salaam, Kilosa, and Mombasa." (A.L.) This bird ranges from the Zambesi valley and Nyasaland north through Mozambique and Tanganyika Territory to the country west of Lake Tanganyika, and to Ankole, Uganda, and to Lamu on the coast of Kenya Colony. A female from Kongolo, Belgian Congo, is slightly smaller than the present female and has the reddish inner margins of the remiges some- what darker. It has a wing length of 83.5 as against 90.0 mm., in the Lumbo female; a culmen length of 19.5 as against 21.0 mm., in the > Journ. f. Ornith., 1921, pp. 137-138. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 255 latter. The Lumbo male has a wing length of 92.0, culraen 21.0 mm. More material may reveal a smaller form in Central Africa. The present two specimens are in fine fresh plumage. ClCHLADUSA GUTTATA GUTTATA (Heuglin) Crateropus guttatus Heuglin, Journ. f. Ornith., 1862, p. 300: Aniop, Bahr el Jebel; Upper White Nile. 1 cf, 1 9 , Morogoro, Tanganyika Territory, 6 July 1917. "Also Dodoma." (A.L.) These two specimens are somewhat intermediate between true guttata and rufipennis, but are nearer the former. Loveridge found this species breeding at Morogoro in February and April. "The mud nest is built upon a branch, or, more rarely, in a fork; it is very deep. The eggs are of a uniform blue colour . . . "-1 Erythropygia leucophrys zambesiana Sharpe Erythropygia zambesiana Sharpe, Proc. Zool. Soc. Lond., 1882, p. 588, pi. xlv, fig. 2: Tete, Lower Zambesi. 1 d\ 1 9 , Lumbo, Mozambique, 16 July 1918. Erythropygia ruficauda iubilaea Grote is a synonym. This form of this scrub-robin occurs from the Zambesi valley, the southeastern part of Northern Rhodesia and Nyasaland, and, along the coast at least to the Rovuma valley. In the Belgian Congo it is replaced by a darker form, ruficauda, and in Uganda by still another race, vansomereni, and in northern Tanganyika Territory by soror. The present two specimens are in very fresh plumage, and are some- what darker and redder above than the colored drawing in the original description (loc. cii.). According to Alexander2 the breeding season at Zumbo, on the Zambesi, is in December. Loveridge also collected this species at Dar es Salaam, Kilosa, and Kinyambwa, Tanganyika Territory. I have not seen these specimens, but judging from localities, they are probably E. I. soror. Bowen3 has recently reviewed the races of this species, with results that agree with the present allocation of Loveridge's specimens. 1 Proc. Zool. Soc. Lond., 1922, p. 846. 2 Ibis, 1900, p. 83. 3 Proc. Biol. Soc. Wash., 47, 1934, pp. 157-158. 256 bulletin: museum of comparative zoology Erythropygia leucophrys vansomereni Sclater Erythropygia leucophrys vansomereni Sclater, Bull. Brit. Orn. CI., 49, 1929, p. 62: Mokia, Ruwenzori, Mountains. 1 9 , Kabare, Bukoba, Tanganyika Territory, 14 January 1923. This specimen, which is in very worn plumage, I identify as van- somereni more by locality than by anything else. I have only one vansomereni with which to compare it — a freshly feathered male from Ruwenzori. The latter bird is darker, more richly rufous above, and is smaller (wing 65.5 as against 69.0 mm., in the present female). Erythropygia hartlaubi Reichenow Erythropygia hartlaubi Reichenow, Journ. f. Ornith., 1891, p. 62: Mutjora, Semliki valley. 1 cf, Ankole, Uganda, 5 September 1919. 1 9 , Ndeza, Ankole, Uganda, 7 September 1919. This scrub robin lives in the elephant grass areas in Uganda. Al- though Sclater1 states its range to be from Cameroon and northern Angola east to western Uganda, it really occurs east to Nairobi in south-central Kenya Colony. E. h. kenia van Someren is probably a synonym, as an example from Mt. Kenya does not differ from these Uganda birds. Erythropygia barbata rovumae Grote Erythropygia quadrivirgata rovumae Grote, Orn. Monatsb., 29, 1921, p. 109: Mbarangandu River, Upper Rovuma. 1 cf, Lumbo, Mozambique, 18 July 1918. This race of the bearded scrubrobin occurs from Inhambane and the Zambesi valley north to extreme southern Tanganyika Territory and southern Nyasaland. In central Tanganyika Territory it merges with a darker browner backed form quadrivirgata which extends from Morogoro and Kilosa to the coast and northwards to the mouth of the Tana River, inland to the Teita and south Ukamba districts in Kenya Colony. In southern Somaliland a small paler brownish-backed race, erlangeri, occurs. The nominate form barely ranges to East Africa, 1 Syst. Avium Aethiop., part ii, 1930, p. 484. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 257 its eastern limits being the western shores of Lake Nyasa and the southern end of Lake Tanganyika. It may be told from the eastern races by its under tail coverts which are isabelline, not white. In his original description of rovumae Grote writes that two speci- mens from the Usambara region are quadrivirgata, but one from Morogoro is intermediate between that form and rovumae. He suggests that the two races merge in the Morogoro region. However, the specimen taken by Loveridge is apparently a typical example of quadrivirgata. The measurements of the present specimen are as follows: wing 84.0, tail 76.0, culmen 21.0, tarsus 26.5 mm. Erythropygia barbata quadrivirgata (Reichenow) Thamnobia quadrivirgata Reichenow, Orn. Centralb., 1879, p. 114: Kipini, lower Tana River. 1 9 , Morogoro, Tanganyika Territory, 23 July 1917. 1 d% Kilosa, Tanganyika Territory, 17 February 1921. The characters and range of this form have already been dealt with, and need not be repeated here. Pogonocichla margaritata orientalis (Fischer and Reichenow) Tarsiger orientalis Fischer and Reichenow, Journ. f. Ornith., 1884, p. 57: Pangani, Tanganyika Territory. 1 adult d\ 1 adult 9 , Uluguru Mountains, Tanganyika Territory, 13 May 1921. 1 juvenal c\ 1 Juvenal 9, Bagilo, Uluguru Mountains, Tanganyika Territory, 8 June 1922. "Also Bungu, Morogoro, and Kilosa." (A.L.) This race is best characterized by the juvenal plumage which is more uniform green above, not as spotted as in guttifer, which form can hardly be told from the present one in adult birds. The young birds have the breast feathers margined with olive green, not with black as in guttifer and keniensis. Of this form P. c. helleri Mearns appears to be a synonym, but inas- much as I have seen no young birds of the latter, it is difficult to be 258 bulletin: museum of comparative zoology certain. There is no question, however, that montanus is a synonym of orientalis. Grote1 has examined the type of montanus and decided it was not distinct from orientalis. This race occurs in the Uluguru and Usambara Mountains, north to Mount Mbololo (?) and to the western part of the Tanganyika-Kenya border. Grote writes that the young of orientalis resemble those of johnstoni but are more greenish above and more spotted with yellow below. Belcher2 writes of johnstoni in Nyasaland, that, " . . . one nest had a pair of young, which were strongly marked with yellow on the back . . . The present two young birds are absolutely without any spots on the upper parts. Since the above was written, Moreau3 writes that the young of Usambara birds are spotted, not plain. Certainly this is not true of the Uluguru birds listed above. Peters and Loveridge4 have recorded juvenal helleri from Mt. Mbololo as being spotted. This may mean that helleri is distinct from orientalis, which would suggest a bewildering array of races in a rela- tively small area. Further study and more material must be gathered before any final answer may be expected. POGONOCICHLA MARGARITATA KENIENSIS Meams Pogonocichla cucullata keniensis Mearns, Smiths. Misc. Coll., 56, no. 20, 1911, p. 9: Mt. Kenya, 10,700 feet. 1 9 , Ngong, near Nairobi, Kenya Colony, 25 July 1919. 1 cf , Nairobi district, Kenya Colony, 24 August 1920. "Also Tumutumu." (A.L.) This race is similar to guttifer of Mt. Kilimanjaro but paler, more golden green, less brownish green on the back. It occurs in the high- lands of Kenya Colony from Mt. Uraguess and Mt. Kenya to Molo, Elgeyu, Escarpment, and Nairobi. It does not occur on Mt. Elgon, where it is replaced by P. m. elgonensis. 1 Journ. f. Ornith., 1921, p. 138. 2 Ibis, 1925, p. 813. 'Proc. Zool. Soc. Lond., 1935 (1936), pp. 880-881. * Bull. Mus. Comp. Zool., 79 1936, p. 179. FRIEDMANN AND LOVERIDGE : EAST AFRICAN ORNITHOLOGY 259 POGONOCICHLA MARGARITATA RUWENZORII (Ogilvie-Grant) Tarsiger ruwenzorii Ogilvie-Grant, Bull. Brit. Orn. CI., 19, 1906, p. 33: e. Ruwenzori; type from Mubuku valley, 7,000 feet. 1 cf, Singo, Ruanda, Uganda, 25 September 1919. Soft parts: iris brown, bill black; feet gray. The Ruwenzori race is smaller and darker on the back than keniensis or guttifer, but nearest the latter. Gyldenstolpe1 considers it the same as intensa Sharpe, but I prefer to follow Sclater's arrange- ment2 in all cases where I have not sufficient material to decide for myself. The specimen is in rather worn plumage. The breeding season is in January. This bird occurs from 6,500-12,000 feet on Ruwenzori, being found in the forest and the lower fringe of the bamboo zone. Irania gutturalis (Guerin) Cossypha gutturalis Guerin, Rev. Zool., 1843, p. 162: Abyssinia. 1 o71, 1 9 , Dodoma, Tanganyika Territory, 22 December 1918. Sclater3 states that this East Asiatic bird winters south along the, "... Persian Gulf to s. w. Arabia, the shores of the Red Sea from Eritrea to Somaliland, and southern Abyssinia; south to the Taveta district of Kenya Colony." Grote4 similarly gives no record south of Taveta. It appears, then, that the present two birds constitute the first records for the species in Tanganyika Territory, and extend the southern limits of the winter range by about 200 miles. The male is in fine fresh plumage; the female is similar but has worn rectrices. Luscinia luscinia (Linnaeus) Motacilla luscinia Linnaeus, Syst. Nat. 10th ed., 1758, p. 184: Europe; re- stricted type locality Sweden. 1 9 , Dodoma, Tanganyika Territory, 23 December 1918. This nightingale winters in East Africa as far south as Nyasaland and the Kafue River in Northern Rhodesia. The specimen collected is in freshly molted plumage. i Kungl. Sv. Vet. Akad. Handlngr., 1924, p. 155. 2 Syst. Avium Aethiop., part ii, 1930, p. 487. 3 Syst. Avium Aethiop., part ii, 1930, p. 491. 4 Mitt. Zool. Mus. Berlin, 16, 1930, p. 49. 260 bulletin: museum of comparative zoology Family SYLVIIDAE. Warblers Sylvia communis communis Latham Sylvia communis Latham, Gen. Syn. Suppl., 1, 1787, p. 287: England. 1 9 , Dodoma, Tanganyika Territory, 21 December 1918. Central Tanganyika Territory is about as far south as the white- throat usually occurs, although it has been taken on rare occasions in Southern Rhodesia and Damaraland. In Kenya Colony, and farther north, it is a very common winter visitor. The present specimen is partly in fresh plumage, rather earlier than usual for the molt to begin. Sylvia borin borin (Boddaert) Motacilla borin Boddaert, Tabl. PI. Enl., 1783, p. 35: ex Daubenton, PI. Enl. 579, fig. 2: Europe. 1 9 , Mwanza, Tanganyika Territory, 3 December 1922. The garden warbler winters throughout the regions covered by the present paper. It is one of the first European birds to arrive in the autumn and one of the last to leave in the spring. The single specimen obtained is in worn plumage. It is large, having a wing length of 77 mm. Hippolais icterina (Vieillot) Sylvia icterina Vieillot, N. Diet. d'Hist., Nat. 11, 1817, p. 194: France. 1 9 , Samumba, Singida, Tanganyika Territory, 25 February 1922. 1 9 , Mwanza, Tanganyika Territory, 6 December 1922. The bird taken at Samumba is in molt. As far as I have been able to discover, the icterine warbler has not been recorded previously from Tanganyika Territory. However, specimens have been taken in Kenya Colony on the north, and at Zumbo on the Zambesi, to the south. Grote1 writes that this palearctic bird has been recorded very seldom from East Africa. i Mitt. Zool. Mus. Berlin, 16, 1930, p. 29. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 261 ACROCEPHALUS ARUNDINACEUS ARUNDINACEUS (Linnaeus) Turdus arundinaceus Linnaeus, Syst. Nat. 10th ed., 1758, p. 170: N. Europe; Danzig, ex Klein. 1 9 , Kilosa, Tanganyika Territory, 3 January 1921. Sclater1 writes that typical arundinaceus winters south to Darfur, Sudan, Eritrea, and southwestern Arabia, "... occasionally to Natal, but birds from South Africa are generally referable to A. a. zarudnyi." For the winter range of zarudnyi, he writes, ". . . Somaliland, Tanganyika (near Dar es Salaam), and perhaps to Nyasaland; also to Natal, Transvaal, and Bechuanaland." It would appear then, that the present specimen should be of the Turkestan race, but it is certainly not different from typical arundinaceus. The Turkestan form zarudnyi is said to be less rufescent, more olivaceous, above, and paler on the flanks and under tail coverts than European birds. I have compared the present specimen with three of the latter and find them identical. Unfortunately, I have no Asiatic examples for comparison. Grote2 has omitted all mention of zarudnyi in his compilation of African records of palearctic birds, and calls them all A. a. arundin- aceus. He writes that the species is a common winter visitor in South Africa, but appears to be scarce in coastal Kenya Colony and Tangan- yika Territory. Van Someren3 finds his Ugandan and Kenyan birds are all typical arundinaceus. Lynes4 likewise finds that his birds from Njombe, southwestern Tanganyika Territory are arundinaceus and not zarudnyi. Recently van Someren5 has obtained zarudnyi on the Juba River. Acrocephalus griseldis (Hartlaub) Calamoherpe griseldis Hartlaub, Abh. Nat. Ver. Bremen, 12, 1891, p. 7: Nguru, Kilosa district, Tanganyika Territory. 1 9 , Kilosa, Tanganyika Territory, 12 February 1921. The Basra reed warbler has had a rather remarkable history. It was first discovered at Nguru in Tanganyika Territory, and, quite nat- 1 Syst. Avium Aethiop., part ii, 1930, p. 501. 2 Mitt. Zool. Mus. Berlin, 16, 1930, p. 27. ' Nov. Zool. 29, 1922, p. 231. " Journ. f. Orn., 82, 1934, Sonderheft, p. 84. 6 Nov. Zool., 37, 1932, p. 374. 262 bulletin: museum of comparative zoology urally, was thought to be an African form of the genus Acrocephalus. The type remained unique until Loveridge obtained two at Kilosa, and Noel van Someren obtained another at Nyambe crater on Mt. Kenya. Then in 1920, Ticehurst discovered an apparently new species in Iraq and named it A. babylonicus and it was not until 1922 that the matter was cleared up by the discovery that babylonicus and griseldis were identical, and that the bird breeds in Iraq and winters in East Africa. It appears to be scarce (or, at least, has been little collected) in both its summer and its winter quarters. Van Someren1 obtained two specimens in December on the Northern Guaso Nyiro River. Aside from the few Kenyan and Tanganyikan localities, this species has also been recorded in the Danakil coastlands of Eritrea as late as April 27 by Madarasz. The present specimen is in somewhat worn plumage. Because of its scarcity in collections, its measurements are here appended — wing 80.5, tail 61.5, culmen 22.0 mm. Acrocephalus baeticatus suahelicus Grote Acrocephalus baeticatus suahelicus Grote, Orn. Monatsb. 34, 1926, p. 145: Zanzibar. 2 (1 = d"), Dar es Salaam, Tanganyika Territory, 25 June 1918. As far as I know, these two constitute the first records for the African mainland for this race, hitherto known only from the islands of Zanzi- bar and Pemba. The present birds have wing lengths of 56 and 58 mm. respectively and therein agree with the figures given by Grote for this race (57-60 mm. as against 63-64 mm. for South African baeticatus). I do not know if Grote's Mikindani birds2 are suahelicus or not. His were the first records for the species in Tanganyika Territory. Aside from the Zanzibar and Pemba birds, the present two are the only ones taken subsequently in that political area. Grote found the breeding season at Mikindani to be in February. The present specimens are in rather worn plumage. ' Nov. Zool., 37, 1932, p. 374. 2 Journ. f. Ornith., 1913, p. 139. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 263 Acrocephalus schoenobaenus (Linnaeus) Motacilla schoenobaenus Linnaeus, Syst. Nat., 10th ed., 1758, p. 184: Europe; restricted type locality, S. Sweden. 2 9 , Kilosa, Tanganyika Territory, 22 April 1921. The European sedge warbler is a regular and common winter visitor and migrant throughout the region represented by the present col- lection. It occurs south as far as Natal and the Transvaal. Phllyoscopus trochilus trochilus (Linnaeus) Motacilla trochilus Linnaeus, Syst. Nat., 10th ed., 1758, p. 188: England (see Hartert, Vog. pal. Fauna, 1907, p. 507). 1 unsexed, Eldoret, Kenya Colony, 9 November 1920. 1 inside, and 13^2 inches in depth. Clutch consisted of three fresh eggs measuring 26 x 19 mm. Pale olive ground-colour, with blotches of purplish-brown and brown grouped around the larger pole; a few scattered elsewhere." Lanius excubitorius bohmi Reichenow Lanius bohmi Reichenow, Journ. f. Ornith., 1902, p. 258: Bogakatani, Tan- ganyika Territory. 1 cf, 1 9, Sagayo, Mwanza, Tanganyika Territory, 28 October 1922. "Also Ankole, Uganda." (A.L.) This race of the great gray shrike differs from the nominate form of the rest of Uganda and the southern Sudan, and from intercedens of Ethiopia and western Kenya Colony, in being darker above with a slight brownish wash on the gray of the upperparts. In size it is inter- » Cf. Proc. Zool. Soc. Lond., 1922, p. 844. 2 Proc. Zool. Soc. Lond., 1923, p. 904. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 303 mediate between the two. The wing lengths of the present specimens are 125 mm., in the male, and 121 mm., in the female. L. e. b'ohmi occurs from the northern end of Lake Nyasa north through western Tanganyika Territory to the Kivu district in the eastern Belgian Congo and to Ruanda and southwestern Uganda even to the Buddu Saza, Masaka district, Ankole). Van Someren1 records this bird as "Lanius excubitorius Rchw.", but this is undoubtedly a slip and was intended to read L. e. b'ohmi. He states that the Ethiopian race intercedens (excubitorius of his paper) is darker gray on the head and mantle than either the Ugandan ex- cubitorius (which he calls princeps) or the more southern b'ohmi. In this he is mistaken; I have examined a series of about 50 birds of all three forms, and b'ohmi is certainly the darkest in color. Lanius mackinnoni (Sharpe) Lanius mackinnoni Sharpe, Ibis, 1891, pp. 444, 596, pi. xiii: Kikuyu, Kenya Colony. 1 9 , Masomuntu Mukubwa, Ruanda, Uganda, 26 September 1919. 1 d\ Rukaya, Mawokota, Uganda, 3 November 1919. Grote2 has recently separated the birds of Cameroon and Spanish Guinea under the name zenkerianus. This form is said to differ from the eastern mackinnoni in having a slightly shorter wing and a notice- ably shorter tail. I have examined birds from Cameroon and from Uganda and Kenya Colony and find them identical in every respect. It is true that my series have been small, while Grote has examined no fewer than 33 of zenkerianus and 11 of mackinnoni, but I cannot help but feel that the deciding factor in the recognition of zenkerianus was not its characters, but the distance by which it is isolated from typical mackinnoni. If mere distance were of any importance in the production of races, then, to be logical we should be doubly suspicious of a form in which the characters are slight in spite of great geographi- cal isolation. Gyldenstolpe3 finds that birds from Central Africa (eastern Belgian Congo), ". . . appear to be slightly larger, the males having wings measuring 85-89 mm. and the females 86-87 mm., against 83 mm. in some males from Cameroon, and 85 mm. in males, 82-85 mm. in 1 Nov. Zool., 29, 1922, p. 123. 2 Ornith. Monatsb., 1924, p. 69. 3 Kungl. Sv. Vet. Akad. Handlngr., 1924, pp. 112-113. 304 bulletin: museum of comparative zoology females from the western parts of Kenya Colony." In other words, Cameroonian and Kenyan birds are more alike than those from the country in between. This shrike has only recently been recorded from Ruanda, the southernmost point at which it had otherwise been reported to be east of Lake Tanganyika having been Bukoba.1 Van Someren2 writes that in Uganda this shrike is not common. "Young in the first plumage were taken in June. These birds build their nests fairly low down in thorny bushes. They are composed of rootlets and twigs, and lined with fibres. The eggs are buff in ground- colour, spotted and streaked with ash and grey-brown. Two are usually laid." In another paper3 van Someren notes that this bird is particularly common in the Elgon district. The fact that Granvik4 never met with it during his extensive field work on Mt. Elgon suggests that this bird may be locally migratory. Its range is as follows: Kenya Colony west of the Rift valley (Kikuyu Escarpment, Nandi, Elgon, Kavirondo, Kaimosi), Uganda, northwestern Tanganyika Territory, Ruanda, the eastern Ituri district of the Belgian Congo, Cameroon, and Spanish Guinea. Lanius collurio (Linnaeus) Lanius collurio Linnaeus, Syst. Nat., 10th ed., 1758, p. 94: "Habitat in Europa;" restricted type locality, Sweden (apud Hartert, Vog. pal. Fauna, 1, 1907, p. 439). 1 , Kilosa, Tanganyika Territory, 1 January 1921. "Also Lumbo, Mozambique." (A.L.) The Kilosa bird is the farthest inland record of this race in Tangan- yika Territory. It may be slightly intermediate between littoralis and congener, but I have seen no specimens of the latter. I have not seen any birds from Lumbo, Mozambique, and am not at all certain that littoralis occurs there. Grote4 writes that minor reaches the coastal 1 Journ. f. Ornith., 1929, Erganzungsband ii, Festschrift Ernst Hartert, pp. 174-183. 2 Ibis, 1916, pp. 392-393. 3 Syst. Avium Aethiop., part ii, 1930, p. 624. * Journ. f. Ornith., 1912, p. 128. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 315 districts at Mikindani but that the form of Cape Delgado is congener. It is therefore not impossible that the Lumbo birds are also congener. P. a. littoralis is the smallest of the East African forms of this shrike, the wings being under 75 mm. in length. It is also whiter underneath than the inland form minor or than the race of Uganda and the 1 akes district— emini. It is restricted to the coastal country of the northern half of Tanganyika Territory and southern Kenya Colony north to the Tana River. The female from Dar es Salaam is a young bird with a brownish bill. POMATORHYNCHUS AUSTRALIS MINOR (Reichenow) Telephonus minor Reichenow, Journ. f. Ornith., 1887, p. 64: Kagehi, southeast corner of Lake Victoria. 1 cf, Ngong Forest near Nairobi, Kenya Colony, 23 July 1919. This is the form of the inland regions of East Africa from North- central Tanganyika Territory (Mwanza, Unyamwesi, Usambara, and Kilimanjaro districts) north to the Northern Guaso Nyiro River in Kenya Colony. All the forms of this shrike are denizens of the thorny scrub of the bushveld and Acacia savannas. The present race breeds from May to July and from September to November. POMATORHYNCHUS AUSTRALIS EMINI (Reichenow) Telephonus australis emini Reichenow, Orn. Monatsb., 1893, p. 60: Bukoba, n. w. Tanganyika Territory. 1 , Uganda, 10 November 1919. 1 cf, Tumutumu, Kenya Colony, 13 October 1920. If kikuyuensis Mearns be found to be recognizable (which is pos- sible), the Tumutumu bird would have to be referred to it. The Uganda specimen is a little less bluish above than the Kenyan one, but is in fresher plumage than the latter. The former is also larger than the latter. Cinnyris loveridgei Hartert Cinnyris loveridgei Hartert, Bull. Brit. Orn. CI., 43, 1922, p. 49: Uluguru Mountains, Tanganyika Territory. 2 cf , 3 9, Bagilo, Uluguru Mountains, Tanganyika Territory, 10 May-9 June 1922. These specimens are all topotypes of this distinct form. Sclater1 lists loveridgei as a race of C. regius, but I doubt if this expresses its true status as well as a binomial name. Loveridge2 has already published on some of these specimens and there is nothing new to add here. Chalcomitra amethystina kirkii (Shelley) Cinnyris kirkii Shelley, Monogr. Nectarin., 1876, p. 273, pi. lxxxv: "Zambesi distr.", probably Shupanga. 1 cf , Dar es Salaam, Tanganyika Territory, 15 June 1918. 1 9 , Dar es Salaam, Tanganyika Territory, 7 February 1919. "AlsoBungu." (A.L.) This race occurs along the coast of eastern Africa from the Zambesi mouth to the Pangani River, and inland to Mashonaland, Nyasaland, 1 Syst. Avium Aethiop., part ii, 1930, p. 698. * Proc. Zoo). Soc. Lond., 1923, pp. 900-901. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 339 and southwestern Tanganyika Territory. The crown patch of the male has a yellowish green sheen; the allied races kalckreuthi and doggelti have a bluish-green sheen. Chalcomitra amethystina doggetti (Sharpe) Cinyris doggetti Sharpe, Ibis, 1902, p. 116: Ravine, Kenya Colony. 1 cf , Nairobi, Kenya Colony, 31 July 1919. This specimen has the crown patch unusually blue even for doggetti and is therefore a little misleading in that it makes the difference between this form and kalckreuthi and kirkii seem greater than it really is. The bird is in fine, fresh plumage. Van Someren1 reports this form as common in the Nairobi area. He obtained young birds there in May and June. Chalcomitra senegalensis aequatorialis (Reichenow) Cinnyris aequatorialis Reichenow, Orn. Monatsb., 7, 1899, p. 171: Bukoba. 1 cf, 1 9, Kabura, Mawokota, Uganda, 26 August 1919. "Also Chantwara." (A.L.) The present form of the scarlet-breasted sunbird is a common bird in western Kenya Colony and in Uganda. It nests from April to July and from October to January. The two specimens obtained are in rather worn plumage. Chalcomitra senegalensis lamperti (Reichenow) Cinnyris senegalensis lamperti Reichenow, Journ. f. Ornith., 1897, p. 196: Moshi, Tanganyika Territory. 1 9 , Nairobi district, Kenya Colony, 6 September 1920. This race is not too well differentiated from aequatorialis, but the females of the lamperti are more grayish, less yellowish below. Chalcomitra senegalensis inaestimata (Hartert) Cinnyris gutturalis inaestimata Hartert, in Ansorge's "Under the African Sun," 1899, App. p. 351: "East Africa;" type in the Tring Museum from Dar es Salaam. » Ibis, 1916, p. 443. 340 bulletin: museum of comparative zoology 1 adult cf, 1 immature d\ Morogoro, Tanganyika Territory, 3 July 1917. 1 adult 9 , Mombasa, Kenya Colony, 31 May 1918. 3 immature d\ 2 adult 9 , Dar es Salaam, Tanganyika Territory, 7 June 1918. 1 immature d\ Morogoro, Tanganyika Territory, 7 August 1918. "Also Bungu, Kilosa, and Lumbo." (A.L.) The adults are in fresh plumage; the young ones are in molt. Loveridge1 found this bird nesting at Kilosa and Uonga on 15 Janu- ary, 12 February, 19 May, 18 July and 27 December. Chalcomitra veroxii fischeri (Reichenow) Cinnyris fischeri Reichenow, Journ. f. Ornith., 1880, p. 142: Mozambique. 1 d\ Mombasa, Kenya Colony, 23 May 1918. This sunbird is restricted to the coastal belt of Eastern Africa from Southern Mozambique north to Lamu in Kenya Colony. The present specimen is in slightly worn plumage. Cyanomitra verticalis viridisplendens (Reichenow) Cinnyris viridispoendens Reichenow, Journ. f. Ornith., 1892, pp. 54, 132: Bukoba, Tanganyika Territory. 1 d% Ruanda, Uganda, 28 September 1919. 1 9 , Tumutumu, Kenya Colony, 15 October 1920. "Also Kabare, Bukoba." (A.L.) The green-headed sunbird is a forest inhabiting species. It ranges from Ruwenzori and the Kivu district east through Ruanda and Uganda to the wooded areas of western Kenya Colony to Mt. Kenya and to Tumutumu. Both the present specimens are in fairly fresh plumage, the male especially so. Granvik2 has recorded considerable color variation in his birds from Londiani and Mt. Elgon. The present male has the throat and breast greenish with a slight bluish tinge. 1 Proc. Zool. Soc. Lond., 1923, p. 900. * Journ. f. Ornith., 1923, Sonderheft, p. 21G. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 341 Cyanomitra olivacea neglecta Neumann Cyanomitra obscura neglecta Neumann, Journ. f. Ornith., 1900, p. 297: Kibwezi, Ukamba, Kenya Colony. 1 9 , Uluguru Mountains, Tanganyika Territory, 24 May 1921. 1 9 , Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 1 d\ Bagilo, Uluguru Mountains, Tanganyika Territory, 10 May 1922. The birds from the Uluguru Mountains are slightly larger and greener above than the Usambara specimen, and therefore the former approach ragazzi to some extent. In fact, in a report on a collection of birds from these two mountain ranges1 the Uluguru birds were called ragazzi. A subsequent study of more ample material destroys the definiteness of this conclusion but unfortunately leaves the whole problem of the distribution and characters of neglecta and ragazzi in a rather unsettled state. These three birds are in fairly fresh plumage. Cyanomitra olivacea changamwensis (Mearns) Cinnyris changamwensis Mearns, Smiths. Misc. Coll., 56, no. 14, 1910, p. 4: Changamwe, Kenya Colony. 1 cf , Mombasa, Kenya Colony, 25 May 1918. This specimen is very large for changamwensis, having a wing length of 61 mm., thereby agreeing with the figures presented by Granvik2 who does not consider this form recognizable. All the races of the olive sunbird are rather poorly differentiated, but on the whole, in series, changamwensis is as good as any, and more distinct from neglecta and ragazzi than those two forms are from each other. Anthreptes collaris zambesiana (Shelley) Anthodiaeta zambesiana Shelley, Monogr. Nectarin., pt. 2, 1880, p. 243, pi. iii: Shupanga, Zambesi. 1^,1 9 , Lumbo, Mozambique, 15 July 1918. "Also Dar es Salaam." (A.L.) i Ibis, 1928, p. 92. 2 Journ. f. Ornith., 1923, Sonderheft, p. 217. 342 bulletin: museum of comparative zoology The reference to the original description of this bird as given above is taken from Sclater.1 However, in Shelley's work there is included a list of dates of publication of the various parts, from which we learn that the description of zambesiana appeared in part 1, issued July 28, 1876. In the monograph, as finally bound together, the pages are renumbered, but Anthodiaeta zambesiana is on the 345th page, not the 243rd. The Zambesi collared sunbird occurs along the coast of eastern Africa from Mozambique to the Pangani River in northeastern Tanganyika Territory. The present specimens are fully adult and are in slightly abraded plumage. Anthreptes collaris elachior Mearns Anthreptes collaris elachior Mearns, Smiths. Misc. Coll., 56, no. 14, 1910, p. 5: Changamwe, Kenya Colony. 1 , Rutaka, Ruanda, Uganda, 5 October 1919. The female is in worn plumage; the male is freshly feathered. This weaver occurs from Ruwenzori east across Uganda to Mt. Elgon. In the north Kavirondo country it merges with P. c. abyssinl- cus. It occurs northward into the Sudan, west as far as Darfur. The breeding season in Uganda is in March and April and in October. Ploceus rubiginosus rubiginosus Ruppell Ploceus rubiginosus Ruppell, N. Wirbelth., Vog., 1840, p. 93, pi. xxxiii, fig. 1: Abyssinia. 1 o*, Suna, Singida, Tanganyika Territory, 28 February 1922. 1 cf , Kinyambwa, Dodoma, Tanganyika Territory, 27 April 1922. "Also Mahaka." (A.L.) The chestnut weaver occurs from central Tanganyika Territory, north through Kenya Colony to southern Ethiopia. Schuster1 found this species two days march south of Dodoma. The birds were engaged in nesting at the time (3 March). The present specimens, taken in February and April, are in nuptial plumage, and may well have been breeding when killed. Ploceus pelzelni tuta (Bangs and Phillips) Icteropsis pelzelni tula Bangs and Phillips, Occ. Papers Bost. Soc. Nat. Hist., 5, 1925, p. 177: Busisi, Mwanza district, Tanganyika Territory. 1 d\ Kabura, Mawokota, Uganda, 19 August 1919. 1 9 , Ndeza, Ankole, Uganda, 10 September 1919. This large race of the slender-billed weaver occurs from the Mwanza area, north, west of Lake Victoria, to Ankole and southwestern Uganda generally. On the southeastern slopes of Ruwenzori it inter- grades with typical pelzelni, which differs only in its smaller size. 1 Journ. f. Ornith., 1926, p. 723. 356 bulletin: museum of comparative zoology The present speeimens are rather small for tuta having wing lengths of 62 and 63 mm., respectively. Both are in fairly fresh plumage. The breeding season is probably somewhat irregular. Nests have been found in July. Ploceus pachyrhynchus pachyrhynchus Reichenow Ploceus pachyrhynchus Reichenow, Orn. Monatsb., 1, 1893, p. 29: Kerevia, in the Semliki valley. 1 cf, 1 9, Kabare, Bukoba, Tanganyika Territory, 12 January 1923. These two speeimens are in somewhat abraded plumage. The compact weaver is chiefly a West African bird, but ranges across the Congo to Uganda east to Kisumu on the Kenya border and to Bukoba on the Tanganyika boundary. A single female has been taken in southwestern Ethiopia (Omo river) and separated as a race, omoensis, by Neumann.1 Van Someren2 writes that he has examined the type of omoensis and finds it quite like some Ugandan specimens. He therefore regards it as an untenable form. In Uganda this weaver is fairly common. Van Someren3 found the birds breeding in May in the tall elephant-grass. Ploceus ocularius crocatus (Hartlaub) Hyphantornis crocata Hartlaub, Abhandl. Nat. Ver. Bremen, 7, 1881, p. 100: Magungo, Victoria Nile. 1 cf, Kabura, Mawokota, Uganda, 18 August 1919. The Uganda spectacled weaver occurs from extreme western Kenya Colony west across Uganda and the Congo to Cameroon and Angola. It has the forehead, cheeks, and sides of the throat bright yellow, while the more eastern race suahclicus is brownish in those parts. The present specimen is in fairly fresh plumage. In Uganda the nesting season is from April to June. 1 Journ. f. Ornith., 1905, p. 342. 2 Nov. Zool., 29, 1922, p. 144. 3 Ibis, 1916, p. 412. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 357 Ploceus ocularius suahelicus Neumann Ploceus ocularius suahelicus Neumann, Journ. f. Ornith., 1905, p. 339: Lewa, Usambara. 1 cf, 1 9, Lumbo, Mozambique, 31 September 1918. 1 cf , Kilosa, Tanganyika Territory, 27 January 1921. "Also Mombasa, Bungu, and Dar es Salaam." (A.L.) The Kilosa bird is a juvenal specimen not fully grown. It resembles the female in color, but has a light brownish bill. It could not have been out of the nest more than three weeks when collected. The breed- ing season must be quite prolonged as Fischer found occupied nests late in August, and van Someren records them in June, July, and November. Ploceus nigricollis melanoxanthus (Cabanis) Hyphanturgus melanoxanthus Cabanis, Journ. f. Ornith., 1878, pp. 205, 232: Mombasa. 1 d\ Dodoma, Tanganyika Territory, 24 December 1918. "Also Mahaka." (A.L.) Dodoma is one of the southernmost localities from which this weaver has been recorded. The present specimen is peculiar in that it has a few pale yellow feathers among the black ones in the mantle and the lesser upper coverts of the left wing. It has a noticeably larger bill (culmen 20 mm.) and slightly longer wing than any comparable specimens from Kenya Colony and southern Ethiopia. More material may indicate a large subspecies at the southern end of the range of melanoxanthus. Ploceus aureoflavus aureoflavus Smith Ploceus aureoflavus A. Smith, 111. Zool. S. Afr., Aves, 1839, text to pi. xxx, fig. 1 : "W. Africa;" probably Zanzibar. 1 cf , Kilosa, Tanganyika Territory, 21 January 1921. "Also Dar es Salaam." (A.L.) ! Sclater1 considers castaneiceps and bojeri as races of aureoflavus, but I cannot agree with him as all occur together in the Taveta area. That 1 Syst. Avium Aethiop., part ii, 1930, p. 747. 358 bulletin: museum of comparative zoology they are closely allied is unquestionable. It is one of those cases when taxonomic procedure is inadequate to express the truth. The range of aureoflavus is from Zanzibar and the coastal belt of northern Tanganyika Territory and of extreme southern Kenya Colony, inland to the Teita and Taveta regions south to Morogoro and Kilosa. The present specimen from Kilosa is the southernmost record for the species and constitutes a considerable extension of known range. According to Sclater, the birds of Kilosa and Morogoro are castaneiceps and not aureoflavus, but the present individual is certainly different from castaneiceps from Taveta. The bird is in fresh plumage. I have not seen Meise's form reicherti, from Mbamba Bay, Lake Nyasa. Ploceus castaneiceps (Sharpe) Hyphantornis castaneiceps Sharpe, Cat. Bds. Brit. Mus., 13, 1890, p. 448, pi. xiii, fig. 5: Taveta. 1 9 , Morogoro, Tanganyika Territory, 23 March 1917. This specimen is referred to this form more by elimination than by anything else. It has very yellowish underparts and therefore cannot be aureoflavus. This leaves castaneiceps and bojeri to be considered, and as the former is known to occur at Morogoro while the southern limit of the range of the latter is considerably to the north of that area, the geographic evidence favors the identification here employed. The present specimen is subadult, which makes it all the harder to be sure of its determination. Ploceus bojeri (Cabanis) Hyphantornis bojeri Cabanis, in von der Decken, Riese, 3, Vogel, 1869, p. 32: Mombasa (ex MS. Finsch and Hartlaub). 1 d\ Mombasa, Kenya Colony, 30 May 1918. "Also Dar es Salaam." (A.L.) The specimen is in fresh, adult plumage. The golden weaver is a very common bird in the coastal districts of Kenya Colony and northern Tanganyika Territory. I have not seen the Dar es Salaam bird but it it is really bojeri, it constitutes the southernmost record of the species. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 359 Ploceus xanthops camburni (Sharpe) Hyphantornis camburni Sharpe, Bull. Brit. Orn. CI., 10, 1890, p. 35: Mt. Kenya. 2 tf, Kabale, Ruanda, Uganda, 20 September 1919. 1 9 , Nairobi district, Kenya Colony, 8 September 1920. 1 Zool. Anzeig., 78, 1928, p. 114. ' » Journ. f. Ornith., 1926, p. 725. 368 bulletin: museum of comparative zoology yellow in the anterior portions; the underparts are whiter than in a female of taha and the breast is marked with yellow, not washed with tawny as in ladoensis. The central Tanganyikan birds may well be an undescribed race. The dimensions of the two specimens are as follows: male — wing 59, tail 33, culmen 13; female — wing 59, tail 31, culmen 13 mm. The female is in worn plumage ; the male is in somewhat abraded breeding dress. Van Someren has recently1, recorded ladoensis from Eldama Ravine and from Lake Rudolf, and states that the, " . . . record of the race intercedens from Baringo probably refers to this race." While these data help to bridge the enormous gap between the Dodoma birds and those of the Sudan and Ethiopia, still I feel that the Tanganyikan specimens are not ladoensis, and perhaps not intercedens either. Urobrachya axillaris phoenicea (Heuglin) Coliuspasser phoeniceus Heuglin, Journ. f. Ornith., 1862, p. 304: Sobat River, Sudan. 1 cf, 1 9 , Masomunta Mukubwa, Ruanda, Uganda, 26 September 1919. 1 d\ Kome Island, Mwanza, Tanganyika Territory, 22 November 1922. U. a. media Sharpe, and U. a. neumanni Neunzig are regarded as synonyms. A number of authors maintain that media is valid, but I follow Sclater's arrangement2 as I have not the material wherewith to form an opinion of my own. The present form occurs from the Nile Valley in the Sudan (and northeast to Sennar), south through Uganda to western Kenya Colony, the eastern Belgian Congo, and the adjacent parts of Tan- ganyika Territory (south to Mwanza and Umbugwe). As mentioned under zanzibarica, there is one uncertain record from Uhehe. The September male and female are in fresh nuptial plumage; the November male is in worn breeding dress. Urobrachya axillaris zanzibarica Shelley Urobrachya zanzibarica Shelley, Proc. Zool. Soc. Lond., 1881, p. 586, pi. xxx, fig. 1 : type in Brit. Mus. from Malinda, that is, Malindi, Kenya Colony. 2 c?, 2 9 , Tindiga, Kilosa, Tanganyika Territory, 24 January 1922. i Nov. Zool., 37, 1932, p. 323. 3 Syst. Avium Aethiop., part ii, 1930, p. 764. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 369 One of the males is molting into breeding plumage; the other is in full nuptial dress. This race of the fan-tailed widow bird is chiefly coastal in its range, occurring from Lamu in Kenya Colony south to the Rufiji River in Tanganyika Territory. It gets inland to the Usambara Mountains and Kilosa. Schuster1 records a specimen listed as "phoenicea," from Uhehe. I do not know if Uhehe is inhabited by zanzibarica, by axillaris or by phoenicea. U. a. zanzibarica is distinguished from phoenicea by its larger bill. U. hildebrandli Sharpe and U. nigronolata Sharpe are synonyms of U. a. zanzibarica. COLIUSPASSER ALBONOTATUS ALBONOTATUS (Cassin) Vidua albonotata Cassin, Proc. Acad. Nat. Sci., Phil., 4, 1848, p. 65: Port Natal, that is, Durban. 1 cf , Muhalala, Dodoma, Tanganyika Territory, 13 March 1922. "Also Mahaka." (A.L.) The single specimen here listed is in worn breeding plumage. Sclater2 gives the northern limits of the range of albonotatus as the Rovuma River and Nyasaland, but it occurs north to Dodoma in Ugogo. In the last named area it meets with (but, as far as I know, does not intergrade with) eques. COLIUSPASSER ALBONOTATUS EQUES (Hartlaub) Vidua eques Hartlaub, Proc. Zool. Soc. Lond., 1863, p. 106, pi. xv: Kaseh, that is, Tabora, Tanganyika Territory. 2 d\ 1 9 , Nairobi, Kenya Colony, 22 May 1919 and 7 September 1920. The September specimen is in prenuptial molt; the May birds were a mated, breeding pair. This race of the white-winged whydah occurs in most of Kenya Colony (except the coastal and arid areas), Uganda and adjacent parts of the eastern Congo, and in northern Tanganyika Territory. The breeding male has the upper lesser wing coverts chestnut brown, instead of yellow as in albonotata. The breeding season appears to be very prolonged and irregular. i Journ. f. Ornith.. 1926, p. 725. 2 Syst. Avium Aethiop., part ii, 1930, p. 766. 370 bulletin: museum of comparative zoology COLIUSPASSER ARDENS ARDENS (Bodclaert) Fringilla ardens Boddaert, Tabl. PL Enlum., 1783, p. 39: Cape of Good Hope, ex PL Enl. 647. 2 d1, Morogoro, Tanganyika Territory, 23 May and 4 June, 1917. 1 9 , Rukaya, Mawokota, Uganda, 2 November 1919. "Also Kilosa and Kome Island." (A.L.) I have examined a number of specimens from Tanganyika Territory and also from South Africa and find Reichenow's form tropica1 to be untenable. The chief character of tropica is the color of the throat band which is supposed to be more reddish, less orange than in ardens. One of the present males has a deep scarlet throat collar; the other has an orange one. The cut-throat whydah occurs from eastern South Africa north to north-central Tanganyika Territory, Uganda, and the eastern and southern Belgian Congo, and Rhodesia. Sclater2 does not mention Tanganyika Territory in its range, but Schuster3 and others have noted it from several Tanganyikan localities, to say nothing of the fact that tropica was described from that country. The breeding season in Tanganyika Territory is from January to June, according to Schuster. Coliuspasser ardens suahelica (van Someren) Penthetria laticauda suahelica van Someren, Bull. Brit. Orn. CL, 41, 1921, p. 121: Nairobi River. 1 cf, 1 9 , Nairobi, Kenya Colony, 1 May 1919. The Kenya red-naped whydah is a common bird throughout the interior of Kenya Colony. In the coastal districts it is replaced by teitensis, which in turn, is very similar to typical ardens. Both specimens are in breeding plumage. The nesting season is rather indefinite and prolonged. This race differs from ardens in that the breeding male plumage has a red crown and nape connected postauricularly with the throat band, while in the nominate form only the throat band is red. ' Vog. Afr., 3, 1904, p. 135: type from Karema, Tanganyika Territory. 2 Syst. Avium Aethiop., part ii, 1930, p. 767. Uourn. f. Ornith., 1926, p. 726. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 371 Drepanoplectes jacksoni Sharpe Drepanoplectes jacksoni Sharpe, Ibis, 1891, p. 246, pi. v: Masailand, near Lake Nakuru. 2 d\ Nairobi, Kenya Colony, 1 May 1919. One male is in full nuptial plumage ; the other is in fairly fresh off- season dress. Jackson's whydah inhabits the grassy areas of the high interior of Kenya Colony from Kikuyu to Nandi north to Lake Baringo and Mt. Kenya. The breeding season is in May, June, and July. Spermestes cucullatus scutatus Heuglin Spermestes scutatus Heuglin, Journ. f. Ornith., 1863, p. 18: Dembea, Ethiopia. 3 9 , Nairobi, Kenya Colony, 7 October 1915. 1 d\ 1 9, Morogoro, Tanganyika Territory, 11 June 1917. 1 cf, 1 9 , Dar es Salaam, Tanganyika Territory, 7 June 1918. 1 d", 1 9, Lumbo, Mozambique, 31 July 1918. "Also Kilosa, Tumutumu, Tabora, and Kabare." (A.L.) The Ethiopian bronze mannikin is common and widely distributed throughout the regions represented in the present collection, except in Uganda, where it is replaced by the nominate form. Three of these birds are young (one each from Morogoro, Dar es Salaam, and Nairobi). At Kakoma and Zanzibar Bohm found nests with eggs and young from the middle of April to the end of May. Schuster1 writes that in Central Tanganyika Territory the birds breed throughout the year. He found nests in September, March, and April. In Nyasaland Belcher2 writes that some birds nest as early as August but most wait for the rains. He found nests with eggs as late as May. Spermestes bicolor stigmatophorus Reichenow Spermestes stigmatophorus Reichenow, Journ. f. Ornith., 1892, p. 46: Bukoba, Lake Victoria. 1 cf , 1 9 , Chantwara, Bukoba, Tanganyika Territory, 5-7 January 1923. "Also Kabare, Bukoba." (A.L.) i Journ. f. Ornith., 1926, p. 726. 2 Birds of Nyasaland, 1930, p. 325. 372 bulletin: museum of comparative zoology These specimens are practically topotypical of stigmatophorus, which is a rather doubtful form. It differs from poensis in being slightly browner, less glossy on the back. However, it may be noted that specimens of this race may be browner, as van Someren1 even suggests considering stigmatophorus and nigriceps conspecific, "... mostS. stigmatophorus showing a strong brownish tinge on the mantle." This bird occurs in the clearings and around the edges of the forests in Uganda and adjacent parts of the Congo, Kenya Colony, and Tan- ganyika Territory. It seems to be less numerous than S. cucullata cucullata but has similar habits and habitat preferences. Spermestes nigriceps nigriceps Cassin Spermesies nigriceps Cassin, Proc. Acad. Nat. Sci. Phil., for 1852, p. 185: Zanzibar. 1 a", 1 9 , Ngong, Kenya Colony, 23 July 1919. "Also Morogoro and Kilosa." (A.L.) These two specimens are in worn plumage. The rufous-backed mannikin occurs throughout the eastern terri- tories under consideration in this paper, but its range does not extend to Uganda. It is the least abundant of the species of its genus, but is widespread and not uncommon. The birds probably breed throughout the year. A nest was found in July near Mombasa by Hildebrandt and Kalkreuth. Amauresthes fringilloides (Lafresnaye) Ploceus fringilloides Lafresnaye, Mag. Zool., 1835, pi. xlviii: "India;" Liberia, apud Hartlaub, Syst. Orn. Westafr. p. 147. 1 d\ 1 9 , Kilosa, Tanganyika Territory, 26 April 1922. "This bird is considered rare and I never came across it un- til April 1922 when a perfect specimen was picked up dead near a water tank without any signs of injury. I showed it to my collector and told him to spend the day searching for others, but not to kill more than half-a-dozen. He returned a few hours later with these saying he had met with the birds near a wall in a maize plot and could have obtained more than the half-dozen if he had wished." (A.L.) i Nov. Zool., 29, 1922, p. 154. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 373 These two specimens are paler on the upper back and slightly less deep brownish on the wings and lower back than a lone female from Liberia (topotypical fringilloides). Whether or not the eastern birds are separable is a matter that requires much more material to decide. The magpie mannikin occurs throughout Mozambique, Nyasaland, and Tanganyika Territory, and in southeastern Kenya Colony, west (not in the interior of Kenya Colony or in Uganda) to Senegal. It is nowhere very numerous, but Roehl was able to obtain no fewer than 36 specimens during his residence in the western slopes of the Usambara range.1 Fischer found a nest with six eggs in Zanzibar. Belcher2 found a nest on March 27 in Nyasaland. Euodice cantans meridionalis (Mearns) Aidemosyne cantans meridionalis Mearns, Smiths. Misc. Coll., 61, no. 14, 1913, p. 4: Indunumara Mountains, Kenya Colony. 1 cf, 1 9, Kinyambwa, Dodoma, Tanganyika Territory, 8 April 1922. These specimens appear to constitute a new southern extension of the known range of the Kenya silver-bill. Previously it was known only from as far south as the Kilimanjaro region (Donya Ngai) and from Zanzibar. E. c. tavctcnsis van Someren appears to be a synonym of meridion- alis. If it were distinct, the present specimens would have to be identified as iarctcnsis, both because of geography, and because of their white underparts. The supposed characters of dorsal coloration and chin spots do not seem to hold good. Odontospiza caniceps (Reichenow) Pitylia caniceps Reichenow, Orn. Centralbl., 4, 1879, p. 139: Massa, Tana River. 1 c?, 1 9, Kinyambwa, Dodoma, Tanganyika Territory, 7 April 1922. "Also Shanwa." (A.L.) The gray-headed silver-bill occurs from central Tanganyika Terri- torv north through the interior of Kenva Colonv to southern Shoa, 1 Recorded by Grote, Journ. f. Ornith., 1921, p. 128. 1 Birds of Nyasaland, 1930, pp. 326-327. 374 bulletin: museum of comparative zoology northern and northwestern Uganda. It is nowhere common and its habits are largely unknown. The female is a young bird and has no white spots on the chin, upper throat, and cheeks. The male is molting the tail feathers and is otherwise in rather worn plumage. These specimens agree with others from Kilimanjaro, Kenya Colony, and Shoa. NlGRITA CANICAPILLA SCHISTACEA Sharpe Nigrita schistacea Sharpe, Ibis, 1891, p. 118: Sotik, Kenya Colony. 1 9 , Chantwara, Bukoba, Tanganyika Territory, 26 December 1922. 1 d\ Bukoba, Tanganyika Territory, 12 January 1923. "Also Kabare, Bukoba." (A.L.) These specimens are topotypes of V. sparsimguttata Reichenow1 but are not different from Kenyan examples of schistacea. Gyldenstolpe2 recognizes sparsivigidtata rather hesitantly; Sclater3 considers it a synonym of schistacea. The present form of the gray-headed negro-finch occurs throughout Uganda (west to Ruwenzori) and the eastern Ituri district, Belgian Congo, east to northwestern Tanganyika Territory and to western Kenya Colony, to the western escarpment of the Rift Valley. East of the Rift Valley, it is replaced by diabolica, a race with less whitish on the crown, and with the white marks on the upper wing coverts less like round spots, more like transverse terminal bars. The specimens here listed are in slightly worn plumage. This bird lives in the forests and in the clearings. Cryptospiza reichenowi sanguinolenta Vincent Cryptospiza reichenowi sanguinolenta Vincent, Bull. Brit. Orn. CI., 53, 1933, p. 148: Mlanje Mountain, southern Nyasaland. 1 d\ 1 9, Uluguru Mountains, Tanganyika Territory, 14 May 1921. 1 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 5 May 1922. The present specimens are in somewhat abraded plumage. This race is known to inhabit the mountains of Tanganyika Terri- 1 Journ. f. Ornith., 1892, p. 132. 2 Kungl. Sv. Vet. Akad. Handlygr., 1924, p. 52. 3 Syst. Avium Aethiop., part ii, 1930, p. 775. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 375 tory (.Usambara ; Uluguru; Uzungwe; Ukinga; and Poroto Mountains); and of Mozambique north of the Zambezi (Cholo Mt.; Namuli Mt.); and the Mlanje Mountain of Nyasaland. (ryptospiza salvadorii subsp. nov.? 1 9 , Mbeta, Uluguru Mountains, Tanganyika Territory, 24 July 1922. This single specimen is obviously C. salvadorii and not C. reiche- nowi, but it does not fit any of the races of the former. Sclater1 does not record the species from Tanganyika Territory at all, but Shelley2 notes that, "... in its most southern range the species has been met with by Dr. Stuhlmann at Uluguru." The bird is darker below than ruwenzori, but is olive-green, not brown as in the Nyasaland form australis. It may belong to a hitherto unknown race intermediate between these two, but more material is needed to make certain of this. For the present the bird may best be designated as above, and its absolute determination left for a more propitious occasion. The bird is in fairly fresh plumage; the amount of red on the lower back is intermediate between borcalis and ruwenzori; the crown is duskier olive green than in most specimens of either of these. The wing is short, only 50 mm., as against 55-56 mm., in females of ruwen- zori and 56 mm., in borcalis. PlRENESTES MINOR MINOR Shelley Pyrencstes minor Shelley, Ibis, 1894, p. 20: near Zomba. 1 cT, Uluguru Mountains, Tanganyika Territory, 26 May 1921. 1 o71, Bagilo, Uluguru Mountains, Tanganyika Territory, 30 May 1922. The range of the Nyasaland seed-cracker is from Beira, Mozambique, to Nyasaland and to the Uluguru Mountains. In the last named locality it meets with the large-billed race fro mmi. Probably the latter inhabits the more open country while minor lives in the moister forested areas. Aside from the Uluguru range it is known from only one other Tanganyikan locality — Sanji, Mahenge. Belcher3 writes as though the adult male has the back, wings, middle 1 Syst. Avium Aethiop., part ii, 1930, pp. 779-780. 2 Birds Afr., 4, 1905, p. 277. » Birds Nyasaland, 1930, pp. 327-32S. 376 bulletin: museum of comparative zoology of abdomen, and under wing and tail coverts, black, but in this he is mistaken, as minor is a brown and red, not a black and red, plumaged species. PlRENESTES MINOR FROMMT Kothe Pyrenestes minor frommi Kothe, Orn. Monatsb., 19, 1911, p. 70: Kitungulu, Urungu. 1 d\ Uluguru Mountains, Tanganyika Territory, 16 May 1921. The Uluguru Mountains are the northernmost area where the large- billed form of P. minor occurs. This specimen was at the Tring Museum where Chapin examined the series there and is mentioned by him1 in his paper — " ... P. m. minor likewise occurs in the same mountains, as shown by a young bird . . . secured by Loveridge. So the distribution of these two races of the East African species offers a problem at least as complex as in the case of P. oslrinus. That the small-billed race here again inhabits more wooded districts is probable at least, for Claude Grant, who collected it near Beira, tells us: 'It frequents densely wooded localities . . .' ..." Chapin points out that if additional collecting from the southern end of Lake Tanganyika should, "... reveal only adults of the ostrinus collection, then frommi would probably become a synonym of oslrinus or replace maximus, and the large race of minor would require renaming." The bird is in fairly worn plumage. Amadina fasciata alexanderi Neumann Amadina fasciata alexanderi Neumann, Bull. Brit. Orn. CI., 23, 1908, p. 43: Waram, Hawash River. 2 cf, Dar es Salaam, Tanganyika Territory, 24 June 1918. 1 cf , 1 9 , Kinyambwa, Dodoma, Tanganyika Territory, 8 April 1922. A.f. Candida is so very slightly marked as a race that I now consider it identical with alexanderi. The Dar es Salaam birds are browner above than those from Dodoma, but this may be due partly to wear as the former specimens are in fresh plumage and the latter are in abraded feathering. The cut-throat finch occurs in the northern half of Tanganyika Territory, Kenya Colony, Ethiopia, Somaliland, and Eritrea. i Bull. Amer. Mus. Nat. Hist., 49, 1924, pp. 430, 441. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 377 Ortygospiza atricollis dorsostriata van Someren Ortygospiza atricollis dorsostriata van Someren, Bull. Brit. Orn. CI., 41, 1921, p. 115: s. Ankole. 1 a" , Kabare, Bukoba, Tanganyika Territory, 13 January 1923. I have seen no material of 0. a. ugandae and so follow Sclater1 in considering it a synonym of dorsostriata. Van Someren2 writes that ugandae has white around the eye and on the chin, while dorsostriata has no white there. However, there seems to be some variation in this character, as van Someren writes that males of the present race have a little white on the chin, but not around the eye. The present specimen has no white in either place. 0. a. ugandae was described from Mumias, North Kavirondo, Kenya Colony. Therefore, if it is to be synonymized with dorsostriata, the range of the latter must be extended eastward to the Mumias region. Sclater has forgotten to do this and gives merely, "Western Uganda, including Ankole and Toro." It should also be extended to include extreme northwestern Tanganyika Territory. On the other hand, ugandae may be looked upon as nearer to mulleri (it appears to be a group of intermediates between dorsostriata and mulleri), and the dis- position of the name may be a matter of opinion, some synonymizing it with dorsostriata, others with mulleri. The present specimen is molting the rectrices and is otherwise in worn plumage. A female in breeding condition was taken at Butiti, Uganda, on 2 October by van Someren.3 Hypargos niveoguttatus (Peters) Spermophaga niveoguttata Peters, Journ. f. Ornith., 1868, p. 133; Inhambane. 1 d\ 1 9 , Kilosa, Tanganyika Territory, 1 January and 12 February 1921. "Also Bungu and Morogoro." (A.L.) Peter's twin-spot occurs sparingly and locally from Mozambique north throughout Tanganyika Territory to central Kenya Colony (Meru near Mt. Kenya). It also occurs in Nyasaland, Rhodesia and the Katanga. 1 Syst. Avium Aethiop., part ii, 1930, p. 783. = Nov. Zool., 29, 1922, p. 155. s Ibis, 1916, p. 424. 378 bulletin: museum of comparative zoology Sclater1 writes that it occurs only in the coastal districts of Kenya Colony (Lamu and Teita) but the species ranges inland to Meru. //. n. macros pilotus Mearns is a synonym (type examined and com- pared with birds from Mozambique, Gazaland, Nyasaland, and Tanganyika Territory). Pytilia afra (Gmelin) Fringilla afra Gmelin, Syst. Nat., 1, pt. 2, 1789, p. 905: Angola. 1 d\ 2 9 , Kilosa, Tanganyika Territory, 14 January, 18 January, 25 March 1921. The yellow-backed pytelia occurs throughout the northern half of Mozambique, north through East Africa to southern Ethiopia, and west through Rhodesia and the southern Congo to northern Angola. It is local and uncommon everywhere. The female collected on 25 March was shot off its nest which con- tained 4 eggs. Loveridge2 published this nesting record, but recognized P. a. cincreigula at the time. This form does not seem to be valid and may therefore be discarded. The present females are grayer below, less greenish or yellowish, than in Ethiopian birds; the present male has the hind crown and occi- put more yellowish, less grayish than in comparable Ethiopian examples. Pytilia melba grotei Reichenow Pytelia melba grotei Reichenow, Journ. f. Ornith., 1919, p. 227: Kionga, near mouth of Rovuma River, Tanganyika Territory. 1 9 , Morogoro, Tanganyika Territory, 23 July 1917. 1 cf , Lumbo, Mozambique, 6 August 1918. 1 cf , 1 9 , Kilosa, Tanganyika Territory, 14 January 1921. The Lumbo specimen is a topotype and was collected at the same time as the type of mosambica van Someren; now considered as a synonym of groiei. The Kilosa and Morogoro birds constitute the northernmost records for this race. The Morogoro female is in the "percivali" color phase with dark gray head and throat. This race and belli come close together in central Tanganyika 1 Syst. Avium Aethiop., part ii, 1930, p. 785. 2 Proc. Zool. Soc. Lond.. 1923, p. 903. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 379 Territory: — grotei ranges along the coast and inland to Kilosa and Morogoro, while belli ranges from Uganda and northwestern Tan- ganyika Territory eastward to Dodoma, less than 100 miles west of Kilosa. These specimens are in rather fresh plumage. Pytilia melba belli Ogilvie-Grant Pytelia belli Ogilvie-Grant, Bull. Brit. Orn. CI., 21, 1907, p. 14: S. E. Ruwen- zori; type from Mokia, Toro district, Uganda. 1 cf, 1 9 , Dodoma, Tanganyika Territory, 24 December 1918. These specimens constitute an extension of the known range east- ward as far as Dodoma. The male has white lores as in the supposed species, P. percivali van Someren, but the female is in typical belli plumage. Both are somewhat abraded. Lynes1 found this weaver breeding in March in the Iringa district. Lagonosticta rubricata haematocephala Neumann Lagonosticta rubricata haematocephala Neumann, Orn. Monatsb., 16, 1907, p. 168: Songea, east (not north) of Lake Nyasa. 1 cf, Kilosa, Tanganyika Territory, 14 January 1921. 1 9 , Uluguru Mountains, Tanganyika Territory, 14 May 1921. These specimens, neither of which is fully adult, appear to be best referred to haematocephala, although not quite typical of that race. They are of interest because of the locality. The form was previously known from Nyasaland, Gazaland, Mozambique, to coastal Tangan- yika Territory and to the Katanga. The present examples are the first from the interior of north-central Tanganyika Territory. They are in somewhat worn plumage; the male is either subadult or wrongly sexed as it resembles the female in plumage. Lagonosticta rubricata taruensis van Someren Lagonosticta jamesoni taruensis van Someren, Bull. Brit. Orn. CI., 40, 1919, p. 54: Tsavo. 2 d\ Kilosa, Tanganyika Territory, 13 January 1921 and 17 September 1922. 1 Journ. f. Orn., 82, 1934, Sonderheft, p. 124. 380 bulletin: museum of comparative zoology Sclater1 considers taruensis a race of L. rubricata. While I follow him in this regard, I am not at all certain that he is correct. The fact that these specimens were taken in the same place as some of L. r. haematocephala indicates that the two may well be specifically distinct. The color of the back in taruensis and jamesoni is so different from that in hildebrandti, haematocephala, etc., that it would be not unnatural to consider the reddish backed birds one species {jamesoni) and the olive- brown backed forms, another species {rubricata). If they are all kept as one specific unit, then the ranges of taruensis and haematocephala meet in the Kilosa area. The present race has the back even more reddish than in jamesoni, almost as much so as in senegala. One of the birds is immature and is much more brownish, less reddish generally above and below. Lagonosticta senegala kikuyuensis van Someren Lagonosticta senegalla kikuyuensis van Someren, Bull. Brit. Orn. CI., 40, 1919, p. 55: Nairobi. 1 9 , Tumutumu, Kenya Colony, 20 October 1920. 1 immature unsexed, Kilosa, Tanganyika Territory, 22 February 1922. The adult specimen obtained is in poor plumage. This fire-finch occurs commonly throughout the interior of Kenya Colony from the Mt. Kenya area south to the Kilosa district, Tan- ganyika Territory. The Kilosa bird is unfortunately immature, but seems to fit kikuyu- ensis better than any other race, although it is slightly intermediate between that form and somaliensis. It extends the known range of kikuyuensis southwards by about 250 miles. The species breeds throughout the year. Lagonosticta senegala somaliensis Salvadori Lagonosticta somaliensis Salvadori, Mem. Accad. Torino, 44, 1894, p. 557: Somaliland. 1 d\ 1 9, Morogoro, Tanganyika Territory, 2 April and 6 September 1917. "Also 5 birds from Dar es Salaam." (A.L.) The "male" is probably a female as it has the plumage of that sex. 1 Syst. Avium Aethiop., part ii, 1930, p. 789. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 381 These birds constitute an extension of range for somaliensis which was previously known only as a coastal bird in northeastern Tangan- yika Territory. They are not wholly typical, but show a very slight approach towards kikuyuensis but are much nearer the former. Lagonosticta senegala rendalli Hartert Lagonosticta senegala rendalli Hartert, Nov. Zool., 5, 1898, p. 72: Upper Shire River. Loveridge collected five of these fire-finches at Lumbo, Mozambique, for the Nairobi Museum, where they now are. Coccopygia melanotis kilimensis Sharpe Coccopygia kilimensis Sharpe, Cat. Birds Brit. Mus., 13, 1890, p. 307: Kilimanjaro. 1 d\ 1 9 , Morogoro, Tanganyika Territory, 2 July 1917. 1 9 , Uluguru Mountains, Tanganyika Territory, 28 May 1921. "Also Nairobi, Bagilo, and Kilosa." (A.L.) The Kenya yellow-bellied waxbill occurs in suitable areas through- out the regions under discussion in this report, except Uganda and adjacent portions of the eastern Belgian Congo, northwestern Tan- ganyika Territory and the north Kavirondo country of western Kenya Colony, in which areas a paler race, nyanzae, replaces it. The present specimens are in somewhat worn plumage. They have rather short wings, but are matched in this regard by some birds from Kilimanjaro (topotypes). In Nyasaland Belcher1 found it nesting in February and March. Estrilda astrild minor (Cabanis) Habropyga minor Cabanis, Journ. f. Ornith., 1878, p. 229: Voi River, Kenya Colony. 1 adult 9 , Uluguru Mountains, Tanganyika Territory, 28 May 1921. 1 adult d", 1 juvenal 9, Bungu, Usambara Mountains, Tanganyika Territory, September 1921. 1 adult 9 , Bagilo, Uluguru Mountains, Tanganyika Territory, 22 May 1922. « Birds Nyasaland, 1930, p. 332. 382 bulletin: museum of comparative zoology The Bagilo bird is dark above and approaches the Nyasaland form, cavendishi. I have seen no material of Grote's form, litoris, described from the mouth of the Ruvu River1 and therefore accept Sclater's disposition of it as a synonym of minor.2 Estrilda astrild massaica Neumann Estrilda astrild massaica Neumann, Journ. f. Ornith., 1907, p. 596: Njoro, Kenya Colony. 1 cf, 1 9 , Nairobi, Kenya Colony, 28 September 1920. This is the common waxbill of the interior of the southern half of Kenya Colony. It differs but slightly from the coastal form minor, but has the chin and cheeks slightly less whitish than the latter race. Both specimens are in worn plumage. The breeding season is from March to July and from November to January. Estrilda astrild nyanzae Neumann Estrilda astrild nyanzae Neumann, Journ. f. Ornith., 1907, p. 596: Bukoba, Lake Victoria. 1 cf, 1 9 , Kabura, Mawokota, Uganda, 25 August 1919. 1 cf, 1 9 , Buchosa, Bukoba, Tanganyika Territory, 30 November 1922. The Uganda waxbill differs from massaica in having the underparts more suffused with pinkish, the bars obsolete on the breast, the upper- parts slightly more grayish, less brownish, and in being larger. The November specimens are in worn plumage; the August birds are in fairly fresh feathering. This bird is very common in Uganda where it breeds from March to July and from November to January. Estrilda subflava clarkei (Shelley) Coccopygia clarkei Shelley, Bull. Brit. Orn. CL, 13, 1903, p. 75: Richmond Road, Natal. 1 d\ Morogoro, Tanganyika Territory, 30 November 1918. 1 cf, 1 9 , Nairobi, Kenya Colony, 29 September 1920. 1 9 , Kilosa, Tanganyika Territory, 12 January 1921. 1 Journ. f. Ornith., 1919, p. 301. ! Syst. Avium Aethiop., part ii, 1930, p. 795. FRIEDMANN AND LOVERIDGE: EAST AFRICAN ORNITHOLOGY 383 The southern zebra waxbill occurs, in the territories under con- sideration in this report, in Mozambique, Tanganyika Territory, and Kenya Colony. In Uganda it is replaced by the nominate race. The birds breed more or less throughout the year, although chiefly from January to August. Occasionally the birds use old nests of other species, but more frequently they build for themselves. Estrilda roseicrissa roseicrissa Reichenow Estrilda roseicrissa Reichenow, Journ. f. Ornith., 1892, p. 47: Bukoba. 1 unsexed, Kabura, Mawokota, Uganda, 20 August 1919. The rosy-flanked waxbill occurs in western Uganda, and northern Tanganyika Territory south to the Kivu district. The single specimen obtained has the crown and occiput more gray- ish than the back, while two from Nyanza on the northeastern shore of Lake Tanganyika have the head brownish, concolorous with the back. It therefore appears that the present bird is subadult. Gylden- stolpe1 describes a similar subadult bird. Estrilda nonnula nonnula Hartlaub Astrilda nonnula Hartlaub, Journ. f. Ornith., 1883, p. 425: Kudurma, Bahr el Ghazal. 1 9 , Chantwara, Bukoba, Tanganyika Territory, 26 December 1922. 1 o% Kabare, Bukoba, Tanganyika Territory, 17 January 1923. The black-crowned waxbill occurs in the western parts of Kenya Colony and of northern Tanganyika Territory, west through the northern Congo and southern Sudan to Cameroon. Van Someren2 writes that birds from, "... South Elgon are rather whiter below, not tinged with creamy, and have the breast and flanks more washed with grayish . . . ' I have seen a number of specimens from Kaimosi and find them similar to others from Bukoba, and from Ruwenzori and from the Lualaba. In Uganda this bird is fairly common in the native shambas, in the coffee plantations, and in the wooded grasslands. It breeds throughout the year. 1 Kungl. Sv. Vet, Akad. Handlngr., 1924. p. 61. 2 Nov. Zool., 29, 1922, p. 164. 384 bulletin: museum of comparative zoology ESTRILDA ERYTHRONOTOS DELAMEREI Sharpe Estrilda delamerei Sharpe, Bull. Brit. Orn. CI., 10, 1900, p. 102: Athi River, Kenya Colony. 1 , wing 130, tail 125. Due to the marked difference in size between the sexes, any size differences indicated should be confirmed by additional material. Notes on Nyctipolus nigrescens (Cabanis) Birds from the south side of the Amazon average very slightly larger than Guiana specimens, as the following wing measurements show. Lower Amazon 4 c? 142-150 Guiana 3 c? 139-140 Lower Amazon 19 144 Guiana 3 9 139-146 These differences are insufficient for formal description. N. maculo- sus Todd (Proc. Biol. Soc. Wash., 1920, p. 76) the type of which we have examined, cannot possibly be a distinct species. It is not quite fully adult, and the greater degree of the white spotting on the pri- maries and rectrices is exactly the same character that distinguishes N. whitelyi (Salvin) from nigrescens, except that the white on the rectrices is terminal as in nigrescens, not subterminal as in whitelyi. In other words maculosus Todd is a single individual which varies GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 423 towards whitelyi. A Cayenne trade-skin before us approaches macu- losus in having a narrow whitish tip to the outermost rectrix. There are really three subspecies as follows : 1. typical nigrescens (Cabanis). Male: white spot on primaries 2, 3, 4; white tip on both webs of rectrices 2 and 3, rarely with an oblong tip of white on inner web of outermost (a maculosus character). Female: primaries black; no white tips to rectrices. 2. whitelyi (Salvin). Male: white spot on primaries 1, 2, 3, 4; white on rectrices 2 and 3 subterminal, on inner web only (a maculosus character). Female: fulvous bar replaces white spot on first primary. 3. duidae subsp. nov. Type: No. 147,396, Mus. Comp. Zool.; d" ad.; Valle de los Monos, alt. 725 m., Mt. Duida, Venezuela; Nov. 9, 1928; coll. Olalla. Much smaller white spots on primaries 3 and 4 only; white tips to rectrices half the extent of nigrescens; much silvery grey freckling on scapulars and secondaries. Female: darker below, with smaller spots on the chest. In this species the amount of grey on the pileum, the presence or absence of a superciliary stripe and the relative distinctness of the barring of the abdomen are age characters. All have been claimed as specific. Notes on Caprimulgus rufus (Boddaert) The receipt of two males from northeastern Brazil (Rio Tapajoz, Tauary and Pinhy), which presumably represent true rufus, enables us for the first time to give some notes on the variations of this little known bird. It proves necessary to revise radically the diagnoses of the two currently recognized races. Neither Bangs nor Ridgway ever saw true rufus. Ridgway's description of rufus is based on very different birds from Venezuela, Colombia and Panama, and his diagnosis of otiosus Bangs from St. Lucia is valid for this same mate- rial, but requires revision when compared with true rufus. We recog- nize the following races. 1. typical rufus Boddaert. Guianas and northeastern Brazil. A large bird, wing of 3 d* 165-174, 1 9 184. General coloration darker and more rufescent; spotting on the inner webs of primaries relatively slight; barring on under tail coverts relatively heavy. A Bahia speci- men in the American Museum also belongs here. 2. rufus rutilus Burmeister. As large as typical rufus, wing 181-185 mm.; general coloration similar to rufus, but separable at a glance 424 bulletin: museum of comparative zoology by the paler, less rufescent plumage; the pectoral collar almost pure white in part; spotting on primaries more extensive; under tailco verts sometimes nearly immaculate; 1 cf 1 9 ad. ; from Tucuman, Argentina April 9, 1916 and Nov. 14, 1918. 3. rufus otiosus Bangs. Island of St. Lucia, Lesser Antilles. Per- haps very slightly larger than typical rufus; general coloration much less rufescent, particularly on lower underparts ; spotting on primaries and undertailcoverts as in rufus. It follows, therefore, that this race is far less distinct from typical rufus than previously supposed. 4 cf 2 9 examined, the males distinctly larger than the two males of rufus, the females about the same size as rufus, 188-189 mm. 4. rufus minimus subsp. nov. Type. No. 114,053, Mus. Comp. Zool.; 9 ad.; Panama City, Panama, May 6, 1904; coll. W. W. Brown. Characters. Much less rufescent than rufus in general coloration, and smaller, wing of 3 9 169-174 mm.; chin and throat in adult more tawny, consequently paler than in rufus, the pectoral collar also tawnier and darker, consequently less contrasted with the throat. Two males and four females, Panama, Colombia and Venezuela (San Julian and Merida). Venezuela birds may prove separable from Panama birds. Remarks. As perhaps can be inferred from the diagnosis above, some of the racial characters alleged by Ridgway to distinguish otiosus from rufus are matters of maturity. The two females from Colombia and Venezuela are immature birds. In these the abdomen is lighter and more buffy, thickly barred with blackish. The under tailcoverts are less barred than in adults, and there is far more spotting on the inner webs of the primaries. These last two points are consequently not racial characters. It was these two birds that Bangs and Ridgway both took to represent typical rufus. Nomenclatural points are somewhat confused. Antrostomus rutilus Burmeister ("Brazil"; we designate southeastern Brazil) was appar- ently proposed without knowledge of the prior rufus. Pelzeln at one time maintained that rufus should be replaced by rutilus on the ground that the wretched plate in the "Tables Planches Enluminees" was not really identifiable. This point was accepted by Sclater (P.Z.S., 1866, p. 586), who quotes Pelzeln in litt. and who suggested that rufus be called rutilus, and who described ornatus from Rio de Janeiro. This was based on a misconception of tail characters, and is a pure synonym of rutilus Burmeister. Pelzeln apparently changed his mind two years later in Ornith. Bras., p. 13. Here appears Antrostomus GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 425 cortapau ex Natterer ms. The type locality would be the first locality mentioned, Engenho do Capitao Gama, which is fifteen leagues from the city of Matto Grosso. (cf. Itinerary, p. X). An important foot- note states that two specimens from Bahia and the type of cortapau differ from a Para specimen in being yellower in general coloration. This seems to us to apply clearly to the pale extreme as represented by the Tucuman specimens cited above, and suggests a southern pale subspecies from most of southern Brazil southward. We feel it is much safer to call this bird rutilus Burmeister rather than describe the Argentine bird as new, until evidence to the contrary appears. Bahia birds may prove to be variable intermediates, as the only one seen by us is clearly inseparable from the Rio Tapajoz birds. Hydropsalis climacocerca canescens subsp. nov. Type. No. 173,621, Mus. Comp. Zool.; cf adult; Lago Grande south bank of the Amazon, west of Rio Tapajoz, Para, Brazil, Sept. 11, 1932; coll. A.M. Olalla. Characters. Resembling typical climacocerca of far upper Amazonia, but the adult male notably paler and greyer above, less buffy and white below; the spotting on the wings as a rule pure white rather than buffy or even rusty; black streaking on pileum narrower; adult female apparently slightly paler below, but the light bars on all but the outer- most rectrices paler and greyer, less buffy and ochraceous. Material Examined, climacocerca: Manacapuru, Rio Solimoes, 3 cf 1 9 (Carnegie Museum), canescens: Santarem, 14 cf 5 9 and Rio Tapajoz, 1 cf (Carnegie Museum) ; also Lago Grande 2 cf 2 9 , and Rio Tapajoz, 1 cf • Remarks. In default of Peruvian topotypes, we assume that the series from the Rio Solimoes represents true climacocerca. Even if it should prove not to be so, there can be little doubt that our bird represents a distinct form, as a very different, a remarkable rich and buffy bird occurs on the Rio Purus much nearer our area. On the north bank of the Amazon at Obidos 1 cf 4 9 in the Carnegie Museum repre- sent another very distinct subspecies which connects schomburglci of British Guiana with climacocerca. We have here a graphic illustration of perfect modern series proving extensive geographic variation pre- viously unsuspected in a little known bird. Mr. Todd is about to describe the two races mentioned above of which the M.C.Z. possesses no material. A word about trifurcata Tschudi, first proposed in Wiegmann's 426 bulletin: museum of comparative zoology Archiv., 1844, as a nomen nudum, but validated by Tschudi in his Fauna Peruana ex Natterer MS. Sclater and Salvin as the first re- visers (P.Z.S., 1866, p. 193) restricted this name to the first locality mentioned, viz. Lower Ucayali, Peru, thus making it a straight syn- onym of climacocerca. It consequently makes no difference that Nat- terer obtained specimens on the Rio Madeira, which might prove to belong to the form here described. Pharomachrus pavoninus viridiceps subsp. nov. Type. No. 47,852 Mus. Comp. Zool.; cf ad.; lower Amazon, Brazil; coll. C. M. Calverly. Characters. Resembling typical pavoninus Spix of upper Amazonian Brazil and eastern Ecuador, but upper tail coverts not reaching to tip of tail, instead of slightly surpassing the tail, in fully adult males; head green with little or no golden reflexions, instead of golden or bronzy golden in sharp contrast to green back; female darker green and browner below. Remarks. Whether pavoninus is specifically distinct from auriceps Gould or not is largely a matter of opinion. The two birds differ only in the color of the bill and in size. As regards the latter point, however, pavo?ii?ius is distinctly smaller than true auriceps, 175-185 mm., versus 193-200 mm., but is little if any smaller than auriceps heliactin of western Ecuador. Typical pavoninus is a little known bird in collections, and definite localities are few. It has never been definitely recorded from Lower Amazonia. We have a 9 from the Rio Tapajoz, Tauary, and 2 cf ad and 2 9 from the Thayer Expedition to Brazil in addition to the type. While none of these latter birds have a definite locality, they belong here on the basis of their characters, and Newton Dexter who collected most of the birds, was in Lower Amazonia only. Galbula leucogaster viridissima subsp. nov. Type. No. 173,977, Mus. Comp. Zool.; c? ad.; Rio Tapajoz, Pinhy; May 8, 1933; Olalla Bros. Characters. Differing from typical leucogaster of Surinam in being strikingly greener, less coppery above and on chest, and central tail feathers averaging about 10 mm. longer. Remarks. This southern extreme from the south side of the Amazon River is distinct at a glance. True leucogaster has a very extensive range, as we have specimens from Mt. Duida, Venezuela and the Rio Purus in north central Brazil, which are not satisfactorily separable. 5 cf 2 9 of the new form examined. GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 427 Westward a very rapid transition takes place to chalcothorax Sclater, currently treated as specifically distinct. This much larger bird with purplish coppery upperparts and chest is represented in the collection by four specimens from the Rio Suno which agree perfectly with the beautiful colored plate of the type in Sclater's Monograph. From the Rio Curary further east in Amazonian Ecuador, however, we have two specimens, which are perfect intermediates between typical leucogaster and chalcothorax. These birds make it quite im- possible to maintain the latter as specifically distinct. We do not believe that the description of a purely intermediate race is warranted, until evidence is available to show that these characters occupy a reasonably extensive range. 343. Ramphastos tucanus oblitus subsp. now Type. No. 174,070, Mus. Comp. Zool.; c? ad.; RioTapajoz,Tauary; May 7, 1933; A. M. Olalla. Santarem (Chapman and Riker, as R. erythrorhynchus). Characters. Resembling typical tucanus Linnaeus in size, but upper tailcoverts sulphur yellow with darker, more orange tips, or nearly uniform bright orange yellow; strikingly different from typical tucanus, which has uniform lemon or gamboge yellow upper tailcoverts; resembling cuvieri in having the bill largely blackish, rather than sanguineous red. Remarks. Any proper treatment of one of these Toucans involves a thorough review of the entire group, beset with numerous complexi- ties, both systematic and nomenclatural. Fortunately four great and keen students of neotropical birds have written notable critiques on this situation in recent years, in each case reporting the evidence of new and interesting material examined by them. These are Hellmayr in a series of famous papers on the birds of Amazonian and north- eastern Brazil; Hartert in two papers in Novitates Zoologicae for 1902 and 1925; Chapman in his reports on the birds of Colombia and Ecuador; and Zimmer in his report of the birds of the Marshall Field Peruvian Expedition. Two outstanding facts of great biological interest have emerged as a result of these studies: — (1) Chapman proved that in upper Amazonia two species, which he called cuvieri and culminatus, occurred together, differing only in size and proportions and exactly alike in coloration. (2) Hellmayr (Novit. Zool., 1905, p. 299) showed that two species occurred together in the Guianas, which differed only in the color of the bill. This point seems to have been overlooked or ignored by all 428 bulletin: museum of comparative zoology workers since, whose collections from the Guianas were either in- adequate or erroneously determined. Another factor, which has obscured the issue in the past, is the fading of the color of the bill of the living bird in death, and the proper allowances to be made for this, when there are both specific and sub- specific characters in the color of the bill. The outstanding illustration of this is the old key in the Catalogue of Birds, Vol. 19, where there were three "species" as follows: a. Bill red, etc erythrorhynchus Gmelin b. Bill black, etc. a. Bill with a reddish basal spot inca Gould b. Bill uniform black cuvieri Wagler As a matter of fact these three "species" boil down to two sub- species. It is quite possible that there may be some average bill color difference in life between erythrorhynchus Gmelin (= monilis P. L. S. Miiller; = tucanus Linnaeus) of Guiana and lower Amazonia and cuvieri Wagler of upper Amazonia, the bill of the former often being described as "dark sangineous red" and the latter by Hoffmann as "blackish." In dead birds, however, this difference often disappears. Thus a series before us from the Rio Tapajoz all taken the same year (1933) would key down to all three "species," though we are convinced that only one subspecies is involved. All that can be said then of dead birds is that many specimens of erythrorhynchus clearly show they must have had red bills, while others appear all blackish; in cuvieri all specimens seen and all others recorded would seem to show a blackish bill with at most a reddish basal spot. It was Zimmer, who recently emphasized these bill characters and suggested that inca Gould might disappear from nomenclature. We agree heartily. There is a nomenclatural moral to these points. The identification of old names had better be made with extreme caution by some expert thoroughly familiar with the complexities of the problem. The brief descriptions of these old names were based on earlier plates of the greatest crudity. We should remember that (1) author and artist were entirely unaware of details, which we now know are of specific or racial importance and (2) we do not know whether the artist was making a crude painting of a living bird, or whether he was making an equally crude reconstruction of a dead bird. To continue our systematic summary, we have then the following species and subspecies: A. tucanus Linnaeus. Bill without the concavity on the side of the maxilla below the culmen ; of relatively large size. GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 429 1. tucanus tucanus. Bill dark sanguineous red; upper tail coverts lemon or gamboge yellow; wing of adults, 225-229 mm.; culmen, 142-146 mm. The Guianas and northeastern Brazil. Birds from Para approach the next in having sulphur to orange upper tail coverts. Birds from the Rio Xingu are still another connecting link (fide Zimmer). 2. tucanus oblitus Griscom and Greenway. Bill blackish, with at most a reddish basal spot; upper tail coverts rich sulphur or orange yellow to orange; size similar to the last. Amazonian Brazil (right bank of Rio Tapajoz presumably to left bank of Rio Xingu). 4 d1 1 9 examined. 3. tucanus cuvieri Wagler. Similar to the last in bill and color characters, but much larger; wing of cf 242-258 mm., bill 190-217 mm. Wagler's type locality was "Brasilia versus flumen Amazonum." Helhnayr, however, compared speci- mens from the Rio Madeira with Wagler's type and found them identical. We consequently restrict the type locality to Borba, Rio Madeira. Known definitely from western and northern Matto Grosso, the Rio Madeira, the Rio Purus and the Rio Negro to the Rio Tapajoz, west to the eastern base of the Andes, Bolivia to Colombia. Birds from Mt. Duida, Venezuela are intermediate between cuvieri and oblitus. The bill is as long as the tail or longer. Large series examined. B. aurantiirostris Kartert. Bill without the concavity on the side of the maxilla, just as in tucanus. Coloration and size exactly as in typical tucanus (including lemon yellow upper tail coverts), but bill fiery or orange red in life instead of dark sanguineous; in death pale yellowish instead of dark reddish or blackish. Known onlv from Surinam and British Guiana (where much commoner than tucanus) and Venezuela. As shown by Helhnayr (he. cit.) this bird exists side by side with tucanus in Surinam, and he suggested that the light billed bird needed a name. Hartert (Novit. Zool., 1925, p. 143) pro- posed monilis aurantiirostris for this bird, overlooking or dis- believing Hellmayr's evidence. The Penard collection from Surinam contains both species from the vicinity of Paramaribo, and there is no difficulty in separating the two birds, including stubby billed juvenals. We have also seen three specimens from British Guiana in the American Museum. That institution has no specimens of true tucanus from British Guiana. We must here sav a few words about the nomenclature. We 430 bulletin: museum of comparative zoology heartily agree with Peters (Auk, 1930, p. 405) that Linnaeus' description and references for tucanus are clearly referable to this species in the "formenkreis" sense. But it is not definitely identifiable as either of the two species here defined. We have the right to apply the name, and consequently propose Surinam as a restricted type locality, and restrict the name tucanus to the bird with the dark sanguineous red bill. In this way no nomen- clatural upsets are made. We might add that monilis P. L. S. Miiller, erythrorhynchus Gmelin and haematorhynchus Berlepsch and Hartert are clear synonyms of tucanus Linnaeus, and we definitely restrict the first two to the darker billed bird. C. culminatus Gould. Type locality restricted by Hellmayr to Rio Solimoens, Brazil. Bill with a marked concavity on the side of the maxilla below the culmen. Much smaller than tucanus cuvieri with the bill shorter than the tail. Upper Amazonia to the eastern base of the Andes, Bolivia to Colombia. Wing of adults, 190-212; bill 111-149. The only possible subspecies of culminatus is the dubious osculans Gould, which, however, seems to have no definite range of its own. It would represent this species in the Guianas, however, if the specimens in the British Museum from British Guiana and Oyapock, Cayenne are authentic. In bill structure and color it agrees with culminatus in having the maxillary groove and the pale yellow culmen in sharp contrast with the blackish sides. Notes on Pteroglossus aracari As Hellmayr has shown, typical aracari ranges from the south bank of the Amazon south to Bahia. We have an adequate series from Bahia and Rio de Janeiro. It is the bird with a paler yellow breast, and the narrow black culminal ridges. Still further south, we have three specimens of aracari from Santa Rita, Minas Geraes (George Sceva), and Sao Paulo, Fazo Cayoa and Valparaiso. These birds show a remarkable convergence in body color to castanotis australis Cassin, but are immediately separable in having the bill characters of aracari, and not those of the specifically distinct castanotis, which we also have from Valparaiso, Sao Paulo. In the latter species the bill is a deeper orange, the black culminal ridge is broadly triangular basally, and the upper mandible has an elongated band of black along the cutting edge. The new subspecies of aracari is almost identical in coloration with the black-headed variation of castanotis australis. It may be known as GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 431 Pteroglossus aracari vergens subsp. nov. Type. No. 156,885, Mus. Comp. Zool.; cf ad.; Valparaiso, Sao Paulo; June 30, 1931 ; coll. by Lima; ex Museu Paulista in exchange. Characters. Differing strikingly from typical aracari in having chin and upper throat dark chestnut brown broadly connected with the side of the head and auricular region, which are also dark chestnut brown. In this respect exactly resembling castanotis australis, in which the chestnut is, however, a lighter shade. In addition to the totally different bill characters discussed above, differing from castano- tis in color in having olive green instead of chestnut thighs. Selenidera maculirostris hellmayri subsp. nov. Type. No. 174,105, Mus. Comp. Zool; cf ad.; Rio Tapajoz, Boim; Jan. 12, 1933; A. M. Olalla. Characters. Resembling gouldi (Natterer) of Para, but in both sexes black blotch at base of upper mandible more restricted; black blotch near -tip of lower mandible surviving only as an obscure jagged line. Remarks. Hellmayr (Novit. Zool., 1910, p. 400) called attention to this character in birds from the Rio Madeira, remarking that this difference should be confirmed by additional material. The birds before us do strikingly confirm it, and we take pleasure in naming the new form after Dr. Hellmayr. 3 cf 6 9 examined. Chrysoptilus punctigula pallidior subsp. nov. Type. No. 174,228, Mus. Comp. Zool.; cf ad.; Rio Amazonas, Lago Grande; September 6, 1932; A. M. Olalla. Characters. Differing from typical punctigula of Cayenne and Suri- nam (series from Cayenne, Surinam and Obidos examined) in being generally paler above and below, upper parts lighter, more golden brownish; light bars of outer tail feathers paler; chest paler in ground color and more orange, less crimson tinged, the belly paler and purer yellow, less greenish yellow; round black spots on chest and breast usually larger, more distinct and more abundant. 1 cf 5 9 from the type locality; 7 cf 7 9 , Santarem (Carnegie Mus.). We are not aware of any report of a series of this woodpecker from the Lower Amazon, with the exception of those recorded by Snethlage, who did not have typical punctigula for comparison. The characters of this subspecies have nothing to do with guttatus (Spix). 432 bulletin: museum of comparative zoology Cerchneipicus tinnunculus angustus subsp. nov. Type. No. 171,161, Mus. Comp. Zool., d71 ad.; Caxiricatuba, right bank of Rio Tapajoz, Para, Brazil; August 12, 1932; coll. A.M. Olalla. Characters. Differing from typical tinnunculus in the narrow black barring of the upper parts and the rufous and black barring of the whole of the outer tail feathers; differing from occidentalis Hargitt of Amazonian Peru (Upper Ucayali) in that the barring on the upper parts is greatly reduced and the individual bars are always much narrower; in extreme specimens the upper half of the back is almost uniform rufous. Material examined, occidentalis: fine series in the Carnegie Museum from the Rio Solimoes; angustus: the type, and 3 cf 1 ? from San- tarem in the Carnegie Museum. Nasica longirostris australis subsp. nov. Type. No. 104,401, Mus. Comp. Zool.; cf ad.; Santarem, Brazil; Oct. 2, 1882; coll. unknown. Characters. Adults readily distinguishable from true longirostris of the north bank of the Amazon by the lighter upper parts, more fox colored, less chestnut brown; belly slightly more fulvous or tawny, less buff; in immature stages, the difference in the underparts is more intensified. Remarks. Immature specimens of Nasica are readily recognized by the paler underparts, the presence of faint barring and cross vermiculations on the belly, and by the paler and larger appearing white guttate markings on chest and abdomen, due really to the paler less blackish borders to the spots. Fully adult specimens with no vermiculations on the belly seem scarce, the majority of our series being transitional. When specimens of comparable age are compared the characters given above are readily appreciable. Immature speci- mens of australis, however, do not have the ground color of the belly tawnier than in adults of typical longirostris. An adult and an immature from the base of Mt. Duida in Venezuela agree absolutely with the series from near Obidos. An adult and an immature from Puerto Indiana, Amazonian Ecuador are nearer australis. Dendrocopus longirostris Vieillot was based on a Levaillant plate from "Brasil." There being two races in Brasil, we designate Boca-do- Igarape-Piaba. near Obidos, as the restricted type locality of typical longirostris. Nasica nasalis Lesson was based on the same Levaillant plate. GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 433 Ancistrops strigilatus cognitus subsp. nov. Type. No. 174,474, Mus. Comp. Zool.; c? ad.; Tauary, right bank of Rio Tapajoz, Para, Brazil; Oct. 16, 1932; coll. A.M. Olalla. Characters. Differing from typical strigilatus of the Rio Solomoes (17 specimens) in being more ochraceous brown and paler above, less greyish olive brown (even in worn specimens); as a result the shaft streaks are much less contrasted. Below much buffier throughout, particularly on the throat, which is less spotted with dusky; the eye stripe is buffy rather than whitish as are the ear coverts. Material examined: strigilatus: Rio Solimoes, 17 specimens (Car- negie Museum); Rio Purus, 21 specimens (Carnegie Museum), cog- nitus: the type and in the Carnegie Museum, Santarem, 1 cf ; Rio Tapajoz, Miritituba, 2 cf 1 9 ; Villa Braga 2^1 ?. Remarks. A series from the Rio Purus (21 specimens) resembles true strigilatus, but averages minutely greyer and paler above, a difference which we do not consider sufficient for formal description. Philydor erythropterus diluvialis subsp. nov. Type. No. 174,480, Mus. Comp. Zool.; cf ad.; Caxiricatuba, right bank Rio Tapajoz, Para, Brazil; Aug. 9, 1932; A. M. Ollala. Characters. Differing from typical (?) erythropterus in being slightly browner, less ashy above, most noticeable on rump and upper tail coverts ; below obviously buffier (on chest) and sides, flanks and under tail coverts more brownish, less ashy. Material examined. Typical (?) erythropterus: east Ecuador, 2; Rio Solimoes, 4 (Carnegie Mus.); Rio Purus, 10 (Carnegie Mus.). diluvialis: the type and 3 c?, Villa Braga, left bank of Rio Tapajoz (Carnegie Mus.). Remarks. Our series is obviously a different subspecies from a series from further west. There is no knowing, however, whether these birds represent true erythropterus (Bogota collections) or still another form. XlPHOLENA LAMELLIPENNIS PALLIDIOR Sllbsp. nov. Type. No. 175,166, Mus. Comp. Zool.; 9 ad. breeding; Santarem, Rio Tapajoz; July 15, 1932; A. M. Olalla. Characters. Adult males inseparable, but breeding females com- pared with breeding females of the typical form (and non-breeding with comparable birds) notably paler below, more cinereous, less sooty on throat and chest, more whitish on belly and abdomen. 434 bulletin: museum of comparative zoology We note the following measurements for the wing of males: — Para, 117-122; Rio Tapajoz, 115-117. There are no size differences in the females. A series of 7 cf 5 9 from Rio Tapajoz, Pinny, and Santarem. TODIROSTRUM LATIROSTRE SENECTUM Subsp. nov. Type. No. 175,819, Mus. Comp. Zool.; 9 ad.; Boca do Igarape- Piaba, near Obidos, Brazil; March 6, 1933; coll. A. M. Olalla. Characters. Differing from typical latirostre (Pelzeln) of Borba, Rio Madeira, southward and westward, in being slightly greyer, less bright green above, wing bars and edgings paler, less rusty buff; below much greyer on throat and chest, less olive and yellow on flanks and belly; differing from caniceps Chapman in being browner on the pileum and lacking the extension of the grey of the pileum of this form on to the hindneck and mantle; wing bars buffier, less rusty; whiter, not so grey below. Material examined, latirostre: Matto Grosso, 2 cf, series of 18 from Rio Purus and Rio Solimoes (Carnegie Mus.); caniceps: 1 cf, Curaray, east Ecuador; senectum: 2 9, type locality; 5 cf , Obidos (Carnegie Mus.); 5 c\ 2 9 , Santarem (Carnegie Mus.). Remarks. Never previously recorded from the State of Para. We have seen no topotypes from the Rio Madeira, but note that Hellmayr, who has, refers Matto Grosso and Rio Solimoes birds to the nominate form. MOLOTHRUS BONARIENSIS RIPARIUS Subsp. nOV. Type. No. 176,543, Mus. Comp. Zool; 9 ad.; Pinhy, Rio Tapajoz, Brazil; June 11, 1933; coll. A. M. Olalla. Material examined. 13 cf, 3 9, Rio Tapajoz, various localities east bank; 1 9, Lago Grande west of Rio Tapajoz. Also 8 cf, 4 9 in Carnegie Museum. This series makes it quite impossible to follow Friedmann's treat- ment (Auk, 1927, p. 500) in which he extends the range of typical bonariensis northward to the south bank of the Amazon. This dis- position of the case is also followed by Hellmayr (Orn. northeastern Brazil, 1929, p. 273), who abandons his attempt to "revive" M. bonariensis sericeus (Licht.) on the basis of size differences only. The main difficulty of both these authors seems to have been that they had very few specimens and practically no adult females. It is, of course, the females that show best the subspecific characters. Curiously enough Madame Snethlage did have a series from the GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 435 Amazon, which she referred erroneously to "atronitens Cabanis," now minimus Dalmas. The average wing length she gives for seven males, 107 mm., is just right for our own series, whereas the males of true minimus are less than 100 mm. Hellmayr extended the range of minimus Dalmas south to the Amazon on the basis of a 9 collected by Natterer at Cajutuba, just east of Para, which measured 99 mm. Other females from this region, however, range up to 104 mm., and this wing length for females of 99-104 mm. is again just right for our females from the Tapajoz. This again has nothing to do with the dwarf female of minimus, which is less than 90 mm. There are really three very distinct races, and the one from the Amazon is in no sense an intermediate. 1. bonariensis (Gmelin). Type locality, Buenos Aires. Wing of male averaging 114.5 mm. Adult females sooty brown, appreciably lighter below than above; very uniform below, chin and throat but little or not at all lighter than chest. Juvenal males not heavily streaked below. North to Ceara, Piauhy and Maranhao. Synonyms are sericeus (Lichtenstein), minor (Spix), violaceus (Wied) and murinus Pelzeln. 2. riparius Griscom and Greenway. Wing of male averaging 107 mm. Adult female quite different from bonariensis, slightly blacker, less brown, above, relatively paler below; also sootier, less brown, below, the chin and throat abruptly paler than chest, more contrasted than any other South American race, east of the Andes. Two immature males are heavily streaked blackish and buffy below, a plumage we cannot match in any other race. Lower Amazon Valley, westward to the Rio Tapajoz and Obidos. Judging by the recorded measurements of adult males, birds from the Rio Madeira and the Rio Purus do not belong here. Adult females should be examined for a final decision as to their status. 3. minimus Dalmas. Type locality, Tobago. A dwarf race, wing of adult males under 100 mm. Female, wing under 90 mm.; paler below, a warmer, less sooty brown than bonariensis, which is conse- quently intermediate in general coloration below. This race does not range south to the Amazon valley, but its southern limits remain to be determined. Notes on Tangara cayana (Linnaeus) A good series of lid71 15 9,2 imm. from Santarem, leads to some unexpected discoveries. In the first place this common tanager is practically unknown on the south bank of the Amazon. In the second 436 bulletin: museum of comparative zoology place it is almost identical with typical cayana of the Guianas, and shows no approach whatever to huberi of Marajo Island or flava of eastern and southern Brazil. In the third place there would appear to be a most unusual type of post mortem color change in this species. At first glance we supposed that our Rio Tapajoz series represented a very distinct subspecies as compared with a good series of cayana from Cayenne, British Guiana and eastern Venezuela. The Brazilian males have a brighter blue chin and throat, and this blue is continued downward as a wash over much of the silky, old gold breast and abdomen. In typical cayana this blue is duller and more ashy; there is at most a faint grayish wash on the center of the breast, but the underparts have a distinct lavender tone, lacking in our Santarem series. Much to our surprise, however, two old skins from Santarem are quite indistinguishable from the older northern birds and show none of the characteristics of the recent Santarem series. We conclude, therefore, that some post mortem change has taken place. There are also color differences in the females and immature. Our recently killed Brazilian birds are brighter colored below, yellower, less buffy and ashy than northern birds, with more of a silky gloss. But by analogy with the adult males we are suspicious that these differences might also be due to post mortem change. There proves to be a surprising amount of size variation in this species, best brought out by the table below of wing measurements. 1 cf "Cayenne" 75 2 cf British Guiana (interior) 75-78.5 2 cf Surinam (coast near Paramaribo) 66.5-67 8 cf east Venezuela 71-76 9 cf Rio Tapajoz, Brazil 68-72 It will thus be seen that Cayenne and British Guiana birds are large, Venezuela birds slightly smaller, Brazilian birds still smaller, and the Surinam bird a dwarf in sharp contrast with the large size of its geographical neighbor. In addition to this the Surinam birds are distinctly darker above, the female darker and more green than more recently collected Venezuelan specimens, eliminating any possi- bility of a post mortem change. On these characters of size and color, therefore, we propose: Tangara cayana littoralis subsp. nov. Type. No. 145,451, Mus. Comp. Zool., cf ad.; June 29, 1921; near Paramaribo, Surinam; Penard Coll., and typify true cayana (Gmelin) as the larger bird in the interior of French and British Guiana. GRISCOM AND GREENWAY: NEW NEOTROPICAL BIRDS 437 We do not favor the description of the Brazilian bird on size charac- ters alone. While it is a geographic extreme, too much of the range of cayana is occupied by intermediate specimens, and the examination of larger series from the interior of French and British Guiana might reduce the apparent large size of birds from this area. Habia rubica hesterna subsp. nov. Type. No. 176,738, Mus. Comp. Zool.; cf ad.; Pataua, right bank of the Rio Tapajoz, Para, Brazil; June 26, 1933; coll. A. M. Olalla. Characters. Differing from peruviana Taczanowski of far upper Amazonia in averaging slightly paler below, most obvious in the paler grey veiling on the abdomen and the much paler under tail coverts; throat in adult males pinker and less scarlet. Material examined, peruviana: large series (21 specimens) from the Rio Solimoes and Rio Purus; also 5 c? Villa Braga, left bank of Rio Tapajoz (all in Carnegie Museum), hesterna: Santarem, 6 cf ad., 7 cf imm., 6 9 (Carnegie Museum); Rio Tapajoz, right bank, 1 c? 5 9. Remarks. The Rio Tapajoz is very definitely the dividing line between the new form and what is presumably true peruviana. Hell- mayr cannot separate series from the Rio Madeira and westward in Amazonian Brazil from another series from northwestern Peru. The Villa Braga birds agree perfectly with those from the Rio Purus and Solimoes. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXXXI, No. 3 ANTS MOSTLY FROM THE MOUNTAINS OF CUBA By William Morton Wheeler CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM May, 1937 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the Bulletin, Vols. I to LXV, LXVI, No. 1 & 2, LXVII to LXXIX No. 1, 2, 3 & 4, and LXXX, No. 1, of the Memoirs, Vol. I to LIV No. 1, 2 & 3. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXXXI, No. 3 ANTS MOSTLY FROM THE MOUNTAINS OF CUBA By William Morton Wheeler CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM May, 1937 No. 3. — Ants Mostly From the Mountains of Cuba By William Morton Wheeler While studying the distribution of the Coleoptera in the mountains of Cuba during the summer of 1936, Dr. P. J. Darlington generously- collected for me such Formicidae as he encountered. The highlands to which for the most part he confined his explorations are the Trinidad Range of Santa Clara Province in Central Cuba and the Sierra Mae- stra, Sierra del Cobre, Gran Piedra Range, Sierra de Purial, north of Imias and the Yunque de Baracoa, in the Province of Santiago de Cuba (Oriente), at the eastern end of the island. The collection is noteworthy, first, for the number of forms (9 species, 8 subspecies, 3 varieties) which he has added to the known Cuban fauna and especially for the fine series of Macromischa, of which he collected, mainly by beating foliage, no less than 13 forms, of which 10 are new to science, and second, for his discovery of a new Cylindromyrmex. This re- markable genus and the subfamily Cerapachyinae to which it belongs were not previously known to have representatives in the Antilles. This is an outstanding achievement, considering the intense interest of collectors in the beautiful ants of the genus Macromischa and the number of species of this genus already described from Cuba. Since the list of the known species in my paper of 1931 is already antiquated, I append a revised list to the present publication. In my former account I raised Mann's subgenera Croesomyrmex and Antillaemyrmex to generic rank, but the recent discovery by Dr. Aguayo of a subspecies of C. wheeleri Mann with vestigial epinotal spines, and the discovery by Dr. M. H. Smith of a subspecies of M. isabellae Wheeler without epinotal spines, necessitates a return to Mann's original conception. The list of Macromischa sens. lat. now comprises 53 species, 12 sub- species and 14 varieties and therefore nearly quadruples the number of forms cited in Emery's "Genera Insectorum" list of 1921. Dr. Darlington's collection emphasizes the very local distribution of these ants in Cuba and suggests that the careful collector, especially in the many unexplored recesses of the mountains of Oriente, may succeed in bringing to light other members of this exquisite genus. Subfamily CERAPACHYINAE Cylindromyrmex (Hypocylindromyrmex) darlingtoni sp. nov. Worker. Length 5.6-6 mm. Head suboblong, one third longer than broad without the mandibles, as broad in front as behind, with straight, parallel sides and angularly 442 bulletin: museum of comparative zoology excised posterior borders. Eyes small, flat, nearly as long as the great- est diameter of the antennal scapes, one third as long as their distance from the posterior corners of the head and situated at its posterior two-fifths. Ocelli minute but distinct. Mandibles large and convex, with evenly rounded external borders, the apical borders broad, with about 10 subequal, low, blunt teeth. Antennal scrobes, clypeus and frontal lobes of the usual conformation. Frontal area distinct, elongate -lanceolate. Antennal scapes less than three times as long as broad; first funicular joint as broad as long, joints 2-7 more than twice as broad, 8th as broad as long, the large terminal joint as long as the 9th and 10th together. Thorax short, somewhat less than twice as long as broad, parallel-sided, semicircularly rounded behind, with rounded humeri and short neck; promesonotal and mesoepinotal sutures distinct, interrupting the sculpture but not impressed. In profile the thoracic dorsum is nearly flat, the sides of the pronotum marginate, of the meso- and epinotum submarginate, the declivity of the latter abrupt, forming nearly a right angle with the base. Petiole distinctly longer than broad, with a blunt tooth at each of its anterior corners, the node distinctly widened behind and in this region with rounded, convex sides; in profile the anterior surface is flat and perpendicular, the dorsal surface evenly convex. The anteroventral process of the petiole is large, subrectangular in profile, and laterally compressed. Postpetiole about one and two-thirds times as broad as the petiole and a third broader than long. Gaster long but not much broader than the postpetiole. Pygidium truncated posteriorly and beset with numerous spinules. Legs short, the femora and tibiae broad and flattened, the hind tarsi decidedly longer than the hind tibiae, which are furnished with a long and a short pectinated spur. Shining; mandibles finely striate and coarsely punctate apically, smooth at the base; head including the scrobes, thorax, petiole and postpetiole rather finely striate, the striae occasionally interrupted by sparse, elongate, piligerous punctures ; pleurae and first gastric segment more finely striated, but the latter only on the anterior half where it is also coarsely and sparsely punctate. Remaining gastric segments, scapes and legs smooth, with numerous fine, superficial punctures. Erect hairs yellowish, sparse, delicate, moderately long, of uneven length. The anterior border of the gula bears a conspicuous fringe or tuft of bristles, and there are some long, sparse hairs along the ventral surface of the lateral borders of the mandibles and a few long deflected hairs on the anterior, truncated surface of the clypeus. Pilosity sparse on the legs, most abundant on the tarsi. Gastric segments distinctly pubescent. wheeler: ants from cuba 443 Black; tips of scapes, funiculi, legs and pygidium brownish red, femora dark brown, terminal funicular joint and tarsi, except the basitarsi, paler and more yellow. Female. Length 7.5-8 mm. Closely resembling the worker. The head is not longer in proportion to its width but the eyes are much larger though only feebly convex, as long as their distance from the posterior corners of the head. Ocelli larger than in the worker. Scapes broader, with their anterior border more strongly excised at the base. Thorax long; pronotum, without the neck, subtrapezoidal, nearly twice as broad as long, with straight, submarginate, posteriorly diverging sides. Promesonotal suture semi- circular, distinct and impressed. Mesonotum and scutellum small and flat. Gaster longer than in the worker owing to a lengthening of the individual segments. Wings short (5 mm.) Sculpture very much like that of the worker but the striae on the posterior portions of the mesonotum and scutellum and dorsal portion of the mesopleurae feeble or absent, so that these regions are smoother and more shining. Gaster more densely punctulate than in the worker. Pilosity and color as in that caste. Wings grayish hyaline, veins dark brown, pterostigma black. Described from four workers and two females taken by Dr. Darling- ton June 30, 1936, in decayed wood on the Gran Piedra Range, Oriente, at an altitude between 2000 and 3000 ft. This species is peculiar in the structure of the mandibles in the worker and female, and in the retention of distinct promesonotal and mesoepinotal sutures by the former. Dr. A. E. Emerson has sent me from Barro Colorado Island, in the Panama Canal Zone, all three castes and pupae of Santschi's C. parallelus, described in 1932 from a single dealated female. The worker has small, flat eyes and belongs, therefore, to my subgenus Hypo- cylindromyrmex. The wings of the female and male are grayish hyaline, with distinctly infuscated tips. The pupae are naked, that is, not enclosed in cocoons! Santschi believes that my subgenus Meta- cylindromyrmex may not be valid, because the two pectinate spurs of the hind tibiae, on which it was based, are present also in striatus Mayr and brasiliensis Emery, which I assigned (1924) to Cylindro- myrmex sens. str. Examination of a long series of striatus sent me by Dr. Wolfgang von Hagen from Ecuador confirms this opinion. Unless, therefore, the workers of godmani Forel and boliviae Wheeler, known only from female specimens, are found to have some other significant character, the subgenus will have to be synonymized with Cylindro- myrmex sens. str. 444 bulletin: museum of comparative zoology I subjoin a revised dichotomic table for the indemnification of the 1 1 species of Cylindromyrmex now known. 1. Workers 2 Females 10 2. Eyes small and flat; at least the first gastric segment more or less striated (Subgen. Hypocylindromyrmex) 3 Eyes large and convex; gastric segments smooth and estriate (Subgen. Cylindromyrmex) 7 3. Head nearly twice as long as broad; apical borders of mandibles edentate. Length 8 mm. (Brazil) longiceps Ern. Andre Head not more than one and one-half times as long as broad; mandibles dentate 4 4. Tergite of first gastric segment striated only at the base 5 Tergite of first gastric segment entirely and base of second partially striated 6 5. Body brownish red, with gaster and anterior portion of head black; mandibles 6-toothed; eyes very small, with only about 15 facets. Length 4.5 mm. (Brazil) brevitarsus Santschi Body black; mandibles larger and more convex, 10-toothed; eyes larger, with more numerous facets. Length 5.6-6 mm. (Cuba) darlingtoni sp. nov. 6. Dark brown; scapes twice as long as broad. Length 5.5-6.5 mm. (Venezuela) meinerti Forel Black; scapes three times as long as broad. Length 5.5 mm. (Panama) paraUelus Santschi 7. Strise and interstrise of head, thorax and pedicel very coarse and regular 8 Strise and interstrise of head, thorax and pedicel finer and less regular. Length 6-7.3 mm. (Brazil) brasiliensis Emery 8. Legs stout; tibiae, except at their tips, whitish or pale ivory yellow 9 Legs more slender, entirely black, except the terminal tarsal joints. Length 4.5-6 mm. (Galapagos Islands) williavisi Wheeler 9. Frontal carinas not strongly developed ; eyes situated at the middle of the sides of the head. Length 6-6.5 mm. (Costa Rica) schmidti Menozzi Frontal carinas large, prolonged as strise to the occiput; eyes dis- tinctly behind the median transverse diameter of the head. Length 7 mm. (Peru, Ecuador) striatus Mayr 10. Postpetiole and gaster smooth and estriate 11 Both petiole and postpetiole entirely striate above 12 wheelek: ants from cuba 445 11. Femora yellow; fore tibiae more than twice as long as broad; wings brownish. Length 10 mm. (Bolivia) bolivice Wheeler Femora reddish black; fore tibiae scarcely twice as long as broad; wings violaceous. Length 12.5 mm. (Panama, Ecuador) godmani Forel 12. Petiole and postpetiole coarsely and evenly striate; gaster smooth and estriate throughout 13 At least the first gastric segment more or less striate 14 13. Head one and one-half times longer than broad; legs slender and black, except the terminal tarsal joints. Length 7 mm williamsi Wheeler Head shorter; legs relatively stout; tibise, except their tips, pale ivory yellow. Length 7 mm striatus Mayr 14. Gaster with striation only at the base of its first tergite; mandibles 10-toothed. Length 7.5-8 mm darlington sp. nov. Tergite of first gastric segment at least entirely striated 15 15. Mandibles 7-toothed; tergite of second gastric segment entirely striated, third segment with striae at the base; wings grayish hyaline, with inf uscated tips Length 9 mm. . parallelus Santschi. Mandibles indistinctly dentate; tergite of second gastric segment with striae only in the middle; second segment smooth; wings uniformly inf uscated. Length 9.5 mm schmidti Menozzi. Subfamily PONERINAE Thaumatomyrmex cochlearis Creighton A single worker of this rare and extraordinary ant recently taken by Dr. Marston Bates and Dr. Graham Fairchild with the Berlese funnel at Limones Saboruco, Soledad. The type specimen in Dr. Creighton's collection is from Mina Carlota, also in Santa Clara Province. Euponera (Trachymesopus) succedanea (Roger) Seven workers from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-14, 1936 and the Sierra de Purial, near Imias, July 25, 1936, and a worker and female from the Gran Piedra Range, Oriente, 2000- 3000 ft., June 30, 1936. Ponera opaciceps Mayr subsp. jamaicensis Aguayo A single worker from the Sierra del Cobre, Oriente, 3000-3800 ft., July 7, 1936, agreeing closely with Aguayo's type from Jamaica in the 446 bulletin: museum of comparative zoology Museum of Comparative Zoology. I have raised this form, described as a variety, to subspecific rank because it differs from opaciceps almost sufficiently to be regarded as a distinct species. Ponera trigona Mayr subsp. opacior Forel One dealated female from the south side of Pico Turquino. Sierra Maestra, 1500 ft., June 1936. Leptogenys puncticeps Emery Twelve workers from Los Llanos, Eastern Oriente, 1000-2000 ft., July 16-20, 1936. Odontomachus haematoda L. subsp. INSULARIS (Guerin) var. pallens Wheeler Three workers, two females and two males from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-14, 1936. Subfamily PSEUDOMYRMINAE Pseudomyrma pallida F. Smith A single small worker from the Yunque de Baracoa, Oriente, July 13, 1936. Pseudomyrma elongata F. Smith subsp. subatra Wheeler and Mann. A single worker from the coast below Pico Turquino, Sierra Maestra, June 26, 1936, agreeing well with the cotypes from Haiti. Subfamily MYRMICINAE Pheidole punctatissima Mayr subsp. jamaicensis Wheeler var. praetermissa Wheeler Two soldiers and a worker from Los Llanos, Eastern Oriente, 1000- 2000 ft., July 16-20, 1936. Previously known only from Haiti. Pheidole similigena sp. nov. Soldier Length 2.5-3 mm. Related to Ph. floridana Emery. Head large, subrectangular, with- out the mandibles distinctly longer than broad and distinctly narrower wheeler: ants from cuba 447 in front than behind, with distinct occipital inpression and rather deeply and angularly excised posterior border. Eyes small, at the anterior fifth of the sides, which are rather convex. Clypeus ecarinate, moderately convex posteriorly, its anterior border broadly and very feebly excised in the middle and narrowly sinuate on each side. Man- dibles convex but with nearly straight external borders, their apical borders with two large terminal and two small basal teeth. Frontal area distinct, subtriangular, rounded behind; frontal carinae rather strongly diverging posteriorly where they are twice as far apart as each is from the lateral border of the head, as long as the antennal scapes and forming the mesial borders of narrow, shallow, scrobe-like im- pressions. Antennae slender; scapes curved at the base, reaching only to the middle of the sides of the head; clubs longer than the remainder of the funiculi; joints 2-8 slightly longer than broad. Pronotum half as broad as the head, forming with the mesonotum a convex, hemi- spherical mass, with rounded humeri and feeble mesonotal torus; epinotum much lower, as broad as long, marginate on each side, rather concave and sloping in the middle; the spines stout, not very acute, suberect, as long as their basal diameter but shorter than their distance apart. Peduncle of petiole rather short, anterior and posterior faces of node concave, its superior border rather thick and blunt, emarginate when seen from behind. Postpetiole more than twice as broad as long and as broad as the petiolar node, its sides produced at the middle as acute angles, which are not conulate as in floridana. Gaster broadly elliptical, about as large as the head. Femora only moderately thick- ened. Shining, with sparse, piligerous punctures, which are largest on the mandibles and posterior half of the head; somewhat more than the anterior half of the head rugose, the rugae fine, sparse, and diverging on the front, coarser, more reticulate and longitudinal on the cheeks and also coarse but more oblique or concentric around the antennal insertions. Scrobe-like impressions nearly smooth, with only a few short, longitudinal rugules. Remainder of body smooth, except for the small, scattered, piligerous punctures. Pilosity yellowish, erect or suberect, moderately long and abundant ; shorter and more oblique on the legs, scapes, sides and ventral surface of head. Black or dark brown; cheeks, antennae, clypeus, neck and man- dibles, except their apical borders and tips, testaceous; legs, including the coxae, more brownish yellow, middle of femora and tips of antennal scapes brown. 448 bulletin: museum of comparative zoology Worker. Length 1.4-1.5 mm. Head subreetangular, broader in front than behind, with feebly convex sides and slightly concave posterior border. Eyes at the anterior third of the sides, shorter than their distance from the anterior corners. Mandibles with mnltidenticulate apical border and long apical tooth. Clypeus convex behind, with medially straight, trans- verse anterior border. Frontal area distinct, triangular, not impressed. Antennal scapes extending about twice their greatest diameter beyond the posterior border of the head. Thorax shaped as in the soldier, but the promesonotum is less convex; mesoepinotal impression deep, base of epinotum slightly convex in profile. Node of petiole narrow, sub- conical, without emargination. Postpetiole subglobular, as long as broad, with the sides very feebly angulate anteriorly, less than half again as broad as the petiolar node. Legs rather long. Smooth and shining, with minute, sparse, piligerous punctures, more distinct on the mandibles. Cheeks and front longitudinally rugulose. Pilosity whitish, like that of the soldier, but sparser, short and bristly on the antennae and legs, long on the dorsal surface of the body. Black or dark piceous; mandibles, antennae and remainder of legs paler, yellowish brown, except the middle portions of the femora which are dark brown. Described from three soldiers and four workers from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-11, 1936. Though obviously related to Ph. floridana, this ant differs in so many characters that I have not hesitated to describe it as a distinct species. Its status may change when some myrmecologist undertakes a revision, now long overdue, of the large and difficult fiavens group of Neotropical Pheidoles. Spelaeomyrmex urichi Wheeler A single worker taken at an altitude of 3500 ft. in the Sierra de Purial, north of Imias, in leaf mold. This minute ant was previously known only from the type specimens collected in nests of the Guacharos (Steatornis), in a cave in Trinidad. Very recently Prof. A. S. Pearse has sent me two specimens which he found on bat guano in a cave in Yucatan. Its occurrence in leaf-mold in Cuba is not surprising since other minute cave-insects are known to live occasionally in similar situations, but we should expect it to be more abundant in some of the many bat-infested caves of the island. wheeler: ants from cuba 449 SOLENOPSIS GEMINATA (Fabr.) A few workers of the typical dark form of this common ant from the south side of Pico Turquino, Sierra Maestra, 3000-3500 ft., July 7, 1936. MONOMORIUM CARBONARIUM (F. Smith) subsp. ebeninum Forel Single worker and female specimens from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-14, 1936. Crematogaster (Orthocrema) steinheili Forel A single worker from the coast below Pico Turquino, Sierra Maestra, July 25, 1936. Crematogaster (Acrocoelia) sanguinea Roger var. atavista Mann. Two workers from the Sierra de Purial, north of Imias, July 25, 1936. Crematogaster (Acrocoelia) sanguinea var. torrei Wheeler Seven workers from the coast below Pico Turquino, Sierra Maestra, July 25, 1936. Macromischa purpurata Roger Numerous workers from the following localities: Yunque de Baracoa, 1000-1800 ft., July 13, 1936; Sierra del Cobre, 3000-3800 ft., July 7, 1936 and the Sierra de Purial, near Imias, 3500 ft., July 20, 1936. Macromischa manni Wheeler var. villarensis Aguayo Four workers from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-14, 1936, agree closely with Aguayo's cotypes in the Museum of Comparative Zoology. Apart from its shape, the petiolar node of this variety is shorter and its anterior surface rises more abruptly from the peduncle than in the typical manni. Macromischa squamifera Roger Five workers from Buenos Aires, Trinidad Mts., 2500-3000 ft., June 9-14, 1936. Macromischa scabripes Mann var. brunneipes var. nov. I have not seen specimens of the typical scabripes, described from Saetia, Oriente, but two workers taken by C. T. Ramsden at Guan- tanamo, in the same province, May 7, 1914, and loaned me by the 450 bulletin: museum of comparative zoology Philadelphia Academy of Science agree closely, except in color, with Mann's figure and description. The brownish red head has distinct greenish reflections and the thoracic dorsum is metallic green passing into metallic violet on the pleurae, the epinotal spines are red, the femora, tibiae and tarsi brown instead of black. The pale pilosity seems to be longer and more abundant than in the typical scabripes. The rounded transverse rugae of the thorax in the variety here de- scribed extend as in the type to the anterior end of the pronotum, that is, to the very base of the neck. The middle and hind basitarsi, which are nearly as long as the tibiae, are distinctly compressed. Macromischa darlingtoni sp. nov. Worker. Length 4.3-4.6 mm. Closely related to scabripes Mann but decidedly larger. Head subrectangular, distinctly longer than broad, as broad through the posterior as through the anterior corners, with feebly convex sides and medially placed, convex eyes, the posterior border sinuate. Mandibles convex, with five stout, subequal teeth. Clypeus flattened, its anterior border transverse and broadly sinuate in the middle. Frontal area indistinct. Antennae stout; scapes curved basally, not reaching to the posterior corners of the head ; second funicular joint longer than broad, half as long as the first; joints 3-7 slightly broader than long; 8 as long as broad; club well-developed, 3-jointed, the terminal as long as the two preceding joints together. Thorax shaped as in scabripes, distinctly flattened above and without promesonotal and mesoepinotal sutures; epinotal spines closely approximated basally, straight, some- what stouter but scarcely longer than in scabripes. Petiole with similar peduncle, which is as long as the node, but the latter is quite unlike that of scabripes, being squamiform, rising perpendicularly in profile or with slightly concave anterior face, with rounded summit and long, convex posterior slope; seen from above it is transverse and subtriangular, fully three times as broad as the peduncle at its nar- rowest point; from behind higher than broad, subelliptical, its apex slightly narrowed but rounded. Postpetiole campanulate, broader than long. Gaster small, elongate-elliptical, acutely pointed posterior- ly; sting very long. Legs long, with strongly incrassated femora and tibiae; middle and hind basitarsi compressed. Shining; mandibles rugose-punctate; clypeus longitudinally, an- tennal foveae concentrically rugose; remainder of head with delicate, undulating, interrupted rugules, which on the front gradually diverge posteriorly and tend to fade out on the occiput. Neck transversely wheeler: ants from cuba 451 shagreened; pronotum with sparse, irregular, transverse rugules and coarse, scattered punctures; remainder of thoracic dorsum with a series of about 21 coarse, rounded (costate), transverse rugae, like those of scabripes, becoming longitudinal on the sides of the meso- and epinotum. Sides of petiolar node with 6-7 similar rugae, its posterior surface very smooth and shining, like the postpetiole and gaster. Posterior border of postpetiole coarsely shagreened, gastric segments with scattered, piligerous punctures. Femora tuberculate, but not as strongly as in scabripes; tibiae, basitarsi and antennal scapes finely, longitudinally striate. Hairs white, erect or suberect, bristly, rather short and uneven, covering the whole body, most abundant on the head and thorax, sparse on the gaster, scarcely shorter on the appendages. Mandibles, clypeus, frontal carinae and antennal foveae brownish yellow; teeth and basal border of mandibles black; remainder of head red, with vivid purplish and golden reflections; neck reddish yellow; sides and anterior quarter or third of pronotum metallic purple, declivity of epinotum and tips of its spines and metasternal angles yellowish red; remainder of thorax vivid metallic green. Node of petiole, postpetiole and gaster black, the extreme base of the first gastric segment and sting reddish or yellowish; sides of petiolar node metallic green; peduncle of petiole, coxae, trochanters and basal two- thirds or three-fourths of femora orange yellow; antennae and re- mainder of legs black. Described from seven specimens taken by Dr. Darlington on the coast below Pico Turquino, Sierra Maestra, June 26, 1936. This very beautiful species is readily distinguished from scabripes by its larger size, more shining head and thorax, much more vivid coloration, the absence of costate rugae on the pronotum, partially orange femora and very different petiolar node. Macromischa opalina sp. nov. Worker. Length 2.8 mm. Very similar to darlingtoni but much smaller. Head proportionally longer, slightly broader in front than behind, anteriorly with nearly straight, subparallel sides, straight posterior border and more convex and slightly more anteriorly placed eyes. Clypeus with entire, rounded anterior border. Antennal scapes proportionally shorter, not reaching the posterior corners by a distance equalling the length of the eyes; funicular joints 3-7 also proportionally shorter. Thorax, pedicel, gaster and legs much as in darlingtoni, but the petiolar peduncle is 452 bulletin: museum of comparative zoology shorter and the node narrower and when seen from behind terminating above in a more distinct point. Clypeus and head smoother, the latter bearing only feeble traces of longitudinal rugules. Pronotum also smooth, the transverse rugae on the meso- and epinotum anterior to the spines only 16 in number, those on the meso- and metapleurae less coarse, those on the epinotal declivity even finer, those on the sides of the petiolar node less pro- nounced than in darlingtoni. Pilosity and coloration much as in darlingtoni, but the former less abundant on the body. Declivity of epinotum and ventral portions of meso- and metapleurae metallic purple like the pronotum; hind femora brown with black bases and apices; remainder of legs colored as in darlingtoni. A single specimen from the coast below Pico Turquino, Sierra Maestra, June 21, 1936. Perhaps this form should be regarded as an extreme subspecies of darlingtoni, but it is so easily recognizable that it seems preferable to give it specific status. M. darlingtoni and opalina belong to the group of superb Macro- mischas comprising splendens Wheeler of the Bahamas and the Cuban scabripcs and squamifcra. The members of this splendens group, as it may be called, agree in having very thick, tuberculate femora and dorsally depressed, transversely rugose or costate thoraces. Macromischa platycnemis sp. nov. Worker. Length about 4 mm. Head subrectangular, without the mandibles only slightly longer than broad, slightly broader behind than in front, with nearly straight sides and posterior border and rounded posterior corners. Mandibles convex, with five large, blunt teeth. Clypeus flattened, with a median cannula, its anterior border emarginate in the middle and sinuate on each side. Eyes convex, at the middle of the sides of the head. Frontal area distinct, impressed. Antennae stout, their scapes reaching nearly to the posterior corners of the head; second funicular joint longer than broad, half as long as the first joint; joints 3-8 broader than long; club well-developed, 3-jointed, its terminal joint as long as the two preceding joints together. Thorax short, about two and one-half times longer than broad; in profile evenly and rather strongly convex dorsally, without promesonotal and mesoepinotal sutures, the declivity of the epinotum perpendicular, the spines long, closely approximated basally, directed backward, obliquely outward and upward, their tips very slightly upturned. Peduncle of petiole very long and slender, wheeler: ants from cuba 453 more than four times as long as the node, when seen from above, slightly enlarged at the middle where the spiracles are situated; in profile with the posterior thinner than the anterior half. Petiolar node squamiform, in profile rising perpendicularly from the peduncle, narrowly rounded at the summit and less abruptly sloping and some- what convex behind ; from behind as high as broad, semicircular above and rather abruptly contracted below at the sides ; in dorsal view more then twice as broad as long, three times as broad as the peduncle, broadly convex behind and roundly triangular in front, the sides bluntly marginate. Postpetiole campanulate, distinctly longer than broad, its sides straight and subparallel behind. Gaster small and very slender, nearly three times as long as broad, with very long sting. Femora and tibiae incrassated, but less strongly than in the two pre- ceding species; femora nontuberculate, both the tibiae and basitarsi distinctly compressed. Mandibles, antennae, head and thorax subopaque; legs and remain- der of body shining. Mandibles striate-punctate; clypeus with a few delicate longitudinal rugules, mainly on the sides; head longitudinally rugose, the rugae fine and diverging on the front, coarser and more reticulate on the occiput, sharp and concentric in the antennal foveae. Both the scapes and funiculi of the antennae microscopically reticulate. Thorax coarsely and regularly, longitudinally rugose both dorsally and laterally, the rugae rounded and vermiculate, occasionally anastomosing, irregularly reticulate on the epinotal declivity; neck transversely rugulose. Gaster very smooth, with sparse piligerous punctures; legs superficially and very finely reticulate. Hairs white, erect or suberect, moderately long, of uneven length, abundant on the head, antennae and legs, sparse on the thorax and abdomen. Head red, clypeus and mandibles more yellowish, teeth of latter black; antennae brown; neck brownish red; thorax metallic blue- green, epinotal spines reddish brown; peduncle of petiole, tips of coxae, trochanters, extreme bases of femora and sting yellow, remaining parts of legs brown, except the terminal tarsal joints which are paler and more reddish. Described from a single specimen from the coast below the Pico Turquino, Sierra Maestra, June 26, 1936. This fine species resembles some of the non-metallic species of the genus, such as allardycei Mann and affinis Mann in the conformation of the thorax, but differs from all the species of Macromischa sens, str., except purpurata Roger, in the great length and tenuity of the petiolar peduncle. 454 bulletin: museum of comparative zoology Macromischa chloana sp. nov. Worker. Length 2.5-3 mm. Related to M. bruneri Mann but clearly distinct. Head shorter, only slightly longer than broad, broader behind than in front, with more convex sides and vertex, medially placed eyes, broadly rounded posterior corners and straight or very feebly sinuate posterior border. Mandibles convex, with five rather small teeth. Clypeus convex, with median carinula, its anterior border projecting and rounded and with feeble but distinct median emargination. Frontal area large, distinct, triangular and impressed. Antennae rather stout; scapes nearly reaching the posterior border of the head; second funicular joint distinctly longer than broad; joints 3-8 slightly broader than long; the 3-jointed club well-developed, its terminal joint decidedly thicker than the two basal joints. Thorax without promesonotal and meso- epinotal sutures, broadest through the pronotum, which is evenly rounded anteriorly and laterally, with well-developed neck; sides of meso- and epinotum subparallel ; dorsal outline in profile convex at the anterior end of the pronotum and feebly convex thence to the epinotal spines which are rather closely approximated basally, as long as the nearly perpendicular declivity, straight, with moderately acute tips, direct backward, outward and somewhat upward. Petiole like that of bruneri, with similar node, as long as the peduncle and rising anteriorly at a less pronounced angle than in a finis Mann. The dorsal surface in profile is more rounded and less truncated than in either of these species ; seen from above the node is short, scarcely longer than broad and less laterally compressed. Postpetiole nodiform, subhemispherical, slightly broader than long and about one-fourth broader than the petiolar node, more convex than in affinis and bruneri. Gaster pro- portionally large, subelliptical, about one and one-half times as long as broad, its anterior border concave at the insertion of the post- petiole; sting well-developed. Femora and tibiae incrassated but distinctly less so than in bruneri and affinis; basitarsi somewhat compressed. Subopaque; occiput, petiolar node, postpetiole, gaster and legs shining. Mandibles striatopunctate; clypeus longitudinally rugulose. Head finely and evenly reticulate with superimposed longitudinal rugae, which diverge on the front but converge somewhat towards the occiput where they become distinctly reticulate like the rugae on the cheeks and gula. Pronotum irregularly and vermiculately, longi- tudinally rugose; mesonotum and base of epinotum regularly, trans- versely rugose, the rugae of the meso- and metapleurae longitudinal wheeler: ants from cuba 455 and somewhat finer; epinotal declivity smooth and shining. Petiole and postpetiole, except the dorsal surfaces of their nodes, finely and evenly reticulate, with oblique rugae as in bruneri and affinis. Gaster superficially, legs and antennal scapes more sharply reticulate. Erect hairs white, coarse, of uneven length, not abundant except on the scapes, shorter on the head and appendages than on the thorax and abdomen, absent on the pleurae. Mandibles, clypeus, a small area of the front, thorax, peduncle of petiole, ventral portions of its node and of the postpetiole, brownish red ; head black, with strong, metallic green reflections, which are also visible, though feebler, on the thorax. Antennae dark brown, except the clubs, which are black, as are also the gaster and dorsal portions of the petiolar node and postpetiole. Legs dark brown, except the tips of the coxae, the trochanters and the bases of the femora, which are brownish yellow. Described from ten specimens taken by Dr. Darlington at Buenos Aires, in the Trinidad Mts., May 9-16, 1936, at altitudes between 2500 and 3500 ft. Though closely related to bruneri and affinis this species can be readily distinguished by its less incrassated femora and tibiae, greenish head and thorax and transversely rugose meso- and epinotum. Macromischa affinis Mann The typical form of this species is represented in the Museum of Comparative Zoology by four cotypes from Felton, Oriente, Cuba. It was not taken by Dr. Darlington but he secured instead the fol- lowing three undescribed subspecies: Macromischa affinis umbratipes subsp. nov. Worker. Length nearly 3.6 mm. Rugae of the head, especially on the front and vertex, much feebler than in the typical affinis, but coarser on the thoracic dorsum and more undulating or vermiculate. Epinotal spines more divergent and dis- tinctly turned upward at their tips. Postpetiole decidedly broader in proportion to its length. Head, thorax and pedicel more brownish red, postpetiole, node of petiole and femora dark brown, almost castaneous; tips of femora, entire tibiae and basitarsi black, dark brown in the fore legs; antennae brownish black throughout. A single specimen from the coast below Pico Turquino, Sierra Maestra, June 26, 1936. 456 bulletin: museum of comparative zoology Macromischa affinis laeta subsp. nov. Worker. Length about 3.4 mm. Peduncle somewhat longer than in the typical affinis and the pre- ceding subspecies; epinotal spines more elevated; sculpture of both head and thorax finer, the thoracic rugae especially much finer and more numerous. Yellowish ferruginous; head slightly darker and more reddish; mandibles, clypeus, cheeks, neck, epinotal spines, petiolar peduncle, gaster and legs yellow; antennae, tips of femora, tips and bases of tibiae and entire basitarsi black; postpetiole and node of petiole ferruginous like the thorax. Lateral borders of first gastric segment and ill-defined bands at the posterior borders of the first, second and third segments, brown. One specimen from the Sierra de Purial, north of Imias, 3500 ft., July 25, 1936. Macromischa affinis maerens subsp. nov. Worker. Length 3.3-3.6 mm. Head distinctly shorter, with more convex sides than in the pre- ceding forms, almost semi-circularly rounded behind the eyes. Epi- notal spines long, usually horizontal, with upturned tips. Sculpture very coarse, the rugae variable, on the head dense and longitudinal, interconnected by reticulations, diverging rather strongly on the front and vertex, equally coarse on the thorax, most regular and longi- tudinal on the pleurae, transverse anteriorly on the pronotum, on the mesonotum and base of epinotum rather regularly longitudinal in some specimens, in others irregular or vermiculate at least on portions of these regions; concentric or oblique rugae on sides of petiolar node and postpetiole strong. Deep castaneous brown; antennae, head, nodes of pedicel and the gaster black; mandibles, bases of antennal scapes, peduncle of petiole, coxae, terminal tarsal joints, sting and apical segments of gaster paler, reddish brown. Described from 18 specimens taken in June, 1936 on the south side of Pico Turquino, Sierra Maestra, at altitudes between 1500 and 5000 ft. This or a very similar form was seen by Mann. He mentions (1924) a single worker of affinis which he collected in the Sierra Maestra at an altitude of 2000 to 3500 ft. as being "considerably darker than those of the type series, with the head and thorax dark reddish brown and only the bases of the femora pale." Macromischa bruneri Mann I have not seen the cotypes of this species, but the three following forms taken by Darlington are obviously subspecies of bruneri, which wheeler: ants from cuba 457 is smaller than affinis, with more rectangular head, much shorter epinotal spines and the node of the petiole longer, more compressed laterally and with sloping instead of subperpendicular anterior surface. Macromischa bruneri proxima subsp. nov. Worker. Length 2.8-3 mm. Head and thorax subopaque, finely and densely punctate, the former with a few longitudinal rugules just behind the frontal carinae and along the internal orbits, and in some specimens with a few in- terrupted rugules on the vertex; the thorax regularly longitudinally rugose on the sides, the pronotum above with finer, irregular or vermiculate rugae, those on the sides of postpetiole and petiolar node sharp and oblique. The color of the head, thorax and pedicel is paler than in the typical bruneri, being yellowish brown, the gaster, mandi- bular teeth and antennal funiculi, except the first joint, black, the legs, including the coxae, the mandibles, clypeus, antennal scapes and first funicular joint, clear yellow; tarsi and sting pale brown. Described from 12 specimens taken on the south side of Pico Tur- quino, 1500-5000 ft., June 11, 1936. Perhaps this form is merely a color variety. Mann describes the typical bruneri as being "brownish red, the gaster darker and the legs lighter than the rest." Macromischa bruneri ornatipes subsp. nov. Worker. Length 3 mm. Both the head and thorax more coarsely and more longitudinally rugose throughout than in the preceding subspecies. Head, thorax, and pedicel red, the head a little darker, gaster black with the base of the first segment reddish; legs yellow, the tips of the femora, bases and tips of tibiae and entire basitarsi dark brown ; scapes and first funicular joint yellow, remaining funicular joints black. Four specimens from the Sierra del Cobre, 3000-3800 ft., July 7, 1936. Macromischa bruneri imitatrix subsp. nov. Worker. Length 2.8-3.4 mm. Head somewhat larger and broader than in the preceding forms, with stouter antennae, the scapes shorter, not reaching the posterior border of the head by a distance equalling the diameter of their tips. Epinotal spines variable, in some specimens as long as the declivity and with slightly upturned tips, in others straight and distinctly 458 bulletin: museum of comparative zoology shorter. Head, thorax and pedicel subopaque, the head punctate, finely and somewhat reticulately longitudinally rugulose; thorax similarly sculptured but the rugae coarser and more vermiculate and in some specimens very irregular or obsolete on the mesonotum and base of epinotum. Dark brown ; head and gaster and in some specimens also the antennae, or at least their clubs, black; peduncle of petiole and legs, except the swollen portion of the femora, pale and more piceous brown than the thorax. Described from 9 specimens taken on the north side of Pico Tur- quino, Sierra Maestra, at altitudes between 4500 and 6000 ft., June 17, 1936. In color and sculpture this subspecies resembles aifinis maerens. There is a similar resemblance between bruneri ornatipes and aijinis laeta but the differences in size, shape of head, length of epinotal spines and shape of the petiolar node show that these parallel forms really belong to distinct species. Macromischa azteca Wheeler var. maya var. nov. Worker. Differing from the typical azteca in the petiolar node which is distinctly higher and shorter through its base, with its anterior face in profile rising with a distinctly larger and more rounded obtuse angle from the dorsal surface of the peduncle. The brown bands at the posterior borders of the gastric segments are usually absent or when present faint or obsolescent. Female. Exhibiting the same differences in the petiole as the worker. The brown bands of the gaster are very distinct but only half as broad as in the typical form of the species. Described from numerous specimens from four colonies which I found March 22, 1935, nesting in dead twigs of small trees at Moca, on the western slope of the Volcan de Atitlan, Guatemala, at an altitude of 3500 ft. The typical Mexican form was also taken from twigs by Dr. E. Skwarra. Tetramorium guineense (Fabr.) A single worker from the coast below Pico Turquino, Sierra Mae- stra, June 26, 1936. Rogeria scabra Weber One wTorker from Diaquiri, Oriente, June 30, 1936. wheeler: ants from cuba 459 Cryptocerus (Cyathomyrmex) varians F. Smith A single soldier from the coast below Pico Turquino, Sierra Maestra, June 26, 1936. Atta insularis Guerin A soldier and minor worker from the coast below Pico Turquino, Sierra Maestra, June 26, 1936. Trachymyrmex jamaicensis cubaensis subsp. nov. Worker. Length 4.3-4.5 mm. Differing from other forms of the species in the longer and more acute anterior and posterior occipital teeth, longer and very slender epinotal and posterior mesonotal spines. The epinotal spines are almost filiform and slightly curved forward at their tips. The color, too, is peculiar, being opaque, coal-black, with a bluish bloom, the mandibles, antennae, legs including coxae, epinotal spines and metas- ternal angles, ventral portion of petiole and posterior portion of post- petiole, dull red. Three specimens from the coast below Pico Turquino, Sierra Maestra, June 26, 1936. No form of T. jamaicensis has been found in Cuba hitherto, though several subspecies and varieties are known from Jamaica, Porto Rico, the Virgin Islands, Hispaniola, St. Vincent and the Bahamas. Subfamily FORMICINAE Camponotus (Tanaemyrmex) santosi Forel Three winged females and a worker minor from Buenos Aires, Trinidad Mts., 2500-3000 ft., May 9, 1936. Camponotus (Tanaemyrmex) santosi pazosi Santschi A major and a minor worker from the south side of Pico Turquino, Sierra Maestra, 3000-5000 ft., July 13, 1936, and a minor worker from Yunque de Baracoa, July 13, 1936. Camponotus (Tanaemyrmex) ramulorum Wheeler var. A single minor worker from the south side of Pico Turquino, 3000- 5000 ft., June 1936, differs in coloration from the known Cuban forms of this species, but cannot be given a name till the major worker has been seen. 460 bulletin: museum of comparative zoology Camponotus (Myrmobrachys) planatus Roger A single worker from the coast below Pico Turquino, Sierra Maestra, June 26, 1936. Camponotus (Myrmeurynota) gilviventris Roger In his original description of the workers of this ant Roger (1863) mentions considerable variation in color. He says that the "head and thorax are opaque black, the gaster red or black, the legs sometimes brown, the antennal scape either entirely black or half or entirely, usually pale red." Mann, who redescribed the species from specimens which he took at Mina Carlota, in the Trinidad Mts., and Pinares, in Oriente, says nothing about these variations. His specimens, repre- sented by three major workers in the Museum of Comparative Zo- ology, are described as having the dorsal integument as well as the hairs of the gaster red, "with a black blotch at middle of base of first segment and blackish margins of the other segments," and, I may add, black legs, with brownish tarsi. Dr. Darlington's material, consisting of five, unfortunately small, series of workers from as many localities, show clearly the variations described by Roger from specimens collected by Gundlach, probably from different colonies or localities. The form described by Mann should, of course, be regarded as the typical form of the species. It is represented in Dr. Darlington's material by five major workers and a single minima worker from Yunque deWacoa, 1000-1800 ft., July 13, 1936. The other specimens belong to the following varieties to which neither Roger nor Mann assigned names: Camponotus (Myrmeurynota) gilviventris var. refectus, var. nov. Worker. Differing from the typical gilviventris in having the dorsal as well as the ventral integument of the gaster black throughout, the mandibles, cheeks and anterior border of the clypeus in the largest specimens only slightly tinged with red, the legs black, the antennae dark brown or blackish, with the basal half or two-thirds of the scapes red. A major and a media worker from Buenos Aires, Trinidad Mts. (type locality), 2500-3500 ft., May 9-11, 1936, and a major and two minor workers from Los Llanos, Eastern Oriente, 1000-2000 ft., July 16-30, 1936. wheeler: ants from cuba 401 Camponotus (Myrmeurynota) gilviventris var. renormatus var. nov. Worker. Also with the gastric integument black throughout but with the mandibles, antennae and legs red, the terminal joints of the funiculi darker in the larger individuals. Four mediae from the coast below Pico Turquino, Sierra Maestra (type locality), June 26, 1936, and a media and minor from the Sierra del Cobre, Eastern Oriente, 3000-3800 ft., July 3, 1936. Camponotus (Myrmemyrnota) thysanopus sp. nov. Worker media and minor. Length 3-4 mm. Closely related to albistramineus Wheeler of Haiti but smaller. Head trapezoidal, shorter, not longer than broad, less narrowed anteriorly and with distinctly less convex cheeks and posterior border. Eyes more convex, the carinae between their posterior orbits and the corners of the head less pronounced. Antennae slender, scapes extend- ing about two-fifths their length beyond the posterior border. Thorax shaped much as in albistramineus, but lower, dorsally and laterally less convex and with distinctly more sloping epinotal declivity. Petiolar scale narrower but thicker anteroposteriorly, as thick above as below, with blunter and seen from behind less broadly rounded posterior border. Sculpture like that of albistramineus, but the gaster is shining above as well as below with its dorsal surface sharply transversely shagreened instead of reticulate and with more pronounced transverse piligerous punctures. Pilosity similar to that of albistramineus but the flattened snow- white hairs are more abundant on the dorsal surface, especially on the head and thorax, and the erect hairs are more obtuse and more flat- tened, like the appressed hairs, than in the Haitian species. Moreover, the flexor border of each femur bears a more conspicuous fringe of these longer hairs. Black, with red mandibles and antennae, like albistramineus, but with the legs also red instead of black; coxae darker and more brown- ish. Described from three media workers and a minor from the Sierra de Purias (type-locality), north of Imias, 3500 ft., July 25, 1936, and a single media from the Sierra del Cobre, 3000-3500 ft., July 3, 1936. These specimens indicate that the type of albistramineus is a media worker and not a minor, as stated in my paper of 1936. 462 bulletin: museum of comparative zoology Camponotus (Manniella) micrositus sp. nov. Worker minor. Length 3 mm. Differing from the described species of Manniella in its very small size and in color. Head subtrapezoidal, narrowed anteriorly, with convex sides and very convex vertex and posterior border. Eyes large and moderately convex, half as long as their distance from the anterior corners. Mandibles narrow, with oblique, apparently 4-toothed apical borders. Clypeus feebly and evenly convex, ecarinate, with the an- terior border projecting and broadly rounded in the middle, narrowly sinuate on each side. Frontal area transverse, indistinct; frontal carinae rather straight, strongly diverging posteriorly. Antennae slender; scapes extending about one-fourth their length beyond the median occipital convexity of the head. Thorax narrower than the head, short and high, about twice as long as broad, with distinct but not impressed promesonotal and without traces of a mesoepinotal suture, from above subtriangular, broad through the pronotum, rapidly narrowing to the posterior end of the epinotum, which is less than half as broad as the pronotum, the pleurae distinctly and evenly convex. From above the pronotum is twice as broad as long, semi- circular anteriorly, somewhat flattened above and submarginate at the sides; the thorax in profile evenly and rather strongly convex as far back as the epinotal declivity which is sloping, distinctly concave and as long as the distance from its anterior end to the promesonotal suture. Petiolar scale rather thick, as thick above as below, with flattened anterior and posterior faces and very blunt superior border, which is broadly rounded when seen from behind. Gaster about as large as the thorax. Legs moderately long, fore femora incrassated. Shining; mandibles subopaque, very finely granular. Head, thorax, petiole and appendages delicately reticulate, the meso- and meta- pleurae more coarsely and therefore appearing more opaque. Gaster finely, transversely shagreened or striolate. Erect hairs yellow, sparse, rather short, bristly and of uneven length, on the legs present only at the tips of the femora. Appressed hairs, or pubescence sparse, coarse, long on the pronotum and first gastric segment, shorter and more numerous on the clypeus and cheeks. Head, thorax, petiole and first gastric segment red, remainder of gaster black; mandibles, posterior portion of head and first gastric segment more brownish red; legs and antennae somewhat paler and more reddish yellow. Described from a single specimen taken on the coast below Pico Turquino, Sierra Maestra, July 26, 1936. wheeler: ants from cuba 463 Notwithstanding its minute size and aberrant coloration I believe that I am right in referring this ant to the subgenus Manniella. At least there is no other subgenus of Camponotus into which it would fit more naturally. Prenolepis gibberosa Roger The large, typical form of the species, with opaque gaster, as deter- mined by Aguayo. Nineteen workers from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9-14, 1936, and the south side of Pico Turquino, Sierra Maestra, 3000-5000 ft., June 1936. Prenolepis gibberosa albimaculata Santschi This small form with conspicuous white knees, and terminal anten- nal and tarsal joints, erroneously described by Mann as the typical gibberosa of Roger and redescribed by Santschi as a variety of that species, deserves, in my opinion, to rank as a subspecies. According to Mann, it is, unlike the typical gibberosa, confined to Oriente. Seven workers from the Sierra de Purias, near Imias, 3500 ft., July 25, 1936 and Yunque de Baracoa, Oriente, July 13, 1936. Nylanderia anthracina (Roger) Twelve workers from the south side of Pico Turquino, Sierra Maestra, 3000-5000 ft., June 1936, and the Sierra de Purial, north of Imias, 3500 ft., July 25, 1936. Nylanderia myops (Mann) Fourteen workers from Buenos Aires, Trinidad Mts., 2500-3500 ft., May 9, 1936, and the Sierra del Cobre, Oriente, 3000-3800 ft., July 7, 9136. A List of the Described Forms of Macromischa Sens. Lat. Subgenus Macromischa Roger affinis Mann (1920) g Cuba subsp. laeta Wheeler (1937) 8 Cuba subsp. maerens Wheeler (1937) § Cuba subsp. umbratipes Wheeler (1937) y Cuba allardycei Mann (1920) £ 9 Bahamas androsana Wheeler (1905) £ Bahamas 464 bulletin: museum of comparative zoology annedens Wheeler (1931) 9 Mexico azteca Wheeler (1931) 3 9 d" Mexico var. maya Wheeler (1937) § 9 Guatemala barbouri Aguayo (1931) 8 Cuba bruneri Mann (1924) 8 Cuba subsp. imitatrix Wheeler (1937) § Cuba subsp. ornatipes Wheeler (1937) § Cuba subsp. proximo, Wheeler (1937) § Cuba chloana Wheeler (1937) 8 Cuba creightoni Mann (1929) § 9 Isle of Pines darkngtoni Wheeler (1937) § Cuba dissimilis Aguayo (1932) § Cuba flavitarsis Mann (1920) § 9 & Guatemala foreli Aguayo (1931) (= petiolata Forel) 3 9 Mexico fuscata Mann (1920) U Guatemala isabellae Wheeler (1908) Q 9 c? Porto Rico subsp. mutica H. M. Smith (MS) 8 Porto Rico laevissima Wheeler (1911) § Mexico lucayensis Forel (1901) 8 Bahamas luciliae Mann (1935) 8 Guatemala manni Wheeler (1931) £ Cuba var. villarensis Aguayo (1931) 8 Cuba melanocephala Wheeler (1931) § 9 Cuba myersi Wheeler (1931) 8 9 Cuba opalina Wheeler (1937) 3 Cuba pastinifera Emery (1894) § 9 cf Bahamas, Cuba var. opacipes Wheeler (1905) 8 Bahamas platycnemis Wheeler (1937) 8 Cuba porphyritis Roger (1863) § Cuba var. jaumei Santschi (1931) § Cuba var. laiispina Wheeler (1931) § Cuba purpurata Roger (1863) § 9 Cuba sallei Guerin (1852) Santo Domingo subsp. haytiana Wheeler and Mann (1914) § 9 cf .Haiti subsp. opacinoda Wheeler (1931) § Haiti salvini Forel (1899) § Panama var. obscurior Forel (1899) § Panama scabripes Mann (1920) 8 Cuba var. brunneipes Wheeler (1937) § Cuba scandens Mann (1922) 8 Honduras schwarzi Mann (1920) 8 Cuba wheeler: ants from cuba 465 skwarrae Wheeler (1931) 9 9 Mexico splendens Wheeler (1905) U 9 c? Bahamas squamifera Roger (1863) 8 Cuba subditiva Wheeler (1903) S 8 Texas violacea Mann (1924) g Cuba Subgenus Croesomyrmex Mann aguayoi Wheeler (1931) g Cuba var. archeri Wheeler (1931) 5 Cuba var. bierigi Santschi (1931) U Cuba subsp. natenzoni Aguayo (1931) y Cuba fcarroi Aguayo (1931) § Cuba bermudezi Wheeler (1931) 9 Cuba var. gracilis Aguayo (1932) § Cuba var. mutabilis Wheeler (1931) U Cuba gundlachi Wheeler (1913) 8 Cuba to Roger (1863) 5 Cuba var. rufithorax Wheeler (1931) § Cuba var. tristis Wheeler (1931) 3 Cuba lugens Roger (1863) (3 9 Cuba poeyi Wheeler (1913) S Cuba subsp. rugiceps Aguayo (1932) 5 Cuba punicans Roger (1863) £ Cuba versicolor Roger (1863) § 9 Isle of Pines wheeleri Mann (1920) 8 9 Cuba subsp. petri Aguayo (1931) § Cuba Subgenus Antillaemyrmex Mann albispina Wheeler (1908) U 9 Culebra Island var. pallipes Mann (1920) § 9 Mona Island ciferrii Menozzi (1930) 8 Santo Domingo flandula Wheeler and Mann (1914) S Haiti floridana Wheeler (1931) g Florida pidchella Emery (1894) 8 9 St. Thomas terricola Mann (1920) g 9 Cuba forrei Aguayo (1931) 8 Cuba ^qo AUG 16 1937 Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXXXI, No. 4 INSECT POLYEMBRYONY AND IT? GENERAL BIOLOGICAL ASPECTS By Filippo Silvestri With Four Plates CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM August, 1937 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the Bulletin, Vols. I to LXV, LXVI, No. 1 & 2, LXVII to LXXIX, No. 1, 2, 3, 4. 5, 6 & 7, LXXX, No. 1, 2, 3, 4, 5, 6, 7 & 8, LXXXI, No. 1, 2 & 3, and LXXXII, No. I, of the Memoirs, Vol. I to LIV No. 1, 2 & 3. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXXXI, No. 4 INSECT POLYEMBRYONY AND ITS GENERAL BIOLOGICAL ASPECTS By Filippo Silvestri With Four Plates CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM August, 1937 No. 4. — Insect Polyembryony and its General Biological Aspects* By Filippo Silvestri Polyembryony means the origin of two or more individuals instead of one from a single egg. This process was first known (1879) in an earthworm, in which the egg produces two individuals instead of one; afterwards for some small marine Bryozoa (1893) whose eggs each produce numerous individuals by a sort of fission. The phenomenon attracted more attention when Marchal (1902-05) discovered it in some hymenopterous insects, which are parasites on other insects, and Silvestri discovered the most remarkable case known up to the present, that of another insect parasite, Litomastix truncatelhis. A brief historical summary of the studies on Insect Polyembryony follows. Historical summary Bugnion published in 1891 a paper on Ageniaspis fxiscicollis (Dalm.)1 treating the postembryonic development, the anatomy and the habits ; in the first part of the paper he also called attention to the novel fact that the embryos were placed "dans un tube flexueux qui flotte dans la lymphe de la chenille (of Hyponomeuta) a cote de l'intestin." He did not suspect the origin of all the embryos of a chain from one egg, and considered the entire tube (containing the embryos) as an eggs, derivative : the epithelium existing around the tube as a formation of the embryos, serosa and amnios, and the substance in which the embryos are immersed (called, later, trophamnios by Marchal) as originating from the eggs, material comprised between the embryos and the amnios. He could not be precise as to the time of the ovide- position in the host, and made a supposition which proved erroneous, but honestly at the end of his paper wrote: "malgre ses imperfections j'espere que mes recherches en susciterons bientot de nouvelles et de plus completes sur l'organisation et les moeurs si dignes d'interet des Hymenopteres parasites." The hope expressed by Bugnion was realized by Marchal, who studied the same species of parasite and some others and with diligent researches (1898-1904) cleared up various points left uncertain by his predecessor and, above all, discovered polyembryony in Ageniaspis fuscicollis and in Platyg aster zosine2 Walker. ♦Address delivered as part of the Harvard Tercentenary Conferences. 'Syn. Encyrlus fuscicollis Dalm. 2Syn. Polygnotus minuius Marchal. 470 bulletin: museum of comparative zoology The principal and new results of Marchal's researches were the following: 1) the egg of Ageniaspis fuscicollis is laid in the egg of Hyponomeuta malineUus Z., but continues its development into the larva. 2) in the developing egg (where he observed a first stage with 5 nuclei: 4 small and 1 larger and of very different appearance) there are distinct embryonal nuclei and the so called "paranucleus"; from the former (together with the respective cytoplasm) originate all the embryos of the chain, from the second the nucleus of the trophamnios. He described the further development, the paranucleus (tropham- nios), the adventitious cyst, and made interesting considerations on sex determination and, generally, on the polyembryony discovered by him in the Hymenoptera and in other animals; but, as Bugnion before him, did not recognize the true nucleus in the oocyte, and considered as the nucleus what is really the nucleolus (or oosome). He did not study the maturation and the fecundation of the egg, and therefore was not in a position to define the different origin of the embryonal cells and the paranucleus. The gaps left by Marchal were filled a short time later (1905) by Silvestri with a first study on Litomastix truncatellus, in which he established the exact structure of the complete oocyte, the formation of the polar bodies, the fecundation, the first stages of cleavage, the localization of the sex primordium. Moreover he followed the entire polyembryonal development, which resulted somewhat differently from that of Ageniaspis, concluding with the formation of about a thousand embryos, which for the most part transform into normal larvae, destined to give adults, and in very small part to originate sterile larvae, called asexual. In 1907 Silvestri published his researches on the polyembryonal development of Ageniaspis fuscicollis and a subspecies of the same (A. fuscicollis prasincolaSilv.), and confirmed what he had discovered in Litomastix regarding the egg structure, the maturation, the fecunda- tion and the origin of trophamnions. He investigated the monem- bryonal development of Chalcididae, also parasites of insects, contrast- ing the differences with polyembryonal forms. Later (1908) the first stages of development of another Chalcid (Prospaltella), with con- tinous parthenogenesis, demonstrated that in it the egg has no oosome (nucleolus) and gives a single polar body only, whereas the other monembryonal parasites studied have normal egg structure. Martin (1914) took up the study of Ageniaspis fuscicollis, and added a few particulars on the development of the oocyte in the ovarium, but otherwise confirmed what Marchal and Silvestri had made known. SILVESTRi: INSECT POLYEMBRYONY 471 Hegner (1914) studied the origin of the sexual cells determinant in Copidosoma gelcchiae, and described the development of the oocyte in the ovarium, concluding that it was derived from the fusion of two eggs, and that the oosome (nucleolus) was the nucleus of one of the two eggs. This mistake was immediately (1914) corrected by Silvestri in a note on Copidosoma buyssoni Mayr, and later by Hegner himself. Silvestri, continuing the study of the biology and development of parasitic Hymenoptera (1915-1916), extended his researches to other species of Chalcididae and a species of the Proctotrypidae also of mon- embryonal development (Platygasler dryomyiae Silv.), and found some new particulars, but confirmed the results of the preceding researches on the egg structure, polar bodies, presence of oosome (nucleolus) in the bisexual Chalcididae and its absence in the unisexual Chalcididae and Proctotrypidae. Patterson (1915) published a first paper on the polyembryonal Hymenoptera, treating Copidosoma gelcchiae, but he studied the ad- vanced stages only, which are similar to those of Litomastix, adding some notes and figures of abortive embryos. The same author under- took afterward a series of researches on Litomastix floridanus1 Ashm., which began in 1917 and were concluded in 1921. In collaboration with Porter he studied the spermatogenesis for the first time, but for the egg structure, maturation, fecundation, segmentation, forma- tion of asexual larvae, he confirmed completely the discoveries of Silvestri, merely giving a slightly different interpretation of the nature of the asexual larvae, and admitting the origin of males from fecun- dated eggs and of females from virgin eggs. In North America, Leiby also undertook the study of the develop- ment of Copidosoma gelcchiae, with results similar to those known for Ageniaspis, Litomastix, and Copidosoma buyssoni, but he should have been able to ascertain the reduction of the number of chromosomes during the maturation from 16 to 8. The same author in collaboration with Hill studied the development of Platygaster hiemalis Forbes (1923) and of Platygastcr zosine Walker2, finding in the former an egg struc- ture, maturation, and fecundation similar to that described by Silvestri in Platygastcr dryomyiae, and in P. hiemalis a monembryonal develop- ment, erroneously considered as in part twinning. In 1931 polyembryony was discovered in Macrocentrus gifuensis Ashm. of the Braconidae, a parasite of Pyrausta nubilabis Hiibner, by Parker. He was not able to work out the maturation and the first >Syn. Paracopidosomopsis floridanus. 2Syn. Plalygaster vernalis. 472 bulletin: museum of comparative zoology stages of cleavage, but ascertained that the parasite lays the egg in the first larval stage and described a polyembryonic process somewhat similar to that of Litomastix. Daniel (1932) published a paper on polyembryonal development in Macrocentrus ancylivorus Rohwer, but stated that "only one parasite larva matures from each host larva." Sketch of the development of Litomastix truncatellus. There is no point in giving here all the particulars of development of the polyembryonic species. I advise those particularly interested to consult the papers of Marchal, Silvestri, Patterson, Leiby and Parker.1 I shall therefore present a brief sketch of the development of Litomastix truncatellus before discussing the general biological aspects of insect polyembryony. This wonderful creature Litomastix (Figs. 2-3) is a small black insect about 1mm. in length including the wings. It lives in the adult stage only a few days, taking for nourishment some water and a small quantity of sweet substance, secretions of plants or animals such as plant lice. The male insect in adult life has the sole function of inseminating the female, but the latter, whether fertilized or not, is capable of depositing eggs, which in either case are able to develop. If the eggs are inseminated females only are produced, as in the honey bee; if the eggs are not inseminated the offspring are all males. The instinct of the female from the moment she emerges from the skin of the host leads her to lay eggs; therefore she is seen crawling around on the leaves and branches of plants on which these animals spend their life with the head low and the antennae extended forward and touching the surface with their tips, beating them like microscopic hammers in order to find what they want for depositing their eggs; that is, the eggs of another insect, a moth of the genus Plusia, which lives on the leaves of the cabbage and some other herbaceous plants. As soon as a female in her persistent search has touched an egg of the moth, she rapidly turns a little about it, mounts on it, stops, pushes down the ovipositor, a short needle-shaped abdominal appen- dage, thrusts it through the shell of the egg, and lays one egg in the egg of the moth. She turns a little on the moth egg, and continues her search for other eggs until all her eggs — about 100 — are deposited. This continues with some intervals during several days. Generally •A book on the subject of insect polyembryony is in preparation, and will be published as soon as researches on polyembryonal Macrocentrus are concluded. SILVESTRi: INSECT POLYEMBRYONY 473 one or two eggs of Liiomastix are laid in one egg of the moth. Both the host egg and the parasite egg continue to develop. There are species of parasites, such as the Trichogramma of the European corn borer, which eat the material of the host eggs, consuming all of it, and transform themselves under the shell of the egg, from which they emerge by making a small hole through it, but the egg of Liiomastix develops gradually first in the egg (Fig. 8), later in the body of the caterpillar, and concludes its development only in the last stage of the caterpillar's life. The extraordinary process of polyembryony consists in the case of Liiomastix in the fact that the egg in the first stage of its develop- ment (cleavage) does not, like most eggs, form a relatively small number of cells which remain together in groups and gradually differentiate in parts, organizing one individual; instead it multiplies in enormous numbers and forms an elongated asymmetric cellular body which is constricted by an external membrane derived also from a part of the egg, and by a sort of strangulation (Figs. 20-21) is separ- ated into pieces numbering about a thousand. All these small pieces, or rather groups, are formed in the body of the caterpillar, and attach themselves to the walls of the various internal organs of the cater- pillar and there absorb their nourishment, which is the lymphatic material of the host. Each group grows into an embryo which at last transforms into a small wormlike larva about }/2 mm. long when completely developed (Fig. 23). By this process, which has been very briefly recapitulated, more than 100 individual larvae of Liiomas- tix develop from one egg within first the host egg and later the moth Plusia. They remain under the skin of what once was a caterpillar, and transform into pupae, deforming the appearance of the body. Through the skin of the carcass (Figs. 5-7) the small oval puparia are visible. Each contains a pupa which will transform into an adult ; this will come out well winged to begin another generation. From each egg of Liiomastix originate about 1000 adults, but since two or more eggs are often deposited, more than 3000 adults have sometimes been counted emerging from one host. In connection with this extraordinary process of development, Silvestri discovered very peculiar phases in the development of the egg cell, more particularly also the forma- tion of a number of asexual larvae, which live a few days, lacking sexual cells and sometimes other organs. The life cycles of Liiomastix from egg to egg in the summer of South Europe last about 30 to 40 days. Here in America there is another Liiomastix (L. floridanus) whose life historv is similar. 474 bulletin: museum of comparative zoology The more important scientific results of the researches on Litomasiix iruncatellus are: 1. The egg in the primary oocyte stage (Fig. 9) has a very thin chorion with a small micropyle at the anterior (cephalic) end, the ooplasm very poor in deutoplasm, the nucleus very small and placed in the anterior part of the egg, and an oosome (or nucleolus or sexual cell determinant) bigger than the nucleus and placed in the posterior half of the egg. 2. The egg maturation (Figs. 10-12) is typical, and is the same for both fecundated and parthenogenetic eggs, and the polar bodies remain in the ooplasm (Figs. 13-17) forming the paranucleus. 3. The egg cytoplasm divides into two parts, one of which con- taining the paranucleus will form an involucrum for the other part, which is destined to give rise to the embryonal cells (Figs. 19-22). 4. The oosome or nucleolus (Fig. 18)) is distributed to a few embryonal cells destined to become sexual cells. 5. During the constriction and separation of the groups of embry- onal cells, the majority contain sexual cells and will give normal sexual larvae (Fig. 23); a small number will lack sexual cells and will give very different asexual larvae (Figs. 24-25). Having concluded our brief summary of the study of insect poly- embryony, it now remains to consider its contribution to points of general biological interest: 1. egg organization; 2. egg maturation and fate of polar bodies; 3. egg fecundation; 4. egg cleavage; 5. origin of two kinds of larvae from the same egg; 6. causes of poly embryonic development; 7. sex determination; 8. sexual cells determination. 1. Egg organization We know from the studies of egg structure and development (published especially from 1900 on) of numerous species of Metazoa belonging to different classes, that the eggs, in conformity with their behavior in giving rise to new individuals, can be divided into two groups; those having a mosaic structure, viz. localization of future parts of the embryo, and eggs capable of regulation, totipotent. (Between the two extremes, naturally, there are intermediate types.) Eggs manifest their potentiality, generally, after fecundation; but those of some species before insemination and maturation, when the oocyte of the first order is just completed, show a different structure of the cytoplasm in the various distributions of place and of quantity of SILVESTRi: INSECT POLYEMBRYONY 475 the protolecithe and of the deutoplasm, or a different coloration of ooplasme zones, either a natural coloration, or an artificial one obtained by means of vital staining after the Vogt method. The egg of polyembryonic Hymenoptera, when the oocyte of the 1st order is complete, is, as we have seen, characterized in all the species by its smallness (about 20 to 150 jj.), with a cytoplasm lacking or almost lacking deutolecithe, and having, as an homolecitic egg type, the small yolk and fat granules scattered through it, nearly homo- geneously except along the sides. Moreover the egg of polyembryonic Chalcididae and Braconidae is provided with an oosome as well as a nucleus, in the posterior half, which is the sex primordium; therefore this kind of egg shows, besides the usual polarity and symmetry com- mon to the insect's egg in its form, micropyle and nuclear1 position, a determinate sexual region, which will remain enclosed by the first sexual cell, which is one of the four in the 2d stage of segmentation. This segregation of the sex primordium early in the egg, was also demonstrated by me for the eggs of Chalcid and Braconid parasites with monembryonal development, and by other authors particularly for the egg of Diptera (Kahle, 1908; Hasper 1911), of Coleoptera (Hegner 1909) and in Crustacea (Haecker 1897, 1903; Amma 1911). In Sagitia (Elpatievsky 1909, Biichner 1910) this type of segregation always begins to be visible during the development of the 1st order oocyte. I could not detect any sign of it in a Chalcid (Prospaltella berlesei How.), which is a continuously parthenogenetic (thelytocous) insect. The eggs of Proctoirypidae Platigasterinae either poly-or mono- embryonic, which are smaller (15-20^ in length and 8/x in width) and destitute of deutolecithe, do not reveal any oosome; therefore they have an ooplasme completely (at least apparently) equipotent. 2. Egg maturation and fate of polar bodies Inseminated egg. The study of the chromosome equipment of the polyembryonic species of the Chalcididae, Braconidae and Proc- totrypidae has not yet been adequately made, but the egg maturation phases have been followed in all the species and can be defined as being similar (except the chromosome number) to that of other insects, and like typical Metazoa; therefore the result of maturation for these •It is understood that the bilateral symmetry of the egg is not perfect, because the principal axis of symmetry does not pass exactly through the middle of the nucleus, but in the 1st order, the oocyte is very near to that position, as is also the case with the oosoma. 476 bulletin: museum of comparative zoology particular Hymenoptera is also to reduce the diploid number of nucleus chromosomes of the 1st order oocyte to haploid. Parthenogenetic egg. The unfertilized eggs of the same species of polyembryonic Hymenoptera form two polar bodies, as in the fer- tilized; therefore at the end of maturation they have a haploid pro- nucleus. The great difference between polyembryonal and most1 of the monoembryonal species lies in the fate of the polar bodies, and also in the separation of that part of the ooplasm containing them, from the part comprising the 1st nucleus of segmentation and the oosome. Until 1905, the year in which I published the discovery of the fate of polar bodies in Litomastix truncatellus, it was known that the polar bodies with their set of chromosomes, and a very small quantity of cytoplasm, were extruded from the egg; they were justly considered as abortive eggs because of their lack of cytoplasm and other parts peculiar to eggs. In some species of Metazoa (Limax, Kofoid 1895, and Meisen- heimer 1896; Asterias, Andrews 1898 and 1899), the polar bodies ex- truded from the egg were observed to remain visible longer, to grow larger, and, in the case of Asterias to form pseudopods similar to that of blastomeres, but afterwards to degenerate without showing any functional activity connected with the embryonal development. Polar bodies had been observed in some Crustacea (Hacker 1895) to remain in the ooplasm, as with many insects (Henking 1888, 1890, 1892) ; in the eggs of some species they had been noted to make hernia on the surface, and in eggs of other species to unite with the chro- mosomes in a small chromatinic group. But they were always destined to degenerate and to vanish without taking any part in the develop- ment of such an insect. Petrunkewitch alone (1902), after having fol- lowed the exact formation of the polar bodies in the parthenogenetic eggs of bees, had noticed that the 2d polar body copulated with the inner part of the divided 1st polar body, giving rise to the male germ cells. But it was demonstrated that owing to his incomplete series of stages he was wrong, since the polar bodies of the bee also degenerate. In Litomastix, as in the other polyembryonic Chalcididae, we have, at least up to the present, the unique fact that the polar bodies remain- ing in the ooplasm unite with the chromatin and form a triploid polar nucleus, which will divide repeatedly mitotically, or the 3 nuclei of the polar bodies can form one or two groups (Ageniaspis), or all 'I write "most" and not all, because in the monembryonal Platygaster dryomyiae the fate of the polar bodies is the same as in the polyembryonal species. SILVESTRi: INSECT POLYEMBRYONY 477 three can remain separated and produce nuclei, which pass first to the quiescent stage and afterward grow larger, dilate, branch out and can divide directly, like the polar ooplasm (called now trophamnios) in which they are imbedded. The fusion of the polar bodies was a fact already known in many species of insects. An attempt to demonstrate their behavior up to the 3d division of segmentation was described by me (1915) for Encarsia parthenopea (Chalcididae), but in this case also the ooplasm takes no visible part in the division of the polar bodies, which degenerate. In the polyembryonic species of Hymenoptera, with the reconstitu- tion of a polar nucleus by fusion of the two nuclei of the 1st polar body with that of the 2d polar body, there is, in the cytoplasm, an incipient activity connected with that of the nuclei. The whole egg cytoplasm becomes divided into two fields: one as the result of the energetic action of the polar nucleus (or nuclei), and one as the result of the action of the 1st segmentation nucleus. Owing to the continuing and increasing nuclear activity of these two cytoplasmic fields during the first segmentation division, the ooplasm divides into 2 parts. Considering this fact in its general biological aspect, it demonstrates that the egg of these Hymenoptera can lose a third or more of its cytoplasm, and will maintain unaltered the potentiality of develop- ment, a fact ascertained experimentally in or for the merogonic de- velopment of the eggs of Nemertina (Wilson 1903) and Echinidi (Hertwig O. et R, 1887). In these the egg was divided by one or another method into two or more pieces, one nucleate and the others enucleate. Normal development of the nucleate piece ensued as well as the enucleate piece, if a spermatozoon was introduced into it, there- by supphing a nucleus. The egg of these polyembryonic Hymenoptera, which we have shown to be supplied with polarity, symmetry, and sexual localization, is otherwise isotropous for the organo-formative substances, because it can lose a great part of its cytoplasm without preventing later embryonic development. Perhaps, during the formation of polar bodies, there is a differentiating exchange of qualitative substances between the polar and the embryonic ooplasm. From a general point of view, the separation of a notable part of the ooplasm together with the polar bodies to form a protective involucrum in Litomastix1, or both protective and assimilative in Copidosoma, Ageniaspis, Macrocentrus, Platygaster, demonstrates a morphogenetic 'The polar involocrum (trophamnios) can have a metabolic function during the first days of development in Litomaslix also, but afterward it seems quite certain, has only a protective one. 478 bulletin: museum of comparative zoology and functional plasticity which is innate in the organism in all its stages, viz: the polar bodies retain their primitive function in these species of polyembryonie Hymenoptera. Later, in the adaptive per- fecting of such species as internal parasites, the polar bodies are diverted to another, new, function, together with a part of the ooplasm. Thus in Litomastix, the 3 haploid polar nuclei unite to form a triploid nucleus, which is capable of regular mitotic division. Meanwhile in the egg of Copidosoma, Ageniaspis, M acrocentrics, and Platygastcr, the nuclei derived from the polar bodies are able to increase, forming polymorphic nuclei capable of various unequal direct divisions. In both cases the polar ooplasm remains undivided, but in the first, after the 6th division of segementation, it becomes a laminar sheet; in the second it becomes an involucrum with a notable thickness in com- parison to the encircled material. 3. Egg fertilization The known polyembryonal Hymenopterous species are all arrhenotic and have a fertilizing process similar to that of other Metazoa. The fecundation of these polyembryonal insects confirms in a general way the theory that fecundation has the essential function of carrying half of the chromosomes to the egg, which has extruded the same number in the polar bodies, to reestablish in this manner the diploidism in the first nucleus of segmentation and to bring in this manner genes of another individual, to favor the amphimixis. In these Hymenoptera fecundation also has the function of determining the female sex. Other theories of the function of fecundation, such as that urging its necessity for transforming the egg from a monocentric to a dicen- tric system (Dalcq 1928), have no absolute general value, because in these Hymenopterous species, bees, and all other animals with a facultative parthenogenesis and an arrhenotic parthenogenesis, the unfertilized egg, which has also formed and separated the 2d polar body and has (at least until proof to the contrary) a haploid nucleus, is able to develop in exactly the same manner as the inseminated egg, forming a normal spindle after the first division of segmentation. 4. Cleavage of the egg The cleavage (or segmentation) of the egg of the polyembryonie species of Hymenoptera, as we have seen, takes place regardless of whether the first nucleus of segmentation be derived directly from the SILVESTRi: INSECT POLYEMBRYONY 479 female pronucleus alone, or from the conjugation of the female and male pronuclei. Cleavage is always total, equal (or adequal) and synchronous also in the somatic blastomeres, but somewhat hetero- chronous, in comparison with the somatic (in the Chalcididae and Braconidae) in the sexual blastomeres which include the oosoma. It is completely synchronous in the Proctotrypidae, in which such a determinant sex-germ has not yet been distinguished. This kind of cleavage is found neither in other insects nor in Hymenoptera of the same families, therefore this type of cleavage may be considered an adaptation of these eggs to polyembryony and not to the parasitic life or to the small size of the egg, because we know of many Hymen- optera, living as internal parasites, with a very small egg which can be laid in the egg or larva or pupa of the host insect, but which does not present a cleavage extending to the cytoplasm; the cleavage is limited to the nuclei and leaves the cytoplasm undivided. We have seen in the polyembryonal Chalcididae that all the oosome remains in one of the blastomeres of the 1st division of cleavage (Ageniaspis) or in one of the blastomeres of the 2d division. As it is now certain from the studies made of its destiny in monembryonal species that the oosome is the visible determinant of the sexual cells, it is noteworthy that the differentiation of these cells from the somatic takes place from the 1st or 2d division of the egg cleavage. In the monembryonal species, provided also with an oosome, such a differen- tiation takes place much later: at the stage of blastoderma or a little forward. I stress this precocious differentiation of the two kinds of blasto- meres, somatic and sexual, because I shall revert to it again in con- sidering the subject of the asexual larvae. In certain species of Platygaster, where the egg is always destitute of oosome during cleavage or blastoderm formation, it has been possible to distinguish the cells destined to form the reproductive organs from those of the soma. 5. Origin of two kinds of larvae (sexual and asexual) from the same egg The precocious distinction of somatic and sexual cells in the develop- ment of monembryonal species has an organogenetic individual value, but in the polyembryonal species it takes another larger aspect; it offers a solution of a biological problem of great interest. (1) if, during the cleavage of the egg, some blastomeres, certainly destined to form 480 bulletin: museum of comparative zoology the soma, are separated from the sexual, are they able to give rise to a complete embryo and by means of regulation to form sexual cells also, forming in this manner a complete organism? (2) Can the somatic cells, already differentiated as such since the first cleavage divisions, although lacking potentiality for the formation of sexual cells, origi- nate sterile individuals similar to or somewhat different in form from normal individuals and capable of short life at least? In the numerous experiments of blastotomy, performed by various authors from Chabry (1887) to Rund (1925), no animal species with a precocious differentiation of oosome was ever selected. Therefore, although very interesting in relation to other problems, these experi- ments do not present answers to the questions enunciated above; on the contrary, in the development of Litomastix, we have a case which is worthy of careful consideration, although we have no absolute demonstration of my reasonable interpretation. During the cleavage of the egg, as we have seen, the sexual blasto- meres have a delayed multiplication compared to that of the somatic; therefore these prevail numerically over the others. After the great primitive morula is formed from well defined cells of two kinds, protected by the polar involucrum, there begins to be manifest a first constriction of the primitive embryonal mass in such a manner that this becomes elongated, strained first in one part, later along various points of the longitudinal axis, and later still in a trans- verse direction also, until it forms an embryonal mass, which is divided by the mechanical action of the involucrum against the centrifugal pressure of the embryonal cells in active favorable multiplication, into small secondary groups, constituting the polygerminal complex, which can show in a median longitudinal section more than 60 units. During these haphazard constrictions and divisions, the primitive embryonal mass is divided into groups of embryonal (somatic and sexual) cells, which can divide again haphazardly into other groups. These at last will form isolated embryos, which (in Litomastix) will total a thousand (and more, asexual larvae comprised) for each egg. In such separation of groups there is no regularity in the disposition of the somatic embryonal cells together with the sexual, or regularity in the separation of groups, which is haphazard in itself. Therefore there will be groups including somatic and sexual cells, and groups composed of somatic cells only, because they are the more numerous. The groups embracing both kinds of embryonal cells give rise to embryos and normal larvae, destined to transform into adults. The groups composed of somatic cells only do not disintegrate; on the contrary they develop SILVESTRi: INSECT POLYEMBRYONY 481 precocious individual larvae, having form, structure and dimensions different from those of the normal larvae, to which I have given the name of asexual larvae. These have been described in preceding pages, but here they must be considered from a general biological viewpoint which will be corollary to the above exposition, always supposing that my interpretation is in accord with the facts or at least approximately so. Litomastix with this manner of development, giving origin to two kinds of different larvae, proves the possibility that groups of somatic cells, precociously separated at the blastomere stage from the sexual cells, have the potentiality of forming an individual, complete for some organic systems and incomplete for others, though having a general body shape which gives it the appearance of a perfect organism at a determined stage and for a determined order of insects. No entomologist, acquainted with the larvae of parasitic Hymenop- tera, can equivocate in ascribing the asexual larva of Litomastix to such an order of insects; but nobod}' who had followed only the develop- ment of the typical larva of that species and had observed it at the last stage, could have been able to attribute the asexual larva to the same species of insect, without having the experimental proof which I was obliged to obtain (1905) and which Patterson repeated (1921). I maintain that this fact is a very interesting one for the biologist, and that it must be kept in mind in other researches in the development of Hymenoptera and experiments of blastotomy in other Metazoa, because it demonstrates that in the somatic cells (always, at least, if precociously differentiated as such) not only formative substances, the sexual-cells' determinant, are absent, but with the same substances has some organizing potentiality for one or more other systems also (in the case of Litomastix malpighian tubes, respiratory and circula- tory systems). The experimental work on insects by a number of researchers (Oudemans 1899, Kellogg 1904, Meisenheimer 1909, Regen 1909, Kopec 1911-13), who employed castration and the grafting of gonads of one sex into the body of individuals of the opposite sex, and above all the gynandromorphism offered by the comparable experiments of Nature, render the conclusion inevitable that in insects the body with its various parts, sexual cells excepted, is independent of the gonads and furthermore that by these neither hormones or other substances are elaborated so as to be (Steche 1912; Geyer 1913) capable of direct action on the somatic parts of the same insect. It is sufficient to to remember that in the same perfect gynandromorphous individual, 482 bulletin: museum of comparative zoology where half the body has male gonads and the opposite half female gonads, each half of the body has all the characteristics of the male and of the female without any interference between the male gonad and the female soma and vice versa, the diametrical opposite to what is proved for the vertebrates. In the case of the asexual larva, Nature has carried on an advanced experiment and has separated groups of somatic cells from the primi- tive morula or from the secondary morulas. Such groups have demon- strated the potentiality of independent development from the sexual cells and their capacity to form asexual larvae, which are viable for some days at least, though not complete in all the organic systems. This proves that the somatic cells in insects really possess an inde- pendence of development from the sexual cells, but that they cannot produce an organism perfect in all its parts. Having shown that in many Chalcididae, Braconidae and Proc- totrypidae the larva has various stages, generally 4 to 5, of which the earlier may be very different from the later, it would be possible to maintain that the form of asexual larvae is an atavism, namely, that it represents the larva of ancestors of the genus Litomastix, in which the normal larva had a first stage which was free and similar to the asexual larva. At a later time in the evolution of the species, with a more special adaptation as an internal parasite, with rapid develop- ment in relation to the abundance of nourishment offered by the host, the asexual larva should have been suppressed in the actual normal development, though remaining in a latent stage in some genes which are dominated by the normal rhythm of the development of the whole organism, when complete with genital organs also. In some embryos, with the sexual cells absent, the somatic characteristics, latent in the genes of the somatic cells, could become free to develop and produce an ancestral larva of the first stage, but would be incapable of trans- forming into a larva of the following instars due to the lack of the sexual cells and of other indicated systems. Therefore it should cease developing and die after a few days (about 6 to 10 in Litomastix) . The function of larvae of the 1st stage, which have a body slenderer than that of the typical larva of Hymenoptera in the last stage, more agile and better adapted to diffuse itself through the host tissues, having also well-developed labial glands, can be considered of some usefulness to Hymenopterous parasites in that mechanical action is increased by the strong mandibles and the chemicals of the labial glands, which discharge the secretion into the haemolymph of the host. SILVESTRi: INSECT POLYEMBRYONY 483 Therefore the reappearance of such larvae of Litomastix truncaiellus can be considered useful also to the species. If this should not really be so, especially for Litomastix, it might be possible to suppose that the asexual larvae represent the larva of the first instar, which, in antiquo, would normally be present in the ancestor of Litomastix. The formation of asexual larvae induces one to think of the free- martin of Ruminantia. Lillie (1916 etc.) published an important revi- sion of numerous cases and demonstrated that the sterility of the free-martin is caused by the hormones developed by the genital organ of the male twin, which, entering the circulation of the other individual, inhibit the normal development of the sexual organ of the female; but, though this may be a good explanation of numerous cases, other cases of monozygotic twins could be referred to a very unequal separation of sexual cells between the twins or a complete defect in one of the two, similar to what was sustained by Hart (1910) and with some variations by Cole (1916). 6. Causes of polyembryonal development The polyembryonic Hymenopterous species, as far as is known up tg the present time, can produce from one egg a few individuals in some species, in others 15, in others 70, in others some hundreds, and in others a thousand and more. What are the causes of this phenomenon? We remember that such species are all parasites of other insects and that, the species of Macrocentrus excepted, the eggs are laid in the eggs of the host, where the development begins and continues in the larva of the host or (in some species) in the chrysalis also. Referring to the known experiences of Loeb (1894) with the eggs of Echinidi, and especially to the work of Bataillon (1900-1901) with the development of eggs (2 hours after fecundation) of Petromyzon planeri, in which he got the dissociation of the blastomeres and their develop- ment to the point of producing complete double embryos, Marchal (1904) concluded that variation of osmotic pressure was the cause of insect polyembryony. Now if we examine particularly the case of Litomastix, we must keep in mind that the egg is surrounded by the almost impermeable chorion and develops at the expense of its own nutritive material until the 6th division of segmentation, and its ooplasm is not in immediate contact with the content of the egg of Plusia. After this stage, which occurs about 7 hours after laying, it loses the chorion and remains free and in direct contact with the tissues and the nutritive material of the 484 bulletin: museum of comparative zoology host, which it uses to accelerate development and to reach an uncom- mon (for eggs of parasites Hymenoptera) increase of the primitive morula. But the embryonal cells are already completely surrounded by the polar ooplasm (trophamnios). From that stage of development on, the nutriment taken from the host, for its abundance and quantity, has an exciting action on the embryonal cells, which, being also separated by the cytoplasm, have a great total surface, to which a more active exchange corresponds. Such an augmentation of the embryonal cells produces a disharmony in the form of the total mass, which must acquire an irregular shape. The cause therefore of the unusual cell-augmentation, in which a disharmony of shape of the whole mass results, is the first step toward the process of dissociation, which appears later and originates through hypernutrition of the blastomeres, furnished by the elements contained in the host egg and elaborated by this also. The cleavage cells, because of this extraordinary augmentation, tend to break by centrifugal pres- sure the resistance of the surrounding polar involucrum and of the host tissue also, in which the polygerminal mass can be fixed. For me, in conclusion, the cause of polyembryony in Litomastix is the favorable nutriment as to quantity and quality, which induces an extraordinary activity of multiplication of the cleavage cells, either somatic or sexual, and the centripetal force of the separative tro- phamnios of groups of those cells. In the case of Litomastix the trophamnion is reduced later to a simple membrane, which surrounds each separated group also. These groups, finally, can develop har- monically in embryonal unity. The blastomeres of Litomastix, either somatic or to a lesser degree sexual, have a very great evolutionary capacity; therefore, immersed as they are, in a very favorable nutritive fluid, they multiply rapidly and tend to spread just as embryonal cells of pieces of tissues placed in adapted culture "in vitro;" they are prevented from an isolated dissociation by the polar involucrum, and they develop against it a centrifugal force. Meanwhile the involucrum produces constrictions, which first strangle the primitive germinal mass and later divide the same in groups. It remains to find the cause of the formation of blastomeres, distinct in the cytoplasm also after the first cleavage division in Litomastix. If we examine the egg structure during cleavage in the other polyem- bryonal species of Hymenoptera; and in various monembryonal Proctotrypidae, which, contrary to the usual rule, also have a cleavage extended to the nuclei and not to the ooplasm, we find as a constant SILVESTRi: INSECT POLYEMBRYONY 485 difference between one group and the other that the group with total cytoplasmic cleavage is provided with a polar involucrum (tro- phamnios) and the group with multiplication of nuclei only is destitute of such an involucrum; therefore it seems probable that the presence of a cytoplasmatic involucrum is the direct or indirect cause of the total cleavage. However, in seeking the reason why some Hymenopterous species of parasites of insects have acquired embryonal morphogenetic potency distinct from that of other species of the same families of Hymenoptera living as internal parasites of insects and sometimes in the same species of insects and having very small eggs also, it is possible to suspect that the cause is inherent in the diverse capacity of reaction which similar protoplasms can have, though placed in apparently similar condi- tions. It is remarkable that this particular potentiality of the egg has been acquired in similar conditions by some species of different families and yet not by all the species of a family (f.e. Chalcididae) or by all the species of a genus (f.e. Platygaster). The polyembryony of Hymenoptera is, from what we have seen, a real separation of blastomeres in groups, taking place at various stages of the morula of various dimensions, from Platygaster to Litomastix. Each group comprises, in the Chalcididae and Braconidae, blastomeres of two kinds: somatic and sexual or of one kind only, at least ap- parently, (Proctotrypidae) and therefore equipotent. A polyembryony similar to that of Hymenoptera, namely caused by separation into groups of primitive embryonal cells, was described by Harmer (1893-1900) for some very distinct genera (Crisia, Lichenopora, TubuUpora) of Bryozoa. In these Metazoa the egg in the ovicella remains suspended by a kind of funicle and later is surrounded by a follicular tissue, which can be compared physiologically, though not morphologically at least so far as known, to a trophamnios. The egg first forms, by total cleavage, a germinigenous mass. This increases rapidly through multiplication of its cells, always surrounded by the follicular tissue, and, under this centripetal action and the centrifugal pressure of its own cells, deforms, strangulates in various points, and produces buds or dissociation of embryonal cells groups, which at last develop into distinct embryos, each giving a larva. In Allolobophora caliginosa Savigny, f. trapezoides A. Duges (= Lum- briens trapezoides) Kleinenberg described a kind of gemmation of the gastrula producing 2 embryos. This twinning process is very different from the polyembryony of the Hymenoptera and of the Bryozoa; very different also from that of some Mammalia (Edentata), which 486 bulletin: museum of comparative zoology has been extensively studied, though the first cleavage stages are not yet very well known. In these Vertebrata (Fernandez, Newman and Patterson 1907, Patterson 1913, Newman 1917) it has been proved that from an egg a simple embryo develops till the blastodermic vesicle is complete; afterward in the case of Dasypus (= Tatusia) novemcincta the wall of the blastodermic vesicle, which has been pushed against the vegeta- tive pole, will become thinner in correspondence to the said pole, meanwhile inspissating at the two opposite (right and left) sides, from which first evaginations of the primary embryos arise. From each of these will take place the budding of secondary embryos, forming in this manner 4 embryos. These, later, remain surrounded each by an amniotic sac, but all four by the same chorion. In Mulita (Dasypus hybrida) similarly, 12 embryos can develop from one egg; in other genera of Edentata, allied to Dasypus also, the development is mon- embryonal, or in Eiiphractus (E. villosus) 2 twins develop, but from 2 eggs, though surrounded by the same chorion, for the reason that the 2 eggs remain contiguous in the uterus and develop in the same position. The cause of polyembryony in the Edentata is difficult to determine. Newman (1917) attributes it to activation of more morphogenetic centers, consequent to the relenting of metabolism in the central or animal pole of the blastodermic vesicle, a principle enunciated by Child on the basis of numerous studies of regulation of the form, especially in Cerianthus (1903-1905). Newman also maintains that polyembryony in the Edentata must be in relation to some special type of simple uterus, with the so-called germ-layer reversion and the reduction of the apical pole of the blastodermic vesicle with the great enlargement of the extra-embryonic cavity. In any case, the process of polyembryony in the Edentata and the Anellids also is very different from that of Hymenoptera. 7. Sex determination I must premise that the chromosomic set of the egg and of the spermatozoon of any species of polyembryonal Hymenoptera is not well-known to date. Therefore it is not possible to take into considera- tion the special chromosomic mechanism which regulates the produc- tion of males and females in any detail. The following general facts have been definitely observed in all the species: (1) the parthenogenetic eggs form 2 polar bodies, as well as the inseminated eggs; (2) the polar SILVESTRi: INSECT POLYEMBRYONY 487 bodies separate completely from the female pronucleus, and (3) from the female pronucleus, which becomes the first nucleus of segmenta- tion, there can originate many individuals, more than a thousand, all males. Patterson and Porter (1917) and afterward Patterson and Hamlet (1925) published results of researches demonstrating that the males of polyembryonal origin are all haploid ; therefore it should have been ascertained for these Hymenoptera that, haploidism is confined to the male sex, and diploidism obtained by the process of fertilization to the female sex; furthermore, that sex determination takes place in the egg with the complete act of maturation for the male, and with maturation plus fecundation, namely the conjugation of male and female pronuclei, for the female; and that such a determination has value for one as for a thousand or more individuals derived, as in the case of polyembryony, from the same egg, proving that sex determina- tion is very precocious and very tenacious, if not unchangeable. Patterson advanced the view that sex determination of hymenopterous individuals, originating by polyembryony, could be changed at least in certain cases, basing his statement on the presence of one or a few males among more numerous or many females in polyembryonal broods. Although it must not be absolutely excluded that among many females, originating from one egg, there could appear, for exceptional causes, a haploidism limited to few specimens, we do not have, so far, valid grounds for admitting a change of sex in a part, though minimal, of the same brood, because that which seemed of more value to Patterson, namely the appearance of one or few males in the broods of Platygaster felti, cannot be taken into consideration, as it is not proved to be due to polyembryony in this species. Therefore polyembryony confirms what is known for animals having a monoembryonic development of the classic type of the honey bee: (1) that the egg after maturation remains provided with a pronucleus which has male characteristics; (2) that the spermatozoa in such Meta- zoa carry a pronucleus of female characteristics, which has a dominant potency also ; (3) that the sex is determined in the egg before or during the essential process of fertilization and that it cannot be changed. It is only possible to admit that, under very exceptional conditions, the chromosome set of the sexual cells in their repeated divisions (which in the case of Litomastix can be more than a thousand) may lead to a change from diploid to haploid, or at least be without the chromosomes which determine the prevalent sex. But further proof is needed to validate such a supposition. 488 bulletin: museum of comparative zoology 8. Determination of the Sex Cells So far as we know, both in the polyembryonal Hymenoptera and the monembryonal, the chromosomes set, which has the first cleavage nucleus, derived from the female pronucleus alone or from its conjuga- tion with the male pronucleus, determines the sex, which, as we have seen in the preceding paragraph, must be considered immutable (or nearly so). But between the somatic and the sexual cells no difference has been detected so far, either in the number of chromosomes or in the presence of one of the two kinds of chromotinic reduction, as observed in 1887 by Boveri with the classic researches on Ascaris, or recently (1908) in the dipterous insect (Miastor) by Kahle. Therefore, though not excluding the possibility that in the Hymenoptera with polyem- bryonal development there may exist a chromatinic reduction in the nucleus of somatic cells, for the present this does not appear to have been observed by any author. On the contrary, in the Chalcididae and in the Braconidae we have full evidence to show that sex cells are determined by a plasmatic factor, namely the so-called oosome (or egg nucleolus). It is necessary to note that notwithstanding the plasmatic nature of the oosome, advanced by me and confirmed by Patterson, Leiby, and more particularly by Gatenby and Hegner, the latest methods of technique for differentiating the cytoplasmatic inclusions should be applied in order better to define the nature of the oosome. But at present it appears to be certain that in various families of different orders of insects, it is clearly of the determinant sex cells. Therefore the researches on polyembryonic Hymenoptera afford further evidence, already forthcoming, really numerous in Metazoa from the Chaetognatha (Sagitta) to the Arthropoda, of the importance of such an oosome in differentiating or organizing the sexual cells, because in these polyembryonic Hymenoptera more than a thousand individuals arise from the same egg and inherit part of the same oosome. This argument is also of great general biological interest, and is worthy of further careful comparative study in many forms. I recall that I myself, though using the same methods of preservation and coloration applied in other Chalcididae, could not detect the oosome in a contin- uously parthenogenetic species of Prospaltella (P. berlesei How.), to which genus bisexual species belong, also having an egg provided with an oosome. Therefore it will be important to find out in this case when and in what way the sexual cells differ from the somatic. The absence of oosome in said species should not be connected with their thelyto- SILVESTRi: INSECT POLYEMBRYONY 489 cous parthenogenesis, because in the paedogenesis of Miastor, the larvae, which produce other female larvae, form a polar body alone, as in Prospaltclla berlesei, and the egg is notwithstanding furnished with an oosome. Perhaps there is another nucleo-plasmatic mechanism, for which further search is necessary. 490 bulletin: museum of comparative zoology BIBLIOGRAPHY Allen, B. M. 1906. The Origin of the Sex-cells of Chrysemys. Anat. Anz., 29. Amma, K. 1911. Ueber die Differenzierung der Keimbahnzellen bei den Copepoden. Arch. Zellf., 6. Baehr, W. B. von. 1909. Die Oogenese bei einigen viviparen Aphididen und die Sperma- togenese von Aphis saliceti, mit besonderer Beriicksichtigung der Chromotinverhaltnisse. Arch. f. Zellf., 3, pp. 269-333, Taf. 12-15. Balbiani, E. G. 1864. Sur les mouvements qui se manifestent dans la tache germinative chez quelques animaux. Comp. Rend. Soc. Biol. Paris. Banta, A. M. 1916. Sex intergrades in a species of Crustacea. Proc. Nat. Acad. Sci., 2, pp. 578-583. Bataillon, E. 1900. Blastotomie spontanee et larves jumelles chez Petromyzon Planeri. Comp. Rend. Acad. Sci. Paris, 120, pp. 1201-1203. 1910. Pression osmotique de l'oeuf et polyembryonis experimentale. Ibid, 130, pp. 1480-1482. Boveri, T. 1887. Uber Differenzierung der Zellkerne wahrend der Furchung des Eies von Ascaris megalocephala. Anat. Anz. 2. Brandes, G. 1898. Germinogonie, eine neue Art der ungeschlechtlichen Fortpflanzung. Zeitsch. f. die gesammten Naturwissensch. Halle, 70, pp. 420^422. Brandt, H. 1936. Ueber die Aenderung des Geschlechtsverhaltnisses bei Insekten und ihre Ursachen. Arb. Phys. Angew. Ent. Berlin-Dahlem, 3, n° 3, pp. 218-221. Brindley, H. H., and Potts, F. A. 1910. The effect of parasitic castration in insects. Science, N. S., 32, n° 832, p. 836. Btjchner, P. 1910. Keimbahn und Ovogenese von Sagitta. Anat. Anz., 35. 1910. Die Schicksale des Keimplasmas der Sagitten in Reifung, Befruch- tung, Keimbahn, Oogenese und Spermatogenese. Festschr. R. Hertwig., 1, Jena. SILVESTRi: INSECT POLYEMBRYONY 491 1918. Vergleichende Eistudien 1. Die akzessorischen Kerne des Hymen- optereneies. Arch. f. Mikroskop. Anat., 91. II. Abt. pp. 1-202, Taf. I-X, 31 Texfig. Bonn. BtJGNION, E. 1891. Recherches sur le developpement postembryonnaire, l'anatomie et les moeurs de YEncyrtus fuscicollis. Recueil Zool. Suisse, 5, pp. 435-534, pi. xx-xxv. Chapin, Catharine L. 1917. A microscopic study of the reproductive system of foetal free- martins. Journ. Exp. Zool., 23, n° 2, pp. 453-478, 16 text figs. Crampton, H. E. 1899. An experimental study upon Lepidoptera. Arch. Entw.-Mech., 9, pp. 293-298, pi. 11-13, 13 text figs. CUENOT, L. 1899. Sur la determination du sexe chez les animaux. Bull. Scient. France et Belgique, 32, pp. 462-535. 1903. L'ovaire du Tatou et l'origine des jumeaux. Comp. Rend. Soc. Biol., 60, pp. 1391-1392. DONCASTER, L. 1914. Chromosomes, heredity and sex: a revision of the present state of the evidence with regard to the material basis of hereditary trans- mission and sex-determination. Quart. Jour. Micr. Sci., N.S., 59, n° 236, pp. 487-521, 4 text figs. Daniel, D. Mc. 1932. Macrocentrus ancylivorus Rohwer, a polyembryonic Braconid parasite of the Oriental fruit Moth. New York State Agr. Expt. Stat., Tech. Bull. N° 187, 101 pags., 18 figs. Geneva, N. Y. Driesch, H. 1892. Entwickelungsmechanische Studien. Zeitsch. f. Wiss. Zool., 53, pp. 160-183. Elpatiewsky, W. 1910. Die Entwickelungsgeschichte der Genitalprodukte bei Sagitta. 1. Die Entwickelung der Eier. Biol. Zeitsch., 1. Ferriere, C. 1926. Note sur un Chalcidien a developpement polyembryonique. Rev. Suisse Zool., 33, pp. 585-596, figs. 1-7. Fink, D. F. 1926. The biology of Macrocentrus ancylivora Rohwer, an important parasite of the Strawberry leaf-roller (Ancylis comptana Froehl.). Jour. Agr. Research, 32, pp. 1121-1134; illus. 492 bulletin: museum of comparative zoology Friederichs, K. 1906. Untersuchung Qber die Entstehung der Keiblatter und Bildung des Mitteldarms bei Kafern. Nova Acta, Abh. K. Leop.-Carol. Akad. Nat., 85, No. 3. FUNKHOUSER, W. D. 1915. Life history of Thelia bimaculata Fab. (Membracidae). Ann. Entomol. Soc. Amer., 8, n. 2, pp. 140-151, pi. 9. 1917. Biology of the Membracidae of the Cayuga Lake Basin: Cornell Univ. Agric. Exp. Stat., Memoir 11, pp. 181-445, pi. 23-44, 6 text figs. Gahan, A. B. 1933. The Serphoid and Chalcidoid Parasites of the Hessian Fly. U. S. Dept. Agr. Miscell. Public. N° 174, 147 pags., 32 figs. Washing- ton, D. C, Dec. 1933. Ganin, M. 1869. Beitrage zur Erkenntniss der Entwickelungsgeschichte bei der Insecten. Zeitsch. f. Wiss. Zool., 19, pp. 381-449, pi. xxx-xxxiii. Gatenby, J. B. 1918. Polyembryony in parasitic Hymenoptera; a review. Quant. Jour. Micr. Sci., 63, pt. 2, pp. 175-196. Geyer, Kurt. 1913. Untersuchungen liber die chemische Zusammensetzung der Insektenhaemolymphe und ihre Bedeutung fur die geschlechtliche Differenzierung. Zeitsch. f. Wiss. Zool., 105, pp. 350-499, pis. 20-22, 58 text fig. GlARD, A. 1898. Sur le developpement de Litomastix truncatellus (Dalman). Bull. Soc. Entomol. France, p.. 127-129. GOLDSCHMIDT, R. 1917. A further contribution to the theory of sex. Jour. Exp. Zool., 22, n° 3, pp. 593-611, 53 text figs. Haecker, V. 1897. Die Keimbahn von Cyclops. Arch. Mikr. Anat., 45. Harmer, S. F. 1893. On the occurrence of embryonic fission in cyclostomatous Polyzoa. Quart. Jour. Micr. Sci., 34, pp. 199-242, 3 pi. 1895. Preliminary note on embryonic fission in Lichenopora. Proc. Roy. Soc, 57, p. 188. 1896. On the Development of Lichenopora verrucaria Fabr. Quart. Journ. Micr. Sci., 39, pp. 71-144, pi. vii-x. SILVESTRi: INSECT POLYEMBRYONY 493 Hegner, R. W. 1916. The Effects of Removing the Germ-Cell Determinants from the Eggs of some Chrysomelid Beetles. Biol. Bull., 16. 1909. The Origin and Early History of the Germ Cells in some Chry- somelid Beetles. Journ. Morph., 20. 1914. Studies on germ cells. III. The origin of the Keimbahn-Deter- minants in a parasitic Hymenopteron, Copidosoma. Anat. Anz., 46, n° %, pp. 51-69. 1914. Studies on germ cells. I. The history of the germ cells in insects with special reference to the Keimbahn-determinants. II. The origin and significance of the Keimbahn-determinants in animals. Journ. Morph., 25, n° 3, pp. 375-499. 1914. The germ-cell cycle in animals. Macmillan Company, New York. 1915. Studies on germ cells. IV. Protoplasmic differentiation in the oocytes of certain Hymenoptera. Jour. Morph., 26, n° 3, pp. 495- 535. 1917. The genesis of the organization of the insect egg. Amer. Natural., 41, n° 611, pp. 641-661; n° 612, pp. 705-718; 12 text figs. Hoy, W. E. 1916. A study of somatic chromosomes. I. The somatic chromosomes in comparison with the chromosomes in the germ cells of Anas tristis. Biol. Bull., 31, n° 5, pp. 329-363, 10 text figs. Howard, L. O. 1891. The biology of the hymenopterous insects of the family Chal- cididae. Proc. U. S. Nat. Mus., 14, pp. 567-588. 1891. The methods of pupation among the Chalcididae. Insect Life, 4, pp. 193-196. 1906. Polyembryony and the fixing of sex. Science, N. S., 24, n° 625, pp. 810-818. 1907. Polyembryony and the fixation of sex. Proc. Ent. Soc. Wash., 8, n° 3-4, pp. 145-148. Howard, L. O. 1919. Two new instances of Polyembryony among the Encyrtidae. Science, 49, .pp. 43-44. 1925. An obvious new case of polyembryony. Science, 62, p. 308. Kahle, W. 1908. Die Paedogenesis der Cecidomyiden. Zoologica, 55. Kammerer, P. 1912. Ursprung der Geschlechtsunterschiede. Fortschr. d. Naturwiss. Forsch. (Abderhalden), 5, pp. 1-240. 494 bulletin: museum of comparative zoology Kellogg, V. L. 1904. Influence of the primary reproductive organs on the secondary sexual characters. Jour. Exp. Zool., 1, n° 4, pp. 601-605. Kleinenberg, N. 1879. The Development of the Earth-Worm, Lumbricus trapezoides Duges. Quart. Journ. Micr. Sci., 29, p. 206. Kopec, Stefan. 1911. Untersuchungen iiber Kastration und Transplantation bei Schmetterlingen. Arch. Entw.-Mech., 33, pp. 1-116, pi. 1-4, 19 text fig. 1913a. Untersuchungen iiber die Regeneration von Larvalorganen und Imaginalscheiben bei Schmetterlingen. Arch. Entw.-Mech., 37, pp. 440-472, pi. 12-14, 6 textfig. 1913b. Nochmals iiber die Unabhangigkeit der Ausbildung sekundarer Geschlechtscharaktere von den Gonaden bei Lepidopteren (Fiih- lerregenerationsversuche mit Kastration und Keimdriisentras- plantation komniniert). Zool. Anz., 43, n° 2, pp. 65-74. KORNHAUSER, S. I. 1919. The sexual characteristics of the membracid, Thelia bimaculata (Fanr.). I. External changes induced by Aphelopus theliae (Gahan) Journ. Morph., 32, pp. 531-636, illus. Ktjlagin, N. 1898. Beitrage zur Kenntniss der Entwicklungsgeschichte von Platy- gaster. Zeitch. f. Wiss. Zool., 62, pp. 195-235, pi. x-xi. Leiby, R. W. 1922. Biology of the goldenrod gall-maker, Gnorimoschema gallae- solidaginis Riley. Jour. New York Entom. Soc, 30, n° 2, pp. 81-94. 1922. The polyembryonic development of Copidosoma genechiae, with notes on its biology. Jour. Morph., 37, pp. 195-285. Leiby, R. W. and Hill, C. C. 1923. The twinning and monembryonic development of Platygaster hiemalis, a parasite of the Hessian fly. Jour. Agr. Research, 25, pp. 337-349. 1924. The polyembryonic development of Platygaster vernalis. Journ. Agr. Research, 28, pp. 829-839. 1926. The origin of mixed broods in polyembryonic Hymenoptera. Ann. Ent. Soc. Amer., 19, pp. 290-299. LlLLIE, F. R. 1917. The free-martin; a study of the action of sex hormones in the foetal life of cattle. Jour. Exp. Zool., 23, n° 2, pp. 371-452, 29 figs. SILVESTRi: INSECT POLYEMBRYONY 495 LOEB, J. 1894. Ueber eine einfache Methode zwei oder mehr zusammengewach- sene Embryonen aus einem Ei hervorzubringen. Pfluger's Archiv., 54. Marchal, P. 1897. Les Cecidomyides des cereales et leurs parasites. Ann. Soc. Ent. France, 66, pp. 1-105, pi. i-viii. 1898. La dissociation de l'oeuf en un grand nombre d'individus destincts chez VEncyrtus fuscicollis. Comp. Rend. Acad. Sci. Paris, 126, pp. 662-664; Comp. Rend. Soc. Biol. ser. 10, 5, pp. 238-240; Bull. Soc. Entom. France, pp. 109-111. 1899. Comparaison entre les Hymenopteres parasites a developpement polyembryonnaire et ceux a developpement monoembryonnaire. Comp. Rend. Soc. Biol. ser. 11, 1, pp. 711-713. 1902. Observations sur la biologie des Hyponomeutes. Bull. Soc. d'etudes et de vulgarisation de la Zoologie agric. de Bordeaux, 1, fasc. 4, pp. 13-26. 1903. Le cycle evolutif du Polygnotus minutus (Lindm). Bull. Soc. Entom. France, pp. 90-93. 1904. Le determinisme de la polyembryonie et le determinisme du sexe dans la polyembryonie specifique des Hymenopteres. Comp. Rend. Soc. Biol., 56, p. 468. 1904. Sur la formation de l'intestin moyen chez les Platygasters. Ibid, 56, p. 1091. 1904. Recherches sur la biologie et le developpement des Hymenopteres parasites. I. La polyembryonie specifique ou germinogonie. Arch. Zool. Expt. et Gen. ser. 4, 2, pp. 257-335, illus. 1906. Recherches sur la biologie et le developpement des Hymenop- teres parasites. II. Les Platygasters. Ibid., 4, pp. 485-640, pi. xvii-xxiv. Martin, F. 1914. Zur Entwicklungsgeschichte des polyembryonalen Chalcidiers Ageniaspis (Encyrtus) fuscicollis Dalm. Zeit. f. Wiss. Zool., 110, pp. 419-479. Meisenheimer, J. 1909. Experimentelle Studien zur Soma-und Geschlechts-Differenzie- rung. Erster Beitrag. Gustave Fischer. Jena. pp. 1-147, 2 taf., 55 Textfig. Morgan, T. H. 1909. A biological and cytological study of sex determination in Phyllox- erans and Aphids. Jour. Exp. Zool., 7, n° 2, pp. 239-352, 1 pi., 23 text figs. 496 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY Morrill, C. V. 1910. The Chromosomes in the oogenesis, fertilization and cleavage of Coreid Hemiptera. Biol. Bull., 19, n° e, pp. 79-126, 2 pi., 12 text figs. Noack, W. 1901. Beitrage zur Entwickelungsgeschichte der Musciden. Zeitschr. f. Wiss. Zool., 70. OUDEMANS, J. Th. 1898. Falter aus castrirten Raupen, wie sie aussehen und wie sie sich benehmen. Zool. Jahrb., Abt. f. Syst., 12, pp. 71-88, pi. 3-5, 2 Text fig. Parker, H. L. 1924. Recherches sur les formes post-embryonnaires des Chalcidiens. Ann. Soc. Ent. France, 93, pp. 261-392, illus. Parker, H. L. 1931. Macrocentrus gifuensis Ashmead, a polyembryonic Braconid para- site in the european Corn Borer. U. S. Dept. Agr. Tech. Bull., N° 230, March, pp. 62, figs. Patterson, J. T. 1913. Polyembryonic development in Tatusia novemcincta. Jour. Morph., 24, n° 4, pp. 559-662. 1915. Observations on the development of Copidosoma gelechiae. Biol. Bull., 29, N° 6, pp. 333-373. 1917a. Studies on the biology of Paracopidosomopsis. I. Data on the sexes. Biol. Bull., 32, n° 5, pp. 291-305. 1917b. Maturation and Fertilization. Biol. Bull., 33, n°. 2, pp. 57-62. 1918. Asexual larvae. Biol. Bull., Vol. 35, (1918) N°. 6, pp. 362-377. 1919. Polyembryony and sex. Journ. Hered., 10, n°. 8, pp. 344-352. 1921a. Sex-ratios in Platygaster. Amer. Nat., 55, n°. 637, pp. 180-183. 1921b. The development of Paracopidosomopsis. Journ. Morph., 36, pp. 1-69. 1927. Polyembryony in animals. Quart. Review Biology, 2, pp. 399- 426. Patterson, J. T. and Porter, L. T. 1917. Studies on the biology of Paracopidosomopsis. II Spermatogenesis of males reared from unfertilized eggs. Biol. Bull., 33, no. 1, pp. 38-51. Patterson, J. T. and Hamlett, G. W. D. 1925. Haploid males in Paracopidosomopsis. Science, 61, p. 89. SILVESTRi: INSECT POLYEMBRYONY 497 Puppini, G. 1930. Contribute) alia conoscenza dell' Anarsia lineatella Zeller e appunti sulla Recurvaria nanella Hubn. Boll. Labor. Entom. R. Istut. Sup. Agr. di Bologna, 3, pp. 182-220, 2 pis., 18 figs. Riley, W. A. 1907. Polyembryony and sex determination. Science, n.s., 25, n°. 629, pp. 106-107. RlTTER, R. 1890. Die Entwickelung der Geschlechtsorgane und des Darmes bei Chironomus. Zeitschr. f. Zool., 50. Ruud, G. 1925. Die Entwickelung isolierter Keimfragmente friihester Stadien von Triton taeniatus. Roux' Arch. Entwick., 105. Sarra, R. 1915. Osservazioni biologiche sull' Anarsia lineatella Z., dannosa al frutto di mandorlo. Boll. Lab. di Zool. gen. e agr. della R. Scuola Sup. d'Agricoltura in Portici, 10, pp. 51-65. 1918. La Variegana e di suoi parassiti. Boll, del Lab. di Zool. gen. e agr. della R. Scuola Sup. dAgricoltura in Portici., 12, pp. 175-187. Silvestri, F. 1906. Contribuzioni alia conoscenza biologica degli Imenotteri parassiti. I. Biologia del Litomastix truncatellus (Dalm). Boll. Lab. Zool. Sc. Agr. Portici, 1, pp. 17-64, 13 figs., pis. i-v. 1907. Contribuzioni, etc. II; Sviluppo dell' Ageniaspis fuscicollis (Dalm.). III. Sviluppo dell' Encyrtus apidivorus Mayr. IV. Sviluppo dell' Oopthora semplidis. Boll. Lab. Zool. Sc. Agr. Portici, 3, pp. 30-84, 52 figs., pis. i-ii. 1914. Prime fasi di sviluppo del Copidosoma Buyssoni (Mayr), Imen- ottero Calcidide. Anatomischer Anz., 47, pp. 45-56, 30 figs. 1915. Struttura dell' ovo e prime fasi di sviluppo di alcuni Imenotteri parassiti: I.-V. Boll. Lab. Zool. Sc. Agr. Portici, 10, pp. 66-88, 4 figs., pis. i-vi. 1916. Sulla maturazione dell' ovo, fecondazione e formazione del tro- phamnios nel Platygaster dryomyae Silv. (Imenottero procto- trypidae). Rend. Ace. Lincei (5) 25, pp. 121-128, 2 figs. 1923a. Contribuzioni alia conoscenza dei Tortricidi della Querce. (I-II.) Boll. Lab. Zool. gen. agr. Sc. Agr. in Portici, 17, pp. 41-107, 47 figs. 1923b. Contribuzioni alia conoscenza degli insetti del Nocciuolo. III-VI. Boll. Lab. Zool. gen. agr. R. Sc. sup. Agr. in Portici, 17, pp. 221- 301, 2 pis., 49 figs. 498 bulletin: museum of comparative zoology Steche, O. 1912. Die "sekundaren" Geschlechtscharaktere der Insekten und das Problem der Vererbung des Geschlechts. Zeitschr. f. Indukt. Abstamm. u. Vererb., 8, no. 3, pp. 284-291. Tandler, J. and Gross, S. 1913. Die biologischen Grundlagen der sekundaren Geschlechtscharak- tere. Berlin, Julius Springer. 169 p., 23 text fig. VOUKASSOVITCH, P. 1927. Observations biologiques sur le Macrocentrus abdominalis Fab., Braconide parasite. Comp. Rend. Soc. Biol. Paris, 96, pp. 379- 381. 1929. Contribution a l'etude de Macrocentrus abdominalis F. et de ses parasites. Ann. Soc. Ent. France, 98, pp. 163-187, illus. Waterston, J. 1920. On a New Polyembryonic Encyrtid (Chalcidiodea) Copidosoma tortricis, sp.n., bred from the Strawberry Tortrix Moth. Ann. App. Biol., Cambridge, 7, n° 1, Sept. pp. 1-5, 5 figs. Wheeler, W. M. 1910. The effects of parasitic and other kinds of castration in insects. Journ. Exp. Zool., 8, pp. 337-438. EXPLANATION OF PLATES PLATE 1 Silvestri — Insect Polyembryony. PLATE 1 Fig. 1. Plusia gamma: adult. Figs. 2-3. Female of Litomastix laying the egg in the egg of Plusia, seen laterally and from the dorsum. Fig. 4. Larva of Plusia gamma. Figs. 5-7. Carcasses of larvae of Plusia gamma full of cocoons of Lito- mastix. Fig. 8. Sagittal section of a developing egg of Plusia: a amnios, c chorion, s serosa, o (also the other black balls) egg of Litomastix. BULL. MUS. COMP. ZOOL. Silvestri. Insect Polyembryony. Plate 1 . vVL-i'f-W 4 8 PLATE 2 Silvestri — Insect Polyembryony. PLATE 2 Fig. 9. Egg completely developed: c chromatin of the nucleus of the oo- cyte of first order, d chorion, m micro pyle, n oosome (or nucleolus). • Fig. 10. Egg half an hour after deposition: a nucleus of the first polar body in anaphase, b nucleus of the oocyte of second order in anaphase, n oosome, s head of spermatozoon. Fig. 11. Egg at a little later stage than the preceding: a1 and o2 nuclei derived from the first polar body; b1 second polar body, b'~ female pronucleus, n and s as above. Fig. 12. Egg at a little later stage than the preceding: s male pronucleus, I2 female pronucleus, the other signs as above. Fig. 13. Egg with first cleavage nucleus in mitosis. Fig. 14. Egg with two cells of segmentation (N1 and N2, n oosome) and the undivided polar ooplasme with the polar nucleus (p) derived by fusion of the 3 nuclei of the polar bodies. Fig. 15. Egg a little more advanced with the two segmentation (embryonal) cells in mitosis. Fig. 16. Egg with four embryonal cells, one of which has received the entire oosome. Fig. 17. Egg a little more advanced in development, with the embryonal cell containing the oosome in metaphase and the other three in anaphase and the polar nucleus in anaphase also. Fig. 18. Egg with the embryonal part composed of 14 cells, of which two only have the oosome substance scattered in the cytoplasm and are repre- sented as separated also in the upper right part in comparison with a cell lack- ing oosome. BULL. MUS. COMP. ZOOL. Silvestri. Insect Polyembryony. Plate 2. PLATE 3 Silvestri — Insect Polyembryony. PLATE 3 Fig. 19. Egg about seven hours after deposition: the polar part (P) has 32 nuclei, of which 30 are readily visible; and the embryonal part (E) is com- posed of about 30 cells. Fig. 20. Optical longitudinal section of an egg at a later stage, having a great number of polar nuclei and embryonal cells. Fig. 21. Optical longitudinal section of a polygerminal mass: a sexual cells, b somatic cells, r involucrum (trophamnios). BULL. MUS. COMP. ZOOL. Silvestri. Insect Po-Yembryony. Plate 3. * ■* H^jy?«Mi 2 I PLATE 4 Silvestri — Insect Polyembryony. PLATE 4 Fig. 22. Longitudinal section of a complex of secondary germinigenous masses and of some monembryonal: CA adipose tissue of the host larva, cc sexual cell of a secondary germinigenous mass, cm2 secondary monembryonal mass, cp external involucrum of the monembryonal and of the germinigenous masses, cs somatic cells; im? internal involucrum of the secondary monembry- onal masses, m2 embryonal morula of the secondary monembryonal masses. Fig. 23. Sexual larva of Litomastix showing the internal organs also in transparency: A imaginal disk of the antennae, a and b imaginal disk of wings, E supraoesaphageal ganglion, G gonads, I middle intestine, N ganglionar chain, P posterior intestine, S labial glands, T malpighian tubes, 1-3 imaginal disks of the feet. Fig. 24. A sexual larva of Litomastix showing the internal organs. Fig. 25. The same seen superficially with the addition of the parietal muscles: N ganglionar chain, C fat cells, I intestine, S labial glands. BULL. MUS. COMP. ZOOL. Silvestri. Insect Polyembryony. Plate 4. /? • » *" >/ h 1, INC. AUG 2 3 1984 :iDGE STREET : VFOWN, Harvard MCZ Library 3 2044 066 303 546