f\) j- \r ( | | LJ HARVARD UNIVERSITY cal LAS LIBRARY OF THE Museum of Comparative Zoology } bf , a! 7 * 7 j F a | jie a | ee | “LS —ta vu ; Bs) " iD | Tr | i i Pee POSTIELA NOG a 7 : ; 4 ; Pa ee ee POSTIELA NOS. 1250 MARCH 1950 — JUNE 1961 MUS, Comp. ZOOL LIBRARY APR 23 1969 HARVARD UNIVERSITY, PEABODY MUSEUM OF NATURAL HISTORY YALE UNIVERSITY, NEW HAVEN, CONN. CONTENTS List of Papers 5 Author Index 8 List of New Genera 9 List of New Species 9 asteot New Subspecies. 2%. oe. san6. seeks: | 10 List of Maps and I]lustrations — 11 et ed radere > Ole No. nw Or =~ LIST OF PAPERS Notes on Indian Birds, III. Birds from Assam. S$. Dillon Ripley. March, 1950. Rare Migration and Wintering Records from the Yucatan Peninsula. Raymond A. Paynter, Jr. March, 1950. A Small Collection of Birds from Argentine Tierra Del Fuego. S. Dillon Ripley. April, 1950. A New Tanager from Mexico. Raymond A. Paynter, Jr. May, 1950. A Large Pyenodont from the Niobrara Chalk. Joseph T. Gregory. December, 1950. Notes on Indian Birds, IV. Some Recently Collected Birds from Assam. S. Dillon Ripley. February, 1951. New Passerine Birds from the Indo-Chinese Subregion. H. G. Deignan. May, 1951. Bassariscus in’ Miocene Faunas and “‘Potamotherium Lycopotamicum Cope.” Joseph T. Gregory and Theodore Downs. May, 1951. Birds Collected and Noted Round Dhahran, Saudi Arabia, and Bahrein Island. S. Dillon Ripley. May, 1951. Three Birds from the Mountains of Muscat. S. Dillon Ripley. November, 1951. Geographical Variation in Garrulaw Sannio Swinhoe. H. G. Deignan. March, 1952. A New Genus of Thrush from Eastern Africa. S. Dillon Ripley. April, 1952. The Thrushes. S. Dillon Ripley. September, 1952. A New Race of Black-Throated Babbler from Assam. S. Dillon Ripley. December, 1952. Typothorax Seutes from Germany. Joseph 'T. Gregory. May, 1953. Typothorax and Desmatosuchus, Joseph 'T. Gregory. June, 1953. Notes on Indian Birds, V. S. Dillon Ripley. December, 1953. Three New Birds from the Yucatan Peninsula. Raymond A. Paynter, Jr. March, 1954. Birds from Gough Island. S. Dillon Ripley. July, 1954. » 6 38. LIST OF PAPERS Notes on Indian Birds, VI. Additional Comments on the Wren-Babbler, Spelaeornis. S. Dillon Ripley. July, 1954. A New Fruit Pigeon from the Philippines. S. Dillon Ripley and D. 8S. Rabor. February, 1955. Additions to the Ornithogeography of the Yucatan Penin- sula. Raymond A. Paynter, Jr. April, 1955. A New White-Throated Spinetail from Western Brazil. S. Dillon Ripley. October, 1955. A Thresher Shark from Long Island Sound. James E. Morrow. December, 1955. Avifauna of the Jorullo Region, Michoacan, Mexico. Raymond A. Paynter, Jr. March, 1956. Cuban Bird Notes. S. Dillon Ripley and George E. Watson, 3rd. May, 1956. Meteorites in the Collections of Yale University. Kurt Servos. September, 1956. The Species of Notharctus from the Middle Eocene. Peter Robinson. January, 1957. A Redefinition of the Subspecies of Fodiator Acutus. James E. Morrow. February, 1957. Notes on the Horned Coot, Fulica Cornuta Bonaparte. S. Dillon Ripley. February, 1957. New Birds from the Western Papuan Islands. 8. Dillon Ripley. March, 1957. Additional Notes on the Horned Coot, Fulica Cornuta Bonaparte. S. Dillon Ripley. June, 1957. On a New Species of Anisops (Hemiptera, Notonectidae) from the Moluccas. G. E. Hutchinson. November, 1957. Notes on an Additional Example of the Fruit Bat, Sco- tonycteris Ophiodon Pohle. Alvin Novick. March, 1958. Notes on Indian Birds, VII. S. Dillon Ripley. April, 1958. On the Doubtful Validity of T'achypleus Hoeveni Pocock, an Indonesian Horseshoe Crab (Xiphosura). Talbot H. Waterman. June, 1958. A Note on the Firethroat and the Blackthroated Robin. S. Dillon Ripley. September, 1958. Comments on Birds from the Western Papuan Islands. 5S. Dillon Ripley. April, 1959. ~ LIST OF PAPERS On Makaira Nigricans of Lacépede. James E. Morrow, Jr. May, 1959. A New Species of Grammatostomias (Family Melano- stomiatidae) from the Western North Atlantic. James E. Morrow, Jr. May, 1959. Birds from Djailolo, Halmahera. S. Dillon Ripley. Sep- tember, 1959. Character Displacement in Indian Nuthatches (Sitta). S. Dillon Ripley. December, 1959. Two New Birds from Angola. S. Dillon Ripley. January, 1960. Sinopa from the Cuchara Formation of Colorado. Peter Robinson. February, 1960. Notes on the Embryolog gy and Evolution of the Mega- 4 podes (Aves: Galliformes). George A. Clark, Jr. Febru- ary, 1960. Fossil Amphibians from Quarry Nine. Max K. Hecht and Richard Estes. June, 1960. Additions to the Avifauna of Northern Angola, I. S. Dillon Ripley and Gerd H. Heinrich. July, 1960. Rodents and Lagomorphs from the Miocene Fort Logan and Deep River Formations of Montana. Craig C. Black. January, 1961. Results of Research in the Antofagasta Ranges of Chile and Bolivia. Luis E. Pena and Ruth Patrick. June, 1961. The Avifauna of Mount Katanglad. S. Dillon Ripley and D. S. Rabor. June, 1961. AUTHOR INDEX Black, Craig C., 48 Clark, George A., Jr., 45 Deignan, H. G. 7, 11 Downs, Theodore and Joseph T. Gregory, & Estes, Richard and Max K. Hecht, 46 Gregory, Joseph T., 5, 15, 16 Gregory, Joseph IT’. and Theodore Downs, 8 Hecht, Max K. and Richard Estes, 46 Heinrich, Gerd H. and 8. Dillon Ripley, 47 Hutchinson, ‘G. E.., 33 Morrow, James E., Jr., 24, 29, 39, 40 Novick, Alvin, 34 Patrick, Ruth and Luis E. Pena, 49 Paynter, Raymond A., Jr., 2, 4, 18, 22, 25 Pena, Luis E. and Ruth Patrick, 49 Rabor, D. S. and 8. Dillon Ripley, 21, 50 Ripley, Sz Dillon, 1,3, 6, 9,10, 12, 13, 14,47 lO oO BL, GOs Oi GO, 41, 42, 43 Ripley, S. Dillon and Gerd H. Heinrich, 47 Ripley, S. Dillon and D. S. Rabor, 2/, 50 Ripley, S. Dillon and George E. Watson, 3rd, 2¢ Robinson, Peter, 28, 44 Servos, Kurt, 27 Waterman, Talbot H., 36 Watson, George E., 3rd, and S. Dillon Ripley, 26 . 30 LIST OF NEW GENERA Comobatrachus aenigmatis, 46: 6 Comonecturoides mirshi, 46: 8 Modulatria, 12: 2 Niglarodon, 45: 2 LIST OF NEW SPECIES Amphora atacamana, 49: 50 Amphora boliviana, 49: 51 Amphora carvajaliana, 49: 52 Anisops sylvia, 33: 1 Dikkomys woodi, 48: 13 Eumys eliensis, 48: 7 Erythrura coloria, 50: 18 Grammatostomias circularis, 40: J Hadrodus marshi, 5: 1 Megalagus dawsoni, 48: 17 Monarcha julianae, 3S: 9 Navicula atacamana, 49: 45 Navicula carvajaliana var, carvajaliana, 49: 45 Navicula luisti, 49: 48 Navicula pseudosepta, 49: 49 Nectarinia sororia, 43: 2 Niglarodon koerneri, 48: 3 Paciculus montanus, 48: 10 Ptilinopus arcanus, 21: 1 Serinus mindanensis, 50: 18 LIST OF NEW SUBSPECIES Acridotheres cristatellus fumidus, 1: 4 Aepypodius arfakianus misoliensis, 31: 1 Ammomanes deserti insularis, 9: 6 Anthus similis travancoriensis, 17: 2 Arborophila torqueola interstincta, 6: 1 Collocalia esculenta nubila, 41: 4 Crateroscelis murina fumosa, 31: 3 Dendrocolaptes certhia legtersi, 18: 1 Dendrocopos darjellensis fumidus, 6: 3 Ducula badia carolinae, 17: 1 Dumetella glabrirostris cozumelana, 18: 3 Eos squamata attenua, 31: 2 Galerida cristata thomsi, 10: 1 Garrulaxw sannio comis, 11: 3 Garrulax sannio oblectans, 11: 3 Gerygone magnirostris occasa, 31:3 Horeites flavolivaceous alewanderi, 6: 6 Indicator xanthonotus fulvus, 6: 2 Nectarinia sericea mariae, 3S: 13 Oligura castaneo-coronata ripleyt, 7: 3 Parus major vauriet, 1: 2 Pellorneum ruficeps vocale, 7: 2 Pericrocotus flammeus gonzalesi, 50: 8 Phrygilus unicolor ultimus, 3: 10 Phylloscopus fuscatus mariae, 6: 5 Piranga roseo-gularis tincta, 4: 1 Platyrinchus mystaceus timotheit, 18: 2 Prinia gracilis anguste, 9: 10 Psalidoprocne albiceps suffusa, 43: 1 Pycnonotus leucotis dactylus, 9: 8 Pyrrhula leucogenys coriaria, 50: 17 Spelacornis chocolatinus chocolatinus, 20: 4 Spelaeornis chocolatinus nagaensis, 6: 4 Stachyris nigriceps coei, 14: 2 Sturnus contra sordidus, 1: 38 Turnia nana insolata, 47: 2 Nanthotis chrysotis austera, 31: 4 10 List OF MAPS AND ILLUSTRATIONS MAPS “sag se localities of Gerd Heinrich in North Angola 1957- 58, doetihen orientation of species of Sitta, 42: 6 Jorullo region, Michoacan, Mexico, 25: 12 Map of areas visited by Luis Pena, 49: 7 Mindanao Island, 50: 7 ILLUSTRATIONS Alopias vulpinus from Long Island Sound, 24: 3, fig. 1 A portion of the malpais seen from Jorullo, 25: 11, fig. 2 Comobatrachus aenigmatis, 46: 15, 17, pl. 1, figs. 2, 4, 6, pl. 2, fig. 2 Comonecturoides marshi, 46: 17, pl. 2, figs. 3, + Comparative series of urodele eae 46: 19, pl. 3, figs. 1-9 Diatoms from the alimentary tract of Phoenicoparrus jamesi (Sclater), 49: 57,.pl. 1, figs. 1-12 Eobatrachus agilis Marsh, 46: 15, 17, pl. 1, figs. 1, 3, 5, pl. 2, fig. 1 Head of Fulica cornuta, 30: 5 Horned Coot contrasted with Larus serranus, 30: 3 Horseshoe crabs (Xiphosura), 36: 15, 15; 17, pls. I, II, IU, figs. 1-11 Inacaliri marshes and tolar region, 49: 9 Jorullo from the plain at La Puerta, 25: 11, fig. Laguna Colorada in Bolivia, 49: 11 Male “parina chica,” Phoenicoparrus jamesi (Sclater), 49: 27 Notharctus gracilis (Marsh), 28: 25, 27, pl. 1, figs. 1, 2, 4-8, pl. HH, figs. 2, 3,.6 Notharctus robustior Leidy, 28: 27, pl. I, figs. 1, 4-5 > mils Notharctus tenebrosus? 28: 25, pl. 1, fig. 3 Notharctus tenebrosus Leidy, 28: 25, 27, pl. I, figs. 9-10, pl. LD, fig: Phoenicoparrus jamesi (Sclater), 49: 5 Senor Pena at Laguna Verde, 30: 3 11 ibe LIST OF MAPS AND ILLUSTRATIONS Serinus mindanensis and Erythrura coloria, 50: 5 Sinopa cf. vulpecula Matthew, 44: 2 Toconce and Paniri volcanoes, 49: 16 Traversay drawing of Makaira nigricans, 39: 2, fig. 1 Unidentified anuran, YPM 1394, 46: 17, pl. 2, figs. 5, 6 Unknown reptile, YPM 1568, 46: 15, pl. 1, figs. 7, 8 Where Io YALE PEABODY MUSEUM oF Natur:t History Number 1 March 10, 1950 New Haven, Conn. NOTES ON INDIAN BIRDS III BIRDS FROM ASSAM S. Ditton RieLrey I have recently been checking over specimens of birds from Assam, some kindly presented to me by Mr. Salim Ali, and others recorded by me on a recent survey for the Assam Govern- ment. The following notes appear worthy of setting down here. Alcippe rufogularis Comparison of a freshly-collected specimen of the Red- throated Tit-babbler from Gabru, Belsiri River, north Assam, on the edge of the Bhutan Duars (type locality of rufogularis), with a fresh specimen from Tezu, near Sadiya in the Mishmi Hills, shows that there are two races of this species as follows: 1. Alcippe rufogularis rufogularis Mandelli (type locality Bhutan Duars). Characters: brown above and paler, more gray below. Range: Bhutan Duars and northern Assam north of the Brahmaputra east presumably to the Dihang. 2. Alcippe rufogularis collaris Walden (type locality Sadiya). Characters: much darker on the crown and rufescent on the back. Flanks more heavily washed with brown. Range: northern Assam east of the Brahmaputra and Dihang, south to Manipur. Only fresh specimens of this species are worth comparing, as is the case with some other members of the genus. I am most grateful to Mr. H. B. Usher for help in comparing my material with the types and other specimens in the British Museum. 2 Postilla Yale Peabody Museum No. 1 Parus major I have examined thirteen fresh skins from Nepal, Bengal, and Assam as well as older skins borrowed with the courteous cooperation of the authorities concerned, from the U. S. Na- tional Museum and the American Museum of Natural History. I agree with Ticehurst (Jour. Bombay Nat. Hist. Soc. 36, 1933, p. 921) that nipalensis differs from cinereus of Java by the greater amount of white on the second outer rectrix. On the mainland there appears to be a continuous cline in color with topotypical nipalensis from Nepal being somewhat paler on the back than birds from Bengal or Burma. Birds from northern Assam have a much reduced area of white on the second outer rectrix, comparable in this respect to cinereus, but they differ from that form in being more suffused with grayish-smokey on the flanks. I, therefore, propose: Parus major vauriei subsp. nov. Type: ¢ ad. (Peabody Mus. Nat. Hist. Yale No. 9334), collected December 21, 1946, by S. Dillon Ripley at Chabua, Northeastern Assam. Diagnosis: from nipalensis this race differs by having re- duced white patches on the second outer rectrices and a darker smokier wash on the flanks. From cinereus this race differs by the smokier wash on the under surface. From ambiguus this race differs by having somewhat more white on the second outer rectrix and by being darker, smokier on the flanks. From decolorans it differs by smaller size. Measurements of type: wing 59, tail 53.5, culmen 10.5, white area on second outer rectrix (measured on inner web) 9mm. Range: Northeastern Assam. I have not been able to de- termine from fresh material the extent of the range of this form into central Assam. It is possibly another of the races which attain a climax of saturation in Lakhimpur. Mar. 10, 1950 Notes on Indian Birds III 3 Remarks: it gives me great pleasure to name this race for Dr. Charles Vaurie of New York, who has been most helpful with identifications on numerous occasions. Two new races of Sturnidae Sturnus contra sordidus subsp. nov. Type: ? ad. (S. Dillon Ripley Coll. No. 1802, deposited in Peabody Mus. Nat. Hist. Yale), collected November 21, 1946, by Salim Ali at Sadiya, Northeastern Assam. Diagnosis: from contra contra Linnaeus, described from “India” and hereby restricted to Calcutta, Bengal, this race differs by having the streaklets on the shoulders much reduced, and quite lacking on the nape. In color these streaklets are sepia rather than vinaceous or drab. This race also differs from contra in the much darker underparts which are deep vinaceous-drab in tone. The thighs also are streaked with black rather than dark brown as in the nominate race. From superciliaris and floweri this race differs as does contra, in not having the forehead streaked with white, in lacking the broad supercilium of those races, and in being very dark on the back. Measurements of type: wing 120, tail 71, bill (from skull) 32. Range: Northern Assam from Dibrugarh and Margherita north to the foothills around the Brahmaputra gorges and east through the Lohit Valley. North Cachar birds are some- what intermediate, showing a cline in coloration between contra and sordidus, but are better placed in contra. Remarks: Whistler (Jour. Bombay Nat. Hist. Soc. 36, 1933, p. 725) expresses doubt about the race dehrae Baker, set up for the more western and southern population of this star- ling in India. I agree with him after having examined freshly- collected Indian specimens from Nepal and Central India, and suggest that dehrae be made a synonym of contra. 4 Postilla Yale Peabody Museum No. 1 Acridotheres cristatellus fumidus subsp. nov. Type: ¢ ad. (S. Dillon Ripley Coll. No. 1803, deposited in Peabody Mus. Nat. Hist. Yale), collected November 21, 1946, by Salim Ali at Sadiya, Northeastern Assam. Diagnosis: from fuscus (restricted to east Bengal by Baker) this race differs by being darker, more sooty on the upper parts particularly on the rump, and darker, more smokey on the abdomen and belly. In fuscus this area is pinkish-ashy; in fwmidus the pale color is much reduced in area in most speci- mens, and darker in tone. The thigh coverts also are blackish rather than dark vinaceous-gray as in the nominate form. From grandis this race differs in the color of the bill and in the darker plumage. Range: Assam in north Cachar and north to Lakhimpur and the Mishmi Hills. Remarks: there is some variation in the extent of the color of the underparts in this form, but none are as pale on the abdomen as Indian specimens. The literature on this species is extensive and somewhat confusing. Much remains to be learned about the relationship of the species and its sibling, albocinctus in Burma. N.B.: Previous notes in this series appeared in JourNat Bomspay Narurat History Socrety 47, No. 4, August 1948, and Zootocica 33, pt. 4, December 1948. SEP eB 1950 YALE PEABODY MUSEUM oF Naturat History Number 2 March 27, 1950 New Haven, Conn. RARE MIGRATION AND WINTERING RECORDS FROM THE YUCATAN PENINSULA Raymonp A. Paynter, JR. Osborn Zoological Laboratory While collecting ornithological specimens from October, 1948 to August, 1949 in the territory of Quintana Roo, on some of the islands off the east coast of the peninsula, and for a short while in the state of Yucatan, a number of wintering and migrating North American species were taken. Some of these specimens constitute new, or rare, records for the penin- sula and cast additional light on the winter range and migra- tion routes of common North American birds. The specimens referred to hereinafter are incorporated in the collections of the Peabody Museum. A comprehensive report on my Yucatan Peninsula collection is in preparation. Numenius americanus Bechstein Long-billed Curlew The Long-billed Curlew has been recorded from Cozumel Island by Salvin (Ibis, 1889: 379) but never from the main- land of the peninsula. However, on March 31 while at Vigia 2 Postilla Yale Peabody Museum No. 2 Chico, an abandoned village on Ascension Bay, Quintana Roo, a flock of five curlews was seen flying about a quarter of a mile off shore. On April 7, again at Vigia Chico, eight curlews were found on a sand-bar a short distance from the shore. They could be studied with binoculars with ease but were too far out to be collected. Catoptrophorus semipalmatus inornatus (Brewster) Western Willet The only example of this species seen was a male which was collected in southern Quintana Roo at Xcalac on Febru- ary 3. The willet has been recorded from Yucatan by Lawrence (Ann. Lye. N. Y., 9: 210, 1878) and on Cozumel Island by Salvin (Ibid.: 379). Ridgway (Bull. U. S. Nat. Mus., 50 (8): 317, 1919) has listed these records under C. s. semipalmatus but apparently he did not examine the specimens and merely assumed only the eastern race occurred on the peninsula. It seems probable that both races do occur there but, until the old specimens can be examined, inornatus is the only race definitely recorded. Steganopus tricolor Vieillot Wilson’s Phalarope Wilson’s Phalarope has never been recorded from the Yuca- tan Peninsula. There are numerous records from central Mexico and the migration of the species through the lower portion of the peninsula is not unexpected. The only specimen seen was a lone male which was taken on May 19 at Laguna Chacanbacab (also called Laguna Alton), a large shallow body of water in Quintana Roo near the border of Campeche at the Mar. 27, 1950 Yucatan Records 3 base of the peninsula. The date is unusually late for a bird so far south (see, ¢.g., Bent, Bull. U. S. Nat. Mus., 142: 28, 1927) but the specimen was exceedingly thin and may have been unable to migrate farther north. Caprimulgus carolinensis Gmelin Chuck-will’s-widow On April 5, at Cayo Culebra, Ascension Bay, a single speci- men of this species was secured. No others were seen or heard. It has never been found on the peninsula or nearby islands and presumably does not winter there. During the first week of April the arrival of a wave of migrants composed of a number of diurnal species was noted and presumably this species was among them. Tyrannus tyrannus (Linnaeus) Kingbird On April 3 we began field work on Cayo Culebra and secured two specimens of this species from the great number which was present. The Kingbird was not found on the mainland previous to this date but, upon our return on April 7, it was seen quite frequently during the months of April and May whenever we were in rather open country. Both Salvin (Ibid.: 362) and Boucard (Proc. Zool. Soc. Lond., 1883: 448) reported the species abundant in northern Yucatan in the same months. yy Bombycilla cedroruwm Vieillot Cedar Waxwing A very rare species on the mainland of the peninsula. A single waxwing was seen feeding in the sea-wrack at Xcalac on February 2 and another bird was seen and collected at 4 Postilla Yale Peabody Museum No. 2 Tabi (an Indian village about twenty miles northwest of Carrillo Puerto, Quintana Roo) on March 12. The only pre- vious mainland record was a bird which was taken at Izalam, Yucatan in February, 1879 (Boucard, [bid.: 442). Griscom (Amer. Mus. Novit. No. 236: 4, 1926) found the species abundant on Cayo Centro, Chinchorro Bank in January, 1926, but it was not present during my field work in February, 1949. V Vireo virescens virescens Vieillot Red-eyed Vireo One specimen was taken at Carrillo Puerto on April 8 and another at Chetumal, Quintana Roo on April 14. These were the only ones seen. The Red-eyed Vireo apparently passes through the Yucatan Peninsula in early April and has moved north by the time Vireo v. flavoviridis returns in large numbers in mid-April. Boucard (Ibid.: 441) records a specimen taken at Silam, Yucatan, in November by Gaumer but he notes, “No specimens sent to me.” It would seem that the record is an error since no one has collected this vireo during the fall or winter months in spite of extensive collecting. The only pre- vious record was a bird heard singing on April 3 by Cole at Chichen Itza (Bull. Mus. Comp. Zool., L (5): 136, 1906). Limnothlypis swainsonii (Audubon) Swainson’s Warbler A specimen taken on February 12, forty-six kilometers west of Chetumal, is the third record of this species for the Yucatan Peninsula. Although rare, it seems to occur regularly in the lower portion of the peninsula where little collecting has been done. Mar. 27, 1950 Yucatan Records 5 Helmitheros vermivorus (Gmelin) Worm-eating Warbler The Worm-eating Warbler has been recorded twice before from the mainland and once from Cozumel Island. The fourth record for the region was secured forty-six kilometers west of Chetumal on February 17. Vermivora pinus (Linnaeus) Blue-winged Warbler Boucard (Ibid.: 440) reported the only previous record of this species which presumably was collected in Yucatan. On March 12, while at Tabi, a single specimen was seen and collected. Vermivora peregrina (Wilson) Tennessee Warbler A male was collected on Cayo Culebra, Ascension Bay, on April 6. The few existing records for this species are from Cozumel Island. It has not yet been found on the mainland. Dendroica tigrina (Gmelin) Cape May Warbler This warbler regularly winters in the West Indies. The only previous Mexican records are both from the Yucatan Penin- sula. Boucard (Ibid.: 440) reported a bird in Yucatan and Peters collected one in Quintana Roo (Auk, X XX: 387, 1913). I secured one on Cayo Norte, Chinchorro Bank on February 4. Its presence on Chinchorro is not surprising since these small islands have great West Indian affinities, as does Cozumel Island to the north. 6 Postilla Yale Peabody Museum No. 2 Dendroica caerulescens caerulescens (Gmelin) Black-throated Blue Warbler One specimen taken on Cozumel Island on January 12. This is the second record of this species for the island, again indicating the island’s affinities with the Antilles. Dendroica castanea (Wilson) Bay-breasted Warbler One of the most interesting discoveries resulting from the work on the Yucatan Peninsula was the presence of the Bay- breasted Warbler in great numbers in early May. The only other Mexican record I have been able to discover was a bird taken at Tehuantepec (Lawrence, Bull. U. S. Nat. Mus. 4: 15, 1876). Unfortunately, I was present in Chetumal only two days during the migration of the species and my data is not so abundant as might be desired. Previous to the days in Chetumal I was in the city of Merida, and after that time in the rainforests in the middle of the peninsula. In neither local- ity did I see the species. However, on May 6 two specimens were taken at different localities in the low second growth out- side the town of Chetumal. On the next day one specimen was taken, on a short trip to the outskirts of the town, and five more were seen mixed in with a flock of yellow warblers (Den- droica aestiva) and Magnolia Warblers (Dendroica magnolia). It would appear then, from the lack of records from central Mexico, that the species migrates through Central America and up through the Yucatan Peninsula and then across the Gulf of Mexico. Mar. 27, 1950 Yucatan Records a Piranga olivacea (Gmelin) Scarlet Tanager The Scarlet Tanager appears to be another species which migrates through the peninsula in the manner of the Bay- breasted Warbler. The first specimen seen was collected on April 1 outside of Chetumal. Field work took me to regions less suitable for the species during the month but on May 5 another specimen was taken at Chetumal. There are only a few other records for the peninsula although Boucard (Jbid.: 443) states it is common near Merida, presumably in the spring. Lesa LIBRARY SEP 22 195i HARVARD MVERSTY YALE PEABODY MUSEUM or Naturat History Number 3 April 3, 1950 New Haven, Conn. A SMALL COLLECTION OF BIRDS FROM ARGENTINE TIERRA DEL FUEGO S. Ditton Rieiey In the spruig of 1948, Mr. Stephen Sanford generously suggested that a representative of the Peabody Museum of Yale accompany his party on a brief visit to Tierra del Fuego for collecting purposes. Unfortunately the group was some- what delayed in arrival due to difficulties with weather and planes, but in mid-April they arrived at Rio Grande on the east coast. Mr. Sanford was accompanied by his wife, Mr. Van Campen Heilner and Mr. Migdalski. The first part of their stay was at Estancia “Sara”? near Rio Grande where the country was very open and treeless, but Mr. Migdalski was later able to go farther north to the Estancia of the Bridges family where there were trees in some number. During his actual collecting time April 27th to May 20th, Mr. Migdalski was able to collect 78 specimens of 35 species, a difficult assignment due to the severe autumnal weather and many of the species having already migrated north. One advantage, however, was the fact that all the birds were in fresh plumage, although an additional difficulty was the heavy deposit of fat which all specimens carried. Some species showed definite gonadal enlargement, presumably correlated with the migratory season, and heavy fat deposition. 2 Postilla Yale Peabody Museum No. 3 The grateful thanks of the Museum must go to Mr. and Mrs. Sanford for assisting so generously in securing this valuable material, as well as to the authorities of the American Museum of Natural History and the Museum of Comparative Zoology for permission to examine specimens in their care. Color notes are by Mr. Migdalski. List of the Species Phalacrocoracidae Phalacrocorax brasilianus brasilianus (Gmelin): Brazilian Cormorant. A female cormorant was shot on May 16. Mr. Migdalski has noted the soft parts as: “iris brown; eyelids yellow; bill, upper mandible dorsally brown, laterally greenish-yellow, lower mandible greenish-yellow, skin of throat yellow; legs black.” This bird is in immature plumage. It measures: wing 226 mm., tail 149 mm., culmen 44 mm. Ardeidae Nycticorax nycticorax obscurus Bonaparte: Chilean Night Heron. A male taken May 20 is of this dark-bellied form. Soft parts: ‘iris red; ocular skin greenish-yellow; bill, upper mandible brownish-black, yellowish-green at the base, lower mandible distally for 1/3 of the length brownish-black, remainder yel- lowish-green; legs, suffrago, posterior tibio-tarsus, anterior tibio-tarsus greenish-black; feet black, pads yellow.” Wing: 349. Threskiornithidae Theristicus caudatus melanopis (Gmelin): Black-faced Ibis. A single female of the ibis of Tierra del Fuego was collected Apr. 3, 1950 Birds from Argentine Tierra del Fuego 3 April 29. Soft parts: “iris red; bill and facial skin black; legs purplish-red.” Anatidae Cygnus melancoryphus (Molina): Black-necked Swan. A male Black-neck was shot on April 30. It has a wing measurement of 436. Soft parts: “iris brown; bill, upper mandible dark slate, tip fleshy gray; facial skin red; legs and webs fleshy.” Coscoroba coscoroba (Molina): Coscoroba Swan. A male Coscoroba taken May 4 is in partial immature plum- age. A few feathers on the crown and a considerable area of the upper back, a few scapulars, tertials, median wing coverts and upper tail coverts are tipped with brown. The primaries are blackish terminally. This immature plumage is very swan- like. Soft parts: “iris light brown, bill reddish purple, legs and webs pinkish-flesh.” Wing: 145. Not common. Found only on inland ponds. Lophonetta specularoides specularoides (King): Crested Duck. The most abundant duck at this season. Three males were skinned out of a good number shot on April 28 and 29. Two males are in very worn plumage. One bird is in extremely fresh plumage indicating the closeness of the moult. Soft parts: “iris red (2), light brown (1); bill, upper mandible bluish-black, lower pinkish-orange; legs and webs brownish- gray.” Wing: 257.5, 259 (worn), 2765. Chloéphaga poliocephala Sclater: Ashy-headed Goose. A male Ashy-head weighing five pounds was taken on May 1. Normally these birds leave about the seventh of April, and Mr. Migdalski consequently found this the least common of t Postilla Yale Peabody Museum No. 3 the geese. It is in fact the least common species in the Rio Grande area. Soft parts: “iris dark brown; bill black; tibio- tarsus orange with black mottling; feet, middle toe and webs black, inner and outer toes black with orange sides.” Wing: 363. This bird is in worn plumage although the tail and wing feathers have been largely freshly moulted in. Chloéphaga rubidiceps Sclater: Ruddy-headed Goose. A pair and a young female were taken April 27 and 29. This, the second most common species in the area, is said to leave normally by April 15. They return in early September and usually begin to breed about October 24. The clutch size ranges from 4 to 11 eggs. Birds are flocking for migration by about April 10. One adult weighed 4 pounds 8 ounces. These birds are in wing moult. They measure: wing ¢ 325, 2 314, im. 2 329. The adult female seems to be growing a new set of primaries, while the males’ are being partially shed. The tarsus of these specimens measures: 66.5, 64, im. 65. Chloéphaga picta picta (Gmelin) : Upland Goose, Caiquen. This is the most numerous species in the San Sebastian region. An adult and an immature male and an adult female were collected on May 1 and 2. These are the Barred Upland type, the predominant type in this area. Mr. Migdalski failed to find the white males during his stay. Local information claimed that most males were of the barred type, and that the pure white birds were a color phase. This species arrives in the area in early September and seldom begins breeding before October 17. The average clutch is seven eggs which take the normal incubation period. By the end of February most of the young can fly. Migration begins the last week in April and by May 20 nearly all birds have left. Those few that stay over the winter usually suffer from frozen feet and become very thin. Local estimates report that seven to ten geese eat as much grass as one sheep. Apr. 3, 1950 Birds from Argentine Tierra del Fuego 5 Geese are said actually to be drawn by the sheep, as they prefer cropped grass. Thus the density of the geese is greater in the sheep grazing areas. Where they are heavily concen- trated the amount of goose excrement is said to become so great that sheep are driven away. Consequently for many years now the upland geese have been a pest and a bounty is paid for their destruction, at the rate of 4 cents (Argen- tine) an egg, 15 cents per gosling and 50 cents for an adult bird. In 1947-8 alone the Estancia “Sara” staff at Rio Grande broke 35,000 eggs, and another nearby estancia broke 75,000 eggs. The cook and her husband on one estancia during that season collected 4,000 eggs, 900 young and 400 adults for the bounty. These specimens measure: wing ¢ 427 (worn), 2 892; tail 6 177, 2 156; culmen ¢ 33, 2 32; tarsus 3 84, 2 77. Weight ¢ 6 pounds 4 ounces, 2 6 pounds. Chloéphaga hybrida hybrida (Molina): Kelp Goose. A pair of Kelp Geese were collected at Viamonte on May 11 and 15. They were wary and difficult to approach. Migdalski never saw them farther than 60 yards from the shore, once or twice on sheltered small ponds near the beach. The “Kelpers” as they are called locally, arrive in March and leave in September in contrast to the other species. Soft parts: “iris dark brown, bill black, legs, feet and webs bright yellow; iris brown, bill fleshy, legs, feet and webs bright yellow.” Wing: ¢ 381, 2 344. Both birds are freshly moulted and weigh: ¢ 5 pounds 12 ounces; 2 4 pounds 8 ounces. Tachyeres patachonicus (King): Flying Steamer Duck. Two females were taken May 1 and 4. One is subadult and is in moult. The reddish throat patch is prominent in this bird. Wing: ad. 296, subad. 260. Soft parts: “iris dark brown; bill (ad) greenish-yellow basally, distal half bluish- 6 Postilla Yale Peabody Museum No. 3 black, (im) yellowish-green, tip black; legs and feet yellow, webs grayish-black.” The flightless species does not occur so far north. Anas versicolor fretensis King: Southern Gray Teal. A single male, taken April 29, measures: wing 219. Soft parts: “iris dark brown, bill, upper mandible black, basal half excluding culmen yellow, lower mandible bluish-gray; legs and feet greenish-gray, webs black.” Anas georgica spinicauda Vieillot: Brown Pintail. A female was collected April 27. The tail is worn but the wings are freshly moulted and measure: 229. The bird was very fat and greasy. Soft parts: “iris dark brown; bill, upper mandible yellow, culmen and tip black, lower mandible yellow, tip black; legs and toes gray, webs black.” Anas flavirostris flavirostris Vieillot: Yellow-billed Teal. With the Crested Duck, the commonest species at Rio Grande. Birds were taken the end of April, all fat, freshly moulted and in good condition. Wing: ¢ 179 (m), 2 191, 192. Soft parts: “iris dark brown; bill, upper mandible yellow, culmen and tip black; lower mandible yellow, tip black; legs and feet gray, webs black.” Anas sibilatriz Poeppig: Chiloe Widgeon. A single male, freshly moulted, was collected April 29. Accipitridae Buteo polyosoma polyosoma (Quoy and Gaimard): Rufous- backed Buzzard. Two varicolored immature birds, presumably a male and female, were secured on May 2 and 15. Soft parts: “iris Apr. 3, 1950 Birds from Argentine Tierra del Fuego fi creamy-brown, yellowish-brown; bill greenish-gray, black tip to upper mandible; cere greenish-yellow; legs yellow.” Wings: $ 2? 440, 2 ? 500. Falconidae Milvago chimango chimango (Vieillot) : Chimango. Three females were collected at Rio Grande in late April. They are tame confiding birds. Soft parts: “iris dark brown; bill horny-gray, cere fleshy ; legs bluish-gray.” Wing: 283-304. Polyborus plancus plancus (J. F. Miller): Carancho or Caracara. A male was collected at Rio Grande May 4. Soft parts: “iris light brown; bill creamy-white, touch of light blue at the base; cere orange; ocular skin orange; legs yellow.” Wing: 415. Falco sparverius cinnamominus Swainson: Chilean Kestrel. A female with a wing measurement of 205 was taken May 2. Haematopodidae Haematopus leucopodus Garnot: Magellan Oyster-catcher. A pair were found at Viamonte May 3 and 16. They measure: wing 6 261, 2 256. Both are rather worn. Soft parts: “iris orange red; eyelids yellowish-orange; bill, upper mandible dark brown, reddish-orange at base, lower mandible basally orange red, distally dark brown; light grayish with fleshy touches.” Haematopus ater Vieillot and Oudart: Quoy’s Black Oyster- catcher. A male taken May 3 has a wing measurement of 271. Soft 8 Postilla Yale Peabody Museum No. 3 parts: “iris orange red; eyelids orange red; bill red; legs light creamy-flesh.” Thinocoridae Attagis malouinus (Boddaert): White-bellied Seed Snipe. Two males, a female and a specimen in alcohol were taken in late April and May. These birds are in fresh plumage, wing ¢ 165, 175, 9172. Soft parts: “iris brown; bill dark horn; legs creamy gray.” Charadriidae Oreopholus ruficollis (Wagler): Slender-billed Dotterel. A male with a wing measurement of 167 was collected May 3. This is a very late date for this Dotterel. No other shore birds were taken or seen. This bird was extremely fat. Laridae Larus marinus dominicanus Lichtenstein: Kelp Gull. A single male of April 27 has soft parts: “iris brownish-gray ; eyelids pinkish-orange; bill yellow, distal third of lower mandible reddish-orange; legs greenish-cream.” Strigidae Bubo virginianus nacurutu (Vieillot): Magellan Horned Owl. A female with a wing of 350 was shot at Viamonte May 11. Soft parts: “iris yellow; bill and cere black.” Picidae Campephilus magellanicus (King): Magellan Woodpecker. A pair were taken in a patch of forest fifteen miles inland from Viamonte on May 19. The female has slightly enlarged Apr. 3, 1950 Birds from Argentine Tierra del Fuego 9 ovaries. Soft parts: “iris orange; bill and legs grayish-brown.” Wang) 6) 221.5; 2216. Furnariidae Aphrastura spinicauda spinicauda (Gmelin): Thorn-tailed Creeper. Common in the trees at Viamonte. Soft parts: “iris brown; bill, upper mandible black, lower whitish-flesh, tip black; legs greenish-brown, pads greenish-yellow.” Wing: 4 64, 65, 2 60-62. This race is more pure brown, less rufous than specimens from Southern Chile, which should be separated as twpinieri. Pygarrhicus albogularis (King): White-throated Tree-Runner. Another common spiney-tail about Viamonte. Soft parts: “iris brown, dark brown; bill, upper mandible black, lower grayish-white, dark tip (¢), light grayish-white, dark tip (2°); legs brown.” Wing: 4 85-88, 2 83-85. Tyrannidae Aolmis pyrope (Kittlitz): Fire-eyed Pepoaza. Two males taken May 13 and 14 were the only tyrannids seen. Soft parts: “iris red.” Wing: 120, 122.5. Muscicapidae Turdus falcklandti magellanicus King: Magellan Robin. Some of these thrushes, taken between May 11 and 15, are 10 Postilla Yale Peabody Museum \ No. 8 coming into breeding condition. All specimens are very fat and freshly moulted. Wing: 6 187.5-144, 2 1381, 132. Fringillidae Phrygilus unicolor ultimus, subsp. nov: Tierra del Fuego Plumbeous Finch. Type: ¢ ad. (Yale Peabody Museum No. 9335.) collected at Viamonte, Rio Grande, Tierra del Fuego, Argentina, on May 16, 1948, by E. C. Migdalski. Description: similar to wnicolor (type locality Tacna, restricted by Hellmayr) but larger, the adult females with somewhat darker, more blackish streaks. From tucwmanus and inca this race differs as does unicolor. From grandis these birds differ by being lighter gray in color in the male, and in the females by being lighter, less rufescent on the back, with lighter streaking. Measurements: wing 6 98.5, 2 94.5, 95: tail ¢ 72.5, ? 65, 66: culmen (from skull) ¢ 13, 2 12.5 (2). A series of eleven birds from Chile and northern Argentina measure: wing 6 88-92.5, 2 81-88; tail ¢ 60-68.5, 2 59-61.5; culmen ¢ 1-12, 12-71 T-11.5. Remarks: These birds are in fresh plumage. Soft parts: ‘iris brown; bill, upper mandible dark horn, lower light horn; legs brown.” Zonotrichia capensis australis (Latham) : Patagonian Sparrow, Chingolo. Two males were collected in mid-May. Both birds have con- siderable streaking on the crown. Wing: 78, 84. Icteridae Notiopsar curaeus (Molina): Chilean Blackbird. Not uncommon about Viamonte in mid-May. One male had slightly enlarged testes. Wing: ¢ 137, 1388, 2 129, 1382. Apr. 3, 1950 Birds from Argentine Tierra del Fuego 11 Trupialis militaris militaris (Linnaeus): Red-breasted Star- ling, Pecho Colorado. Common at Rio Grande in April and May. Soft parts: “iris brown; bill, upper mandible dark horn, lower light grayish- horn; legs slate gray (some with a brownish tint).” YALE PEABODY MUSEUM oF Naturau History Number 4 May 22, 1950 New Haven, Conn. A NEW TANAGER FROM MEXICO Raymonp A. Paynter, JR. Osborn Zoological Laboratory Piranga roseo-gularis tincta subsp. nov. Type: ¢ ad. (No. 9153, Peabody Mus. Nat. Hist., New Haven), collected at Chetumal, Territory of Quintana Roo, Mexico, Nov. 12, 1948, by Raymond A. Paynter, Jr. Diagnosis: Compared with typical roseo-gularis the male differs in being more saturated gray above, grayer on the breast, and washed with buffy on the abdomen. Compared with cozumelae it is equally dark above but with a reddish tinge, slightly deeper pink on the throat, and faintly streaked with pink on the breast and upper abdomen. The tail is shorter than in either race. The female of this form is more buffy below than roseo- gularis and lighter on the pileum than typical cozwmelae. Immature specimens of both sexes are highly variable and are inseparable from roseo-gularis and cozumelae. Measurements: Ridgway (Bull. U. S. Nat. Mus. 50 (2):99) states that cozumelae has a longer tail and a shorter wing than roseo- gularis. However, in my series of specimens, the difference between the mean tail length of the males of tincta and the means of the other two races is the only statistically significant interracial differ- ence in external measurements. P<.02. Adult Males Adult Females Standard] No. Standard error of | Speci- error of Character Race tincta roseo-gularis cozumelae 2 Postilla Yale Peabody Museum No. 4. Three adult males of tincta collected in Chetumal during Novem- ber and December had a mean weight of 24.83.21 grams, whereas two males of the same race from Carrillo Puerto taken in March weighed 22.90+.40 grams. The difference between the two means is statistically significant. P<.02. A female collected in January at Chetumal weighed 22.00 grams. The only weights available for roseo-gularis are two males taken at Xcan and Kantunil Kin in April which weighed 23.40 and 22.60 grams, and two females from Tabi and Kantunil] Kin, taken in March and April, which weighed 23.20 and 20.70 grams. A male and female of cozumelae, taken in January, weighed 22.30 and 22.90 grams respectively. Range: The more humid regions of the central and southern portions of the Yucatan Peninsula including Campeche and Quintana Roo. The range of roseo-gularis should be amended to include only the more arid portion of the peninsula, roughly consisting of the entire state of Yucatan and the northern tip of Quintana Roo. Material Examined: 54 specimens of all races. roseo-gularis: 26 specimens from Chichen Itza, Yuc., Xocempich, Yuc., Xbac, Yuc., “Yucatan,” Kantunil Kin, Q. Roo, Xcan, Q. Roo, and Tabi, Q. Roo. tincta: 19 specimens from Chetumal, Q. Roo, Carrillo Puerto, Q. Roo, Acomal, Q. Roo, Palmul, Q. Roo, Chunyaxche, Q. Roo, and Pacaitun, Camp. cozumelae: 9 specimens from Cozumel, Q. Roo. Discussion: The specimens from Pacaitun, Campeche, approach roseo-gularis clinally in the coloration of the back but they have been placed with tincta because of their shorter tails. The variations in the weights of the males of tincta are extremely interesting. Greater weight is possibly an additional racial character of tincta. From the few available data the birds from Chetumal are definitely more heavy than those from Carrillo Puerto. It may be that Chetumal is an area where the racial characters are most strong and Pacaitun, Carrillo Puerto, etc., are areas which show a gradation toward roseo- gularis. Variations in color and tail length seem to support this hypothesis. Weight variation may be seasonal but exam- ination of the gonades indicates that it is not correlated with the development of these organs. I wish to thank the authorities of the Museum of Compara- tive Zoology, the Chicago Natural Museum, and the American Museum of Natural History for permitting me to examine material in their care. hatelle LE PEABODY MUSEUM or Natura History may 1/1951 wravera «= yMIseTTY Number 5 December 29, 1950 New Haven, Conn. A LARGE PYCNODONT FROM THE NIOBRARA CHALK JosEPH T. GREGORY Among the fossils collected on the Yale Scientific Expedi- tion of 1872 are fragments of the skull and dentition of a large pycnodont fish. These were found by O. C. Marsh in the Cretaceous chalk exposures along the Smoky Hill River in Kansas on November 6, 1872. They are of particular interest as an example of extreme reduction of the dentition in an aberrant member of this family of durophagous fishes, and also because of their unusually large size. The specimens confirm the distinctness of a genus described by Leidy from the Cretaceous of Mississippi. It is quite fitting that the species should be named for their discoverer. CLASS PISCES (OSTEICHTHYES) Order Pycnodontoidea FAMILY PYCNODONTIDAE Hadrodus marshi new species Type: Premaxillary, left and part of right splenials, and frag- ments of skull roof of one individual, Y.P.M. Catalogue of Vertebrate Palentology, no. 1950. Type locality: “South side Smoky Hill River, 2 miles east of North Fork.” This places it in Logan Co., Kansas, about five miles west of Russell Springs. 2 Postilla Yale Peabody Museum No. 5 Formation and age: Probably upper Niobrara Chalk, early Senonian. Diagnosis: Two-thirds the size of Hadrodus priscus Leidy, premaxillaries shorter and much higher than in that species and not excavated anteriorly; anterior prehensile tooth smaller than posterior. Splenials with 4 rows of irregularly oval teeth, some of which bear apical cusps; teeth of the lateral row slightly larger than the others; 4 to 5 teeth in each row. DESCRIPTION Premaczillary: A left premaxillary lacking the dorsal ex- tremity is tall and short, of fairly stout proportions, more similar to Gyrodus (Hennig, 1906, pl. X) than to such forms as Proscinetes |Microdon]|. There is no trace of a horizontal process along the border of the mouth. It is about twice the size of that of the large specimen of Gyrodus circularis Agassiz figured by Hennig. The median surface is straight and bears throughout its length a suture for the opposite premaxillary ; these bones must have been closely united throughout their length, in a normal fashion, not diverging as they have been restored in Gyrodus (Hennig, 1906, p. 148, pl. X). On its posterior margin is a large oval, vertically elongate depression, its upper end merging with the lateral surface of the bone. Two large, bicuspid, prehensile teeth are ankylosed to the oral margin. These differ from those of H. priscus figured by Leidy (1873, pl. 19, figs. 17-20) in somewhat greater disparity in size, and in the less distinct groove separating the cusps on the outer surface of the crown. They lack the posterior concavity characteristic of most pycnodont “incisors.” The premaxillary bone itself differs markedly from Leidy’s figure in its relatively greater height—which may be due to incompleteness of that specimen—and in the absence of an excavation in the anterior border. Leidy (1873, p. 294) interpreted the excavations above the roots of the teeth as spaces for developing replace- ment teeth. In the opinion of most students of the pyenodonts Dec. 29, 1950 A Large Pycnodont 3 (cf. Woodward, 1895, p. 194) there was no tooth replacement. Neither the form of the cavity in the premaxillary of Hadrodus nor its remoteness from the dentigerous border suggests that it was an alveolus; however, Saint-Seine (1949, p. 121) has observed unworn replacement teeth in just this position in Proscinetes |Microdon| sauvanausi Thiollitre. Hence Leidy’s inference may be correct, although the mechanism of replace- ment and ankylosis of the teeth to the premaxillary remains an enigma. A more plausible interpretation is offered by Smith- Woodward (1895, p. 193) who suggests that the excavation lodged the nasal capsule. Measurements of the Premaxillary mm. Maximum length, anteroposterior ..................... 26.0 Height as preserved, including” teeth: 2222. 5...../2s6a.- 66.0 Wenrthwotwhirst) too bly s..).smierscsraescie ns siciaseciaete mies cles 10.1 VACUA O PCTS Ea COOL 5 a tie etapateccetbicteisy seas 6 oe aa lio laiclt eta atone 7.3 engin or Second. LOGY 2/2). 20 ore siajc ste ee Weices 3 (ete ewe Sal ela 12.4 Wadthvofssecondstoothans adc ae eects ere eiacicteinie 8.7 B Figure 1. Hadrodus marshi, n. sp. Type specimen, Y.P.M. 1950. A. Medial view of premaxillary showing interpremaxillary suture and pocket for olfactory capsule in posterior border. B. Lateral aspect of premaxillary. Kees 4 Postilla Yale Peabody Museum No. 5 Splenials: A large left splenial with extremely deep anterior symphysis and only moderate coronoid process, and the pos- terior part of its left homologue show that the lower jaw of this genus differed in detail from other pycnodonts. It is only slightly longer than that of G. circularis, but much deeper, especially anterior to the front teeth where it reaches its maximum depth. The symphysial area is short, deep, with a straight posterior boundary. Its lower fourth forms a nearly round facet separate from the remainder of the denticulate suture. Possibly this small area met the opposite splenial, and the coarser suture was with the dentary. If so, the latter bone was further reduced than in Mesturus (Woodward, 1895, pl. 15; Saint-Seine, 1949, p. 107, fig. 38) or Gyrodus (Hennig, 1906, pl. X and Weitzel, 1930, p. 93), but perhaps no more than in Proscinetes [Microdon] (Saint-Seine, p. 112, fig. 41). There is no indication of contact with the dentary on the lateral surface of the splenial, except possibly at the extreme front. This is a characteristic pycnodont condition and sup- ports the reference of Hadrodus to this Order in spite of considerable differences in dentition. The large size of the upper prehensile teeth suggests that lower incisors should likewise have been prominent. It is pos- sible that the dentary was larger than suggested above and the lower jaw as a whole about twice the size of that of Gyrodus circularis, with proportions similar to that species. Four rows of irregularly oval teeth with one to three cusped crowns are present. Unlike other pycnodont genera, the lateral row contains the largest teeth; they are subequal in size, about 9 mm. in their long diameters, separated by spaces of about 1 mm. The most posterior bears three cusps in a straight line along its crown; the second from the front bears an obscure single, laterally placed cusp. In the second and third rows the teeth are slightly smaller and more variable in shape. A single large tooth with two apical cusps forms the fourth, innermost row. Variability in both number and shape of the teeth is indicated by the fragment of the right splenial in Dec. 29, 1950 A Large Pycnodont 5 Figure 2. Hadrodus marshi, n. sp. Type specimen, Y.P.M. 1950. A. Left splenial, lateral aspect. B. Left splenial, medial aspect. C. Right splenial, dorsomedial aspect. x 1. 6 Postilla Yale Peabody Museum No. 5 which the posterior tooth of the lateral row, as displayed on the left side, is absent, and the teeth of the second row are nearly as large as those of the lateral row. The crowns of the teeth are smooth except for the papilla- like tubercles at the apex. In this they differ markedly from Gyrodus and are more similar to Gyronchus [Mesodon] or Proscinetes [Microdon]. There is no trace of the tendency toward transverse broadening of the teeth seen in Coelodus or Anomoeodus. Similar papillae are present on the crowns of tritoral teeth of Acrotemnus faba Agassiz. That species, how- ever, differs from Hadrodus in the much greater transverse width of its tritoral teeth, in the presence of a marked trans- verse ridge along their crowns, and in having a group of papillae adjacent to but not upon this ridge line. In Hadrodus the papillae are upon the ridge, if any is present, and tend to be oriented anteroposteriorly if more than one papilla occurs. Measurements of the Splenials mm. Anteroposterior length (reconstructed from both) ...... 120 Depthrmeiront of crushing: teeth. {22 --. eas. en eae 48 Anteroposterior length dental battery .................. 47 Roofing bones: Several fragments of thick skull bone orna- mented by closely but irregularly spaced, rounded tubercles are present. Most probably they are portions of opercular bones, although insufficient borders remain to determine their exact position. None shows traces of canals of the lateral line system. Some fragments show a lower radiating type of sculpture near the thin margins such as Hennig (1906, p. 161) describes on the preoperculum of G. circularis. The thick tu- berculated Jayer of ganoine above extremely cancellous bone is characteristic of pycnodonts; Hadrodus shows coarse tu- berculation commensurate with its large size. Such strong sculpture is found in Gyrodus and in Gyronchus [Mesodon] Dec. 29, 1950 A Large Pycnodont 7 hoeferi, the earliest appearing pycnodont; other members of the family are said by Hennig (Ibid., p. 179) to have considerably weaker tuberculation. DISCUSSION Dr. David H. Dunkle has called my attention to the re- semblance of this specimen to certain semionotids. Bifid pre- hensile teeth are characteristic of Dapediwm, whereas the incisors of pycnodonts have single crowns, concave internally. Dapedium also has crushing palatal teeth. However the shape of the premaxillary of Hadrodus differs greatly from that of Dapedium in its great vertical and short horizontal extent, and also in the absence of surface ornamentation. Conceivably the premaxillary of Hadrodus could be derived from that of the early Jurassic Dapediwm by shortening and dorsal ex- tension, but as Dapediwm had already attained a deep body and relatively high, short skull without vertical elongation of the premaxilla, it seems unlikely that such a change would have occurred. Aside from the form of the incisors, there is no reason to postulate this relationship. The form of the splenial teeth, particularly the develop- ment of one or a few papillae on the crown, is not unlike that of some species of Lepidotes such as L. mantelli Agassiz. Wide and irregular spacing of the splenial teeth, and lack of dif- ferentiation of these teeth into rows of small and large tritors are most unlike normal pycnodonts and far more like Lepidotes. Also, the deep anterior portion of the splenial is suggestive of that genus. No trace of tooth succession can be found, however, and the shape of the premaxillary bone is very unlike Lepidotes in which there is a well-developed alveolar ramus along the oral margin and a slender ascending process arising from the anterior end (Saint-Seine, 1949, p. 138, fig. 161, p. 140). Nor is there any trace of a separate coronoid bone such as occurs in the semionotids. The splenial alone forms the major portion of the lower jaw and its coronoid process, as 8 Postilla Yale Peabody Museum No. 5 in other pycnodonts. Thus the resemblances lack detail in- dicative of relationship and may reasonably be ascribed to convergence. Only one other species of pycnodont is known from the Niobrara formation, Micropycnodon kansensis (Hibbard and Graffham). This is a small fish, scarcely one-third the size of Hadrodus marshi, which may readily be distinguished by its more typically pycnodontid splenial dentition, with two rows of small teeth lateral and one row internal to the principal row of enlarged crushing teeth. Coelodus streckert Hibbard from the underlying Carlisle shale of Kansas is also of Turonian age. Pycnodonts are more numerous in the lower Cretaceous of the Gulf of Mexico embayment, several genera having been reported (Williston, 1900, Gidley, 1913). The type locality and horizon of Hadrodus priscus Leidy are uncertain; Columbus, Mississippi, is on the Eutaw forma- tion but only a short distance from the base of the Selma Chalk. The horizon may well be equivalent to the Niobrara and close to that of H. marshi. Whether the characters here used to seperate these species are valid remains to be deter- mined by future discoveries of more complete material from these and other localities. Differences in the form of the pre- maxillary seem sufficient for specific distinction of the two forms. Although it is difficult to estimate the size of the fish from such fragments as are available, especially when the pro- portions of the genus are not accurately known, it seems worthwhile to point out that Hadrodus may well have been the largest of the pycnodonts. If its proportions were similar to those of Gyrodus circularis, it may have exceeded a meter in length. The premaxillary is twice the size of that of a large specimen of G. circularis described by Hennig, and the splenial exceeds those of that species by 30 to 50 per cent. Hadrodus shows the most reduced and specialized dentition thus far known among the pycnodonts. Reduction in number of Dec. 29, 1950 A Large Pycnodont 8) teeth, increase in size of the external row and corresponding decrease in importance of the next to mnermost row of the splenial teeth, and development of a diastema between teeth of the dentary and splenial are all divergent from the general trend of pycnodont evolution, and separate Hadrodus sharply from all other described genera. Closest resemblances, in denti- tion, appear to be with Gyronchus [Mesodon] and certain species of Proscinetes [Microdon], in which the crushing teeth are irregular in size and distribution. It is interesting to note that the dental evolution of the pycnodont line leading from Gyronchus to Hadrodus parallels that of the placodont rep- tiles from Paraplacodus through Placodus and Cyamodus to Henodus (von Huene, 1936). Four main types of dentition have evolved among the pycno- donts. Eomesodon, the earliest form, and Gyronchus |Mesodon] have smooth crowned crushing teeth arranged in irregular rows and uneven in size. In Mesturus and Proscinetes | Micro- don] the teeth attain regular arrangement in longitudinal rows; their crowns are smooth or with apical pits. This type of dentition persists into the Eocene Pycnodus. Gyrodus has similar rows of teeth, but the crowns are ornamented with concentric rings of mamillary papillae. Coelodus shows diver- gence in the transverse broadening of the enlarged teeth, beginnings of which may be observed in some species of Prosci- netes. Anomoeodus may represent a further development of this line, with degeneration of the lateral rows of teeth. Finally, Hadrodus has greatly reduced the number of teeth and shifted emphasis from internal to external rows. It will be most in- teresting to discover the vomerine dentition which accompanied this modification. ACKNOWLEDGMENTS I am indebted to Dr. David H. Dunkle of the United States National Museum for critical comments and advice. The il- lustrations were prepared by Miss Shirley Glaser. 10 Postilla Yale Peabody Museum No. 5 REFERENCES Dunkle, D. H. and Hibbard, C. W. 1946. Some comments upon the structure of a pycnodontid fish from the Upper Cretaceous of Kansas. Univ. Kan. Sci. Bull., vol. 31, pt. I, pp. 161-181, 3 pls. Gidley, J. W. 1913 Some new american pycnodont fishes. Proc. U. S. Nat. Museum, vol. 46, pp. 445-449. Hennig, E. 1906. Gyrodus und die Organisation der Pyknodonten. Pa- laeontographica, Bd. 53, pp. 137-208, pls. 10-13. Hibbard, C. W. 1939. A new pycnodont fish from the Upper Cretaceous of Russell County, Kansas. Univ. Kan. Sci. Bull., vol. 26, pp. 373-375, 1 pl. Hibbard, C. W. and Graffham, A. 1941. A new pycnodont fish from the Upper Cretaceous of Rooks County, Kansas. Ibid., vol. 27, pp. 71-77, 1 pl. Leidy, Joseph. 1857. Notices of some remains of extinct fishes. Proc. Acad. Nat. Sci. Phila., 1857, pp. 167-168. 1873. Contributions to the extinct vertebrate fauna of the Western Territories. Rept. U. S. Geol. Surv. of the Territories (F. V. Hayden), vol. 1, pp. 14-358, 37 pls. Saint-Seine, P. de. 1949. Les Poissons des Calcaires lithographiques de Cerin (Ain). Nouvelles Archives du Muséum d’Histoire Naturelle de Lyon, Fasc. II, vii + 351 pp., 26 pls. Weitzel, K. 1930. Drei Reisenfische aus den Solnhofener Schiefern von Langenaltheim. Abh. Senckenbergischen Naturforschenden Gesellschaft, Bd. 42, pp. 85-113. Williston, S. W. 1900. Cretaceous fishes—selachians and pycnodonts. The University Geological Survey of Kansas, vol. 6, pp. 237-256, 1900. Woodward, A. S. 1895. Catalogue of fossil fishes in the British Museum (Natural History), part III, 544 pp., 18 pls. iy TU CoERAN 18 8" i a ache Pine why: oe ; he ¥, f v 4 { ’ { ; s At eae? 4. : ») ‘i ’ j Ar 4 } a) 7 . j ‘ i iff Pha r A — batilla oF Natura History Number 6 February 28, 1951 New Haven, Conn. NOTES ON INDIAN BIRDS IV* SOME RECENTLY COLLECTED BIRDS FROM ASSAM S. Ditton RIPLey During a collecting trip this autumn and winter in Assam’s eastern Naga Hills and Manipur, a number of interesting specimens were secured which seem to warrant preliminary description prior to further publication. The localities of all these forms are within the Naga Hills District or the Chief Commissioner’s District of Manipur (formerly Manipur State), and will be dealt with in detail in a later paper. I am most grateful to the authorities of the British Museum, the United States National Museum, and the American Museum of Natural History for allowing me to examine comparative material in their care. Arborophila torqueola interstincta, subsp. nov. Type: ¢ ad. (Yale Peabody Museum No. 12006) collected November 30, 1950, by S. Dillon Ripley on Mt. Zephu, 93 miles east of Kohima, eastern Naga Hills, Assam. Diagnosis: from torqueola of the Sikkim Himalayas this race differs by being more heavily and distinctly barred on the back and inner wing coverts in both male and female * Previous papers in this series have appeared in the Journat, Bompay Natourat History Society 47, 1948, p. 622; Zootocica 33, 1948, p. 199; and Postiiua, 1950, No. 1. 2 Postilla Yale Peabody Museum No. 6 plumage. The lower parts in both sexes are richer and darker chestnut, and richer and darker rufous-buff on the thighs and upper under tail coverts. Compared to batemani of Mt. Victoria and the Chin Hills, this form differs as does torqueola, lacking the greater degree of chestnut on the sides of the neck and the greater area of chestnut on the scapulars. From griseata of Tonkin this popu- lation differs in having richer, darker chestnut streaking on the flanks with more pronounced and distinct white drops on the centers of the feathers, and with the black central patch on the feathers of the under tail coverts much reduced in extent. Measurements (mm.): wing tail culmen (from skull) 4d ¢ 144.5-156 53-62 19-21 2 150 56 19 Range: Upper Chindwin River drainage area in eastern Naga Hills of Assam and Burma. Indicator xanthonotus fulvus, subsp. nov. Type: ¢ ad. (Yale Peabody Museum No. 12001) collected December 11, 1950, by S. Dillon Ripley at Pfutsero, eastern Naga Hills, Assam. Diagnosis: from zanthonotus of the Himalayan Range this race differs by being darker, more blackish on the upper parts and darker, more blackish on the abdomen, thighs and under tail coverts. The streaking of the abdomen though blackish, is less in extent, thus less prominent. On the forehead the golden patch extends somewhat less far back on the crown, and the edging to the feathers of the back and scapulars is reduced. Soft parts: iris brown; bill yellowish-horn, distal half of upper mandible and lower mandible brown; feet grayish-brown. Weight: 29 grams. Feb. 28, 1951 Notes on Indian Birds IV 3 Measurements (mm.): wing tail culmen se) 90 57 11 S 86 56 10 Range: Naga Hills, Margherita (?) Assam, and Myitkina District, north Burma. Remarks: An opportunity to examine the material in the British Museum showed at once the existence of a dark eastern race of the Yellow-backed Honeyguide. The only specimen from Burma is the one recorded by Smythies (Ints, 91, 1949, p. 645) with which my type agrees. An additional character of this race may be slightly smaller size, but more material would be needed. I include Margherita in the range of the form as Stuart Baker’s sight record (Fauna Bair. Inpia IV, 1927, p. 182) presumably refers to this form. My male specimen is moulting out the feathers of the nape. Both near Pfutsero, 28 miles east of Kohima at 6000 feet altitude, and on the slopes of Mt. Japvo, 10 miles southeast of Kohima at 7000 feet, we found cliffs abounding in wild bee nests, and both were investigated for Honeyguides. Birds were present, but the height of the nearby trees and the thick vegetation as well as the comparatively short duration of our stay made collecting very difficult. The Angami Naga name is “Mephi Tsu Kelie Para.” The body of the type is preserved in alcohol. Dendrocopos darjellensis fumidus, subsp. nov. Type: ¢ ad. (Yale Peabody Museum No. 12002) collected November 9, 1950, by S. Dillon Ripley on Mt. Japvo, Naga Hills, Assam. Diagnosis: from darjellensis this subspecies differs by be- ing darker, more smoky on the underparts, particularly the lower throat and breast and with a more richly colored vent patch. The nuchal patch also is darker. There appears to be a tendency in these birds to heavier streaking below, although I am not sure whether this character would hold in a large 4 Postilla Yale Peabody Museum No. 6 series. Specimens of fwmidus are smaller also than typical darjellensis, although Burmese examples, which in color repre- sent darjellensis, are also small. Measurements (mm.): wing tail culmen 3 126.5 83.5 32 22 123, 126 76 31, 32 Range: higher hills in Cachar, Naga Hills and Manipur from 5000 to 9000 feet. Spelaeornis chocolatinus nagaensis, subsp. nov. Type: 6 ad. (Yale Peabody Museum No. 12004) collected November 12, 1950, by S. Dillon Ripley on. Mt. Japvo, Naga Hills, Assam. Diagnosis: compared to chocolatinus this race is much more olivaceous-brown, less rufous above, the lores, cheeks and sides of the head grayish-brown rather than reddish-brown. The throat is white, the breast pale brownish-white, narrowly spotted, with terminal blackish edgings to the feathers. There is dimorphism in this subspecies, females being much more rufescent below, in this approaching the two existing unsexed specimens of chocolatinus although by no means matching them. The type and paratype of chocolatinus may thus both be females. In any case the fine spotting on the underparts and the white throat at once distinguish this race from the nominate form. From oatesi this race differs by being more grayish-brown about the head, and the spots and terminal edgings on the feathers of the lower surface being strikingly different, much finer and more delicate in pattern and form. From reptatus this subspecies differs by having a white throat and far more pronounced spotting below. Feb. 28, 1951 Notes on Indian Birds IV 5 Measurements (mm.): wing tail culmen 466 49-52 41-44 13-14 200 49, 52.5 42 (2) 12.5, 18 ”: 48 41.5 12.5 Range: Naga Hills. Remarks: In my review of this genus (Aux. 67, 1950, p. 390) I had no material from the Naga Hills. Thus it may be now supposed that longicaudatus must occur from the Khasia Hills southeast across southern Cachar without enter- ing the eastern Barail Range, for it apparently does not occur on Mt. Japvo. Phylloscopus fuscatus mariae, subspec. nov. Type: ¢ ad. (Yale Peabody Museum No. 12005) collected October 19, 1950, by S. Dillon Ripley at Moirang, Manipur. Diagnosis: upper parts dark olive brown, rump hair brown, a distinct supercilium varying in tone from cinnamon-rufous to ochraceous-buff, lores and ear coverts blackish-brown, cheeks and sides of throat ochraceous-buff mixed with hair brown, the latter becoming predominant on the sides of the breast. Throat and center of breast whitish to pinkish-buff. Flanks and thighs cinnamon brown, under tail coverts and under wing coverts ochraceous-buff. Outer edges of tail feathers tinted with olive green. Rictal bristles three extending nearly to end of nasal groove. Nasal hairs extending nearly to end of nasal groove. Measurements (mm.): wing tail culmen 32 3 57.5-63.5 50-55 10.5-11 9 57 43 (m.) 10 Wing formula: 2= 8. Sixth primary emarginate on the outer web. First primary exceeds the coverts by 12-14 mm. Third, fourth or fifth primary longest. 6 Postilla Yale Peabody Musewm No. 6 Soft parts: (different labels note varieties in color) iris brown; bill, upper mandible brown, dark brown, black; lower mandible basally yellow, distally brown, yellowish-brown, brown tip, yellowish horn; feet brownish-yellow, light brown, greenish-brown; pads yellow. Weight: 3 ¢ 7-8.5 grams; ?8 grams. Range: four specimens were taken at Kanglatongbi and Moirang in Manipur. Remarks: This form is closest to Phylloscopus fuscatus from which it differs in darker coloration, a richer, more ochraceous-buff tone to the cheeks, sides of the under parts, and under wing and tail coverts, a shorter bill, and, presum- ably, a slightly different wing formula, the third primary equalling or exceeding the fifth and fourth primaries. The subspecies may have been overlooked in collections due to the difficulty of identifying willow warblers in general, and foxing. The specimens collected by us were in low bushes and long grass in swampy areas, behaving rather like Phylloscopus subaffinis in this respect. They were often near cultivation. It is undoubtedly a wintering bird and should be looked for in similar situations in the northern plains of Assam. It gives me great pleasure to name this subspecies of willow warbler after my wife who worked tirelessly as a member of the recent Yale-Assam Expedition. Horeites flavolivaceous alexanderi, subsp. nov. Type: @ ad. (Yale Peabody Museum No. 12003) collected November 21, 1950, by S. Dillon Ripley on the Phek-Meluri Road, 60 miles east of Kohima, Naga Hills, Assam. Diagnosis: from flavolivaceous of the Himalayas this race differs by being much darker below, more olive buff with dark buffy breast and flanks. Compared to weberi of Mt. Victoria this race is richer, more olive buff on the breast and flanks. Compared to both preceding races, this form is darker above Feb. 28, 1951 Notes on Indian Birds IV ui and intermediate in size apparently, the bill larger than webert. From the Shan form, imtricatus, these birds may be dis- tinguished by being much darker, more buffy below and darker above. Measurements (mm.): wing tail culmen 2 g 48, 51 538.5, 54 12, 12.5 Soft parts: iris brown; bill black, base of lower mandible pinkish-horn; lower mandible yellowish, tip brown; feet flesh, pale brown. Weight: 6, 7 grams. Range: eastern Naga Hills. Remarks: This is a difficult species to collect, skulking in high grass and light second growth scrub. The call is a wren- like “‘tsick.” The area where we found this race is in the Chindwin drainage, so it may well extend into the Naga Hills of Burma. This form is named for my friend Horace Alexander, the well-known field student of Indian birds who accompanied me on part of my journey into the eastern Naga Hills. ay ata vai Out RN te ve ont at i) i W, : ah, | , nite mts ist | ; Sy TAMAR NOM ya ae i j tet : waite, syed Hit: Hh a Hal c/a iea 2 RM HR ae Bi i et \ 1 j Su nan Wie nat i A | dg” OMI OC iD Pur aty Pie WOM Ra Non 5 ye ult haat eras TANG Raye SPLAT: etnies we MOAG ban. yi Litt Oh jareeatit Heist Te, AMEE! Pes yh 7 | , a i ty ‘ { Pty Aor PF itn in (int : aU Pt Veh colby " ny . aati! eA ne we wi i ft, i ra Ae 1 AT pst one Xi erat KG Hebd 5 eegpneteg tt a 7: oe Mors y! ‘aikanjruds ¢ Git As nals ha AAR 4 HOWL 9 ie Ht ‘om BA CAD hy <4 ¥ i , ' ' t a vy ny q ey Vy Io aby 4) ef ees it att, A os bee. "F mes M4 Me AU ay f ‘ mn brig, ( ei A Rodd 5 ee SSA *. o Ve i ‘ J d ), Lik pu reel ) } i , y } : : i i. vii " rea nih : ah it: one i pe A bf Pi ue i iy Le te ‘Ais ¥, vai} ys a] ih Nai voy ‘Nai WAM dh ‘Sain Ae it le ie “Abii ‘iy Aes Var bibviacay asia, Gs in . veld! a Bd i, arth ; baie we aut aN wi a | bh BY: yt ont wih oe her Path COD, EA SR Th (tue ap) 1h my) 4 \ i i i i gel a ee ya ‘ TC in i iy) 4) , my ; i+) im iy i ial an TP Ae EL a i nary! Prorc) Seen Wes et att ranean Hit hay Ue Ls vypite: ‘hia : Wi Annie aH} Ant i i iH) nb dtd ait Y Tit AIAN SI wih, «Ni Mae INA a as ' Te a {i é i ; ; Hh) biel i} ei ry ' ’ i ‘ ‘ } Iho N¢ it) Wr hte i re vi vi si Men i ‘i . Kon ii : v ii mi i vane ust i 3 Aaa L ’ } ‘ i a Witt aril ’ | Laan? j . tN m M wh ti be aries es ay “ mei i ke hia di ra a MUS. COMP. ZOOL. LIBRARY elf UN 1 0 1952 CALLLLA HARVARD UNIVERSITY | YALE PEABODY MUSEUM |__UMVESE or NaTurAL History Number 7 May 10, 1951 New Haven, Conn. NEW PASSERINE BIRDS FROM THE INDO-CHINESE SUBREGION* H. G. Dricnan Aware of my interest in the races of the babbling thrush, Pellorneum ruficeps Swainson, Dr. Dillon Ripley has sent for my examination five specimens of this bird recently collected by him in the hill country of eastern Assam. Three of them, from the Naga Hills District, prove to agree very well with P. r. chamelum Deignan and serve to extend northeastward the range of this form, which is otherwise known from the Garo Hills, Khasi Hills, and Cachar Districts. The remain- ing two, from Manipur, are the first I have seen from that District, and are apparently sufficiently distinct from all other described populations of Assam and Burma to justify the erection of yet another subspecies, which, at Dr. Ripley’s kind invitation, is named below. For the loan of material for comparison with that in the Yale Peabody Museum and the United States National Mu- seum, I am indebted to Dr. Dean Amadon and the authorities of the American Museum of Natural History. * Published with permission of the Secretary of the Smithsonian Institution. 2 Postilla Yale Peabody Museum No. 7 Pellorneum ruficeps vocale, subsp. nov. Type: ¢ ad. (Y.P.M., No. 12007) collected at Kangla- tongbi (ca. lat. 24°59’N., long. 93°54’E.), elev. 2933 ft., Chief Commissioner’s District of Manipur (formerly Manipur State), October 19, 1950, by S. Dillon Ripley (original number 33). Diagnosis: while inseparable by characters of the under parts from P. r. chamelum (Cachar District), the new form differs from chamelum by having the forehead, crown, and nape chestnut (rather than rufous) ; the blackish-brown cen- ters to the feathers of the uppermost back obsolescent (rather than sharply defined) ; the olivaceous brown of the remaining upper parts deeper in tone. From P. r. hilarum (Pakkoku District, Burma), it differs in having the forehead, crown, and nape chestnut (rather than rufous); the blackish-brown centers to the feathers of the uppermost back more clearly defined; the olivaceous brown of the remaining upper parts much deeper in tone; the under parts more strongly washed with buff, and with the central streaks of the feathers of the breast and sides of the abdomen broader and more numerous. Range: the valley of central Manipur. Remarks: I have given detailed comparisons of the new race only with the two that occur nearest its range, one to the west and northwest of Manipur, the other to the southeast. From such more distant forms as ripleyi (Lakhimpur District south of the Brahmaputra) and stageri (Myitkyina District, Burma), it is immediately separable by its obsolescent (not well-defined) dark centers to the feathers of the uppermost back, as well as by other characters. Ef. I have for some time been aware that the population of Oligura castaneo-coronata (Burton) inhabiting Szechwan and northwestern Yunnan could not properly be combined with May 10, 1951 New Passerine Birds 3 either of the recognized races, the nominate one from the Himalayas or O. c. abadiei (Delacour and Jabouille). Again I am indebted to the authorities of the Yale Peabody Museum and the American Museum of Natural History for the loan of comparative material that enables me to define the characters of the new form. Oligura castaneo-coronata ripleyt, subsp. nov. Type: ¢ ad. (U.S.N.M., No. 296605) collected in the Likiang Mountains, Yunnan Province, China, in June 1923, by Joseph F. C. Rock (original number 584). Diagnosis: from O. c. castaneo-coronata separable by sig- nificantly greater length of wing and tail (see measurements below) and the slightly paler and brighter orange-rufous of the pileum. From O. c. abadiei distinguishable by slightly greater length of wing and tail and the distinctly paler and brighter orange- rufous (without brownish cast) of the pileum, which is, moreover, sharply defined from the olive green of the mantle, rather than insensibly intergrading with it. Measurements (mm.): O. c. castaneo-coronata (16 specimens) Wing: 45-50 (avg. 16 spec.: 47.25) Tail: 21-26 (avg. 12 spec.: 23.5) O. c. ripleyit (7 specimens) Wing: 52-57 (avg. 7 speg.: 55.4) Tail: 28-32 (avg. 7 spec.: 30.3) O. c. abadiei (4 specimens) Wing: 50-53 (avg. 4 spec.: 51.5) Tail: 27-29 (avg. 4 spec.: 28) 4 Postilla Yale Peabody Museum No. 7 Remarks: Delacour (Isis, 84, 1942, p. 515), removing this species from the genus Tesia, has established for it the new generic name Chlorotesia (there misspelled Chorotesia; but see ibid. the seventh line below), in the belief that Oligura, like Tesia, has Tesia cyaniventer Hodgson for genotype. This is, however, not the case. Oligura of Hodgson first appeared in Gray’s ZooLocicaL Misceviany, 1844, p. 82, as a nomen nudum, with mention of Oligura (Tesia) cyaniventer and O. flaviventer. The genus was first properly diagnosed in PRocEEDINGS OF THE ZooLoGicaL Society oF Lonpon, 13, 1845, p. 25, with reference to the same two species, in reverse order. In THe Genera oF Birps, /, 1849, p. [156], G. R. Gray affirms that T'[esia]. castaneo-coronata (Burton), of which Tesia flaviventer Hodgson is listed as a synonym, is the geno- type of Oligura Hodgson, as he does again in CATALOGUE OF THE GENERA AND SUBGENERA OF Birps ConTaINnEeD IN THE British Museum, 1855, p. 31 (where, for the first time, he treats Oligura as a valid genus). Thus, since 1849, Sylvia? castaneo-coronata Burton has been the genotype of Oligura Hodgson, by subsequent designation of G. R. Gray. Having shown that Oligura Hodgson is properly applied to Sylvia? castaneco-coronata Burton, I must now discuss its homonym, Oligura Riippell. The former first appeared in August 1845, the latter in “1845,” and it is not possible to prove that Hodgson’s name has priority of publication. The genotype of Riippell’s name is T'roglodytes micrurus Riippell — Sylvietta brachyura micrura (Riippell), a mere sub-species of Sylvietta brachyura Lafresnaye, the genotype of Sylvietta Lafresnaye, 1839. Riippell’s name never achieved wide currency and can almost certainly never in the future be brought into use; Hodgson’s, on the other hand, has been employed by authors for a century. In the circumstances, it seems best simply to assume that Hodgson’s name antedates Riippell’s (as I have done above), and to consign Delacour’s Chlorotesia to its synonymy. S~NA-N r Mex: ~4 anveyy 4 MUS. COMP. ZO6L. | : / / LIBRARY oslt aA JUN 1 0 1959 YALE PEABODY MUSEUM |! — PAAvars URIVERS oF Naturat History UNIVERSITY | Number 8 May 10, 1951 New Haven, Conn. BASSARISCUS IN MIOCENE FAUNAS AND “POTAMOTHERIUM LYCOPOTAMICUM COPE” JosEPH T. GrEGoRY AND THEODORE Downs INTRODUCTION Cope described a fragment of the lower jaw of a small carnivore from the “Loup Fork of Cottonwood Creek, Oregon,” as Lutrictis ? lycopotamicus (1879, p. 67). The type has been lost, but was figured (Cope - Matthew, 1915, pl. 119c, figs. 5 and 5a). Matthew (1904, p. 254) corrected the generic reference to Potamotherium (as Cope himself also had done at the time the plate was prepared), and noted the loss of the type and absence of other specimens. He considered it a small species and later (1915, loc. cit.) suggested that it was related to Sthenictis. Also, in 1915, he gave the locality as Pawnee Creek, Colorado, a lapsus calami. In 1922 Thorpe referred two specimens in the Yale Peabody Museum collec- tions to this species. Although these are from the Niobrara River fauna of Nebraska, they may well contain the key to the identity of the Oregon form. Restudy of the specimens described by Thorpe and examina- tion of additional material from the Niobrara River fauna and of fragments from the Crooked River region in Oregon 2 Postilla Yale Peabody Museum No. 8 lead us to the conclusion that they belong to a genus of pro- cyonid carnivores, which is inseparable from the living Bas- sariscus Coues on the basis of lower jaws and teeth alone. A maxillary, described below, which appears referable to the same species, differs markedly from the Recent form, how- ever, and suggests that were more known of these animals, a distinct genus might be indicated. Potamotherium Geoffroy, of which Lutrictis Pomel is a synonym, is a European otter dis- tinguishable from Bassariscus (and from these fossils) by its larger size, stouter jaw, more anteriorly placed single mental foramen, much shorter M., more posteriorly situated metaconid of M,, and characters of skull and upper dentition too numer- ous to mention here. Bassariscus PARVUS HALL FROM THE Nroprara River Fauna Direct comparison of the specimens described by Thorpe (1922, pp. 444-445: Y. P. M., Nos. 12825, 12834) and an additional lower jaw (from locality V3218, U. C. M. P., No. 33147) [see Stirton and McGrew, 1935, p. 127], with Bas- sariscus astutus (Lichtenstein) shows close agreement in such important features as the straight and slender horizontal ramus of the lower jaw; four premolars, the first single rooted ; presence of two mental foramina situated beneath P, and P;; and the form of the lower carnassial with a high trigonid, including a well developed metaconid, and a basined heel. The form of the premolars in Yale Peabody Museum No. 12825 also agrees with Bassariscus. These specimens differ from B. astutus and agree with the type of B. parvus from Cedar Mountain, Nevada, in the greater crowding of the premolars and relatively shorter trigonid of M,. One specimen from locality V3218, University of California Museum of Paleon- tology No. 29225, shows as much crowding as the type of B. parvus. None of the 13 specimens of B. a. raptor (Baird) in the Museum of Vertebrate Zoology which were examined show this condition. Niobrara River specimens are below the average May 10, 1951 Bassariscus In Miocene Faunas 3 length of B. astutus, although within the range of variation of the recent species, as shown by the following tabulation. The length of M, in the type of B. parvus is less than in any of the recent specimens, although the deviation is not significant. Measurements in Millimeters Length of M, Coefficient Number Observed Standard of Specimens Extremes Mean Deviation Variability B. astutus (Hall, 1927) 40 6.9-8.0 749 + 0.04 0.265 +.0.030 3.54 + 0.40 B. parvus, type 6.8 B. parvus, Niobrara River 2 7.0-7.3 7.15 Length talonid M, Coefficient Number Observed Standard of Specimens Extremes Mean Deviation Variability B. astutus (Hall, 1927) 4.0 2.4-3.0 2.70 + 0.02 0.1504 0.017 5.56 + 0.62 B. parvus, type 2.7 B. parvus, Niobrara River 2 2.4-2.5 2.45 Errors are standard errors. A maxillary with P?-M?, U.C.M.P. No. 31983, from locality V3218, Niobrara River fauna, (fig. 1) occludes so well with U.C.M.P. No. 33147 that it may have come from the same individual. It differs from a series of 13 specimens of B. astutus raptor in: P* crowded by P*; P* with relatively larger parastyle, with smaller protocone, and without hypo- cone; M! with somewhat stronger parastyle; M* with para- style more prominent and hypocone deflected more posteriorly ; infraorbital foramen more elongate dorsoventrally. The num- ber and height of cusps and general shape of the teeth other- wise resembles B. a raptor. This maxillary bears considerable 4 Postilla Yale Peabody Museum No.8 resemblance to that of foxes in the absence of a cusp posterior to the protocone of the upper carnassial, in the strong para- styles on M! and P*, and in the vertical enlargement of the infraorbital foramen. It differs from that of the kit fox, Vulpes macrotis arsipus (Elliot) (4 specimens from Yuma Co., Ari- zona, in M.V.Z.) in: more crowded P*-P*; shorter carnassial shear; more prominent parastyle and less developed hypocone of M’, the latter cusp not so deflected posteriorly ; protoconule and metaconule less developed on M'; M? proportionately shorter and with less developed cingula and hypocone. Figure 1. Bassariscus parvus Hall. Left maxillary, U.C.M.P. No. 31983, x2. Drawing by Owen J. Poe. As McGrew pointed out (1938, pp. 326-327) the principal distinctions between Bassariscus and the primitive fox, Pseu- docynodictis, lie in the presence of a posterointernal cusp and more anterior protocone on the upper carnassial, relatively larger metaconule of M', and greatly reduced cingula and absence of hypocone on M? in the former genus. The specimen May 10, 1951 Bassariscus In Miocene Faunas 5 here described resembles Pseudocynodictis gregarius (Cope) in the absence of a posterointernal cusp on P*, large parastyle of M', narrow anteroposterior diameter across the protocone of M', somewhat vertical infraorbital canal, and nearly similar size; it differs in the relatively lesser width across the proto- cone of P*, weaker hypocone of M', and undeveloped inner cingulum (smaller hypocone) of M?. Nothocyon lemur (Cope) differs more markedly in having a still more prominent meta- conule and larger hypocone on M' (thus approaching Procyon) and greater anteroposterior length of the inner part of M?. If correctly associated with the Bassariscus-like jaws, this specimen reveals that the late Miocene B. parvus retained a primitive, essentially canid pattern in the upper dentition although the lower jaws are scarcely distinguishable from the recent B. astutus. 'Two other procyonid genera, Cynarctus and Cynarctoides, (McGrew, 1938) lack the postero-internal cusp of the upper carnassial, but these have progressed much farther from the primitive condition typified by Pseudocyno- dictis in their molar pattern. Confirmation of the association of these specimens would probably justify erection of a new subgenus for Bassariscus parvus and related forms in which the upper carnassial lacks a fourth cusp, but material here described does not warrant proposal of a new name. The dis- tinctness of B. parvus from B. astutus, not demonstrable on features of the lower jaw alone, is supported by the tentative association of this maxillary dentition with its unique charac- ter combination. ? Bassariscus LycopoTamicus (Corre) FROM OREGON As figured by Cope the type jaw was slender and straight like that of Bassariscus, and contained four premolars in life. Cope described the trigonid of the carnassial as low, and the illustration shows it worn down almost to the level of the talonid. It is difficult, especially in the absence of the type specimen, to judge whether this wear could have been pro- 6 Postilla Yale Peabody Museum No.8 duced on a Bassariscus tooth with its tall trigonid, or whether the tooth was originally lower crowned as in Sthenictis. The type of Bassariscus antiquus matthewi (Merriam), U.C.M.P., No. 12539, is a heavily worn specimen and the trigonid and talonid of M, are worn to nearly the same level. Although P, of this specimen is broken at the crown it still shows less wear than M,. A similar difference in relative wear is apparent in the figure of Potamotheriwm? lycopotamicum, so it seems possible that the type of that species could have been a Bas- sariscus. Unfortunately the number of molars behind the carnassial cannot be determined as the specimen is broken off directly behind M,. A specimen from Paulina Creek, Oregon, in the collection of Yale Peabody Museum, No. 14313, bears much resemblance to Bassariscus. Its ramus is straight and slender but broken off in front of the greatly defaced M,;. Mz is absent. The talonid of M, is preserved and has a basin with distinct ento- conid and hypoconid. Although somewhat smaller than the B. parvus specimens from Nebraska there is little in this frag- mentary material to distinguish it from them. Paulina Creek is in the Crooked River region and could be either a Mascall (Miocene) or Rattlesnake (Pliocene) locality. The locality data given by Cope for P. ? lycopotamicum likewise is inade- quate to identify the source formation. No other specimens have been found to verify the location and even in 1907 Mer- riam and Sinclair (p. 195) pointed out that mixture of ma- terial from the Mascall and Rattlesnake formations is easily possible. It thus seems that “Potamotherium” ? lycopotamicum is probably referable to Bassariscus, and may have come from either Miocene or Pliocene. The limited material available is insufficient to demonstrate its affinities with other species. OruEer MiocENE OCCURRENCES OF BASSARISCUS Fossil cacomistle remains have been found in the Lower Snake Creek fauna (B. antiquus Matthew and Cook), the May 10, 1951 Bassariscus In Miocene Faunas 4 Virgin Valley fauna (B. antiquus matthewi Merriam), and Cedar Mountain fauna (B. parvus [Merriam] Hall). B. antiquus was distinguished from the Recent B. astutus by its larger paraconids on M, and M,, and by the slightly wider heel of M,. Merriam (1911, p. 246) sought to establish a new genus, Probassariscus, on these characters, but Hall (1927, p. 438) has pointed out that the variability within the Recent genus is such that the fossils should not be accorded more than subgeneric distinction. B. antiquus matthewi was not satisfactorily distinguished from the Snake Creek species, and Hall (loc. cit.), although recognizing the possibility that better material might reveal differences, maintains that the fossils can not even be shown to be subspecifically distinct. B. parvus Hall (B. nevadensis Merriam, 1916, non G. S. Miller, 1913) was considered by Merriam to be nearly indistinguish- able from the Recent “miners cat” of California, and Hall based most of his distinctions from Bassariscus astutus on the crowding of the premolars and size of trigonid of My). The material here described suggests that the species may be valid. At present then, the following extinct species of Bassariscus can be recognized: B. antiquus Matthew and Cook, Lower Snake Creek, Virgin Valley B. parvus Hall, Cedar Mountain, Niobrara River ? B. lycopotamicus (Cope), Mascall or Rattlesnake All are founded upon lower dentitions and differ only in minute characters from the living Bassariscus astutus (Lichtenstein). The lower teeth of this genus have undergone extremely little change since Oligocene time. The upper molars also are con- servative, the principal distinction from Pseudocynodictis be- ing a reduction of M?, but the upper carnassial has become 8 Postilla Yale Peabody Museum No. 8 modified in the Recent genus through addition of a postero- internal cusp and greater development of the internal cingu- lum on P*. A specimen from the late Miocene Niobrara River fauna suggests that this feature had not been acquired at that time. Oruer AMERICAN SPECIES REFERRED TO PoTAMOTHERIUM Potamotherium still appears in some faunal lists of North America* so it seems advisable to point out that those Ameri- can fossils which have been identified with this European otter are not at all related to it. Brachypsalis pachycephalus Cope was referred to Potamotherium by Hay (1902, p. 768); it is a far larger and stouter animal than P. valetoni (the geno- typic species) and the type of a now well-known American genus of Mustelinae (not Lutrinae). Potamotherium lacota Matthew from the Pliocene of South Dakota is likewise larger and referable to Brachypsalis; it appears close to B. modicus Matthew. As shown above, Potamotheriwm ? lycopotamicum Cope is probably not a mustelid at all but Bassariscus. ACKNOWLEDGEMENTS We are grateful to Dr. R. A. Stirton for assistance and permission to examine pertinent specimens in the University of California Museum of Paleontology, and to Dr. Alden H. Miller for access to comparative material in the Museum of Vertebrate Zoology. Valued comments concerning the material or manu- script have been received from Donald E. Savage, Robert W. Fields, Wann Langston, and Walter Wheeler. Owen J. Poe, artist of the University of California Museum of Paleontology, prepared the illustration. * Not, however, in Simpson’s Classification of Mammals. Bassariscus In Miocene Faunas May 10, 1951 6ESCIL ‘ON Ad WO ATTA UISITA snnbyun *g STEP ‘ON ‘WdA YI2I19 eulned U——_. (adog) ———_— addy, snovupjodooh) *g ¢ rag Ov V'é oP 3:0 : GL i Th 18 Gh "9 T8 i 69 0°9 gh ae VS GG reat) : VS os S6 ; Gs Le 8'9 A ee) OL rend 1% oe GG oF oP ae Lv VG ST mas 1G es 8's “ UP oo6 CT a0 oor 89L61 GG8cI PESSI LYLEE ‘ON 'd ‘WOT %“ON'Wd A ‘ON'NWd A “ON ‘dW ureyuno jy MA 38 snured ssouyory J, “| ye snurea Jo yydaq “AL ye snurea Jo yydoq *q ye snuesr jo yy}doq ®q ye snuier jo yydoq ee teeeees pruopey yqsue, ‘xoidde pruoye} ye yypIA "+ puostay} ye yprM oe oe iepag ‘eddy, W——— vuney Joarty vaerqoiN ——————” —__—__——— ey snawnd ‘g ——__———__ sel I9MO'T SIOPIUT|[TTAY UL SPUIULIANSBI]AT Yysus] ysozeois "PW * UQPIaA ysozyeois *g * yySugl ysoqzeors *g * YPIA ysoqyvois *g * yydug] jsozyeois &g " YIPIA ysoyvaid “gq * yQsugl ysozyeois 7g 10 Postilla Yale Peabody Museum No. 8 Measurements in Millimeters U.C.M.P. No. 31983, upper dentition Greatest Anteroposterior anteroposterior Transverse diam. across diameter diameter protocone SRA C Nyala egal etaloievete etsy ate a iers 4.2 2.1 | Ea eS SRA are Sh AL 6.5 4.0 J lig eet ves atthe as cig eee 5.2 Us 3.7 Pa SCREEN daha sm EVIE ORE ETA 3.2 4.9 2.2 BisLioGRAPHY Cope, E. D., 1879. Observations on the Faunae of the Miocene Tertiaries of Oregon. Bull. U. S. Geol. and Geogr. Survey of the Territories, vol. 5, pp. 55-69. Cope, E. D. and Matthew, W. D., 1915. Hitherto unpublished plates of Tertiary Mammalia and Permian Vertebrata. Amer. Mus. Nat. Hist., Monograph series No. 2. Hall, E. R., 1927. Species of the mammalian sub-family Bassariscinae. Univ. Calif. Publ., Bull., Dept. Geol. Sci., vol. 16, pp. 435-448. Hay, O. P., 1902. Bibliography and catalogue of Fossil Vertebrata of North America. U. S. Geol. Survey, Bull. 179. Matthew, W. D., 1904. New or little known mammals from the Miocene of South Dakota. Amer. Mus. Expedition of 1903. Am. Mus. Nat. Hist., Bull., vol. 20, pp. 241-268 [p. 254]. Matthew, W. D. and Cook, H. J., 1909. A Pliocene fauna from western Nebraska. Amer. Mus. Nat. Hist., Bull., vol. 26, pp. 361-414 [p. 377, fig. 6]. McGrew, P. O., 1938. Dental morphology of the Procyonidae with a description of Cynarctoides gen. nov. Field Mus. of Nat. Hist., Geol. Series, vol. 6, no. 22, pp. 323-339. Merriam, J. C., 1911. Tertiary mammal beds of Virgin Valley and Thousand Creek in northwestern Nevada, Part II, Vertebrate Faunas. Univ. Calif. Publ., Bull., Dept. Geol., vol. 6, pp. 199-304. Merriam, J. C. and Sinclair, W. J., 1907. Tertiary faunas of the John Day Region. Univ. Calif. Publ., Bull., Dept. Geol., vol. 5, pp. 171-205. Stirton, R. A. and McGrew, P. O., 1935. A preliminary notice on the Miocene and Pliocene mammalian faunas near Valentine, Nebraska. Amer. Jour. Sci. (5), vol. 29, pp. 125-132. Thorpe, M. R., 1922. Some Tertiary Carnivora in the Marsh Collection, with descriptions of new forms. Amer. Jour. Sci. (5), vol. 3, pp- 423-455. “ae a : } en ae ls y i ee if en vane it as i ee, Dd Ck Le ie Das aie aM ath i ie 2 alt A ir real ; f ey ay + i a tie vis id) an Ie les ALN Natty { b : fl t rR Z Fi C 1. Or eT, | rr folie ee OUSE We gil fhe hh a ltd Ree, pte Sa LTS, lie er " eae ri . “ed i: ; iy ? ‘ , 2 = i ty yy ' ae _ aie: it } 1 Ht ay. © a AV 7 Le Wa hee a | (ie at Wet) em * a | etsy i ah ea) i ; a Shy : iA An ae 7 rs ri peed fA 2\ ip + ‘ , is 7 } ry Ae ie re oy an Mtg ae ; if : i any iw xa ‘ -! ' i ant 5 ‘te = Ray * ' a ST | i. ’ a oe pa tl aS cy ee Co Manes La Ls eh de ear a ie? ‘ F ' , ; ie 2 xy he a) Oh a ae ere ad { Hes Sh 0! ‘ag aie, 70 ae rae ce oe : 4 a os er eet Ae i, ‘ve ih TURE ee ey eg as Ue We Bs ai AO hee Aleem! ey hie BO ' ‘ y ; r Hy en fl Tr) ily Pita ‘y.. ae ’ a y de ht ey ae a) Wii aera | rs ¥ ve ip, ; ’ wh, ALP an ih ‘ : Vern « m.\ “1 il | cite BANE aoa He i 7 Opts: 5 aN hag te Ae : Pros : J ae ay 4 fe - aa : ft: i i oe : f : NOt aoa Ea a - 2 LATA v4.8 ua ae sii ne i y ' ki ce Puy f Ae side nur W ane: Vn ioe a) y i Ce ee jai) tie Bae ccna ; ’ PR ar id ath any aes TAN CE ae ie hae Ware i 4, ) 5 } P ‘ hey Atta) bs A ox \ yy, FET Al 7," ein f } ¥ ve a ee i Nae ily toe, ah MELD Sb i ie ieee j " \ % hi } ‘ : f 4 s i F 4 Noa sl ‘ ‘ ‘ h " i » } ‘ we. ‘ ' y y bg - f { . | fais vant Sia Fi) i ee Me Pea TD \ pale ee ani" ’ ( nits Wy | ; : iS Iran ‘ j vp nie sift Pe ALIN ANY ey ‘KARTE MOT HOA x7 Olt tas , , Lf iif ath! / ' } i i 7 om ay ¥ . ok *) i, ‘ A \ ; ty Kee 4 “ ou it Ps at ene 7! | } ‘ ‘ it at ‘ , a Ah abe ah AEN ss 1 i i { R ws ‘ ’ a) | ; ’ ” 7 \ ih . i - 4 j i { ’ rye ( sannio. 2. Garrulax sannio oblectans, subsp. nov. Type: ¢ ad. (U.S.N.M., No. 277649) collected near Ipin [Suifu], elev. 1400 ft., southwestern Szechwan Province, China, November 22, 1923, by David C. Graham. Diagnosis: separable from G. s. sannio (Fukien) by having the several browns of the pileum, mantle, rectrices, throat and breast, belly, and under tail coverts more saturate and strongly rufescent (not olivaceous). From G. s. comis (Yunnan), it differs in the same characters as from G. s. sannio, but also by having the postocular stripe brownish black (not light rufescent brown) and by having the supercilium unsullied white or creamy (not suffused with rufescent brown above and behind the eye). Range: Chinese province of Szechwan, elev. 1,000-6,000 ft., and northern Kweichow. 4 Postilla Yale Peabody Museum No. 11 Remarks: Garrulax sannio may be considered an autochthon of southern China, which probably originated as a species in northwestern Yunnan or southeastern Sikang, where is found a race (comis) that shows in the adult certain charac- ters (such as the light rufescent-brown postocular stripe) that appear in other races only in the immature plumage. From this center it has pushed out in two directions: west- ward across northern Burma as far as Manipur and adjacent parts of Assam (albo-superciliaris), where it is rare, and southeastward along the rivers as far as the Southern Shan States, northern Laos, northernmost Annam, and Tongking, in all of which it is common at elevations from 2800 to 4900 feet. Specimens from these southeastern regions represent popu- lations of individuals variably intermediate between topo- typical comis and topotypical sannio (Fukien), which there- fore cannot definitely be named. The majority, however, from localities west of the Black River-Red River divide (western Tongking) are nearer comis, while the majority of those from east of the divide are nearer sannio; this line might then, simply for convenience, be considered the boundary between the two forms. It may be supposed that the species next advanced north- eastward through the southeastern provinces of China, where, as the race sannio, it now occupies the hills of Kwangsi, Kwangtung, Fukien, Kiangsi, and northeastern Hunan (Yo- yang [Yochow]). Having reached the valley of the Yangtze in Kiangsi and Hunan, and by now accustomed to elevations much lower than those of the ancestral homeland, the species readily ad- vanced westward up the great river into the lowlands of Szechwan, here to give rise to the race oblectans (birds from southwestern Hupeh are already oblectans > sannio). Thus, while the ranges of comis and oblectans are contigu- ous, it may be seen that these two are farthest apart in distance traveled from the original home, and, by the same March 26, 1952 ‘Garrulax sannio Swinhoe 5 token, of all Chinese populations farthest apart in external characters. Interbreeding of the two races, with consequent masking of their differences, is inhibited by the fact that they occupy distinct altitudinal ranges. That the geographi- cally distant G. s. sannio should be the intermediate between them is accounted for by the history of specific expansion I have hypothesized above. So far as is now known, the range of G. s. albo-superciliaris is wholly isolated from those of its Chinese cousins and, as might be expected, this is a very distinct form by the deep (scarcely rufescent) brown of its pileum, the cold dark olivaceous brown of its mantle, and the strong vinaceous wash over its entire under parts. Its postocular stripe is blackish brown in the adult. Since the birds of this genus are subject to alteration of color by wear, I should mention that my diagnoses have been based wholly upon fresh-plumaged adult specimens. The pos- sibility of post-mortem change has also been considered and discounted, since skins of G. s. comis taken twenty-nine years ago do not differ significantly from one collected in 1945, and the oldest specimens of each race are roughly equivalent in age and should therefore have altered to the same degree. For the record, I should state that I have examined 40 examples of comis from Yunnan and Sikang, taken in 1923, 1928, 1929, 1980, and 1945; three of sannio from Fukien, taken in 1923 and 1930; forty of oblectans from Szechwan, taken in 1921, 1922, 1923, 1924, 1925, 1928, 1929, 1931, 1932, 1933, and 1934; three of albo-superciliaris, taken in 1950. Of the birds of Laos and Tongking, intermediate between comis and sannio, I have had 36, collected in 1924, 1929, 1930, 1938, 1939, and 1941. iy * / bs Aired el fet TRL es Rua eR A Ape he | ah hig earl hy ied Pega ard ie aad hi 1% a Maddie io Tie een PP avai’ Lat aye te aes sake VO) PT Leen ead Ad 1 + } RNY Soe find 5 ie 1h ST eR ES Shy ot! ’ * bic to? Pa ie. ht Oe Ree eee Pam Vs iy SAL ae ee ay Mithril) PPAR Ae Mm cis ahh) 1 Aide Rae ARS Ae 1 eae tay Fh 3 mi he i eae wh j 4 cdl vil iw ita ert. ity eis dhe bee Vey 4 A" ems | : ‘ 4 7 Vili td Ete ee eT ae y not ra th ee [ rity ; F ie yt 4 4 ve | i h § f ae 4 Ml -. i ¥ 4 + ' i a va oe es 4 ¥, j Bin! iy . ; } A i An bee » i / Nd uf ‘ } if i 1 . 7 Atk ht as Ap) t ’ Pi i ie : ¥ - fpr Wie : 7 Ls 7 é 7 \ - : 4 ' ras f +, ‘ ; A 7 ‘ | bea | ae, awl ips . * Cn 4 Aba pane Anh re fi aot ee a ; a ivy } ‘ o & a / i Be Ma eA | Rees ae 2) ie et Peas 4 Mais ; AY ne ee Pay ihs q A / ae : ‘ yee ae 1. 4 a Ne he r ' 7 : ; | : pee tes p) ma j ea. ata. at : aA oe j A x “7 i Wy as A i ; Ma Pont [ere a, i ee A ati Ain’ & : * i . é J j ry 16: ) ’ \) ) i! .% te PAL heath RP wet ie sie tay an Were ke ty id.) bares hia Ont Mine eh Eby an sili die: Sr pha de ae a hae ay ee pees é * pot \ j ‘ 4 | 7 ; tH 7 a ve 7 4 4 s i ( 6 } A TeAst i * nh § } y \ ! tin ae rat ‘Airn iS hy _N/- NUS. COMP. ZO8L. LIBRARY UN i OF 1952 HARVARD UNIVERSITY prion YALE PEABODY MUS or NaturaL History Number 12 April 2, 1952 New Haven, Conn. A NEW GENUS OF THRUSH FROM EASTERN AFRICA S. Ditton RireLtey In connection with a revision of the thrushes, I have examined specimens of T'urdinus stictigula Reichenow through the courtesy of the authorities of the American Museum of Natural History. This rare and seldom observed species found only in the hills in parts of Tanganyika was described as a babbler and placed in Illadopsis by W. L. Sclater (Systema Avium “thiopicarum, 1932 :363). In his revision of the bab- blers (L’Oiseau, 1946, 16:13) Delacour has pointed out that the Spot-throat is certainly a thrush and not a babbler in the lengthening of its narrow bill, and its long and slender tarsus and toes. M. Delacour placed the species provisionally in Cossypha, noting that it bore a slight resemblance to anomala and archeri. Virtually nothing was known of the habits of the Spot- throat until Moreau published his observations (Ibis, 1932: 672 and 1938:302). He noted particularly its thrush-like habits and beautiful voice. He further described the nest, eggs, and nestlings, all reasonably thrush-like, but unfor- tunately the nestlings were destroyed before he could observe the juvenal plumage. In its uniform coloration, spotted throat, and wing and tail formation, stictigula differs notably from the members of the genus Cossypha, and I therefore propose the erection of a new genus. Modulatrix, n. gen. Type.—Turdinus stictiguia Reichenow 1906. This genus is similar to Cossypha Vigors 1825 in its narrow bill and its long tarsi and toes; but the wing, which is rounded, has the sixth primary longest rather than the fifth and the first primary is less than two-thirds the length of the second (shorter than in species of Cossypha). The tail, which is shorter than the wing and of 12 feathers, is slightly rounded rather than squared, the individual rectrices being somewhat pointed. The color pattern of Modulatriaz is distinc- tive uniform deep olive brown on the crown and back, with a faint tendency to terminal barring on the feathers of the center of the back. The tail is rich rufescent. Below, the throat is buffy-gray with terminal black spots; the abdomen is rufous. The flanks are dark brown. There is no tendency to a white eyebrow, crown or presuperciliary spot or stripe, so uniform a feature of Cossypha. Measurements of three males are: wing, 79-81; tail, 71-72; culmen, 16; and tarsus, 29-30 mm. Two forms of this genus have been described: 1. Modulatrix stictigula stictigula (Reichenow) 2. Modulatriz stictigula pressa (Bangs and Loveridge) “Moduatria,” a smger Tertull. pstilla MUS. COMP. ZOOL. LIBRARY. YALE PEABODY MUSEUM APR 238 1968 HARVARD UNIVERSITY. Number 13 September 18, 1952 New Haven, Conn. or NaturaL History THK THRUSHES S. Ditton Rietey “The merle, the mavis, and the nichtingale, With mirry notis mirthfully forth burst,” from May by Gavin Douglas, 1513. The thrushes are one of the most cosmopolitan of the groups of birds. References to the nightingale as Procne rather than as Philomela occur in the work of Apollodorus, about sz. c. 140. By the time of Ovid, 3. c. 43-18 a. p., Philomela had become the nightingale and so continued through early legend and poetry. The Persian bulbul of the eastern poets was un- doubtedly the nightingale which was supposed to pour forth its strain of melody while pressing its breast against a rose thorn to ease its heart’s pain. Thrushes of numerous sorts have been noted as superlative songsters in many countries since earliest times. By the Nineteenth Century Seebohm (1881) and other stu- dents of birds were inclined to list the thrushes as the most highly developed among the birds on account of their singing qualities and the development of the plantar tendons. Newton (1893) citing Cabanis (1847) and W. K. Parker (1872-1873) disagreed, and later arrangers from Stejneger to Stresemann 2 Postilla Yale Peabody Museum No. 13 have listed the crows and their allies as the highest group of the Oscines. The work of Seebohm (op. cit.), however, and that of Seebohm and Sharpe (1898-1902) have certainly been, historically, the most important monographs on the thrushes. These authors felt, as I do, that the thrushes represent a subfamily, the Turdinae, closely allied to the warblers, Sylviinae, from which they differ by not possessing a complete double molt, and by having in nearly all cases a spotted plumage in the young. Thrushes also, in most cases, have a booted rather than a scutellated tarsus. But even these rather generalized characters tend to have exceptions. Perhaps in such closely related groups as those within the Passeres, exceptions tend to prove the rule. Hartert (1910) also listed the thrushes as a subgroup within the Muscicapidae, and this arrangement has been followed recently by Mayr and Amadon (1951). The principal objections to such a classification have been expressed by Witherby, Jourdain, Ticehurst, and Tucker (1938). In actual effect the objections to listing the thrushes as a subfamily may be reduced to the simple fact that the group is a very large one numerically, and that to lump them together with such other large groups as the warblers, babblers, and flycatchers into a single family may be an exceedingly unwieldly arrange- ment. This of course is true, but against this must be balanced the lack of well-defined characters, morphological or otherwise, to separate these groups. If categories, and especially higher categories in the Class Aves are to have comparable validity to those in other classes of animals it would be well to be some- what more conservative in these matters than many systematists have been in the past. It has been amply demonstrated in recent years that many of the minutiae, the characters set up by avian taxonomists to create such categories as subgenera, genera, subfamilies, and the like, based on the relative differ- ences between the length of the tarsus and the wing or tail, the length of bristles versus the bill length and so on, are extremely plastic in nature. Too much reliance on such relative- ly trivial morphological characters serves rather to obscure than to define relationships. My inclination then to list the thrushes as a subfamily is September 18, 1952 The Thrushes 3 based not only on the consideration of morphological and evolutionary characters, but also on the desire to be consistent. The characters separating the groups of the Muscicapidae are so few and inconstant that to keep the groups as separate families serves only to damage the concept of the term, family. Within the Muscicapidae, the differences between the numerical- ly largest subfamilies may be listed as follows: Juvenile Double Rictal Subfamily plumage molt Tarsus bristles weak, Muscicapinae spotted uncommon scutellate strong Timaliinae unspotted uncommon scutellate present scutellate Sylviinae unspotted usual primarily distinct booted or Turdinae spotted uncommon scutellate variable RELATIVES Besides the above closely related subfamilies, the dippers, Cinclinae, seem to me to be nearly related to the thrushes as suggested by Stejneger (1905), and I am inclined to agree with Mayr and Amadon (op. cit.) in making them a subfamily of the Muscicapidae. The mockingbirds, Mimidae, recently have been considered a separate family, but formerly were included as a subfamily of the wrens (Coues and others). Perhaps both wrens and mock- ingbirds should be placed in the Muscicapidae as suggested recently by Mayr and Amadon (op. cit.). The tarsus of the mockingbirds is barely scutellate. The internal anatomy of this group differs only slightly in the narrowness of the anteorbital region of the skull, the rather narrow descending process of the nasal bone and the slightly differently shaped coracoid, sternum, and the narrow pelvis. In Dumetella the maxillo-palatines ap- proach the thrushes exactly in shape, rather than being clavi- form as in the other mimine genera. These differences are no 4 Postilla Yale Peabody Museum No. 13 more compelling than those between other subfamilies of the Muscicapidae. Ridgway (1907) erected a family for Zeledonia, the aber- rant wren-thrush of Costa Rica on the basis of its possessing 10 tail feathers. The young also are not really spotted, hav- ing only a tendency to pale edgings on the feathers of the lower parts. In view of the diversity of the thrushes, and from the appearance, anatomy (Pycraft, 1905), and of what is known of its habits, I consider Zeledonia to be a relict member of the Turdinae, and a geographical representative allied to the short-wings, a group of South Asian chat-thrushes of the genus Brachypteryz. The hedge-sparrows (Prunella) are kept by most authors as a separate family. Field and museum study inclines me to the belief that these birds are a subgroup within the thrushes as proposed by Baker (1924). The hedge-sparrows have scutel- late tarsi, but the structure of the tibiotarsus in Sazicoloides seems to me intermediate and leading directly to Prunella. The musculature of the cheek in the hedge-sparrows puts this genus into the chat-like thrushes according to Beecher (personal communication), as does the immature plumage which is streaked or mottled. I consider, therefore, that the hedge-sparrows are a group of the chat-like thrushes which has become secondarily bunting-like in its food adaptations. ACKNOWLEDGMENTS I am most grateful to the authorities of the American Museum of Natural History and the United States National Museum who have loaned me material in their care. Several individuals have been particularly helpful in discussions, notably Dr. J. P. Chapin who has generously shared with me his wealth of knowl- edge on the African thrushes. Without his advice I should have been unable to consider many of these peculiar genera. The comments of Messrs. Amadon, Beecher, Deignan, Delacour, Friedmann, and Peters have all been sought and freely given. Their advice has been most welcome and instructive. September 18, 1952 The Thrushes ii DistTRIBUTION The thrushes would seem to be primarily of Palaearctic origin, I believe. The greatest number and complexity of forms occur in the Old World. Africa, also, has been an important evolutionary center in this subfamily, especially in the chat- like forms. From Eurasia there have been several invasions into the New World, even reaching outlying groups such as Hawaii and isolated Tristan da Cunha. In the other direc- tion invasions have penetrated into the Australian region, and into Polynesia. New Zealand apparently has not been reached. Turnagra, formerly considered the New Zealand thrush, now appears to be nearer the shrike-billed flycatchers (Oliver, 1945). No continental area, however, is without some member of the thrush group. ARRANGEMENT In general the thrushes fall into two large groups or tribes, the chats and the true thrushes. The name for the second group is a convenient one in that the word, thrush, conjures-up to the mind a generalized bird of a definite size; a bird which might be a robin in the United States, a song thrush in England, a mistle thrush, fieldfare, or blackbird in many parts of Europe and northern Asia, or their equivalents in such a variety of places as Tierra del Fuego and the Solomon Islands. These birds are all of a roughly equivalent size and shape and even have roughly similar types of song. The chats on the other hand are in general small, skulking, often very difficult to see or observe if they belong to jungle-living species, and do not generally convey the impression of being thrushes at all except to the initiated, aside from those species which happen also to possess fine songs. Cuat THRUSHES The chats seem to me to represent the earlier forms of the subfamily. They are all smaller; many are more specialized 6 Postilla Yale Peabody Museum No. 138 for tropical or subtropical habitats; they are less uniform as a group, and generally present the impression of containing more end-lines of evolution and differentiation among them, more relicts. The exceptional cases of unturdine-like characters such as unspotted young, scutellate tarsus, prominent bristles and so forth occur among the chat-like aggregation. As a tribe, then, they would seem to have had a longer history. Some genera are transitional between thrushes and warblers. Others resemble the flycatchers. In general the chat tribe might be characterized as being smaller, possessing more slender legs, more diverse nesting habits, weaker song, and a tendency to brighter, more variegated plumage in adult and young. Diverse as they are, however, they seem to stand closer together as a group, distinct from the larger, more compactly evolved true thrushes. Mr. William Beecher’s anatomical researches on the musculature of the cheek of passerine birds inclines him, also, to this division into two main groups within the Turdinae (personal communication). The largest development numerical- ly of the chat thrushes has taken place in Africa and may well have commenced there for all that we know geologically or climatologically speaking. The most primitive of the chat thrushes would seem to consist of two types, the short-wings, personified by Brachypteryx and Zeledonia, which I feel are equivalent tropical relict forms of the Old and New World, and the genus Erythropygia, which seems rather like a link between the warblers and the thrushes. Near Erythropygia may come the aberrant Drymodes of Australia and New Guinea, a relationship suggested to me by Dr. Amadon. I had previously thought these scrub-robins nearer the true thrushes, the group with closer geographical connections. In pattern of coloration there are many similar- ities even though the longer tail in Drymodes, longer than the wing in all cases, is certainly proportionally different from that in Erythropygia. From such warbler-like origins have developed the robins, Erithacus, with their bewilderingly diverse complex of relatives in Africa, the magpie-robins, Copsychus, the fork-tails, Enicu- rus, the solitaires, Myadestes, of the New World, an early suc- cessful invader of that area with its own offshoot, Phaeornis, September 18, 1952 The Thrushes 7 and its Old World relict relatives, Cochoa (?), and possibly Stizorhina (?). The remaining subdivisions of the chat thrushes are more uniform, running from the redstarts, Phoenicurus, again through an array of African relatives, to Cercomela and the chats, Saxicola, beyond which in the evolutionary scale would seem to lie the wheatears and the rock-thrushes. Low down in the chat thrush group lies the small twig leading off through Saxicoloides to Prunella. True THRUSHES The true thrushes consist of a more uniform group of birds in size, almost all of which have fine powers of song. These birds range from open heath in the temperate and subarctic regions to deepest tropical jungle. They have strong tarsi, strong bills lacking frontal hairs in most cases, and they molt from immature plumage in the first autumn directly into the adult plumage. These thrushes have certain seemingly more primitive off- shoots among them, notably Myiophoneus of the Indo-Malaysian region, several isolated genera in Celebes and New Guinea, and at least three separate invasions of the New World. The earl- iest of these would seem to be represented by the forest thrush of the West Indes, Cichlherminia, and the Tristan da Cunha thrush, Nesocichla, which I feel shows New World rather than Old World affinities. A second invasion has produced the distinctive birds grouped currently in the genera, Hylocichla and Catharus, birds of the forest or the forest edge which in their habits and song seem almost intermediate between the nightingale-robin group of the Old World, and the true thrushes. A third rather inter- esting minor invasion in terms of end results has left two stranded species of the Zoothera assemblage in western North America and Mexico. A final invasion has populated the whole of both continents with a multitude of true thrushes of the genus T'wrdus which I would rank as the most highly developed genus of the subfamily. Whatever the varying points of view 8 Postilla Yale Peabody Musewm No. 13 may be as to its place in the evolutionary sequence, it is certain- ly the most widespread of all the genera. Tue GENERA Sharpe (1903) listed 75 genera in his “Hand List” in the family Turdidae. Excluding Turnagra and Ephthianura, and adding eight genera placed by him in the Timaliidae and Sylvi- idae, namely Drymodes, Chaétops, Cataponera, Myiophoneus, Arrenga, Brachypteryx, Heteroxenicus, and Agrobates raises his total to 81. In the list following I propose that the number recognized be 46, a reduction of 45 per cent. 1) CHAT THRUSHES Among the chat thrushes several generic names seem more usefully dropped than retained. The genus Heterowxenicus Sharpe, a new name for Drymochares Gould, was separated from Brachypteryx on the basis of having a tail less than twice the length of the tarsus. The type of Heteroxenicus cruralis is now considered a subspecies of the species B. montana of most authors, so unimpressive does this tail character appear. The same character was used for the species stellatus, type of Drymochares. The genus Heinrichia was erected by Stresemann for the species calligyna discovered on Celebes by Heinrich. It is similarly, if somewhat somberly colored to Brachypterya, but happens to be somewhat larger than the other species. I can glean no indication of its habits or structure being in any way different from a typical Brachypteryz. In the genus Erythropygia I include T'ychaédon, a renaming by Richmond of Sharpe’s Aedonopsis, erected for the species signata of South Africa. In color pattern and habits this brown robin-chat seems to belong with the scrub-robins and should be placed there. The fact that the culmen in signata is longer Turdus ), Zoothera Nesocichla fa’ = Cathorus ll we Cichlherminia Cataponera Amalocfchl Oenanthe ee Pentholaea Ec Geomalia Saxicola Myiophoneus Myremecocichla Cercomela Chaetops Emarginata Thamnolaea Phaeornis Prunella S Myadestes Cochoa Enicurus Saxicoloides Sialia Myiomela Grandala frania Phoenicurus se SATE Copsychus Hodgsonius Alethe ctchladusa Pinarornis Ree Cercotrichas LJ ~ Erithacus Cossypha

? 2? ”? ? 9 > September 18, 1952 The Thrushes 21 Brachypterys montana saturata Salvadori >? 2? ”? erythrogyna Sharpe ‘ss montana Horsfield ‘i floris Hartert II. Zeledonia coronata Ridgway wT. Lrythropygia coryphaeus coryphaeus (Lesson) abbotti Friedmann* hamertoni Ogilvie-Grant* leucophrys lecoptera (Riippell) eluta Bowen vulpina Reichenow brunneiceps Reichenow soror Reichenow jugens Bowen* vansomerni Sclater ruficauda Sharpe zambesiana Sharpe munda (Cabanis) pallida Benson sclateri Grote pectoralis A. Smith leucophrys (Vieillot) a kabalii White hartlaubi Reichenow galactotes galactotes (Temminck) 4G minor (Cabanis) syriaca (Hemprich and Ehrenberg) familiaris (Menetries) paena benguellensis Hartert ” — damarensis Hartert paena A. Smith leucosticta collsi Alexander ‘< leucosticta (Sharpe) * reichenowi Hartert barbata erlangeri Reichenow* ‘a greenwayi Moreau* quadrivirgata (Reichenow) rovumae (Grote) 3 barbata (Finsch and Hartlaub) wilsoni Roberts* signata (Sundevall) > 2? >? Postilla Yale Peabody Museum No. 13 LE Drymodes brunneopygius brunneopygius Gould pallidus (Sharpe) superciliaris colcloughi Mathews i beccarii (Salvadori) superciliaris Gould is ‘ brevirostris (De Vis) nigriceps Rand Vi Enthacus erythrothorax erythrothorar (Hartlaub) gabonensis (Sharpe) zanthogaster (Sharpe) mabirae (Jackson) sharper sharpei (Shelley) . usambarae (Macdonald)* gunningi gunning (Haagner) is 3 bensoni (Kinnear) * sokokoensis (van Sommern) cyornithopsis cyornithopsis (Sharpe) lopezi (Alexander) if ge acholiensis (Macdonald) * bangsi (Friedmann) houghtoni (Bannerman) aequatorialis (Jackson) Cea ai ruwenzorii (Ogilvie-Grant) guttifer (Reichenow and Neumann) elgonensis (Ogilvie-Grant) keniensis (Mearns) macarthuri (van Sommern) orientalis (Fischer and Reichenow) johnstoni (Shelley) transvaalensis (Roberts)* sf lebombo (Roberts) margaritatus (Sundevall) swynnertoni (Shelley) rubecula melophilus Hartert rubecula (Linnaeus) vanthothorax Salvadori and Festa sardus Kleinschmidt superbus Koenig witherbyi Hartert atlas Lynes 9? > > 2? 3? »? 2? >) September 18, 1952 The Thrushes 23 Erithacus rubecula tataricus Grote 33 + caucasicus Buturlin ‘fi 4 hyrcanus Blanford sibilans (Swinhoe) luscinia (Linnaeus) megarhynchos caligiformis (Clancey and von Jordans )* megarhynchos (Brehm) re ie corsa (Parrot)* ‘i i hafizi (Severtsov) i 4 lusciniodes (von Jordans)* 44 ms africana (Fischer and Reichenow) akahige akahige (‘Temminck) + tanensis Kuroda kobayashii (Momiyama) calliope camtschatkensis (Gmelin) iy cs beicki (Meise) ‘i ‘i calliope (Pallas) svecicus svecicus (Linnaeus) a a namnetum (Noirmoutier) occidentalis (Zarudny) cyaneculus (Wolf) luristanicus nom. nov. pallidogularis (Zarudny )* i A abbotti (Richmond) tianschanicus (Tugarinov)* kaschgariensis (‘Tugarinov)* 2? ? ‘< * altaicus (Sushkin) 4 ‘i saturatior (Sushkin) s is weigoldi (Kleinschmidt) przevalskii (‘Tugarinov)* kobdensis (Tugarinov)* pectoralis pectoralis (Gould) oe a confusus (Hartert) és a ballioni (Severtsov) tschebaiewi (Przevalski) si ruficeps (Hartert) if wickhami (Baker) pectardens (David) 29 2? 2 Erithacus svecicus magnus (Zarudny and Loudon), 1904, is preoccupied by Philomela magna Blyth, 1833 (a nomen nudum in synonymy with Luscinia luscinia Linnaeus), 24 Postilla Yale Peabody Museum No. 13 Erithacus hachisukae nom. nov.* a cyane cyane (Pallas) 3 ” — bochaiensis (Shulpin) brunneus (Hodgson) komadori komadori (Temminck) im ” namiyei (Stejneger) “ f subrufus (Kuroda) cyanurus cyanurus (Pallas) C + rufilatus (Hodgson) practicus (Bangs and Phillips) i. ‘ pallidor (Baker) albocoeruleus (Meise) ussuriensis (Stegmann) chrysaeus chrysaeus (Hodgson) . ie whistleri (Ticehurst) i if vitellinus (Stresemann) indicus indicus (Vieillot) A ” — yunnanensis (Rothschild) formosanus (Hartert) - hyperythrus (Blyth) johnstoniae (Ogilvie-Grant) ? >> >? ») Vi: Cossyphicula roberti roberti (Alexander) ” rufescentior (Hartert) VII. Cossypha insulana insulana Grote a granti Serle kungwensis Moreau polioptera polioptera Reichenow * ie tesmanni Reichenow nigriceps Reichenow bocagei Finsch and Hartlaub archeri Sharpe ‘ isabellae isabellae G. R. Gray * - batesi (Bannerman) natalensis natalensis A. Smith ms ¥ hylophona Clancey* garguensis Mearns “4 dichroa (Gmelin) semirufa semirufa (Riippell) > ”? ” ” ”? »” 3 As Marquess Hachisuka has noted (in litt.) Erithacus obscura Berezow- ski and Bianchi, 1894, is preoccupied by Cyanecula obscura Brehm 1831, a synonym of Erithacus svecicus cyaneculus (Wolf), 1810. September 18, 1952 The Thrushes Cossypha semirufa donaldsoni Sharpe 2? intercedens (Cabanis) heuglini heuglini Hartlaub ic ” subrufescens Boacage intermedia (Cabanis) cyanocampter cyanocampter (Bonaparte) si . periculosa Sharpe % os bartteloti Shelley caffra iolema Reichenow i ” kivuensis Schouteden caffra (Linnaeus ) namaquensis Sclater albigularis albigularis (Reichenow) maclounii (Shelley) 2? >? 2? >? >? >? njombe Benson anomala (Shelley) a humeralis (A. Smith) niveicapilla niveicapilla (Lafresnaye) ‘ ss melanonota (Cabanis) 2 albicapilla albicapilla (Vieillot) 4 x, giffardi Hartert yi iy genderuensis Reichenow omonensis Sharpe sharpei sharpei G. R. Gray i ” erythronata Lavauden imerina Hartlaub >”? ? WITT: Modulatrix suictigula stictigula (Reichenow) pressa (Bangs and Loveridge) IX. Cichladusa guttata guttata (Heuglin) > rufipennis Sharpe arquata Peters i ruficauda (Hartlaub) 2? X. Alethe castanea castanea (Cassin) 2? a9 woosnami Ogilvie-Grant poliophrys Sharpe fiilliborni fiilliborni Reichenow is usambarae Reichenow poliocephala poliocephala (Bonaparte) castanonota Sharpe carruthersi Ogilvie-Grant >> 25 porotensis (Bangs and Loveridge) 26 Postilla Yale Peabody Museum No. 13 Alethe poliocephala akeleyae Dearborn diademata (Bonaparte) choloensis choloensis Sclater ii fe namuli Vincent* sharpei (Shelley )* XI. Xenocopsychus ansorgei Hartert 5.1 Cercotrichas podobe podobe (P. L. S. Miiller) melanoptera (Hemprich and Ehrenberg) (?)+ XIII. Pinarornis plumosus Sharpe XIV. Copsychus saularis saularis (Linnaeus) ceylonensis Sclater andamanensis Hume prosthopellus Oberholser amoenus (Horsfield) erimelas Oberholser nesiotes Oberholser zacnecus Oberholser nesiarchus Oberholser masculus Ripley pagiensis Richmond javensis Chasen and Kloss problematicus Sharpe pluto Bonaparte | ts adamsi_ Elliott mindanensis (Gmelin) seychellarum Newton albospecularis albospecularis (EKydoux and Gervais) ¥ is pica Pelzeln inexpectatus Richmond malabaricus malabaricus (Scopoli) i = indicus (Baker) ‘i i leqggei (Whistler) ‘f ‘a albiventris (Blyth) 2? > 4'Two specimens of this form in the collection of the Bombay Natural History Society from S. W. Arabia have brownish quills and a rufous patch on the inner webs, but they are very worn specimens. September 18, 1952 The Thrushes 27 Copsychus malabaricus interpositus (Robinson and Kloss) »”? minor (Swinhoe) mallopercnus (Oberholser) Hf tricolor (Vieillot) melanurus (Salvadori) opisthopelus (Oberholser) javanus (Kloss) omissus (Hartert) “3 ochroptilus (Oberholser) heterogynus (Oberholser) eumesus (Oberholser) Ks abbotti (Oberholser) suavis Sclater nigricaudus (Vorderman) stricklandii Motley and Dilwyn barbouri (Bangs and Peters) luzoniensis luzoniensis (Kittlitz) >? parvimaculatus (McGregor) superciliaris (Bourns and Worcester) 9? niger niger (Sharpe) 2? cebuensis (Steere) pyrropygus (Lesson) XV. TIrania gutturalis (Guérin) XVI. Phoenicurus alaschanicus (Przewalski) ) >> erythronotus (Eversman) caeruleocephalus Vigors ochruros gibraltariensis (Gmelin) aterrimus von Jordans* af ochruros (S. G. Gmelin) semirufus (Hemprich and Ehrenberg) phoenicuroides (Horsfield and Moore) zerophilus Stegmann ¥ rufiventris (Vieillot) phoenicurus phoenicurus (Linnaeus) a algeriensis (Kleinschmidt) samamiscus (Hablizl) turkestanicus Zarudny* hodgsoni (Horsfield and Moore) 2? 2»? 28 KVIT: XVIII. XIX. XX. Postilla Yale Peabody Musewm No. 13 Phoenicurus frontalis frontalis Vigors ‘3 ”\ \wsinae THartert* schisticeps schisticeps (Gray) e ‘ beicki Stresemann auroreus auroreus (Pallas) a % leucopterus Blyth moussieri (Olphe-Gaillard) erythrogaster erythrogaster (Giildenstadt) maximus Kleinschmidt fe si grandis (Gould) leucocephalus leucocephalus Vigors % 3 pamirensis (Zarudny and Moltchanow ) i bicolor (Ogilvie-Grant) ‘o fuliginosus fuliginosus Vigors i i affinis (Ogilvie-Grant) i frontalis frontalis (Blyth) 7? >? orientalis (Delacour and Jabouille) Hodgsonine phoenicuroides phoenicuroides (Gray) " ichangensis Baker Myiomela leucura leucura (Hodgson) 4 i cambodiana (Delacour and Jabouille) diana sumatrana (Robinson and Kloss) ie ” diana _ (Lesson) ie frontalis (Blyth) Grandala coelicolor coelicolor Hodgson ‘ ‘ florentes Bangs and Peters Sialia sialis sialis (Linnaeus) " ” grata Bangs episcopus Oberholser fulva Brewster azurea Swainson? meridionalis Dickey and van Rossem* mexicana mexicana Swainson 4 4 australis Nelson* anabelae Anthony occidentalis Townsend ”? »”» ”? > 5 Replaces guatemalae Ridgway. September 18, 1952 The Thrushes 29 Sialia mewicana bairdi Ridgway * —_ currucoides (Bechstein) XXI. Enicurus scouleri scouleri Vigors . i. fortis (Hartert) velatus sumatranus (Robinson and Kloss) ai q velatus Temminck ruficapillus Temminck immaculatus (Hodgson) schistaceus (Hodgson) leschenaulti indicus Hartert ‘ s sinensis Gould a % frontalis Blyth borneensis (Sharpe) is ii leschenaulti (Vieillot) chaseni de Schauensee maculatus maculatus Vigors a ‘4 guttatus Gould is (i omissus Rothschild a 7 robinsoni Baker XXII. Cochoa purpurea Hodgson a viridis Hodgson azurea beccarii Salvadori is < azurea (Temminck) 2”? XXIII. Myadestes townsendi townsendi (Audubon) i a calophonus Moore obscurus obscurus Lafresnaye e ‘ oberholseri Dickey and van Rossem occidentalis Stejneger cinereus Nelson insularis Stejneger a elisabeth elisabeth (Lembeye) ‘a x retrusus Bangs and Zappey genibarbis solitarius Baird ‘ fe montanus Cory dominicanus Stejneger genibarbis Swainson sanctae-luciae Stejneger sibilans Lawrence ¢ ralloides ralloides (d’Orbigny) q ‘k plumbeiceps Hellmayr venezuelensis Sclater > ” 30 Postilla Yale Peabody Museum No. 13 Myadestes ralloides candelae de Schauensee i - coloratus Nelson melanops Salvin unicolor unicolor Sclater i ” veraepacis Griscom pallens Miller and Griscom leucogenys leucogenys (Cabanis) i : gularis (Salvin and Godman) peruvianus (Hellmayr) re . chubbi (Chapman) - leucotis (Tschudi) coracinus Berlepsch 9) >? >? > ”? >? XXIV. Phacornis obscura obscura (Gmelin) lanaiensis Wilson rutha Bryan oahensis Wilson and Evans myadestina Stejneger palmeri Rothschild XXV. Stizorhina fraseri fraseri (Strickland) ” — rubicunda (Hartlaub) i ” — vulpina Reichenow i finschii (Sharpe) XXVI. Neocossyphus rufus gabunensis Neumann ” — arrhenii Lonnberg rufus (Fischer and Reichenow) poensis poensis (Strickland) ” praepectoralis Jackson granti Alexander > 2? > ” XXVIII. Emarginata sinuata (Sundevall) * schlegelii benguellensis (Sclater) x * namaquensis (Sclater) is oe schlegelii (Wahlberg) * i. 4 polluz (Hartlaub) XXVIII. Cercomela fusca (Blyth) % dubia (Blundell and Lovat)* < melanura melanura (Temminck) September 18, 1952 The Thrushes 31 Cercomela melanura neumanni nom. nov.°® >»? >> > 9 ”? >> ”? >? familiaris lypura (Hemprich and Ehrenberg) aussae Thesiger and Meynell* airensis Hartert ultima Bates* scotocerca scotocerca (Heuglin)* spectatriz S. Clarke furensis Lynes* turkana van Someren enigma Neumann and Zedlitz sennaarensis (Seebohm)* omoensis (Neumann) falkensteini (Cabanis) angolensis Lynes gambagae (Hartert) hellmayri (Reichenow) galtoni (Strickland) familiaris (Stephens) liibberti (Reichenow) a) sordida sordida (Riippell)* erlangeri Reichenow* schoana Neumann ernesti Sharpe hypospodia Shelley olimotiensis Elliott* XXIX. Saxicola rubetra (Linnaeus) macrorhyncha (Stolicza) insignis (Gray) oh dacotiae dacotiae Meade-Waldo 2? >? 29 ? murielae Bannerman torquata theresae Meinertzhagen hibernans (Hartert) rubicola (Linnaeus) maura (Pallas) desfontainesi Blanchet* archimedes Clancey* variegata (Gmelin) indica (Blyth)? 6 The name erlangeri Neumann and Zedlitz, is preoccupied in the genus Cercomela. 7I include indica, as I believe the breeding form of the Himalayas differs from the Palaearctic race, maura. This whole species seems in need of critical examination and revision, but I have not seen enough material. 32 2? a? >? > >> Postilla Yale Peabody Museum Sazicola torquata przewalskii (Pleske) yunnanensis (LaTouche) stejnegeri (Parrot) jebelmarrae Lynes » a avillaris (Shelley) promiscua Hartert moptana Bates* nebularum Bates adamauae Grote salax (Verreaux) sibilla (Linnaeus) voeltzkowi Grote ¢ tectes (Gmelin)* delacouri David-Beaulieu* caprata bicolor Sykes® burmanica Baker caprata (Linnaeus) francki Rensch cognata Mayr aethiops (Sclater) belensis Rand jerdoni (Blyth) ferrea Gray gutturalis gutturalis (Vieillot) luctuosa Bonaparte* armenica Stegmann* ig i leucura (Blyth) gabrielae Neumann and Paludan “\ i felix Bates albofasciata Riippell pallidigula (Reichenow) robusta (Tristram) torquata (Linnaeus) ankaratrae Salomonsen nilgiriensis Whistler af iy atrata (Blyth) albonotata Stresemann pyrrhonota (Vieillot) i? 7 fruticola Horsfield No. 18 wahgiensis Mayr and Gilliard 8 Stresemann (Ibis, 1952, 94:520), shows that Gmelin’s name must be used rather than borbonensis Sclater. 9I include rossorum (Hartert) in bicolor, vide Dementiev (Systema Avium Rossicarum, Paris, 1935, p. 254). September 18, 1952 The Thrushes 30 XXX. Pentholaea albifrons albifrons (Riippell) 4 pachyrhyncha Neumann frontalis (Swainson) limbata Reichenow clericalis Hartlaub melaena (Riippell) 9 > 29 9? XXXI. Thamnolaea cinnamomeiventris cinnamomeiventris (Lafresnaye) albiscapulata (Riippell) subrufipennis Reichenow bambarae Bayes cavernicola Bates* coronata coronata Reichenow ‘ ‘a kordofanensis Wettstein semirufa (Riippell) XXXIT. Chaetops frenatus frenatus (Temminck) As aurantius Layard DOCK TTT. Myrmecocichla tholloni tholloni (Oustalet) 3 chaboti (Menegaux and Berlioz) bifasciata (‘Temminck) aethiops aethiops Cabanis ‘i by buchanani Rothschild “s « sudanensis Lynes cryptoleuca Sharpe formicivora formicivora (Vieillot) 3 minor Roberts nigra nigra (Vieillot) ss ” stoehri Sclater arnotti leucolaema Finsch and Reichenow 9? 2? collaris Reichenow arnotti (Tristram) harterti Neunzig XXXIV. Ocenanthe tractrac tractrac (Wilkes) * ik albicans (Wahlberg) isabellina isabellina (‘Temminck) ‘3 ie bottae (Bonaparte) frenata (Heuglin) 2? 9? 34 Postilla Yale Peabody Museum No. 13 Ocnanthe isabellina heuglini (Finsch and Hartlaub) ” campicolina (Reichenow) xanthoprymna xanthoprymna (Hemprich and Ehrenberg) chrysopygia (de Filippi) kingit (Hume) oenanthe leucorhoa (Gmelin) ‘ nivea (Weigold) oenanthe (Linnaeus) virago Meinertzhagen Ai seebohmi (Dixon) phillipst (Shelley) deserti oreophila (Oberholser) ” —— atrogularis (Blyth) deserti (Temminck) homochroa (Tristram) hispanica hispanica (Linnaeus) 4 melanoleuca (Giildenstidt) finschit finschii (Heuglin) i barnesi (Oates) picata (Blyth) lugens boscaweni Bates ” — lugentoides (Seebohm) persica (Seebohm) lugens (Lichtenstein) ” halophila (Tristram) vaurei Meinertzhagen* lugubris lugubris (Riippell) a schalowi (Fischer and Reichenow) monacha ('Temminck) alboniger (Hume) pleseliazing pleschanka (Lepechin) cypriaca (Homeyer) leucopyga leucopyga (Brehm) ernesti Meinertzhagen leucura leucura (Gmelin) ” —— riggenbachi (Hartert) syenitica (Heuglin) monticola monticola Vieillot iF atmorii (Tristram) # albipileata (Bocage) moesta moesta (Lichtenstein) ‘4 theresae Meinertzhagen 2? >)? > 2? ? September 18, 1952 The Thrushes 35 Oenanthe moesta brooksbanki Meinertzhagen ‘ pileata pileata (Gmelin) ‘ i livingstonii (Tristram) XXXV. Sazicoloides fultcata cambaiensis (Latham) stuartbakeri Koelz intermedia Whistler and Kinnear fulicata Linnaeus’® leucoptera (Lesson) XXXVI. Prunella collaris collaris (Scopoli) ii iN subalpina (Brehm) sj if montana (Hablizl) ii ‘; rufilata (Severtzov) whymperi (Baker) nipalensis (Blyth) fe ‘4 tibetana (Bianchi) ripponi Hartert kwenlunensis (Buturlin)* erythropygia (Swinhoe) himalayana (Blyth) rubeculoides rubeculoides (Moore) iy 4 muraria (Meinertzhagen) strophiata sirotensis Koelz* 2? >? a i jerdoni (Brooks) i + strophiata (Blyth) ij ‘+ multistriata (David)* ? montanella (Pallas) fulvescens fulvescens (Severtsov) i ef dresseri Hartert dahurica (Taczanowski)* 2? 9? +) i ocularis (Radde) if a fagani (Ogilvie-Grant)* 4 atrogularis atrogularis (Brandt) Si a huttoni (Moore) koslowi (Przewalski) 10 Stresemann (Jbis, 1952, 94:521), points out that fulicata must be restricted to Pondichery on the peninsula of India. Ptymatura, restricted by Whistler (Jour. Bomb. Nat. Hist. Soc., 1935, 38:286) to Pondichery thus becomes again a synonym of fulicata. 11 Rufirenter Swainson 1831, based on “Le traquet 4 queue striée” of Levaillant (Oiseaux d’Afrique, pl. 188) is hereby restricted to Pondichery also. This leaves Lesson’s name (1840) for the Ceylon bird. 36 Postilla Yale Peabody Museum Prunella modularis modularis (Linnaeus ) 9 a occidentalis (Hartert) hibernica Meinertzhagen lusitanica Stresemann enigmatica Dunajewski* orientalis (Sharpe) . obscura (Hablizl) rubida (Temminck and Schlegel) immaculata (Hodgson) True THRUSHES XXXVII. Monticola rupestris (Vieillot) 22 explorator explorator (Vieillot) 2? brevipes brevipes (Waterhouse) 9? hebrideum Meinertzhagen tenebriformis Clancey* No. 13 pretoriae Gunning and Roberts* rufocinereus rufocinereus (Riippell) >? sclateri Hartert tenuis (Friedmann) angolensis Sousa saxatilis (Linnaeus) gularis cinclorhynchus (Vigors) ” — qularis (Swinhoe) rufiventris (Jardine and Selby) solitarius solitarius (Linnaeus) hy longirostris (Blyth) scortecciti Moltoni* behnkei Niethammer* magnus (LaTouche) > 2? ea philippensis (P. L. S. Miiller) pandoo (Sykes) madoci Chasen XXXVIII. Myiophoneus blighi (Holdsworth) melanurus (Salvadori) glaucinus castaneus Ramsay ”? > borneensis Sclater robinsoni Ogilvie-Grant horsfieldii horsfieldii Vigors glaucinus (Temminck) September 18, 1952 The Thrushes 37 M gaciphonens horsfieldi insularis Gould caeruleus turcestanicus Zarudny fs = teminckii Vigors eugenei Hume caeruleus (Scopoli) crassirostris Robinson dicrorhynchus Salvadori v flavirostris (Horsfield) XXXIX. Geomalia heinrichi heinrichi Stresemann a yi matinangensis Stresemann XL. Zoothera schistacea (Meyer) ii dumasi dumasi (Rothschild) He ” — joiceyi (Rothschild and Hartert) interpres interpres (Temminck) i ‘ leucolaema (Salvadori) minima (Hachisuka) erythronota erythronota (Sclater ) ie p dohertyi (Hartert) ir i mendeni (Neumann) a wardi (Blyth) cinerea (Bourns and Worcester) peroni peronii (Vieillot) ‘a ” — audacis (Hartert) everetti (Sharpe) sibirica sibirica (Pallas) " a davisoni (Hume) citrina citrina (Latham) a ” — eyanota (Jardine and Selby)” melli (Stresemann) aurimacula (Hartert) se ” innotata (Blyth) andamanensis (Walden) gibson-hilli_ (Deignan) > > »” a albogularis (Blyth) »” 2 rubecula (Gould) 3 ” orientis (Bartels)* aurata (Sharpe) piaggiae piaggiae (Bouvier) 12 Comparison of material in New York inclines me not to recognize the race, amadoni (Biswas), 1951. Postilla Yale Peabody Museum No. 13 Zoothera piaggiae hadii (Macdonald) * kilimensis (Neumann) rowei (Grant and Mackworth-Praed)* y if williamsi (Macdonald) * oberlanderi (Sassi) gurney gurneyi (Hartlaub) otomitra (Reichenow) usambarae (Neumann) raineyi (Mearns) * chuka (van Somern)* princet princei (Sharpe) cameronensis (Sharpe) graueri (Sassi)* at “a batesi (Sharpe) crossleyi crossleyi (Sharpe)* ‘4 % pilettei (Schouteden)* naevia naevia (Gmelin) ih ” — meruloides (Swainson) pinicola (Sclater) spiloptera (Blyth) ws talaseae (Rothschild and Hartert) margaretae (Mayr) andromedae (‘Temminck) mollissima whiteheadi (Baker) is i simlaensis (Baker) griseiceps (Delacour) mollissima (Blyth) ‘3 dizoni (Seebohm) dauma imbricata Layard x ” — neilgherriensis (Blyth) dauma (Latham) miharagokko (Momiyama)* toratugami (Momiyama)* aurea (Holandre) major (Ogawa) horsfieldi (Bonaparte) choiseuli (Hartert) eichhorni (Rothschild and Hartert) papuensis (Seebohm) % ” — machiki (Forbes) ii ” — -heinet (Cabanis) lunulata (Latham) 2? 2? >? a? > > a> 2? September 18, 1952 The Thrushes 39 XLI. XLII. RIT: XLIV. XLV. Zoothera dauma halmaturina (Campbell) 2? macrorhyncha (Gould) monticola monticola Vigors a 7 atrata Delacour and Greenway* marginata marginata Blyth cH 7 parva Delacour i terrestris (Kittlitz)* Amalocichla sclateriana sclateriana De Vis a a occidentalis Rand incerta incerta (Salvadori) = a olivascentior Hartert . = brevicauda (De Vis) Cataponera turdoides turdoides Hartert a ne tenebrosa Stresemann f i abditiva Riley x m heinrichi Stresemann Nesocichla eremita eremita Gould ‘ i gordoni Stenhouse* Cichlherminia Vherminierit Vherminieri (Lafresnaye)* ra a lawrencit Cory dominicensis (Lawrence) sanctae-luciae (Sclater) 22 2”? 2? >> Catharus gracilirostris gracilirostris Salvin ¥ fg accentor Bangs griseiceps russatus Griscom a ‘ griseiceps Salvin phaeopleurus Sclater and Salvin aurantiirostris clarus Jouy of is melpomene (Cabanis) bangsi Dickey and van Rossem costaricensis Hellmayr aurantiirostris (Hartlaub) ci * birchalli Seebohm insignis Zimmer fuscater hellmayri Berlepsch % . sanctae-martae Ridgway fuscater (Lafresnaye) caniceps Chapman >) 3? ? bed 29 > 2? 40 Postilla Yale Peabody Musewm No. 13 Catharus fuscater mentalis Sclater and Salvin occidentalis olivascens Nelson > > fulvescens Nelson occidentalis Sclater alticola Salvin and Godman frantzii Cabanis >? 2? 2? dryas dryas (Gould) maculatus (Sclater) mexicanus (Bonaparte) cantator Griscom fumosus Ridgway fuscescens fuscescens (Stephens) > salicicolus (Ridgway) fuliginosus (Howe)* 22> minimus minimus (Lafresnaye) Fi bicknelli (Ridgway) ustulatus ustulatus (Nuttall) 2? swainsoni (Tschudi) 5 almae (Oberholser) " oedicus (Oberholser) * guttatus guttatus (Pallas) 2? nanus (Audubon) 4 slevint (Grinnell) sequoiensis (Belding) polionotus (Grinnell) 4 auduboni (Baird) favoni (Bangs and Penard) crymophilus (Burleigh and Peters) mustelinus (Gmelin) XLVI. Turdus bewsheri bewsheri Newton ”? 3? comorensis Milne-Edwards and Oustalet olivaceofuscus olivaceofuscus Hartlaub >? xanthorhynchus Salvadori nigrilorum nigrilorum Reichenow >? poensis Alexander olivaceus chiguancoides Seebohm 2? saturatus (Cabanis) adamauae Grote* bocagei (Cabanis) centralis Reichenow pelios Bonaparte September 18, 1952 The Thrushes 41 Turdus olivaceus stormsi Hartlaub * si williami White graueri Neumann swynnertoni Bannerman smithi Bonaparte* transvaalensis (Roberts) * olivaceus Linnaeus abyssinicus abyssinicus Gmelin’ ‘i i baraka (Sharpe) oldearni Sclater and Moreau* bambusicola Neumann polius Mearns* elgonensis (Sharpe) deckeni Cabanis uluguru Hartert roehli Reichenow ¢ 4 helleri (Mearns)* nyikae Reichenow* milanjensis Shelley libonyanus verreauxi Bocage if ‘e tephrinus Oberholser™* tropicalis Peters niassae Rensch* costae Rensch* tephronotus tephronotus Cabanis iy ‘ australoabyssinicus Benson* menachensis Ogilvie-Grant e ludoviciae (Phillips) litsipsirups simensis (Riippell) is litsipsirupa (Smith) késteri Neumann* stierlingi (Reichenow) fischeri natalicus Grote <4 ‘ fischeri Hellmayr* . i belcheri Benson* s dissimilis Blyth hortulorum Sclater unicolor Tickell cardis Temminck J} > 2? ”? 13 I am much indebted to Dr. Chapin for suggestions on the arrangement of this and the following species. 14 Replaces cinerascens Reichenow, a name which is preoccupied in the genus Turdus. Postilla Yale Peabody Museum No. 13 Turdus albocinctus Royle BS torquatus torquatus Linnaeus ss i alpestris (Brehm) amicorum Hartert ‘ boulboul (Latham) merula ticehursti Clancey ij ” merula Linnaeus cabrerae Hartert azorensis Hartert agnetae Volse* mauritanicus Hartert algirus (Madarasz) aterrimus (Madarasz) insularum Niethammer* syriacus Hemprich and Ehrenberg maximus (Seebohm) intermedius (Richmond) mandarinus Bonaparte sowerbyi Deignan nigropileus (Lafresnaye) spencet Whistler and Kinnear simillimus Jerdon 7 cs bourdillont (Seebohm) ‘ i kinnisti (Blyth) poliocephalus erythropleurus Sharpe i if indrapurae Robinson and Kloss léseri de Schauensee javanicus Horsfield biesenbachi Stresemann* x ‘i fumidus S. Miller i r whiteheadi (Seebohm) seebohmi (Sharpe) albiceps Swinhoe thomassoni (Seebohm) mayonensis (Mearns)* mindorensis Ogilvie-Grant nigrorum Ogilvie-Grant 2? > i" s kelleri (Mearns) au 4 malindangensis (Mearns)* A i celebensis (Biittikofer) hygroscopus Stresemann sterlingi Mayr schlegeli Sclater September 18, 1952 The Thrushes 43 Turdus poliocephalus deningeri Stresemann versteegi Junge keyssert Mayr aN # erebus Mayr and Gilliard papuensis (De Vis) canescens (De Vis)* heinrothi Rothschild and Hartert renellianus Mayr bougainvillei Mayr kulambangrae Mayr ‘e a vinitinctus (Gould) poliocephalus Latham xzanthopus Forster pritzbueri Layard albifrons (Ramsay) efatensis Mayr if ‘A becki Mayr malekulae Mayr whitneyi Mayr placens Mayr vanikorensis Quoy and Gaimard ruficeps (Ramsay) i M layardi (Seebohm) - m tempesti Layard xi “ hades Mayr vitiensis Layard samoensis Tristram chrysolaus orii Yamashina i“ % chrysolaus Temminck celaenops celaenops Stejneger Me # yakushimensis (Ogawa)* niveiceps (Hellmayr)* rubrocanus rubrocanus Hodgson ‘ x gouldi (Verreaux) kessleri Przewalski ‘i feai (Salvadori)* pallidus Gmelin obscurus Gmelin ruficollis atrogularis Jarocki * ie ruficollis Pallas naumanni naumanni Temminck i of, eunomus Temminck pilaris Linnaeus 4A Postilla Yale Peabody Musewm Turdus musicus coburni Sharpe > 15 Mayr (Ibis, a? musicus Linnaeus?® ericetorum hebridensis Clarke 2? VISCLUOTUS >” catherinae Clancey* ericetorum Turton philomelos Brehm nataliae Buturlin* viscivorus Linnaeus reiseri Schiebel deichleri Erlanger theresae Meinertzhagen* bithynicus Keve-Kleiner* transcaspius Zarudny* bonapartei Cabanis mupipennis mupipennis Laubmann 2? conquisitus Bangs* swalesi (Wetmore) aurantius Gmelin ravidus (Cory) plumbeus rubripes Temminck coryi (Sharpe) schistaceus (Baird) plumbeus Linnaeus ardosiaceus Vieillot verrillorum (Allen) (ee ae flavipes Vieillot venezuelensis (Sharpe) polionotus (Sharpe) melanopleurus (Sharpe) zanthoscelus Jardine leucops Taczanowski No. 18 chiguanco chiguanco Lafresnaye and d’Orbigny > anthracinus Burmeister nigrescens Cabanis fuscater cacozelus (Bangs) > ”? gigas Fraser quindio Chapman 1952, 94:532-534) suggests using musicus for the song thrush and iliacus for the redwing, on the basis of a reéxamination of Linnaeus’ Editions of the “Systema” and “Fauna Svecica,” a suggestion with which I am heartily in accord. Perhaps this question can be finally voted on and an opinion rendered by the International Commission on Zoological Nomenclature. September 18, 1952 The Thrushes 45 Turdus fuscater gigantodes Cabanis >? »”? > ockendeni Hellmayr fuscater Lafresnaye and d’Orbigny serranus infuscatus (Lafresnaye) 4 cumanensis (Hellmayr) atro-sericeus (Lafresnaye) serranus Tschudi nigriceps nigriceps Cabanis % subalaris (Seebohm) reevet Lawrence olivater olivater (Lafresnaye) ” — caucae (Chapman) sanctae-martae (Todd) ptaritepui Phelps* paraquensis Phelps* duidae Chapman roraimae Salvin and Godman maranonicus Taczanowski fulviventris Sclater falcklandu falcklandii Quoy and Gaimard ‘i magellanicus King rufiventris juensis (Cory) if rufiventris Vieillot leucomelas albiventer Spix 1 leucomelas Vieillot cautor Wetmore amaurochalinus Cabanis ignobilis differens (Nelson) plebejus Cabanis ignobilis Sclater goodfellowi Hartert and Hellmayr 5 debilis Hellmayr murinus Salvin ts arthuri (Chubb) lawrencii Coues fumigatus bondi Deignan personus (Barbour) aquilonalis (Cherrie) fumigatus Lichtenstein haurwelli Lawrence colombianus Hartert and Hellmayr parambanus Hartert obsoletus Lawrence >> 229 2? >? Postilla Yale Peabody Museum No. 13 Turdus haplochrous Todd f. nudigenis umbrinus Griscom i grayi Bonaparte tamaulipensis (Nelson) casius (Bonaparte) incomptus (Bangs) nudigenis Lafresnaye ‘3 ti extimus Todd maculirostris Berlepsch and Taczanowski jamaicensis Gmelin albicollis assimilis Cabanis uf "i renominatus Miller and Griscom rubicundus (Dearborn) leucauchen Sclater parcolor Austin* atrotinctus Miller and Griscom oblitus Miller and Griscom cnephosus (Bangs) coibensis Eisenmann* daguae Berlepsch paraguayensis (Chubb) a ‘ albicollis Vieillot crotopezus Lichtenstein contemptus Hellmayr spodiolaemus Berlepsch and Stolzmann « 4 berlepschi Todd phaeopygus Cabanis phaeopygoides Seebohm minusculus (Bangs) rufo- palliatus rufo-palliatus (Lafresnaye) graysoni (Ridgway) rufitorques Hartlaub migratorius migratorius Linnaeus ¥ nigrideus Aldrich and Nutt achrusterus (Batchelder) caurinus (Grinnell) propinquus (Ridgway) phillipsi Bangs confinis Baird September 18, 1952 The Thrushes AT LirerATURE CITED Baker, E. C. Stuart, 1924. The fauna of British India; Birds. 2d ed., 2:187. , 1930. Ibid., 7:100. Bannerman, D. and Priestley, J., 1952. An ornithological journey in Morocco in 1951. Ibis, 94:427. Bond, James, 1940. Check-list of birds of the West Indies. Acad. Nat. Sci., Phila., p. 108. Cabanis, J. L., 1847. Ornithologische Notizen. Arch. f. naturges., 13:186-256, 308-352. Chapin, J. P., 1948. The systematic position of Xenocopsychus ansorgei. Auk, 65:292. deBalsac, H. and Mayaud, N., 1951. Sur la Morphologie, la Biologie et la Systematique de Cercotrichas podobe (P. S. L. Miiller). Alauda, 19:137-151. Delacour, J., 1942. The whistling thrushes (genus Myiophoneus). Auk, 59 :246-264. , 1946. Les Timaliinés. L’Oiseau et la Revue Francaise d@ Ornithologie, 16:13. Delacour, J., 1946. Notes on the taxonomy of the birds of Malaysia. Zoologica, pt. 1, 31:3. , 1951. Commentaires, Modifications et Additions a le Liste des Oiseaux de l’Indochine Francaise, (II). L’Oiseau et la Revue Francaise d’Ornithologie, 21:30. and Mayr, E., 1945. Notes on the taxonomy of the birds of the Philippines. Zoologica, pt. 3, 30:112. Dorst, Jean, 1950. Considerations Systematiques sur les Grives du Genre Turdus. L’Oiseau et la Revue Francaise d’Ornithologie, 20:212-248. Hartert, E., 1902. Novitates Zoologicae. 9:325. , 1910. Die Vogel der Palirktischen Fauna. Berlin, Heft. 5, 6, pp. 640-761. Lack, David, 1946. The life of the robin. London, pp. 144-147. Lowe, P. R., 1923. Notes on some land birds of the Tristan da Cunha group collected by the “Quest” Expedition. Ibis, 11th ser., 5:523-529. and Amadon, D., 1951. A classification of recent birds. Amer. Mus. Nov., no. 1496, 42 pp. Meinterzhagen, R., 1951. Some relationships between African, Oriental, and Palaearctic genera and species with a review of the genus Monticola. Ibis, 93:451. Newton, Alfred, 1893. A dictionary of birds. London, I:72. Oliver, W. R. B., 1945. Avian evolution in New Zealand and Australia. Emu, 45:148. 48 Postilla Yale Peabody Museum No. 13 Parker, W. K., 1872. On the structure and development of the crow’s skull. Monthly Microscopical Jour., pp. 217-226, 253. , 1873. On the development of the skull in the genus Turdus. Monthly Microscopical Jour., pp. 102-107. Pycraft, W. P., 1905. Ibis, 8th ser., 5:1-24. Ridgway, R., 1907. The birds of North and Middle America. Bull. U. S. Nat. Mus., Wash., pt. 4, 50:1-179, 885. Ripley, S. Dillon, 1952. A new genus of thrush from eastern Africa. Postilla, Apr. 12, no. 12, 2 pp. Sclater, W. L., 1930. Systema Avium Aethiopicarum. 11:446. Seebohm, H., 1881. Catalogue of the birds of the British Museum. 5:640-761. and Sharpe, R. B., 1898-1902. A monograph of the Turdidae. London, 2 vols. Sharpe, R. B., 1879. Catalogue of the Passeriformes or perching birds in the collection of the British Museum. London, 4:2. , 1881. Catalogue of the birds in the British Museum. London, 6:368. » 1903. A hand-list of the genera and species of birds. London, 4:111-184. Stejneger, L., 1905. The birds of the genus Cineclus and their geographical distribution. Smith. Misc. Coll., 47:421-430. Vaurie, Charles, 1949. Notes on the bird genus Oenanthe in Persia, Afghan- istan, and India. Amer. Mus. Nov., no. 1425. ,» 1950. Variation in Oenanthe lugubris. Ibis, 92:540-544. Whitaker, J. I. S., 1905. The birds of Tunisia. London, pp. 61-64. Whitherby, H. F., Jourdain, F. C. R., Ticehurst, N. F., and Tucker, B. W., 1938. The Handbook of British Birds. 2:1. Wolters, H. E., 1950. Akad. Verlag. Gust. and Portig. K.-G. Leipzig. p- 1130. pe iilla YALE PEABODY MUSEUM oF Naturat History Number 14 December 15, 1952 New Haven, Conn. A NEW RACE OF BLACK-THROATED BABBLER _ FROM ASSAM MUS. COMP. (GUL. | LIBRARY S. Ditton RiP.ey AR 24 1953 RAAVARD {| UNIVER SHY The Mishmi Hills in northeastern Assam, India, were so violently devastated by the great earthquake of August 1950 that whole hillsides for miles along the narrow steep valleys have been denuded of soil and vegetation. Centuries will be needed in some areas to restore even an approximate habitat for the fauna. That this fauna is in many respects unique was abundantly shown by the Smithsonian-Yale Expedition of 1946-1947, the results of which were discussed in “The Birds of the Mishmi Hills” by Mr. Salim Ali and myself (Jour. Bombay Nat. Hist. Soc., 48(1) :1-37, 1948). It is a sad fact that the fate of many of these little known bird and animal species will probably remain unknown for an indefinite period of time to come. Meanwhile, reéxamination of specimens of the Black-throated Babbler from the Mishmi Hills, at the suggestion of Mr. H. G. Deignan, prompts the recognition of another population of this species as follows: Stachyris nigriceps coei, subsp. nov. Type: é ad. (Y.P.M., No. 9585), collected January 4, 1947, by S. Dillon Ripley at Dreyi, Mishmi Hills, northeastern Assam, India. Diagnosis: from typical nigriceps of Nepal and the Hima- layas ranging east into northern Assam this form differs by generally darker tone of plumage, and by having a blackish unstreaked throat and very slightly darker ear coverts. From spadix of Cachar and the Khasia Hills this popula- tion differs by being notably darker with a more blackish throat and dark, really seal-brown, ear coverts. Compared with coltarti, the subspecies found in Margherita, the Naga Hills, and north Burma, coei differs by having dark brown rather than rufous-brown ear coverts, and by being a purer, less rufescent brown below. Range: Mishmi Hills, northeastern Assam, India. Remarks: in describing the subspecies, spadiz (Bull. British Ornith. Club, 68 :89-90, 1948), I left the Mishmi Hills popula- tion unnamed as an intermediate. Recent collections in the Naga Hills in 1950 of coltarti demonstrated anew the fal- lacy of this course and the necessity for recognizing this population. It gives me great pleasure, therefore, to name this new subspecies for Yale’s notable benefactor and collector of ornithologica, William Robertson Coe. MUS. COMP. ZOOL. LIBRARY a) | FEB 3 1954 ost A | papeaan YALE PEABODY MUSEUM oF NatTuraL History Number 15 May 12, 1953 New Haven, Conn. TYPOTHORAX. SCUTES FROM GERMANY JosEPH T. GREGORY Among a small lot of Triassic fossils from Wiirttemberg presented to Professor O, C. Marsh by Dr. Eberhard Fraas is a lateral dorsal armor plate of a pseudosuchian, Yale Peabody Museum no. 3694, which bears a small pyramidal spine. Similar plates were figured by H. von Meyer (1861, p. 341-342, pl. 43, figs. 4-7) and attributed to “Belodon.” These plates so closely resemble the corresponding portions of the armor of T'ypothorax meadei Sawin from the Dockum formation of Texas that familial or even generic affinity is suspected. Inasmuch as no horned pseudosuchian has hitherto been recognized from Europe, they deserve particular notice. Description: The spine was directed outward and _ back- ward, but rose little or not at all above the level of the reptile’s back, which suggests a thoracic position. The dorsal surface of the plate is slightly convex, triangular in outline as preserved. The medial portion is broken off so that its full width and precise shape cannot be determined. Traces of the smooth, narrow, anterior border which was overlapped by the plate ahead of it are present; the remainder of the dorsal surface is covered by a weak sculpture of ridges and grooves radiating from a pitted area above the junction of the lateral and dorsal portions of the plate, which might be termed the base of the spine (fig. 1). 2 Postilla Yale Peabody Musewm DeEscrIPTION OF ILLUSTRATIONS Left lateral thoracic dermal scute of pseudosuchian allied to Typothoraz, Y.P.M. no. 3694. From Keuper formation near Stuttgart, Wiirttemberg, Germany. All figures X 1. Figure 1. Dorsal surface Figure 2. Anterolateral surface The anterior and posterior edges of the spine are acutely angulate (fig. 3). An angular ridge also runs inward on the lower side of the spine for less than a centimeter from the apex and then disappears into the rounded surface which joins the anterior and posterior faces of the spine and merges with the lateral face of the plate. Typothorax scutes from Germany 3 FEB 3 1954 Figure 3. Posterolateral view Figure 4. Internal surface The lateral portion of the plate is very short antero- posteriorly but projects down conspicuously from the nearly flat dorsal portion, almost at right angles to the latter, below the base of the spine. Its medial surface (fig. 4) consists of an anterior inwardly directed narrow band which abruptly turns upward to merge with the posterior face of the spine. A shallow depression lies medial to the base of the spine at the posterior internal junction of the lateral and dorsal parts of the plate. The dorsal section is flat internally. | 4 Postilla Yale Peabody Museum Along the anterior face of the bone (fig. 2) a slight ridge branches from the angle between anterior and dorsal surfaces and roughly divides the face into equal parts, a lower with radiate sculpture and an upper “spine” area with weaker orna- ment of pits. The posterior face of the spine has a weak con- cavity leading to the internal surface at the angle between lateral and dorsal sections of the plate. Measurements: Length, normal to posterior margin ......... 60 mm. Length, anterolateral corner to tip spine .... 82 mm. Height, dorsal surface to tip lateral process .. 46 mm. Comparisons: The plates from Germany appear to differ from those of T'ypothorax meadei Sawin in the somewhat shorter spine which does not curve backward so markedly at its tip as do those of the lateral dorsal plates of that species. Also it lacks any indication of the faint dorsal ridge from the tip of the spine, which occurs on the Texas specimen. It differs from the lateral plates of Desmatosuchus in its smaller size, the broader base to the spine which is indistinguishable from above from the whole dorsal surface of the plate, and in the fine radial sculpture rather than coarse irregular pitting on the dorsal surface. These features are essentially those which distinguish T'ypothorax from Desmatosuchus. It would appear to differ from Stegomus as Typothorar does, in the much greater development of the laterodorsal spine and in the reduced size of the lateral portion of the plate. From Stagonolepis it apparently differs in the development of a strong spine and in the more acute angle between dorsal and lateral faces. This is somewhat uncertain, for although the limited assemblage of plates figured by Huxley did not include any like the lateral dorsals of Typothoraa, it is not impossible that such existed. The mid-dorsal plates of the two genera are quite similar. T'ypothorax differs from Stagonolepis in its ventral armor, which consisted of separate small quadrangular plates (Sawin, 1947, p. 232) instead of the articulating scutes of the latter genus (Huxley, 1877, p. 10-11). Typothorax scutes from Germany 5 Dermal “skin plate” armor was first associated with phy- tosaurs by H. von Meyer in 1861. At that time he figured a large number of the trapezoidal, longitudinally ridged plates which have since come to be known as mystriosuchid, and also a few elongate-rectangular plates bearing knob-like eminences and showing a coarser sculpture. On the same plate with these long plates (von Meyer, 1861, pl. 43) he illustrated a lateral dorsal plate extremely similar to the one described above. All of these specimens were referred to “Belodon’”’ (ibid, p. 337- 342), but it is clear that there was no association with the phy- tosaur skeletons; they merely were found in the Stubensand- stein which also produced the phytosaurs. E. Fraas (1896, p- 16) definitely described the elongate median dorsal plates and attributed them to Phytosaurus kapffi. Von Huene (1911, p- 103) affirmed the association of this type of plate with Phytosaurus kapffi, and contrasted them with the mystrio- suchid plates. In North America this quadrangular type of plate has been found principally associated with pseudosuchians; Desmato- suchus and T'ypothoraz have such plates as the median elements of their dorsal armor; they are found with a pseudosuchian type pelvis and vertebrae in University of Michigan Museum of Paleontology no. 13950, described as a “phytosaur” by Case (1932) at a time when that term was also applied to Desmatosuchus-like forms. Occasionally such plates have been found near phytosaur skulls (Camp, 1930, p. 89, and a plate, Y.P.M. no. 3695, found near a Machaeroprosopus gregori skull at San Jon, New Mexico), but never demonstrably in association with them. Camp has expressed his skepticism over the association of this type of plate with Phytosaurus kapffi, and the presence of these unmistakably pseudosuchian lateral armor plates in the Wiirttemberg Triassic strongly suggests that the specimens figured by von Meyer actually belonged to Typothoraz or a closely related genus which has otherwise escaped detection in the German deposits. Without more material it is impossible to decide whether the plates from the German Keuper represent Stagonolepis, Typothoraz, or another as yet unknown genus of pseudosuch- 6 Postilla Yale Peabody Museum ian. Consequently to propose a name for this almost unknown form at this time would be most improper. However, the exist- ence of such a creature seems well established; its presence serves to strengthen the faunal similarity between the late Triassic faunas of Europe and North America. BrsLioGRAPHY Camp, C. L., 1930. A study of the phytosaurs with descriptions of new material from Western North America. Univ. Calif. Mem., vol. 10, 174 p., figs. 1-49, pls. 1-6. Case, E. C., 1932. A perfectly preserved segment of the armor of a phytosaur with associated vertebrae. Univ. Michigan, Contrib. Mus. Paleontology, vol. 4, no. 2, p. 57-80, 6 figs., 8 pls. Fraas, E., 1896. Die Schwabischen Trias-Saurier nach dem Material der Kgl. Naturalien-Sammlung in Stuttgart zusammengestellt. Festgabe des K6niglichen Naturalien-Cabinets in Stuttgart zur 42. Versammlung der Deutschen geologischen Gesellschaft in Stuttgart, p. 1-18, pls. 1-6. Huene, F. von, 1911. Beitrige zur Kenntnis und Beurteilung der Para- suchier. Geologische und Palaeontologische Abhandlungen. n.f. Bd. 10, p- 67-121, pls. 12-19. Huxley, T. H., 1877. The crocodilian remains found in the Elgin sand- stones, with remarks on the ichnites of Cummingstone. Memoirs of the Geological Survey of the United Kingdom, Mon. III, p. 4-58, (quarto), pls. 1-16. Meyer, H. von, 1861. Reptilien aus dem Stubensandstein des oberen Keupers. Palaeontographica, Bd. 7, (1859-1861), p. 253-346, pls. 28-47. Sawin, H. J., 1947. The pseudosuchian reptile Typothorax meadei. Jour. Paleontology, vol. 21, p. 201-238, figs. 1-13, pl. 34. 2) vs Vth ye 4 ae be hi en sn 4 ie i ui : 1a ee te ete , PD Meh! Satis tae ap ig 4 alr n ¥ : i ae © » 4 , { a ufo af « *?. ‘ i eli 7 ipo a> H sd 1 7 By ' pe een! y ee 7 ) se ‘ Be Pet ) eee 24. 1a t's drome. ¢@ \#e . . o s = = ‘ Sis) ic : ‘ j Be ‘ 2 ? / a ’ 4 * T v \ ; . . . a ay / ws ve é P eo ude’ 7 j ae AY a 7 Poe 2 . . a e bs ne e y } e Pal ne patilla YALE PEABODY MUSEUM = =————— oF NaturaL History Number 16 June 3, 1953 New Haven, Conn. TYPOTHORAX AND DESMATOSUCHUS JosePH T. GREGORY Much uncertainty has been expressed by students of North American Triassic reptiles over the relationships and possible synonymy of the pseudosuchian genera T'ypothorax Cope (1875R), Episcoposaurus Cope (1887A), and Desmatosuchus Case (1920B). Still another genus belonging to this group, Acompsosaurus Mehl (1915), has been described, but its rela- tionships to the better known forms have never been adequately determined. In the course of preparing faunal lists of the Dock- um formation it was necessary to face the problem of nomencla- ture of these reptiles; the inadequacy of some of the early de- scriptions led me to examine Cope’s types (one of which had never been illustrated), and to compare these with the more com- plete specimens described by Case (1922B) and Sawin (1947). From this study it is evident that the type of Episcoposaurus haplocerus Cope is a specimen of the large, horned genus well known as Desmatosuchus Case and quite unlike EF. horridus Cope, the type of Episcoposaurus. Desmatosuchus, the type species of which becomes D. haplocerus (Cope), is a valid genus quite distinct from T'ypothorax Cope, which is best known from the Texas species 7’. meadet Sawin. Cope’s original type of Episcoposaurus horridus is hopelessly mixed with bones of other individuals, some of which were referred by him, and later by von Huene (1915A), to T'ypothorax, and which include char- 2 Postilla Yale Peabody Museum No. 16 acteristic dorsal armor of that genus. Although it is not demonstrable from the type material, the similar proportions and size of the limb bones originally described as Episcoposaurus horridus by Cope to those of T'ypothorax meadei Sawin, and the intimate association of these bones with larger armor plates of T'ypothoraaz, strongly suggests that E. horridus Cope actually is a synonym of T'. coccinarum Cope. The pelvis and associated fragments of Acompsosaurus wingatensis Mehl are clearly pseudosuchian but are not diagnostic portions for generic identification within this Family. There is some sug- gestion that they may belong to T'ypothoraa. Discussions of this problem with Professors C. L. Camp and E. C. Case have stimulated this attempt to solve a persistent taxonomic puzzle. It is a great pleasure to record the assist- ance rendered by many colleagues in the course of this study. Repeated opportunities to examine Cope’s types and other col- lections from the region of Gallina, New Mexico, at the Ameri- can Museum of Natural History have been given me by Dr. Edwin H. Colbert, with whom I have profitably discussed many aspects of this problem. Dr. Horace G. Richards of the Acade- my of Natural Sciences in Philadelphia graciously permitted me to study the type of Episcoposaurus haplocerus and to bor- row portions of that specimen for illustration. Through the kindness of Dr. E. C. Case of the University of Michigan, I have been able to study the type of Desmatosuchus spurensis closely and to benefit from his long experience collecting in the Triassic of western Texas. Professor A. S. Romer has permitted me to examine an undescribed skeleton of T'ypothorar in the Museum of Comparative Zoology at Harvard. The illustrations were prepared with great care by Miss Shirley Glaser. Bibliographic citations correspond to those used in O. P. Hay’s “Catalogue and Bibliography of Fossil Vertebrates of North America.” Museum locations of specimens are abbreviated as follows: A.M.N.H. American Museum of Natural History, New York City, New York A.N.S.P. Academy of Natural Sciences, Philadelphia, Pennsylvania June 38,1953 T'ypothoraxr and Desmatosuchus 3 M.C.Z. Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts [Gish University of Michigan, Museum of Paleon- tology, Ann Arbor, Michigan Y.P.M: Peabody Museum of Natural History, Yale University, New Haven, Connecticut Taxonomic History During the summer of 1874, E. D. Cope accompanied one of the parties of Lt. G. M. Wheeler’s Geographical Survey west of the 100th Meridian in northwestern New Mexico (Cope 1875U).' In the region of Gallina, New Mexico, he collected a few scraps of reptilian bones which he described (Cope 1875R) as T'ypothorax coccinarum Cope. Largely on the basis of these he correctly determined the age of the strata as Triassic. Cope’s original description of T’ypothorax mentions in order: a frag- ment of a jaw which he recognized as phytosaurian; dermal scutes; part of vertebral centrum; and the head of a femur. A phytosaur tooth also was associated. A second specimen in- cluding part of the top of a skull, pitted dermal bone like the type, and a single keeled scute was doubtfully referred to the species (Cope 1875R, p. 266). David Baldwin collected new material from the Gallina Creek locality in 1881 which formed the basis for a more detailed discussion of T'ypothoraa and the description of E piscoposaurus horridus by Cope several years later (1887A). Typothorax was diagnosed on the basis of dermal plates, ribs, and femur; the jaw fragment was excluded from the type, which Cope restricted to the dermal scutes with regular round pitting, figures 4, 5, and 9 of plate 22, Cope 1877K. This emendation of the type appears perfectly valid, for the restricted type is more homo- geneous than the original, yet is strictly part of it. Skin plates of large size attached to each other by matrix (Cope believed 1An illuminating account of this trip is given by G. G. Simpson in, “Hayden, Cope, and the Eocene of New Mexico.” Acad. Nat. Sci. Phila., Proc., 103, p. 1-21, 1951. ene ‘FEB 3 1954 HARVARD. 4 Postilla Yale Peabody Museum No. 16 them fused to ribs) and a complete femur of small size from the 1881 collection were referred to T'ypothorax; but smaller, keeled dermal plates, a large and massive femur, bones of the forelimb, and some vertebrae were made the type of Episcopo- saurus horridus. These specimens (A.M.N.H., nos. 2710-2713) are intimately mixed, partly still in matrix, the two types of dermal plates occurring together but not in original position. There is no evidence as to which limb bones are associated with each type of armor; if size is regarded as a criterion, the large femur of Episcoposaurus should belong with the T'ypothorax armor, and the smaller femur would be associated with the keeled plates. It is more than likely, however, that the two types of skin plates came from different parts of the same animal, and that the larger hind limb bones are associated with them. The small femora which Cope and von Huene regarded as T'ypothorax may well belong to one of the small phytosaurs which occur in the deposit, or else are those of somewhat smaller individuals of T'ypothorax. The size difference is easily accounted for on the basis of growth, the greater curvature of the shaft and twisting of the ends are harder to evaluate, for in these Triassic clays bones are frequently deformed in vari- ous ways. I do not regard the differences as proof of original diversity in form. Limb bones of M.C.Z., no. 1488, from the Ghost Ranch above Abiquiu in northwestern New Mexico are intermediate in size and similar to the “J'ypothorax” material in form. Alternatively, the large femur may be incorrectly as- sociated, though this seems unlikely in view of Sawin’s discoveries. Von Huene redescribed and figured the material in the American Museum in 1915 (Baldwin’s collection of 1881). He pointed out the difficulty of determining the association of bones and also of determining which specimens were in fact the types. By error he regarded the specimens which formed the basis of Cope’s 1887 redescription of T'ypothorawx as the type of that species, and included with it fragments of tibia, metatarsals, and scapula which Cope (1887A, p. 210) had re- garded as of uncertain reference. Although realizing the un- certainty of association of the various dermal plates and other bones, von Huene attributed to T'ypothoraxr a number of small June 38,1953 Typothoraxr and Desmatosuchus 5 scutes bearing conical central eminences (1915A, figs. 7-10) in addition to the flat, shallowly pitted plates. Some of these he regarded as caudal, others as lateral to the costals. (Com- parison with the articulated armor of M.C.Z., no. 1488 and T'ypothorax meadei Sawin shows that they are from proxi- mal and medial parts of the tail.) These plates are sculptured with ridges and grooves radiating outward from the central boss exactly like those which von Huene and Cope attributed to Episcoposaurus (cf. von Huene 1915A, fig. 24). Von Huene rejected the association of fore and hind limbs of Episcoposaurus on the grounds that the bones were too dis- proportionate in size; he restricted the type of E. horridus to the large femur and referred the small forelimb bones to a mystriosuchid phytosaur, possibly ““Belodon”’ scolopax Cope. However, the femur of 7’. meade is much longer and stouter than that referred to T. coccinarum. If Sawin’s figures 7 A and B are compared with von Huene’s figures 1 (Typothorax coc- cmarum) and 12 (Episcoposaurus horridus) it will be seen that the femur of the Texas pseudosuchian is far less curved than that attributed to Typothorar and much more like that of Episcoposaurus. Moreover the dimensions of the bones agree better with the latter genus. Measurements in millimeters of limb bones Typothorax Typothorax Episcoposaurus Typothorax coccinarum sp.? horridus meadei (AMNH 2710) (M.C.Z. 1488) (AMNH 2713) (Texas 31185-84) Length femur 22 25.8 31.5 (est.) 33 lacks condyle Length humerus Ry 17.5 22 21 Length radius ae ae ne 14 Length ulna Be 13.0 16.1 18 6 Postilla Yale Peabody Museum No. 16 The humerus and ulna of 7’. meadei are quite similar to those belonging to the type collection of Episcoposaurus horridus which von Huene (1915A, p. 493, 499, figs. 25-27) rejected as too small to belong with the femur of that animal. Their proportions suggest that Cope may have been correct in his reference of the small forelimb and large femur to the same animal (Episcoposaurus horridus). But the probable association of the Episcoposaurus femur (lectotype of E. horridus) with unmistakable armor of Typo- thorax, and the further likelihood that the supposed distinctive armor of Episcoposaurus is merely that of the tail rather than thorax or abdomen, and finally the intimate association of the type specimen of Episcoposaurus horridus with bones refer- red by Cope as well as subsequent students to T'ypothoraz coc- cinarum, all suggest that only one species is present. If this be so (it is probably incapable of absolute proof), Episcopo- saurus is a synonym of T'ypothorar having been established upon remains of the same species. Smaller femora referred by von Huene to T'ypothoraz coc- cinarum probably belong to younger individuals of that species, but in no case have any crucial bearing on the taxonomic problem as they are not primary types. W. F. Cummins of the Second Geological Survey of Texas began explorations of northwest Texas in 1889. In 1890 he found vertebrate fossils in the Dockum formation and in 1891 collected near Dockum the specimen which Cope described as Episcoposaurus haplocerus. Cope himself accompanied Cummins on a collecting trip along the east side of the Staked Plains in 1892, securing additional material including fragments of pseudosuchian armor which he referred to T'ypothoraxr (Cope 1893A, p. 17). The thick, coarsely sculptured armor plate and large lateral horns of E. haplocerus are obviously so similar to Desmatosuchus spurensis Case as to leave no doubt of specific identity. The type localities are only a few miles apart and at about the same level in the Dockum formation. In 1917 Professor E. C. Case of the University of Michigan discovered a pseudosuchian skeleton in Crosby County, Texas, which he described in 1920 as Desmatosuchus spurensis. Later June 38,1953 Typothoraxr and Desmatosuchus ii Case (1929B, p. 43) suggested that Desmatosuchus might prove to be a synonym of Episcoposaurus; Sawin (1947, p. 233) was also of this opinion. This view undoubtedly arose from com- parison with Cope’s description of E. haplocerus, which is indeed the same as Desmatosuchus. But this animal, which will have to be known as Desmatosuchus haplocerus (Cope), differs widely from Episcoposaurus horridus, the type species of the genus Episcoposaurus. As has been pointed out above, the latter is probably a synonym of T'ypothorax coccinarum., A full description of the type of E. haplocerus is given below, with comparisons to Case’s excellent account of Desmatosuchus. Desmatosuchus differs from T'ypothorax most obviously in the coarser and unequal pitting of the dermal armor plates and in the greatly enlarged horn over the shoulder region. In 1915 Dr. M. G. Mehl of the University of Wisconsin described the pelvis, sacrum, and other fragments of a pseudo- suchian from a “red shale series near the base of the Mesozoic section” at Fort Wingate, New Mexico, and called the fossil Acompsosaurus wingatensis. The specimen is not readily com- parable with the better known members of the group; some points in the description suggest T'ypothorax, others Desmato- suchus. 'There is no satisfactory indication that it represents a distinct genus. (I have not been able to examine this specimen. ) The extensive collections from Howard County, Texas, by the Texas Bureau of Economic Geology, W. P. A. Paleontolog- ical Survey, under the supervision of Grayson Meade in 1940 included two skeletons of a pseudosuchian described as T'ypo- thorax meadet by Sawin (1947). The species differs in minor details from T'. coccinarum Cope and provides by far the most complete picture thus far obtained of the anatomy of these reptiles. Other specimens of T'ypothoraaz have been obtained in eastern Arizona and northwestern New Mexico by the University of California, the Museum of Comparative Zoology, and Yale Pea- body Museum, but these have contributed little toward the un- derstanding of the family. Thus far Desmatosuchus has been found only in the Crosby County area of Texas. 8 Postilla Yale Peabody Museum No. 16 Systematic Revision Class REPTILIA Order THECODONTIA Family Stagonolepidae Genus Typothorax’ Cope 1875R Typothorar Cope. Acad. Nat. Sci. Phila., Proc. 27, p. 265, 1875; type species by original designation T'ypothorax coc- cinarum Cope. Episcoposaurus Cope. Am. Philos. Soc., Proc. 24, p. 213-217, 1887; type species by original designation Episcoposaurus horridus Cope. Quadrupedal archosaurian reptiles 214 to 3 meters long, with short, pointed heads, depressed bodies enclosed in dermal armor of overlapping bony plates, and with forelimbs much smaller than hind limbs. Skull pointed, flat-topped, overlapped by nuchal armor; the upper temporal opening laterally situated, lateral opening low, not completely known. A large antorbital fenestra. Mandible edentulous anteriorly and possibly covered by horny beak. Dorsal armor of two rows of overlapping plates, medial pair of plates flat, wider than long except in anterior cervical series, flat or with low conical or pyramidal eminence near center of posterior edge; lateral plates angulate with dorsal and lateral flanges meeting at sharp angle below base of projecting lateral spines; no enlarged, hornlike shoulder spines; surface of scutes covered with shallow, round, uniform sized pits about 14 cm. in diameter, except on smooth anterior articular flange and on bosses and spines which are covered with fine punctation; armor plates relatively thin (5 to 8 mm. thick). Ventral armor of small polygonal plates in regular rows narrower near the midline than laterally, with anterior flange for articulation with overlapping plates similar to dorsal series ; pitting of ventral plates similar to dorsal armor. Tail enclosed in rings consisting of four keeled plates each rising to an angu- 2The name Typothoraz is derived from the Greek rvzos, a model or image, and @dépat, breastplate, in allusion to the shallow pitting of the dermal scutes which resemble a hammered surface. June 38,1953 Typothorax and Desmatosuchus 9 lar posterolateral point; sculpture consisting of punctate sur- face on boss and weak pits or grooves radiating from boss over remainder of outer surface except articular flanges. Distal caudal plates elongate rods with posterior projecting spikes. Pelvis resembling that of Phytosauria but with ilium higher and more elongate anterior and shorter posterior iliac spines. Pectoral girdle poorly known, the glenoid an open, posteriorly directed notch as in crocodiles, coracoids rounded medially and not elongate as in Crocodilia ; dermal shoulder elements unknown. Humerus with ends expanded, at 45° angle, shaft slender; rear limb much longer and more massive than forelimb; femur nearly straight (slightly sigmoid) with strong 4th trochanter; tarsus with crocodiloid astragalus. Feet pentadactyl, toes of manus short and weakly clawed, those of pes with stout claws: metatarsal V hook shaped. Typothoraz is readily distinguished from Desmatosuchus by the absence of enlarged curved horns on the dorsal armor over the shoulder, by the regular round pitting of its armor, by relatively thin dermal plates, and by its slightly smaller size. Typothorax coccinarum® Cope 1875R Typothorax coccinarum Cope. Acad. Nat. Sci. Phila., Proc., 27: p. 265, 1875. Cope, E. D. 1877K, p. 29-30, pl. 22, figs. 1-9. Cope, E. D. 1887A, p. 210-213, pl. I. von Huene, F. 1915A, p. 485-490, figs. 1-10. Episcoposaurus* horridus Cope. Am. Philos. Soc., Proc., 24, p. 213-217. 3 The trivial name coccinarum was given by Cope from the Latin coc- cineus, scarlet colored, referring to the red-beds from which the specimen was derived. 4Cope gives no hint of the derivation of Episcoposaurus; two suggestions are possible. Latin episcopus, bishop + saurus, in allusion to the resem- blance of some of the conical caudal scutes to a bishop’s mitre. Alterna- tively, as Cope regarded the animal an ally of the phytosaurs, the literal Greek derivation ei, over + ckozely to look at + cavpa, cavpos, a lizard; a reptile which looks over or upward, in allusion to the high and upwardly directed orbits of phytosaurs is possible. The specific name, horridus, was derived from the Latin horrere, to bristle or tremble with dread, to be terrible. 10 Postilla Yale Peabody Museum No. 16 Type: U.S.N.M., no. 2585, dermal scutes (Cope 1877K, pl. 22, figs. 4, 5, and 9.) Collected by E. D. Cope, October 5, 1874. Type locality: “Triassic red beds of the western side of the Sierra Madre on Gallinas Creek” (Cope 1877K, p. 28). This site has been relocated by Camp (1930B, p. 143) as at Cerro Blanco, north of Gallina, New Mexico. It lies near the center of the N¥% sec. 9, T. 23 N., R. 1 E. New Mexico Principal Meridian. Chinle formation, Upper Triassic. Type of Episcoposaurus horridus: A.M.N.H., no. 2713 (formerly 2307). Two caudal vertebrae (proximal and distal); humerus; two ulnae; femur lacking condyles; proximal part of tibia; distal part of fibula; caleaneum; a number of dermal bones. Splenial possibly associated. Von Huene (1915A, p. 492-493) designated bones of hind leg as lectotype. From same locality as type of Typothorax coccinarum. Col- lected by David Baldwin, April 12, 1881. The only diagnostic features of the type of T'ypothoraz coc- cinarum are the thin, flat, dermal plates ornamented with numerous rather small, shallow round pits. A single keeled scute in the original collection was regarded by Cope (1875R, p. 266) as of uncertain reference. The later collection by Bald- win (A.M.N.H., nos. 2710-2713) contained both flat plates which Cope referred to T'ypothorar (1887A, p. 211) and keeled plates which he ascribed to Episcoposaurus (ibid. p. 216- aii). Both Cope (1887A) and von Huene (1915A) emphasized the difference in size and shape of the femora as distinctions between T'ypothoraz and Episcoposaurus. At first sight the massive straight femur of the latter appears quite different from the small sigmoid femora attributed by Cope to T'ypothoraz. But aside from the question of reference of these specimens, discussed on a previous page, the similar shape of the head of the two bones, and the intermediate character of the femora of Typothorax meadei Sawin and M.C.Z., no. 1488 makes refer- ence to the same species at least reasonable. The surprisingly small forelimb of “Episcoposaurus” is now known from T’, meadei to be characteristic of T'ypothoraz, and the difference in shape and pattern of the dermal plates appears to be con- trolled by their location on the body; the flat plates which Cope regarded as typical of T'ypothorax belonging to the median June 38,1953 T'ypothorax and Desmatosuchus 11 dorsal series of the back, the keeled plates of Episcoposaurus (fig. 17) belonging to the caudal series. Typothorax meadei® Sawin Typothorax meadei Sawin. Jour. Paleontology, 21, p. 201-238, 1947. Syntypes: Univ. Texas, Bur. Econ. Geol., Coll. no. 31185-84A, “frag- mentary skull, a poorly preserved vertebral column, appendages, and dermal armor susceptible of reconstruction from the anterior cervical region to the proximal caudal. Associated with this specimen were numerous small dermal buttons and plates referable to the ventral and appendicular armor,” and no. 31185-84B, fairly complete skull, portions of nuchal plates, fragments of dorsal plates, incomplete vertebral column, major limb bones, fragmentary remains of the girdles. Three other specimens referred. Collected by Grayson Meade and W. P. A. Paleontological Survey, 1940. Type locality: Univ. Texas, Bur. Econ. Geol., loc. 31185, Quarry 3A, 3 miles north of Otis Chalk, Howard Co., Texas, Dockum formation. These specimens belong to a rather wide and flattened animal with a short pointed pseudosuchian skull, prominent, backward- ly directed spines along the edge of the dorsal armor, large rear limbs and relatively weak forelimbs, and rather crocodiloid feet. Sawin (1947, p. 233) distinguished the species from T’ coccinarum Cope on the basis of (1) pyramidal instead of conical eminences on the median dorsal plates and (2) smaller size. A further distinction between the Texas specimen and Typothoraz coccinarum is the uniform presence of posteromesial eminences on the rear borders of the median series of plates in T’. meadet. On T’. coccinarum these plates are flat. As pointed out above, the limbs have the proportions and form of those of Episcoposaurus horridus. Sawin’s illustra- tions of the dorsal armor suggest a radial pattern on the median plates, and the presence of keeled bosses on these plates is distinctly more like the type of Episcoposaurus than that of Typothoraz. The admixture in this animal of the supposed characters of the two genera further supports the evidence of their identity. 5 The species was named for the collector, Dr. Grayson E. Meade. 12 Postilla Yale Peabody Museum No. 16 Our knowledge of the range of variation in these reptiles is far too meagre to permit a reasonable assessment of the biological validity of these species. It would be equally unwise to unite them in spite of these tangible differences or to flatly assert that the differences were unquestionably due to genetic isolation. The excellence of the Typothorar meadei specimens in comparison to other material of the genus is ample justifica- tion for retention of the specific name in the absence of proof of identity with another form. Typothorax cf. coccinarum Cope Typothorax is represented in collections of Yale Peabody Museum from the middle part of the Dockum formation west of San Jon, New Mexico. Two fragments of median dorsal plates with characteristic shallow pitting were found during the excava- tion of a Machaeroprosopus gregort skull (along with several other specimens which could not possibly have belonged to that animal), and the fragmentary weathered remains of much of a carapace (Y.P.M., no. 3696) were collected nearby. None of these plates show any trace of a boss or tubercle on the median series. Short posteriorly directed spines at the angles of the lateral plates are suggested by a few fragments. With grave doubts, a large thin median scute of a mid-dorsal series (Y.P.M., no. 3695) found near the same Machaeroproso- pus skull is referred to Typothorax (fig. 16). Its shape and thinness suggest this genus, but the ornamentation consists of radial ridges and grooves arranged around the very low round conical boss, which lies slightly behind the middle of the plate and rises less than a millimeter above its general surface. Inas- much as one of the typical T'ypothoraz plates mentioned above was found only a few inches from this plate, and as both show the same prominent anterointernal projection one may wonder whether the difference in sculpture is anything more than an artifact of preservation and preparation. The upper surface of the peculiar scute appears damaged. It (Y.P.M., no. 3695) is generally similar to those from the Keuper of June 38,1953 Typothoraxr and Desmatosuchus 13 Wiirttemberg attributed by von Meyer (1861, pl. 43, p. 3387- 342) and Fraas (1896, p. 16) to Phytosaurus kapffi. Rectangular median plates with similar sculpture charac- terize the specimen (U.M., no. 13950) from the Dockum forma- tion on Cerita de la Cruz Creek northwest of Amarillo, Texas, which Case (1932A) referred questionably to Phytosaurus. Pelvis and vertebrae of that specimen are of pseudosuchian rather than phytosaurian type, as Case realized (ibid. p. 71- 74) ; the dermal armor is more like T'ypothorax than Desmato- suchus and may be tentatively referred to the former. The posterior portion of a T'ypothoraz skeleton was collected from the Chinle formation at the Ghost Ranch on Canjilon Creek northwest of Abiquiu, New Mexico, by a party from Harvard University. Professor A. S. Romer most kindly per- mitted me to examine this specimen (M.C.Z., no. 1488; also other bones, no. 1487). It is important because the typical flat Typothorax plates of the body are associated with keeled scutes on the tail. Also its size is intermediate between the various specimens described by Cope as T'ypothorax coccinarum and Episcoposaurus horridus. The femur is 25.8 cm. long, essentially straight and stout like that of EF. horridus but with a well developed 4th trochanter. The tibia is very broad and massive, 12.5 cm. long. Another tibia from the same locality, no. 1487, is 13.5 cm. The humerus is 17.5 cm. long to the ulnar condyle; an ulna is 13.0 cm. If my interpretation of the speci- men is correct the ventral armor consists of transverse bands of plates which overlap in the same fashion as the dorsal armor, with an articular flange on the anterior external edge of each plate. Two narrow rows of plates along the midline are flanked by an uncertain number of wider scutes; all have shallow round pits like the dorsal armor, arranged in a faintly radial pat- tern about a central point. The latter is not raised as a boss above the surface. Camp (1930B, p. 3) reported T'ypothoraz locally abundant in the upper Chinle formation of Arizona and Utah, and rare in the lower part of that formation. Camp, Colbert, McKee, and Welles (1947, p. 4) list “Stagonlepis,” Typothorar, and Episcoposaurus in the Lower Chinle fauna of Arizona and Utah, and Typothorax in the Upper Chinle of northwestern 14 Postilla Yale Peabody Museum No. 16 New Mexico. I have collected characteristic armor of the genus from the lower Chinle near St. Johns, Arizona, but am not able to determine whether early and late species can be differentiated. Desmatosuchus® Case Desmatosuchus Case. Jour. Geology, 28, p. 524-529, 1920. Type species by monotypy Desmatosuchus spurensis Case = Episcoposaurus haplocerus Cope. Large quadrupedal pseudosuchians, 3 meters or more in length, with short-snouted skull, depressed body covered by heavy bony armor, the lateral plates over the shoulders pro- longed into curved, hornlike spines. Limbs and feet unknown but presumably crocodiloid. Dorsal armor distinguished from that of Typothorazx by its greater thickness, much coarser and less regular pitting of the exposed surface, and by the great elongation of the laterodorsal spine over the shoulder. Desmatosuchus haplocerus (Cope) Episcoposaurus haplocerus Cope. Am. Philos. Soc., Proc., 30, p. 129-181, 1892J. Wilson, J. A. 1950, p. 113-114, figs. 1-3. Desmatosuchus spurensis Case. Jour. Geology, 28, p. 524-529, figs. 1-4, 1920. Case, E. C. 19214, p. 133-147, pl. 3 (endocast) Case, E. C. 1922B, 26-48, figs. 7-20, pls. 5-10. Case, E. C. 1929B, p. 50-51, fig. 21. 6 The name is derived from the Greek décua, Séouaros, a band or fetter, and govxos, a crocodile, in allusion to the encircling bands of armor plates. The specific name spurensis was given for the town of Spur, Dickens Co., Texas; haplocerus comes from the Greek dos, simple, and xepas, @ horn, referring to the hornlike spines of the armor plates. Professor E. C. Case tells me that he formed the name by analogy with Desmatochelys Williston. June 38,1953 Typothorax and Desmatosuchus 15 Type of D. spurensis: U.M., no. 7476, skull; an associated skeleton belong- ing to the same individual includes the greater part of the vertebral column, ribs, fragments of the pelvis, and dermal armor of the back. Collected by E. C. Case in 1917 and 1919. Type locality: Near the east bank of Blanco or Catfish River, about one- half mile east of the crossing of the old mail road from Spur to Crosbyton, Crosby County, Texas. Type of Episcoposaurus haplocerus Cope: A.N.S.P., no. 14688; a sacral and two caudal vertebrae, right scapula, ribs, and about 30 demal plates. Collected by W. F. Cummins, July 20, 1891. Type locality: Near windmill in top pasture 3 miles north of Dockum, Dickens Co., Texas. Distinctive characters include the short-snouted pseudo- suchian skull with broad parietals and the temporal fenestrae lateral in position. It is obviously similar to, though not identical with, the skull of “T'ypothorax” meadei, which unfortunately is also difficult of interpretation in the postorbital region. The vertebrae differ from those of phytosaurs in the lower neural spines of the cervical, lumbar, and sacral regions, in the more projecting and lower parapophyses of the thoracic series, and in the somewhat lesser expansion of the dorsal ends of the neural spines in the thoracic area; the expanded tips of the spines are carried back farther than in Machaeroprosopus, however. Ribs are broad, stout, with a median supporting ridge run- ning along their internal surface. The pelvis is imperfectly known. A referred specimen, U.M., no. 7470, has a stronger anterior process of the ilium than phytosaurs. Pseudosuchian features are shown by the glenoid region of the scapula. Most distinctive of Desmatosuchus is the development of the dorsal armor with enlarged hornlike spines above the shoulder region. The armor consists of median and lateral paired plates. The median series of plates are rectangular, wider than long, and bear a smooth anterior facet over which the anterior plate moved, elsewhere they are ornamented by coarse, shallow pit- ting and by a median posterior raised boss. The lateral dorsal plates are angulate, extending from the median series out to the side of the back and thence downward along the flank. At the 16 Postilla Yale Peabody Museum No. 16 angle, a stout spine projects outward. These spines are long in the cervical region, reach a maximum in the curved shoulder horn, and then are abruptly shorter. It seems very likely that the ilium, U.M., no. 7322 (Case, 1922B, pl. 138A) from Sand Creek in Crosby County, Texas, and the pelvis and vertebrae, U.M., no. 7470, from the head of Holmes Creek, Crosby County, belong to Desmatosuchus. Case (1929B, p. 48-50) has pointed out reasons for such refer- ence; the different form of ilium in T'ypothorax meadei (Sawin, 1947, p. 218, fig. 5A) makes confusion with this form unlikely. Case (1929B, p. 43) and Sawin (1947, p. 233) have sug- gested that Desmatosuchus may prove to be a synonym of Epis- coposaurus. These statements seem to have arisen from compari- sons with E. haplocerus Cope. This species, attributed by Cope to his genus Episcopo- saurus, was based upon a dorsal and two caudal vertebrae, a right scapula, ribs, and 30 dermal plates, found by Cummins in 1891, near Dockum, Texas. Only the armor is at all com- parable with the other type material. It has never been figured, although Wilson (1950) gave drawings of topotype material (Texas Bur. Econ. Geol., no. 18569) which was supposedly part of the original specimen. Through the kindness of Dr. Horace G. Richards of the Academy of Natural Sciences in Philadelphia, I have been permitted to examine and illustrate the type material. To one familiar with these pseudosuchians the remains are obviously so close to Desmatosuchus spurensis Case as to leave no doubt of specific identity. The type localities are only a few miles apart, and at about the same level in the Dockum formation. The “single dorsal vertebra” mentioned by Cope (figs. 6, 7, 8) is a sacral, to judge by the massiveness of the rib which abuts against the side of the centrum as well as the short transverse process. Its centrum is slightly wider than tall, with moderately flaring, shallowly concave faces and an evenly rounded ventral surface without trace of keel. The neural canal was narrow and rather deeply grooved into the upper sur- face of the centrum in the middle position. On one side part of the heavy neural arch is preserved adjacent to the head of the sacral rib. This structure occupies the posterior half of the vertebra, and the rib facet stands out from the side of that body with smoothly curved outline. In front of it the side of the centrum is excavated, behind the flaring anterior June 3, 1953 Typothorax and Desmatosuchus Mi rim. This rim is marked by vertically elongate facets on either side which can only be interpreted as supports for the head of the expanded rib of the adjacent vertebra. Cope mentions the presence of rib facets at each end, a most unusual feature. As first sacral ribs are generally larger than the second, it seems most reasonable to assume that this is the second sacral vertebra and that these elongate facets supported the en- larged first sacral rib. The broad centrum, absence of twin ventral keels which characterize the second sacral of Machaeroprosopus, (Camp, 1930B, p. 65), and the curving upper surface of the transverse process and second sacral rib, suggest Acompsosaurus and the specimens U. M., 13950 and 7470 described by Case (1932A, p. 67-68, figs. 5-6). Unfortunately little of the sacrum was preserved in the type of Desmatosuchus spurensis, but this vertebra seems to differ from those of phytosaurs in a manner similar to other parts of the column of that animal. Two other vertebrae were considered by Cope to be caudals; one of these, which he described (1892J, p. 130), may well be a proximal caudal. Its transverse processes arise from the middle of the body of the centrum below the level of the neural canal. The ventral surface is flattened, and meets the lateral surfaces with an abrupt though rounded angle. These angular ridges terminate in facets for chevrons, as do those of phyto- saurs. Anterior and posterior faces are flat or nearly so, and as Cope indicates, somewhat taller than wide. The third vertebra in the collection, which was not described by Cope, consists of only a broken centrum, quite narrow and compressed, more like those of phytosaurs. Its association with the remainder of the speci- men is questioned. Great thickness characterizes the fragment of scapula (figs. 4, 5) which is all of the shoulder girdle preserved. Cope noted the normal inward curvature of the ventral portion, below the glenoid and acromion. How- ever, the bone is not necessarily thinner here, as he said, for a sub- stantial portion of the medial surface is broken away. Likewise the coro- coid suture was undoubtedly more extensive than the small area preserved next to the glenoid. A prominent tubercle for the long head of the triceps muscle lies 8 cm. above the glenoid on the posterior edge of the blade. Above this the blade widens and thins, rather symmetrically. Thinness of the dorsal edge, especially anteriorly, suggests that most of the blade is preserved. Comparison of this specimen with the fragment of the glenoid region of Desmatosuchus (Case 1922B, fig. 19) is difficult as there is little in common between them. The prominent acromion and thick oval base of the blade appear similar. It is also evident that the glenoid must have had something of the helical form indicated in Case’s figure. These features seem sufficient to establish the association of this kind of scapula with the more characteristic dermal plates. Numerous features of the scapula distinguish it from that of phytosaurs, especially slight concavity of the anterior profile, the short massive blade, and the pronounced acromion process. It differs from the scapula of Stagonolepis (Huxley 1877, pl. x, fig. 1) in greater thickness, much greater separation of glenoid and triceps tubercle, and stronger acromion. The scapula of Typothorax meadei as figured by Sawin (1947, fig. 3) ap- 18 Postilla Yale Peabody Museum No. 16 June 38,1953 Typothorar and Desmatosuchus 19 pears to be more expanded above, but also has a prominent acromion and triceps tubercle. These forms are by far closest to one another in char- acters of this bone. The rib fragments with flattened external surface and inner convexity are quite similar to those of Desmatosuchus (Case, 1922B, fig. 14), and separable from the narrower and more rounded ribs of phytosaurs. By far the most distinctive portions of the type of Episcoposaurus haplocerus Cope are the plates of dermal armor. Three transverse series of plates are figured herewith, and also two other isolated plates of different form. All plates are ornamented in their flatter portions by coarse, irregularly round pits up to one centimeter in diameter. The tuberos- ities and spines are coarsely punctate. No trace of radial arrangement of the sculpture can be detected. The plates are characterized by greater thickness than phytosaur plates of similar size, or than the plates of Typothorax. All are incomplete; the anterior and most of the posterior edges are broken away on the more anterior series so that the smooth articular flanges are not preserved; these are clearly shown by the more posterior plates. As Cope pointed out, the plates of each transverse row were suturally united. Each row consisted of 2 pairs of plates, the median flat, the lateral, angulate and spinebearing. Both Case’s reconstruction of Desmato- suchus and Sawin’s of Typothorax show the median series increasing in width posteriorly to the lumbar region and then gradually decreasing. In Desmatosuchus the cervical plates are thicker than those farther back, have more nearly a right angle between the dorsal and flank portion of the lateral plates, and greater ventral development of the lateral plates. The type of H. haplocerus agrees in showing a decrease in the angle between the dorsal and flank portions of the lateral plates as the median plates increase in width, and a reduction in thickness of the median series as they increase in width. Aligning the plates on these characters gives a progres- sive series almost suggestive of contiguity. By far the largest lateral spine is on the most anterior of these series; accordingly the preserved portion may be compared with the 5th to 9th series of Desmatosuchus as restored by Case (1922B). The right median and lateral plates are present in the most anterior preserved series which is that bearing the enlarged shoulder horn (figs. Desmatosuchus haplocerus (Cope). Type specimen of Episcoposaurus haplocerus Cope, A.N.S.P., no. 14688. All figures x 1%. Figure 1. Dorsal view posterior cervical series of dermal armor, bearing shoulder horn. 2. Median view of lateral, horn-bearing, plate of posterior cervical series, shown in figure 1. . Anterior view of same series as figure 1. . Posterior view of scapulocoracoid. Lateral view of scapulocoracoid. . Left lateral view of sacral vertebra. . Anterior view of sacral vertebra. . Posterior view of sacral vertebra. DAAnX PS w 20 Postilla Yale Peabody Museum No. 16 ANN - oon et ee emt eww ewe nee ee we rr me ee NN pes June 38,1953 T'ypothorax and Desmatosuchus 21 1-3). The median plate is longer than wide, bears a round tubercle 1% cm. in height posterior to its center, and is strongly concave from side to side ventrally, but convex anteroposteriorly below, particularly at the sutural edges which are lenticular in outline. The right lateral plate has very little dorsal extent, and this is entirely covered by the base of the horn (fig. 3). Its flank projection is extensive, and at right angles to the dorsal part. The horn itself rises in continuity with the lateral surface of the plate, its axis sloping outward at an angle of 25° from the vertical, and its medial edge reaching the suture with the median plate. Its base is longer than wide, and slightly flattened medially. Toward the tip of the preserved portion the beginning of a backward curve is apparent. In comparison with the large horn of the type of Desmatosuchus spuren- sis Case, this plate differs in the upward direction of the spine, and in its more rapid tapering, suggesting lesser length. Possibly these are in part due to individual variation. I am inclined to regard the angle of the horn as better established on Cope’s type than in Case’s specimen. The second preserved series, which may well fall next behind the first and thus be the sixth in Case’s animal, is represented by the left median plate and the conjoined right median and lateral plates (figs. 9, 10). A faint line of pores on the inner surface and of irregular small pits above mark the course of the fused suture between them. The medial plates are similar to that of the previous series save for slightly greater width. Their lenticular longitudinal section is shown in figure 11. The lateral plate, in contrast to that of the preceding series, has an obtuse angle between dorsal and flank portion, considerably greater dorsal extension, and a much smaller horn base. The lateral extent of the plate and length of horn cannot be safely inferred from the broken remains. The diameter of the horn is not greater than that in the 3rd series to be described below. This series differs from that lying behind the large horn of Desmatosuchus in retaining essentially the same thickness of plates. An incomplete median plate and articulating horn-bearing scute (figs. 12, 13) differ from the two preceding series in the appreciably thinner bone. Dimensions of this series suggest that it could have immediately followed the one just described. The bone is appreciably thinner than in plates of the cervical series, and does not thicken greatly at the sutures. The boss and sculpture of the median plate are quite similar to those of the last (?) cervical series, but the distance from boss to lateral border of the plate is greater. Also, the boss may be closer to the posterior edge, although Desmatosuchus haplocerus (Cope). Type specimen of Hpiscoposaurus haplocerus Cope, A.N.S.P., no. 14688. All figures x 1. Figure 9. Anterior view of anterior thoracic series of dermal armor, con- sisting of paired median and right lateral plates. 10. Dorsal view of same segment as figure 9. 11. Median view of right median plate of series shown in figures 9 and 10. 12. Anterior view of a more posterior segment of the thoracic armor. 13. Dorsal view of segment shown in figure 12. 14. Dorsal view of median plate from a more posterior dorsal series than figures 12 and 13. 22 Postilla Yale Peabody Museum No. 16 15. Machaeroprosopus sp. Phytosaur pelvis found near type locality of Episcoposaurus haplocerus by Cope in 1892. A.N.S.P. x ¥,. this is not certain as all edges are badly broken. The lateral plate bears a short conical spine directed both upward and outward. Its anteroventral and posterodorsal surfaces bear flattened facets in the portion preserved. The lateral flange of the plate forms almost a right angle with the dorsal portion, and appears to have been fairly extensive from the thickness of the broken edge. The suture between plates of this series is somewhat irregular, in contrast to the straight sutures between the cervical plates. Other incomplete median plates of the dorsal region are preserved, one of which is illustrated in figure 14. These plates show a pronounced depressed flange devoid of sculpture along the anterior margin; the round conical boss lies close to the posterior edge, and the sculpture is very coarse. The thickness, away from the boss, is about 2 cm. Included with the type of Episcoposaurus haplocerus is a phytosaur ilium (fig. 15) of the type figured by Case (1922B, fig. 27 C; 1927D, U.M., no. 7244). A field label accompany- ing it says “pelvis supposed to be Episcoposaurus haplocerus. Found 50 yards from type specimen. Windmill, Top Pasture. Coll.: E. D. Cope.” As Cope accompanied Cummins in 1892 this must have been obtained a year later than the type. Cope did not mention it in his description of E. haplocerus; the June 38,1953 Typothorax and Desmatosuchus 23 16. Cf. Typothorax sp. Median dorsal armor plate from Dockum formation west of San Jon, New Mexico. Y.P.M., no. 3695, x %. 17. Caudal scute from type material of Episcoposaurus horridus Cope, A.M.N.H. no. 2713. Specimen figured by von Huene, 1915, fig. 24. A. Dorsal, and B. medial views. x %%. distance of 50 yards is too great to permit association with the remainder of the specimen, and its form is unlike that which has been found more intimately associated with pseudosuchian remains. The ilium probably belongs to Machaeroprosopus (Camp, 1930B, p. 78-79, fig. 16) which occurs abundantly in 24 Postilla Yale Peabody Museum No. 16 the Dockum of this area. The length of the spine of ilium is 219 mm. Desmatosuchus haplocerus differs from Typothorax meadei in: 1. Thicker armor plates, especially anteriorly. 2. Coarse pitting of plates with no trace of radial arrange- ment. 3. Tuberosities and horns rounded instead of angular. 4, Larger size. 5. Median dorsal plates have central tuberosity behind middle of plate but not reaching posterior margin. 6. Fifth lateral plate, situated over shoulder, produced into long backward curving horn. It differs from “Episcoposaurus horridus” in: . Larger size. . Thicker armor plates. . No radial pattern to sculpture. . Bosses on median dorsal scutes not keel-like. oO Ff WwW NY Ee . Probably in presence of shoulder horns. There can be no possibility of generic identity between these forms. Acompsosaurus’ Mehl, 1915 Acompsosaurus Mehl. Science, n.s.,41 , p. 735, 1915. Type species by monotypy Acompsosaurus wingatensis Mehl. Sci- ence, n.s.,41 , p. 735, 1915. An imperfectly known genus which resembles Desmatosuchus in form of pelvis but has T'ypothoraaz-like dermal plates. Pos- sibly a synonym of T'ypothoraw. 7The name Acompsosaurus is derived from the Greek ay, lacking or without, xouwos, elegant, and gavpos, a lizard, hence a reptile lacking elegance. It was given, according to Mehl, because of the massive construc- tion of the pelvic girdle. The species was named for Fort Wingate, New Mexico. June 38,1953 Typothorar and Desmatosuchus 25 Acompsosaurus wingatensis Mehl Acompsosaurus wingatensis Mehl. Mehl, Science, n.s., 41, p. 735, 1915. Mehl, M. G., Toepelman, W. C., and Schwartz, G. M., 1916, p- 33-39, figs. 12-14, pl. III. Type: Univ. Wis., no. 3811, pelvic girdle and fragments of vertebral centra, ribs, dermal plates, phalanges. Collected by M. G. Mehl and G. M. Schwartz, 1914. Type locality: Region of Fort Wingate, New Mexico, in red shale series near base of Mesozoic section (no. 2 of section). The pelvis is characterized by a forwardly projecting spine of the ilium, deep vertical apronlike pubis, and moderately elongate ischium. The acetabulum is imperforate. Case (1929B, p. 51-52) has pointed out its resemblance to certain pelvic remains found in the Dockum formation of Texas; there is good reason to refer these to pseudosuchian, perhaps Des- matosuchus. The associated dermal plates, some of which are closely related to the ribs like those of T'ypothorax, resemble that animal in lacking any trace of the keels or spines such as are found on phytosaur plates, and also in their circular pitting. The pits are also described as deep, which suggests Desmatosuchus. Acompsosaurus may well be a synonym of T'ypothoraa, al- though the poorly preserved types make such determination difficult. Its relationships, at least, appear closer to T'ypothorax than to Desmatosuchus, if characters of the dermal plates are regarded as significant. The form of the pelvis differs from that of T'ypothorax meadei, but there may be errors in Sawin’s reconstruction of this region from incomplete materials. Possibly Acompsosaurus is really a third distinct type of pseudosuchian, but this seems most doubtful. Bibliography Camp, C. L., 1930B, A study of the phytosaurs with description of new material from western North America: Univ. Calif., Mem., v. 10, p. 1-174, pls. 1-6, figs. 1-49, 1 map. 26 Postilla Yale Peabody Museum No. 16 Camp, C. L., Colbert, E. H., McKee, E. D., and Welles, S. P., 1947, A guide to the continental Triassic of northern Arizona: Plateau, Mus. No. Ariz., v. 20, no. 1, p. 1-8. Case, E. C., 1920B, Preliminary description of a new suborder of phyto- saurian reptiles with a description of a new species of Phytosaurus: Jour. Geology, v. 28, p. 524-535, figs. 1-6. , 1921A, On an endocranial cast from a reptile, Desmatosuchus spurensis, from the upper Triassic of western Texas: Jour. Comp. Neurology, v. 33, p. 133-147, 3 pls. , 1922B, New reptiles and stegocephalians from the Upper Trias- sic of western Texas: Carnegie Inst. Wash., Pub. 321, p. 7-84, 14 pls., 33 figs. , 1927D, A complete phytosaur pelvis from the Triassic beds of western Texas: Univ. Mich. Contr. Mus. Geology, v. 2, no. 12, p. 227-229, 1 pl. , 1929B, Description of the skull of a new form of phytosaur, with notes on the characters of described North American phytosaurs: Univ. Mich. Studies, Mem., Mus. Paleontology, v. 2, p. 1-56, figs. 1-24, pls. 1-7. , 1932A, A perfectly preserved segment of the armor of a phyto- saur with associated vertebrae: Univ. Mich., Contr., Mus. Paleontology, y. 4, no. 2, p. 57-80, figs. 1-6, pls. 1-8. Cope, E. D., 1875R, The geology of New Mexico: Acad. Nat. Sci. Phila., Proc., v. 27, p. 263-267. , 1875U, Report on the geology of that part of northwestern New Mexico examined during the field season of 1874 by E. D. Cope, palaeontologist and geologist: Ann. Rept. Geogr. Explor. and Survey west of 100th Meridian, by G. M. Wheeler; Appendix LL, Ann. Rept. Chief of Engineers, p. 981-1017, (p. 61-97 of separate report) pls. ii, Vv, Vi. , 1877K, Report on the extinct Vertebrata obtained in New Mexico by parties of the expedition of 1874: Rept. Geogr. Survey west of 100th Meridian by Lt. G. M. Wheeler, v. 4, pt. 2, p. 1-365, pls. 22-83. , 1887A, A contribution to the history of the Vertebrata of the Trias of North America: Am. Philos. Soc., Proc., v. 24, p. 209-228, 2 pls. , 1892J, A contribution to the vertebrate paleontology of Texas: Am. Philos. Soc., Proc., v. 30, p. 123-131. , 1893A, A preliminary report on the vertebrate paleontology of the Llano Estacado: 4th Ann. Rept. Geol. Survey Texas, 1892, pt. 2, p. 1-137, pls. 1-23. Fraas, E., 1896, Die Schwibischen Trias-Saurier nach dem Material der Kgl. Naturalien-Sammlung in Stuttgart zusammengestellt. Festgabe des Koniglichen Naturalien-Cabinets in Stuttgart zur 42. Versammlung der Deuteschen geologischen Gesellschaft in Stuttgart, p. 1-18, figs. 1-10, pls. 1-6. June 38,1953 Typothorax and Desmatosuchus 27 Huene, F. von, 1915A, On reptiles of the New Mexican Trias in the Cope collection: Am. Mus. Nat. Hist., Bull, v. 34, p. 485-507, figs. 1-64. Huxley, T. H., 1877, The crocodilian remains found in the Elgin Sand- stones, with remarks on the ichnites of Cummingstone: Geol. Survey, United Kingdom, Mem., Mon. 3, p. 1-52, pls. 1-16 (quarto). Mehl, M. G., 1915, New reptiles from the Trias of Arizona and New Mexico: Science, n.s., v. 41, p. 735. Mehl, M. G., Toepelman, W. C., and Schwartz, G. M., 1916, New or little known reptiles from the Trias of Arizona and New Mexico with notes on the fossil-bearing horizons near Wingate, New Mexico: Univ. Okla., Bull., n.s., no. 103, p. 1-44, figs. 1-16, pls. 1-3. Meyer, H. von, 1861, Reptilien aus dem Stubensandstein des oberen Keu- pers: Palaeontographica, Bd. 7, p. 253-346, pls. 28-47. Sawin, H. J., 1947, The pseudosuchian reptile Typothoraw meadei: Jour. Paleontology, v. 21, p. 201-238, figs. 1-15, pl. 34. Wilson, J. A., 1950, Cope’s types of fossil reptiles in the collection of the Bureau of Economic Geology, The University of Texas: Jour. Paleontology, v. 24, p. 113-115, figs. 1-3. —— g 3 re | oem Jy. S j ARY FEB ef ac Perey Laie Ar YALE PEABODY MUSEUM =—————— oF Naruratu History Number 17 December 18, 1953 New Haven, Conn. NOTES ON INDIAN BIRDS. V* S. DILLON RIPLEY When my wife and I were in the Naga Hills in 1950, we collected two specimens of the Grayheaded Imperial Pigeon which I subsequently considered to represent the form griseicapilla, recorded by Baker (1928, Fauna or Bnririsu Inpi1a, 5:204) from southeastern Assam, and extreme eastern Bengal. The Imperial Pigeon was the only species of this genus seen by us in the Naga Hills, where pigeons of this impressive size and beauty now seem rare, no doubt due to the assiduous attentions of the Nagas themselves. Subsequent comparison of these specimens with birds both in the Peabody Museum collection and in the American Museum of Natural History in New York, from northeast Burma, Tenasserim, Thailand, and Indochina shows that the birds from the Naga Hills are distinct as follows: Ducula badia carolinae subsp. nov. Type: 2 ad. (Y.P.M., no. 12042), collected December 9, 1950, by S. Dillon Ripley at Phek, eastern Naga Hills, Assam, India. *Previous papers in this series have appeared in 1948, Jour. BomsBay Nat. Hist. Soc., 47:622; 1948, Zootocica, 33:199; 1950, Postirra, No. 1; and 1951, Postizua, No. 6. 2 Postilla Yale Peabody Museum No. 17 Diagnosis: From insignis, the subspecies of the eastern Hima- layas and adjacent foothills and the Khasia Hills, this subspecies differs by having the forehead and crown over- laid with gray, the vinous or lilac-gray color of the mantle reaching only to the hind nape and neck. The wing coverts and edges of the secondaries, and especially the lower back and rump are distinctly gray, a brighter, more light color than the mouse-gray, or dull brownish-gray of insignis. From griseicapilla, this form differs in the noticeably paler, more grayish tone of the scapulars, edges of the secondaries, lower back and rump. This lighter more pure gray tone seems to invade the tail also, the terminal band being paler, more pure gray, although this may be due to comparative age of the specimens examined. Certainly no other specimens in the long series examined by me throughout the species have as light pure grayish colors in the areas above listed. Measurements: 6 23; wing 242, 240; tail 181, 173; culmen 21.5, 25.5 mm. Soft parts: iris, gray; bill, coral or carmine- cherry basally, distally brownish-horn; feet, coral or carmine-cherry. Range: Eastern Naga Hills, and probably Cachar and Manipur south through the hills to east Pakistan as cited by Baker (op. cit.) for the westernmost range of griseicapilla, al- though no specimens have been available for comparison. Remarks: It gives me very great pleasure to name a new sub- species of this magnificent pigeon in honor of Mrs. William Robertson Coe. Among an interesting collection of birds sent to me recently by Mr. N. G. Pillai of the Travancore-Cochin government are four specimens of a pipit which I should like to describe as follows: Anthus similis travancoriensis, subsp. nov. Type: 2 ad. (Y.P.M., no. 23327), collected by N. G. Pillai on the road to Muthukuzhi, about 4500 feet altitude in the Ashambu Hills, April 15,.1952, Travancore-Cochin State, southern India. Notes on Indian Birds. V 3 Diagnosis: From similis similis of Bombay, Mysore, and Madras, this form differs in being uniformly darker above and below and with a much larger area of dark brown on the inner web of the penultimate tail feathers. The feathers of the upper surface are clove-brown edged, in fresh plumage, with dark tawny-olive. Below, this population is cinnamon rather than buff. The edgings to the outer tail feathers are darker, tawny-olive rather than wood-brown. In size there seems to be no difference. Measurements: Type—wing 89.5, tail 75.5, culmen 17.5 mm. A male molting into fresh post-juvenal plumage is too small to measure. Remarks: These specimens are so much darker than any other pipit belonging to the species found in India that I cannot understand how the single Travancore bird mentioned in the Ornithological Survey of that State (1936, Jour. Bompay Nar. Hist. Soc., 38:764) was not commented upon at least. Were it not for the similarity in plumage pattern with similis and the locality, it would be easy to confuse these birds with one of the dark African forms. In this connection it is worth pointing out that in the specimens of travancoriensis examined, there is an important difference from typical similis. Baker (op. cit., p. 278) in his key to the Indian pipits, separates trivialis, hodgsoni, and sordidus (= similis) from nilghiriensis on the basis of a very small pale tip to the inner web of the penultimate tail feather. The specimens of travancoriensis have large pale tips, nearly or more than a third of the total length of the tail. Of course they differ from nilghiriensis in the uniform tone to the plumage. The Rufous Babbler, T'urdoides subrufus, of Travancore— Cochin is a far more richly colored bird than the neighboring populations from the western Ghats of Bombay, Goa, parts of Madras and Mysore. Through the kindness of Mr. Greenway I have been able to borrow the type of Lafresnaye’s “*Timalia” poecilorhyncha (Neilgh the Museum of Compara- tive Zoology (Harvard)MUD Kis’ &peein T agrees adequately LIBRARY FEB 9. 1954 | 4 Postilla Yale Peabody Museum No. 17 with present-day specimens of typical swbrufus and so should be considered a synonym. However, Sharpe (1883, Car. Bos. Brit. Mus., 7:390) described Argya hyperythra from Madras on exactly the characters represented by my present series of Travancore—Cochin birds. I should like to revive this name, therefore, fixing the type locality at Palghat, and list the following populations: a). Turdoides subrufus subrufus (Jerdon), type locality, Wynaad. Synonym, Timalia poecilorhyncha Lafresnaye, type locality ‘“‘Neilgherries” hereby restricted to the northern slopes of the Nilgiris, as this species does not ascend to the summits of those hills. Range: Bombay in the western Ghats from Mahableshwar south, Goa, Coorg, western Mysore, and western Madras south to the northern slopes of the Nilgiri Hills, and east to the Shevaroys. b). Turdoides subrufus hyperythrus (Sharpe), type locality, restricted to Palghat. ‘Range: Southwestern Madras and Travancore-Cochin. ~e2 | osle i | aes YALE PEABODY MUSEUM or Natura History Number 18 March 6, 1954 New Haven, Conn. THREE NEW BIRDS FROM THE YUCATAN PENINSULA. RAYMOND A. PAYNTER, Jr.* In the course of an ornithogeographic survey of the Yucatan Peninsula the following new subspecies were found. It is be- lieved that the description of these three races completes the long list of Peninsular endemics, with the possible exception of several forms for which there is still inadequate material. Dendrocolaptes certhia legterst subsp. nov. Type: ¢ ad. (Y.P.M., No. 8471), collected March 4, 1949, by Raymond A. Paynter, Jr., at Carrillo Puerto, Quintana Roo, Mexico. Diagnosis: closest to D. c. sancti-thomae, the contiguous race, but considerably more pallid ventrally and slightly more pallid dorsally ; the distinctive rufescent tone of the under- parts is almost lacking, the gray of the chin is lighter, and the pileum is less richly rufescent. Range: known only from Carrillo Puerto and Tabi, in central Quintana Roo; probably ranges northward on the Penin- sula to the limit of the rain forest. *Museum of Comparative Zoology, Cambridge, Massachusetts 2 Postilla Yale Peabody Museum No. 18 Remarks: this is another example of a species characteristic of the rain forest which responds to the drier conditions of the Peninsula by becoming more pallid. Specimens from southern Campeche and southern Quin- tana Roo exhibit an approach toward this new form. It is a pleasure to name this race for Mr. D. B. Legters, a resident of Yucatan, who has been of inestimable assist- ance in the field and who collected many of the specimens used in these studies. Specimens examined: D. c. legtersi—four males and one female from Carrillo Puerto and Tabi, Quintana Roo. D. c. sancti- thomae—88 specimens, of both sexes, from Nicaragua. Honduras, British Honduras, Guatemala, and Mexico, in- cluding Veracruz, Oaxaca, Chiapas, Campeche, and southern Quintana Roo. Platyrinchus mystaceus timothei subsp. nov. Type: & ad. (Y.P.M., No. 13735), collected February 25, 1951, by Raymond A. Paynter, Jr., 24 km. NW. Xtocomo, Quintana Roo, Mexico. Diagnosis: nearest to P. m. cancrominus but dorsally lighter, more olive rather than brown; ventrally paler yellow, the breast band less well-defined, and the streaking reduced. Range: rain forest in Quintana Roo and Campeche, Mexico, in Petén, Guatemala, and in British Honduras. Remarks: specimens from Petén and from southern and central British Honduras are slightly less pale than those from Quin- tana Roo and Campeche. This race is dedicated to the memory of Timothy H. Laughlin, who spent the last months of his short life as my companion and assistant in Yucatan, and to whom I shall always be greatly indebted. Three New Birds from the Yucatan Peninsula 3 Specimens examined: P. m. timothei—six males and two fe- males from Agua Blanca, Km. 21 on the Chetumal-Bacalar Road, 46 km. W. Chetumal, 24 km. NW. Xtocomo, Car- rillo Puerto, Tabi, and Chacalal, Quintana Roo; one female and one unsexed from La Tuxpefia, Campeche; five males and two females from Uaxactun, Petén; three males and one female from Manatee Lagoon, Toledo District, and Cayo District, British Honduras. P. m. cancrominus—21 specimens of both sexes from Nicaragua, Honduras, Guate- mala, and Mexico, including Veracruz and Tabasco. Dumetella glabrirostris cozumelana subsp. nov. Type: ¢ ad. (Y.P.M., No. 8786), collected January 6, 1949, by Raymond A. Paynter, Jr., on Isla Cozumel, Quintana Roo, Mexico. Diagnosis: differs from the nominate form in having a longer and slightly heavier bill. Range: Isla Cozumel, Quintana Roo. Measurements: D. g. cozumelana—the culmen, from the base, of seven males ranges from 22.5 to 25.0 mm., with a mean of 23.64 + 0.29 mm.; one female 22.5 mm.; D. g. glabri- rostris—seven males from the mainland of Quintana Roo and from Half-moon Cay, British Honduras range from 21.5 to 22.0 mm., with a mean of 21.79 + 0.09 mm., and five females from Quintana Roo, Campeche, and Yucatéan range from 21.0 to 22.5 mm., with a mean of 21.60 + 0.26 mm. The difference between the means of the males of the two forms is statistically significant (P < 0.01). Remarks: I am unable to recognize any character which war- rants maintaining the monotypic genus Melanoptila for this species. In structure, size, and behavior this is merely an all-black species of Dumetella. Although not a strongly marked race, as are so many of those endemic to Isla Cozumel, its characters appear to be consistent. + Postilla Yale Peabody Museum No. 18 Slightly heavier weight may be an additional character of D. g. cozwmelana, although the data are insufficient to prove the suggestion. Four males from Isla Cozumel weighed 39.4, 40.3, 40.3, and 41.8 grams; a female 41.3 grams. Two males of D. g. glabrirostris from Quintana Roo weighed 36.8 and 38.1 grams; two females from Quin- tana Roo and one from Campeche 35.3, 36.1, and 31.6 grams respectively. Specimens examined: D. g. cozwmelana—seven males, one fe- male, and one unsexed from Isla Cozumel, Quintana Roo. D. g. glabrirostris—four males, three females, nine unsexed from Chetumal, 24 km. NW. Xtocomo, Carrillo Puerto, Ch’ich’, Isla Holbox, and Isla Mujeres, Quintana Roo; one male and two unsexed from “Yucatan,”’? Chichén Itza, and Xocempich, Yucatan; one female from 2 km. N. Agua- da Seca, Campeche; two males and two females from Belize and Half-moon Cay, British Honduras. For lending me necessary comparative material under their care I am obligated to E. R. Blake of the Chicago Natural History Museum, to H. Friedmann of the United States Na- tional Museum, to J. D. Macdonald of the British Museum (Natural History), to R. W. Storer of the Museum of Zoology, University of Michigan, and to J. T. Zimmer of the American Museum of Natural History. batalla ce YALE PEABODY MUSEUM or Naturau History Number 19 July 9, 1954 New Haven, Conn. BIRDS FROM GOUGH ISLAND S. Dillon Ripley Recently the Yale Peabody Museum has been fortunate enough to secure a small collection of birds from Gough Island in the South Atlantic Ocean, south of Tristan da Cunha. These specimens were secured through the intercession of Mr. R. Upton, now of the Bechuanaland Protectorate, formerly of Tristan da Cunha. It is of peculiar interest that these birds should come to Yale, as one of the first collections of birds from Gough, secured by Mr. George Comer, was formerly in the possession of Mr. G. E. Verrill, son of Professor A. E. Verrill of Yale; reported upon in the Proceedings of the Con- necticut Academy of Arts and Sciences (1895, 9:430-477), and certainly, at least, passed through the doors of the Pea- body Museum. Except for the type of the Gough Island gal- linule, now in the American Museum of Natural History in New York, the whereabouts of the Comer collection at the present time remains a mystery. The Museum’s grateful thanks are due to Mr. Wilmarth S. Lewis and to Mr. W. Sheffield Cowles for help in securing these interesting specimens. I am also grateful to Dr. Robert Cushman Murphy of the American Museum for showing me specimens in his care. 2 Postilla Yale Peabody Museum No. 19 Daption capensis (Linnaeus) : Cape Pigeon. A male was taken on Gough October 12, 1952. The species has not previously been recorded from the neighborhood of this island. Fulmarus glacialoides (A. Smith): Antarctic Fulmar. This species also is newly recorded from Gough. A male was collected September 13, 1952, and measures: wing 330, tail 120.5, culmen 43.5 mm. Pachyptila forstert (Latham) : Broad-billed Prion or Whale Bird. A female, September 5, 1952. Wing 204, tarsus 34, middle toe and claw 38 mm. Bulweria macroptera macroptera (A. Smith) : Great-winged Petrel. A male, December 15, 1952. Wing 302 mm. This specimen is somewhat more grayish about the throat and forehead than a July female from Tristan da Cunha. Although this bird was presumably not taken during the breeding season (July on Tristan), it appears to be the first definite record for Gough Island. Bulweria incerta (Schlegel) : Atlantic Petrel. A male and a female taken December 15, 1952, have wing measurements of: ¢ 318, 2 326 mm. These specimens are ap- parently the first recorded from Gough Island. Unfortunately the condition of the gonads was not stated. Compared to July specimens they appear to be in very slightly more worn, brownish plumage. Bulweria brevirostris (Lesson): Kerguelen Petrel. A male and female, December 15, 1952. These birds meas- ure: wing ¢ 257, 2 265; tail ¢ 110, 2107; exposed culmen } No. 19 Postilla Yale Peabody Museum 6 26.5, 2 28.5; tarsus 6 38.5, 2 37 mm. The collection of these specimens on Gough, although unaccompanied by data on their breeding condition lends credence to the original sup- position that the Kerguelen Petrel might breed in the neighbor- hood of Tristan da Cunha (Salvin, 1896, Cat. Bds. Brit. Mus., 25:410), and later reinforced by the collection of a female in January 1946 on Inaccessible Island (Roberts, 1948, Ann. Transvaal Mus., 21:60). Gough Island is far enough south (lat. 40°19’S.), to lie within the Subantarctic Zone, and the date of collection (December) corresponds to that for the sea- son of nestlings in down, on Kerguelen Island, the only present- ly known breeding place for this rare species. These birds match approximately in size those recently reported from Kerguelen by Milon and Juanin (1953, Oiseau, 23:17). Bulweria mollis mollis (Gould) : Soft-plumaged Petrel. Three specimens taken on December 15, 1952, measure: wing 6 249, 2 (2) 260; tail ¢ 111.5, 2 112, 120; culmen é 28, 2 28,29; tarsus ¢ 34, 2 36, 37 mm. Fregetta grallaria melanoleuca Salvadori: Tristan Storm Petrel. A pair taken December 15, 1952, have wing measurements of $ 171, 2 155 mm. Pelecanoides urinatrix dacunhae Nicoll: Tristan Diving Petrel. A female, December 15, 1952, measures: wing 117.5, tail 37, culmen 16 mm. This specimen confirms the occurrence of this subspecies on Gough Island. Puffinus assimilis elegans Giglioli and Salvadori: Tristan Shearwater. A female, December 15, 1952, measures: wing 190, tail 65, culmen 26, tarsus 42 mm. + Postilla Yale Peabody Museum No. 19 Gallinula nesiotis comert (Allen): Gough Island Cock. The Gough Island Cock, so-called, was originally described by J. A. Allen (1892, Am. Mus. Nat. Hist., Bull. 4:57) as differing from nesiotis of Tristan (now extinct?) in having greatly re- duced areas of white on the edges of the wings and the flank feathers. In size and structure the two forms appear to be so close that it seems useful to list them as subspecies rather than species. A pair and a downy young female were collected on Decem- ber 15, 1952. The adult birds measure: wing ¢ 145.5, 2 141.5; tail ¢ 63, 2 67; culmen (with shield) ¢ 42.5, 2 40; tarsus 6 48.5, 2 47.5; mid-toe with claw ¢ 64, 2 67. The soft parts of these birds appear to be as recorded by Clarke (1905, Ibis, 5(8) :258-259), the frontal shield and basal two-thirds of the bill being bright coral red, the distal third yellow. The legs in these specimens are red splotched with greenish yellow, the feet rather greenish yellow, the pads, nails, and posterior margins of the tarsi being blackish. The downy young bird, previously undescribed, is exactly similar to a downy young gallinule or moorhen. It is covered with black down and has black legs and feet. The bill is horny yellow, the upper mandible having a median black band and a black tip. The lower mandible is horny yellow, the basal half of the gonys and the tip being black. Allen (op. cit.:58) created the genus Porphyriornis for the Tristan and Gough Island gallinules on the basis of combining the short thick bill and oval nostrils of “Jonornis”=Porphy- rula, with the coloration of Gallinula. Actually the nostrils are oval, set in a nasal depression in Porphyrula, just as they are in Gallinula. The bill is stouter in the Gough and Tristan species than it is in a typical gallinule, but this is a feature of island species in any case, and its shape, and that of the frontal shield are virtually identical. These island birds, as well as the gallinules, both belong to the group which have narrow lateral membranes on the toes as pointed out long ago by Sharpe (1894, Cat. Bds. Brit. Mus. 23:6), and in fact the only striking morphological differences are the reduction in No. 19 Postilla Yale Peabody Museum 5 size of the wings in connection with flightlessness, and the heavier, more rugged appearing feet and tarsi, usually a corol- lary development. It appears to be a question, then, whether the genus Por- phyriornis should be maintained. Peters (1934, “Check-list of the Birds of the World,” 2:206, footnote) united Jonornis with Porphyrula feeling that the minor external differences of the species concerned were not of generic significance. Porphyrula differs from Gallinula in lacking lateral membranes on the toes, in having a bright plumage in the adult, differently colored young, and a posteriorly pointed frontal shield. These char- acters add up to a cumulative factor which may be considered to imply generic value. The sole character of “Porphyriornis,” aside from relative proportions, is flightlessness. Flightlessness, whether verified in fact or not, has not been thought of as havy- ing generic importance in the case of Rallus wakensis, for example, (Rothschild, 1903, Bull. Brit. Ornith. Club, 13:78). In the case of ducks, flightlessness is not considered to have generic value in itself. The Auckland Island and Campbell Island flightless teal have been made subspecies of the New Zealand Brown Duck, Anas aucklandica, by Delacour and Mayr (1945, Wilson Bull. 57:20, 39). Purple gallinules have been shown to occur rather commonly on the Tristan group of islands as occasional vagrants. An immature male and female in the Yale Peabody Museum col- lection were taken on Tristan in May and June 1952. (See also, Hagen, 1952, “Birds of Tristan da Cunha, Results of the Norwegian Sci. Exped. to Tristan da Cunha 1937-1938,” No. 20: 199-201). If Purple gallinules can wander from the New World so easily to Tristan, it seems quite possible that the Tristan and Gough Island gallinules represent an endemic population derived from vagrant Gallinula of New World origin. Rowettia goughensis (Clarke): Gough Island Bunting. Two pairs of this interesting species were secured December 15, 1952. They measure: wing ¢ 98.5, 100.5, 2 96 (worn), 6 Postilla Yale Peabody Museum No. 19 104; tail ¢ 76.5, 81, 2 81.5, (w.); culmen ¢ 18,19, 2 18,19; tarsus ¢ 29, 30, 2 27, 31 mm. On the label it is noted that the birds were seen from sea level to 1800 feet. Certainly in outward appearance Rowettia seems of New World origin, markedly similar in color pattern to Melanodera as pointed out by Lowe (1923 Ibis, 5(11) :511-513). vg aa i a1) 7 on hata @ a rte Fares Wee ath ee ta RL a 2 YALE PEABODY MUSEUM or Natura History Number 20 July 9, 1954 New Haven, Conn. NOTES ON INDIAN BIRDS. VI. ADDITIONAL COMMENTS ON THE WREN- BABBLER, SPELAEORNIS S. Dillon Ripley At the suggestion of Dr. Walter Koelz, I have assembled a number of specimens of Spelaeornis, the small Wren-babbler whose status I reviewed in 1950 (Awk, 67 :390-391), and again in 1952 (Jour. Bombay Nat. Hist. Soc., 50:492-494, and col. pl.). Dr. Koelz has kindly loaned me a series of 10 specimens, which with my own series and 16 specimens of chocolatinus and longicaudatus from the British Museum and five specimens of longicaudatus in the American Museum of Natural History has given me a total of 38 examples of these two species for study. I am most grateful to Dr. Koelz as well as to the authorities of the Institutions concerned for permission to examine this material. Dr. Koelz’ original question concerned my _ statement (1950 and 1952) that I had examined a specimen of Spelaeor- nis longicaudatus from Kedimai, Manipur, and that thus this species overlapped S. chocolatinus in range. As a result, being sympatric, they must be listed as separate species. Dr. Koelz felt (personal communication) that the differences between the species were so slight that they must be considered as all one. bo Postilla Yale Peabody Museum No. 20 On further examination I find I must stick to my original statement that these are two valid species, and it may be worth- while here to list the differences between them. A. §. longicaudatus. This species described from the Khasia Hills, occurs as far east as Manipur (one specimen known and reexamined). It is rather olive brown above, each feather particularly on the head and upper back, margined narrowly with black; the rump, tail, and outer edges of the wings and wing coverts tinged or edged with rich chestnut or rufous. In most, but not all, specimens (perhaps partly due to poor make- up of the skins) there is a small, buffy, ashy or white streak just over the eye. The lores, cheeks, and ear coverts are ashy. Below there is a small white chin spot and a number of the feathers of the lower breast and abdomen are terminally white, or tipped with white, particularly along the distal end of the shaft and inner margin of the feather. The effect is to produce an irregular patch of white on the abdomen. The rest of the underparts tend to be deep buff to ferruginous buff, rather rufous buff along the flanks. The feathers of the sides of the throat and upper breast have pale buffy shaft streaks, roughly elongated and diamond-shaped in outline. Higher on the sides of the neck, the feathers have subterminal buffy cres- cent-shaped spots, producing a slightly scaled appearance. B. §. chocolatinus. This species occurs from Cachar and the Naga Hills south to Manipur, the Chin Hills, Bhamo and the Shan States in Burma, north to Yunuan, and in Tonkin. The nominate form described from Kedimai, Manipur, differs from longicaudatus by being deeper, darker olive brown above, but with similar narrow black terminal margins on the head and back. The upper parts including the rump and upper tail coverts and tail tend to be rufous in females, which show con- siderable dimorphism in this regard. All sexed specimens which show this rufous suffusion are females. The lores, cheeks, and ear coverts, and a circumocular ring are ashy. The underparts show some variation, females tending to have the white areas reduced to a chin spot or small patch, and an abdominal patch similar to that in longicaudatus. The majority of specimens, No. 20 Postilla Yale Peabody Museum 3 however, 13 out of 17 or 76 per cent, including all the males and one female (in which the buffy throat patch is very ex- tensive and light in color), have large areas of white or palest buff on the throat extending continuously down to the abdo- men and belly, so that the underparts may appear largely white with ashy sides of the neck, and buff or olivaceous flanks and under tail coverts (see the colored plate, 1952, op. cit. in which the female is the upper figure, the male the lower). The feathers of the sides of the neck and flanks in typical chocolatinus have narrow white shaft streaks opening out near the tip to a narrow whitish fork enclosing a black terminal spot. The white or buffy feathers of the lower throat and breast have this same pattern, which on the white feathers is indicated simply by a terminal black spot. These differences are summarized below: Pattern of Species Upperparts Superciliary Underparts underparts A: brownish small, uniform, with reduced, pale usually white submen- shaft streaks present tal spot and on sides patch on abdo- men B. dark olive absent apparently sex- prominent or rufous ually dimorphic, streaks and brown ranging from black spots on largely white breast and (4), to large- sides ly buff (9 ) Measurements: wing-tail index, wing mean tail mean per cent culmen A. 20 (449 9) 49-55 5284041 45-55* 50.14.1386 90-100 11-14 mm. B. 16 (4 @ 9 9) 46-52 5014043 41.5-47.5 48.84 0.43 81-93 * One juvenal female from Mawphlang, Khasia Hills, in the Koelz collec- tion has a tail measurement of only 43.5 mm. 12-14 mm. 4 Postilla Yale Peabody Museum No. 20 Comparison of the means of the tail length of these two samples gives a figure for o, of 1.53 which is statistically signi- ficant although somewhat high. It appears, therefore, that there is a distinct likelihood that any specimen of, species B., i.e., chocolatinus, will tend to have a shorter tail measurement than specimens of the typical form of species A., i.e., longicaudatus. | Finally the examination of all available specimens of choco- latinus from Manipur and the Naga Hills has convinced me that sexual dimorphism is a prominent feature of this species, that the type and other known specimens of chocolatinus from Manipur, although unsexed, are most probably females. In addition Dr. Koelz’s series from the Naga Hills contains three strongly rufous colored females, far richer than any in my original series when I described nagaensis (Postilla, 1951, No. 6:4), with much reduced white areas on the lower parts. As these specimens are inseparable from chocolatinus, I feel that all should be combined under that name as follows: Spelaeornis chocolatinus chocolatinus (Godwin-Austen and Walden) Pnoepyga chocolatina Godwin-Austen and Walden, 1875, Ibis, 5(8) :252. (Kedimai, Manipur.) Elachura haplonota Baker, 1892, Ibis 4(6) :€2. (Hangrum, N. Cachar.) Spelaeornis chocolatinus nagaensis Ripley, 1951, Postilla, No. 6:4. (Mt. Japvo, Naga Hills.) Range—Assam in the hill ranges of north Cachar from Hangrum east to the Naga Hills in the Japvo area and east to Pfutsero at least, south to Kedimai in Manipur, presumably above 5500-6000 feet, in evergreen forest. oslt Ve nov 785 pz : YALE PEABODY MUSEUM |p 7 4 ° . st , P. \ if pte yk ( 5 ; 4 wy . "i 7 yo os » a ‘ 7 } LP s bf ry a sb, i . cA Y ely i), 5 ea ‘ ’ At | ote Ca iu! A ey bd > } Ul ; 4 4 : , , | a af A Oe UE a AS es ee Ps ue » Pra 1h om SORA YINE a! AMEN Pas Sect me | Angi Tgaisg Yi ae oi ais AA) Mabl ic att oy i. Wu as Loy wiped AO EI Main, THe Ww alee Hoe, a bat See Le A Lie) USER A Peet ye heen ede ate A eee Wap a: bi yea a fr rr et ONE eee Ceo Tun Z. f iT j VOR dh vi OUR bd A (jee Phebe: EVER GR Eee ah ih ee AL iY m i { v a nm iy iy . \ oe ead gaan ee . OM ali hu ge ete . eh is hs na Pa fi 1% iy iV "4 pnt 7 4 } J 7 j i; { a : Ww af vi Tey ‘ ays i vag Uae hy i ‘i hese , 7) ia : Y ‘ 4 ny ; a is i] 7 ey ' ve Wei: Pay Winer ua ¥ sale recite 7 , a BP UO tod rear if PAA Ue wall h, val AN yr Kune : ? reat WwAy. i olety al eeu The TPA le ae ee Fl folly Oe ve ee Ws hen j bale , hy * UA ais rR a HA wi r Ais ts) i river ie a i ey y AYA) LA Y ; / We oy " \ ix Tah i mA tH a es i ont vk \es Pi ay. ' wo? 28 VIA Ay eae, ae a Ve ’ ex = IN: re Por | AYE Ries bor HARVARD | UNIVERSITY ene YALE PEABODY MUSEUM oF NatTurAL History \ Number 27 September 28, 1956 New Haven, Conn. METEORITES IN THE COLLECTIONS OF YALE UNIVERSITY Kurt Servos! ABSTRACT More than 400 localities where meteorites fell are recorded in this first list compiled in more than fifty years. The list combines the catalogues of two major collections housed in Yale University: the Peabody Museum Collection and the Carl Bosch Collection, which has been catalogued and is now being published for the first time. INTRODUCTION H. S. Washington (1897, p. 83-87) was the last person to compile a catalogue of the meteorites in Yale collections. When he compiled that list the entire collection was a part of the Peabody Museum of Natural History and his list was essentially a revision and moderniza- tion of the catalogue that E. S. Dana had prepared in 1886. During the fifty years after Washington’s compilation of the catalogue some accessions were added to the collection which did not find their way into other catalogues. In 1949, however, a great ccllection of minerals, the Bosch Collection, was provisionally deposited in the University and the meteorites in that collection are here listed for the first time. A short note previously called attention to the noteworthy specimens in that collection (Servos, 1954). ACKNOWLEDGEMENTS Professor Horace Winchell advised me in the compilation of this list and to him I express my sincere appreciation. Professor Harrison S. Brown and Mr. Walter Nichiporuk, both in the California Institute 1 Present address: New York State Museum, Albany, New York. 2 Postilla Yale Peabody Museum No. 27 of Technology, and Dr. E. P. Henderson of the U. S. National Mu- seum gave valuable advice. Two undergraduate students, Richard F. LaGanza and Andrew N. Jergens, Jr., spent many hours working over the Peabody Museum Collection. The Department of Geology in Yale University generously gave financial aid in the form of the William E. Ford Scholarship and for this I express my profound gratitude. EXPLANATION OF THE List oF METEORITES The data in this list include merely the locality name, number of the specimen and the weight, with appropriate postillations in some cases. We have refrained from listing the synonymy, date of fall or find, and type of meteorite because that information is readily avail- able in such standard references as Prior’s Catalogue of Meteorites, as revised by Hey. The assignment of serial numbers to specimens was arbitrary and these numbers do not represent either the number of falls or the number of individual fragments, but rather the approximate number of accessions to the collections. The meteorites in the Bosch Collection are identified by the prefix M (here listed under each individual local- ity) ; those in the Peabody Museum Collection have the prefix P; and the few meteorites in the Brush Mineral Collection are distinguished by the prefix B. The guide for nomenclature of localities is Hey (1953). The synonymy established in that catalogue is followed here. Some obvious geographical errors appearing on the original labels have been cor- rected in making this list although, of course, the basic catalogue of the collections shows also the original data. Prior’s Catalogue was useful in resolving such errors. Many of the specimens in the Bosch Collection were originally acquired by trade or purchase from other collections or from supply establishments. In order to keep errors and confusion at a minimum, when these specimens can be traced directly to their source, the weights reported for these specimens are those given on the labels even though they may disagree slightly with the present weights of the specimens. For specimens that have more than one label showing different weights, the lower one is recorded here. The specimens in the Peabody Collec- tion, on the other hand, were re-weighed and any discrepancy between the weight recorded here and the weight of the corresponding specimen in a previous list indicates that a portion of the specimen has been withdrawn from the Collection for trading or other use. Unless other- wise indicated, weights are given in grams. Sept. 28,1956 Meteorites in the Collections of Yale Univer Srrciric Nores In REFERENCE TO THE LIST Mivcduiil Three pallasites from Mexico (M123, 2.29 g.; M551, 38 g.; and M552, 26.5 g.) in the Bosch Collection, found in 1893, have labels with insufficient information to assign them to specific localities. About 26 specimens in the Peabody Museum Collection lack labels which would permit them to be assigned to specific localities. _ The assignment of specimen P286 to Santa Rosa is made with more than a moderate amount of trepidation and uncertainty. The specimen was given by Wm. Huland, Esq. to Benjamin Silliman. H..S. Washington, in his Catalogue,’ wrote that “G.J. Brush suggests that this is a specimen of the Otumpa Iron” but the original label has not been recovered. In an older catalogue, compiled in December 1868 by Professor G. J. Brush,” this specimen is described “Marked Meteoric Iron found in Columbia So. Am from Wm. Huland Esq London to B. Silliman. May this not be a specimen of the Tucuman (Otumpa) Iron? G.J.B.” Specimen P421 was acquired by Professor George R. Wieland in Mexico and was donated by him. Its assignment to Morito is uncertain but probably correct and is based on Wieland’s reference to it as the “Humboldt Iron.’ The total weight of the Homestead fall in the Peabody Collection is 36,252 grams; the largest individual of this fall weighs 11,960 grams. St. Augustine’s Bay, Madagascar (P22) is listed despite the fact that it does not appear in most modern catalogues. Because informa- tion in the literature concerning the St. Augustine’s Bay, Madagascar (P22) iron is scanty, the find is here tentatively listed under meteorites. The specimen labelled Feroe Islands, North Sea (P204) is listed tentatively with the meteorites although precise information is lacking. ‘We consider Asheville (P20a) and Black Mountain (P20b) a “paired” fall but we list them separately because the information on their original labels does not permit more specific assignment. Both the Bosch Collection and the Peabody Museum Collection contain some specimens of pseudometecrites which are not listed here poe the information they offer has questionable value. * Catalogue of the Collection of Meteoric Irons, Yale University, Peabody Museum. 1896. (MS in the Peabody Museum.) - *-Meteoric Iron in Mineralogical Cabinet of Yale sea 1868. eke in the Yale Peabody Museum.) > Reference made at the time of donation. 4 Postilla Yale Peabody Museum No. 27 WEIGHT WEIGHT LOCALITY IN GRAMS’ LOCALITY IN GRAMS Apmire, Lyon County, Kansas, ArisPE, Sonora, Mexico U.S.A. M288! . 255). 2.00 ates ae 20 MEL O42). Ste Stata tele ota cale ahebe miey sie eye 73.8 W402 is «cists, oys. rave als ever e eenetete 590 IE ASSIS nhs baie eels Ge pewrote' = Gieoaels 85 Artincton, Sibley County, IWADGE Boi cks aie copesttc enlaces 400 Minnesota, U.S.A. Acen, Lot-et-Garonne, France IPD 28) We caisee ote clelasele cue rete eterna 10 EMO one wierd aterm eet cas toale(e es 0.80 AsHEvILLE, Buncombe County, IPO85. sale cake gees tte eee qT North Carolina, U.S.A. Arxswortn, Brown County, P20a (vial with fragments) . .36 Nebraska, U.S.A. Avusurn, Lee County, Alabama, a We as ee nee ee 11.2 U.S.A. P162 (1 pc—90 g., 1 vial— | o:) Creer 2 OE Benet Read eis 93 AvcustinovKA, Ekaterinoslav, Axsarpour, Saharanpur district, Ukraine tadin MG08" 2. dens) See 36.4 2 ys a ey ee or Ree CT OE er 1 1 Vl POE TOO CCG aoe 6.2 Anant, Modenaseltaly MUG: 6.0: ss e:oars acsicteyee rere eae 130 BilOS bs bn 28,c5 f Ie eae 0.9 SEED co Pn ae ee = IE ALS ye cons haze vaaconeh 15 W228 ges odes ee 80.25 BrenHam Townsnip, Kiowa County, M235. ..):003. 0 ee eee 31.88 Kansas, U.S.A. MBB? «ic a0 tee Sane ghee 122 VAST Dia aeate cierartecias siceehs ceases oe 249.2 M253*.c..Len Ree 605 1 Se Pee Ie ec 119 PUOBD ce casersssictaxciccoretescncnele eee 937 IMRT cena oe sera ote iwah sora onsa vests 28.55 Pldse: hats eee 323 ie eee eo eet A ee 175 PIO8d <5... 5s 318 INLD Oe oreo oe te ALON coe oe, s 113 Piss “0. eee Al IVE er cite eons, of afiows mickacoselews tere ow eXe 1.7 Pi0st. Co eee 26 P1O1 (4 pes—266 g., 567 g., PLOS¢° 2.5565. 27. 2 eee 500 701.5 g., 1263.5 g.). .2798 Posh" 5. 6s oc Lee ee 841 IPAOT Desert 2 oes: cet ce eee 21.5 P103i (bottle with POL (Gast) s. oe o-oo 4.8 shavings) ......... 1056.5 IPZOAE TELS Om ctl Sart eee or 3671 P108j..-(Hilings), : 34.4, 3ee 20.5 LBS he Mn os Sacrum DS aaa 101.5 Ibs. Pl0sk (filings) ......::ceemen 10 Serer scvateistehe. sete fs stars Seeks es 20.55 P1031 (shavings) ........... 20 Brincewater, Burke County, P103m\, (shavings) toce eee 25.5 North Carolina, U.S.A. P103n (bottle with MOLD E Raped crisintias oki tet OULD Shavings) oie:sre aie 1660 LEONG Awe berets Gaile eid atoree 19 P1030 (graphite nodule) ...721 Bur-Guetval, Bur-Hagaba district, PAO nee eee shee 374.6 kg Italian Somaliland BGISS&: vans lnwsk ott tee eee 429.5 WEGAT oS ate eine hace ae e,F nee 17 Canton, Cherokee County, Burtineron, Otsego County, Georgia, U.S.A. New York, U.S.A. PIL «el cilitiaton eieeerees 82 PU Geis shee etace ek 723 Cape Grrarpeau, Cape Girardeau Buscuuor, Zemgale, Latvia County, Missouri, U.S.A. MBS ATE (EAR). 27k oaks ewer Be 2.6 P56 (10 fragments, 18 g., WEG wists ctoyat aie nse piaiae sede ters 1.2 main mass, 1460 g.) 1478 Sept. 28, 1956 WEIGHT LOCALITY IN GRAMS Carre or Goop Hors, Cape Province, South Africa IVD A Tare sare Mar stare ha rerate ai akaebonates otoistos 6.79 1 Edy Orn CORO PRC NCR PRC A 8 Cartton, Hamilton County, Texas, U.S.A. UVES? (Gh ve sos asodconn oe 13.43 MUEUGS Bh Eh 5.6 880 ss) 3S a Pahe tae ne DE 57.5 LOOM te dee ettete ss hem oe ae aie 172.5 CartTHacE, Smith County, Tennessee, U.S.A. IMI Z BS orate ae ie lojeisterstare oto ters Petoievartt oe 12 MEQ DG ats neces fare icicles ae see OSs 35.7 LL I a RR LA 370 DA a res scbaravststagers Ehalwfeyenste’e ee oe 105 Castatia, Nash County, North Carolina, U.S.A. MISS itareN crs tio alte Mh ectass oat ns 0.2 ELT Baws s oiahoveisfovsiels, tert +... .248 Castine, Hancock County, Michigan, U.S.A. PAA GOR cha tee nis eae anton 16 CuHanpaKaPur, Berar, Central Provinces, India IEG Ga eer e cia ee GA ane AO 2 40 CHANTONNAY, Vendée, France MSD Ohere sre ree teen ae 89.59 MS9IGRE Rect re tone nie 176 EDR CEES) 3. teers + oie, sae 12.5 P283 (2 pes—3 g., 41 g.) ....44 Cuarnorre, Dickson County, Tennessee, U.S.A. AOS + (turMINgS)) WS. os. 1.5 CuHarsonvittE, Meung, Loiret, France INISG REE tae eer rs nh tS en 0.26 PABA ACA PCS) wes kia. eects es 16 Cuassicny, Haute Marne, France MVD Diese cte acne ni orate acne tare ate 18) CHATEAU-RENagpD, Montargis, Loiret, France 1 De Sa cerca pets Sows SOR a eae 41.34 IR I As eieahs CREUSET AIC ist cee ea 807 IVES 8 erent eek io Pesce ics eae 26 ZOO; (CS) nevis din. slave's wees 91 Meteorites in the Collections of Yale University “f WEIGHT LOCALITY IN GRAMS CHESTERVILLE, Chester County, South Carolina, U.S.A. 1, WT gee ae et prea gana 12.49 PA Poe Ree Gtatetets upckctesereronei oe 751 Cuinavutita, Guatemala IME GDS eter tn ere io) Cuvupaberos, Jimenez, Chihuahua, Mexico MASS ated ucdcnns leit rence eee 2375 Coanuinta, Mexico M250' “Kort Duncan” ©... ...- 66.63 M2b1 >) Bonanza, et eae 293 M254 “Santa Estate,” “COUCH ELEGHGs eis 53.0 INEZ DSA. oer tetas ot Mane eosin 189.7 49s Bonanza a. eno aa 48 J ELA ie ttt Seca Atel a te gy rl 1508 GAD Soir te oc isnte Rican ae 1502 PGA Cera. cae eb ae oe eae 2707 P88 “Sanchez Estate,” “Couch Iron” (38 g., BuO Retina eee AT P22, Kort Duncan?) = 2-54-01 29 P122 “Fort Duncan” (dust)..1 Cozisa, Pampa of Santa Barbara, Antofagasta, Chile MSOs carlin Re acs te ote eee 35.04 IBZ 3B al sar kere Fe eee Swen Bere 67 Corp Boxxkevetp, Cape Province, South Africa IVI AB ta cuca ARS Ao INA orate eet 1.6 P229 (2 pes—13.1 g., OG aoa). 1 oh tne 109.5 Cottescipout, Terni, Umbria, Italy M118 (2 pes—7.2 g., 10.09 g.) 17.29 CooKevitte, Putnam County, Tennessee U.S.A. IME GOR rece eset rcp ee aie es 85 Coorertow x, Robertson County, Tennessee, U.S.A. Zoillanen(QusCS))) oes ener 120 P23 TDs uci ce eave repaid aca tane eee 695 8 Postilla Yale Peabody Musewm WEIGHT LOCALITY IN GRAMS Cossy’s CreEEK, Cocke County, Tennessee, U.S.A. INEZ pe). aye iaseravere din) steve ciel svete (atenstate 1 12 I PO emine Wa nme Once Oo 917 P17b (small fragments) ...... 8.2 Pl7c (small fragments) ....262 Costi1ta Prax, Taos County, New Mexico, U.S.A. WWIGUY Hee SB a GiH dO SOO CRO. c 135 12 UFC eee Eo eG ree tater eRe Orc 172.5 Covert, Osborne County, Kansas, U.S.A. INGA Sie sienis acre chai. cine ateeioc 1365 P405 Cras Orcuarp Mountains, Rockwood, Cumberland County, Tennessee, U.S.A. IVE SNE ccvsre see eregas, State savecveieyete 115.6 UTR eC Rise a ete aie § oie arcs ee 8 P22 Rees ne ard carn oer savor 96 CranzBerry Prarns, Poplar Hill. Giles County, Virginia, U.S.A. WEG 2 eee rc suelo tors ae ie snatalers ais el 22 CranpourneE, Victoria, Australia WEDROM Says steselniers efic, Si\ereuale che fine ale 1 P134 “Beaconsfield” ........ 148 |g 1 i 0) Se 109 Cross Roaps, Wilson County, North Carolina, U.S.A. Ge rrave iaaate sere seyaueve ie =, 02 eee 8 Cuiuison, Pratt County, Kansas, U.S.A. MIT AG rage ccaiertelsc savavcisvtnnie mite etersae tare 85.6 CuMBERLAND F atts, Whitley County, Kentucky, U.S.A. IMAGO Rit. eites siteclneis ss ct oe 155 Datton, Whitfield County, Georgia, U.S.A. 30) GW ie Rae ence oa oe Cae PRE a 8.2 P94 Danpapur, Gorakhpur district, India a erates t alse create Saree 82 Danvitte, Morgan County, Alabama, U.S.A. OS ahs arene wlonecarekilec ote em ll No. 27 WEIGHT LOCALITY IN GRAMS Deep Sprines, Rockingham County, North Carolina, U.S.A. IME AG ais, 0 ao, ntokminles “ya sjevere ae ee 27.4 P42 Deport, Red River County, Texas, U.S.A. M264 \ 205 caioeaeaietlt Cte ee 300 DescusriporaA, Catorce, San Luis Potosi, Mexico MAIS oil ehiccc ects oan pees 363 P58 DuurmsaLa, Kangra district, Punjab, India MGR oo ridin vein 3 ie Og ae 217.5 MEAG oo. 60d sins sane ee eee 59 P2338 (2 pes)... .:<< eee 93 Dsati-PenciIton, Ngawi district, Java M69» 3.05 waned) «cee 1.27 P228 (2 pes—0.5 g., 240 g.) .240.5 Drake Creek, Nashville, Sumner County, Tennessee, U.S.A. MS4) 2c ost 053s ee eee 1.2 P220 (2 pes—4 g., 426 g.) ..430 Duet Hix (1854), Walnut Mountains, Madison County, North Carolina, U.S.A. W298. ools cistet cnyctstaane ace eee 14.17 PAG, «52 5, iches' stu.csee oietnrceei eae 4224 PADD, © icc ieirare.5)0% ass cheet ee 45 PAV CS soso cine oc 2s ho oe ee 22 Dunprum, County Tipperary, Ireland PG arwisponeteus ose. 6 onesie he eee 69 Duncannon, Scott County, Virginia, U.S.A. M262) Aico cero waco eee eee 250 Durata, Punjab, India MSGi as50.55.6 ae Pes se oe 0.4 P2is (Gopea)). 2.0.02 a aoa 2.5 Eacte Sration, Carroll County, Kentucky, U.S.A. M62) sn) hel eostearsoc ae ene eee 76 MATA ii 0 atten! gallo an ee 60 POD «an .ivG rete oe Ge tales Ae 70 Sept. 28, 1956 WEIGHT LOCALITY IN GRAMS EnRENBERG, Yuma County, Arizona, U.S.A. AO eet reine ores peter obiaya lel ne ented 11 E.socen, Bohemia MEZZO ares sel skecace chee aleve 5 8) aces loko Ooms 12 IMR Eis a oisis oar nrc ane 8 Ex Carrran Rance, Lincoln County. New Mexico, U.S.A. IEA esa vorscapens are svar-eniona¥ ov ors ior aleteners 59.84 P113 Exim Creex, Admire, Lyon County, Kansas, U.S.A. M76 (2 pes—7.34 g., 8 g.) ..15.34 PFT Oleerpeave escent rst oyevaveve oie 0 3) eho 14 EnsisHetm, Alsace, France VTE Sime Nerercsets cise teroveieneisyeceseiy acaithatays 1.2 JETS) (FE) TED) poop ourdiscosnee 18 Ereuero, Brava, Italian Somaliland, East Africa VEG Di yiies eset ae aneccte sn ust asoeue Woe 143.55 NVUAD Gros oyes tava Xove es oc’ s) eevehen so Rieke 183.5 USAGE are crit sce eile we oeraisins 232 IPA0G eeetice © tees oo See 57 Erxtesen, Magdeburg, Prussia AO Mamie Mere actos Sits. d coe 13 Estracapo, Hale County, Texas, U.S.A. IMG YS 7 AR a ears AE) ts et ee ae 162.87 IVTAS One ih fofasessa actoretraie oe sueteclele 310 VAT Ota thn velo ocioisitoe ssie mer tees 1.8 NEDLO! - SNR Tea Poet oe 1.27 ZOO) sais viarect sais eters eee ein ee 410 EsTHERVILLE, Emmet County, Iowa, U.S.A. eae eS PCS) Anco. Seek 40.26 OS Oe ee an ae 9.5 M477 (2 pes—S8 g., 22 g.) ....30 EIGBE G2 ES) Weise cs = 22 oA « 30 PGB EA CCTS” PES)! Sexee(scldsa se 48449 Baap) (IO pes yrs. Po .cc Oe 93.05 Farmincton, Washington County, Kansas, U.S.A. M70 Meteorites in the Collections of Yale University 9 WEIGHT LOCALITY IN GRAMS Farmineron—Continued MBB, , 2 fc See ekityanteteen © ahah ore eee 23.9 P230 (3 pes—3l g., 45 g., ZIG He alee ila hoes 292 Ferro: Istanps, North Sea P2048. ee Gaia elas lemiie ess ounlencisye 2 Finmarken, Arctic Norway M204) 23sec tee Uh Sito estate 244.5 U.S.A. MGS 7. aa a. en roe seer «lei ae 46.4 IMU s con ops Seatnia etagaten arn onohoy a) 9 sleet aie 10 MAO8 i cinavaciee tale ocie stale 01s) ae 680 Forest Crry, Winnebago County, Iowa, U.S.A. MAGE -gstuere he i eeeiaw woe e 204 GAO wae a8 aig Sores, Slate © 5 suet 720 1 Cl ee Ae nd bya ec ae 166 ) 0 A pe eee ee eco 27.3 MERGES 2 eyed se ctrotes ee ar ee ele 8 11 ee Pa EPO io Emote orc 47.6 P37a (8 pes—7 g,, 14 g., (OT np osebecboaree: 95 Pass (991 pes) <.-.-5-«.. 28066.5 BOZEs iets sore eee alae 22.7 U.S.A. P196 U.S.A. 1): ins te oO OOD od Deca ce 150.5 EZ Ge ) Mukerap? 5222 necite 435.5 M289 “Goamus Farm” ...... 134 M294 -“Mukerop?2. 2222). - 380 M297 “Great Fish River” ...195 M209" Mukeropiie tase ie 53.1 M300 “Lion River” (2 pes— Alig e 2G: cul) i setit/s)cler= 30.3 M3803 “Amalia Farm” ...... 563.55 M304 “Grindorn Farm” ...2790 NISOS eee MuKerop!s -jeciie lo -l-tr 296 M309 “Goamus Farm” ...... 270 MATA Nien opin tote) 1-0 852 M416 “Great Fish River” ..1400 M417 “Amalia Farm” ..... 1290 M579 “Mukerop” .. ...<. ca. 100 lbs. oon Pe LAOW RIVER? 552 Ben oee oe 0.5 La Grance, Oldham County, Kentucky, U.S.A. UG Merscrsetters.o sia ais elas wetee eels 28.79 GOR Hite ciclo vein dani conver. 46 L’A1cie, Orne, France WEDS Mery oh cielo civic es eisdeerah ots 115.45 IMPS SSi amy stattereeae ene etas bee 20 P199 (2 pes—2 g., 907 g.) ..909 Lanck, Vendéme, Loir-et-Cher, France IMSL Singers See terer os noah erste ie 36.7 MBGSR ere peer arms wae setsine ch 35 1 EFA. Came oi cee eR ea ese etc 26 Langon, Bouches-du-Rhoéne, France VEZ ORR eh Renate Sela tnt bi ial Pats 0.64 La Primiriva, Santa Catalina, Desert of Tarapaca, Chile MSA re ces tte tele ne Pace 25 LénArré, Saros, Czechoslovakia, MDT 2 nccnienPiene oa aoe aa es 18.4 MOT Orincccnts ctbeachiecne aoe 39.69 Ole tin toph tion toaster fe Shots loan vices nets 121 Le Pressom, Indre-et-Loire, France JMC setae At boa A ne eR ate 0.1 ZT DEP Scr T ee eerie suacahelneroree 4 Lexincton County, South Carolina, U.S.A. IP GSE y eee cra eo Dae aE 49.5 Lime Creex, Claiborne, Monroe County, Alabama, U.S.A. PAC Vs 5 ee aan a a 59.94 PUD eee ae cot ehe aki Ge kes 64 Limerick Counry, Ireland PES (oat ene iokte oithecdloven< oak 0.5 WEIGHT LOCALITY IN GRAMS Lissa, Bunzlau, Bohemia IMB SAS A are ieee ee 48.38 M506 Aaa Seat 11 Lrrrtz Pryey, Pulaski County, Missouri, U.S.A. 1 ed L(Y (ae aS ine eI ie Gok or oet se 24 Lrxna, Dvinsk, Latvia Ju OA NSO OMT AIAD ASRS OC BEE tbe 8.7 Locust Grove, Henry County, Georgia, U.S.A. Pattee Reet o GRO GIa ENT BOE cue 187.22 IPAS UE oelere cide ortic ist Pee oe 1824 Lone Istanp, Phillips County, Kansas, U.S.A. 1 Ga SPR rae etc en 138.3 MB 2 ie apse IS eR Oe ooo 296 P2209). (ee siak Aiea Mewes rons Oe 150 Lostrown, Cherokee County, Georgia, U.S.A. PQs eeenctag acti oniitare aeee 3 Lucky Hr, Bellevue, St. Elizabeth, Jamaica M267 (2 pces—15.04 g., EA A Ve 36.24 M280) vrttactacet eerie crs ce ero 3.8 Luis Lorez, Socorro County, New Mexico, U.S.A. M2 ODOR Eis ers) nee ae ee es 34.5 1 2 GS ie St A uN Ee oe een ar abe A ed 57 Luororax, Viborg, Finland IVT reo Ma pct Sat ors Sai ee 0.5 Macdvu, Rio Grande do Norte, Brazil VIL DAs Seda Ware sine ae oieeiosin’ 2.75 PLS Gk tae ecclesia 9.5 Mapoc, Hastings County, Ontario, Canada 1 223 Se Saree a i era ear ae Hon se 26 Macura, Arva, Czechoslovakia M288) 215.2 Ae ee os 79.95 MZ OB ite A avavree oleae eet ienc fe eee 8.8 MB0T6 cock nteroee ecciomnenstes 105 PS Bie et Cee e sere IE 822 14 Postilla Yale Peabody Museum No. 27 WEIGHT WEIGHT LOCALITY IN GRAMS’ LOCALITY IN GRAMS Macura—Continued Mezo-Manaras, Transylvania PES Dy rieistetee ei atatote cel een tae rtatets 300 WG DEG Sincmiron ps onccrd ccc 2 oc: 66.79 Pl8c (2 small fragments— lu lofi? Weanentornomonmcce cols. 32.6 Bite Vihear terre ote 8 PIAA oo s0cienseey eee 20 Pisd “(dust)) 622.4225. «few 2.8 Micuet, Olviopol, Kherson, Mansuoom, Bengal, India Ukraine WEIS

“Werier’) (ities stein 0.14 MEE BB 5% Bors 0 ale onc sstveleumterate retains 252.55 NESS (C18) Poesy ees ile. sia 86.8 NELGB sisters Ri stattote races epee ash ct otekees 2.4 INE AD Gtk crsisvasiam lorelzjcrvisacientty auere 87 INEAG Bi ois asians sol orsusislarsteiasy wesc 54.5 WEROO yoke Specie: sora snob bia stale? 17 ME U2 i cases sci phe anoe otal oicioh Sie 6 fens 141 M529) visits iézeie slo toualeveverde nie she ein 400 Eso iemey 330) of eon ae So oeric 675 ERO ire tatay tarot ohoteh ate Metodel deh al eet 69.15 byt tte Soe ie Ceca connc ar nis ace thd Pvuauios, Copiapo, Atacama, Chile 6) FE IIOR O RID GOO Ane PAI 70.1 JOR) (74 TX)! geoveasedomonnc 39 Purnam County, Georgia, U.S.A. INES) xort opatvorsioniay ain io eke tants ios cya ost 2.2 WEBS Oost erotegs suet situa aio we) wate otehere 22 UGS 2.5 otcieara less ciaeeafamers esialeiecera 290 EUG) (QUSE) "os wre ws eiaysts we jereters, 2 1.2 BUD ere erste ar orsicnsloseietsierts otectererse 82.5 QuENccouk, Bassein district, Lower Burma IMULG5i Ss Fe eka oc i eels 2.09 RaxovxkKa, Tula, Russia PAIPANS sc tensteis, crate spores eyorern scene 0.3 Rep River, Texas, U.S.A. BA OM Ay cspctahsy ct aetotecvethelova asses 1635 lbs. PAOUD), CEHIPS) ia ates a/c «: 5 a sterols 16.3 BAO, | (dU8t) posnc snk 2.ek ees 2.3 Reep Crry, Osceola County, Michigan, U.S.A. IMB2Gipe tse cactebevi geet avalen ata 122.9 Renazzo, Cento, Ferrara, Italy VEL GIY 755 Soret ata Syst oie neonate. careers 0.85 18 Postilla Yale Peabody Museum WEIGHT LOCALITY IN GRAMS Ricuarpron, Stark County, North Dakota, U.S.A. BOB ciclo Sra cetcinis wntarnaas telson 198 PAL5. Wa sein Ss eklerdaiee Gisictatne 1428 Ricumonp, Chesterfield County, Virginia, U.S.A. P129 Rocuester, Fulton County, Indiana, U.S.A. P202 (2 pes—6 g., 8 g.) ....14 Ropro, Durango, Mexico WETS Ree SG RE OR ROO eR ice 163.05 VHA DD tee elas ia iaiereionsnstianevekois vesareveete 100 RoegourNE, Hamersley Range, North-West Division, Western Australia IVE ZOU eek eletsyscuctene Beets sickens ie ake 815 INES OO bie oc troealals amir Se Make 265 12) Ok Bo cree Cer et RIOR caoinie 33 Rosario, Honduras MD ites a vaeee ett yeu tats saveyencicusihavotetete 8 U.S.A. P416 Rurr’s Mountain, Lexington County, South Carolina, U.S.A. IBS BR cas orto SERIO Erno OCI: tens AT AS 1s) ACER a ES Seme CRTR 510 Russet Guitcu, Gilpin County, Colorado, U.S.A. Deh acitohetet notenc AER sate eaters 121 Sacramento Mountains, Eddy County, New Mexico, U.S.A. WW BOE aicca'd o Beeeraieastand ole neal mete 123 IMIS a cette eiclsictetst teste eerie 195 M336 (2 pes—2.07 g., (CoN ee eee 136.02 TAG) Go os) Sah reves over tte aoer Ae Ce 250 Me Ai hon spogas Pak cave spe terest tre ohne 4692 Sr. Aucustine’s Bay, Madagascar PD Din 5 acetyl Savane sata the u Sr. Caprats-pE-QUINSAC, Gironde, France LIS eee. eee MD UAE 0.45 No. 27 WEIGHT LOCALITY IN GRAMS St. Francois County, Missouri, US-A- PATO adisvwowiee diss eee ee eee 68 Sr. Genevieve County, Missouri, U.S.A. MSS9 a i:5scefejacesdva% oicto-s Gus ee 413 PYAS seuss eusisyaie essere 174 Sr. Marx’s Mission Station, Transkei, Cape Province, South Africa M176 (2 pces—2.57 g., 26.45 ig) i cts coma 29.02 Str. Mesmin, Aube, France IMDET8) sale ssetes, 4: creke apoceterevee erie 8.45 M500. aces sisi chante Sheresaeta at eee 5.6 Sr. Micuet, Finland M463) 00 Sotioonwinenee see 95 Satine Townsurp, Sheridan County, Kansas, U.S.A. 10 0}: OE SS ree cy a) 66.9 IMT 89: 2. taveiessrta dice cota erteteroreeeie 146.1 M523 a0 sPcd.csa oo Re eee 43.2 Sart Laxe Crry, Utah, U.S.A. P2ZT -2 eee 47 WLS PARE apes nn sdtvey cheeate BIAEAES o Ror 16.2 B305* eka oko ee 17.35 Vicarano, Ferrara, Italy Wuirman, Grant County, M204 ote a res. dake Soa 44.29 Nebraska, U.S.A. TO A eR ef, 1398 M61.’ > (8)-pes) Suds 3.59 P2AO) @ Bec e Gtorarar Merete seorsieris Grenee 87.5 Wicuira County, Texas, U.S.A. Vouk, Poitiers, Vienne, France M386) >. oles cece ete cee ieee 46.21 WSS: WORE te pom TO Oe Ces cleteraice 0.55 P89° seeks ck ebicc wae eee 167 MEO Se Sete has heise ne ae chee Wutamette, Clackamas County, PAG wd PCS) sae see eee ee 74 Oregon, U.S.A. M381 (2 pes—34.09 g., Waconpa, Mitchell County, 242.5 95) 4. oa ee 276.59 Kansas, U.S.A. Wiiamstown, Grant County, M41 (2 pes—11.57 g., Kentucky, U.S.A. DEO (aps tories tere 26.64 M889 | os 5 éciseiesas cs eee 38.7 INUAGi rity Re meee cere 2 lee sels 38.47 PIG. 2 oooh ee eee 80 IVES D GE oj cris sae y havc atcclees eae ate 46.5 Worp Corrace, Thwing, P2262) Ci PCs) s onsck se ee ees 160 Scarborough, Yorkshire, England Watxer County, Alabama, U.S.A. P78" oc. tacine eke Be) eee 1 MBG Ais eeu cterce d as Sales orsial sees 7.92 PD Ba ei Fetes eeroaieee eters as Soe 350 Yaroon, Nellore, Madras gine 728) UN a Altre boinc 21 Mis. hSteon crags losers er even cue eee 0.1 Warrenton, Warren County, ; s PAT Bia ok reidieceraiis, «ts anon ere oe 27 Missouri, U.S.A. s Phy. (see apo ee 6 gs UST Ore Sou we Gare TOTES as 187 Division, Western Australia : M384 (2 pes—1.46 g., Weaver Mountains, Wickenburg, , : 101.33 g:)" s.2 ge oeee 102.79 Maricopa County, Arizona, M391. . cago. d oe aes eee 756.8 Ce P147 124 WVESS iter ete te tenawale te caeectne acts 8:4. TTS et Pa ines ) oe Werranp, Welland County, Ontario, Canada ZasorziKa, Zhitomir, Volhynia, | MBS eee ene te mics terme 44.67 Ukraine IMSS ee ie eesti eae. 28 P206 -cxdcccdeows ack eee eee 22 MESES ee ae ere ee reer ene tee ate ere 55 Zacatecas, Mexico ; POA E Rene eee ae eee es 57 M382 (2 pes—I5 g., Wessety, Hradisch, Moravia, 2U Dg )escte ees 36.75 Czechoslovakia Ma90" coc iwasaeerentemeee vat 7a0 WE) x(Sipes) Mace. te 0.27 Zaviw, Zvornik, Bosnia, Yugoslavia — Weston, Fairfield County, M247 5 2. ee 74.95 Connecticut, U.S.A. IMi29) escera at wo) cts ede onto ree tne 69 MBBS. 43) Pes). “eis o5\s:e% arcane 7.78 11 Pe ain ectaine aid COD AGO nad OIC 5.82 24 Postilla Yale Peabody Museum No. 27 WEIGHT WEIGHT LOCALITY IN GRAMS LOCALITY IN GRAMS moldavite kamacite 0 eee crea eacecticactctes AOR Cie 16.03 P9254 Cranbourne ........ 2 M444 (2 pes—0.80 g., rhabdite ORE ry esac aes i 40.37 B17? ‘France .6. 2... seer 0.6 TE BS ae I ISIC OE 12.11 schreibersite MAAG iS ic cceet cote s Meee Sens 23.98 P9252 Sao Juliao de MAA 205 Soe Sk eee ne wets cae 5.85 Moreira \<-..4 12.8 MAAS I oe cetera a x Saner te cers 21.73 taenite RISAS- (A pes) assess. es 29.59 P9255 Cranbourne ........ 2 IBGS4A: (Se emoces ct ome sseres 6.95 P9255) Toluca 2ss.-.--- eee ee 2.5 Ill. Impactite II. Meteoritic minerals impactite (15 pes) cohenite P103z Meteor Crater, P9253 Cranbourne ........ 1.5 Coconino County, ero MAPETA. dees. 5 one 1 Arizona, U.S.A. ..24.57 REFERENCES Dana, E. S., 1886, Catalogue of the Collection of Meteorites in the Peabody Maseum of Yale College: Am. Jour. Sci., (3), v. 32, Appendix ff. p. 246. Hey, M.H., 1953, Catalogue of Meteorites: London, British Museum (Natural His- tory), 4382 p. Servos, Kurt, 1954, Meteorites in the Carl Bosch Collection of Minerals, Yale Uni- versity: Geochimica et cosmochimica acta, v. 5, p. 299-300. Washington, H.S., 1897, Catalogue of the Collection of Meteorites in the Peabody Museum of Yale University: Am. Jour. Sci. (4), v. 3, p. 83-87. = ni BS. ra. 7 ai Rab asuce f J aL | | i JAN 2 9 1957 | ostilla yHIVERS YALE PEABODY MUSEUM oF NaTurAL History Number 28 January 21, 1957 New Haven, Conn. THE SPECIES OF NOTHARCTUS FROM THE MIDDLE EOCENE Peter Rosrnson’* Granger and Gregory (1917) recognized eight species of Notharctus from the middle Eocene of North America. These were all recorded from the Bridger formation. In the light of recent studies of natural populations this seems too high a number. The material has been restudied to discover if there are significant differences in morphology between Notharctus specimens from the same locality ; if populations from various localities differ significantly within a given member; if there are stratigraphically separable variations; and if non-Bridger middle Eocene Notharctus specimens belong to the same species as Bridger specimens. These studies show that there are only three valid species of Notharctus in that part of the middle Eocene, represented by Bridger B, C, and D. Osborn (1902, p. 191) considered that the genus Notharctus Leidy was separable from the lower Eocene genus Pelycodus Cope by the following characters, “Jaw stout. Symphysis typi- cally coéssified. Superior molars quadrate, with pronounced hypocone; a mesostyle.” This distinction breaks down: Pely- codus jarrovi has quadrate molars and a pronounced hypocone. The genera Pelycodus and Notharctus need thorough revision. Since lower jaws are the commonest mammalian remains, the statistical studies have been based mainly on them. The mor- 1 Department of Geology, Harpur College, Endicott, New York. en 2 Postilla Yale Peabody Musewm No. 28 phology of all teeth has been studied. Species are differentiated solely on dental characters, as there is not sufficient skeletal material for any statistical study. I have used the following abbreviations: A.M.N.H. ... American Museum of Natural History Sif Si" Ree ee Yale Peabody Museum Univ. Wyo. .. University of Wyoming U.S.N.M. ... United States National Museum OR ea <3 oi 2 observed range tn ae ates number in sample 1 ae eee ee mean Vine! Meee (with a superscript or subscript) molar S ...2:/8... standard deviation RUE ae weer coefficient of variability LAS ae ae width of trigonid WWy tall. 2.4 > = width of talonid Wimax «.. a5 maximum width ] Sa iene length eee ah tite: (with a superscript or subscript) premolar ACKNOWLEDGEMENTS I wish to thank the American Museum of Natural History for the use of the museum’s collection and for the photographs of their specimens. Dr. George Gaylord Simpson and Dr. Bobb Schaeffer of that institution were very kind in discussing strati- graphic problems. Dr. Paul McGrew of the University of Wyoming generously lent me two specimens in his care and dis- cussed the Morrow Creek locality with me. Dr. David Votaw of the Yale University mathematics department criticised some of the statistics. Drs. Stuart A. Northrop and Vincent Kelley of the University of New Mexico have read the manuscript and offered criticism; Dr. Kelley especially helped in the prepara- tion of the illustrations. Dr. Joseph T. Gregory of the Yale University geology de- partment has helped and guided this project since he first sug- gested it in 1952. I am greatly indebted to my wife for typing the manuscript. Foe som. 200L | LiBRARY | | JAN 29 1957 HARVARD UNIVERSITY January 21,1957 The Species of Notharctus STRATIGRAPHIC OcCURRENCE THE BRIDGER FORMATION The Bridger formation of southwestern Wyoming was di- vided by Mathew (1909) into five members, A-E in ascending order. Wood (1934) included A and B in the Black’s Fork member, C and D in the Twin Buttes member, and omitted the unfossiliferous E from the discussion of faunal correlation or member rank. The specimens of Notharctus that have been studied come from B and C and D. Twin Buttes member localities listed below have produced 81 specimens of N. robustior and 6 specimens of N. gracilis: Beaver Creek Henry’s Fork Hill Birch Creek Lane Meadows Burnt Fork Lonetree Burnt Fork Post Office Sage Creek Dry Creek Summers Dry Creek Henry’s Fork Twin Buttes Henry’s Fork Divide Black’s Fork member localities yielded 86 specimens of N. tenebrosus and 7 specimens of N. gracilis: Black’s Fork Six miles south of Granger Church Buttes Grizzly Buttes Cottonwood Creek Little Dry Creek Granger Station Millersville Five miles south of Granger _ Five miles east of Millersville In Bridger B, N. gracilis and N. tenebrosus occur at all levels. In Bridger C and D, N. tenebrosus is replaced by its descendant N. robustior and N. gracilis is present and appar- ently unchanged. N. tenebrosus and N. gracilis also occur at the same localities in Bridger B and N. robustior and N. gra- cilis are found together in Bridger C and D. There are no changes in N. tenebrosus or N. robustior from locality to lo- 4 Postilla Yale Peabody Museum No. 28 cality. There is apparently no change in N. robustior between the Bridger and Washakie formations; there is no significant variation in N. gracilis over its entire geographic or vertical range. THE HUERFANO FORMATION More than 600 feet of middle Eocene rocks are present in the upper part of the Huerfano formation of south central Colorado. The upper strata contain a fauna which is younger than the Lost Cabin and older than the Bridger B and there- fore could represent stratigraphically the sparsely fossiliferous Bridger A. The lack of Bridger A fossils makes direct cor- relation impossible. The upper Huerfano has many species which are of lower Eocene affinity and many of middle Eocene relationship. Notharctus is represented by N. nunienus, a spe- cies usually found in lower Eocene deposits. It seems, there- fore, that it would be better to omit it from this discussion. THE GREEN RIVER FORMATION N. gracilis has been collected from two widely separated localities in the Green River formation. One locality is in the Morrow Creek member, Green River Basin of Sweetwater County, Wyoming. The other locality is questionably in the Evacuation Creek member, Uinta Basin, Uintah County, Utah. The relationship of these localities to the Bridger formation is not known exactly: McGrew (personal communication) con- siders the Morrow Creek locality to be upper middle Eocene; Burke (1935) considered the questionable Evacuation Creek locality perhaps equal to the lower Bridger. Dane (1954) places the Evacuation Creek member directly below the Uinta formation. This would indicate an upper Bridger (upper middle Eocene) age. THE WASHAKIE FORMATION Middle Eocene deposits in the lower Washakie formation of south central Wyoming are termed Washakie A. These de- posits are equivalent to Bridger C and D and contain N. ro- bustior. No specimens of N. gracilis have been reported from the lower Washakie. January 21,1957 The Species of Notharctus 5 SysTEMATIC REVISION AND DEscRIPTIONS Family ADAPIDAE Trouessart Subfamily NorHarcTINAE Trouessart Genus Notharctus Leidy Type species: Notharctus tenebrosus Leidy Notharctus gracilis (Marsh) Plate I, figs. 1-2, 4-8, Plate II, figs. 2-3, 6 Hyopsodus gracilis; Marsh, 1871, p. 242 ?Microsyops gracilis; Leidy, 1872A, p. 20 Notharcius gracilis; (Marsh) Cope, 1872, p. 471 Smilodectes gracilis; (Marsh) Wortman, 1903, p. 362 Notharctus matthewi; Granger and Gregory, 1917, p. 847-848 Pelycodus relictus; Gregory, 1917, p. 631 Notharctus gracilis; (Marsh) Troxell, 1926, p. 423-428 Type: Y.P.M. #11800. Broken left lower jaw with P,-M,. From Grizzly Buttes, Wyoming. Bridger B. Lower middle Eocene. Range: Entire middle Eocene, found in Green River formation of Wyom- ing and Utah and in upper and lower Bridger formation. Hypodigm: A.M.N.H. #12011 (Type of P. relictus and N. matthewi), 11471, 13030, 12566; Y.P.M. #11800, 12904, 12965, 12966, 12969, 12970; Univ. Wyo. #965, 966. Some numbers have more than one specimen included and several have both upper and lower dentitions, and right and left side of upper and/or lower tooth rows. Description: Smaller than NV. tenebrosus and N. robustior. Two ridges enclosing a basin on posterior half P;. Ridge runs from the hypoconid of M, to the protolophid. Single external cusp on P*. Discussion: the lower third premolar has a single central cusp. Three ridges are present on the cusp: one runs forward and joins the cingulum; the other two go posteriorly and join the cingulum at the corners, forming a basin between them. A small cusp is present on the cingulum between the junc- tions of the posterior ridges. There is a swelling on the lingual cingulum directly below the position where the meta- conid would arise. The fourth lower premolar has a large protoconid lo- cated slightly forward of center. It has a smaller meta- conid which seems to have been formed by the fusion of a projection from the protoconid and a swelling on the lin- gual cingulum. This projection occupies the place of the interior posterior ridge on P;. There may be a small “hy- poconid.” 6 Postilla Yale Peabody Museum No. 28 Bridger B |BridgerC,D Figure 1 Frequency diagrams of trigonid widths of first and second lower molars of Notharctus gracilis specimens from lower and upper Bridger faunas and Morrow Creek (Green River) fauna. January 21,1957 The Species of Notharctus 7 The first two lower molars are not morphologically dis- tinct from those of N. tenebrosus. They are considerably smaller. There may or may not be a paraconid. If it is pre- sent, it is largest on M,. The third lower molar is quite distinct. On the M, of Notharctus a ridge joins the proto- conid and metaconid. In N. gracilis this ridge is connected to the hypoconid by a ridge running from the hypoconid to this ridge (see pl. 1, fig. 2), whereas in N. tenebrosus and N. robustior it joins a ridge running back from the protoconid. The fourth upper premolar bears a single external cusp and a single internal one. The external cusp shows no evi- dence of dividing. The first and second upper molars are similar to, but smaller than, those of N. tenebrosus. The third upper mo- lar has a rectangular outline but lacks a hypocone. It has a prominent lingual cingulum and a prominent mesostyle. This species is distributed throughout Bridger time. It has been found in the Black’s Fork and Twin Buttes mem- bers of the Bridger formation and Morrow Creek and Evacuation Creek members of the Green River formation. It is not common. Leidy considered Marsh’s “Hyopsodus” gracilis to be the same as his Microsyops gracilis, and named the latter for that reason (1872). He later changed his mind and con- sidered it separate (1873). In his original description he noted that if H. gracilis = M. gracilis, Marsh’s type would take precedence. It appears from illustrations that M. gra- cilis is distinct. This is fortunate for if it were not, No- tharctus gracilis would become the type of the genus Microsyops, thus producing much confusion. Notharctus tenebrosus Leidy Plate I, figs. 9-10, Plate II, figs. 7-8 Notharctus tenebrosus; Leidy, 1870, p. 114; type: U.S.N.M. No. 3752 Limnotherium tyrannus; Marsh, 1871, p. 43; type: Y.P.M. No. 11856 Hipposyus formosus; Leidy, 1872, p. 37; 1873, p. 90; type: U.S.N.M. No. 3757 Tomitherium rostratum; Cope, 1872, p. 470; type: A.M.N.H. No. 5009 Thinolestes anceps; Marsh, 1872, p. 205; type: Y.P.M. No. 11786 A and B PLP €9°0 98'S g S'SI- LSI rl eW-1WA 19° 02'0 egg v gg - og M o0) GL’e 90 Lo'P 7 vP - OF T eA “ Zerg ze"0 109 = 8 9-99 M 3 1O'R 61'0 OL'F L 6 - VP iN cW 4 cee 82'0 €2'¢ L gg - OF M GZ'9 LZ'0 Zev g oy - ge aT cA Wee ene aie eis I Z SL A rd-id G2'F Zz Tr - T'P M Ove bd Ze - 08 a rd S aoe 4 a i . 3 I 6% aI d = HLIAL Uddan ~ ers Lg'0 €2'ST g Gct- TST i | *W-'W s 61ST + 06'S 9600 + 120 690'0 + L¥'S Il ge- Oe xvul M 3 669°0 * LOS ZPO'O + SI'0 60'0 + 80'9 OL 9 - 99 a *W RY 126°0 + ZL S0'0 + ST'0 PFO'O + SL II Ob - TE TPM > OFS'T + Ga°g £40'0 + 6T°0 190°0 + 9's Il EG =e I7M : NS 998° + IT'9 9900 + 08°0 ¥60'0 + ¥9'F II qo - eb a W 3 OLL'O # 80° LZ0'0 + I1'0 L200 + PFS 6 98 - BE TPM. S L86'0 * §9°F 180°0 + F1'0 eb0'0 + IT'e II ee - 8S TM 2 se1'0 + ST's €90'0 + F1'0 L¥0'0 + 09°F 6 8b - Sh af TW x eene eeee ee I oe U2. T ¥a-hy POL 1Z'0 06% Il C8 - Ls M LG'S €8'0 gg'e Il Gy - $s oT ‘d 1g" Z1'0 9% g vo" Le M PLL Za'0 80°S ¢ TS - SZ a *d A S W N uo ee HLAGL WAMOT (suv) syvonsh snjoumy,O Ay AO SUALAWIITIP, NI SLNAWAUASVAW tT HIaVL January 21,1957 The Species of Notharctus 9 Limnotherium affine; Marsh, 1872, p. 207; type: Y.P.M. No. 11795 Notharctus osborni; Granger and Gregory, 1917, p. 848; type: A.M.N.H. No. 11466 . anceps; (Marsh) Granger and Gregory, 1917, p. 849 . affinis; (Marsh) Granger and Gregory, 1917, p. 850 . tyrannus; (Marsh) Granger and Gregory, 1917, p. 851 . tenebrosus Leidy; Granger and Gregory, 1917, p. 851 . pugnax; Granger and Gregory, 1917, p. 853; type: A.M.N.H. No. 11461 Type: U.S.N.M. No. 3752. Right lower jaw with C,, P.-M;, from Black’s Fork, Bridger Basin, Wyoming. Bridger B. Lower middle Eocene. a23245 Range: Lower middle Eocene of Bridger formation. Hypodigm: A.M.N.H. #5009, 11449, 11452, 11454, 11456-57, 11460-61, 11463-11467, 11469, 11472, 11475, 12002, 12568, 12569, 12572, 12575, 12578, 12583, 12586, 13022, 13024-13027, 13029, 13031, 13130, 14567, 14568, 18985-87, 18989, 18990; Y.P.M. #11786, 11795, 11856, 12151, 12153, 12911, 12923, 12932, 12935, 12939, 12941, 12948, 12957-12959, 12963; U.S.N.M. #3752, 3757 (Casts of these specimens are available in Peabody Museum.) Some numbers include several specimens. Description: Size larger than NV. gracilis. Posterior ridge on P, divides halfway down protoconid. No ridge from hypoconid to center of metaconid-protoconid ridge on M;. P* has two external cusps or indi- cations of them. Discussion: The third lower premolar has a single cusp with anterior and posterior ridges. The posterior ridge divides about halfway down the cusp. The interior branch joins the cingulum at the postero-interior corner; the exterior branch joins internally to the external corner. The en- closed basin is small and does not show on worn teeth. The fourth lower premolar has a more pronounced “hy- poconid” than that on the P, of N. gracilis. It is joined to the protoconid by a ridge which divides the heel of the tooth into more or less equal parts. In worn specimens, the heel appears as a shelf. The first lower molar teeth are similar to those of N. gracilis, except for larger size. On M, the ridge running forward from the hypoconid joins a ridge running back- ward from the protoconid. It does not join the protoconid- metaconid ridge. There is rarely a complete external cin- gulum on the molars of N. tenebrosus. Occasionally it ex- tends backwards onto the heel. The fourth upper premolar of N. tenebrosus shows a divided external cusp, or furrows running down the internal 10 Postilla Yale Peabody Museum No. 28 Rane, [aaiea Pecmrom Pereneespmione Pine go 2] 3 & 5 -6 tf 8 9 aaa CHOON VED: pu > ‘w © “No Figure 2 Frequency diagrams of trigonid widths of M, and M, of Notharctus tenebrosus from various localities in Bridger B beds. January 21,1957 The Species of Notharctus sa and external sides of the cusp suggesting a division (see pl. 1, fig. 9). The third upper premolar is near an equila- teral triangle in shape. The upper molars are larger than those of N. gracilis, but smaller than the average molars of N. robustior. The M® is not rectangular. Granger and Gregory (1917) included Tomitherium ros- tratum Cope and Hipposyus formosus Leidy in their de- scription and concept of N. tenebrosus. This writer also includes N. anceps (Marsh), N. affinis (Marsh), N. tyran- nus (Marsh), N. osborni (Granger and Gregory), and N. pugnax Granger and Gregory. Length in Condition of millimeters ————paraconids———__-, M,-M, M, M, M, N. tenebrosus ........ 18.5 strong weak weak IM, GUTKY DEED. “3 Goode Geode 20.7 present absent absent HVE ROSDOTIAEL 2215, 5a ci o10.0 16.6 small vestigial absent INES ONCO DS yt. 's:caiatapeciere's 6 17.5 present vestigial absent INEM OINAS Ns ls tite wle.c'e) ie 18.0 present vestigial absent INE YT ONMUS asa. 5:01) 18.0 (est.) low absent absent After Granger and Gregory, 1917 Listed above are some of the characters which Granger and Gregory considered pertinent in their classification of the lower Bridger Notharctus specimens now included in N. tene- brosus. As stated above, in Notharctus if a paraconid is pres- ent it is always more pronounced on the first molar. The data quoted here (taken from types) bear this out. On five of the types there is no paraconid on M,; on two, there is none on M.; on one there is a strong paraconid on M,. Development of the paraconid is variable, under influence of a gradient which diminishes to the rear. Modern studies of populations have shown that there is a variation in living forms of the same species. There is no reason why this should not be true for fossil forms and Simp- son and others have applied this reasoning to studies of fossils (Simpson and Roe, 1939). The range in lengths of the lower molar series of the types is 16.6-20.7 mm., a distance of 4.1 No. 28 Postilla Yale Peabody Museum 12 = aos | L8'0 g9°ST 8 TLI- 2°61 i eN-1N €0°6 eo'0 Les Or 99- 2g M ERGs ee AO se OO > 00 er PRU Se = eee pe = ee eee OVI €8'0 POL oT us - 39 M G9'8 1¢'0 OL'S 9T 99-19 *y 7 1201 99°0 99 9 l= $9 M £08 PPO LYS LI 89 - Lib Pt uA eaey ae wes Ae aes dead 00°9 v 99 - FS M 02 G7 Gh - 0'F Py rd 02'F z ov M GL’e z Se - Le T od HigaL Wddn LESS F FSI POV'O + FS Lg'0 + LO'ST GT Gunn LOL 41 *W-"W €86'0 + 86°L 1h0'0 + E80 LG0'0 + ZIP rE Lb- Peg xvol M ror L + SPIT LOU'0 + $80 PSO + GEL Ze $6 - 29 al "WH 8860 + LI'OL 940'0 + 8¥'0 990'°0 + LL'b Pg 6¢ - OF Te1M 8Z0'L + 89°01 ChO'0 + L¥FO ¥790'0 + IVb £S wg - 98 IYM 888'0 + FI'6 €s0'0 + S¢'0 GL0'0 + 96'S 75 OG ai *W 890'l + ZOOL 9F0'0 + PFO 990'0 + S8'b cP gg - 8s Te7M 8L0°T + 0801 PFO'O + TF'0 190°0 + 868 CP To - oe ITM 98L'0 + PSL ChO'O + SF'0 190°0 + $1 LY OL - oe aT 'W fs 2 eee eS ee eee 1¢°6 1g'0 92'S ¥L S'S - 6% M S16 2 1r'0 6F'F GT eG - $s a ‘d Lg‘O1 820 69'S Or 08 - OZ M oS'L 0g'0 28's 6 3 - SS ry ot A S WwW N uo HLAAL YAMO'T 6 WIaVL AGIAT 8nsoiqgaua} 8NJIADYFO AT LO SUALAWITTIP, NI SLNAWAWMOSVAYL January 21,1957 The Species of Notharctus 13 mm. The observed range of all specimens is 16.1-22.2 mm. or 6.1 mm. The lengths of the individual teeth all vary at least 1 mm. between limits. Assuming that all teeth were packed end to end, the possible range (from observed specimens) is 16.4-— 23.5 mm. As there is some overlapping of teeth in a row, the figures for natural rows are smaller. Notharectus tenebrosus? Plate 1, fig. 3 A specimen of Notharctus in the American Museum of Natu- ral History (#11481) has a peculiar lower third molar. The length of the tooth is less than that of the second molar. It lacks the enlarged hypoconulid. Due to its similarity to typical specimens of N. tenebrosus in the characters of its anterior molars and premolars, it is classified as N. tenebrosus?. Gran- ger and Gregory thought that it might represent the jaw of Aphanolemur gibbosus, but no teeth are known for this species. It is considered an aberrant form of N. tenebrosus, and has not been included in the hypodigm of that species. A.M.N.H. #11481, Grizzly Buttes, Bridger Basin, Wyoming, Bridger B M, M. M TOC Sh ice eens aacae a Re ee areas 5.3 5.? 5.1 DWiptnh axe) Gee te ores Sates soiree 3.4 \O 58 reo) se che cae ARI Ot uae anes 3.6 AW ae tall ei eiuere ie hake ectetust ere ote 3.2 3.6 Notharctus robustior Leidy Plate II, figs. 1, 4-5 Notharctus robustior; Leidy, 1872, p. 364; type: U.S.N.M. No. 3750 Telmatolestes crassus; Marsh, 1872, p. 206; type: Y.P.M. No. 11782 Hipposyus robustior; (Leidy), 1873, p. 93 Notharctus crassus; (Marsh) Granger and Gregory, 1917, p. 854-856 Notharctus robustior Leidy; Gazin, 1934, p. 71 14 Postilla Yale Peabody Museum No. 28 Type: U.S.N.M. No. 3750. Henry’s Fork, Bridger Basin, Wyoming. Bridger Cor D. F. V. Hayden collection M,. Range: Upper middle Eocene of Bridger formation (Bridger C and D) and of Washakie formation (Washakie A) of southwestern Wyoming. Hypodigm: A.M.N.H. #11451, 11982-11987, 11990-93, 11995-97, 11999, 12000, 12003, 12005-12010, 12564, 12565, 12567, 13023, 13133, 13134; Y.P.M. #11782, 12900, 12905, 12908, 12909, 12911-12916, 12918, 12920, 12925- 12931, 12933, 12934, 12937, 12940, 12945, 12953; U.S.N.M. #3750 (Cast in Peabody Museum. Many numbers have several specimens included, especially in the Peabody Museum collections.) Description: Larger than NV. tenebrosus, otherwise quite similar. P* has pronounced double peak to external cusp. Discussion: It is extremely difficult to tell some specimens of N. tenebrosus from some of N. robustior. The two ranges overlap and where the age data are unknown, the determina- tion can be difficult, if not impossible. A few pertinent measurements of N. tenebrosus and N. robustior are quoted below for comparison. N. tenebrosus Range in millimeters Mean N IGenvaidn Nils Goaaneasosncdbe00r 6.2-9.3 7.385 + 0.154 32 SW is IM aR ee at eyarscsaysteds Srrineheiens 3.4-4.7 4.12 + 0.057 34 \Ny tanitgoil ik, aesoauoonenodc 3.2-5.1 3.98 + 0.061 45 Woirigowid: Ma. 62s. Soi asi fon 3.6-5.7 4,41 + 0.064 54 N. robustior Range in millimeters Mean N Ibarra MES sooopeacoaoosoHEce 7.7-9.6 8.66 + 0.068 39 \ Wo TENG IRS hes po cap pons opo-o 4.3-5.7 5.01 + 0.045 41 \MY “Uitbsorie lily Goosndoonerane 4..2-5.4 4.72 + 0.041 AT \hY qenyorncl WS Goce comegapec 4.6-6.3 5.42 + 0.050 62 January 21,1957 The Species of Notharctus 15 ]HenrysFork Twin Buttes [Henrys Fork [Burnt Fork Henrys Fork Divid PQ Hilt. Wtr.M, = AZ o oe -3 4 5 -6 -7 8 -g Oo a :2 a) 4 = ° od —~o 0@ VO o ~o OHNO uF uv Olpo Figure 3 Frequency distributions of trigonid widths of M, and M, of Notharctus robustior from various localities in upper Bridger formation. Specimens from both Bridger C and D divisions are included at all localities. 16 Postilla Yale Peabody Museum No. 28 The most significant distinction between these species is the width of M;. The ratio of the difference of the means to the standard error of the difference (*/o4) is 12.59, a highly sig- nificant figure. N. robustior certainly arose from N. tenebrosus but the samples shown here prove there is a statistical difference in the populations. A similar test proves that there is no dif- ference in the populations of N. gracilis during Bridger time. Specimens of N. robustior from the Washakie formation fall near the middle of the range of the Bridger specimens. They have been included in the statistical analysis of the species. The measurements are, for the characters used above: A.M.N.H.* Mean of all #13133 #13134 N. robustior specimens* Li. I ae LER ence tego Rea eR es Fe eee 8.7 8.4 8.66 Waa Ml toc protic te sect 4.4 4.7 5.01 W trigonid M, ..:..5:2.%. 4.6 4.5 4.72 Wittrigonid’ Mi) 2o...o- so. 5.3 5.2 5.42 VARIATION Histograms for measurements of teeth of N. tenebrosus are distinctly bimodal. The separation of the peaks is not great; 1 mm. is the maximum. The areas under the peaks are not equal, but are not sufficiently unequal to be significant. It is my opinion that this bimodality is due to sex differences. The animals are morphologically similar ; one would expect a morph- ological dissimilarity if more than one species was present. The coefficients of variability for the lower teeth vary from 11.45 mm. for the length of M, to 7.54 mm. for the length of M,. The average is 9.772 mm. This is high but not too high for a variable species. It has been pointed out that rarely do more than one species of a genus inhabit a given area and when two congeneric species do live in one habitat, one greatly outnum- bers the other. This is the case in both the upper Black’s Fork member and in the Twin Butte member of the Bridger forma- tion. N. gracilis is one-tenth as common as N. tenebrosus or N. robustior. * All measurements given in millimeters. 17 The Species of Notharctus January 21, 1957 pa es ee ES eS ene 0Z'F 18'0 61°61 6 FOZ- 6'LT TI eN-N Z9'°9 60 OWL 61 61- 8S M Goh Ir'0 g9°¢ SI 29 - Fy “Al eN 6S°L 69°0 60°6 GS ZOI- 89 M cro r'0 z8'9 GS Pe" 6ie aii sWW e8'°% $20 PPS 61 88 - SL M Ze 1z'0 PL'9 02 ah SHG oT iN ra eins ae ae ae waa 68°F ZE'0 62h L GL - $9 M eh'9 ze'0 86h €I rg - OF | rd SI's 62'0 09° g 09 - Sg M €9°FI 99°0 €9'P g yo - TP T sd HiLaaL Widdn oo' 60 + GZ'2S GT 9°SZ- 6'0Z Al "WW- "TA 0829 TL’¢ Z80'0 + 62'0 GhO'O + 10'S IP LG-@p xeor MW : PSS'0 + E8'b 8h0'0 + ZF'0 890°0 + 99'8 68 96 - Lh ay "WW 8gc'0 + 929 1g0'0 + ¢¢'0 PFO'O + 79'S $9 G9 - 6P Teh 1g9°0 + 61'L Gg0'0 + 68°0 0S0'0 + ZFS 29 e9 - OF T3M 80L'0 + S6'L 0S0°0 + 9¢°0 TL0°0 + F0'L ~9 gg - 19 ry "WW S6V'0 + LYS 9Z0°0 * 62'0 0F0'O + 92'S ag 69 - Lv TRUM 029'0 + S6'S 6Z0°0 * 820 T¥0'O + LP LP wo - OF TM IL8'0 + 29'8 €F0'0 + SF'0 190°0 + 69°9 0g gy - 9's ali TW CHOhDsO Mier ane ee eeee aera 60'P 810 10°F LI Gr- os M GO'L Or'0 LOS SI 9 - 6 oT fd Hig} $20 60° fh GE - GS M STL 1g'0 ce'h 8 Lb - 8s T ®d A S Ww N uo HiIGdL YLMOT (AGIa]) w01sngot snjouMYyJO AT JO SUTLAWITTIL NI SLNAWAMASVAY & WIGdVL 18 Postilla Yale Peabody Museum No. 28 When the deposition of the Bridger formation began, the area had just emerged from the Green River Lakes. Notharctus tenebrosus was a new species in the Bridger B time. The source area for this species could have been any one of several inter- mountain basins of the Wyoming, Colorado, Utah area. The Bridger environment was similar to the source area, but different enough that N. tenebrosus had to adjust to it. It is natural for a species to be variable in a new environment. Simp- 6 © WN. fenebrosus O WN. robustior Oo Oo =~ Oo ” oO @0 5 @ Or OS® @ ° O°000 £ O = ©0000 ‘e COMO : 628089 = oo) OO = °o 009 ° ° ° 4 ° Wp M, (millimeters) 5 6 Figure 4 A scatter diagram comparison of anterior and posterior tooth widths of first lower molars of Notharctus tenebrosus and NV. robustior. son states (19538, p. 140), “In a group not already at the se- lective optimum, selection should, given a sufficient store of appropriate variation, move the group to that optimum.” January 21,1957 The Species of Notharctus 19 Notharctus tenebrosus was a species not at the selective opti- mum; neither was its descendant, N. robustior, evidently, but it was closer to it. The variation of the N. tenebrosus — N. robustior line de- creased in time. The average size of N. robustior was larger than that for N. tenebrosus; however, the maximum size of N. robustior is within the probable limits for N. tenebrosus. Oo Oo 6 Oo oOo lolexe) OO 00°00 foXo) 0000 1@) eke) ©70} (© wa ©00O ° a Orcior@ 5% rexore) E O @c00 Oo = O © ©o°o ° £ ©" Oo O fav] ° => ° o oo = 0 oOo 3. ° oo ° 4 000 W, Mo (millimeters a p 2 (r ey Figure 5 Scatter diagram comparison of anterior (trigonid) and posterior (tal- onid) widths of M, of NV. tenebrosus and N. robustior. N. robustior is therefore a more highly adapted species than its ancestor N. tenebrosus. The scatter diagrams for anterior/posterior widths of M, and Mz, of N. tenebrosus—N. robustior on figures 4, 5 show that 20 Postilla Yale Peabody Museum No. 28 the variability is greater for N. tenebrosus. The maximum values of measurements of N. robustior are only 0.3 mm. greater for M, and 0.6 mm. greater for M, than the maximum values for N. tenebrosus. N. robustior represents the maximum size for the evolutionary line that could live in the environment at the time. N. robustior also has somewhat bimodal histograms. Why the tendency toward bimodality should be less is not known al- though it is partly caused by the general tendency for a de- crease in variability. Perhaps the smaller sex was tending to become as large as the other for adaptive reasons. It is interesting to contrast the data for N. gracilis with that of the N. tenebrosus —N. robustior line. N. gracilis re- mained practically unchanged during the entire middle Eocene. The greatest coefficient of variability for the lower molars is 6.11 mm.; the smallest 2.97 mm. The average is 4.437 mm. These data were compiled from 16 specimens from Bridger B, C, and D, and the Green River Morrow Creek locality. The recorded geographic range of N. gracilis includes south- western Wyoming and northeastern Utah; N. robustior is found only in the Bridger and Washakie basins of Wyoming. N. tenebrosus was restricted to the Bridger basin. At its maxi- mum, the area of the N. tenebrosus — N. robustior group was less than half that of N. gracilis. In none of the localities in which N. gracilis occurs is it common. The combined wide dis- tribution and infrequent occurrence of N. gracilis is interesting Possibly N. gracilis inhabited the higher areas and only oc- casionally visited the plain. If this were true it would account for its ability to spread to other areas more easily than N. tenebrosus — N. robustior. The statistical data for these species have been compiled fully only for measurements of the lower teeth as these are the most common remains. Certain problems were encountered in measuring. The measurement of a tooth length will be dif- ferent if measured alone or in place in a row. The maximum error for this is perhaps 0.2 mm. This would not be considered harmful for large teeth (Simpson, 1949A). When dealing with an animal which has small teeth, the chances of a significant error are increased. For example, the mean length My, of N. January 21,1957 The Species of Notharctus 21 robustior is 7.04 + 0.071 mm. The means of four groups of M.’s, determined by their association with other teeth, are as follows: Lengths in millimeters of M, IMErenisolated). 0 soak detect ae caus Mean= 7.19 N= T10 Vet Mila Are cste tac yerate 3 ctavevae sraiale vol af Seete are 1% 7.08 26 MIEN VU name reye Soins novels ciate ora’ so nD Giclees 6.79 12 ire VicersWDettt syatarciscevey cuctonco ste ieee atara are 7.10 16 The means of these subgroups are within one standard de- viation from the means of the entire sample. This is not too significant but it suggests that single teeth would give larger parameters (in absolute values). 22 Postilla Yale Peabody Museum No. 28 REFERENCES CITED Burke, J. J., 1935. Preliminary report on fossil mammals from the Green River formation in Utah: Carnegie Mus. Annals, v. 25, p. 13-14. Cope, E. D., 1872. Third account of new Vertebrata from the Bridger Eocene of the Wyoming Territory: Am. Philos. Soc. Proc., v. 12, p. 469-472. Dane, C. H., 1954. Stratigraphic and facies relationships of upper part of Green River formation and lower part of Uinta formation in Duchesne, Uintah and Wasatch counties, Utah: Am. Assoc. Petroleum Geologists Bull, v. 38, p. 405-425. Gazin, C. L., 1934. On the priority of specific names for the upper Bridger Notharctus: Jour. Mammology, v. 15, p. 71. Granger, W. and W. K. Gregory, 1917. Revision of Eocene primates of the genus Votharctus: Am. Mus. Nat. History Bull., v. 37, p. 841-859. Gregory, W. K., 1917. Genetics vs. Paleontology: Amer. Naturalist, v. 51, p. 631. Leidy, Joseph, 1870. Descriptions of Palaeosyops paludosus, Microsus cuspidatus, and Notharctus tenebrosus: Acad. Nat. Sci. Phila. Proc., v. 22, p. 114. 1872A. Remarks on fossils from Wyoming: Acad. Nat. Sci. Phila. Proc., v. 24, p. 19-21. 1872B. Remarks on some extinct mammals: Acad. Nat. Sci. Phila. Proc., v. 24, p. 37-38. - 1872C. On the fossil vertebrates of the early Tertiary forma- tion of Wyoming: U. S. Geol. Survey of Montana, etc., F. V. Hayden, Report for 1871, p. 364. 1873. Contributions to the extinct vertebrate fauna of the Western Territories: U. S. Geol. Survey of the Terr., F. V. Hayden Report, v. 1, pt. 1, p. 93, pl. 6, fig. 40. Marsh, O. C., 1871. Notices of some new fossil mammals from the Tertiary formation: Am. Jour. Sci. (3), v. 2, p. 35-44. 1872. Preliminary description of new Tertiary mammals, pts. II, III, and IV: Am. Jour. Sci. (3), v. 4, p. 202-224. Matthew, W. D., 1909. The Carnivora and Insectivora of the Bridger Basin, middle Eocene: Am. Mus. Nat. Hist. Memoirs, v. IX, pt. VI, p. 291- 567. Osborn, H. F., 1902. American Eocene primates and the supposed rodent family Mixodectidae: Am. Mus. Nat. Hist. Bull., v. 16, p. 169-214. Picard, M. D., 1955. Subsurface stratigraphy and lithology of Green River formation in Uinta Basin, Utah: Am. Assoc. Petroleum Geologists Bull., v. 39, p. 137-163. Simpson, G. G., 1940. Mammals and land bridges: Wash. Acad. Sci. Jour., v. 30, p. 137-163. 1949A. A fossil deposit in a cave in St. Louis: Am. Mus. Novitates, No. 1408. 1949B. The meaning of evolution: New Haven, Yale Univ. Press. 1953. The major features of evolution: New York, Columbia Univ. Press. January 21,1957 The Species of Notharctus 23 Simpson, G. G. and Anne Roe, 1939. Quantitative zoology: Numerical con- cepts and methods in the study of recent and fossil animals: New York, McGraw-Hill. Troxell, E. L., 1926. Smilodectes or Notharctus: Am. Jour. Sci. (5), v. 11, p- 423-428. Wood, H. E., 1934. Revision of the Hyrachyidae: Am. Mus. Nat. Hist. Bull, vy. 67, p. 181-295. Wortman, J. L., 1903. Studies of Eocene Mammalia in the Marsh Collec- tion, Peabody Museum, pt. II. Primates: Am. Jour. Sci. (4), v. 16, p. 345-368. 24 Postilla Yale Peabody Museum No. 28 Plate I Figure 1. Notharctus gracilis (Marsh). Right P*-M*, A.M.N.H., no. 12011, part of type specimen of VV. matthewi Granger and Gregory. Grizzly Buttes, Wyoming, Bridger B horizon. Crown view of teeth, x3. Figure 2. Notharctus gracilis (Marsh). Left lower jaw with C,-M,, A.M.N.H., no. 12011, part of type of NV. matthewi Granger and Gregory. Crown view of teeth, x3. Figure 3. Notharctus tenebrosus? Right jaw with P,-M,, A.M.N.H., no. 11481, Grizzly Buttes, Wyoming, Bridger B. Crown view of teeth, x2. Figures 4-8. Notharctus gracilis (Marsh). A.M.N.H. no. 13030, Bridger B formation, Cottonwood Creek, Wyoming. Figs. 4-7, crown views. Fig. 4, right P,;-M.; Fig. 5, left M,-M,; Fig. 6, left M’-M*; Fig. 7, right P?-M*; Fig. 8, externo-lateral view of right maxillary P'-M*, same specimen as fig. 7. All x2. Figures 9-10. Notharctus tenebrosus Leidy. Right maxillary with P*, M’, and M,. A.M.N.H., no. 13027, Grizzly Buttes, Wyoming, Bridger B formation. Fig. 9, externo-lateral view; Fig. 10, crown view. Both x2. Photographs by permission of the American Museum of Natural History. January 21,1957 The Species of Notharctus 25 Plate I 3 “ PAs 26 January 21,1957 The Species of Notharctus Plate II Figure 1. Notharctus robustior Leidy. Y.P.M., no. 11782, right lower jaw with P,-M,, type of Telmatolestes crassus Marsh. Henry’s Fork, Wyoming, Bridger C or D. Crown view of teeth, x2. Figures 2-3. Notharctus gracilis (Marsh). Y.P.M., no. 12970, Twin Buttes, Wyoming, Bridger C or D. Fig. 2, right maxillary with P?-M°; Fig. 3, left jaw with P,-M,. Crown views, x2. Figures 4-5. Notharctus robustior Leidy. Y.P.M., no. 12905, Lone Tree, Wyoming, Bridger C or D. Fig. 4, left maxillary with M?-M*; Fig. 5, left lower jaw with P,-M,. Crown views, x2. Figure 6. Notharctus gracilis (Marsh). Type specimen, Y.P.M., no. 11800, Grizzly Buttes, Wyoming, Bridger B. Left jaw with P,-M,, crown view, x2. Figure 7. Notharctus tenebrosus Leidy. Y.P.M., no. 11786A, Grizzly Buttes, Wyoming, Bridger B. Cotype of Thinolestes anceps Marsh. Left maxillary with M’-M®, crown view, x2. Figure 8. Notharctus tenebrosus Leidy. Y.P.M., no. 11786B, Grizzly Buttes, Wyoming, Bridger B formation. Cotype of Thinolestes anceps Marsh. Left jaw with P,-M;, crown view, x2. Photographs by permission of the Yale Peabody Museum. January 21,1957 The Species of Notharctus 27 Plate II 7 ¥, é re A . aa ; a : ms hs al a ‘As ar : ; ~ er (hy! ae j pets ; : ‘ 7) ith “ied aE 1S cat te ae la iy at ee. aL fo. ; ieee | i + ‘ ; aot lees at? ' Ont Ay a tx Mace vias | es «dy aetentah ; ei? - ’ ey a, y tres mae is dam f a ; ; , ‘ ‘ ? 5° one ~- — “ . wat > ows ‘ i 7 | 7 7 ¢ mA - ext ', 5 , i an ; : 4 4 th LA (ey Me W ee 4 : 7 i 4 cadet Ky , ie : Re ry ty f : ; had ; ° " 2 " ? + , x 3 a .) ? rs ¥ 7 ie r f , ey al Ri é¥ 1 1 N= o * age oy » el a ‘ ‘ ~~ a x a Dives q 4 \ U . ee : me + h% ; (oar Fat. wy e wks Pr ri > y . — 1 OMEN eee -_- os. ~~ AS YALE PEABODY MUSEUM oF NaTurRAL History Number 29 February 12, 1957 New Haven, Conn. A REDEFINITION OF THE SUBSPECIES OF FODIATOR ACUTUS JAMES E. Morrow* ABSTRACT Analysis of characteristics of 11 Atlantic and 54 Pacific specimens of Fodiator acutus showed that the two subspecies fF. a. acutus and F. a. pacificus could not be distinguished on the basis of the original diagnosis, but that such distinction could be made on the basis of the eye and the snout expressed as a percentage of head length. It was also shown that F. a. pacificus Bruun and Hemiexocoetus caudimaculatus Fowler are syn- onyms of Fodiator acutus rostratus (Giinther). INTRODUCTION The primitive flying fish species, Fodiator acutus (Cuvier and Valenciennes), was divided into two subspecies, F’. a. acutus from the Atlantic and F. a. pacificus from the Pacific, by Bruun (1933) on the basis of 38 or 39 vertebrae in Atlantic specimens and 41 vertebrae in a single specimen from the Pacific, “along with a number of other smaller, yet distinct differences in pro- portions and fin-ray characters.” Breder and Nichols (1934) noted vertebral counts of 39 in two specimens of F. acutus from the Pacific coast of Panama and therefore rejected the validity of the new subspecies. They also considered pacificus to be a nomen nudum because of the incomplete description. Later, Bruun (1935) gave a complete tabular description of * Bingham Oceanographic Laboratory 2 Postilla Yale Peabody Museum No. 29 the type of F. a. pacificus, comparing it with four specimens from the Atlantic Ocean off the coast of Angola, and with the type of Giinther’s Exocoetus rostratus (Giinther, 1866) from Hawaii. Bruun’s differential characteristics may be summa- rized as follows: ATLANTIC PACIFIC IOs el BAG. 2 ae oO ney Aral ns ae sh an 1:41 9 12 ATCA EG: aR ade Rea, SS, SERIO Le 11 10 Re@ctonal Taye. et oe cals 14 — 16 13 Precdlorsal Scales tek. Soe ew ble) ae 21—24 24 — 26 MeGHeUE aes eo em fe Loe aude 38 — 39 4] Giilvrakers?.. ese to. keer, cee 7+ 22 8 + 24 Subsequently, Breder (1938, p. 12-16) accepted the validity of this diagnosis, and noted several additional characters, as shown in his key: “A. Gill rakers 7-+22; predorsal scales 21 to 24; dorsal 10 or 11; anal 11; body depth 4.1 to 4.8; dorsal in- sertion 1.32 to 1.38; interorbital 3.6 to 3.7; pectoral rays 14 to 16. Fodiator acutus acutus AA. Gill rakers 8-+24 or 25; predorsal scales 24 to 26; dorsal 9 or 10; anal 10 or 11; body depth 4.85 to 6.0; dorsal insertion 1.31; interorbital 3.9 to 4.0; pectoral rays 13. Fodiator acutus pacificus.” However, a specimen of F. acutus taken by the Yale South American Expedition at Cabo Blanco, Peru, on April 1, 1953, could not be ascribed to either subspecies on the basis of the characteristics given by Bruun and by Breder. This at once raised the suspicion that the two subspecies might not be valid, but might be merely the result of examining too few specimens. Accordingly, a study of a larger number of individuals was attempted. As far as Pacific specimens were concerned, the re- sult was gratifying, but, although nearly every museum in the United States and western Europe was canvassed for material, the combined efforts of all could produce but 11 specimens from the Atlantic Ocean. Feb. 12, 1957 Subspecies of Fodiator Acutus | {ARS ARD ; MATERIALS The following material has been available. The numbers in parentheses indicate the number of individuals included in each sample. Pacific Material Bingham Oceanographic Collection, Peabody Museum of Nat- ural History No. 704 (3) No locality 1004 (9) Concepcion Bay, Mexico 1011 (5) Concepcion Bay, Mexico 1012 (5) Concepcion Bay, Mexico 1040 (2) San Jose del Cabo, Baja California 1101 (3) Las Perlas Islands, Gulf of Panama 1193 (1) San Felipe Bay, Baja California British Museum (Natural History) No. 1898.12.31.40-41 (2) Sta. Elena Bay, Ecuador 1903.5.15.298 (1) Panama 1938.12.12.44 (1) Galapagos Islands 1938.12.12.45 (1) Galapagos Islands 1938.12.12.46-47 (2) Galapagos Islands 1939.7.10.17 (1) Galapagos Islands 1939.7.10.18 (1) Tehuantepec, Mexico United States National Museum No. 82006 (5) Chame Point, Panama 119729 (2) Concepcion Bay, Mexico Chicago Natural History Museum —No. 2580 (2) Gulf of California 41516 (1) Galapagos Islands 41517 (1) Galapagos Islands 41703 (1) Galapagos Islands 49217 (3) Bahia Honda, Panama Academy of Natural Sciences, Philadelphia No. 7484 (1) No locality 7508 (1) Mazatlan, Mexico (Type of H. caudimacu- latus Fowler) UNIVERSITY 4 Postilla Yale Peabody Museum No. 29 Atlantic Material American Museum of Natural History No. 9028 (2) Angola British Museum (Natural History) No. 1906.8.24.128 (1) Angola 1938.10.10.23 (1) Nigeria 1938.10.20.24 (1) Nigeria Chicago Natural History Museum No. 4913 (2) Caribbean Sea, 14° 30’ N, 80° 30’ W Zoologische Staatsinsitut und Zoologische Museum Hamburg Nos. H1, 2,8,4 (4) West Africa (Measurements from Bruun, 1935). There have thus been 54 specimens from the Pacific Ocean and 11 from the Atlantic available for study. It early became ap- parent that a considerable degree of allometric growth existed in most body proportions among the smaller individuals, but that this allometry had largely ceased at a standard length of about 100 mm. Consequently, the smallest specimen utilized in the comparison of these factors was 95 mm. in standard length. Hence, there were 39 Pacific and nine Atlantic speci- mens available for this purpose. Even though the size of the Atlantic sample still leaves much to be desired, it is more than twice as great as the amount of material that was available to Bruun. All the material, regardless of size, was utilized in analysing meristic characters. Discussion In the analyses of the various characteristics of the two groups, it was found that only two of the characteristics on which the subspecies were erected actually showed a significant difference between the groups (P<0.02). These characteristics were the number of predorsal scales and the width of the in- terorbital space. For the latter, it was found that a greater Feb. 12, 1957 Subspecies of Fodiator Acutus 5 degree of separation could be effected by expressing the width of the interorbital as a function of the length of the head rather than as a function of the standard length. In addition, there may be a difference in the number of pectoral rays. However, as will be shown later, the differences shown in these characters are not sufficient to warrant subspecific rank for the two groups. On the other hand, the length of the snout and the anterior- posterior diameter of the eye, in relation to head length, appear to be valid characters upon which the subspecies may be based. The number of pre-dorsal scales varies between 20 and 25 in the Atlantic material, and between 22 and 28 in the Pacific, with mean counts at 22.3 and 24.5 respectively (Fig. 1A). It will be observed, however, that the standard deviations of the two distributions overlap by an amount equal to about 33 per cent of the smaller. Such a high degree of overlap is indicative of a corresponding degree of intergradation and is generally considered as indicating less than subspecific differentiation. The width of the interorbital space, expressed as a percentage of the head length (Fig. 1B), ranges between 24.7 and 28.8, with a mean at 27.0 in the Atlantic material. In the Pacific specimens, however, these values are somewhat lower, the range spreading from 22.2 to 26.8, with the mean at 25.4. This dis- tribution is skewed by the inclusion of a single specimen whose interorbital was only 22.2 per cent of the head length, the next lowest value being 24.0 per cent. Eliminating the single low value produces a considerable reduction in the range of the sample and almost entirely eliminates the skewness, but has no other marked effect. However, whether the sample is taken whole or without the one individual, the standard deviations again show considerable overlap, actually equal to 58.7 per cent of the smaller standard deviation. Clearly, this cannot be considered as indicating subspecific differentiation. The distribution of the counts of pectoral fin rays in the Atlantic material is such that graphical representation is mean- ingless. A total of 16 counts showed 15 fins with 14 rays each and one with 16 rays. In the Pacific material, 94 counts showed 4.3 per cent of the sample with 12 rays, 63 per cent with 13 rays, and 33 per cent with 14 rays. The mean counts were 14.1 for the Atlantic specimens and 13.3 for the Pacific. 6 Postilla Yale Peabody Museum No. 29 The distribution in the Atlantic material suggests that in a larger sample, more lower counts would be expected. There- fore, although the number of rays in the pectoral may actually be a good indicator of the subspecies, this cannot be deter- mined from the present samples. Some most interesting results were derived from a compari- son of the snout as percent of head length in the two samples. The Atlantic and Pacific groups are clearly different (Fig. 1C), the former having a mean value of 36.3 per cent as com- pared with 39.6 per cent for the latter. The standard devia- tions of the two distributions do not overlap at all, even though the upper limit of range of the Atlantic sample includes the whole standard deviation of the Pacific material. This skewness of the Atlantic material is caused entirely by two individuals from the Caribbean Sea, whose snout lengths were 39.7 per cent and 41.4 per cent of their head lengths. As may be seen from a comparison of the upper and middle distributions in Fig. 1C, inclusion of these two specimens (the only ones from the western Atlantic) results in a discontinuous distribution. When the two Caribbean specimens are removed from considera- tion, the Atlantic material has the distribution shown in the upper line of Fig. 1C. There can be no doubt that the upper Atlantic group and the Pacific group are different. It also appears as though the Caribbean material may represent a population that is distinct from that of the west African coast, although the material is not large enough to draw reasonably reliable conclusions. However, this possibility is certainly not out of the realm of probability, especially when it is realized that F. acutus is generally confined to coastal waters and is therefore extremely unlikely to undertake long movements back and forth across the Atlantic. This is precisely the situation that could lead to separate, distinct races on either side of the ocean. It will be seen from Fig. 1C that the standard deviations of the Pacific sample and the whole Atlantic sample do not quite meet, indicating a separation of a little more than 84 per cent. This degree of separation seems fully adequate to indicate sub- specific differentiation. Comparing the west African material alone with the Pacific specimens. they are separated by a large Feb. 12, 1957 Subspecies of Fodiator Acutus 7 gap, indicating a pronounced degree of differentiation. With a larger sample, this could be interpreted as showing full speci- fic rank for each group, but with the number so small this as- sumption is not warranted here. The diameter of the eye, again expressed as a percentage of the head length (Fig. 1D), also provides a reliable character for the distinction of the two subspecies. The Atlantic material covers a range from 27.1 to 31.5 per cent, with the mean at 29.2 per cent. By contrast, in the Pacific material the eye is only 22.7 to 28.1 per cent of the head length, with the mean at 25.4 per cent. As with the length of the snout, the standard devia- tions of the two samples do not meet, showing that here, too, the samples represent separate subspecies. SYNONYMY The synonymy of the subspecies of Fodiator is no less in- volved than are the characters upon which the separation may be based. Bruun named his Pacific subspecies pacificus, and in his 1935 paper noted that there also occurred in the Pacific the at-least-nominal species, Hemiexocoetus caudimaculatus Fowler and Exocoetus rostratus Giinther, both of which are clearly Fodiator. The type (and only) specimen of H. caudi- maculatus is a juvenile, so that comparison with the adult ma- terial used here would be of no value. However, this specimen does not differ in any significant way from other juveniles of comparable size that are labelled Fodiator acutus. Bruun (1935) gave measurements of the type specimen of Ewxocoetus rostratus. It is undoubtedly a Fodiator, and almost certainly be- longs in the Pacific subspecies. Thus, the eye is 24.3 per cent of the head, a Pacific characteristic; the snout, at 36 per cent of the head, could fall within either category; pectoral rays 13, while not a clear cut character, also suggests the Pacific sub- species ; and the locality of capture, the Hawaiian Islands, pro- hibits any other conclusion. 8 Postilla Yale Peabody Museum No. 29 CoNCLUSIONS The subspecies of Fodiator acutus have been shown to be inaccurately diagnosed, but analysis of various characteristics permits a redefinition of the subspecies, as follows: A. Atlantic Ocean. Eye 29.2% (27.1 — 31.5%) of head: snout 35.1% (33.6 — 36.9%) of head in west African specimens, 40 — 41% in Caribbean. Fodiator acutus acutus (Cuvier and Valenciennes) AA. Pacific Ocean. Eye 25.4% (22.7 — 28.1%) of head; snout 39.6% (36.1 — 42.7%) of head. Fodiator acutus rostratus (Giinther) Feb. 12, 1957 Subspecies of Fodiator Acutus 4) LITERATURE CITED Breder, Charles M., Jr. 1938. A contribution to the life histories of Atlantic ocean flying fishes. Bull. Bingham oceanogr. Coll., 6 (5): 1-126. Breder, Charles M., Jr. and John T. Nichols 1934. On the significance of vertebral counts in exocoetid taxonomy. Proc. biol. Soc. Wash., 47: 37-43. Bruun, Anton F. 1933. On the value of the number of vertebrae in the classi- fication of the Exocoetidae. Vidensk. Medd. dansk naturh. Foren. Kbh., 94: 375-384. 1935. Flying fishes (Exocoetidae) of the Atlantic, system- atic and biological studies. Dana Rep., 6: 1-108. Giinther, Albert 1866. Catalogue of the fishes in the British Museum. Vol. 6, xv + 368 pp. Brit. Mus. (Nat. Hist.) 10 Postilla Yale Peabody Museum. Feb. 12, 1957 LEGEND FOR FIGURE Figure 1. Distribution of several characteristics in Fodiator acutus. A. Pre-dorsal scales. Upper line, Atlantic sample; lower line, Pacific sample. B. Interorbital width expressed as percent of head length. Upper line, At- lantic; lower line, Pacific. C. Snout as percent of head length. Upper line, West African specimens; middle line, all Atlantic material (dashed base line shows discontinuity in sample); lower line, Pacific specimens. D. Eye as percent of head length. Upper line, Atlantic; lower line, Pacific specimens. tt a ae a > ee 7 - faa a Se en ar oF La - MUs. COMP. ZOOL. p) i; // LIBRARY [ostilla ae YALE PEABODY MUSEUM HARVARD or Natrurat History UNIVERSITY Number 30 February 28, 1957 New Haven, Conn. NOTES ON THE HORNED COOT, FULICA CORNUTA BONAPARTE* S. Ditton RieLey In May, 1956, while on a visit in South America on behalf of the International Committee for Bird Preservation, I was fortunate enough to mect Senor Luis E. Pena G. through the good offices of Dr. R. A. Philippi, the well-known Chilean orni- thologist. Senor Pena, an entomologist, has recently spent two seasons in the highlands of northern Chile, in the mountains of eastern Antofagasta just south of the Bolivian border. His trip has been well described in his own words (1954). On his trips, Senor Pena has been lucky enough to observe one of the rarest and least known birds of the world, the Horned Coot, Fulica cornuta, The other naturalist who has seen them, also a Chilean, is Senor William R. Millie, who found the species and the first known nests on Laguna Grande in the high region of Huasco in 1936, 1945, and 1946. Some of Millie’s observations are given in “Las Aves de Chile” by Goodall, Johnson and Philippi (1951). Both these gentlemen have been kind enough to furnish me with notes on the species, and from Senor Pena, I have obtained a pair of his specimens which are now in the Yale Peabody Museum Collection. Fulica cornuta was described by Bonaparte from a single specimen collected in the highlands of Bolivia, in 1853, and for many years this skin remained unique. It was illustrated in a drawing of the head in a paper by Sclater and Salvin (1868). Hellmayr and Conover (1942) list eight specimens of the * An informal synopsis of this paper was presented at the Seventy-fourth Stated Meeting of the American Ornithologist’s Union in September, 1956. Ir Postilla Vale Peabody Museum No. 30 species in museum collections, all from the highlands of south- ern Bolivia or northwestern Argentina in Tucuman. Besides these, there was apparently one other specimen collected at Laguna Blanca, Catamarca Prov. Argentina, 1918, now in New York, and five specimens from northern Chile subsequently taken by Millie and Pena, making a total of fourteen in mu- seum or private collections. Fulica cornuta and Fulica gigantea, the Giant Coot, are the two largest coots in the world, measuring up to 19 or 20 inches in total length. The Giant Coot also has a restricted range to the north of the Horned Coot, in the cordillera of central and southern Peru, Bolivia and extreme northern Chile. The two species are apparently allopatric. The Horned Coot, however, differs from all other members of its family in the extraordinary wattle which arises in the frontal area. Where all other species of the tribe of coots and gallinules characteristically possess a horny shield or cutan- eous structure in the area of the forehead, posterior to the bill this species possesses a wattle which is identical in both sexes, and which is flanked by a pair of smaller wattles. The central, large wattle, which in our specimens measures 6 29, 2? 33 mm. in total length in the dry skins, is fleshy and perhaps somewhat extensible or erectile, although neither Pena nor Millie noticed this in life. In these specimens, in breeding condition taken on October 9, 1955, the wattle is far more developed than in the other museum skins which I have examined; some of which cer- tainly were immature birds. It may be, however, that out of the immediate breeding cycle, the wattles shrink in size. The female in Dr. Philippi’s collection, secured by Millie and said to be on the nest, has a poorly developed fleshy protuberance, about the size of the drawing of the type in Sclater and Salvin (op. cit.), and similar to three immature specimens in the American Museum of Natural History’s collection in New York. The two small flanking wattles in our specimens stand erect on either side of the large wattle, and measure; ¢ 4, 2 6 mm. in length. All three wattles terminate in tufts of thick celluloid- like filoplumes, further extending the length of the structure by some 15-20 mm. In addition, the large wattle which lies forward over the bill, pointing in an anterior direction, is coy- Feb. 28, 1957 Horned Coot, Fulica Cornuta Bonaparte 3 3 rde, 4200 m. alt. Ve Senor Pena at a nest site on Laguna One of Senor Pena’s specimens of the Horned Coot contrasted with Larus serranus for size. Photos by L. E. Pena G. } Postilla Yale Peabody Museum No. 30 ered irregularly on the dorsal surface with small tufts of down, All of this is well illustrated in the accompanying plate by Robert Clem. A strange feature which is apparently more obvious in life, for Millie’s field notes mention it specifically, is a small patch, appearing white in life, lying below the wattle, at the base of the maxilla. Under examination with low power magnification, this is revealed to be a fleshy caruncle, the rugose skin dis- tended into minute patches which appear to be filled with a fatty mass, for in the dried skin, the color has changed from white to a dull, pale yellow. The bill color, as shown in the plate, appears to be olive- yellow in life, brightening towards the base of the mandibles to dull orange. There is a black patch along the culmen, wider at the base including the depression in which the external nares lie, and extending out, more narrowly to the tip of the culmen. The other unique feature of the Horned Coot is its nest. Nesting has been observed by both Mille and Pena. Pena found nests with incubated eggs in December on Laguna Verde (alt. 4300 m.) on the slopes of Vulcan Hecar, in the northeast of Salar de Loyoquis. Millie found nests with young in December and January and a nest with eggs in late November. He also found nest building in progress on November 27, 1946, at Lagunita de Encierro (alt. £200 m.). [I quote from Milhe’s letter: “T watched a pair constructing their nest for about three hours. They, too, had selected a sheltered place with comparatively shallow water. They were just finishing the stone structure made of stones of the size of small potatoes, carried there by them in their beaks. On this mound which I later measured to be about 1 mt. in di- ameter and 60 cms. high and which I calculated to con- sist of at least 1. 14 tons of pebbles they then proceeded to place algae* carried to and fro in rapid journeys. *Mr. Millie has subsequently been again to the high cordillera of the Atacama, and at my request sent me on November 24, 1956, a specimen of plant material from the lake, which he asserts the coots use in nest building, and also feed on. IT am grateful to Professor Gilbert M. Smith of Stanford University, who has examined this specimen and reports that it is in fact not alga, but Myriophyllum, a flowering plant, and a far more probable source of nest building material than an alga. No. 30 HorNneEp Coo1 5 Robert V. Clem Head of Fulica cornuta No.-30 ° Postilla Yale Peabody Museum apse, bunboq ul ‘dpuog ‘ojnus02 o21/N7 }O j\seu 9UL ISD “$V [SAB] A9LOM WW OOS b SQUOJS isoun oul Ul scbbS Feb. 28, 1957 Horned Coot, Fulica Cornuta Bonaparte 7 This was accomptished by swimming out to where they found the slime whence they would tip up with head down and come up with a load of the material and carry it to their nest, first making a sort of landing ramp. When I left that evening they had made quite a large portion of the nest. Both birds worked in this home building proj- ect. Several old nests from previous years were also seen and all had these stone mounds as a foundation.” Pena describes a similar nest on Laguna Verde, 40 meters from the shore and in water 40 centimeters deep, covered with vegetable material and based on a truncated cone of stones. A diagramatic sketch of the nest is given. The eggs of Fulica cornuta are roughly similar to those of the Giant Coot, about the size of a turkey’s egg, and vary from 58.5 to 78 mm. in length and 38.2 to 58 mm. in width, stone gray to buff in ground color, speckled or blotched with dark gray or brown. The clutch consists of three to four eggs. Fulica gigantea of Peru, like other members of the family, makes a more typical nest of a mat of floating water weeds. The Giant Coot also nests twice a year, in August and again in December. No other known bird constructs its own island of stones on which to nest. All authors who have studied the nat- ural history of this xerophytic area in the high Andes of northern Chile and Bolivia, speak of the paucity of vegetation. It seems possible to speculate that this stone platform nest habit has evolved in response to the lack of vegetation, and also perhaps to the presence of predators on the shores. The local Black-headed Gull, Larus serranus found breeding on Laguna Verde by Pena was nesting on a projecting stone, two meters out from the shore of the lake. Other associated bird species seen by Penta in December be- sides the gull, were the Andean Flamingo, Phoenicoparrus andinus, and in October, the Junin Grebe, Colymbus occipitalis jumensis, and Puna Teal, Anas versicolor puna and the Andean Crested Duck, Lophonetta speculoides alticola. The October birds, taken by Petia were paired and courting, thus possibly explaining the enlargement of the wattles. Pena reports that the weather is ferocious at this altitude in October, the wind almost ceaseless from the WSW. At 11 a.m. part of the lagoon was covered with ice which he had to break and swim 8 Postilla Vale Peabody Museum No. 30 in to retrieve his specimens. The only moderate season in this area is from December to February, and this probably deter- mines the nesting cycle which differs from that of Fulica gigan- tea. No definite information is available about the territorial be- havior of Fulica cornuta, Goodall, Johnson and Phillipi (t. e.: 187) report 36 nests of Fulica gigantea on all parts of Lake Cotacotani. Neither Mr. Millie nor Mr. Pena have reported more than a single pair of Fulica cornuta on any one lake, though they have spoken of abandoned and old nest sites. From the meager evidence available, it would appear that Fulica cornuta tends to be territorial and ungregarious. SUMMARY From the above observations it appears that the Horned Coot, unlike the Giant Coot, nests only once a year, from the end of November to the beginning of January, that its pre- ferred nesting area is small lakes in northern Chile, southern Bolivia and northwestern Argentina above 12,000 feet, more commonly above 13,000 feet in essentially a xerophytic zone. During the courtship period there is a considerable develop- ment in both sexes of elaborate frontal wattles with an associ- ated local caruncle. Actual courtship behavior has not been observed, but must involve use of these highly developed ap- pendages, as the frontal shield is used in other coots, described by Gulhon (1953). Unlike other birds, Bonaparte’s Horned Coot builds its own Elba, a nest composed of an island of stones erected by the pair, and covered with a mass of plant material. LITERATURE CITED Goodall, J. D., A. W. Johnson, and R. A. Philippi, 1951, “Las Aves de Chile,’ Buenos Aires, 2:184-187. Gullion, G. W., 1953, Condor, 55:169-185. Hellmayr, C. KE. and B. Conover, 1942, Field Mus. Nat. Hist. Zool. Series 13(1), No. 1:422. Pena, Luis E., 1954, “Kxploraciones en la Cordillera de Antofagasta,” Rev. chilena Hist. nat. No. 14, 54:163-186. Sclater, P. L. and O. Salvin, 1868, Proc. zool. Soc. London: 463. ae Wee oe Ce) oe ) Libsant | c7} } te \n py 2 31957) : + Babee! hs \ 3: 2 lil, [ | guyveRsity | CALL a _ UNIVERSE YALE PEABODY MUSEUM or Naturat History Number 31 March 28, 1957 New Haven, Conn. NEW BIRDS FROM THE WESTERN PAPUAN ISLANDS S. Ditton RIpPLtey As a preliminary to further publications on the work of the 1954 Yale Peabody Museum Expedition to the Moluccas, I should like to describe the following new forms of birds col- lected on the trip undertaken by myself and my wife. Following our departure from Netherlands New Guinea in December, 1954, my assistant, Jusup Khakiaj, made a small collection of birds on the Islands of Misool, the Schildpads, Kofiau and Waigeu in early 1955. In addition to funds from Yale, my work was financed by a Fellowship from the Guggenheim Foundation as well as a Grant from the National Science Foundation. To the authorities of these organizations I am deeply grateful. I must also record my thanks to Drs. Dean Amadon and Charles Vaurie of the American Museum of Nat- ural History, and Mr. R. M. de Schauensee of the Philadelphia Academy of Natural Sciences for help in examining specimens in their care. Aepypodius arfakianus misoliensis subsp. nov. Type: 2 ad. (Y.P.M., no. 36560), collected November 22, 1954, by S. Dillon Ripley ten kilometers W.N.W. of Tamulol, Misool Island, Netherlands New Guinea. 2 Postilla Yale Peabody Museum No. 31 Diagnosis: Compared to arfakianus, a series of three speci- mens, have more slender, slimmer bills which appear not so highly arched as in the mainland population from New Guinea. In addition, the feathers of the vent are tipped with slaty-gray, paler and lighter than in New Guinea birds, and the chestnut of the upper and under tail coverts is duller, less rich in appearance. This population appears to be smaller in size also. Measurements: arfakianus 94 8 2 9 wing tail bill (from ext. naris) 260-272 (267.8) mm. 180-149 (141.9) 20-22 (21.4) misoliensis 36 2 2 wing tail bill (from ext. naris) 243, 261.5, 264 mm. 131.5, 141.5 19, 20, 21.5 Range: Misool Island, Netherlands New Guinea. Remarks: The occurrence of this large Bush Turkey on Misool Island as well as Cuvier’s Bush Turkey, Talegalla cuvieri, is a remarkable discovery. It is planned to publish detailed comments on these species at a later date. Eos squamata attenua, subsp. nov. Type: 9? ad. (Y.P.M. no. 36561) collected March 22, 1955, by Jusup Khakiaj on Kamoa I. Schildpad Is. north of Misool. Diagnosis: From squamata this form differs by having a much reduced nuchal collar. Only one specimen of three shows a patch on the nape of the neck, and in all the speci- mens the prominent patch on the foreneck and upper breast extending to the throat is lacking or only lightly indicated with a few purplish blue tips to the feathers. The under sur- face of the tails of these birds tends to be rather brighter and more reddish, more like obiensis. From this form as well as \ March 28, 1957 New Birds from Western Papuan Islandsy\\8 guenbiensis it differs, however, in lacking the purplish oc- cipital spot, the pronounced collar, and, in obiensis, the black scapulars and greater wing coverts. Range: Kamoa, Lophon and presumably in the rest of the group of the Schildpad Is., north of Misool. Crateroscelis murina fumosa subsp. nov. Type: ¢ ad. (Y.P.M. no. 36562) collected Nov. 18, 1954, by S. Dillon Ripley inland from Tamulol, Misool Island, Netherlands New Guinea. Diagnosis: This form is nearest capitalis from Waigeu from which it differs in the male in the head being darker, more blackish brown. Below in both sexes Misool birds are much more reddish on the underparts. Compared to C. m. monarcha from the Aru Islands this population appears to be much more richly colored. Like capitalis the Misool birds are smaller than typical murina of New Guinea with the top of the head in the male darker, more richly colored. wing tail culmen Measurements: 6,22 58, 52.5,55 33, 33.5,35 16, 15(2) Weight: 6,22 15, 12, 13 grams. Range: Misool Island, Netherlands New Guinea. Gerygone magnirostris occasa subsp. nov. Type: ¢ ad. (Y.P.M. no. 36563) collected May 2, 1955, by Jusup Khakiaj on Kofiau Island, Netherlands New Guinea. Diagnosis: From cobana, brunneipectus and conspicillata, this form differs by being much more richly yellow on the abdomen, belly and under tail coverts, in this character ap- proaching the form, rosseliana from the Lousiade Archipel- ago, from which it differs in having the throat whitish and with a brownish tinge on the pectoral area. On the upper- parts it is darker, more brownish olive than the geograph- ically neighbering forms mentioned above. Above it is close to affinis from northern New Guinea, although again it is much more brightly colored on the lower surface than that form. sr ot 4 Postilla Yale Peabody Musewm No. 31 wing tail culmen Measurements : 53.5 37 11.5 mm. Range: Kofiau Island, Netherlands New Guinea. Remarks: This form is described from a single specimen un- fortunately, but a specimen which is so strikingly different from its geographical neighbors that it would seem to re- quire recognition. Xanthotis chrysotis austera, subsp. nov. Type: ¢ ad. (Y.P.M. no. 36564) collected Nov. 15, 1954, by S. Dillon Ripley at Tamulol, Misool Island, Netherlands New Guinea. Diagnosis: From chrysotis this form differs by being darker and possibly somewhat smaller in size. The tone of the upper- parts is darker olive, brownish. The throat is dark gray, tinted at the lower edges with olive green. The lower parts are dark, considerably darker than chrysotis, and much darker than fusciventris. This form is much brighter, more yellowish-tinted on the breast and more olive-tinted on the back than saturatior, differing from that form as does chry- sotis. Measurements: 2¢ 6, 22 % wing tail culmen $ 96,103.5 (moult) $ 83,84 (moult) $ 29,31 ? 92,94 (moult) 273,77 (moult) ? 25.5,26 Weight: ¢ 49, 2 38 grams. Range: Misool Island, Netherlands New Guinea. Remarks: The wing measurements of typical chrysotis fall within the range of 100-110 mm. for males and 95-100 for females. It would appear that the Misool birds are smaller than the form from the mainland of New Guinea, although lack of material and moult in two specimens prevents com- plete certainty in this case. patilla YALE PEABODY MUSEUM oF NatTurRAL Histrory AAz : toe een Lili Biwi H Number 32 June 20, 1957 New Haven, Conn. ADDITIONAL NOTES ON THE HORNED COOT, FULICA CORNUTA BONAPARTE S. Ditton RIPLEY Subsequent to the publication of the notes on this curious species contained in Postilla No. 30, Feb. 28, 1957, new in- formation has come to light which seems worth printing here. Again I am indebted for these added observations to Sr. Wil- ham R. Milhe of Vallenar, Chile, as well as to Sr. A. W. John- son of Santiago, who have recently returned from a trip to the high altiplano of extreme northern Chile and western Bolivia. Mr. Johnson and Mr. Millie found a group of twelve Horned Coots on an artificial lake called Tranque Caritaya, altitude 3600 metres, in the south of the Department of Arica, Chile. This extends the range of Fulica cornuta north more than 500 kilometres from the previously known range in Chile. Lake Caritaya is only 60 miles from Lake Cotacotani where Fulica gigantea has been observed (1951, Goodall, Johnson and Phil- ippi, Las Aves de Chile, 2: 185-188), although it is 1200 metres lower in altitude. At this lake Messrs. Johnson and Millie found three nests on February 9 and 10, 1957, of which one was occupied by a female brooding a clutch of eggs, about one-third advanced in incubation, and the others gave evidence of having been recently occupied. A point of great interest was that the nests were con- structed entirely of vegetation, apparently Myriophyllum, vide Millie, in the usual coot fashion, but that the shape of these nests was similar to those made of stones far to the south, being cone-shaped with the greater part under water ap- parently resting on the bottom, unlike the mat-shaped, largely floating nests of Fulica gigantea and the smaller species. It seems possible to speculate, therefore, that the nests made of stones in the alpine xerophytic zone where Fulica cornuta has previously been observed, which have only a coating of Myriophyllum on the surface, have been developed as a unique nest building habit in direct response to the lack of vegetation, and that where vegetation is abundant this species will build a nest using traditional materials, although in a particular shape, peculiar to itself. Three further points of interest emerge from these observa- tions of extension of range of this species. The presence of several birds on Lake Caritaya implies at least that Fulica cornuta is more tolerant than its occurrence in the south would suggest. Perhaps the local abundance of food and nesting ma- terials allows compression of territories in this situation. In addition it would appear that Fulica gigantea and Fulica cornuta are at least geographically sympatric, although the species may be altitudinally or ecologically isolated in this zone of presumed overlap. Finally, the lateness of this nesting date in February might allow the Horned Coot to nest twice in the year. The Giant Coot nests in August, and again in November-December. Fulica cornuta has been known to breed only from late November to early January, now to February. Further investigation is needed to determine whether a double nesting cycle may occur in this species in the northern part of the range. YALE PEABODY MUSEUM oF Natura. History Number 33 Nov. 1, 1957 New Haven, Conn. ON A NEW SPECIES OF ANISOPS (HEMIPTERA, NOTONECTIDAE) FROM THE MOLUCCAS G. E. Hutcuinson' In a small collection of aquatic insects made by Dr. S. Dil- lon Ripley in the Moluccas, two specimens of the genus Ani- sops are present. One of these is a female specimen from Misool which is possibly referrable to A. stali Kirk.; in the absence of a male it is obviously impossible to make a precise identification. The second specimen appears to represent a new species. Anisops sylvia sp. un. Stramineous with dark venter abdominis, no special pigmen- tation. A moderately small, fairly wide-headed species with a long pronotum, anterior half of body subparallel, widest just before the middle. 6 Head wide (1.84mm), almost as wide as pronotum, about four times the anterior width of the vertex (0.36mm) which is just over twice the width of the synthlipsis. Anterior margin ' Department of Zoolegy, Yale University. ae oslt Wy a S | bs 2 Postilla Yale Peabody Museum No. 33 of head between eyes very slightly rounded and hardly pro- jecting, frontal interocular region hardly visible in lateral view, no longitudinal roll-like ridge between eyes (fig. 1). Pronotum just over twice (1.47mm) the length of the head (0.67mm) as seen from above and three fourths as long as wide (1.95mm) ; sides slightly diverging from anterior to hu- meral angles, posterior margin very slightly flattened centrally ; exposed portion of scutellum about half (0.70mm) the length of pronotun. Facial tubercle low, simple, glabrous, with traces of slight lateral depressions just inside postero-ventral corners of eyes. Labrum very short, much wider than long (fig. 2); prong of third rostral joint very slightly longer than the joint itself, apex subacute. Anterior femur subparallel, suddenly constricted apically, tibia with comb in two sections, of eleven small proximal tooth- like irregularly set elements and sixteen distal lamelliform elements; tibial chaetotaxy as in figure 3; claw one-third the length of tarsus. Intermediate claws equal and regularly curved. Dimensions of joints of legs in millimeters : Femur Tibia Tarsus MuIRerior oS So ae Ae ete ee 1.38(100) 1.2490) 0.86 (62) Mnterme diate rere r 1.60(100) 146(91) 0.66(41) 0.54(34) Postenione si anacw tat oa 2.72(100) 2.16(80) 1.18(43) 0.7327) Length 6.4mm, maximum breadth 2.0mm. INDONESIA: Molucca Archipelago; Gng Sibela, Batjan Is. Sept. 19, 1954. 14. S. Dillon Ripley (holotype P.M.) In the structure of the tibial comb, with a series of short proximal followed by long distal elements, A. sylvia resembles, among the ninety well-known (Brooks, 1951) species of the genus, only A. nigrolineata Lundblad (1933) from Java, India, Nov. 1, 1957 New Species of :Anisops 3 Anisops sylvia, sp. n. holotype 1. Lateral view of head. 2. Labrum and rostral prong. 3. Anterior tibia. Anisops nigrolineata Lundblad, Sohawa, Pakistan 4. ‘Tibial comb of 2. eon Postilla Yale Peabody Museum No. 33 Pakistan, Burma and the Philippine Islands. The new species differs strikingly from A, nigrolineata in lacking the longitudi- nal rounded ridge between the eyes on the antero-ventral region of the head; the seriation of the elements of the comb is even better developed, with more numerous but more reduced distal elements in 4. sylvia than in A. nigrolineata (fig. 4). The new species resembles both A. nigrolineata, and A. paranigrolineata Brooks from the uplands of central and south India, in the excessively short labrum, and in the long pronotum. The three species clearly form a natural group and are certainly more closely allied to each other than to any other members of the genus. I am greatly indebted to my friend Dr. Ripley for the opportunity to describe this species, and to Mrs. Nancy S. Kimball for professional assistance with the figures. REFERENCES Brooks, G. 'I., 1951. A revision of the genus Anisops (Notonectidae, He- miptera). Kans. Univ. Sci. Bull. 34: 304-519. Lundbald, O., 1933. Zur Kenntnis der aquatilen und semiaquatilen Hemip- teren von Sumatra, Java und Bali. Arch. Hydrobiol. Suppl. 12: 1-489. ; A |_WlversiTy oslt ie i YALE PEABODY MUSEUM or NatTurAL History Number 34 March 7, 1958 New Haven, Conn. NOTES ON AN ADDITIONAL EXAMPLE OF THE FRUIT BAY, SCOTONYCTERIS OPHIODON POHLE Auvin Novicx’ In the Peabody Museum collection of bats there is a single specimen of a fruit bat which appears to be closely allied with Scotonycteris ophiodon Pohle (1943) and Scotonycteris ophio- don cansdalei Hayman (1945), both of which are known only from the type specimens. Scotonycteris ophiodon Pohle. Skin and part of skull, Y.P.M. #9442, collected in Liberia, 1928( 7), by G. P. Cooper. Most of the cranial portions of the skull are missing, includ- ing the posterior and ventral borders of the orbits and the zygomata. This specimen is similar to Scotonycteris ophiodon Pohle and to S. 0. cansdalei Hayman in general size and in external and dental characters. Like cansdalei it has conspicu- ous white patches at the posterior angles of the eye and ex- tremely faint and inconspicuous white tufts at the anterior base of the ears (both spots lacking in ophiodon). The white border of the upper lips is conspicuous for only two-thirds of the way forward towards the nostrils but can be traced faintly 1 Department of Zoology, Yale University. 2 Postilla Yale Peabody Museum No. 34 all the way to the nostrils. In ophiodon the white border is said to include the nostrils; in cansdalei it reaches two-thirds of the way to the nostrils. These two forms have otherwise been dis- tinguished by cranial features which can not be assessed in this specimen. Neither author has mentioned the conspicuous yellow- ness of the skin ventral and anterior to the orbit and the small bright yellow patches of fur ventral to the postorbital patches of white seen in this specimen. The skin of the rostrum is faintly yellow; the skin under the jaw and extending back to the breast is also yellow, darker anteriorly and fading to a faint yellow posteriorly. Otherwise Hayman’s description of the fur and color of cansdalei agrees in detail with the present specimen. The measurements of the three specimens are compared in Table 1. The present specimen differs markedly from the other two in total length but this measurement is unreliable in pre- pared specimens. The hindfoot is somewhat longer as are all of the metacarpals while the forearm and pollex are in the same range. The palate is similar in total length but the post dental palate is slightly longer. The breadth m'—m‘ is greater but the breadth e—c and the interorbital constriction are only slightly greater. The mandible is distinctly longer and is also higher at the coronoid. The teeth are all but identical in size and form with those of the previously described specimens. Thus this bat differs most interestingly from cansdalei in having increased yellowness of the skin of the head with the appearance of yellow tufts of fur posterior to the eyes. Furthermore it differs in having a longer hind foot, longer metacarpals, and a longer and higher mandible. REMARKS In 1943, Pohle described S. ophiodon from Bipindi, Camer- oons, as the second species of the genus, previously known only from the genotype, S. zenkeri, whose range included the Camer- oons and Fernande Po. §. ophiodon was characterized by its much greater size and by striking dental peculiarities, of which the most important are the secondary cusps on the inner edges of upper and lower canines and the heightening of the canines and March 7,1958 Scotonycteris Ophiodon Pohle wie cheekteeth; the latter also being provided with prominent in- ner cusps. S. 0. ophiodon also lacks the white spot behind the eye found in zenkeri but agrees in most other respects in mark- ings and color. Pohle considered that some of the characters of ophiodon, particularly those of the cheekteeth, showed affinity to Casinyc- teris argynnis, whose close external similarity to Scotonycteris has been discussed by Andersen (1912). Pohle felt that the palatal characteristics of Casinycteris by which the genus is principally defined were unstable and proposed to relegate the genus to synonomy with Scotonycteris. Hayman (1945), how- ever, has convincingly defended the independent position of the genus Casinycteris. Thus the three species belonging to two genera—Scotonycteris zenkeri, S. ophiodon, and Casinycteris argynnis—present an interesting group. The heightening of the inner cusps of the cheekteeth of ophiodon is a feature very closely approaching the dentition of Casinycteris argynnis, rather than S. zenkeri. The normal palate is shared by the two species of Scotonycteris but not by Casinycteris. The sec- ondary cusps of the canines of ophiodon are found neither in zenkeri nor in Casinycteris. Externally these three are very similar, being distinguished only by details (which may well be variable) of the facial markings. The white ear tuft which is characteristic of all other epomophorine bats disappears in zenkeri and ophiodon but reappears in cansdalei, in the present specimen, and in Casinycteris ; the white spot behind the eye is not found in ophiodon. Yellow postorbital spots appear only in the present specimen. It appears that these three species form a natural group as judged by external appearance, dentition, cranial character- istics, wing membrane insertion, and many other considerations. The material at present is too sparse to attempt a clear analy- sis of the larger group when so few specimens represent most of the forms. It appears that its aberrant palate justifies re- taining the genus Casinycteris. I am not prepared any more than was Hayman to erect a new genus for ophiodon because of its dentition. I feel that its clear dental separation from zenkeri should be emphasized by the erection of a new subgenus. + Postilla Yale Peabody Museum No. 34 Hayman chose to express the position of cansdalei as a sub- species of ophiodon with the comment that the differences might be of specific value. I am reluctant to name a new form from the present specimen in view of its incomplete skull and of its being a single specimen to be compared with only single specimens of both ophiodon and cansdalei. The differences, nevertheless, be- tween the Peabody Museum specimen and the types of ophiodon and cansdalet appear to be slightly greater than the differences between the latter two. For the time, I suggest considering this specimen to be a variant of Scotonycteris ophiodon Pohle. Its discovery in Liberia increases the known range of this fruit bat to include Liberia, Gold Coast, and the Cameroons. LITERATURE CITED Andersen, Knud. 1912. Catalogue of the Chiroptera in the Collection of the British Museum. 2d ed. Vol. I. Mega- chiroptera. London, British Museum of Natural History. Hayman, R. W. 1945. A new Scotonycteris, with notes on other Gold Coast bats. Ann. Mag. nat. Hist., (11)12, 766-775. Pohle, H. 1943. Scotonycteris ophiodon, sp. n., eine neue Art epomophoroider Flughunde. S. B. Ges. naturf. Fr. Berl. 1942 (1943) : 78-87. Or March 7, 1958 Scotonycteris Ophiodon Pohle TABLE 1 All measurements are in millimeters. Where a blank appears, the portion concerned could not be measured. The measurements of ophiodon and cansdalei are taken from Hayman (1945). Scotonycteris S. o. Wo 1a Mile 0. ophiodon cansdalei #9442 leadgand eb odiyigerrtcrivaciea oe ere ase 105 115 143 Aerials ns Baye tapes okay ogeasyarsnelinrs Pr snsnale sors oys 1 ae palpable EAU LOOT torr ateysp ohare aoiers sr ousmetotsy sos esr ere 14 15 19 TATE Wye acsres Se ps cucis ehore 6, oto a ck staiolea aes: 20.5 22 : IRS OTE BETA Be aco) 2) ssi ej oe e a vo eget Svea ran sekehes 75 76 73.8 RON GX aga 5 cies nesevs eee eels Sos roe woe 8s 36.5 31 35.2 ZHOSMeELACATP Aes lc ciereitacr sie-- eieler° 39 35 43.2 SROGMELACAT PAL cciescissryee ee ss c\sres> 54 52 58.5 An sTaetieerna el SoncaoccsopodoKsGocED] 50 49 55.5 HIN WHRCAADE 5 ocoogonoocodccuoenbES 51 50 54.9 Skane totalsleng thrice cet el 36 36 Palation} tomnes foramina =... 2-222. .> 17.6 We 17.7 Palationetombasion yee arriecacec ose ac 12.2 13 ane ost-dental palate. sc. -s-2 5 sor: 6.4 6 7.0 RUOS EMI fe aocewereicustcin aie ove eronsaysisiata: siete ets 9.2 10 9.6 Braincase at zygomatic root .......... 16 15 Prycomaticnbreadtheerenerrecce eerie: 21 22.5 a Breadth! yeaa cas ae cee aer 12.6 12 13.4 Read the — Cx. Ach sie sist evceses = /ehssle/s 1s) hi ae 6.7 7.5 7.8 Breadth of postorbital processes ...... 11.2 14 a Interorbital constriction ............. 6.4 7 7.2 Diametersomonroil near cee 9 9 mye Weng thpoiemandibletase eerie 26.5 26.5 29.5 leightraticononoldieanret-)iiirs mec 11.2 12 14 Wipper tooth mow c¢—m3i eee se. er 11.9 12 11.9 PLCIG NACH Meola we nersievate tects came ees 6.1 DD 6.0 Ve Nts Pt aiavas cc ote cuntvenre cvenire + eientea 3.8 4 4.1 VEIL U RS 5 ayer stos oeisvans eteaeve wre sieig stevens 3.1 3 2.8 Pete htwnite sete nace. os sakes 2.3 2 2.3 ISIGTE NE GG Sones dooosacocusansooocar 4.1 4 4.1 ISIC IE NE 1Ds Sparco sememo ced on oceacs 4.1 4 3.9 SIONS 1D Se eacabocuenosbooadauooe oe 3 3 3.2 ISIDEINE I page manccoemoes soos ssacocE 2.4 2.1 2.2 SIGINT, Gosacuceddabber woah sno cn 1.1 1.2 1.3 Svar = a) Bos acthalahy a“ in apd DAB, oie? . we s ree ees n -_ - - : ere. - = i= 7 \ : i” ; : ‘ile : « ; * wo @ — Ue d a ¢ i 7 1s ¢ ; a 76 bd Q é oe y 2 =-2? » & ro j 4, a . ; ; 7 r ah ( i ee : ‘ > ~ | > = © c ; : =a ANE iL 7 —— a 7 ae = > * & af = 7 , - - - - et. ee =@ * ~ a oumpe- 6g i a va 2 ee ee ee torte ie —" ie 6 ii eel me Ye me: WO eG i ty ip hea 460 1% ¢ 2 we) | a eee ® , : a Pe Ms ry Lstill, ‘~tuad YALE PEABODY MUSEUM oF NATuRAL History Number 35 April 21, 19538 New Haven, Conn. NOTES ON INDIAN BIRDS. VII.* S. Ditton Ripeiery Pycnonotus leucogenys (Gray ) In my forthcoming hand-list of Indian birds, I have treated the White-cheeked Bulbuls as belonging to a single species. I have been interested, therefore, to read a disquisition on this interesting problem by Vaurie (1958, Amer. Mus. Novit. No. 1869 :14-15) whose conclusions differ from mine. Pycnonotus leucogenys leucogenys is a sprightly, cockaded bulbul which ranges along the Himalayas from extreme eastern Afghanistan at medium altitudes up to 8000 feet, following the main river valleys such as the Kunar, Jhelum, etc., moving into such mar- ginal areas as Hazara and Chitral with the advent of warm weather. Thence it extends east to Nepal, Sikkim and Assam north of the Brahmaputra River. It is a wandering species where- ever it occurs in the northwest, addicted to cultivated areas and moving with the seasons. The distribution in the higher latitudes of its range (above 33° N. lat.) seems to be a mar- ginal fluctuating one, implying that the species has reached the limit of its ecotolerance and is probably subject to sporadic occurrences and numerical fluctuations in this area. Such be- havior may be shown some day to account for the two ques- *The preceding number in this series appeared in Postilla, No. 20, July 9, 1954. 2 Postilla Yale Peabody Museum No. 35 tioned sight records of Zarudny in the Tadzhik S.S.R. They may have been vagrants or strays during a warm weather post- breeding wandering season. Whistler (unpublished MSS and Popular Handbook of In- dian Birds, 1941, and subsequent dates) speaks of this form as occurring in Afghanistan, although he doubts its occurrence in his subsequent publication on Afghanistan (1944, Jour. Bombay Nat. Hist. Soc. 44:518) ; Baker also records it (1922, Fauna Brit. India, Birds, 1:390). The Afghan population has recently been described as a new subspecies, picru, by Koelz (1954, Contrib. Inst. Regional Exploration, No. 1:11) who collected a series near Laghman in extreme east Afghanistan in May, 1957. My own feeling is that these birds may well have recently arrived as summer visitors to the area from the lower valleys to the east. They do not appear to be separable from populations farther to the east, vide Vaurie, op. cit. supra, and Traylor (in litt.) who kindly examined specimens in the Chica- go Museum for me. Pycnonotus leucogenys leucotis is a lowland form, reaching 6000 feet in warm weather in the Kandahar area, more addicted to dry arid localities in tropical dry deciduous and semi-desert as well as desert facies. Like leucogenys, however, it frequents cultivated areas and spreads with the spread of gardens and groves, probably thereby extending its range. This form is the common White-eared Bulbul of West Pakistan and western India extending north to Kandahar and southern Afghanistan. West, the species extends to Iran, Iraq, and Arabia. P. l. leucotis lacks a crest and, therefore, appears quite distinct from leucogenys, especially in museum skins. Dr. Bowdler Sharpe was, at one time, inclined to separate the two forms into different genera (vide Whitehead, 1909, [bis:113). My own reaction when I first saw lewcotis and lewcogenys separately in the field was that although in habits and behavior they were identical, nonetheless they were recognizable in terms of plum- age differences as separate species. However, there is an intermediate population which occurs at exactly the meeting place of the two forms, the upper Indus River in North West Frontier Province of West Pakistan in the Bannu and Kohat districts where the Kabul River joins April 21, 1958 Indian Birds. VII. = the Indus. This population has been discussed by Whitehead (1909, op. cit :112-114) under the name Molpastes leucogenys. He does not seem to have been aware that he was discussing Pycnonotus leucogenys hum which differs from leucogenys in possessing a blackish head with reduced olive wash on the back. The range of hum, which was described by Oates in 1889 (Fauna Brit. India, Birds, 1:274), extends up the Kabul River at least to Jalalabad thence east in Bannu and Kohat districts to Cambellpur and Rawalpindi, Jhelum and the Salt Range of West Punjab. Vaurie, on the basis of one specimen (1958, op. cit:16). has given a key to the species which imphes that hwmii is closer to leucotis and so should be listed with it. Specimens in the Yale Collection and others I have examined in London show that Vaurie’s description is at fault and that hwmii is, in fact. an intermediate. Vaurie’s characters are as follows: leucogenys leucotis humii size “larger, more smaller small, but with robust, distinct- a “slightly lar- ly larger bill ger” bill than with more leucotis strongly devel- oped rictal bristles” [The above is not supported by the evidence. Typical leuco- tis has a smaller bill, but when the western, Persian Gulf forms are compared, their size becomes equal to lewcogenys. As is so often the case with chains of subspecies, nearly contiguous forms tend to have more pronounced size differences than those more spatially separated. | color “greenish on the “grayish “identical with back and rump brown” in fresh nominate in fresh plumage leucotis” plumage” | The above is not entirely accurate. The intermediate humit, 4 Postilla Yale Peabody Museum No. 35 in fresh plumage, has a faint greenish-olive tinge on the back, exactly intermediate between leucogenys and leucotis. | crest leucogenys “narrowly “no crest... not leucotis edged with edged... no white . . . more streak” conspicuous above the eye thus forming a distinct super- ciliary streak” humii “crest feathers black, not edged with white. . broader and rounded” |The above does not entirely describe the situation in hwmit which has the feathers of the head exactly intermediate between the condition of leucogenys and leucotis, the feathers slightly lanceolate but somewhat broader, with pale edgings to the feathers above and behind the eye and a distinct short white superciliary streak, more prominent in some specimens than others. | crown color “brown” “black” “black” [In humitt the crest feathers are brownish black, not pure black. | { would prefer to list the variations in these forms as follows: Blackish Super- edging to ciliary Back Cheek feathers Crest Streak Color Patch at nape Vent leucogenys long, present greenish- smallest present lemon brown olive yellow brown humiu moder- present dull inter- present lemon ate, but olive mediate yellow brown- reduced brown ish- black Ieucotis lack- lacking grayish- largest nearly chrome ing, brown absent yellow . black — April 21, 1958 Indian Birds. VII. 5 From the above it will be seen that humit occupies the posi- tion of an intermediate in morphological characters, and occurs in the area where the ranges of lewcogenys and leucotis meet. As the form has intermediate and recognizable characters and possesses a discrete range, it would appear to qualify by every conservative definition as a subspecies linking two adjacent subspecies. The possibility must not be neglected that hum is a hybrid, as Vaurie has noted. If it can be shown to be a hybrid, then it is strange that the 30-odd specimens which have been collected have not shown a great deal more variation. It could be that this is a case of interspecific hybridization as shown by Meise (1936, Jour. f. Orn., 84:631-672) for two species of Passer in the Mediterranean, or Chapin (1948, Evolution, 2:111-126) for Paradise Flycatchers, T'erpsiphone species in Africa. Both these cases have been due perhaps to disturbance of the en- vironment. Similarly, in Ceylon (1946, Spolia Zeylanica, 24:218-220), I have described hybridization in two well marked subspecies of Drongos. Bulbuls hybridize in the area, witness the hybrid specimens taken in Bannu District. between Pyenc- notus leucogenys leucotis and Pycnonotus cafer intermedius. I believe, however, that in this case humiti represents a stable form, possibly derived from an original hybridization between two rather morphologically different ancestors. possibly merely one of a series of widely spaced steps in a discontinuous cline of originally geographically isolated and spatially evolved sub- species. Should the former supposition, with its implication of a stabilized hybrid swarm seem feasible, it would provide a most interesting subject of experimental study, as worthy of re- search as the provocative population known as the Marianas Mallard, Anas oustaleti, which has been described as a possible hybrid swarm by Yamashina, (1947, Pacif. Sct. 11:121-124). THe NAME OF THE WHITE-HEADED BABBLER The names of the White-headed Babbler and its close rela- tive, the Bombay Babbler or Jungle Babbler from Ceylon and Peninsular India have been the subject of considerable eon- fusion over the years. G6 Postilla Yale Peabody Muscum No. 35 Jerdon (1863, Bds. of India, 2:59-63) listed the then-known forms of these babblers from India as: Malacocircus terricolor Hodgson from Bengal, the Nepal lowlands or terai and ad- jacent areas; Malacocircus griseus (Gmelin), in the Carnatic, with perhaps a closely allied form, M. affinis Jerdon, in the south of Malabar; Malacocircus malabaricus Jerdon in the Peninsula, and Malacocircus somervillei (Sykes) in Bombay. He mentions M. striatus Swainson of Ceylon which he says is very close to terricolor. Oates (1889, Fauna of Brit. India, Birds, 1:110-114) lists Crateropus canorus (Linnaeus) for the whole of India, with Crateropus griseus (Gmelin) in Southern India, Crateropus striatus (Swainson) in Ceylon, and Crateropus somervillii (Sykes) in the Western Ghats. This is the same arrangement as that of Sharpe, (1883, Cat. Bds. Brit. Mus., 7:480-483) who listed the same four species with equivalent ranges. . More recently, Baker (1922, Fauna Brit. India, Birds, 1:190-195) listed three species with added subspecies for the area: 1) T'urdoides terricolor terricolor (Hodgson) in north- ern India to Bengal, T'urdoides terricolor malabaricus (Jer- don) in southern India, Turdoides terricolor sindianus (Tice- hurst) in West Pakistan and western India; 2) T'urdoides gri- seus griseus (Gmelin) in southern India north only to a line from Ellore, Secunderabad and Belgaum, and T'urdoides griseus striatus (Swainson) in Ceylon; 3) T'urdoides somervillet (Sykes) from Bombay south to Travancore. So in a period of fifty-nine years, these babblers had been reduced from five spe- cies to three, with at least eight names being used for the various forms. In volumes 7 and 8 of his work (1930), Baker introduced two additional names for the first time (pp. 36, 601). These were the substitute name T'wrdus polioplocamus Oberholser, 1920, a new name for Turdus griseus Gmelin, preoccupied by T'urdus griseus of Boddaert 1783. In addition he listed Turdoides striatus (Dumont) 1823, an earlier name than 7’. striatus of Swainson, 1832, and gave to it the type locality of Ceylon. Unfortunately, Baker had apparently not consulted either the original volume of Levrault’s Dictionnaire from which ~) April 21, 1958 Indian Birds. VII. Dumont’s name came, nor the valuable early work of Jerdon, 1847, Illustrations of Indian Ornithology. In the latter volume (text to plate XIX) Jerdon discusses M. striatus Swainson from Cylon and says: “Tt was founded on a Ceylon species which Mr. Swainson identified with a bird in the Paris Museum labelled Gracula striata. It is Cossyphus striatus of Dumeril (Blyth), and Philanthus striatus of Lesson. “It is not impossible, however, that the striatus of the French Museum is one of the allied species, either terricolor, malabaricus, or orientalis, which Swainson might have readily enough mistaken for it. Lesson says it was from Bengal; if so it is probably terricolor.” The most recent and far more biologically interesting treat- ment of these species is that of Whistler and Kinnear (1986, Jour. Bombay Nat. Hist. Soc. 35:737-741) who point out that there are in fact two species involved. One is the Jungle Bab- bler which ranges from West Pakistan and northern India in the Himalayan foothills, Nepal, east to Bengal and western Assam, and south throughout the Peninsula. The second spe- cies is the White-headed Babbler which occurs in Peninsular India as far north as Belgaum on the west, Chanda in the center, and Andhra in the east to the border of Bastar. In Ceylon also, the White-headed Babbler occurs with, in addition, the Rufous Babbler, which is combined nowadays with the Indian Jungle Babbler as a well marked subspecies. In the species somervillet, Whistler and Kinnear recognize four sub- species: sindianus (Ticehurst), terricolor (Blyth), malabari- cus (Jerdon) and somervillet (Sykes). In the species striatus they list polioplocamus Oberholser and striatus (Dumont). Later Whistler (1944, Spolia Zeylanica, 23:131), discussing the species striatus, reintroduces the name affinis (Jerdon) for the Indian Peninsular race of the species, without saying that he evidently considers this an earlier name than that of Ober- holser, as a substitute for the name griseus (Gmelin), pre- occupied. Out of this history of varying nomenclature and usage, one fact emerges. North of the Peninsula proper of India, in areas 8 Postilla Yale Peabody Musewm No. 35 such as Bengal, only one form of this babbler occurs; it is the one which has normally been named terricolor, ascribed either to Hodgson or Blyth, but of whose proper appelation Jerdon (1845, op. cit.) had expressed a doubt. The original description of Dumont (1828, Dict. Sci. Nat. led. Levrault], 29:268) is of a bird which he calls Cossyphus striatus and of which he remarks that several individuals have been received, collected in “Bengale” by Macé and Dussumier, their size being that of a common thrush, and their color domi- nantly reddish gray. One specimen has some brown transverse striations on the breast, and among the others the striations are longitudinal and paler. In response to a query about specimens at the Paris Museum, Professor Berlioz has kindly written me as follows (in litt.): “We have in our old mounted collection a specimen which possibly might have been the type of Dumont’s description ; but nothing is more doubtful as, in the register, it bears only the mention ‘type’ (but of which author?) and, without any label, the only mention, underneath—by a comparatively re- cent handwriting—‘Bengale, Macé’ . . . the head certainly does not show any sign of lighter, whitish color in front, and the culmen (18 mill.) is intermediate in size between our long-billed specimens from the Central Provinces in India and the short-billed from Southern India and Ceylon. ‘However, the most obvious remark is that it is not the typical white-headed bird with short bill from Ceylon; but nothing can be ascertained about it being Dumont’s type of Cossyphus striatus.” The above at least determines that in fact Cossyphus stria- tus as defined by Dumont came from Bengal (where only one species occurs), and that the sole specimen remaining of those collected by Macé corresponds to the Jungle Babbler, a bird approximately the size of a Thrush (over-all length 250 milhi- meters approximately), upper plumage brown, rather reddish brown on the head and rump, slightly fulvous on the upper tail- coverts, the back with dark streaks and paler shaft-stripes; tail brown, tipped paler and cross-rayed; lores whitish with a narrow black line above them, breast and underparts, fulvous- April 21, 1958 Indian Birds. VII. 9 ashy to whitish, the sides tinged with brown. As lectotype, I select the specimen Paris Museum No. 8.614 labelled “type” from ‘‘Bengale” collected by Macé. The following forms of Jungle and White-headed Babbler should thus be listed in India and Ceylon: Turdoides striatus (Dumont): Bombay Babbler or Jungle Babbler. Pakistan, India and Nepal. Turdoides striatus sindianus (TYicehurst). Crateropus Terricolor sindianus Ticehurst, 1920, Bull. Brit. Orn. CL. 40:156. (Karachi, Sind.) Range.—West Pakistan south to India in Kutch, Rajasthan and Del- hi, intergrading south of this line into neighboring subspecies. Turdoides striatus orientalis (Jerdon). M.(alacocircus) orientalis Jerdon, 1847, Ill. Ind. Orn. text to pl. 19. (‘jungles of the Carnatic and more especially among those of the Eastern Ghauts,” hereby restricted to Horsleykonda, west of Nel- lore.) Range.—Intergrades with the preceding form in the little Rann of Kutch and Saurashtra, thence northeast to Agra, south and east to Narbada River, M.P., intergrades with striatus along the Jumna River in northwestern M.P. and southern U.P. and a line southeast to the Godavari Delta; Madras, Mysore. Turdoides striatus somervillei (Sykes). Timalia Somervillei Sykes, 1832, Proc. Zool. Soc. London:88. (“Ghauts” = Bombay Ghats). Range.—Bombay State from Surat Dangs south along the coast, in- tergrading with the following form in Goa. Turdoides striatus malabaricus (Jerdon). M.(alacocircus) malabaricus Jerdon, 1847, Ill. Ind. Orn. text to pl. 19. (“forests of Malabar and on the sides of the Neilgherries” = 'Tra- vancore vide Whistler 1935, J.B.N.H.S., 38:72.) Range.—Western Mysore and Goa where it intergrades with the pre- ceding form, south through western coastal Mysore and Kerala, in- tergrading with orientalis at Tenmalai near the Ariankivu gap. Turdoides striatus striatus (Dumont). Cossyphus striatus Dumont, 1823, Dict. Sci. Nat. (ed. Levrault), 29:268. (Bengale) M. (alacocercus) terricolor “Hodgson” = Blyth, 1844, Jour. Asiat. Soc. Bengal, 13:367. (Nepal.) Range.—Northern and eastern India from U.P., Jumna River east through the Nepal terai along the Himalayan foothills to Assam, as far as Dibrugarh, southeast in the drainage of the Ganges and Brahmaputra rivers in West Bengal and East Pakistan, to Orissa, - Andhra, and the Godavari delta. 10 Postilla Yale Peabody Museum No. 36 Turdoides striatus rufescens (Blyth). M. (alacocercus) rufescens Blyth 1847, Jour. Asiat. Soe. Bengal, 16:453. (Ceylon) Range.—Low country wet zone and southwestern hill zone of Ceylon. T'urdoides affinis (Jerdon): White-headed Babbler Peninsular India and Ceyon. Turdoides affinis affinis (Jerdon). M. (alacocireus) affinis Jerdon, 1847, Ill. Ind. Orn. text to pl. 19. (peninsula = Kanara dist., Mysore, restricted herewith.) M. (alacocireus) Elliotti Jerdon, 1847, ll. Ind. Orn, text to pl. 19. (Travancore) Turdoides polioplocamus Oberholser, 1920, Proc. Biol. Soc. Washing- ton, 33:84. New name tor Turdus griseus Gmelin, 1789 (Coroman- del Coast, India), nec Turdus griseus Boddaert, 1783. Range—From eastern Bombay (Chanda) and the Godavari River Valley (Ellore and Dumagudiam) south through Andhra, Madras, Mysore, and the drier parts of Kerala to Cape Cormorin, and on the west coast from Mangalore north to the Gaprabha River. Turdoides affinis taprobanus subsp. nov. Malacocircus striatus Swainson 1833, Zool. Mlus ser. 2, 3, pl. 127 and text. (Ceylon.) nec Cossyphus striatus Dumont 1823 = Turdeides striatus (Dumont). Range.—Ceylon. Remarks.—This form differs from the Peninsula population as pointed out by Whistler (1944, Spolia Zeylanica 23:131) having a much grayer wash on the head and body, and with reduced or absent subterminal blackish spots. In addition, Ceylon specimens lack the grayish-white heads, characteristic for the birds of southern India. There is a tendency for Ceylon birds to be larger than South In- dian birds; wing: 104.5-110 mm. vs. 98-104 mm. in the Y.P.M. series. Whistler and Kinnear (1936, Jour. Bombay Nat. Hist. Soc., 35:739) give measurements for Peninsula birds from 94.5-109 mm., so it may be that there is too much overlap to be significant. Type—Y.P.M. No. 20220, g ad. collected 13 September, 1950, by S. Dillon Ripley at Alawna, Ceylon. Certhia himalayana Vigors In the Bulletin of the British Ornithogists’ Club, (42, 1922, p- 140) Col. Meinertzhagen states that the type of Vigor’s Certhia himalayana “was undoubtedly collected in either Garh- wal or Kumaon,” and that the specimen from Pushut, Afghani- stan designated as “type” in Gadow’s handwriting in the Brit- ish Museum collection, probably done when he was writing the April 21, 1958 Indian Birds. VII. dol British Museum Catalogue, Volume 8, must be disregarded. Meinertzhagen then goes on to name the extreme northern and northwestern Himalayan population as Certhia himalayana limes, a name which is considered valid today. Two years later, Ticehurst and Whistler, (1924, [bis :468- 473) discussing the area in which Vigor’s birds were secured, stated that the principal collection came from the western Himalayas which “for our purpose must be considered as the area of Gahrwal (sic) and Kumaon, from Simla to the Nepalese border.” They specifically state that there were no birds from Nepal in the collection. It would thus seem that under the International Rules as amended at Paris and Copenhagen, 1948 and 1953, Col. Mein- ertzhagen’s action in stating that Vigor’s specimens must have come from Garhwal or Kumaon, can be interpreted as that of a first revisor, (see proposed Draft of the Regles by Professor Bradley, 1957, Bull. Zool. Nomencl. 14 (1/6) :115-116, Article 17, “Localities of Origin”). However, it would appear in view of the vagueness of Col. Meinertzhagen’s ascription of locality and the fact that in 1830 when the collection came to Vigor’s hand the geographical area from Garhwal to Kumaon could include the entire sweep of hills from Simla to the Nepal border, exclusive of the independent Raja’s Hill States, that Messrs. Ticehurst’s and Whistler’s action of 1924 of narrow- ing the type locality to the Simla-Almora district is perfectly valid and proper. Such an action is envisaged in Paragraph (a) (3) of the Bradley Draft (op. cit.) ... “priority shall apply between two or more authors attempting to establish a designated typical locality for a species, but subsequent re- striction of a designated typical locality shall be allowable.” I thus feel that Ticehurst and Whistler’s action is correct in this instance, and that they have validly restricted the type lo- cality of Certhia himalayana Vigors to Simla-Almora district, as it is understood today in strict political terms. Vaurie (1957, Amer. Mus. Novit., No. 1855:18) discusses this species and points out that Meinertzhagen’s action ante- dates that of Ticehurst and Whistler, but does not attempt to re-restrict or designate a more specific type locality, apparently being unaware of the vagueness, especially in the usage of the 12 Postilla Yale Peabody Museum No. 35 1830’s of such a term as “Garhwal and Kumaon.” He also casts doubt on my naming (Proc. Biol. Soc. Washington 63:106) of a darker population, Certhia himalayana infima from western Nepal, from whence it had not with certainty been previously recorded. (‘There is a Hodgson specimen, an immature bird labelled “‘Nepal,” in the British Museum collec- tion, of uncertain attribution and useless for subspecific analy- sis.) It would seem appropriate in view of the existence of comparative material at neighboring institutions for Vaurie to have borrowed specimens in order to compare them critically as, so far as I know, the American Museum of Natural History collection lacks material of this species from Nepal or eastern Kumaon. However, Vaurie, accepting a vague type locality of “Garhwal and Kumaon,” merely states that with such a type locality, “the cline is not sufficiently steep to warrant in my opinion the nomenclatural separation of the population from western Nepal from that of neighboring Kumaon or Garhwal. I consider infima a synonym of nominate himalayana.” In this connection, it seems appropriate to record that Rand and Fleming (1957, “Birds from Nepal,” Fieldiana :Zoology, 41(1):121, working with actual specimens, state: “We have three Mussoorie birds . . . for comparison. The Nepal birds are darker on the flanks, back and tail, supporting Ripley’s description for C. h. infirma” | sic: infima|. Mussoorie is cer- tainly within the vague general locality of Garhwal and Ku- maon, and their specimens should serve as a comparative series within the range as defined by any of these authors. I am in consequence listing infima as a valid subspecies in my Indian hand-list, and accepting the Ticehurst and Whistler restricted type locality for the nominate subspecies. te bass riage: Wostilla YALE PEABODY MUSEUM oF NaTurAL History No. 36 June 27, 1958 New Haven, Conn. ON] Dab DOUBTEUL VALIDITY OF TACHYPLEUS HOEVENI POCOCK, AN INDONESIAN HORSESHOE CRAB (XIPHOSURA)"* Tatsot H. WarerMan Department of Zoology, Yale University There are four undoubtedly valid species of living xipho- surans. One of these, Limulus polyphemus (1.), is from the western North Atlantic. The other three have different distri- butions in the western Pacific and the Bay of Bengal. They comprise Carcinoscorpius rotundicauda (Uatreille), Tiachy- pleus tridentatus Leach and T'achypleus gigas (Miller). All of these are well represented and recognized in collections. A third species of T'achypleus, T’.. hoeveni, was named by Pocock (1902) for certain figures in van der Hoeven’s monograph on the Xiphosura (1838; Plate I, figs. 2, 10; Plate II, fig. 14). A key to recent species of Xiphosura is presented as an Ap- pendix and in figs. 9-11. | Respectfully dedicated to my colleague and friend Dr. Alexander Petrunkevitch. 2 These studies were uided by a contract between the Office of Naval Research, Department of the Navy, and Yale University, NR163-091, and by a grant from the Pacific Science Board of the National Research Council. z ‘ae ITY Ul NIVERS oe Postilla Yale Peabody Museum No. 36 According to its definition T’achypleus hoeveni resembles T’. gigas in all known respects except that the median terminal segments of the genital operculum in hoeveni are separate and overlap asymmetrically, instead of being united, symmetrical and contiguous as in gigas. Apparently Pocock himself never saw a specimen of his species, and none have since been reported either in the field or in museums. It is of interest, therefore, to inquire whether or not T'achypleus hoeveni is a valid species, closely related to, but distinct from 7’, gigas. Doubt of such validity has already been raised on rather general grounds by Gravier (1929) in his review of specimens of Niphosura in the Paris Muséum National d’Histoire Naturelle. Present at- tention to the problem was stimulated by the author’s interest in speciation and distribution of horseshoe crabs particularly with relation to their eves and visual physiology (Waterman. 1951; 1953a,b; 1954a,b,c; 1955; Waterman and Enami, 1953; Waterman and Wiersma, 1954). In an attempt to settle this moot taxonomic point, a large number of T'achypleus gigas have been studied to determine whether any individuals could be found attributable to 7. hoevent as defined by Pocock. Specimens examined include the author’s own material collected on Singapore Island in 1952; the collections of the Raffles Museum, Singapore; the Zoolo- gisch Museum, Amsterdam; the Rijksmuseum van Natuurlijke Historie, Leiden: the Universitets Zoologisk Museum, Copen- hagen; and the Museum of Comparative Zoology, Harvard University’. Relevant material was found in only three of ‘Tf should be mentioned that visits or inquiries addressed to a number of other museums revealed that no specimens either of Tachypleus gigas or Tachypleus hoeveni were present: U.S. National Museum, Washington, D.C.; Peabody Museum, Yale University; American Museum of Natural History, New York; Museum Zoologicum Bogoriense, Bogor (formerly Buitenzorg), Indonesia; Institut Océanographique de Nhatrang, Vietnam; Natural History Museum, Manila, Philippine Republic; the National Mu- seum, Sydney and National Museum of Victoria, Melbourne, Australia. In addition, review of the present collections in the British Museum (Nat- ural History) and the Muséum National d’Histoire Naturelle in Paris shows that no relevant material has been acquired since the classic papers of Pocock (1902) and Gravier (1929) which are mainstays of our present knowledge of horseshoe crab taxonomy and distribution. The author is much indebted to the Directors and staffs of these institutions. as well as June 27, 1958 Tachypleus hoeveni Pocock AB ne these; the museums at Amsterdam, Copenhagen and Leiden. Of greatest interest was finding that the Leiden museum has what is undoubtedly the original specimen of a male oper- culum from which van der Hoeven’s Plate II, fig. 14 was drawn (reproduced fig. 1). Because this particular figure was a crucial one cited by Pocock in establishing Tachypleus hoeveni, the specimen in question is presumptive type material for that species. It is hereby designated the lectotype (figs. 3,4). This operculum, which is in alcohol, had been Number 1038 in the Collection of the Zoological Laboratory, University of Leiden. It was transferred to the museum when that collection was dispersed. Other records show that van der Hoeven’s anatom- ical collections were given by him to the Zoological Laboratory at Leiden so that this evidence fits well with the identification of the specimen. When examined by the author in August 1953, the jar con- taining this interesting operculum held among several other labels, a note in the handwriting of Dr. H. D. Bléte, Assistant Director of the Leiden Museum. Translated from the Dutch this reads, “This preparation most probably is the original of Plate II, fig. 14 in J. van der Hoeven’s Recherches sur |’His- toire Naturelle, etc. des Lumules, Leiden 1838. See Pocock, Annee Mac. Natwltist. Vil, Ser, 1X; 1902, p. 264.” Careful comparison of this specimen with the figure in ques- tion leaves little doubt of the correctness of Dr. Bléte’s attri- bution (figs. 1,3). Note particularly that even the edges where the appendage was cut from the rest of the body and the slight scar on the lateral margin at the right agree almost perfectly. Only in the exact proportions of the appendage and degree of overlap between the terminal medial elements does the speci- men differ from the figure. But as the specimen itself had been stapled to a slab of soapstone, some changes in its shape and in the position of its parts would not be surprising in the course of the 115 years since it was drawn. The maximum all those mentioned in the text, for their generous cooperation and assist- ance in examining available material. Thanks are also due to Dr. Fenner A. Chace, Jr. of the U.S. National Museum and Dr. Charles L. Remington of the Yale University Zoology Department for their helpful suggestions concerning the manuscript. 4. Postilla Yale Peabody Museum No. 36 width of this Leiden specimen was 70.5 millimeters while that of the drawing, stated to be natural size, is about 70 milli- meters which provides another important point of agreement. It should be pointed out that Pocock’s interpretation of van der Hoeven’s figures does not agree with the original speci- men in certain respects not visible in the plates. The British zoologist believed that the opercular elements which overlap in T’. hoeveni were separated medially, unlike those of T. gigas which are normally united (fig. 8). However, in the Leiden Museum operculum they are, in fact, not separate in the mid- line. The edges appear to overlap, not because they are free medially, but merely because the appendage is pleated with a double fold in that region. Whether this fold was present in the living animal is not obvious from its appearance. Unfortunately, this single xiphosuran fragment is all that is known to remain of van der Hoeven’s original material. Con- sequently, it is not possible to settle the point directly whether this unique opercular detail is really a valid species character or merely an individual idiosyncrasy. As Pocock correctly pointed out, however, three of the four drawings of the oper- culum labelled Limulus moluccanus Latreille (= Tachypleus gigas) in van der Hoeven’s monograph show identical over- lapping elements in both adult male and adult female specimens (reproduced figs. 1,2,5,6). At its face value this does make the peculiarity seem taxonomically significant. Van der Hoeven’s failure to comment in any way on this anomaly is accordingly the more exasperating. On the other hand, close reading of van der Hoeven’s text reveals (1838, p. 2) that only two spirit specimens of T'. gigas were available to him for the anatomical work reported. Yet measurements of three specimens are given (1838, p. 10). It is not expheit whether two of these are the anatomical subjects, which of them may be merely dried examples, or what the total number of individuals studied may have been. If two complete specimens only were available for the four drawings concerned, it is possible that the same operculum may have been used as a model for the figures of more than one animal. The fact that the operculum, still extant, was already detached from the June 27, 1958 Tachypleus hoevent Pocock 5 whole specimen, when drawn for van der Hoeven’s Plate IT, fig. 14, adds some likelihood to this possibility. Further circumstantial evidence is provided by the mirror image similarity of the opercula in figs. 2 and 10, Plate I. This is carried even to the peculiar extra plates present in the proximal lateral margins of the overlapping sections in both drawings (reproduced figs. 5,6). The collaboration of another artist in drafting Plate I in contrast to the anatomical Plates IT and IT drawn entirely by van der Hoeven himself may also lend credibility to this notion. The extra opercular plates men- tioned above, which are not symmetrical, could also be taken as evidence that this operculum is abnormal; in general all nor- mal xiphosuran external anatomical features are bilaterally symmetrical. One must, therefore, entertain the possibility of Tachypleus hoeveni being merely an abnormal T’. gigas. Some independent but congruent evidence has been found that abnormalities of the median distal plates of the genital operculum are not rare in this animal. In the Amsterdam Museum there is a Tachypleus gigas (No. Xi 1001), collected in East Sumatra by J. C. van der Meer Mohr, which also has overlapping elements at the margin of the operculum (fig. 7), although the specimen is otherwise normal. The overlapping plates are not, however, the median distal elements as in the Leiden specimen but are the lateral distal elements. These are so lobed along their medial margins that the edges lie over one another for a short distance. None of the 8 other specimens of 7’, gigas in the Amsterdam collection show any similar opercular anomalies. Another JT. gigas with a deformed operculum is present among the 11 specimens of this species in the Copenhagen Museum, a male from Penang Island on the west coast of Malaya (collected by the research ship “Galathea”’). As in the Amsterdam specimen, there is a small medial overlap in the lateral distal opercular plates in this animal. In this case, though, the reason for the anomaly is more obvious, since there is a considerable healed wound in the edge of the operculum on the left side, and both prosoma and opisthosoma show dis- torted or missing parts. Further evidence for a widespread occurrence of structural abnormalities in T'achypleus gigas 1s 6 Postilla Yale Peabody Museum No. 36 given by van der Meer Mohr (19834), but no opercular devia- tions are mentioned specifically. In J’. tridentatus, however, Smedley (1931) reported that various specimens differ con- siderably in the degree of separation of the internal opercular branches at their tips. From the specimens examined and here discussed one would conclude that the peculiar fold and other unique details of the van der Hoeven T'achypleus operculum in Leiden are but minor teratological variants of Tachypleus gigas. On the other hand the geographical origin of van der Hoeven’s material in the Moluccas is an element that in all fairness should weigh on the side of the validity of T’, hoeveni. Few, if any, specimens of Xiphosura from these islands are known in museum collections. Consequently, study of Tiachy- pleus from Ceram, Halmahera, and adjacent islands might indeed show that JT’. hoevenit exists as a taxonomically distinct form in these regions. Yet such a circumscript distribution would be unique for a xiphosuran species, since the four definitely known recent forms have wide ranges. Limulus polyphemus (u.) occurs from Nova Scotia to Yucatan on the east coast of North America, a large spread in latitude, covering a shore line several thousand miles long. Tachypleus tridentatus Leach occurs south from the Inland Sea of Japan, along the China coast, in the western islands of the Philippine Republic, in Hainan and at least as far south as Nhatrang in south central Vietnam (Flower, 1901; Smedley, 1929, 1931; Shoji, 1932; Asano, 1942; Water- man, 1958a)*. Tiachypleus gigas (Miiller) overlaps the latter species by occurring in northern Vietnam (Prof. C. Boisson, University of Hanoi, personal communication) and North Bor- neo, extends west to the Orissa coast on the Bay of Bengal and east as far as Torres Strait (Pocock, 1902; Annandale, 1909). Carcinoscorpius rotundicauda (Latreille) has been reported in the southern Philippines, Indonesia, Malaya, the Gulf of Siam, and the Bay of Bengal. ! If the specimens of Limulus longispinus (sie) (= Tachypleus triden- fatus) reported (in lit.) in the Australian Museum, Sydney, are correctly identified, this species occasionally reaches as far as the west coast of Malaya. yr June 27, 1958 Tachypleus hoeveni Pocock 7 On the basis of the evidence at hand one must conclude that Tachypleus hoeveni is a dubious species at best and most likely was named for an abnormal operculum of T'achypleus gigas. But since the original material does not permit a definitive solution of the problem, it is to be hoped that the interest and opportunity of studying the xiphosurans of the Moluccas will develop in the near future to resolve the dilemma more de- cisively. However, a recent attempt to do this failed. At the author’s request, Dr. Dillon Ripley of the Peabody Museum at Yale University, who spent three months of 1954 in the Moluccas collecting specimens of various animals, particularly birds, tried to obtain horseshoe crabs from these islands. According to information he most kindly gathered, than mimi or imi imi, as these animals are called in Indonesia, were known to fishermen in the Moluccas but were said not to occur there. According to these sources the nearest place where such crabs were ordinarily caught was Menado. This is a town on the northern arm of Celebes more than 200 miles westward across the Molucca Passage from Ternate, Tidore, Halmahera and other islands in the group. Not only were no specimens of T'. hoeveni to be obtained even in Menado, but no evidence for the occurrence of any species of Niphosura in the Moluccas themselves was found despite the fact that 7’. gigas and Car- cinoscorpius are well known from other parts of Indonesia. The reported complete absence of these forms in the area concerned is the more baffling since the Moluccas are the first place of occurrence cited for the xiphosurans in the Kast Indies. L’Ecluse in 1605 figured specimens of Cancer moluc- canus, a horseshoe crab sent to Holland reputedly from the Moluccas. Rumphius (1705) in his famous book about the natural history of these islands illustrates a horseshoe crab under the name of Cancer perversus. This animal, he states, was well known by him to occur in the Moluccas (he was work- ing in Amboina in the southern part of the archipelago) and he also had received a specimen of it from Menado. SUMMARY 1. In reviewing material suitable for determining the valid- ity of Tiachypleus hoevenit Pocock, an original fragmentary 8 Postilla Yale Peabody Museum No. 36 specimen apparently used by van der Hoeven for one of the figures cited as the type by Pocock was re-examined in the Rijksmuseum van Natuurlijke Historie in Leiden. This speci- men itself, here designated the lectotype, and Pocock’s mono- graph do not alone permit a decisive conclusion whether the material represents an anomalous T'achypleus gigas or another valid species. 2. Evidence has been obtained from single specimens in the Amsterdam and Copenhagen zoological museums that oper- cular anomalies, comparable to but distinct from the one for which Tachypleus hoeveni was erected, are not rare in T'achy- pleus gigas. . 3. The fact that van der Hoeven’s material came from the Moluceas, a region from which few, if any, xiphosuran speci- mens have since been studied, leaves open the possibility of a taxonomically significant geographic variation in this area. However, a recent search failed to find any Xiphosura in the Moluceas. 4. It is concluded that T'achypleus hoeveni is probably a synonym of 7’, gigas, but this synonymy can only receive its decisive test when substantial series of Indonesian xiphosurans have been studied. 5. A key to recent species of Xiphosura is presented as an Appendix. June 27, 1958 Tachypleus hoeveni Pocock 9 REFERENCES Crrep Annandale, N. 1909. The habits of king crabs. Rec. Ind. Mus., 3:294-295. Asano, U. 1942. On the life history of Tachypleus tridentatus. (In Jap- anese) Botany and Zoology, Pure and Applied. 10:120-124. I/feluse, C. de 1605. Exoticorum Libri Decem. (Antverpae) ex officina Plantiniana Raphalengii, pp. 1-378. Flower, S. S. 1901. Notes on the millipedes, centipedes, scorpions, etc. of the Malay Peninsula and Siam. J. Straits Branch Roy. Asiatic Soe., 3631-48. Gravier, C. 1929. Révision de la collection des Limules du Muséum National d’Histoire Naturelle. Bull. Mus. Nat. Hist. Nat. Paris, Ser. 2, 1:313- Bole van der Hoeven, J. 1838. Recherches sur l Histoire Naturelle et U Anatomie des Limules. Leyden, Luchtmans, pp. 1-48. van der Meer Mohr, J. C. 1934. Sur quelques malformations chez la limule, Tachypleus gigas. Miscell. Zool. Sumatrana, 87 :1-3. Pocock, R. I. 1902. The taxonomy of recent species of Limulus. Ann. Mag Nat. Iist., 9:256-266. Rumphius, G. E. 1705. D’Amboinische Rariteitkhamer. Amsterdam, Halma, pp. 1-340. Shoji, K. 1932. Morphology and biology of Xiphosura. (In Japanese) Fukuoka Nat. Hist. J. 7:28-52. Smedley, N. 1929. Malaysian king crabs. Bull. Raffles Mus., 2:73-78. ————. 1931. Notes on king crabs (Xiphosura). Bull. Raffles Mus., 5371-74. Waterman, TT. H. 1951. Polarized light navigation by arthropods. Trans. N.Y. Acad. Sci., 14:11-14. . 1953a. Xiphosura from Xuong-Ha. Amer. Scientist, 41/:292-302. ————.. 1953b. Action potentials from an arthropod ocellus: the median eye of Limulus. Proc. Nat. Acad. Sci., 39:687-694. . 1954a. Directional sensitivity of single ommatidia in the com- pound eye of Limulus. Proc. Nat. Acad, Sci., 40:252-257. . 1954b. Polarized light and angle of stimulus incidence in the compound eye of Limulus. Proc. Nat. Acad. Sei., 40:258-262. 1954c. Relative growth and the compound eye in Xiphosura. J. Morph., 95:125-158. 10 Postilla Yale Peabody Museum No. 36 Waterman, T. H. 1955. Polarized light and animal navigation. Sci. Amer. 193: 88-94. Waterman, T. H. and M. Enami. 1953. Neurosecretion in the lateral rudi- mentary eye of Tachypleus, a xiphosuran. (Abstr.) Convegno sulla Neurosecrezione, Pubbl. Staz. Zool. Napoli, 24: Suppl., 81-82. Waterman, T. H. and C. A. G. Wiersma. 1954. The functional relation be- tween retinal cells and optic nerve in Limulus. J. Exp. Zool., 126:59-86. June 27, 1958 Tachypleus hoeveni Pocock i APPENDIX KEY TO RECENT XIPHOSURA 1. Tail spine oval or circular in cross section with no marked dorsal ridge. Adult sexual characteristics: Male, first two pairs of walking legs modified as chelate claspers (fig. bILa\)) e* e a eeeehe gee Carcinoscorpius rotundicauda (Latreille) Tail spine triangular in cross section with marked dorsal ridge along most of its length ...... FR Re aN Pt ne ee ee 2. Genital operculum (first opisthosomal appendage) with three endopodite segments on each side (including tab-like ter- minal element) (fig. 1OA). Adult sexual characteristics: Male, first pair of walking legs modified as hemichelate claspers (fig. GIVES 8) iat 2 tia +f: Limulus polyphemus Linnaeus Genital operculum with two endopodite segments on each side (Giitcoammlt() (09) errean Fak puree reat it ag Se els Se 2a cea Ree Re eer 3. 3. Distal endopodite segments of genital operculum over- lapping in midline (according to Pocock not united in midline but are so in lectotype; see text) (fig. 10B). Adult sexual char- acteristics: Male, presumably first two pairs of walking legs modified as hemichelate claspers. Female, presumably first three movable lateral opisthosomal spines long, other three short stubs (fig. 9B) .......Tachypleus hoevent Pocock Distal endopodite segments of genital operculum united and not overlapping along midline (fig. 10C). Adult sexual charac- teristics: Male, first two pairs of walking legs modified as hemi- chelate claspers; anterior margin of prosoma scalloped (fig. 9A). Female, first three movable lateral opisthosomal spines lone mother three short stubs (ip 9B) yo 20 ene a 4. 4. Usually three spines on posterior dorsal surface of opis- thosoma over base of tail spine (fig. 9B); lateral eyes black, no pseudopupil visible; lateral eye length not more than 5-6 per cent of prosoma length along the midline... ... . . EMD ey na As win A LNT oh. Qe D9? Tachypleus tridentatus Leach One spine only on posterior dorsal surface of opisthosoma over base of tail spine; lateral eye brownish, pseudopupil visible; lateral eye length more than 6.5 per cent of opisthosoma length alone thesmuidline’ 2.4205. .2.25.. Tachypleus gigas (Miller) 12 Postilla Yale Peabody Museum No. 36 Prater I Fig. 1. Posterior surface of the genital operculum of a male Moluccan xiphosuran as figured by van der Hoeven (1838, Plate II, fig. 14). The original author does not mention the anomalous double fold in the midline of the central terminal segments and referred this drawing to Limulus moluccanus (= Tachypleus gigas). Pocock (1902) considered this opercular fold, which may also be seen in figs. 3, 4, 5 and 6, grounds for establishing a third species of Tachypleus, T. hoeveni. Maximum lateral extent (width) of this operculum 70 mm. Fig. 2. Posterior surface of the genital operculum of a female Molucean xiphosuran as figured by van der Hoeven (1838, Plate II, fig. 15). This also was assigned by the original author to Limulus moluccanus (= Tachy- pleus gigas). Note that the median margins of the inner and outer terminal segments are neither folded nor overlapping and are normal for the species like those shown in fig. 8. Maximum lateral extent of this operculum 85 mm. Fig. 3. Posterior surface of the genital operculum of a male xiphosuran which is most likely the original specimen from which van der Hoeven’s (1838) Plate II, fig. 14 (fig. 1, above) was drawn. Except for the length width ratio, the two agree closely. Photograph courtesy of the Rijksmu- seum van Natuurlijke Historie, Leiden. Maximum lateral extent of this specimen 70.5 mm. Fig. 4. Anterior surface of the same specimen as shown in Fig. 3. Since this peculiar operculum apparently was the basis for Pocock’s species Tachypleus hoeveni, it is designated as the lectotype pending final clarifica- tion of its validity. The specimen is in the Rijksmuseum van Natuurlijke Historie, Leiden, through whose courtesy the photograph is reproduced. ~o r . 3 © June 27, 1958 Tachypleus hoeveni Pocock 13 Puiate I 14 Postilla Yale Peabody Museum No. 86 Prats II Fig. 5. Part of van der Hoeven’s (1838) Plate I, fig. 2 showing the ventral surface of a mature female Tachypleus (prosoma length 100 mm.). This drawing was identified by the original author as Limulus moluccanus (= T. gigas) but on the basis of the genital operculum with the over- lapping median elements was considered by Pocock to be 7. hoeveni. Fig. 6. Plate I, fig. 10 of van der Hoeven (1838) showing the ventral surface of a male Tachypleus opisthosoma (prosoma length 82 mm.). As in fig. 5 above the overlapping distal moiety of the operculum induced Pocock to include this in 7. hoeveni although van der Hoeven had referred it to Limulus moluccanus (= T. gigas). Pocock used this figure and that shown in fig. 5 above as evidence that the peculiar folds in the operculum shown here in figs. 1, 3, and 5 were not just an individual idiosynerasy since the same thing is shown in drawings of both sexes. Note however, that the opercula in the drawings reproduced in figs. 5 and 6 are mirror images of each other and nearly identical which suggests that one operculum was used as a model in drawing two individuals. Fig. 7. Ventral view of part of the opisthosoma of a mature male Tachypleus gigas (prosoma length 98 mm.) with an anomalous genital operculum showing some overlap of median elements. Note, however, that here the outer terminal segments rather than the central ones as in the Leiden specimen (fig. 3) form the overlapping pair. Also observe that the median edges of these segments are free, not fused, and just folded over as in the other case. Specimen Xi 1001 in the Zoologisch Museum, Am- sterdam, through whose courtesy the photograph is reproduced. Fig. 8. Ventral view of part of the opisthosoma of a mature Tachypleus gigas (prosoma length 95 mm.) with a normal genital operculum showing the smooth median fusion of the central distal segments with no folding and the free edges and absence of overlap in the lateral distal segments. Photograph courtesy of the Zoologisch Museum, Amsterdam. Tachypleus hoevenit Pocock June 27, 1958 16 Postilla Yale Peabody Museum No. 36 Puate III Fig. 9. Adult sexual characteristics in Tachypleus tridentatus. A. Male prosoma showing scalloped anterior margin, dorsal view. X 0.3. B. Female opisthosoma showing lateral movable spines (a@-f) and species character- istic posterior median ones (g, h, i), dorsal view. X 0.3. Fig. 10. Genital opercula showing species characteristic endopodite seg- ments (a, b, ¢), anterior view. A. Limulus polyphemus with three of these segments. X 0.5. B. T. hoeveni with overlapping segment (b) (after van der Hoeven, 1838). X 0.5. C. T. gigas with two endopodite segments. X 0.4. Fig. 11. First claspers of adult male, anterior view of left appendage. A. Carcinoscorpius rotundicauda, chelate. X 1.5. B. L. polyphemus, hemi- chelate, X 0.7. C. T. tridentatus, hemichelate. X 0.7. (Figs. 9-11 drawn by Shirley P. Glaser § ; : ‘om WL ; = =} < va — ‘Cy aX oP i ay CY MUS. COMP. Z00L oslt lla LIBRARY YALE PEABODY MUSEUM. _|MAY 251960 ' HARVARD _ UNIVERSITY Number i 37 September 15, 1958 New Haven, Conn. or NaTuRAL History A NOTE ON THE FIRETHROAT AND THE BLACKTHROATED ROBIN S. Ditton RIpeLey While on a visit to theU.S.S.R. recently, I had the oppor- tunity of examining two specimens of the small chat, the Black- throated Robin, collected by Berezowsky and Bianchi and described by them in 1891 as “‘Larvivora”’ obscura, Both are adult males in fully adult plumage and are in the Zoological Museum in Leningrad. Also in the Leningrad collection is an adult male specimen of “Calliope” pectardens David. Due to the kindness of Dr. A. Ivanov, I was able to examine these specimens closely. Later in London I examined the series of pectardens and the single male obscura which have already been reported on by Goodwin (1956, Bull. Brit. Orn. Cl. 76: 74-75). Mr. H. G. Deignan has also kindly supplied me with information on the fine series of pectardens in the U. S. National Museum. Goodwin and Vaurie (1956, Bull. Brit. Orn. Cl. 76: 141- 143), have published their comments on these two species, bringing forward the opinion that both were color phases of a single species. In connection with studies on the subfamily of the thrushes for the Peters’ “Checklist,” I was anxious to determine this matter to my own satisfaction. The principal problem as to the identity or discreteness of these two populations is lack of specimens showing stages of 2 Postilla Yale Peabody Museum No. 36 plumage. In the British Museum and the U. S. National Mu- seum collections there are fine series of pectardens from Yun- nan, southeast Sikang and southeast Tibet. A single male adult pectardens, perhaps a post-breeding season bird has been taken in southwest Shensi. There are many immature males, two presumed femal ~ 4 - . APY , . re i 7 = ~% y + - : . m =- Ns I _— Pe . » i ~ 4a ~~. , -_ © 2 fg ay - en 4 * «re “— ie © id © “= we - 8, 7 + grip terme jos 4 S 2) mi. }: Aediar ron ih. . ; e . Fr © 2, CALI, a0 B*. rp a) Hees + 57) : ¢ ' 1 S Wet bd . > Ain, a > 7 * ¥ . 4 | 4 fire . \ r = Py - . ks - & 6 ‘ e * ‘ a ad iff 24 Tide wT. 7 ate a Lv ae . ) ° 4 a” ' icv¥% = * sc) a 3 i rm _ % * ha cad {Tonle we i a 5 ; . * ’ = 2 » Jet ‘ ry - a = 6 $ . oa © ? a | ° > Jaze tem §: np 4 if Lecr'F ' ey pet ° “ 7 - 7 id " : i. =k Slats mr? et | i Pr BFTA ARMED: 2 . yi ; \ é , N 7 7 . ORAL ar lakgy leas tee yt ge eee tei eq a) ia _ ° eg) Te as UOMT. ZUUL LIBRARY MAY 251959 yittlla pial YALE PEABODY MUSEUM or NATuRAL History Number 38 April 20, 1959 New Haven, Conn. COMMENTS ON BIRDS FROM THE WESTERN PAPUAN ISLANDS S. Ditton Rievtey 1. Birds from Kofiau Island Opportunities for visiting Kofiau Island (often called Kot- fiao, Kavijave, Kavijaaw, or Poppa in the literature) are few and far between. The island lies nearly ninety miles west of Sorong, regional capital of western Netherlands New Guinea, exposed to strong swells in the monsoon seasons. No boat an- chorage exists and the small population of Besarese fishermen lives primarily on a few offshore rocky islets. Odoardo Beceari visited Kofiau in July, 1875, on a schooner from Ternate, intending to spend several days (1875). His visit was cut short, however, by illness, and he spent only thirty hours there. Fortunately he was able to collect a total of 40 specimens during that time including topotypes of Tanysiptera ellioti and Rhipidura vidua. These forms had been taken in 1867 by David Hokum, an assistant of Mr. Hoedt a professional supplier of birds in Ambon. In 1875 also Bruijn’s collectors from Ternate visited Kofiau, and except for an undated visit by Bernstein, this seems to have been the last ornithological visit to the Island. During 1954 while studying birds in the Moluccas and West- ern Papuan Islands on a field trip,’ my wife and I attempted 'This field work was supported by research fellowship grants from the Guggenheim Foundation and the National Science Foundation as well as funds from Yale and the Vose Fund of the Explorer’s Club of New York. 2 Postilla Yale Peabody Museum No. 38 to visit Kofiau. Neither patrol vessels nor commercial schoon- ers were available, however, during our stay in New Guinea and an attempt to secure the services of an oil company flying boat also failed. Somewhat later my assistant, Jusup Khakiaj, managed to visit Kofiau in 1955 in a seagoing canoe accom- panied by an Arafura bird hunter from Misool. He spent fifteen days from the 25 April to 9 May on the Island and was able to clamber about the rocky foreshore and climb a short way into the heavily forested interior. Kofiau Island is about fifteen miles long, running in an east- west direction. It is heavily wooded, and the present settlements are essentially on the offlying islands such as Djailolo and Deer which he just north of the mainland of Kofiau separated from it by a narrow sheltered channel. Kofiau lies outside of the 200 meter bank which marks the Sahul Shelf and includes such islands as Salawati, just off the New Guinea mainland, and Misool, some thirty miles south south-east of Kofiau. The island has several hills, one nearly a thousand feet tall, named Mata or Boemfoar. The Boo Islets which he about ten miles west of Kofiau in- clude one islet Boo Ketjil an alternative name of which is Popa. This name has been applied to Kofiau in the literature. Beecari in his letter to Salvadori (1875, tom. cit.:707) speaks of “Poppa” as being a misnomer for Kofiau, which he spelt Kof- fiao. David Hokum in 1867 called the island Kavijaaw. Jusup Khakiaj’s collection while small, is of interest, as it appears to be the first made in perhaps eighty years. I am grateful to the authorities of the American Museum of Natural History for permission to examine specimens in their WELENY care. Of the thirty-one known species and subspecies from Kofiau, listed in the following pages it is interesting to note that they fall into these several categories. Species of unknown affinities: one, Ducula species (seen but not collected). Migrants: three, Pluvialis dominica fulcus, Merops ornatus, Halcyon sancta sancta. Uo. LENE. LUUL LIBRAR MAY 2 5 1959 > HARVARD UNIVERSITY April 20, 1959 Birds from Papuan Islands This leaves twenty-seven forms which may be characterk as follows: 1) Forms common to Moluccas and New Guinea; seven (= 27%). Butorides striatus papuensis Pandion haliaetus melvillensis Megapodius freycinet freycinet Chalcophaps indica indica Caloenas nicobarica nicobarica Pachycephala phaionotum Nectarinia jugularis frenata 2) Forms representing New Guinea subspecies (includes Kai and Aru Islands) ; fourteen (= 51%). Ptilinopus rivoli prasinorrhous Ptilinopus viridis pectoralis Macropygia amboinensis doreya Opopsitta diopthalma diopthalma Micropsitta keiensis chlorovantha Geoffroyus geoffroyt pucherani Cacomantis variolosus infaustus Alcyone pusilla pusilla Pitta sordida nova-guineae Coracina tenuirostre miilleri Gerygone magnirostris occasa Monarcha alecto chalybeoce phalus Monarcha guttula Philemon novaeguincae novaeguineae 3) Forms representing Moluccan subspecies; two (= 7%). Coracina papuensis melanolora Dicrurus hottentottus atrocaeruleus 4) Forms intermediate between species of the Moluccas and New Guinea; four (= 15%). T'anysiptera (galatea) ellioti Rhipidura rufiventris vidua Monarcha julianae Nectarinia sericea mariae 4 Postilla Yale Peabody Museum No. 38 From the above it will be seen that while 51% of the Kofiau residents are overwhelmingly of close New Guinea affinity, almost one quarter or 22% represent forms either intermediate or more nearly Moluccan in their affinity, thus corresponding closely with the geographical position of the Island. That 15% of these represent endemisms is a remarkable example of the inherent speciation potential of such an island in such a geographic location. Annotated List of Birds from Kofiau In the following list, I have given the names of the collectors in brackets at the end of the discussion. 1) Butorides striatus papuensis Mayr 2, May 8, 1955, wing 178 mm., culmen 65 mm. This specimen is small compared to Mayr’s measurements (1940), but agrees with at least one specimen, although it was listed as possibly subadult, recorded by Van Bemmel (1948:397). (Khakiaj ) 2) Pandion haliaetus melvillensis Mathews Q April 30. (Khakiaj ) 3) Megapodius freycinet freycinet Gaimard 2? May 2 (Bruijn, Beccari, Khakiaj) 4) Pluvialis dominica fulva (Gmelin) An adult, unsexed, in breeding dress was taken in May. (Khakiaj). 5) Ptilinopus rivoli prasinorrhous Gray ( Beccari) 6) Ptilinopus viridis pectoralis (Wagler ) ( Becear1) 7) Ducula sp.? Jusup Khakiaj reported the presence of a large fruit pigeon on Kofiau. The birds were high up in forest trees, difficult to see, as always, and resisted his collecting efforts. He believes that the species represented is Ducula rufigaster. April 20, 1959 Birds from Papuan Islands 5 8) Macropygia amboinensis doreya Bonaparte é subadult, April 80 (Beccari, Khakia)). 9) Chalcophaps indica indica (Linnaeus ) (Beccari) 10) Caloenas nicobarica nicobarica (Linnaeus ) (Hoedt) 11) Opopsitta diophthalma diophthalma (Hombron and Jac- quinot) (Beccari) 12) Micropsitta keiensis chloroxantha Oberholser ( Beccar1) 13) Geoffroyus geoffroyi pucherani Souancé ?, April 80. Wing 164. (Hoedt, Khakiaj) 14) Cacomantis variolosus infaustus Cabanis and Heine 2, April 25. Wing 117, culmen 20. Iris grayish, bill dark brown, feet yellowish. (Bernstein, Khakiaj) 15) Aleyone pusilla pusilla (‘Temmuinck ) ( Beccar1) 16) Halcyon sancta sancta Vigors and Horstield 2, May 8. (Beccari, Khakiaj) 17) Tanysiptera (galatea) ellioti Sharpe This beautiful kingfisher had a decidedly international intro- duction to the world of natural history. Collected by Hoedt’s collector, presumably David Hokum in 1867, the type speci- men was acquired by Count 'Turati of Milan who sent it to Jules Verreaux in Paris for identification. There it was seen by Mr. Daniel Giraud Elliot of New York who advised that the specimen be sent to Dr. Sharpe in London who described it in 1869. Other specimens from Hoedt reached Leyden. Beccari collected the species for the Genoa Museum, one of the speci- mens of which came into the Rothschild collection and is now in New York. The series collected by Jusup Khakiaj includes six adults, all labelled females, taken May 1-5, and two young birds 6 Postilla Yale Peabody Museum No. 38 labelled males, taken May 1 and 2, in first winter or first basic plumage. The adults are alike in possessing a uniformly white rump and upper tail coverts, in this and the largely white tail bearing a certain resemblance to sabrina of the Moluccas. This population has been kept separate from that complex of populations of the Moluccas, Papuan Islands and parts of the New Guinea mainland, all now included in the species galatea, on the basis of having tail feathers which are not sharply spatulate at the tip. This is essentially true as dem- onstrated by this series. All the adults except one possess broad central tail feathers, narrowing somewhat near their Poeatette Satenerteca B CY) A Fig. | Figure 1. Three states of plumage and tail coloration of Tanysiptera (gala- tea) ellioti, B., contrasted with an adult of a more typical Tanysiptera galatea of the subspecies margarethae, A. April 20, 1959 Birds from Papuan Islands fF terminal ends and tapering slightly to a bluntly rounded and broad tip. In four cases the sub-terminal segment of the tail shaft where the feather shows signs of tapering is dull bluish black or blackish. In one of these birds a very narrow area of the vanes adjacent to the shaft is edged with blue. One adult has almost completely white tail feathers, only a trace of blackish shading appearing on the edge of the subterminal part of the shaft. (Fig. 1). A single adult shows marked polymorphism. While the rump and upper tail coverts are white, and while the tips of the tail feathers are broadly and bluntly rounded, the subterminal area of the tail, representing a third of the total length of the central feathers is distinctly narrowed and the vanes strongly washed with blue. The effect is close to that of 7’. g. sabrina. This specimen is important in demonstrating the persistence of an old ancestral allele in what is not a completely homo- geneous population. One is impressed that this is a species in statw nascendi or as Mayr has called it a form of ‘almost specific rank,” (1942). The young birds are interesting as Sharpe has noted (1892) in that the under parts are washed with “ochre” or rich brown- ish buff, with almost completely reduced marginal edgings of blackish so noticeable in other forms. These young birds have bright blue caps, scapulars and lesser and median wing coverts, and ultramarine upper back and primary coverts, the latter with reduced brownish edging. The feathers of the lower back and rump are largely pale brown with white centers and blackish margins, rather strikingly different from the rump of related forms. The tail feathers are blue above, very pale at the centers, and noticeably broader than young of other forms. In these two birds the breast feathers are very badly frayed, indicating the wear of grubbing for insects in muddy jungle undergrowth. Measurements: Wing 6 2 ad. 101.5-108 (105.5) mm. Mail 6"? ads l53 Gnoult?)-215 (19074). Culmen 6 2 ad. 37-40 wing-tail ratio 5 2 ad. 49, 53, 54, 57, 64% (Hokum, Beceari, Bruijn, Khakia)). 8 Postilla Yale Peabody Museum No. 38 18) Merops ornatus Latham 6 ad. May 6. (Khakiaj) 19) Pitta sordida novae-guineae Miiller and Schlegel gad. May 5. Wing 104. (Beccari, Khakiaj) 20) Coracina tenuirostre miillerti Salvadori (Beccari) 21) Coracina papuensis melanolora (Gray) 6, 2 ad. May 9, Wing ¢ 150.5, 2 149; culmen (from skull) ¢ 31, 2 32. (Beccari, Khakiaj) 22) Gerygone magnirostris occasa Ripley 6 ad. May 2. Type. As pointed out in the original description (1957) this form differs from its geographical neighboring forms, cobana, brun- neipectus and conspicillata from the neighboring islands of Waigeu, Batanta and Salawati; western New Guinea, and the Aru Islands by being much more richly yellow on the under- parts. In the color of the underparts it approaches rosseliana from the Louisiade Archipelago and in color of the upper- parts it is close to affinis from north New Guinea, an interest- ing example of pattern replacement in geographically related forms. (Khakiaj) 23) Rhipidura rufiventris vidua Salvadori and Turati 6 ad. May 5. A topotype of this subspecies, not collected for presumably eighty years. As Beccari points out (1875:707), David Ho- kum collected the original specimen for Hoedt who sent it to Turati. Wing 74.5, tail 75.5, culmen 16. This form differs markedly from gularis the adjacent popu- lation of the Western Papuan Island by being much smaller (wing ¢ 6 83-92), the gray breast band marked with white spots, lacking in gularis, but present in obiensis and kordensis, and by having the abdomen and belly plain white, not washed with pinkish buff as in gularis. (Hokum, Beccari, Khakiaj) 24) Monarcha alecto chalybeocephalus (Garnot) é ad. May 8. (Beccari, Khakiaj). April 20, 1959 Birds from Papuan Islands 9 25) Monarcha guttula (Garnot) (Beccari) 26) Monarcha julianae new species.* Type: 6 ad. (Y.P.M. no. 39235) collected April 26, 1955, by Jusup Khakiaj on Kofiau Island, Netherlands New Guinea. Diagnosis: from guttula this species which is known from a single adult male differs by being slightly larger, wing 81, tail 73.5, culmen 17; compared to a small series of guttula from Misool and Waigeu, ¢ 6, wing 76.5-79, tail 67.5-71, culmen 14-16, |Gyldenstolpe’s measurements (1955) are equiva- lent] and by the following differences in pattern and color: back bluish black rather than gray; wing coverts are bluish- black throughout, in guttula the inner wing coverts are grayish, the greater wing coverts are bluish black with pronounced white terminal spots; below the prominent black bib reduced to a narrow diamond-shaped throat patch, extending in a median point towards the upper breast, the white of the breast ex- tending on the sides to the area below the eyes. Like guttula the tail of this species is black above, and below the outer four pairs of tail feathers are tipped with white, the outermost broadly so, the white area representing about 40% of the length of the feather. This species is much more closely related to what I would prefer to call the lewcwrus superspecies and should be included in it, I believe. This superspecies consists of three additional populations as follows: A) Monarcha everetti Hartert. This small Monarch flycatcher is found only on Tanahdjampea, an island of the Saleyer group south of Celebes and north of Flores, between five hundred and eight hundred miles west of the locus of its nearest relatives in the Moluccan-Papuan re- gion. This species represents, as Rensch points out (1986) an incursion of papuan-australian origin into the lesser Sunda-Celebesian area, an area which is primarily of oriental affinity. I entirely agree with Mayr (1944) that “This species is named, by gracious permission, in honor of Her Majesty, the Queen of the Netherlands. 10 Postilla Yale Peabody Museum No. 38 this species has nothing to do with the widespread Monar- cha trivirgatus as Meise attempted to demonstrate (1929). As Mayr notes, this is an “instance of ill-advised applica- tion of the principle of geographical representation.” Simply because the widely distributed gray-backed scrub- inhabiting Spectacled Monarch happens to be absent from certain islands is no reason for including highly dis- tinctively-plumaged arboreal-type Monarchs in the same species. This species is smaller than its relatives, and differs from them in having a white rump, and by having a pro- nounced white patch on the inner margins of all but the outermost primary, and on all the secondaries making a poorly concealed white wing patch which must be ex- tremely noticeable in flight. In addition, the black throat patch is lke a large bib in shape, extending down onto the upper breast. The tail, which is rounded, has the four outer tail feathers tipped with white, the outermost white for half its length. The female is gray above with whitish lores, the upper tail coverts buffy white and the tail black with whitish tips to the outermost feathers. Below the breast is light ochraceous-buff paling into grayish on the throat and sides of flanks and into dull creamy white on the ab- domen. This female plumage is markedly different from the forms described hereafter. In proportions of tail length to wing length and bill size, this species seems similar to its relatives to the east. In shape, however, the tail is much more rounded, the outer tail feathers being only 76% as long as the central tail feathers. It is also notably smaller; wing 34 4 67.5-69 ; tail 67-70.5 ; culmen 15-16; 2 wing 58.5, tail 60, culmen 14.5. This form is represented as “A” in Figure 2. B) Monarcha leucurus leucurus Gray. This population occurs on the Kai Islands (also spelled Kei or Key) of extreme eastern Indonesia, south of the western end of New Guinea. With loricatus I believe it forms a species. Both are rather large Monarchs, blue black above with April 20, 1959 Birds from Papuan Islands 11 a varying shape of throat patch below which extends nar- rowly onto the upper breast. The outer three pairs of tail feathers are white, some brownish margins occurring on the penultimate and third inner feathers. The fourth pair of tail feathers has a black inner web for the basal one-third of its length. The female is dark bluish-gray on head and upper back, brownish gray on the lower back, upper tail coverts blackish gray, central tail feathers black; below center of throat gray, sides of throat and breast dark orange rufous, paling to warm brown on the flanks; center of abdomen white. Size medium; wing 4é ¢ 75-80; tail 74-77; culmen 17- 18; 2 wing 71, tail 74, culmen 17. This form is represented as “B” in Figure 2. C) Monarcha leucurus loricatus Wallace. This popu- lation is found on the large island of Buru just west of Ceram. Although Stresemann (1914) mentions this form as occurring from the coast to the higher tableland and not higher than 800 meters in altitude, Toxopeus in Sie- bers (1921-22) found it only in the mountains at 1200 meters. This is the largest of the forms, the male blue- black above; below with a black throat patch just reach- ing the upper breast, and with a small patch of bluish black on the sides of the breast just before the bend of Fig. 2 Figure 2. Monarcha everetti, A; Monarcha leucurus leucurus, B; Monarcha leucurus loricatus, C3; Monarcha julianae, D. 12 Postilla Yale Peabody Museum No. 38 the wing. The tail is only slightly rounded, the outer- most feathers being 83-86% of the length of the central tail feathers. The two pairs of outer tail feathers are white with some blackish along the base of the shaft, the third pair with a very narrow (2 mm.) black tip, and the fourth pair with a black tip some 10 mm. in width. The female is brown on the forehead, more grayish, “hair-brown” on the crown and nape, and russet on the back, wing coverts and rump. The tail feathers are brownish black above. Below except for some grayish on the chin and center of the upper throat, the female is warm vinaceous brown. The outer tail feathers are rich buffy brown instead of white as in the female of lewcurus. Size largest; wing 4¢ 6 86-91; tail 72.5-85.5; culmen 17-20; female wing 77.5, tail 75. This form is represented as “C” in Figure 2. fo) D) Monarcha julianae. The single male differs from everetti and leucurus by having a gray rather than bluish- black crown and nape, shading into the white of the neck behind the black auricular patch. Unlike everetti but lke leucurus, the rump is concolorous with the back and there are no white patches on the inner margins of the wing feathers. Below julianae has a small roughly diamond- shaped throat patch, the white of the throat extend- ing laterally forward to below the eyes. The outer tail feathers are tipped with white rather than largely white as in everetti or all white as in leucurus. The tail is some- what rounded, the outermost feathers approximating 85% of the length of the central tail feathers. This new species is represented as “D” in Figure 2. Unfortunately, the female of this new form is unknown. It would be interesting to know if the female is dimorphic as in the lewcurus superspecies, and if so if it is predom- inantly russet in tone as in lewcurus or grayish and isabel- line as in everetti. The fact that julianae can coexist on a small island the size of Kofiau along with the widespread Monarcha guttula (collected formerly by Beccari) is an example of April 20, 1959 Birds from Papuan Islands 1183 how little is understood of the ecology and niche rela- tionships of the Monarcha species. Monarcha guttula is stated by Mayr (1944, t.c.) to belong to the trivirgatus group, although it cannot be considered a member of a superspecies as it overlaps with trivirgatus in the Louisi- ade Islands. Undoubtedly a close analysis of the feeding habits and spatial relations of these species will reveal a great deal about the problem of coexistence and competi- tion. Monarcha trivirgatus in my experience is a bird of scrub, low bushes and substage vegetation, found from the coast up to 2500 feet altitude. Monarcha guttula, at least on Misool Island, was found in the substage and also high up in the lower storey of the canopy forest. Unfortu- nately, Jusup Khakiaj has not noted the position in the forest of the single male of jalianae which he secured. 27) Pachycephala phaionotum (Bonaparte) 22 ad. May 3, 5. (Khakiaj) 28) Dicrurus hottentotus atrocaerulus Gray 22 ad. April 25. Wing 2, 150, 163; bill (using Vaurie’s scale, 1949: 284) 25, 25 mm. These two specimens place the Kofiau birds with the large- billed population of Halmahera and Batjan Islands, rather than with the west New Guinea carbonarius where they had been assigned by previous authors including Vaurie (1949) who had not examined specimens. Thus the spangled drongo of Kofiau belongs to the Moluccan rather than the Papuan form. (Beccari, Khakiaj). 29) Nectarinia sericea mariae, new subspecies Type: ¢ ad. (Y.P.M. No. 39234), collected April 25, 1955, on Kofiau I., Netherlands New Guinea, by Jusup Khakiaj. Diagnosis: compared to cochrani (Stresemann and Palu- dan) of Waigeu and Misool Islands, this form has a pansy- violet rather than steel-blue with a purplish gloss, throat patch. This color is nearer that of typical sericea which, how- ever, is more bluish, merely shaded with aster purple (Ridgway, 1912). The cap color is far more greenish than in sericea or cochrani, approaching in this respect auriceps of the Moluccas 14 Postilla Yale Peabody Museum No. 38 although it is less yellow-green than in that form. In the same way the iridescent color of the wing coverts, rump, and upper tail coverts is more greenish-blue rather than steel blue with a purplish-greenish gloss. This is especially noticeable in the area of the lower back. The single female appears somewhat brighter in color on the yellow underparts, nearer typical sericea than either cochrani or auriceps. In size these birds also approach typical sericea: Wing Tail Culmen 26 59, 61 35, 37 18, 19 Q 54, 32.5 18 A series of cochranit measured by Stresemann and Paludan (1932) showed wing measurements of 54-58, 2 51, and in sericea 6 6 60-64, 2 51.5-53 mm. This Kofiau Island population represents an interesting example of discontinuous geographic variation of the type referred to so exhaustively by Mayr (tom. cit. 1942 :77-84). If the distinguishing characters of the iridescent colors of the male mariae are contrasted with adjacent populations running from left to right as one travels from west to east the follow- ing discontinuous clinal pattern emerges : auriceps, mariae, cochrani sericea, Moluceas Kofiau W. Papuan Is. New Guinea throat bluish pansy-violet steel-blue bluish shaded with color aster purple; des- cribed as “reddish lilac” by Gylden- stople (1955:376) cap golden- “chrysoprase- — bluish- bluish-green, color green green” green, “tyrolite-green” (Ridgway ) “tyrolite green (Ridgway ) rump steel yellowish bluish-green — bluish green with and wing blue blue green with a faint a faint yellowish coverts yellowish suffusion suffusion April 20, 1959 Birds from Papuan Islands il Or Named in honor of my wife, Mary Livingston Ripley. Range: Kofiau Island (Beccari, Khakiaj). 29) Nectarinia jugularis frenata (Miller ) 26 Apr. 24, 27. Wing, 54.5, 57.5. (Khakiaj) 30) Philemon novaeguineae novaeguineae (Miiller) Jusup Khakiaj failed to collect this noticeable bird for the same reasons that he missed securing the fruit pigeon. Both species tend to dwell in the upper heights of the trees. (Beccari). 2. New or noteworthy records from the Western Papuan Islands 1) Procellaria pacifica chlororhyncha Lesson A male of this form from Kabaré, Waigeu Island taken by my assistant, Jusup Khakiaj, on October 8, 1955, appears to be the second record for New Guinea vide Mayr (1941:5). Wing 267, tail 131, culmen (from external nares) 31. 2) Goura cristata minor Schlegel A pair of Crowned Pigeons from Misool, the female in the melanic plumage sometimes encountered in this form, the throat and belly blackish, divided by a narrow smokey-blue chest band, the upper surface of the tail largely black, are consider- ably smaller than birds from Waigeu. In addition, a male from Misool recorded by Mayr and de Schauensee (1939) and a pair of birds in the American Museum collection are similarly small, wing ¢ 6 327-385, 2 2 320, 324. Waigeu birds meas- ure: wing 6 6 350-3865, 2 2 318 (1), 333-353. It is possible that additional material might reveal the existence of a distinct subspecies on Misool, which in several other instances seems to have evoked the emergence of populations with smaller dimensions than on the mainland of New Guinea or neighbor- ing islands, 3) Cuculus saturatus saturatus Blyth A male from Waigeu I. taken Sept. 24, 1955 with a wing measurement of 187.5 appears to belong to this small sub- species. Presumably the immature recorded by de Schauensee (1940) with a wing measurement of only 172 represents satu- ratus rather than horsfieldi. 16 Postilla Yale Peabody Museum No. 38 4) Collocalia vanikorensis granti Mayr Three males of this form were taken on Misool, a new record for that island. Wing measurements 114 (2), 116. In size they seem slightly smaller than typical granti, but are similar in color to that form. 5) Coracina morio incertum (Meyer) A male of this form taken by me at Fafanlap, Misool I. on 27 November, 1954, is an extension of range of this species to that island. 6) Hupetes caerulescens caerulescens ‘Temminck. A male taken at Wasa, Misool I. by Jusup Khakiaj on February 6, 1955, substantiates the old record of Neumann’s type of “occidentalis” as having come from ‘“Waigama’ on Misool. Wasa is not far from Waigama, but in any case the form ranges all over the island as we saw it in dense forest at Tamulol nearer the south coast. 7) Pomatostomus isidori isidort (Lesson and Garnot) An unsexed adult taken in September, 1955, by Jusup Khakiaj appears to be a first record for Waigeu Island. This seems surprising in view of the work of Stein and Bergman. 8) Rhipidura threnothorax threnothorax Miller At Tamulol on Misool, I collected a male specimen of this Fantail on November 14, 1954, which is a new record for the island. It does not differ from mainland New Guinea speci- mens and weighed 19 grams. 3. Birds from Ajoe Isiand Ajoe is the largest of a group of coral islets about twenty- five miles north of Waigeu Island. It is a sandy island, about three miles by a mile and a half in area, rising to a height of nearly three hundred feet. It is covered with scrub and coconut palms and there are several villages of Besarese fishermen. These Besarese people, a mixture of Malay and Biak Island Papuan origin, are noted sailors in the region, and in former days practiced a certain amount of local piracy and free lance smugeling. April 20, 1959 Birds from Papuan Islands i, Jusup Khakiaj collected a few birds on September 1, 1955, on Ajoe, and as the island has not been visited by a collector before, it is worth noting that he obtained the following species : Eos squamata squamata, Merops ornatus, Halcyon sancta sancta, Aplonis mysolensis mysolensis. LITERATURE CITED Beceari, O., 1875, Littera Ornitologica intorno agli uccelli osservati durante in suo recente viaggio alla Nuova Guinea. Ann. Mus. Civ. Stor. Nat. Genova, 7:704-720. Gyldenstolpe, Nils, 1955, Birds collected by Dr. Sten Bergman during his expedition to Dutch New Guinea 1948-1949. Ark. f. Zool. Kungl. Svenska Veten. Ser. 2, 8(2) :182-397. Mayr, E., 1937, Birds collected during the Whitney South Sea Expedition, 33. Amer. Mus. Novit. No. 915:1-11. Mayr, E., 1940, Birds collected during the Whitney South Sea Expedition, 41. Amer. Mus. Novit. No. 1056:6. Mayr, E., 1941, List of New Guinea Birds. Amer. Mus. Nat. Hist. New York. Mayr, E., 1942, Systematics and the Origin of Species. New York: 153. Mayr, E., 1944, The Birds of Timor and Sumba. Bull. Amer. Mus. Nat. Hist. 83(2) :162. Mayr, E., and de Schauensee, R. M., 1939, Zoological Results of the Denison- Crockett South Pacific Expedition for the Academy of Natural Sciences of Philadelphia. Part V.—Birds from the Western Papuan Island. Proc. Acad. Nat. Sci. Phila. 91:147. Meise, W., 1929, Die Vogel von Djampea, Jour. f. Orn. 77:458-460. Ripley, S. D., 1957, New Birds from the Western Papuan Islands, Postilla, NEEM Nos oles: Rothschild, Lord, Stresemann, E., and Paludan, K., 1932, Ornithologische Ergebnisse der Expedition Stein. Novit Zool. 38:127-247. Salvadori, 'T., 1880-1882, 1889, Ornitologia della Papuasia e delle Molucche, Mem. del R. Accad. del Sci. Torino 33 et seq. de Schauensee, R. M., 1940, On a collection of Birds from Waigeu, Notulae Naturae, No. 45:7. 1940, On a collection of Birds from Waigeu, Notulae Naturae, No. 45:7. Sharpe, R. B., 1869, On a new Kingfisher belonging to the Genus Tanysip- tera. Proc. Zool. Soc. London:630. Sharpe, R. B., 1892, Cat. Bds. Brit. Mus. 17:306. Siebers, H. C., 1921-1922, Boeroe Expeditie, Fauna Buruana, Aves: 151. Stresmann, E., 1914, Beitriige zur Kenntris der Avifauna von Buru. Novit. Zool. 21:388. Van Bemmel, A.C.V., 1948, A Faunal List of the Birds of the Molueean Islands. 'Treubia 19 (2) :323-402. Vaurie, C., 1949, A revision of the bird family Dicruridae. Bull. Amer. Mus. Nat. Hist. 93, Art. 4:289. ne Lee ets Be Se ee Dt) Oe be eS | i\ YALE PEABODY MUSEUM or Natura. History OCT 18 1960: // oslt [{ a HARD | UNIVERSIT No. 39 May 2, 1959 New Haven. Conn. ON MAKAIRA NIGRICANS OF LACEPEDE By James E. Morrow, JR. INTRODUCTION The identity of Makaira nigricans has long been a puzzle to ichthyologists engaged in studies of the istiophorids. While undoubtedly a marlin of some sort, the published description and figure (Lacépede, 1803: 688—691, Pl. 13, fig. 3) are such that it has been impossible to identify the fish with any known species. Lacépede himself never saw the actual specimen, but made his description from the notes and a sketch sent to him by MM. Traversay, Fleuriau-Bellevue, and Lamathe. The first named was the sub-prefect of La Rochelle, the second was said to have been a well-known naturalist of the district, and the third appears to have been a gentleman resident at Ars, on the western side of the Ile du Ré, where the fish was found washed up on a beach after a storm. Recently, through the efforts of Dr. Willard Hartman, Pea- body Museum of Natural History, Yale University, we received photographs of the original sketch (reproduced here as fig. 1) and notes from the hbrarian of the Muséum National d’His- toire Naturelle in Paris. On this drawing, several notes give the dimensions of various parts of the fish. These notes are in two handwritings. One is rather coarse and heavy and con- tains several misspellings. Presumably this is the hand of the person who drew the sketch. The other hand is much finer, aan eA 7AM | | in y Postilla Yale Peabody Museum No. 39 apparently that of a well-educated person, and is presumably the writing of either M. Fleuriau-Bellevue or M. Lamathe. The impression gained by examining the notes in the two hands is that those in the coarser hand were made first, and that those in the finer hand represent additions and corrections. Prof. Georges May, Department of French, Yale University, kindly puzzled out the sometimes rather illegible notations and archaic usages, a fe f ~ 0 dah Figure 1. The 'Traversay drawing, on which Lacépéde based his description ot Makaira nigricans. See text for notations. The small picture at the lower left is the illustration published by Lacépéde. Over the snout of the fish is written ‘*2 pieds de Longure De la Picque ou defense”; along the anterior edge of the dorsal fin, “Shautur 23 pouze”; above the middle of the back, in the better handwriting, “Cette partie se replojoit sur elle méme dans le corps de animal et saillois de 4 ou 5 pouces a l’ex- térieur”’; along the anterior edge of the second dorsal “Eleron 9 pouze”; across the spread of the tail, “hauture de la queue 4 pieds de haut,” and in the better hand, “d’une pointe A Pautre”; below the caudal peduncle, ‘pointe D’os 2 pouze”; Banh bialernncbe « jeeps Vie Souths jMiUd. 3 Ve llBiee. © | May 2, 1959 Mahkaira nigricans of Lacépede along the posterior edge of the anal fin, “Alero 5! 084960 | de deboute” (the last two words quite illegible, but, Prof, sae! suggests that a poorly educated person might hav magna i hry usage of the adverb “debout”) : below the belly, ees entier De Longueur 10 pieds,” and below this, in the better hand ‘‘trois pieds de profil’; along the anterior edge of the pectoral fin, “Longuer 23 pouze.” Notes on the back of the drawing, apparently written by Lacépede and copied for us by Dr. Hartman, read: “Il pesait 365 kylogr. 730 Ibs.” (the latter in pencil in another hand) “Les habitants de l’Ile de Rhé en ont mangé avec plaisir. La chair était un peu seche, non huileuse. Makaira. Makaira noiratre. L’individu dessiné a été pris aupres de la rochelle. La tempéte Vavait jeté sur le rivage. Le sous- préfet Lraversay m’a donné le dessin sur lequel j’écris cette note.” The notes on the separate page, in a hand similar to the finer one on the sketch, read: ‘Notes A ajouter a la description du poisson échoué sur les cotes de l’Isle de Ré en Vend™ an 10.— dont le dessin a été remis a M. de Lacépéde par M. de Tra- versay. Ce poisson n’a point d’event sur la téte comme les Marsouins. L’os de la déffense ressemble assés A Vyvoire. La deffense est ronde, et sans tranchans d’aucun coté. elle est droite, unie et sans sillons. I] n’a point de dents, son palais etoit extre- ment apre a la main. La chair tres blanche, courte, seche et dun gout fade. “Ces renseignmens ainsi que le dessein, ont été donnés a M. Fleuriau-Bellevue par M. Lamathe fils, demeurant a Ars, Isle de Ré M. Lamathe lui a mandé qu'il répondrois tres volontiers a toutes les questions que M. de Lacépede jugerois & propos de lui faire 4 ce sujet.” DISCUSSION Much of the difficulty that has been encountered in attempt- ing to identify any currently known species of marlin with Makaira nigricans is the direct result of two errors made by Lacépede in writing his description. First, he stated **Longueure totale, 330 centimetres,” but the notes on the original drawing clearly say that the length of the body—not the whole fish— was ten pieds (about 3250 mm). On this point, Cuvier (1831: + Postilla Yale Peabody Museum No. 39 290), referring to the figure published by Lacépéde, remarked, “M. de Lacépede a fait refaire le dessin d’aprés ces dimen- sions écrites; mais je crois qu’il a trop raccourci le corps, et que auteur du dessin n’entendait pas comprendre dans les vingt [sic] pieds l’épée ni la Caudale.” Following Cuvier’s hint, it would seem that the posterior reference point in measuring the length of the body was somewhere near the base of the tail. But from what anterior point was the body measured? Com- paring the length of the spear given by Lacépéde (2 pieds, or about 650 mm), with various measurements of snout length in other large marlins indicates that the spear of M. nigricans must have been measured from its tip to the tip of the lower jaw. Both the angle of the mouth and the anterior border of the eye are too far back. Measurements to either of these points from the tip of the snout would, in such a large fish, be on the order of 900 to 1,000 mm or more, rather than 650 mm. The nostril, another possible reference point, is so close to the eye that measurements to this would still be far larger than 650 mm. The most probable reference points for the snout measure- ments are thus the tip of the snout and the tip of the lower jaw. At 650 mm, this length of snout is quite reasonable for a large fish. Lacking evidence to the contrary, it is logical to assume that the anterior reference point in measuring the body length was also the tip of the lower jaw. This results in a standard length of the fish, from the tip of the snout to the tail base, of about 3900 mm. This is indeed a very large fish, but it is not beyond the recorded or reported size for several species of marlins. The second error of transcription in Lacépéde’s article is his statement ‘tune hauteur d’une meter’ (1,000 mm),whereas the note on the sketch reads “trois pieds de profil’? (974 mm). To the ichthyologist, ‘profil’ is not the same as body depth (hauteur). The latter is the straight-line distance from the mid- dorsal to the mid-ventral line, while the former is the distance between these two lines measured around the curvature of the body. For white marlin, our data indicate that the depth varies between 79% and 89% of the profile, averaging 82%. Data supplied by Dr. C. R. Robins, Marine Laboratory, University of Miami, produce an average relationship of a fraction over May 2, 1959 Makaira nigricans of Lacépede 5 82% for 26 specimens of Atlantic blue marlin. Nakamura (1938: 5) reported that the depth averaged between 83% and 86% of the profile in the three Pacific species examined by him. Since M. nigricans was an Atlantic fish, it seems justifiable to take 82.5% of the profile for an approximation to the body depth, or 804 mm. Two other points require comment. It has been suggested that the awkard positioning of the pectoral fins in the sketch was intended to portray the rigid pectoral fins of the black marlin of the Indo-Pacific. Indeed, we ourselves (Morrow, 1959) ad- vanced somewhat the same line of reasoning in seeking to establish the identity of Tetrapturus indicus Cuvier. But there we were dealing with a drawing made by an experienced zoologi- ‘al illustrator. In the present situation, the drawing apparently was made by someone with little education, zoological knowl- edge, or artistic ability. Particularly since the black marlin has never been recorded from the Atlantic ocean, it seems to us far more probable that Cuvier’s (1851: 289) interpretation was correct—that, consciously or no, the artist was influenced by experience with a fish that must have been well-known to him, the broadbill swordfish, and drew the fins as he thought they ought to be rather than as they were. The statement that the meat of M. nigricans was “tres blanche” has also been taken to indicate the black marlin, whose flesh is very white in contrast to the red meat of the Atlantic species. However, we have eaten nearly all the istiophorids”, and can testify from personal experience that the flesh of all these turns white when cooked. Since the rest of the description of the meat obviously refers to its condition on the table, the statement regarding color can only mean that it was well done. The remaining measurements, etc., seem clear enough, and as Lacépéde and the notes agree, there seems to be no need for further comment. We now have a set of dimensions for M. nigricans as shown in ‘Table I, and by exercising only the mini- mum of imagination, it should be possible to reconstruct the beast. The result is shown in fig. 2. * Atlantic and Indo-Pacific blue marlins, black, white and striped marlins, Atlantic and Indian ocean sailfish, Indo-Pacific spearfish. 6 Postilla Yale Peabody Museum No. 39 TABLE I Revised dimensions of Makaira nigricans Lacépede* elo higo ted onsaletinies ee ier 23 pouces 622 mm Wengthyor pectoral tite sere sce 23 pouces 622 mm IDG HY iE NOCH, odé oosniccae Hobceoontec 2.49 pieds 804 mm Length of snout, to Lipmote lower iy awe scat eee 2 pieds 650 mm Length of body, tip of lowerpyawatomtalle baSe@h seer = acre 10 pieds 3250 mm Standandileng ther. 125 mai ae eis el 12 pieds 3900 mm Rails preag stip suo) tip mame eee ete irele 4 pieds 1300 mm eight second Gorsal filles eee ieee 9 pouces 244 mm I ength posterior edge of HAE CNL os oemodaSoaodcnanan nee 15 pouces 460 mm * At the time when Lacépéde wrote, the pouce and pied had not been standardized. In converting to the metric system, we have used 1 pouce = 2.707 em and 1 pied = 0.3248 m. While this may result in some small error in absolute size, it will not affect the proportional relationships. The general facies of the reconstruction resemble those of the blue marlin more closely than those of any other species. Particularly, the robust body and moderately low dorsal fin place M. nigricans with the blue rather than white among the Atlantic species. The pectoral fin is very short, but its length falls at the lower limit of the range for blue marlin (Conrad and LaMonte, 1937: 218), and found also in black and Indo- Pacific blues (see Table II). The shape of the first anal fin is A ne OR ae) Os ~ a“ ———— == ) oe a a Se Ge \ = Wi wey ee = 4 4 10 20 30 40 50 60 70 PERCENTAGE OF STANDARD LENGTH Figure 2. Wakaira nigricans reconstructed according to the revised meas- urements of Table I. See text for discussion. UL popnypoUL JOU *Z' RT yey} Burueyzye Sotoods ut pue Guy Suourooeds UIUL QQGZ LAO [RUOLIPpe OMT, ¢ IBRADAR UL pepnypourl you Seep Uoultood [RUOTIPpe UY) ; uotatoeds uo ATUO poseq suotsodoard osoyy SYZMOAT OLZOUIOT[R JO osneRdg | wl OL (é"1.L) “PBpNULLog eEEl VVG appu Utd aq DAL DYV [VY w8 OL “SQ O07} Loto OSD 4G J&L “SQ G69 PLO JSosLe | SHOUPUL ‘T SB oUrRg (ST-6' 11) eel (6°&Z-0'81) GIG £61 (T9S1-€'201) S21 “SGT 000% L9A0 don *sqt O00T LIAO PIPLODIYT x notipul eT SR JULRES (9'9I-@' PL) Co | “Sq[ 000% 19A0 Clas lectiea it «9 VL “SAL 09ST plOdod oman] OEpURTPY Wo. POPLODIL LOAN aI il suede (° “RIPRASNY 07 ‘SN RD *S Ssuro0G¢ uvIpuy NY Ved “LOULY (O'ST-O'1T) Sb yo sydoa ON TT, “48 «8 SQ I9T PLO. YSOTLR'] “UROULR LLP Po] + BALLOPR I *SOLOZY ‘RIN yo soury S N S838¥O) ‘[}V (9°L1-8'IT) Vel “Usy Lowry . at OOOT L9A0 04 s}dou poyLtoau sy) "TLL -PL SQL 92 PlOdo1 YSosie yy %y » 32 IG UROG{QLER,) “RUSOP] YS SSOLOZY ' RLLOPR IA “ODOUR GT *RaNyLOg ROLLY “touly SYN SySvood "]LV jo (7'02-8'V1) Gh rol ~ eae AS "WU OO6E "T'S “S(T TOS “OOUBRL ST JZIS UYU ET "7 6 % sv wel 1oMoy diy 07 Jou wppu »D DAID YD [IT (612-91) (622-291) (¢°EZ-6'81) (¢°1Z-Z 141) 9°61 961 G'0% L6L O91 oe PSOne sv [eLoqyood YySuo'] (O'91-6 11) o(Q@¥1-z'01) (Z'0Z-L'e1) (FLI-8'Z1) OVE S Si S91 oS OL parent a geste tO) SU [RSLOP JUDLOP] (S46-6'°V9) <(8'0L-£'9F) (S981-F'86) (¢'98-@' 19) O'LL y09 Corl O'R! eS * R [estop }YUS OPT (1'1Z-0'S1) (VEa-G LT) (GOI-GEL) (0'Z2-@' 81) O61 0G OV 9°61 90% “Ty 6 % se yqdog DADZDUE noipul DPig|yy pydin § Upwey U DDD [i wpdmors] DLUDYD I DALY DY [IT DLUDMD [T "SUI[ARUL JO Sotood A9YJO FO VSOy, YIM supaLEhiu DOYYD [YT JO Th Wilda Val DSLIopoRLeyO JO UOStALduto,) 8 Postilla Yale Peabody Museum No. 39 a bit odd for any marlin species, but may be accounted for by the assumption that either the fin was damaged or was not fully erect when measured. Some support for the former may be found in the presence of a stub of bone (point d’os) just be- hind the first anal, indicating that the second anal had been damaged and lost. The second dorsal fin seems rather high, but we can offer no explanation. The weight of M. nigricans, 804 pounds, is rather light for a fish of its assumed length. Either we have made a gross error in assuming this length (which does not seem likely, as the various proportions fit quite well), or it really was a thin fish. According to length-weight relationships derived from various sources (our own data, as well as that published by Gregory and Conrad, 1939, Tables I and III: Conrad and LaMonte, 1937, Table I; Royce, 1957, Appendix Tables 1 and 2), a fish this size should have weighed well over 1,000 pounds. However, the same data also show that individual weights may differ from the computed regression by as much as 40% of the ex- pected value. There are also a number of possibilities which would result in a lighter weight than expected. First, the fish was washed ashore. This fact, coupled with its large size, at once suggests that it was an old, sick, dying animal, quite pos- sibly emaciated. Second, it may well have lain on the beach for a day or two before being weighed. The rapid loss of weight by fishes out of water is so well known as to warrant no further comment. Third, it is quite possible, even probable, that the fish was cut up in order to be weighed more easily, with concomitant loss of body fluids. It may even have been evis- cerated. And finally, it is not improbable that the weight was estimated, without ever putting the fish on a scale at all. We can now compare the reconstructed M. nigricans with each of the currently recognized (or recently described) species of marlin (see Table IT). The depth of the body, at 20.6% of the standard length, agrees with the Atlantic blue marlin (M. ampla), the Indo- Pacific blue (M. mazara), and the black marlin (Istiompaax indicus). The body is much too deep for either the white (J. albida) or the striped (M. audax), and not deep enough for the Bermuda marlin (WZ. bermudae). May 2, 1959 Makaira nigricans of Lacépede 9 The height of the dorsal fin, expressed as a percentage of the depth of the body, also appears to eliminate identification with the white, striped, and Bermuda marlins. In all three of these species, the height of the dorsal fin is greater than the body depth, but in M. nigricans the dorsal height was less than 80% of the body depth. This value agrees well with the figures for the two blue marlins, but is somewhat too high for the black marlin, except for a single specimen. The height of the dorsal as a percentage of the standard length agrees with the two blue marlins and the white marlin. It is far too low for the Bermuda marlin, slightly low for the striped, and a little too high for the black, although two speci- mens of this last species had dorsal heights that overlapped with M. nigricans. The weight of the evidence here, however, favors the blue marlins over the others. The pectoral fin of M. nigricans, at 16° of the standard length, is too short for any known species. Both the black and the Indo-Pacific blue, however, come close to this value at the low end of their ranges, and the Atlantic blue is not far off. Conrad and LaMonte (1937: 218) give a lower limit of 1/6 of the standard length for this last species. The relative length of the pectoral fin seems to rule against identity with either the striped or the white marlins, which have rather long pectoral fins. The length cf the snout, from its tip to the tip of the lower jaw, is 16.7% of the standard length in M. nigricans. This value is only just beyond the upper limit observed in the Indo- Pacific blue, and is within the range of all the others except the striped marlin. M. nigricans, at a weight of 804 pounds and an estimated total length of about 15 feet, was much too large to have been either a white or a striped marlin. Although the rod and reel record for the latter is listed at 692 pounds, the fact that this individual, caught nearly thirty years ago, was so much bigger (see Table II) than any individual taken since, despite greatly increased fishing intensity, suggests that it may have been mis- identified. It is quite possible that the fish was really an Indo- Pacific blue, a species generally not recognized by American anglers and ichthyologists at the time. Be that as it may, the 10 Postilla Yale Peabody Museum No. 39 size of M. nigricans falls well within the recorded or reported range of the two blue marlins and the black marlin. The locality of capture at once rules out identification of M. nigricans with any of the three Pacific species. In the more than 150 years since the discovery of M. nigricans, with thou- sands of marlin taken each year in recent times, not a single individual of any Pacific species has ever been reported from a location in the Atlantic, nor has any individual of an Atlantic species ever been found in the Pacific. (A possible exception 1s found in the Atlantic and Pacific blue marlins, which may or may not represent a single species. They are treated here as two separate species.) Neither the black nor the striped have been found in the Atlantic at any time. Summing up, then, the pectoral fin of M. nigricans is slightly shorter than the shortest pectorals of the Atlantic and Pacific blues, and the black. The gap between M. nigricans and either the striped or the white is rather wider, and there is no com- parative data for the Bermuda marlin. Of the remaining six characteristics considered, three rule out identification with the white marlin, viz: (1) depth of body as percentage of stand- ard length: (2) height of dorsal as percentage of depth: (3) overall size. Five factors militate against identity with the striped marlin: (1) depth of body as percentage of standard length; (2) height of dorsal as percentage of depth: (8) snout length; (4) habitat: (5) overall size. Comparing with the black marlin, three factors indicate non-identity: (1) height of dorsal as percentage of depth; (2) height of dorsal as per- centage of standard length; (with possible exception as noted) : (3) habitat. With respect to the two blue marlins, only one factor, habitat, rules out the Pacific form. All characteristics of M. nigricans are in agreement with those of the Atlantic blue marlin. The few measurements available for the Bermuda marlin absolutely prohibit any possibility of indentifying this form with M. nigricans. Makaira nigricans is thus established as the form currently known as M. ampla, the Atlantic blue marlin. The specific name, nigricans, has priority over ampla by 58 years. Makaira nigri- cans is the type species of the genus. The type specimen itself was eaten. The type material of the species is a sketch, accom- May 2, 1959 Makaira nigricans of Lacépede v1 panied by notes and measurements, now with the original manu- script of Histeire Naturelle des Poissons, volume 8, by Cuvier and Valenciennes, in the library of the Muséum National d’ Histoire Naturelle, Paris. SYNONYMY Makaira nigricans Laépede. 1803, Histoire naturelle des Poissons, Paris, vol. 4, pp. 688-691, Pl. 13, fig. 3. Tetrapturus herschelti Gray, 1838, Ann. Mag. nat. Hist., 7, ole) 0: Tetrapturus amplus Poey, 1860, Memorias sobre la historia natural de la Isla de Cuba, vol. 2, pp. 237, 243-244; ibid., 1861, Meo tie. 2: Makaira perezi de Buen, 1950, Publ. Cient. Serv. oceanogr. Pesca, Montevideo, 5: 171-175. ACKNOWLEDGEMENTS The field work involved in accumulating much of the com- parative data used here has been assisted and supported by Messrs. Wendell W. Anderson, Thomas Shevlin, John K. Howard, Alfred C. Glassell Jr., Al Pflueger, the American Academy of Arts and Sciences, and the Society of the Sigma Xi. Assistance in other directions has been received from Dr. Willard D. Hartman, Prof. Georges May, and Dr. C. Richard Robins. Messrs. Howard and Robins and Miss Francesca R. LaMonte have criticised the manuscript. To all these good friends and institutions, grateful thanks are tendered. 1 Postilla Yale Peabody Museum No. 39 LITERATURE CITED Conrad, G. Miles, and Francesca R. LaMonte. 1937. Observations on the body torm of the blue marlin (Makaira nigricans ampla Poey). Bull. Amer. Mus. nat. Hist., 74 (4): 207-220. Cuvier, Georges. 1831. In Cuvier and Valenciennes, Histoire naturelle des poissons, 8: 287-291, Paris. Gregory, William Kk., and G. Miles Conrad. 1939. Body-forms of the black marlin (Makaira nigricans marlina) and striped marlin (Wakaira mitsukurii) of New Zealand and Australia. Bull. Amer. Mus. nat. Hist., 76 (8): 443-456. Lacépéde, B. G. E. de. 1803. Histoire naturelle des poissons, 4: 658-691, Pl. 13, fig. 3, Paris. Morrow, James E. Jr. 1959. On the identity of Tetrapturus indicus Cuvier, 1831. Copeia (In press). Nakamura, Hiroshi. 1938. Report of an investigation of the spearfishes of Formosan waters. Rept. Taiwan Goyt.-Genl. Fish. Expt. Sta., 70. Translated from the Japanese by W. G. van Campen, U. S. Fish. Wildl. Sery., Spec. Sci. Rept., Fisheries No. 755, 1955. Royce, William F. 1957. Observations on the spearfishes of the central Pacific. Fish. Bull. 724, Fish. Bull. U. S. Fish Wildl. Serv., 57: 496-554. las paren 7A] [a en nel I oct 18 1960 ean) | R \ YALE PEABODY MUSEUM |_HNIERSIY oF NaTuRAL History Number 40 May 28, 1959 New Haven, Conn. A NEW SPECIES OF GRAMMATOSTOMIAS (FAMILY MELANOSTOMIATIDAE) FROM THE WESTERN NORTH ATLANTIC James EK. Morrow Jr.* The genus Grammatostomias is most easily distinguished from other genera of the family by the presence of a streak or loop of luminous tissue on the sides of the body above the lateral row of serial photophores. Within the genus, the form of the loop or streak, and the number of rays in the pectoral fin appear to be valid characteristics upon which the several species can be distinguished. Grammatostomias circularis, new species Figure 1. Grammatostomias circularis, new species. Drawn from the type specimen, 135.6 mm from snout tip to tail base. The skin of the caudal region has been slightly restored in the illustration. Drawn by Miss Shirley Glaser. Study Material. Type Specimen. One specimen, 135.6 mm in standard length, from the western north Atlantic, north of San Juan, Puerto Rico; Yale Peabody Museum of Natural 1793 History, Bingham Oceanographic Collection, No. 3773. * Bingham Oceanographic Laboratory, Yale University. i Postilla Yale Peabody Museum No. 40 Distinctive Characters. G. circularis is separated from the other two species of the genus by the presence of 9 pectoral rays and by the nearly circular form of the lateral loop of lum- inous tissue. Description. Proportional measurements of the type speci- men are expressed as percentages of the standard length unless otherwise indicated. Body: depth 10.5. Head: 15.6 Eye: 2.6; 16.5% of head. Snout: 3.4; 21.7% of head. Interorbital: 4.1; 26.0% of head; ca. 150% of eye diameter. Distance from snout: to origin of dorsal fin 78.3; to origin of anal fin 78.3; to origin of ventral fins 45.8. Dorsal fin: rays 21; length of base 14.6. Anal fin: rays 23; length of base 17.0. Pectoral fin: rays 9. Ventral fin: rays 8. Branchiostegal rays: 10. Serial photophores: Ventral row: I-P 7, P-V 18, V-A 21 (the last two above anal base), A-C 13. Lateral row: O-V 18, V-A 19, 20. Body slender, compressed, depth about 1/10 of standard length. Caudal peduncle about 5% of standard length, strong- lv compressed. Barbel pigmented basally, broken off, the part remaining not quite as long as head. Head about 1/6 of standard length, its dorsal profile gently rounded, premaxilla projecting into dorsal line. Snout longer than eye. Interorbital width greater than snout, about 1 1/2 times eye, slightly convex, with a low, inconspicuous ridge above each eye. Eye round, about 1/6 of head. A small light organ May 28,1959 A New Species of Grammatostomias 3 present on ventral edge of fleshy orbit below [sapecr oot eVe, Postocular organ elongate, its length about 5 times its w oe length less than 1/2 eye, long axis parallel to | upper voaw! A small, vertically elongate luminous spot presen (befgre 4960 | three small spots on upper jaw, one just befpre a organ, a second elongate spot behind postoculdr, afl@'a“fyird round one behind. Branchiostegals 10, a photo shoVAlYERSTY,- brane near base of each ray. Mouth extending nearly full length of head, gape straight. Premaxilla with a large fixed tooth anteriorly, followed by a larger depressible one which is largest tooth in upper jaw. These followed by two rigid, outer teeth, a depressible inner tooth, a rigid outer, a depressible inner, and five rigid teeth to posterior end of premaxilla. Maxillary with about 28 small oblique denticles on posterior part of its ventral margin. Man- dible with a large, rigid fang anteriorly, followed by a minute rigid tooth, a depressible tooth and a rigid outer tooth. Behind these, three inner depressible teeth and three rigid outer teeth, approximately in pairs, inner teeth longer than outer ones. Behind these, 11 small rigid teeth in’ single row, irregular in size, Ist and 4th longest. Vomer without teeth. Palatines with 3 or 4 teeth in single row on each side, 2nd and 4th teeth minute. "Two se of backwardly-directed teeth on tongue. Twelve small single teeth on first gill arch. Pectoral fins close to mid-ventral line, their origins just below posterior edge of gill openings, fins of 9 long, slender, dark-colored rays, several with slim luminous bodies, one with a large thick mass of luminous tissue basally. Pectoral rays about as long as a head. Ventral fins of 8 rays, well developed, originating before middle of standard length. Dorsal and anal origins on same vertical, anal base longer than that of dorsal, both fins with thick sheaths of body skin extending well up on the rays. Caudal forked. Sides of body with a nearly circular line of luminous tissue, its antero-posterior diameter slightly greater than length of post-ocular part of head, extending backwards from gill open- ings (see fig. 1). Vertical extent from near dorsum almost to 4. Postilla Yale Peabody Museum No. 40 lateral row of serial photophores. Luminous line quite even, smooth, without zigzags or noticeable thickenings. Skin smooth, scaleless, marked with vertical rows of tiny photophores, and with numerous small organs scattered over head and body. Color. The alcoholic specimen is dark brownish black. Serial photophores bluish, luminous loop pale violet. Type Locality. North of San Juan, Puerto Rico, 18° 55’ N, 66° 10’ W to 19° 05’ N, 65° 59’ W:: 0 to 225 fathoms. Name. The species is named circularis, with reference to the nearly circular shape of the lateral loop of luminous tissue. Comparision With Other Species. The present species is most easily compared with others in the genus by means of the following key. Key to Species of Grammatostomias la. Sides with a long luminous line from just behind gill cover to behind ventral bases, hooked sharply downward at its anterior end. Pectoral with 5 rays........... dentatus Goode and Bean, 1895* 1b. Sides with a closed loop of luminous tissue. Pectoral rays Oto WL. 2a. Luminous loop elongate, extending posteriorly about to ventral bases, its anterior ventral portion thickened and PNET AO Br cs MAR flagellibarba Holt and Byrne, 1910.** 2b. Luminous loop nearly circular, without thickenings or PMG MS te ae Sire tah gost te Seat ieee circularis new species. Acknowledgements. We wish to express our gratitude to Dr. ichar« . Backus, Woods Hole Oceanographic Institution, Ricl 1 H. Back W ls Hole O grapl Institut who collected the type specimen, for his gift of the same to the Peabody Museum’s Bingham Oceanographic Collection. We are also grateful to Miss Shirley Glaser for her fine drawing. *Lamprotowus angulifer Beebe 1932 is a synonym. ** Lamprotowus paucifilis Regan and 'Trewavas 1930 and Lamprotoxvus phanobrochus Regan and 'Trewayas 1930 are synonyms. MUS. COMP. 7001 LIBRARY APR 2 6 1960 HARVARD UNIVERSITY ostilla YALE PEABODY MUSEUM oF NaTuRAL History No. 41 Sept. 15, 1959 New Haven, Conn. BIRDS FROM DJAILOLO, HALMAHERA' By S. Ditton RIPLEY My wife and I had the opportunity of spending a week in the vicinity of Djailolo on the island of Halmahera from September 2 through 8, 1954, during a trip in the Moluccas sponsored partly by the Guggenheim and National Science Foundations as well as Yale University. Djailolo was visited briefly by Alfred Russel Wallace in 1858 who remarked (1869) on the lack of original forest in the area around Sahoe, and the vast extent of heavy grass and high reeds which made bird study very difficult. We had origin- ally intended to spend several months on Halmahera, but were prevented from doing so by a local outbreak of guerilla activity and so had to content ourselves with a short visit to the Djai- lolo area. In back of the village there is a low, conical, three thousand foot mountain, Mount Djailolo, which has patches of heavy forest on its steep slopes. In addition, an experimental agri- cultural farm at Achango, seven kilometres by road to the north gave us an opportunity to camp in the midst of a varied environment, patches of dense scrub mixed with sago swamp, * Dedicated to Professor Erwin Streseman in honor of his seventieth birthday. 2 Postilla Yale Peabody Museum No. 41 cleared fields, high stands of cultivated trees such as kapok and shade trees used in the plantations of cocoa and nutmeg. Five kilometres farther along this road to the north is a small auxiliary airstrip which occasionally serves the Ternate area. Altogether in 1954 this was a far more rewarding mixture of scrub and second growth forest than one would be led to be- lieve from Wallace’s description. Is this short time we failed to see many of the birds collected by Wallace or later visitors such as Bernstein, Heinrich, or de Haan. However, we made a small collection and a few ob- servations which may be of interest. Several species were seen but not collected as follows: Sterna sumatrana, Djailolo Bay Tadorna radjah radjah. The Radjah Shelduck was flushed out of a sago swamp at Achango, the female uttering the char- acteristic grunting call as they flew. Spizaetus gurneyi. A pair of this hawk-eagle was observed on the slopes of Mount Djailolo. As the birds flew they showed a prominent mirror patch at the base of the primaries. The birds first appeared about 8:30 a.m., circling high over the heavy forest about half way up the mountain approximately 1500 feet above sea level. Their circling was purposeful, taking them ever lower down to lower altitudes, finally to the area of semi-cultivation and scrub until they were lost to view in the lowland coconut plantation. Hornbills, Aceros plicatus ruficollis, were coming into breed- ing condition at this time although still occupying a communal roost in the kapok trees. However, display flights were com- mon. The appearance of the Hawk-eagles seemed to drive them into a frenzy of display. Six hornbills could be seen at one time in the air, circling round and round in tight circles as the hawk-eagles flew by, banking sharply with set wings making a characteristic whirring, drumming sound. Rallus philippensis subsp. Two rails were flushed from the paddy fields at Achango on September 8. They were close enough to spot the grayish-streaked sides, brownish-grayish ios ~VUL ; APR 26 1960} wiry back and reddish streak through the eye eeabieying them as 4 the Banded Rail. Presumably they were migrants into the north Moluccas from Australia and belonged to the population yorki. This species has not previously been recorded from Halmahera vide van Bemmel (1948). Sept. 15, 1959 Birds from Djailolo, Halmahera Centropus bengalensis medius. Seen in long grass at the Djai- lolo airstrip. Species collected. 1.) Accipter novaehollandiae griseogularis (Gray) In a backyard garden at Achango. Ptilinopus hyogastra (Temminck ) iw) wa Often perched on telephone wires. Males were coming into breeding condition ahead of females. The first female with enlarged ovaries was taken in October on Batjan. Weight; Ge culoo, G92 72 2 105, 162) erams, 3.) Ducula bicolor (Scopoli) 4.) Reinwardtoena reinwardtsi reinwardtsi (Temminck) 5.) Macropygia amboinensis batchianensis (Wallace) 2 ovaries enlarged September 7, weight 139 grams. 6.) Geoffroyus geoffroyi cyanicollis (S. Miller) Not breeding. One bird very worn. Weight 3 ¢ 174, 188, 2? 2 160, 188, 190, 222 grams. 7.) Scythrops novaehollandiae Latham The Channel Bill appeared for the first time on September 2, and was seen daily flying north in small groups from two up to ten. The birds appeared to be migrating. A single male had small fruits in the stomach. Unrecorded previously from Halmahera. 8.) Centropus goliath Bonaparte This coucal was heard to give two calls, a curious chuck- ling which sounds like a rail, and a deep low moaning roar. Local name, “Sokukud.”’ a Postilla Yale Peabody Museum No. 41 9.) Otus leucospilus (Gray) This is a large dark owlet, larger than any of the forms of scops (wing of our é 173 mm.), dark above and with pro- nounced dark central streaks on the feathers of the upper surface. In southern Asia the typical call of scops may be likened to “tonk tonk ta tonk.” As Heinrich (1940) has noted, this species calls entirely differently. A single male perched about thirty feet above the ground in dense shade trees at Achango ‘alled a single, rasping “‘kwok” at regular intervals with per- haps ten seconds between each call. The resonance and growling quality of the call sounded more like a barking deer than a bird. Weight 160 grams. Local name, “Goroko.” 10.) Ninow connvivens rufostrigata (Gray) A female collected from a shade tree at Achango had a two-syllabled call like the yapping of a small dog, “ow-wow, > Heinrich (1956) found this owl in the mountains in contrast to our experience. OW-Wow..’ 11.) Caprimulgus macrurus schillmolleri: Stresemann A male in breeding condition weighed 79 grams. Call, the familiar “tock tock” of the species. 12.) Collicalia vanikorensis moluccarum Stresemann A female coming into breeding condition weighed 11 grams. Another female weighed 12 grams. First record of this subspecies for Halmahera vide Van Bemmel (1948). One of these specimens with a wing measurement of 114 mm. seems exceedingly close to the race waigewensis. The other female with a wing of 107 mm. fits closer to moluccarum. 'These speci- mens belong to the vanikorensis assemblage with uniform backs and unfeathered tarsus. Local name “‘putih.” 13.) Collocalia esculenta nubila, new subspecies. Type: 6 ad. (Y.P.M. no. 36966), collected September 6, 1954, by S. Dillon Ripley at Achango, Djailolo, Halmahera, Indonesia. Diagnosis: The single specimen of Glossy Swiftlet taken by me on Halmahera prompted me to examine comparative material of this species. From this it is at once apparent that Sept. 15, 1959 Birds from Djailolo, Halmahera 5 the population found on Morotai, Halmahera, Ternate and Tidore differs strikingly from typical esculenta of Obi, Buru, the southern Moluccas, Celebes (Sulawesi) and New Guinea in being dark below, the abdomen being clouded over. The feath- ers of the abdomen have dark greenish or dark greenish fuscus centers with white edges only. In this character nubila is similar to dodgei of Borneo or bagobo or isonota of the Philip- pines. This population, however, is smaller than these, more northern forms, and also far more glossy on the back, match- ing typical esculenta in this. In addition the abdomen is even more suffused than in the Philippine races. Wing measurements of nubila are; 64 93 - 96; 32 92 - 95; 5 sex indet. 90 - 96 mm. Weight, 1¢, 6 grams. This new form is extremely interesting from a zoogeo- graphic point of view, showing as it does a strong relationship to the populations of the southern Philippines. I am grateful to Dr. Junge at Leyden, Professor Strese- mann at Berlin and Dr. Amadon of the American Museum of Natural History in New York for the loan of specimens in their care. 14.) Hemiprocne mystacea confirmata Stresemann A pair were in breeding condition in early September and weighed 6 76, 2 69 grams. 15.) Ceyx lepidus uropygialis (Gray) Weight; ¢ ¢ 11-20 (mean 16); 2 2 17, 22, 30 grams. 16.) Halcyon diops diops (Temminck) Common in cultivated areas, often on telephone wires. In addition to the character of the breast band in the female, the lack of the white neck ring and so forth, there is a pronounced weight difference between the sexes. A young male which an- swers to the description of the “young female’? in Sharpe (1892) weighed 37 grams. Two ¢ adult weighed 43,45; ? & ad. 65, 65 grams. Local name “Chawahiru.” 17.) Halcyon funebris (Bonaparte) This heavy-set dark brownish green kingfisher, while pos- sessing a plumage pattern of spots and neck ring rather like 6 Postilla Yale Peabody Museum No. 41 the chloris assemblage has a superficial resemblance to the brightly colored winchelli-hombroni group of the Philippines. 18.) T'anysiptera galatea isis Gray Comparison of a small series of Racquet-tailed Kingfishers from Halmahera and Batjan shows that Halmahera birds have an ultramarine crown only very narrowly bordered on the sides with cobalt which forms a superocular stripe. In the Batjan population the cobalt is much more pronounced, being broad, extending onto the crown and making a noticable nuchal ring. G. R. Gray (1860) described isis from “Batchian and Gilolo,” a name which has been merged with margarethae of Heine (1859) from Batjan. I hereby restrict the type locality of isis to Gilolo ( = Halmahera) which thereby becomes avail- able for the Halmahera population. A female called with a single soft mewing note. Weight 6 55, 2 2 64, 78 grams. Local name, “Menyalum.” 19.) Eurystomus orientalis pacificus (Latham) Weight ¢ é 175, 182 grams. 20.) Hirundo tahitica frontalis Quoy and Gaimard One, sex indet. weight 15 grams. 21.) Artamus leucorhynchus leucopygialis Gould Male, weight 46 grams. 22.) Aplonis metallica metallica (‘Temminck ) Starlings were in small flocks in scrub-edge of forest areas about 650 feet above sea level. They made a series of short tinkling calls. Weight 2 60 grams. Local name “Hidis.” 23.) Corvus validus Bonaparte Local name “Bokogk.” 24.) Lycocorax pyrrhopterus pyrrhopterus (Bonaparte) This crow-like bird of paradise occurs from sea level to the tops of the mountain ridges. Normally the call is a very harsh rasping “tschak tschak.” A pair at Achango were sitting closely side by side on a coconut palm frond. Both birds had enlarged gonads on September 7. A third bird was sitting near by. One bird, the male, was making a very deep, low ba . Sept. 15,1959 ~~ Birds from Djailolo, Halmahera 7 “om” sound, evidently a display note. It seemed to swell up and bow slightly as it called. Local name ‘‘Siamit.” 25.) Lalage aurea (‘Temminck) A bird of open scrub; 6 weight 32 grams. 26.) Coracina atriceps magnirostris (Bonaparte) Weight ¢ 150, 2 128 grams. 27.) Coracina papuensis melanolora (Gray) Weight 2 82 grams. 28.) Hysipetes affinis chloris (Finsch) Specimens from Halmahera seem slightly more bronzey, less pure yellow green above and below than those from Batjan. There is no difference in size, however, and this distinction does not seem marked enough to warrant nomenclatoral recognition. Birds were in breeding condition in September. Weight 6 é 38, 41, 2 2 41, 42 grams. Local name “Klaitua.” These bulbuls were often in small family parties of from two or three to nearly a dozen. On one occasion I found Monarcha trivirgatus flocking with them evidently in a mixed feeding association. They occurred in open scrub or dense for- est from sea level up to nearly five thousand feet without evident altitudinal or habitat preference. 29.) Myiagra galatea galatca Gray A male in breeding condition weighed 11 grams. The bill of this specimen shows no evidence of mal-formation but in form resembles typical Monarcha species, being laterally com- pressed, with little of the tumid appearance of Myiagra. No other specimens collected, or examined in the series in the American Museum of Natural History show this appearance, perhaps a mutant gene for bill shape approaching the related monarch flycatchers. 30.) Monarcha trivirgatus bimaculatus Gray Male in breeding condition, weight 15 grams. 31.) Nectarinia sericea auriceps Gray A male had enlarged gonads although a female showed no enlargement. Weight ¢8, 2 7 grams. 8 Postilla Yale Peabody Musewm No. 41 32.) Melitograis gilolensis (Bonaparte) A male with gonads enlarged was taken in substage forest growth at 850 feet above sea level. The bird called with a single harsh rasping note. Weight 54 grams. Local name *Sotosoto.” This honeyeater was always seen as a solitary individual and seemed to show aggressive behavior. On one occasion as I have described elsewhere (1959) I saw a single bird disperse a flock of bulbuls. 33.) Lonchura ferruginosa jagori (von Martens) A male with gonads shghtly enlarged was taken out of a small flock in heavy weeds on the edge of a garden at Achango. Weight 15 grams. Local name “Kotolor” Like Hypsipetes, Collocalia esculenta nubila and perhaps Pitta maxima and Halcyon funebris, the occurence of Lon- chura ferruginosa jagort on Halmahera emphasizes the zoo- geographic link between the northern Moluccan islands and the Philippine Archipelago. Although the predominant early avifaunal influence in the area can be said to have come from the Australian-New Guinea region with these islands contain- ing the most westward extensions of the families of the Birds of Paradise and the Honeyeaters, represented in each case by endemic genera, still the importance of the Philippines should not be underemphasized. It is instructive in this regard to be in Halmahera during the migration as we were, and to note the arrival daily of such species as the warblers, Phylloscopus and Locustella, the flycatchers, swallows, and the others, so many of which have obviously come directly from the islands to the northwest on passage. These connections can only have been adventitious, over water, but the route is there ready to hand. LITERATURE CITED Gray, G. R. 1860, Proc. Zool. Soc. London : 347. Heinrich G. 1940, Journ. f. Orn. vol. 88 (1) : 5. 1956, Jour. f. Orn. vol. 97 : 36. Ripley, S. D. 1959, Evolution vol. 13 (in press). Sharpe, R. B. 1892, Cat. Bds. Brit. Mus. vol. 17 : 254. Van Bemmel, A.C.V., 1948, Treubia vol. 19 (2) : 323 -402 Wallace, Alfred Russel, 1869 'The Malay Archipelago, London 2 vols., re- printed 1922 : 242. mud. UW? POST ees er | APR 2 6 1960 ARVARD UNIVERSITY Salle YALE PEABODY MUSEUM oF NaTurRAL History Number 42 December 20, 1959 |New Haven, Conn. — CHARACTER DISPLACEMENT IN INDIAN NUTHATCHES (Sitta). S. Ditton RIPLEY Is connection with work on the birds of India, I have had occasion to arrange the various nuthatches of the genus Sitta occuring in India in a purely linear listing suitable for a check- list. Arrangement of this sort at once reveals that far more is at stake than a mere listing can indicate. Some aspects of the problem are given in the recent comprehensive treatise by Voous and Van Marle (1958) who have attempted to recreate the distributional history of the several species of Sitta of Southeast Asia. It is quite clear from a study of these nuthatches that several species are involved, that they are all closely related and that they tend to replace each other. However, where these species are sympatric they show character displacement as well as a degree of niche specificity, thus a tendency to special adaptations as discussed by Brown (1958). That closely related species of nuthatches exhibit this type of spe- ciation allowing geographical overlap and ecological specificity has been well illustrated by Vaurie (1950, 1951) in his two papers on Sitta neuwmayer and S, tephronota of west central Asia. These two species of Rock Nuthatches are admitted to be most closely related to each other. Voous and Van Marle (ibid: 54) feel that the Rock Nuthatches are in addition a 2 Postilla Yale Peabody Museum No. 42 central Asian offspring of the common European Tree Nut- hatch adapted to a xeric environment. That these forms could have split off from the tree-inhabiting nuthatch at op- posite geographically isolated ends of two glacial refugia in the southeastern Mediterranean and central Asia respectively seems conceivable. At a later stage the two populations, al- most indistinguishable morphologically and with similar habits and ecological requirements, having spread into contiguous areas have evolved adaptively so that both can coexist in the same range without hybridization. Morphologically the two species differ in the area of overlap by a significant change in bill size. Whereas isolated populations at the ends of the range have similar bills, the overlap populations differ greatly, the Asian tephronota possessing a large bill, the Mediter- ranean neumayer a relatively small bill. In addition whereas both species possess a black facial stripe of equivalent size and length, running from the base of the bill through the eye and back to the nape, the overlap populations differ greatly, the eye stripe in Asian tephronota being much enlarged and prolonged farther backwards, that of the Mediterranean neu- mayer being greatly reduced to a strip between bill and eye passing backwards to just above the auricular region. Thus two species which are virtually sibling species have developed prominent recognition marks and adaptations for food gath- ering sufficient to prevent interbreeding and reduce competition. That competition is a factor seems inevitable from what is known of the habits of these species, both living in xeric areas of cliffs, rocks and low stunted trees. Both occur in the same areas and have been collected together at the same altitude. This biotope would appear to be homogeneously diverse to use Hutchinson’s term (1957) and would confirm his supposition that in stable homogeneously diverse biotopes the abundances of different species are arranged as if the realized niches were non-overlapping. A different series of constants are involved in the Himalayan ranges of India and Pakistan. From the lowlands to the heights of 15,000 feet above sea level, a number of biotopes occur which are essentially heterogeneously diverse, woodland December 20, 1959 serub to omen forest. subtropical, Aw: Dy (AR fh Indian Nuthatches (Sittal)/i i & 0 1BQY | such ecological details as are known: pine, temperat The following species of Sitta may be listed, giving Species Habitat Preference Altitudinal Range 1) 3) 4) Sitta castanea castanea Sitta castanea neglecta Sitta castanea cinnamoventris Sitta castanea cashmirensis Sitta europaea, various subspecies Sitta europaea montium Sitta himalayensis, various subspecies Sitta leucopsis Sitta victoriae Sitta yunnanensis mango groves, orchards, cultivation, tropical deciduous forest tropical lowland forest, also scrub and cultiva- tion masonry walls, gardens, bamboo clumps, scrub, sub-tropical forest mixed forests at all levels, lower parts of trees mixed forests at all levels mixed alpine and fir forests heavy mixed forests, usually deciduous, strong preference for oaks; may descend to under- growth almost exclusively upper parts of trees in fir and pine forest alpine forest, avoids pines barren fir forest association lowlands to 3,500 feet in central India; low- lands to foothills 1,000 ft. in Himalayas lowlands to 2,500 ft. edge of the plains (winter) to 4,500 ft. 4,500-11,500 ft. 4,500-8,500 ft. 4,500-12,000 ft. 5,000-10,000 ft. 7,000-12,000 ft. 7,500-9,200 ft. 9,000--15,000 ft. t Postilla Yale Peabody Museum No. 42 As pointed out by Voous and Van Marle (ibid :59-61) all these species are roughly similar in size and close to each other in pattern. Some wing and bill measurements can indicate relative size: bill Wing (mean length m.m.) (mean m.m.) 1) castanea castanea 19.1 75.5 2 neglecta 19.8 79.9 i cinnamoventris 23.2 $3.7 > cashmirensis 21.8 85.2 2) europaea, various subspecies 18.8 78.6 4 nagaensis 19.1 78.2 iH montium ill 80.2 3) himalayensis himalayensis 15:9 12.3 a australis 16.6 73.6 4) leucopsis leucopsis 21.6 19.2 x przewalskii 17 75.1 5) victoriae 14.5 70.2 6) yunnanensis 17 73 Status of Sitta castanea Sitta castanea with its various allopatric subspecies has often been listed as part of the common Palaearctic Tree Nuthatch species, Sitta europaea. It is obvious that it is most closely related to Sitta europaea and it satis- fies a taxonomist’s criterion by being strictly allopatric, both geographi- cally and ecologically. However, Sitta castanea differs from europaea by having pronounced sexual dimorphism and a strikingly different color pat- tern on the under surface of the males. Its distribution throughout the Gangetic plain and hills of the Indian Peninsula, with the development of a distinct subspecies in the Eastern Ghats, gives strong evidence for a long colonization similar to that of some of the double invasions and relict forms discussed by me previously (1949). I feel that it belongs to an early break-off of europaea stock, separated from that species in time and by the imposition of two other old Palearctic invasions into the Himalayan chain, the earliest Sitta leucopsis, the second Sitta himalayensis. Contra Voous and Van Marle (ibid:53, 55) I believe that the hill populations of cas- tanea, namely cashmirensis, almorae, cinnamoventris and tonkinensis, which are all larger in measurements than the lowland populations, castanea and neglecta (see measurements) have developed from the older lowland stock which has been able to capture and exploit vacant niche space in the ad- jacent hills in post-pluvial times. That this has been done more than once is shown by the isolated eastern, Indochinese population of tonkinensis which has developed a highland form in the same way that the western forms have evolved. Sitta leucopsis and Sitta himalayensis I would have listed these in my zoogeographic study of the Indian avifauna (1959) except that by ranging into the Indochinese and Chinese subregions, these species did not fit my criteria for such a listing. How- ever, I would include them here along with species I did list (ibid:79) such as Zoothera wardi, Parus melanolophus, Sitta formosa and Pyrrhula erythrocephala and P. aurantiaca as Palaearctic relicts. December 20, 1959 Indian Nuthatches (Svtta) 5 Sitta victoriae From the list of measurements it will be noted that this isolated species, confined to the summits of the Chin Hills, Mount Victoria, of western Burma, is relatively small in size. It has been listed in the past as a sub-species of S. himalayensis, indicating obvious affinity, (Voous and Van Marle, ibid: 58) but they as well as other authors have overlooked the fact that himalayensis has been taken on the same mountain. A view of a map should prove instructive here. Surveying the Himalayan chain from Kashmir east to Sikang and northern Indochina, I have in- dicated the zones of ecological overlap between the species under discus- sion. Although the ranges overlap geographically, it can be seen that there is very little overlap among these species: Altitude (feet) 15000 12000 10000 N \ Lda 7000 5000 ADIAOJIIA @ SISUBAD/OWIY ™ SISUQUDUUNA = A i>) Q D ~s Q 3 Aa) Q i>) Q ® “~s Q 3 Aa) Q SISUBJIWYSOI DAaUD{SOD w SW4O0} ||I4Y ‘O9UV0/S02 2 Suo!}D|ndod om} ‘s/sdoanas m wnijuoW Ppud s/isua0boUv DaDdOINI _O Figure 1. Altitudinal Ranges of Species and subspecies of Sifta. G. e victoriae 2 H S ~ is] ™ 8 & Ss uj ae a /EUCODSIS H. <== yunnanens/s castanea europaea N KASHMIR Figure 9 a Map showing geographical orientation of species of Sitta. December 20, 1959 Indian Nuthatches (Sitta) w From the above it can be seen that there are only a few cases of overlap which need to be explained. Among the sub- species of castanea there is no overlap. The apparent range overlap is due to wintering birds of the montane subspecies descending into the range of the lowland form out of the breeding season. In addition A, B and C are allopatric with E, G, and H. The following range overlaps then need some word of explanation. 1. In Kashmir and adjacent areas of West Pakistan and eastern Afghanistan, Sitta castanea cashmirensis overlaps with S. 1. lewcopsis but is ecologically separated, the latter prefering the upper branches of firs and pines, the former the lower branches of mixed deciduous and evergreen hardwoods. These two species then occur within the mosaic elements of a heterogeneously diverse biotope. There is no competition, hence no character displacement. 2. In eastern Assam in the hills south of the Brahmaputra River, although castanea has a separate population, koelzit, it is allopatric with ewropaea, the latter not occuring below 4,500 feet, the former being a lowland and submontane form up to 4,500 feet. However, a population of himalayensis over- laps with ewropaea in these hills and occupies the same alti- tudinal range and much the same biotope. Both inhabit mixed forest. The only known ecological difference is the preference of himalyensis for oaks, rhododendrons, and lower areas of trees. Evidently a degree of interspecific competition exists here. In the contact area both have evolved recognizable dis- tinct populations showing character displacement. The char- acter displacement follows the same general pattern as in the case of the Rock nuthatches. (See Fig. 3, Page 8) As the figure indicates, the overlap population of ewropaea, called nagaensis, has become very pale on the under surface with a slimmer bill and a pronounced facial stripe. The over- lap population of himalayensis, called australis, has become richer buff on the under surface with a stouter bill and a shorter facial stripe. Thus morphological differences empha- sizing both feeding habits and recognition are involved in this interspecific situation. That characters having recognition 8 Postilla Yale Peabody Museum No. 42 A. europaea (not in contact) C. europaea poke D. hAimalayensis (not in contact) (contact Eastern Assam, Mount Victoria.) E. victoriae (contact Mount Victoria) Figure 3. Outline sketch to show differences between sympatric populations of species of Sitta (C., D., E.) compared to two populations not in contact (A., B.). value are developed even between these two fairly distinct species would imply that selection pressure towards plumage pattern differences is heavy. In a similar way North American warblers have evolved marked phenotypic differences as well as behavioral and food gathering adaptations as discussed by Mac Arthur (1958). December 20, 1959 Indian Nuthatches (Sitta) 9 On Mount Victoria south of the Assam Hills, a third sym- patric species is introduced, Sifta victoriae. This species with a range of 7,500-9,200 feet overlaps altitudinally on the moun- tain stopes with europaea, which reaches 8,500 feet, and with himalayensis, with which it is assumed to be most closely re- lated, which also reaches 9,200 feet. It would appear from a biogeographical point of view that wictoriae is an old _ relict species, first to arrive of a double invasion by himalayensis stock. It has a known ecological preference for alpine forest and avoids pines. Himalayensis also avoids pines and co-occurs with victoriae in mixed and alpine forest. Europaea may occur in alpine as well as mixed forest. In this case ewropaea and himalayensis have a wide range in eastern Assam of co-occur- ence and their morphological characters have been described. The third species victoriae differs by being noticably smaller with a more weak and slender bill and different facial pattern. 3. Eastwards in Tibet, Burma, Thailand, Indochina and Yunnan there is little actual overlap. Wherever europaea has invaded the mountains of southeast Asia, castanea keeps at lower altitudes. In the single instance where castanea in- vades the mountains, europaea is absent. Himalayensis occurs at predominantly higher altitudes. 4. The single instance of co-occurence is in the case of three species in Sikang or eastern Tibet. Here are found a population of ewropaea, one of leucopsis, and Sitta yunnanen- sis. Leucopsis has developed rich bright color on the under- parts in contrast to the population of lewcopsis found in the Himalayas. The population of ewropaea, known as montium is intermediate in color of underparts, while the palest of the three is yunnanensis. Leucopsis lacks an eye stripe, possessed by the other two of differing lengths, longer in europaea, and has a short bill, the shortest of the three. In this instance yunnanensis has a relatively long needlelike bill, and europaea an intermediate, stouter bill. From the table of measurements it will be seen that ewropaea montium differs from other popu- lations of the species by being larger with a larger bill, thus showing character displacement in the presence of the other two species. Schafer (1938) reports that ewropaea and yun- 10 Postilla Yale Peabody Museum No, 42 nanensis were found together, although the latter was found only in rather barren, fir forest. He found leucopsis less often, in more open parkland. In the presence of yunnanensis the population of europaea found with it appears to have de- veloped a shorter, stouter bill, larger size, and a longer eye stripe, as well as somewhat more richly colored underparts, all tending towards character displacement. Even leucopsis, not apparently in close competition, would appear to have been affected, so markedly does it differ from the other popu- lation of the species. Wn D. europaea a \ F. /eucopsis —— B j BeOS ‘i Sagat ae ae SNe $ emv|@ god] S$ [5G Iu 2 3 4 5 6 7 8 g 10 \7 af embryonic period egg tooth ae ae hypothetical typical SS DKF, 3 aa galliform l 2 3 4 5 6 7 8 3 io ee eee BS Ne Bey embryonic period period of requisite parental care hatching flying + + TIME IN| WEEKS Figure 1. Diagram of the hypothesized morphological and behavioral development of the megapode compared with that of more typical galli- naceous birds. Ua Dirks eid} if! potlegciTp embryos. According to Hamilton (1952: 375), the remnants Feb. 15,1960 Embryology and Evolution of Mega of the labial groove are shed on the nineteenth day of incu- bation in the chicken. EARLY PLUMAGE SEQUENCES The newly hatched megapode bears primarily pennaceous feathers in contrast to the ‘downy plumage” of more typical gallinaceous birds at hatching. Studer (1877, 1878, 1889) reported finding down feathers at the tips of the pennaceous feathers in Megapodius embryos. Both pennaceous feathers and down were ensheathed. Studer believed that the down feath- ers were lost before hatching. Blaszyk (1935) thought that what Studer had termed a down feather was merely the result of making a section through the distal end of an ensheathed pennaceous feather. Blaszyk (1935) and Becker (1959) found no traces of embryonic down feathers in Megapodius. Neither Blaszyk (1935) nor Becker (1959) commented on a study by Figure 2. a. Dorsal view of the upper jaw of a Talegalla embryo; note the vestigial egg tooth. b. Lateral view of the upper jaw of the same Talegalla embryo; note the labial groove. e¢. and d. Dorsal and lateral aspects of the upper jaw of a chicken at the time of hatching. 4 Postilla Yale Peabody Museum No. 45 Pycraft (1900), who had reported finding down-like rudiments on the tips of the first pennaceous feathers in Megapodius embryos. I have made a preliminary examination of embryonic Talegalla feathers and have found evidence that Pycraft’s account is fundamentally correct. Both the five week old pheasant (Phasianus colchicus ; West- erskoy, 1957: 20) and the newly hatched megapode bear predominantly juvenal plumage. Thus at roughly 8 weeks postconception the megapode and typical gallinaceous birds appear to reach similar levels of plumage development. Wallace (1860) commented that the newly hatched megapode behaves about as maturely as a chicken at one month posthatching. SPECIAL EMBRYONIC ADAPTATIONS Correlated with the prolongation of the prehatching period, certain embryonic adaptations are present in the megapodes. Meyer (1930: 5) reported that in Megapodius yolk weight was about 60% of total egg weight in boiled eggs; this was compared with a figure from the literature of 30% for Gallus domesticus. Megapode eggs are usually incubated with the small end pointed downward. Bellchambers (1921) thought that this aided the escape of the newly hatched bird from the mound. Umanski (1926) reported on the effects of placing incubating chicken eggs on end. The proportion of teratological effects was greater in eggs placed with the blunt end down than in ones with the pointed end beneath. Normally when the chicken egg lies horizontally, the early embryo les in a horizontal plane. Umanski suggested that the teratological forms found in eggs placed upright might result from the failure of the blastoderm and early embryo to reach a horizontal plane. Experiments (Marshall, 1939) have shown a failure in hatching success in chicken eggs unturned during incubation. Megapode eggs usually are not turned and yet hatch relatively successfully. Marshall (1939) concluded that there must be anatomical and/or physiological differences between megapodes and the domestic fowl with respect to turning. One of the presumed embryonic adaptations related to the long embryonic period of the megapodes is a general lack of Feb. 15,1960 Embryology and Evolution of Megapodes 5 coordinated movement of the embryo until shortly before hatching. It is obvious that a megapode at six weeks post- conception (about two weeks prehatching) would not be so active as a chicken six weeks postconception (three weeks posthatching). I suggest that a physiological barrier pre- vents extensive coordinated behavioral activity of the mega- pode embryo during the latter part of the prehatching period. It would be profitable, perhaps, to examine the cholinergic system in the megapode embryo (Boell, 1955: 547). In species using fermentation as a heat source for incubation the presence of aerobic bacteria should presumably greatly deplete the available oxygen supply. It has been suggested that one of the functions of mound regulation in such species is to provide oxygen for the embryos which would otherwise be under anaerobic conditions (Meyer and Stresemann, 1928: 68; Mayr, 1930: 105-106). Megapodius apparently does not open the mound during temperature regulation (Frith, 1959: 57 It would be of considerable interest for studies of develop- mental metabolism if megapodes were found to have an ex- tended period of embryonic anaerobic respiration (Boell, 1955: 522-523). Further studies on the egg tooth, plumage development, and other aspects of megapode embryology and anatomy are cur- rently in progress at this laboratory. ACKNOWLEDGEMENTS Dr. Renate Becker, Mr. H. J. Frith, and Dr. K. Westerskov have kindly given useful references. I am grateful to Mr. H. J. Brithe Dr J. 2. Trinkaus, Dr. Kh. J. Boell, Prot..G. E. Hutch- inson, and Dr. S. D. Ripley for helpful discussion of the problem. I wish especially to thank Dr. Philip S. Humphrey for initially suggesting this study and for his most valuable suggestions and comments. I am greatly appreciative of the extended efforts of Dr. E. T. Gilliard, who collected for the Yale Peabody Museum of Natural History a number of em- bryos of T'alegalla during his recent expedition to New Guinea. Miss Shirley Glaser has generously given of her time in prepa- ration of the figures. The studies reported here were conducted 6 Postilla Yale Peabody Museum No. 45 using the facilities of the Department of Zoology and Peabody Museum of Yale University. SUMMARY A vestigial egg tooth in embryos of T'alegalla is strong evidence that megapodes have evolved from gallinaceous birds which hatched relatively earlier in ontogeny. It is hypothesized that much of the maturation which occurs during the post- hatching development of typical gallinaceous birds occurs before hatching in megapodes. The vestigial egg tooth, the labial groove of the upper jaw, and vestiges of down feathers plus the state of plumage development and behavior in newly hatched megapodes support this hypothesis. Some speculations on possible unusual embryonic adaptations are presented. REFERENCES CITED Becker, R. 1959. Die Strukturanalyse der Gefiederfolgen von Megapodius freyc. reinw. und ihre Beziehung zu der Nestlingdune der Hiihnervégel. Rey. suisse Zool., 66: 411-527. Bellchambers, T. P. 1921. The Mallee Fowl of Australia. Avicult. Mag., (Ser. 13), 12 (2): 19-24. Blaszyk, P. 1935. Untersuchungen iiber die Stammesgeschichte der Vogel- schuppen und Federn und iiber die Abhingigkeit ihrer Ausbildung am Vogelfuss. Von der Funktion. I. Gegenbaur’s Morph. Jahrb., 75: 483- 521. Boell, E. J. 1955. Energy exchange and enzyme development during em- bryogenesis. Pp. 520-555. IN: Willier, B. H., P. A. Weiss, and V. Ham- burger. Analysis of development. Campbell, A J. 1903. The mound-building birds of Australia. Bird-Lore, 5: 3-8. Friedmann, H. 1931. Observations on the growth rate of the foot in the mound birds of the genus Megapodius. Proc. U. S. Nat. Mus., 80 (Art. 1): 1-4. Frith, H. J. 1956. Breeding habits in the family Megapodiidae. Ibis, 98: 620-640, Frith, H. J. 1959. Incubator birds. Scientific American, 201 (no. 2): 52-58. Hamilton, H. L. 1952. Lillie’s development of the chick. 3rd ed. Pp. xv + 1-624. Hill, J. P., and G. R. de Beer. 1950. Development of the Monotremata. VII. The development and structure of the egg-tooth and the caruncle in the Monotremes and on the occurrence of vestiges of the egg-tooth and caruncle in Marsupalia. Trans. Zool. Soc. Lond., 26: 503-544. Marshall, W. 1939. The Brush-Turkey and “turning.” Emu, 38: 489-491. Mayr, E. 1930. Beobachtungen iiber die Brutbiologie der Grossfusshiihner von Neuguinea (Megapodius, Talegallus and Aepypodius). Orn. Monats- ber., 38: 101-106. Feb. 15,1960 Embryology and Evolution of Megapodes 6 Meyer, P. O. 1930. Untersuchungen an den EKiern von Megapodius eremita. Orn. Monatsber., 38: 1-5. Meyer, P. O., and E. Stresemann. 1928. Zur Kenntnis der Entwicklung von Megapodius und Ovyura im Ei. Orn. Monatsber., 36: 65-71. Portmann, A. 1938. Beitrage zur Kenntnis der postembryonalen Entwick- lung der Vogel. Rey. suisse Zool., 45: 273-348. Portmann, A. 1951. Ontogenesetypus und Cerebralisation in der Evolution der Végel und Sauger. Rev. suisse Zool., 58: 427-434. Portmann, A. 1955. Die postembryonale Entwicklung der Voégel als Evolu- tionsproblem. Acta XI Congr. Int. Orn., 1954. Pp. 138-151. Pyeraft, W. P. 1900. A contribution towards our knowledge of the pteryl- ography of the Megapodii. Pp. 483-492. IN: A. Willey. Zoological re- sults ... New Britain, New Guinea, Loyalty Islands . .. Part IV. Cambridge at the Univ. Press. Pycraft, W. P. 1907. On the origin of differences between nestling birds. Proc. Fourth Internat. Orn. Congr. Pp. 454-459. Pycraft, W. P. 1910. A history of birds. Pp. i-xxxi + 1-450. Studer, T. 1877. Ueber die Bildung der Federn bei dem Goldhaarpinguin und Megapodius. Verhandl. der schweiz. naturforsch. Gesellsch., 60: 240-246, Studer, T. 1878. Beitrage zur Enitwicklungsgeschichte der Feder. Zeitschr. wiss. Zool., 30: 421-436. Studer, T. 1889. Embryonalentwicklung der Vogel. IN: Die Forschungs- reise S. M. S. “Gazelle” in den Jahren 1874 bis 1876, Vol. III: 107-124. Umanski, E. 1926. Zur Frage iiber den Einfluss der vertikalen Lage des Eies auf die Entwicklung des Embryos von Gallus domesticus. Zool. Anz., 68: 81-87. Wallace, A. R. 1860. The ornithology of northern Celebes. Ibis, 2: 140-147. Westerskov, kK. 1957. Growth and moult of pheasant chicks. New Zealand Dept. Int. Affairs Wildl. Publ. No. 74: 64 pp. ‘ ae 4 ; t *@ hay eee #8 A= 1) 4 a - i; F ims) 9 “ J Vas * 7 ie af “2 = dy } o ‘ Ry. i é on & - “ * y 4 - - * “ss a J ’ “i , ° ¥ . “dee « <— =. * ia? i ’ i¢ - ° -~ 4 - | p ’ - [ F “ “4 ~eRe> a oe 4 7 rE be 7 an 7 = = @ . i » é MUS. COMP. 700L LiSihas OCT 18 1960 HARVARD UNIVERSITY pate te YALE PEABODY MUSEUM oF Natura. Hisrory Number 46 June 30, 1960 New Haven, Conn. FOSSIL AMPHIBIANS FROM QUARRY NINE by Max K. Hecur! anp Ricuwarp Estes” The original intention of this investigation was to determine the identity of Hobatrachus agilis Marsh. It was soon evident to us, as to other workers, that the type materials represented more than one species. Fragments referred to this form by Moodie (1912, 1914) represent an ilium of a reptile, a femur of a salamander, an unidentifiable fragment of a tibiofibula of a frog and two distinctly different types of frog humeri. Unavailable to us at this time are the vertebra and urostyle illustrated but not discussed by Moodie (1914). Marsh (1887) described this form in the following words: ‘More recently, various bones of small, anourous amphibians (Eobatrachus agilis) have been found, the first detected in any Mesozoic formation.” Moodie (1912) described Marsh’s material and selected the larger humerus as the type (Yale Peabody Mu- seum no. 1862). He stated that the elements represented a form close to Bufo and later (1914) actually placed it in the Bufonidae. Simpson (1926 a and b) merely records the pres- ence of a modern frog in the fauna.” The importance of these specimens is that the frog remains are among the oldest known and the salamander is the earliest record of that group. Appli- 1 Dept. of Biology, Queens College, Flushing 67, New York. 2 Dept. of Paleontology, University of California, Berkeley 4, California. 3 Reig (1957) for unknown reasons referred Hobatrachus to the Discoglos- sidae. 2 Postilla Yale Peabody Museum No. 46 cation of names to such fragmentary material is in part a matter of taste, but the antiquity of the material and its close correspondence to modern forms make it useful to place the material within the established system of classification. The senior author is responsible for the sections dealing with anurans; the junior author for the remainder of the specimens. Class Amphibia Superorder Salientia Order Anura Suborder Aglossa? Family incertae sedis Eobatrachus agilis Marsh latest, hes. 13,75 Holotype: Yale Peabody Museum no. 1862, the distal por- tion of a humerus. Locality: Quarry 9, Como Bluff, Wyoming. Diagnosis: Distinguished from all known frog humeri by the following combination of characters: A) base of the shaft of the humerus perpendicular to the main axis of the humeral ball (eminentia capitata of Gaupp, 1894, henceforth referred to as the ball), B) a deep triangular fossa present (fossa cubitus ventralis) at the upper end of the ball, C) the ball a fully developed spherical articulating surface which is proportion- ately large in size, D) a small olecranon scar which is nearly triangular in form but with its apex nearest the lateral border of the humerus, E) weakly developed epicondyles, the medial epicondyle larger than the lateral epicondyle but reduced in size as compared to other frogs, F) narrowest cross-section (or neck) of the humerus is just above the ball. Description: A broken distal portion of a right frog hu- merus measuring 6 mm. in length, On its distal portion is a completely rounded but abraded ball, with a diameter of 2 mm. The medial epicondyle is a small slightly abraded nubbin medial to the ball. On the opposite side of the lateral epicondyle is a slight ridge with no evident rise or mound. From the two epicondyles, two distinct ridges run proximally on the main JRA Ly ae oe ¥ MR Fi) T 18 1960 June 30, 1960 Fossil Amphibians from Quarry Ni Har Ban yy hid iy shaft of the humerus. Lying between the two ridges is a itiely fossa (fossa cubitus ventralis) which is roughly triangular. The base of the triangular pit is formed by the ball and its deepest area is on the medial side above the ball. It gradually becomes shallower both proximally and laterally. The apex of the triangular fossa is rounded and lies midway between the two epicondyles. The lateral surface of the medial epicondyle forms a weak flange which projects slightly medially. The olecranon scar, on the posterior surface of the ball, is a small triangular area whose apex is the same height as the ball and lies miday between the two epicondyles. The neck of the hu- merus (the area of smallest cross-section) is apparently long and begins far above the ball. There are no indications of a ventral ridge or crest on the neck of the humerus. Comparisons of the fossil with living frogs are based on the following gen- era: (Unless otherwise indicated only one species of each genus has been examined. ) Leiopelma, Ascaphus (Leiopelmatidae) ; Pipa, Xenopus (Pi- pidae) ; Discoglossus, Barbourula, Bombina, Alytes (Discoglos- sidae); Rhinophrynus (Rhinophrynidae), Pelobates (2. spe- cies), Scaphiopus (3 species), Megophrys (38 species) (Peloba- tidae) ; Pelodytes (Pelodytidae) ; Leptodactylus (2 species) ; Batrachophryne, Calyptocephalella, Eupsophus, Physolaema, Telmatobius, Ceratophrys, Eleutherodactylus, Pleurodema, Adelotus, Kyarranus, Limnodynastes, Lechriodus, Helioporus, Rhinoderma (Leptodactylidae); Dendrobates (2 species), (Dendrobatidae) ; Atelopus (Atelopodidae); Bufo (25 spe- cies), Ansonia, Nectophrynoides (Bufonidae); Hyla (10 spe- cies), Acris, Gastrotheca (2 species), Diaglena, Smilisca (8 species) (Hylidae); Pseudis (Pseudidae); Rana (5 species), Arthroleptides (Ranidae); Phrynomerus (Phrynomeridae) ; Astylosternus, Dyscophus, Probreviceps, Kaloula, Uperodon, Gastrophryne (Microhylidae) ; Hyperolius, Rhacophorus, Me- galivalus (Hyperoliidae). Discussion: The humerus of anurans is one of the most easily identifiable structures because of the presence of the prominent ball on the distal end. The basic morphology of the humerus is discussed by Gaupp (1894) and the terminology to be fol- lowed will be based on this work. Unfortunately this classic 4. Postilla Yale Peabody Museum No. 46 study is based only on members of the genus Rana and there- fore many described features of the humerus are characteristic only of that family or even that genus. The main aspects of the morphology of the humerus are amply illustrated in Fig- ures 39-41 of this work. The discussion will be restricted to the distal portion of the humerus. On either side are two epi- condyles, the lateral and the medial. In most frogs the medial is larger and more prominent, whereas the lateral epicondyle is usually small or represented by a slight nubbin. Immediately above the ball there may be a slight or relatively deep depres- sion, the fossa cubitis ventralis (Gaupp 1894, hereafter re- ferred to as the fossa). On the posterior surface of the humeral ball there is almost always a roughened triangular area which will be called the olecranon scar. This represents the area which articulates with the olecranon process of the radio-ulna. Im- mediately above the widened distal end of the humerus, there usually is a neck region which generally has the narrowest diam- eter of the entire humerus. On the proximal end of the humerus of almost all frogs there is a crista ventralis (Gaupp 1894). In many frogs this ridge is quite long and extends onto the neck of the humerus but usually it is absent on the neck region. On the basis of morphology the humerus of Ascaphus and Leiopelma have much in common. There is a distinct fossa and reduced epicondyles in Letopelma. Eobatrachus can read- ily be distinguished from Leiopelma by the more advanced structure of the ball. Ascaphus has a modern type of ball but the fossa is very small and shallow. The nature of the fossa and the expanded lateral and medial epicondyles and their flanges distinguish it readily from Eobatrachus. The Pipidae is characterized by a small but well developed ball, with equally developed epicondyles and a deep triangular fossa. The symmetry of the pipid fossa is much greater than that of Eobatrachus but the fossa is relatively better developed than in any other known living or fossil frog. The ball of EHoba- trachus is much more advanced than either genus although the reduction in size in the pipids may be due to aquatic adaptation and reduction of jumping abilities. The Discoglossids are pre- cluded from relationship to Hobatrachus by the lack of the fossa. Other features are characteristic of the Discoglossidae June 30, 1960 Fossil Amphibians from Quarry Nine 5 which eliminate them from further discussion. The Pelobatidae and the Pelodytidae can be eliminated because there is no sign of the fossa (except a tiny fossa-like depression in Megophrys ) and the apex of the olecranon scar tends to le laterally rather than medially. The condition of the humerus among the lepto- dactyloid frogs (including Leptodactylidae, Dendrobatidae, Atelopodidae, Rhinodermidae following Griffith, 1959) is most variable with the single exception that the fossa is never pres- ent except weakly in Batrachophrynus, which differs from Eobatrachus by the presence of a low ventral crest on the neck region (crista ventralis) and reduced medial epicondyle in the living species. The Pseudidae and Centrolenidae may be dif- ferentiated from Kobatrachus in the same manner as the other leptodactyloid families. The bufonids can be easily distin- guished by the complete lack of the fossa, the generally curved humerus and by the apex of the olecranon scar being more laterally than medially oriented. The distal portion of the humerus of Hylidae is variable, but is usually characterized by the complete lack of a fossa or at best a lunate deep trench just above the proximal border of the ball. The medial epicon- dyle is usually moderately or weakly developed and the lateral epicondyle is variable in size from very small to very large. The ranid humerus can be distinguished from the fossil by the small pit-like fossa which hes just above the ball, very prominent medial epicondyle, laterally oriented apex of the olecranon scar, and by the general curvature of the humerus. The phrynomerid humerus is distinguished by its very small ball, elongate diaphysis, and reduced olecranon scar. The fossa in this form is very shallow, triangular, and extremely short. Both the Ranidae and Hyperolidae have a deep fossa just above the proximal end of the ball. This fossa is distinctly different in its form from those of Kobatrachus. It appears that in both of these families the depression may merely be formed by enlargement of the sphere-like pattern of the ball. Both these families also differ from the fossil by the great development of lateral extensions or flanges from the epicon- dyles, the relatively large size of the medial epicondyle and the lateral position of the olecranon scar. From the above discussion it appears that there is no clear 6 Postilla Yale Peabody Museum No. 46 relationship between Kobatrachus and any of the living fam- ihes of frogs. The large size of the ball, the development of the fossa, the reduced medial epicondyle, the shape and form of the olecranon scar and the perpendicular position of the humeral shaft all indicate a unique association of characters not found in any living or fossil frog seen. The only frogs which approach Eobatrachus as far as the development of the fossa 1s concerned are the Pipidae and perhaps Leiopelma. In all of these the fossa is a symmetrical trough which is not the case in Kobatrachus. In both Xenopus and Pipa the hu- meral ball is very small with relatively large epicondyles, whereas in Kobatrachus the humeral ball is very large and the epicondyles are reduced. Certainly as far as the ball is concerned the humerus is an advanced structure but the de- velopment of the fossa may indicate a more primitive con- dition. The assignment of Eobatrachus to Montsechobatrachi- dae is at best a guess and perhaps it should be considered a more advanced frog than that. Validity of the assignment of Eobatrachus to this family (Romer 1945) cannot be deter- mined from the published material of Montsechobatrachus. Superorder Salientia Order Anura Suborder Neobatrachia Family incertae sedis Comobatrachus aenigmatis, new genus and species Plate 1, figs. 2, 4, 6 Holotype: Yale Peabody Museum No. 1863, the distal por- tion of a frog humerus. Locality: Quarry 9, Como Bluff, Wyoming. Diagnosis: Distinguished from Eobatrachus by its shallower, symmetrical triangular fossa cubitus ventralis and less devel- oped medial epicondyle; similar to some leptodactylid, micro- hylid and hyperolid frogs in the presence of the fossa cubitus ventralis, but distinguished from these groups by the poorly developed medial epicondyle, the medial position of the apex of the olecranon scar and straight shaft of the base of the humerus. June 30, 1960 Fossil Amphibians from Quarry Nine 7 Description: A broken distal portion of a right frog humerus measuring 5 mm. long. At the distal end of the fragment there is a large distinct abraded ball (eminentia capitata) which has a diameter of approximately 1.3 mm. On the medial side there is a small indistinct slightly abraded medial epicondyle and on the opposite side there is no distinct evidence of a lateral epicondyle. The surface of the area of the lateral epicondyle is slightly abraded. The area of each epicondyle forms slight rounded ridges which meet at the base of the neck. Between the two ridges is a fossa the shape of an isosceles triangle whose base is the upper end of the humeral ball. The fossa is shallow; the deepest area being at the upper border of the humeral ball. Posteriorly, the olecranon scar is triangular in form and its apex is slightly higher than the humeral ball. The apex lies midway between each epicondyle. The neck of the humerus is relatively low and begins above the expanded distal end of the bone. Discussion: The relationships of Comobatrachus are appar- ently with the more modern frog families. The development of the ball and the general shape of the fossa indicate no relation- ship to Leiopelmatidae, Pipidae, Discoglossidae, or Pelobati- dae. Among the Neobatrachia the Bufonidae, Atelopodidae, Dendrobatidae, Pseudidae (and other groups now placed in the Leptodactylidae by Griffith, 1959) and Hylidae are pre- cluded from consideration by either the complete lack of a fossa or only the slightest indication of such a structure. The fossa of the Ranidae is merely a lunate cleft above the humeral ball. Among the Hyperolids there is no fossa in Rhacophorus or Megalivalus but a distinct one in Hyperolius. The base of the humerus of Comobatrachus bears a distinct resemblance to Eupsophus (Leptodactylidae), Hyperolius (Hyperoliidae), Probreviceps and Kaloula (Microhylidae). There are distinct differences between the aforementioned modern frogs and Co- mobatrachus. In all the modern frogs the medial epicondyle is better developed and the fossa is distinctly shorter than in the fossil. As a result of these comparisons there is apparently no family of living frogs to which Comobatrachus can be assigned, though it appears to be a member of the more advanced fam- ilies of the Neobatrachia (Reig 1958). It is probable that the 8 Postilla Yale Peabody Museum No. 46 medial epicondyle has been eroded or broken away and if so the humerus would perhaps conform more closely to one of the above genera. Assuming that the epicondyle has not been too badly damaged, it would appear that no family of living frogs would include the features of Comobatrachus. Therefore we can only conclude that it represents one of the more advanced families, possibly something related to the more gen- eralized Leptodactylidae or perhaps a family as advanced as the Microhylidae or Hyperoliidae. On the basis of probability a leptodactyloid affinity appears more likely. Order Urodela Family incertae sedis Comonecturoides marshi, gen. et. sp. nov. Plate 2, figs. 3, 4; Plate 3, fig. 6 Holotype: Yale Peabody Museum 3919, complete right femur. Type locality: Quarry 9, Como Bluff, Wyoming. Diagnosis: Distinguished from living salamanders princi- pally by the presence of endochondral ossification and heavier ossification of the perichondral diaphysis. Description: The femur is characteristically urodele, with narrow diaphysis, expanded and unossified proximodistal ex- tremities, and tiny, anteroventral twiglike trochanter. The head in cross section is rounded dorsally, and has a slight ventro-posterior angle. The tip of the trochanter is missing, and the point of separation of shaft and trochanter is about one millimeter distal to the preserved proximal edge of the head. The trochanter is continued on the diaphysis by a crest which diminishes distally, but remains discrete almost to the pre- served distal edge of the bone. The dorsal surface of the distal end is swollen and pitted for liigamentary attachment. Ven- trally the distal end bears two tiny foramina. The outline of the distal end is oval, slightly concave ventrally and convex dorsally. In cross section, the bone of the shaft is quite thick and there is endochondral ossification within the expanded extremities. Maximum length of femur, 11.5 mm.; maximum June 30, 1960 Fossil Amphibians from Quarry Nine 9 diameter of distal extremity, 38 mm.; maximum diameter prox- imal end, 2.8 mm. Discussion: Femora and humeri of urodeles may be distin- guished easily by the following characters. In cross section, the distal end of the femur is always convex dorsally and con- cave ventrally; both dorsal and ventral edges are convex in humeri, giving a lobate appearance. The humeri always possess a strong bladelike crest ventrally, continuous with the head, and a smaller trochanter is sometimes present dorsally (e.g. in Salamandridae, see Francis 1934, pl. 7, fig. 42). Femora of living families of urodeles are quite distinct, particularly with respect to the outline of the head in cross-section, and to a lesser degree the shape and orientation of the trochanter. The outline of the distal extremity is less characteristic but may also be helpful. Plate 3 shows outlines of femoral heads of all families (except Sirenidae which lack hind limbs) of living urodeles. Each drawing is based on several specimens and is intended to reflect the characteristic shape for each family rather than that of any particular individual. The following material was seen: (numbers in parentheses indicate specimens examined ) Ambystomidae Salamandridae Ambystoma tigrinum (3) Salamandra atra (2) A. opacum (1) S. maculosa (1) Rhyacotriton olympicus (1) Mertensiella caucasica (1) Siredon mexicanum (1) T'aricha granulosa (1) Hynobidae Amphiumidae Hynobius stejnegeri (1) Amphiuma tridactylum (3) Batrachuperus pinchonii A. means (1) (1) Proteidae Cryptobranchidae Proteus anguinus (1) Andrias scheuchzeri japonicus (4) Necturidae Cryptobranchus Necturus maculosus (4) allegheniensis (1) N. punctatus (1) N. beyeri (1) 10 Postilla Yale Peabody Museum No. 46 Plethodontidae Plethodon cinereus (2) Leurognathus marmorata (1) Desmognathus fuscus (2) Pseudotriton ruber (1) The shape of the head of the femur was found to be rela- tively constant in all families except Ambystomidae. Rhyaco- triton resembles Ambystoma, both differ from Dicamptodon ensatus. The proximal ends of femur and humerus are difficult to distinguish in Siredon mexicanum, probably due to lack of ossification. Comparison of the figures will show that the closest resemblance to the Como Bluff specimen is with Nec- turus (considered here as a family separate from Proteus, following Hecht 1957). There is some similarity to Amphiuma, from which it is distinguished by the less sloping posterior border of the head and the slightly different angulation of the trochanter. Characters of the shaft, trochanter, and distal end are shown in Plate 2. Ambystomids have a relatively di- vergent trochanter, often connected proximodistally to the shaft by thin crests or webs of bone. The short stubby femur of the cryptobranchids with its blunt trochanter is easily recognizable, and the outline of the distal end is especially characteristic. Salamandrids often have ossified extremities and the trochanter is faleate with a rounded excavation between trochanter and head. This condition is also found in pletho- dontids, though they may be separated by the proximal outline of the head. Proteus has a very reduced femur, with only a tiny ridge instead of a trochanter. Necturids are characterized principally by the rounded outline of the femoral head, which lacks any prominent crests or angles, and in this respect Comonecturoides most closely resembles this family. Compar- ison with Necturus beyert and especially N. punctatus was difficult due to reduced ossification in limb extremities of these forms. Both of these species show a little more anteroposterior compression of the head of the femur than does N. maculosus, but this is in part due to lack of ossification in the most prox- imal part of the shaft. In perennibranchiate or larval types June 30, 1960 Fossil Amphibians from Quarry Nine ilk only the larger specimens or species are well ossified enough for comarison. Interrelationships of the various families based on the out- line of the femoral head are as follows. The similarity is greatest between hynobiids and Ambystoma, to be expected due to the close relationship between the two groups. The salamandrid outline is easily derived from this type as is the plethodontid. The necturid outline is probably closer to the hynobiid or perhaps the salamandrid than to any of the others (the latter relationship suggested by Noble (1981) on the basis of reproductive structures) and the similarity of Am- phiuma (probably a salamandroid derivative) to Necturus may strengthen this suggestion, though of course no particular weight may be placed on this single character. The stubby outline of the cryptobranchid femur is unlike any other. Class Reptila Order Sauria? Plate 1, figs. ‘7, 8 Yale Peabody Museum 1568.—right ilium. Description: The ilium is a flattened blade, smooth medially, with no indication of a sacrel attachment. Dorsally and ven- trally there are crests developed, giving a lenticular cross- section. Posteriorly these crests disappear and the cross-sec- tion is circular at the tip. Anteriorly there is a prominent crest with a boss laterally for muscle attachment. The acetab- ular area is broken ventrally and no trace of attachment for pubis or ischium remains. A tiny part of the acetabulum is present. Discussion: This bone was first discussed by Moodie (1912), p. 287) who indicated that it was “‘quite peculiar” and would “possibly be sufficiently characteristic to sustain the validity of Professor Marsh’s genus Eobatrachus.” Later (1914, p. 533) he indicated that there were four pits on the articular surface marking the “‘synchondrosteal union of the halves of the pelvis.”” These pits are breakage surfaces; no evidence of the actual articular surface remains. Reference of this bone to the Reptilia indicates that it must be the right ilium with the narrow tip pointing posteriorly, rather than the left bone with 12 Postilla Yale Peabody Museum No. +6 anteriorly pointing tip as Moodie suggested. There is a super- ficial resemblance to raniform frogs, principally due to the size of the dorsal crest, but anuran ilia in general lack the ventral crest and are relatively much longer than this specimen. The short, compressed bladelike shape most closely resembles that of the Sauria. Ila of all families of lizards have been examined, as well as those of other recent and many fossil reptiles. The general shape of the bone conforms most closely among lizards to certain gekkonids (e.g. Thecadactylus) but the latter differ in the less well developed dorsal muscular crest. Breakage of the acetabular region renders further com- parison fruitless; the primary reason for discussion of the bone here is indication of its reptilian nature. SUMMARY AND CONCLUSIONS The type materials of the earliest known North American fossil frog Eobatrachus agilis Marsh are redescribed. The holo- type of E. agilis is a right humerus and the genus is tentatively referred to the Aglossa (Reig 1958). No comparison is pos- sible at this time with Montsechobatrachus and family refer- ence is left open. The other anuran humerus associated with the type is distinctly different and is made the type of a new genus and species Comobatrachus aenigmatis which is referred to the Neobatrachia (Reig, ibid) without family assignment, though it is possible that it is of leptodactyloid relationships. The associated femur is identified as a urodele, incertae sedis, described as Comonecturoides marshi and a similarity to Nec- turidae noted. The associated ilium is not anuran and is prob- ably that of a lizard or closely related reptile. The distinctive characters of frog humeri and urodele femora are discussed. Mook (1918) characterized the environment of the Morrison formation as a broad floodplain with abundant running water. Wieland (1925) suggested a temperate to cool climate, while Simpson (1933) favored a warm to tropical climate. Salaman- ders are primarily North Temperate today and seek cooler, moister habitats. This may indicate a temperate to warm temperate rather than a tropical environment during Morri- son time. June 30, 1960 Fossil Amphibians from Quarry Nine 13 ACKNOWLEDGEMENTS We are most grateful to Dr. J. T. Gregory for allowing us to study the Como Bluff specimens, to Dr. E. H. Colbert for generously providing the facilities of the American Museum of Natural History, and to Mr. C. M. Bogert and Dr. Richard Zweitel, Department of Reptiles and Amphibians, American Museum of Natural History, and Dr. Ernest Willams, Mu- seum of Comparative Zoology, Harvard College, for allowing us free access to comparative material in their collections. The authors wish to thank Drs. Maxim Lamotte, Marvin Wasser- man, Theodore Cohn, and Eric Matthews for gifts of com- parative materials. Research time for the senior author was made possible by National Science Foundation Research Grant and American Philosophical Society Grant which are grate- fully acknowledged. Special thanks are due Chester Tarka for his excellent photographs of the specimens. REFERENCES EKaton, Theodore H. Jr. 1958. An Anatomical Study of a Neotropical Tree Frog, Centrolene prosoblepon (Salientia: Centrolenidae). Univ. Kans. Sci. Bull. Vol. XX XIX, no. 10, pp. 459-472, figs. 1-20. Francis, Eric T. B. 1934. The Anatomy of the Salamander. Oxford, Claren- don Press. xxxi + 381 p., 3 figs., 25 pls. Gaupp, Ernst. 1896. Ecker’s und Wiedersheim’s Anatomie des Frosches. Braunschweig, Viewig und Sohn, vol. 1 and 2, 548 p., 146 figs. Griffith, I. 1959. The phylogeny of Sminthillus limbatus and the status of the Brachycephalidae (Amphibia, Salientia). Proce. Zool. Soc. London, vol. 132, part 3, pp. 457-487, figs. 1-18, pls. 4. Hecht, Max K. 1957. A case of parallel evolution in salamanders. Proc. Zool. Soc., Caleutta, Mookerjee Memor. vol., pp. 283-292. Moodie, Roy L. 1912. An American Jurassic Frog. Amer. Jour. Sci. (ser. 4), vol. XXXIV, pp. 286-288. ————. 1914. The Fossil Frogs of North America. Amer. Jour. Sci. (ser. 4), vol. XX XVIII, pp. 531-526, figs. 1-2 Mook, C. C. 1918. The habitat of the sauropod dinosaurs. Jour. Geol., vol. XXVI, pp. 459-470. Noble, G Kk. 1931. The Biology of the Amphibia. London and New York: McGraw-Hill, xii + 577 pp., 174 figs., map. Reig. Osvaldo A. 1958. Proposciones para una nueva macrosistematica de los anuros. Physis, vol. XXI, pp. 109-118. Romer, Alfred S. 1945. Vertebrate Paleontology. Univ. Chicago Press, 687 p., 377 figs., 4 tables. Simpson, G. G. 1926. The fauna of Quarry 9. Amer. Jour. Sci. (5th Ser.), vol. XII, pp. 1-11, 1 fig. 1933. Paleobiology of Jurassic Mammals. Palaeobiologica, V Band, pp. 127-158, 6 figs. Wieland, G. R. 1925. Dinosaur feed. Science (n.s.), vol. 61, pp. 601-603. 14 Postilla Yale Peabody Museum No. 46 PATE 1 Eobatrachus agilis Marsh, type specimen, YPM 1862 Fig. 1. Ventral view of right humerus Fig. 3. Dorsal view of right humerus Fig. 5. Medial view of right humerus Comobatrachus aenigmatis, n. gen., n. sp., type specimen, YPM 1863 Fig. 2. Ventral view of right humerus Fig. 4. Dorsal view of right humerus Fig. 6. Medial view of right humerus Unknown reptile, YPM 1568 Fig. 7. Lateral view of right ilium Fig. 8. Medial view of right ilium [Present magnification x 10] June 30, 1960 Fossil Amphibians from Quarry Nine 15 16 Postilla Yale Peabody Museum No. 46 PLATE 2. Eobatrachus agilis Marsh, type specimen, YPM 1862 Fig. 1. Lateral view of right humerus Comobatrachus aenigmatis, n. gen., n. sp., type specimen, YPM 1863 Fig. 2. Lateral view of right humerus Comonecturoides marshi, n. gen., n. sp., type specimen, YPM 3919 Fig. 3. Dorsal view of right femur Fig. 4. Ventral view of right femur Unidentified anuran, YPM 1394 (Part of original type of Hobatrachus agilis) Fig. 5. Dorsal view of distal end of tibiofibula Fig. 6. Ventral view of distal end of tibiofibula Present magnification x 10] £ 4 - ry Nine June 30, 1960 Fossil Amphibians from Quar 18 Postilla Yale Peabody Museum No. 46 PLATE 3. Comparative series of urodele femora. Above: outline of right femur, an- terodorsal view. Below: outline of left femoral head in section; the dorsal surface up and the anterior surface to the right. Not to scale. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 9. Me 2 Plethodontidae . Salamandridae Proteidae . Ambystomatidae 1. dAimbystoma 2. Dicamptodon Hynobiidae . Comonecturoides marshi, n. gen., n. sp . Necturidae . Cryptobranchidae Amphiumidae June 30, 1960 — Fossil Amphibians from Quarry Nine 19 MUS. COMP. Z00L OCT 18 1960, HARVARD UNIVERSITY Wostilla YALE PEABODY MUSEUM or NaTuRAL History Number 47 July 1, 1960 New Haven, Conn. ADDITIONS TO THE AVIFAUNA OF NORTHERN ANGOLA I. S. Ditton RipLteEy anp Gerp H. HeErnricH During the course of a year’s visit to Angola for the Peabody Museum in 1957-58 by one of us (Heinrich) a number of interesting new records were made. Specimens have been critically compared in the Peabody Museum by one of us (Ripley) and the advice of Dr. James P. Chapin is here gratefully acknowledged. The accompanying map shows the course of Mr. Heinrich’s travels. Grateful acknowledgement 1s made to the Diamond Company and its Museum for their wonderful hospitality. Hieraaétus africanus (Cassin) A male was shot in a coffee plantation at Rocga Canzele September 26, 1957 by Prince Gustav von Schoenaich Carolath and presented to the collection. Guttera edouardi schoutedent Chapin The occurrence of this form of the blue-spotted guinea fowl in Angola is hereby noted with the presentation to Yale of a female taken by the chief of the post office at Camissombo, September 17, 1958. Numida meleagris marungensis Schalow Cacolo, Somba and Dundo in dry savannah woods. 2 Postilla Yale Peabody Museum No. 47 COLLECTING LOCALITIES OF GERD HEINRICH IN NORTH ANGOLA -1957-58 BELGIAN CONGO Portugalia Bindos eae Luact imo RF. \ \ Roca Canzele ‘ Vila Verissimo Sarmento e \ 1 | Chicapa alls Somba (=Sombo ) Quiculungo® Falls Duque de J ( Braganca ete la Fenrigile de Carvalho (= Saurimo) oo i al { © Vila Luzo ® Silva Porto Nova Lisboa SS | RHODESIA 1 COMES ANGOLA ; +—— 100 kilometers Turnix nana insolata, new subspecies Type; ¢ ad. (Y.P.M. No. 50,087) collected by Gerd Heinrich February 7, 1958, 35 km. west of Camissombo, north- east Angola. Diagnosis: from T'urnixa nana this form, known from two males, differs in color so strikingly as to suggest a new species. Turnix nana is distinguished from the contiguous species sylvatica by triflingly larger size, by the lack of a pronounced median crown stripe and by the presence of blackish bars on the neck and upper chest which may extend incompletely across the fore-neck in males. This subspecies differs from typical nana by being extremely pale, bleached, lacking the dark shading of the crown and solid blackish color of the lower back and rump. In this form the general coloration above is grayish with reduced black patches on the feathers, highly cross-hatched with brown. There is an indistinct median crown stripe, slightly more pronounced than in typical nana, The MUS. COMP. Z00L LIBKANY OCT P8 1960 wing coverts are whitish with a sandy-buff tone. THe ureR!ARD parts are whitish with an incomplete buffy wash on the UOIMERSITY neck and obsolete bars on the sides of the neck. In Size this form is similar. Two males have wing measurements of 80.5 (type), 79; tail 30; culmen 11, 10; weight 46, 44.5 g. Soft parts: iris pale yellow with inner and outer brown rings: up- July 1, 1960 Avifauna of Northern Angola I. per mandible blackish, lower whitish with dark tip; feet pinkish ivory. Remarks: these birds were found on a dry, sandy, grassy plateau at 1100 metres altitude. Specimens loaned to us through the courtesy of the British Museum (Natural His- tory) and the Chicago Museum of Natural History, come from Ndala Tando in the Benguella area of Angola, from northern Rhodesia and from the Kasai. All agree in being similarly and typically dark. Thus this bleached form occupies an isolated savannah area in the center of the range of normal nana. Ewcalfactoria adansonu (Verreaux) A male was taken 25 km. northwest of Nova Gaia in a grassy meadow near a brook December 5, 1957. Sarothrura bohmi béhmi Reichenow The bird collection of the zoological laboratory of the Diamond Company in Dundo contains a specimen of this species, a male, collected November 21, 1957, in the Luachimo valley near Dundo by a native. Length of wing 88 mm., of tarsus 238 mm. Stephanibywx lugubris (Lesson) A single specimen was collected on the roadside between Dundo and Calulo on the northern side of the Cuanza river. Charadrius tricollaris tricollaris Vieillot Taken at Luanda August 1, 5, 1957. Larus fuscus fuscus Linnaeus An immature female from near Luanda, taken August 7, 1957, is new for Angola. 4 Postilla Yale Peabody Museum No. 47 Columba unicincta Cassin Luachimo River, near Dundo, 800 m. altitude; Roca Canzele, north of Quiculungo, 600 m. altitude. This handsome pigeon lives in mature stands of tropical rain forest and tropical gal- lery woods, keeping always to the crowns of the oldest, tallest trees, usually about 100 feet above the ground. Not rare in the coffee woods north of Quiculungo, but difficult to observe and to collect on account of its dwelling above the normal reach of eye and gun. Cuculus canorus gularis Stephens South of Duque de Braganga, near the road to Matete, in a savannah parkwood: single, medium-sized trees and patches of low bushes alternating with open, grassy spaces. This species seems to be very rare in Angola. During two and a half years of work it has been met with only once. When the call of the male was heard for the first time, it seemed to be the call of a dove or a barbet, rather than that of a cuckoo. It has no similarity with the familiar voice of the European C. canorus, being much softer, deeper, more muffled and accentu- ated on the second syllable instead of on the first. The calling begins usually with a few monosyllabic notes like “tuk .... uk .uk....” and then changes over to a fairly short series of duosyllabic calls, each accentuated in the second syllable. The whole song thus sounds like this: “uk ....uk....uk232- ukuk .... ukuk . ukuk . ukuk.” The calling male is perched, well hidden, in the crown of one of the medium sized trees of its biotope, as described above. From a pair, chasing each other, I heard a sharp, ringing call like: “pit pit pit pit pit,” fairly similar to the voice of the female of the European C. canorus. Centropus grill grill Hartlaub Forty km. northeast of Duque de Braganca, 1250 m.;25 km. A | Sane northwest of Nova Gaia, 1250 m. altitude; Kassai River, 40 km. northeast of Canza, 900 m. altitude in open, wide, marshy, flat and treeless stream valleys of the high plateau in stands of tall grass in the drier areas. July 1, 1960 Avifauna of Northern Angola I. 5 In both males collected in December one of the testes was maximally enlarged, the other not at all. In April, in the Kas- sal valley, the species was still in full breeding condition and here, for the first time, the opportunity was found to study its voice which is rather different from the other species of the genus. The call was heard only during the early morning hours, but then many times and from different directions. It is two-syllabic sounding approximately like “‘julup”, dull in sound, but nevertheless loud and audible from afar. This call is usually repeated two to six times in succession and often followed by a short series of monosyllabic sounds, like ‘‘du du - du - du.” These notes are deeper than the main call, more muffled and not so far carrying. One morning Heinrich stalked a bird that was continuously calling in a thicket of 12-foot tall grass. It was perched on a branch of an over- grown, dead bush, some 3 yards above ground, almost un- interruptedly calling as described above. The specimen turned out to be not a male but a female with maximally enlarged ovaries. A few minutes later, and about 40 yards ahead Heinrich flushed and shot the male. It too, had maximally enlarged gonads. But there is not the slightest doubt that the ‘aller was indeed the female, as Heinrich was close enough to observe the movements of the head synchronized with the calls. Myioceyx lecontet lecontet (Cassin) A female was taken at Roca Canzele in heavy forest along a brook on October 1, 1957. Berenicornis albocristatus cassini (Finsch) A specimen taken on the Rio Luachimo near Dundo on May 11, 1958, and another specimen, a juvenile, in the museum at Dundo presumably from Lunda province taken in 1946, extend the range of this form into northeastern Angola. Tricholaema hirsutum chapint Bannerman A male and female were collected near Dundo on the Rio Luachimo in February and May. 6 Postilla Yale Peabody Museum No. 47 Macronyx grimwoodi Benson In the region of Lake Carumbo at 900 m. altitude, one specimen was found on March 20th, in a marshy meadow along a stream. The biotope was not at all different from the usual habitat of M. fiillerborni, several specimens of which were in the same meadow. A careful examination of the whole area showed that the single specimen of M. grimwoodi was a solitary one. Macrosphenus concolor (Hartlaub) The olive longbill was found on the Luachimo river, near Dundo, 800 m. altitude and on the Kassai river, 40 km. north- east Canza at 900 m. altitude. Within the tropical gallery wood this skulking little bird has chosen a rather specialized habitat. It is rarely found in the thickets of the lower floor. Instead it searches the very densest tangles of hanging vines, the webs and veils of lanas which here and there wrap the trunk of crown of a single tree, sometimes creating an almost solid mass of clustered vegetation. Under this excellent cover the small birds would be in- visible if their lively actions and characteristic voice did not betray them. There are two slightly different versions of the warbling song, both, however, equal in timbre and in their somewhat intrusive, eager, hurried delivery. The basic tune of the first version is four-syllabic, and sounds like “tutuhuo” with the strong accent on the penultimate syllable, which is somewhat higher than the others. This four-syllabic tune 1s repeated many times in rapid sequence without the slightest intermission: “tutuhuotutuhuotutuhuotutuhuotutuhuo is The other version is only three-syllabic with the accent on the first syllable: “huititi.” It too is repeated rapidly and eagerly many times: huititihuititihuititihuititihuititi . . . .” In both versions the volume and the speed of the delivery increases slightly and gradually toward the end of the sequence. Heinrich observed the species always in pairs. Sylvietta denti denti (Ogilvie-Grant ) The single collected specimen, a female with slightly en- larged ovaries, was shot May 11th from the crown of a tall July 1, 1960 Avifauna of Northern Angola I. 2 tree in tropical gallery woods, close to the border of the Luachimo river, near Dundo; altitude 800 m. Apalis goslingt goslingi Alexander Found on the Luachimo river, near Dundo, 800 m. altitude in tropical gallery wood. This species keeps always to the im- mediate neighborhood of the river, usually searching the foliage of branches hanging directly over the water or of bushes standing on little islands in the river. The song is easy to distinguish from that of 4. rufogularis and A. alticola, as its basic strophe is monosyllabic; it is re- peated about seven to twelve times in quick succession with even accentuation of each syllable: “tchi tehi tehi tchi tehi tehi tchi.”” Thus the song has a somewhat twittering timbre. Cisticola melanurus (Cabanis ) The discovery of this secies in northeast Angola proves the point made by Chapin (1958, Bds. Belgian Congo, pt. 3:380) that this form is identical with Dryodromas pearsoni Neave which thereby becomes a synonym of Cabanis’ older name. Found about 50 km. southwest of Cacolo, at 1400 m. altitude in the continuous dry woods of the high plateau. In contrast to most of the species of the genus this one dees not depend on grassland or any other dense cover. The birds are, however, more often found in the neighborhood of grassy clearings than in the depth of the forest. Not very elusive and not difficult to observe as they usually dwell in the branches and crowns of lower trees, searching the foliage in the same manner as Apalis species do, for which Heinrich mistook them several times. Disturbed, or excited specimens flip their wings with a purring noise. paella mrecocrry | 4 eer | YALE PEABODY MUSEUM or Natura History Number 48 Jan. 16, 1961 New Haven, Conn. RODENTS AND LAGOMORPHS FROM THE MIOCENE FORT LOGAN AND DEEP RIVER FORMATIONS OF MONTANA Craic C. Biuack', Carnegie Museum, Pittsburgh, Pa. ton) ae The rodents and rabbits described below represent a part of a collection made by Dr. H. J. Koerner for the Peabody Museum of Natural History, Yale University, during the sum- mers of 1935 and 1937 near Fort Logan, Montana. I would like to thank Dr. Koerner for the stratigraphic and locality data used below and Dr. J. 'T. Gregory for the opportunity to study these specimens. I would also like to thank Prof. B. Patterson, Prof. A. E. Wood, and Dr. Mary Dawson for many helpful comments and criticisms. The illustrations are by Mr. James O. Farley and were made possible by a grant from the Gulf Oil Corporation. The following forms are described below: Fort Logan Fm. Niglarodon koerneri n.g. & sp. Eumys eliensis n. sp. Palaeolagus hypsodus Megalagus dawsoni n. sp. ‘This study was begun while the author was Rufus B. Kellogg Fellow from Amherst College. ye) Postilla Yale Peabody Museum No. 48 Deep River Fm. Protospermophilus angusticeps Monosaulaxw cf. M. hesperus Paciculus montanus n. sp. Mookomys ct. M. altifluminus Dikkomys woodi n. sp. Hypolagus sp. Of these forms, Protospermophilus angusticeps (Matthew & Mook, 1988) and Mookomys altifluminus (Wood, 1931) have been previously recorded from the Deep River Formation. In addition, Mesogaulus ballensis (Riggs, 1899) has been reported from the Deep River Formation but it is not repre- sented in the present collection. The following abbreviations are used throughout: A.M.N.H.—American Museum of Natural History, New York Ke Museum of Natural History, University of Kansas Y.P.M.—Peabody Museum of Natural History, Yale University Class MAMMALIA Order RopENTIA Family Aplodontidae Niglarodon*, 1. gen. (Figure la & b) Type species: Niglarodon koerneri®, n. sp. Diagnosis: Jaw slender and less robust than that of Menisco- mys or Sewellelodon; teeth rooted, more hypsodont than in 2From Niglaros Gr. whistle and Odon Gr. tooth. *For H. J. Koerner who made the collection. ® Jan. 16, 1961 Rodents and Lagomorphs 3 Haplomys or Allomys, less so than in Meniscomys and Sewel- lelodon; flexids persisting to, or near, tooth base; molars of equal size; posterior protoconid arm well developed, passing across the molars to prominent mesostylid on M,-M,; prin- cipal cusps prominent with buccal and lingual flexids deep; mesoconid extremely large on P,. Niglarodon koerneri, n. sp. Type: Y.P.M. No. 14024, right mandible with P,-M,, lacking the incisor, coronoid process, and angle. Hypodigm: Type only. Horizon and Locality: Section 4, TION, R5E, Meagher Coun- ty, Montana. Fort Logan Formation, Arikareean. ? A A va | _ Aree. 2 ve = mm f : A -~ bx Joy), ee a ee Ie 7 . gt ig . aa) A Sag a ae ee - me Bi ee -»> 2 2 a Ae 4 * a =, é % ‘ Ee f Figure 1. Niglarodon koerneri n. sp.. Y-P.M. No. 14024, type. A. Right P,-M,, anterior end to the right, X10. B. Lateral view of right mandible, X5. 4, Postilla Yale Peabody Museum No. 48 DESCRIPTION The angle and coronoid process are missing but the condyle is preserved and lies in a plane only slightly above the tooth row. It is not greatly elevated as in T'ardontia and A plodontia. The masseteric ridge is weak in comparison with the other early Miocene genera although it ends in a well-defined tuber- cule below the anterior roots of M,. The coronoid process rises steeply between M, and M,. The mental and dental foramina are as in Sewellelodon and A plodontia. The fourth premolar has five well-defined major cusps. The protoconid and metaconid are joined posteriorly, the antero- flexid cleaving the anterior face of the tooth to within .5 mm. of the anterior root. There is a minute cuspule, that barely breaks the continuity of the slope, present at the base of the protoconid anteriorly. The metastylid is large and separated from the mesoconid and entoconid by a deep mesoflexid di- rected anteriorly. The protoflexid and hypoflexid join buccally and are extremely deep, reaching almost to the base of the crown. The metafossettid is small and shallow. The molars are all nearly similar in structure. The hypo- conid is the largest cusp. The anterior cingulum is separated from the metaconid in unworn teeth by a shallow cleft, but becomes fused with the metaconid slope after little wear. There is a distinct mesoconid on all the molars. The hypofossettid has been cut off on M, and Mz as a very small lake; it is still slightly open on M,. The protoflexid is open in all three molars and would never have been cut off to form a lake. The meta- fossettid is small on all molars but larger than that on P,. The posterior protoconid arm is well developed, extending to the mesostylid but is lower on M, than on M, or Mo; it is separated from the metaconid by a narrow flexid, which would soon have closed off to form a lake on M, and M,, as it has on M*, and from the entoconid by a wider and much deeper mesoflexid. The teeth are worn in such a way that the meta- conid is by far the highest cusp on all the teeth. DISCUSSION Niglarodon is the fifth genus of aplodontid to be described from the Lower Miocene of North America. Haplomys, known Jan. 16, 1961 Rodents and Lagomorphs 5 only from the John Day, is much too low-crowned to be closely related to Niglarodon. Allomys, also known only from the John Day, has become more highly specialized than any other Lower Miocene aplodontid through the development of numer- ous accessory lophs and pits in the upper and lower cheek teeth. Sewellelodon and Meniscomys (see Shotwell, 1958, for the most recent review of aplodontoid evolution) from the Middle and Upper John Day would appear to have the closest relationships with Niglarodon. The crown pattern is basically similar in all three genera with only minor variations. The major differences between these genera are in robustness, height of crown and depth of crown pattern. Of the three, Niglarodon is the most generalized. The teeth are high- crowned, although not as much so as in Meniscomys or Sewel- lelodon, and the depth of crown pattern has kept pace with this increase in crown height, which is not the case in the other two genera. Niglarodon would therefore seem to represent an earlier stage in the aplodontid line leading to Meniscomys and then through Sewellelodon and Liodontia to the recent A plo- dontia. Also, because of its more generalized structure, Nigla- rodon would seem to be closer structurally, although not in time, to the point of aplodontid-mylagaulid divergence than Meniscomys. MEASUREMENTS — — a-p bia Py 3.10 1.80-1.95 M, Wee) 1.60-1.70 M, 2.00 1.75-1.70 M., 2.10 1.75-1.60 Family Sciuridae Protospermophilus angusticeps Matthew and Mook, 1938 * When two transverse measurements are given, the first is that of the protoloph or metalophid, the second is that of the metaloph or hypolophid. All measurements are in millimeters. 6 Postilla Yale Peabody Museum No. 48 Horizon and Locality: Section 25, TION, R5E, Meagher County, Montana. Deep River Formation, Upper Heming- fordian. Referred Specimens: Y.P.M. Nos. 14029 maxillary with M’, 14030 partial maxillary with M'-M?*, 14031 partial left man- dible with M,-M,, 14032 partial left mandible with M,-M,, 14033 partial left mandible with M,-M,, 14034 partial left mandible with M,. DESCRIPTION The six specimens here referred to Protospermophilus an- gusticeps agree well with the type also from the Deep River of Montana. A detailed description of these specimens is de- ferred to a later paper dealing with the evolution of the North American Tertiary Sciuridae. Family Castoridae Monosaulaxv cf. M. hesperus (Douglas, 1901) (Figure 2) Horizon and Locality: Section 1, TON, R4K, Meagher County, Montana. Deep River Formation, Upper Hemingfordian. Referred Specimen: Y.P.M. No. 14035, a right mandible with M,-M;, lacking the anterior portion of the jaw in front of M,, the incisor, and ascending ramus. Figure 2. A. Monosaulax cf. M. hesperus (Douglas), Y.P.M. No. 14035, right M,-M,, anterior end to the right, X7 1/2. Jan. 16, 1961 Rodents and Lagomorphs -~ DESCRIPTION M;, 1s only slightly worn, M, unworn, and M., unerupted. All the molars would show three fossettids with further wear. The hypostriid (sense of Stirton, 1935, p. 392) is relatively shallow and extends only halfway down the crown. There is a small fossettid on all the molars anterior to the parafossettid and another small fossettid posterior to it on M,-Ms. This small posterior fossettid is cut off from the metafossettid and is extremely shallow. The reference of this specimen to Mono- saulax hesperus is based on the presence of the small anterior fossettid. As Stirton (1935, p. 416) observed, however, the species of Monosaulax are not clearly defined and any specific assignment is at best dubious at present. ———— MEASUREMENTS a-p tr. M, BAS 3.00-3.50 M. 3.30 3.20-3.40 M, 2.90 Family Cricetidae Eumys eliensis, n. sp. (Figure 3a & b) Type: Y.P.M. No. 14022, left ramus with I, M,-M3;, lacking the ascending ramus and inferior border. Hypodigm: Type only. Horizon and Locality: Section 28, T11N, R5E, Meagher County, Montana. Fort Logan Formation, Arikareean. Diagnosis: Teeth large in relation to jaw size; teeth progres- sively longer from M, to M,; posterior protoconid arm joining metaconid on M,-M;; no lingual arm of anterior cingulum on M.-M.,,; mental foramen near inferior border of mandible below anterior root of M,. 8 Postilla Yale Peabody Museum No. 48 DESCRIPTION The ascending ramus, angle, and inferior border of the man- dible are missing. Enough of the jaw is present, however, to demonstrate that it is relatively small and slender in relation to tooth size in comparison with other species of Eumys. This jaw is equal in size to that of Eumys brachyodus, somewhat f fn $ y j 4 f ‘ \¢ } ; é ic 4 Ay 4 af \ H A Re Nd i i “ ij Va \ = \ Sih 3 ~~ <" i i ee ae fie: be é - of : J ‘ BS ¥ ‘ J \ 5 b. 4 * ‘% Figure 3. Eumys eliensis n. sp., Y.P.M. No. 14022, type. A. Left M,-M;, anterior end to the left, X7 1/2. B. Lateral view of left mandible, X5. smaller and less robust than that of E. elegans, yet the denti- tion is larger than in either of these two species. There is a small accessory foramen immediately anterior to the mental foramen. The masseteric scar terminates below the middle of M,. Jan. 16, 1961 Rodents and Lagomorphs 9 M, is the smallest tooth of the series. The anteroconid is small, as high as the protoconid, to which it is joimed by a short anterior protoconid arm, but it is well below the meta- conid. The posterior protoconid arm passes postero-lingually and joins the posterior slope of the metaconid and with wear should join with that cusp. There is a very short mesolophid and a short lophid (buccal portion of mesolophid of Wood, 1937, p. 249) which passes buccally from the ectolophid. This crest is also present on M,. The posterior cingulum is sepa- rated from the entoconid by a relatively deep cleft. There is no lingual portion of the anterior cingulum on M, or M;. The posterior protoconid arm on M, and M. would join the meta- conid with further wear. The valley between the protoconid and hypoconid does not open to the buccal side where it is dammed by a thin ridge more so on My, than on M.. There is no mesolophid on M, but the posterior protoconid arm is well developed. The posterior half of M, is constricted, the hypo- conid and entoconid being closely appressed. The anterior and buccal sides of the incisor are rounded while the medial face is flat. The enamel is restricted to the buccal face. DISCUSSION There are several characters of Eumys eliensis that are not to be found in any 6ther species of the genus with which I am familiar. The extreme inferior and posterior position of the mental foramen is unique. The increase in the length of the molars from M, to M,j is also unusual. In all other species M, is generally the longest tooth. And, finally, the large size of the teeth in relation to jaw size is striking. The only previously described specimen which comes close to Ewmys eliensis in tooth size and structure is U.K. No. 8483 described by Galbreath (1953, p. 73) as Eumys sp. This specimen is from the lower part of the Cedar Creek Member of the White River Formation. However, this specimen is described as having a large, heavy jaw, which is decidedly not the case in E, eliensis. E. cricetodontoides, latidens, and spokanensis (White, 1954) are all large species comparable in overall size to E. eliensis but in all three forms the length of the molars decreases from 10 Postilla Yale Peabody Museum No. 48 M, to M,. There are also several differences in crown pattern. There is no lingual portion of the anterior cingulum in E. eliensis, which further separates it from EF. cricetodontoides and latidens. It is probably closest in crown pattern to EF. spokanensis from which it differs in the slight development of the mesolophid on M, and the buccal crest between the proto- conid and hypoconid of My. Furthermore, E. spokanensis is of Middle Oligocene age. As White (1954, p. 410) has pointed out, the intermontaine species of Eumys (cricetodontoides, latidens, and spokanensis, to which may be added eliensis) share certain features not possessed by the plains forms (Eumys obliquidens, elegans, brachyodus, exiguus, and planidens). Until a revision of the genus as a whole is undertaken, however, exact relationships will be extremely difficult to determine. It would appear, never- theless, that EF. eliensis was probably derived from the Middle and Late Oligocene Eumys complex known to have been living to the west of the Fort Logan area. It certainly shares more characters with this group than with the plains forms. MEASUREMENTS a-p cr. I 2.3 2.1 M, 2.50 1.7-1.95 M, 2.60 2.3-2.3 M 2.75 2.25-1.75 Paciculus montanus n. sp. (Figure 4a & b) Type: Y.P.M. No. 14027, partial right maxillary with M'-M?. Hypodigm: Type and Y.P.M. No. 14026, broken right maxil- lary with M' in place and isolated M?-M®. Horizon and Locality: Section 8, TION, R5K, and Section 3, TION, R5E, Meagher County, Montana. Deep River Forma- tion, Upper Hemingtordian, Jan. 16, 1961 Rodents and Lagomorphs 13! Diagnosis: Relatively high crowned; teeth narrow in relation to length; all buccal re-entrants deepening toward centers of teeth; posterior cingulum short; all five crests extremely prom- inent; mesoloph reaching to buccal margin on all molars; no accessary lophs of Ewmys type; no hypocone on M*. Figure 4. Paciculus montanus n. sp. A. Y.P.M. No. 14026, right M'-M?°, anterior end to the right, X15. B. Y.P.M. No. 14027, type, right M’-M?, anterior end to the right, X20. DESCRIPTION The upper molars are all extremely simple when compared with those of Ewmys, Leidymys, or Scottimus. There are no accessory lophs passing from the protocone to the anterior cingulum nor any sign of mesoloph-metacone connections. M! of Paciculus montanus is smaller than those of the various species of HKwmys and than it is in Letdymys. All the teeth are narrower in relation to their lengths than are those of other 12 Postilla Yale Peabody Musewm No. 48 eumyines with the exception of Scottimus. The complete meso- loph passing transversely across M® to the buccal margin is unique for the group. DISCUSSION Paciculus montanus is extremely similar to P. imsolitus (Wood, 1936a) from the Middle John Day of Oregon. It differs somewhat in tooth proportions and is possibly somewhat higher crowned than the earlier species, but could easily have been derived from it. The relationships of the genus are still not clear. The dentition is more conservative than any other eumyine, which makes it difficult to derive Paciculus from any of the known Oligocene species of Ewmys. = MEASUREMENTS .- No. 14027 No. 14026 a-p Ge: a-p [a VW 2.3 1.6-1.5 2.35 lee ale Me een 1.8-1.65 1.80 1.8-1.65 M° 1.50 1.60 Family Heteromyidae Mookomys sp. cf. M. altifluminus Wood, 1931 Figure 5a (Fig Referred specimen: Y.P.M. No. 14036, left ramus with P,-Ms,, lacking ascending ramus and condyle. Horizon and Locality: Section 14, TION, R5E, Meagher County, Montana. Deep River Formation, Upper Heming- fordian. DESCRIPTION This specimen appears to be identical in nearly all respects to A.M.N.H. No. 21360 from the Deep River Beds, 7 mi. south of Logan, Montana. The teeth are more worn but exhibit the same pattern. The only difference between the two is that M, is slightly larger than M, in Y.P.M. No. 14086 (Wood, 1981, p- 4, fig. 4) whereas it is slightly smaller in the type. In this respect it resembles M. parvus from Colorado. Since the type of M. altifluminus and Y.P.M. No. 14036 both come from the Jan. 16, 1961 Rodents and Lagomorphs 13 Deep River Beds near Logan, Montana, reference to M. alti- fluminus rather than M. parvus is to be preferred. The possi- bility exists that M. altifluminus and M. parvus are conspecific but I have not had an opportunity to see the types. Figure 5. A. Mookomys cf. M. altifluminus Wood, Y.P.M. No. 14036, left P,-M., anterior end to the right, X15. B. Dikkomus woodi n. sp., Y.P.M. No. 14038, type, left P,-M,, anterior end to right, X10. MEASUREMENTS ———— a-p tr. ley 85 “10-85 M, 1.00 1.20-1.00 M, SLY) 1.25-1.10 M, 95 1.10-1.00 Family Geomyidae Subfamily Geomyinae Dikkomys woodi,’ n. sp. (Figure 5b) 5 Named for Dr. A. EK. Wood. 14 Postilla Yale Peabody Museum No. 48 Type: Y.P.M. No. 14038, left ramus with I, Py-M,. Hypodigm: Type only. Horizon and Locality: Section 23, T10N, R5E, Meagher County, Montana. Deep River Formation, Upper Heming- fordian. Diagnosis: Crests uniting in centers of teeth; enamel complete; two cusps on anterior lophid of P,, only slight trace of anterior cuspule; anterior column of P, shorter than in D. matthew. DESCRIPTION The jaw is broken through the alveolus of M,. M.-M, are missing together with the angle and the ascending ramus. The masseteric ridge is very prominent, ending below the anterior root of P,. The mental foramen lies antero-ventral to the an- terior end of the masseteric ridge. The diastema is long and shallow. P, and M, are similar to those of D. matthewi (Wood, 1936b) except that the anterior column of P, is not as long as in that species. The teeth are not as worn as those previously described. The anterior and posterior lophids of P, bear two distinct cusps. The two lophids of P, and M, unite just buccal of the center of the tooth to form the typical geomyine H-pat- tern. The buccal re-entrant is shallower than the lingual and is closed on M,. On P,, a shallow valley immediately lingual to the antero-posterior crest just fails to split the anterior lophid in two and there is a small shallow fossettid in the center of the posterior leophid. At an advanced stage of wear the meta- lophid and hypolophid of M, would unite into a single column completely ringed with enamel. DISCUSSION Wood (1986b, p. 26) suggested that Dikkomys would be an “ideal starting point for the evolution of the latter Geo- myinae.” This interpretation was based on the pattern of the lower premolars in which the two lophs unite at the center of the tooth to give a subcircular metalophid and a compressed Jan. 16, 1961 Rodents and Lagomorphs 15 hypolophid as in the later forms. Wilson (1936, p. 28), in discussing Pliosaccomys, said, “the genus cannot be directly ancestral to any existing gopher but in cheek-tooth characters at least, may show a structural stage through which the Geo- myinae have passed.” However, I may observe that worn teeth (Wilson, op. cit. Pl. 2, fig. 5 & 6) are certainly suggestive of those found in Thomomys. The absence of a groove in the upper incisor would also agree with the condition seen in Thomomys, although this is a rather tenuous character. Hibbard (1954, p. 357) pointed out this possible relationship, but at the same time he suggested that the genus Gregorymys might be ancestral to the Geomyinae, stating (1954, p. 357) that “in Gregorymys, the presence of the groove (“sulcus’’), which varies as to position on the upper incisor, the develop- ment of the skull, and the dental pattern seem to indicate an ancestral relationship to, rather than a parallelism with, the Geomyinae.” However, the dental pattern seen in Gregorymys, especially in Py, is far removed from that of any geomyine. There is no indication of a central union of the anterior and posterior lophs. In the contemporary Dikkomys, however, we do find a premolar pattern similar to that seen in T'homomys and Geomys and also in Pliosaccomys. If the suggestion that the geomyine premolars developed from a condition such as that found in Dikkomys and Plio- saccomys is accepted, the next consideration is the derivation of the single column molar of the later forms. The molars of Pliosaccomys are markedly different from those found in Dikkomys. The union of the lophs in Pliosaccomys begins at the buccal margin and spreads inwards until only one column remains. In Dikkomys, on the other hand, the union is first in the center of the tooth. The buccal margins then unite enclosing a lake, the union then proceeding lingually. On the basis of molar structure, it would seem that Pliosaccomys and Dikkomys represent two distinct lines of later Tertiary geo- myine evolution. Which of these lines, if either, is leading to the modern forms it is impossible to say at present. It would appear, however, that the premolar pattern seen in the two genera is the most logical starting point so far known for the recent Geomyinae. 16 Postilla Yale Peabody Museum No. 48 MEASUREMENTS ———_ a-p EE: By 1.45 1.15-1.30 M, 1.50 1.60-1.65 Order LacomorrHa Family Leporidae Palaeolagus hypsodus Schlaikjer, 1935 (Figure 6a) Referred Specimen: Y.P.M. No. 14021, portion of left maxilla with P?-M?. Horizon and Locality: Section 5, T10N, R5E, Meagher County, Montana. Fort Logan Formation, Arikareean. Figure 6. A. Palaeolagus hypsodus Schlaikjer, Y.P.M. No. 14021, left P*-M?, anterior end to left, X10. B. Megalagus dawsoni n. sp., Y.P.M. No. 14023, type, right P*-M’, anterior end to the right, X5. Jan. 16, 1961 Rodents and Lagomorphs ari DESCRIPTION The specimen consists of a portion of the maxilla with P?-M? and the alveolus of M*. P* has the typical abbreviated anteroloph and the persistent J-shaped crescent. The hypos- tria almost reaches the crescent, traversing about a third of the width of the tooth. The hypostriae on P*-M* extend ap- proximately half-way across the teeth and are persistent. Enamel is absent externally and is thin posteriorly on all the teeth. Cement is well-developed, filling the hypostriae on all teeth and the crescent on P*. It does not appear to extend onto the main body of any of the teeth, however. DISCUSSION As Dawson (1958) has pointed out, it is extremely difficult to separate P. hypsodus Schlaikjer (1985) from P. burkei on the basis of isolated dentitions, particularly upper dentitions. Reference of Y.P.M. No. 14021 to P. hypsodus in this case is based first on size and secondly on the flattening out of the buccal face of the anteroloph of P°’. In P. burkei (Wood, 1940, Fig. 97) the buccal face of P’ appears to be of rather uniform slope. Whether this character is constant remains to be seen, but it is apparent in all illustrations of the two species so far published. This occurrence extends the range of the species as given by Dawson (op. cit.) from Wyoming, South Dakota, and Nebraska, into Montana. pea MEASUREMENTS a-p Er RP 1.80 2.00-3.10 | Pei 1.80 3.30-3.00 M? 1.80 3.30-2.90 M? 1.50 2.60-2.10 Megalagus dawsoni’, n. sp. (Figure 6b) ®’ Named for Dr. Mary Dawson. 18 Postilla Yale Peabody Musewm No. 48 Type: Y.P.M. No. 14023, right maxillary with P?-M?’. Hypodigm: Type and Y.P.M. No. 14037, left maxillary with P°-M* poorly preserved. Horizon and Locality: Section 28, TI1IN, R5E, and Section 23, TION, R5E, Meagher County, Montana. Fort Logan Formation, Arikareean. Diagnosis: Buccal roots presumably present (seen on M°); teeth high-crowned ; hypostriae cement-filled, shallow on P*-P*, deeper on M'-M*, passing one-third of the way across the crown. DESCRIPTION The cheek teeth are more high-crowned than in any other species of the genus, and have prominent buccal roots. P* ex- hibits two anterior re-entrants, the buccal being shallow and extending approximately 2.5 mm. down the anterior face of the tooth, the lingual deeper and extending halfway down the anterior face. The hypostriae are shallow and extend only partway down on P*-P*, but are deeper and extend well below the level of the alveoli on M'-M’. The teeth are longer in rela- tion to their width than in any other species of Megalagus. DISCUSSION The reference of these specimens to Megalagus is based upon the presence of buccal roots and the shallow development of the hypostriae on P°-P*. The dentition of M. dawsoni shows several advances over other species of the genus, notably the development of cement and greater hypsodonty. In this regard, M. dawsoni has progressed further than M. primitivus, the only other Miocene species. M. dawsoni represents a further, more advanced level of development, derivable from M. tur- gidus, as is M. primitivus, but distinct from it. Jan. 16, 1961 Rodents and Lagomorphs 19 = _MEASUREMENTS ——_— Y.P.M. No. 14023 a-p bie: 1B 2, 2.1 Be 3.0 2.5-3.4 Ee 2.8 3.5-3.5 M! 2.5 3.3-3.1 M- 2.2 3.1-2.5 Hypolagus sp. Referred Specimen: Y.P.M. No. 14028, partial left maxillary with alveolus of P*, broken P*, and P*-M!. Horizon and Locality: Section 25, TION, R5E, Meagher County, Montana. Deep River Formation, Upper Heming- fordian. DESCRIPTION The specimen is much too poor for definite reference, but on the characters available it would seem to be close to Hypo- lagus vetus. The crenulations of the hypostriae are slight on M', more wavy on P*. ee MBAS URE VRE INED Ss a-p tr Pt ae | 4.2 M'! ee 4.3 REFERENCES Dawson, M. R., 1958. Later Tertiary Leporidae of North America. Univ. Kansas Paleont. Contrib., Vertebrata, Art. 6, pp. 1-75, 2 pl. Galbreath, E. C., 1953. A Contribution to the Tertiary Geology and Paleon- tology of Northeastern Colorado. Univ. Kansas Paleont. Contrib., Vertebrata, Art. 4, pp. 1-120, 2 pl. Hibbard, C. W., 1954. A new Pliocene Vertebrate Fauna from Oklahoma. Papers Mich. Acad. Sci., Arts, and Letters, v. 39, pp. 339-359. 20 Postilla Yale Peabody Museum No. 48 Matthew, W. D. and Mook, C. C., 1933. New Fossil Mammals from the Deep River Beds of Montana. Amer. Mus. Novit., 601, pp. 1-7. Riggs, E. S., 1899. The Mylagaulidae: An extinct Family of Sciuromorph Rodents. Field Columbian Museum, Geol. Ser. 1, pp. 181-187. Schlaikjer, E. M., 1935. Contributions to the Stratigraphy and Paleontology of the Goshen Hole Area, Wyoming. IV. New Vertebrates and the Stratigraphy of the Oligocene and Early Miocene. Bull. Mus. Comp. Zool., v. 76, no. 4, pp. 97-189, 41 pl. Shotwell, J. A. 1958. Evolution and Biogeography of the Aplodontid and Mylagaulid Rodents. Evolution, vy. 12, pp. 451-484. Stirton, R. A., 1935. A Review of the ‘Tertiary Beavers. Univ. Calif. Publ. Bull. Dept. Geol. Sci., v. 23, no. 15, pp. 391-458. White, ‘I’. E., 1954. Preliminary Analysis of the Fossil Vertebrates of the Canyon Ferry Reservoir Area. Proc. U.S. Nat. Mus., y. 103, no. 3326, pp. 395-438. Wilson, R. W., 1936. A Pliocene Rodent Fauna from Smiths Valley, Ne- vada. Carn. Inst. Wash. Publ., 473, pp. 15-34, 2 pl. Wood, A. E., 1931. Phylogeny of the Heteromyid Rodents. Amer. Mus. Novit., 501, pp. 1-19. 1936a. The Cricetid Rodents described by Leidy and Cope from the Tertiary of North America. Amer. Mus. Novit., 822, pp. 1-8. 1936b. Geomyid Rodents from the Middle Tertiary. Amer. Mus. Noyit., 866, pp. 1-31. ——1937. The Mammalian Fauna of the White River Oligocene. Part II. Rodentia. Trans. Amer. Phil. Soc., n.s., 28, pp. 115-269, 11 pl. 1940. The Mammalian Fauna of the White River Oligocene. Part III. Lagomorpha. Trans. Amer. Phil. Soc., n.s., 28, pp. 271-362, 2 pl. bol YALE PEABODY MUSEUM | oF Natura. History Number 49 June 23, 1961 New Haven, Conn. RESULTS OF RESEARCH IN THE ANTOFAGASTA RANGES OF CHILE AND BOLIVIA I. Birds: Luss KE. Pena II. Diatoms: Rutu Parricx Color Plate: Roger Tory PETERSON ae med 7 7 a . ; : pee et ae “ie ok Ce Celotta eT a Tore ee, ‘ie las . _ Po ao " as) tte ‘ ) oi . ieee OA 1 Z je RHODA WES! “Y | > ae yer | nit 7 - - - ~~ f Swiinoradl 7 | rr aay iD) 4oe ? = : - 7 i °F Way ite) ae cm b. ‘ ; Se : in the middle of a plain where of “algarrobo” and “chanares’ corn and alfalfa are cultivated, irrigated by the salt water of the ravine. Titorozo: On the extreme south of the Atacama salt marsh there is a large area of marsh covered with grasses. We trav- eled as far as the Banos de Tilopozo which spring from a hot-water overflow. This place must be ideal for birds dur- ing the summer nesting season, but they are now absent. The 28 Postilla Yale Peabody Museum No. 49 hurricane-force wind made continuing with the jeep impossible because of the hail of sand and rock which had already heavily scratched the windshield. We returned to Santiago on July 26 after a fruitless visit to the salt marshes of Pedernales and Maricunga, more than 250 kilometers to the south of Atacama. Flamingos do nest on these marshes in the summer months. PART TWO Awxnnoratep List oF Brros CoLLECTED OR OBSERVED The list of birds reported here includes observations made in December, 1957, and in January, 1958, on a second visit with Dr. Roger Tory Peterson. Specimens collected are now in the Peabody Museum at Yale University. A series of our material is also in the collection of the Museo Nacional de Historia Natural of Santiago. Pterocnemia pennata tarapacensis Chubb Several pairs were observed in Inacaliri at 8,900 meters, and on the Turi pampa (3,100 meters). One specimen was captured by persons from Inacaliri and carried to the San Pedro station with the object of removing its fat, since it is highly desired as a remedy for rheumatism and arthritis. Be- fore this, however, I took the following measurements: length from the point of the bill to the end of the tail, 144 cm; wing, 50 cm; bill, 7.2 cm; tarsus, 33 cm. On examining the contents of the stomach, I found it replete with roots and thick stems of the tola. Tinamotis pentlandi Vigors This species of “partridge” is somewhat frequent in the high regions of the Antofagasta range between 3,500 and 4,500 meters. A slightly wounded specimen attacked me by flying against my face as I approached to photograph it at a dis- tance of 3 meters; later it fled, tossing off a large quantity of fetid excrement. An example was captured in Inacaliri at June 25,1961 Antofagasta Ranges of Chile and Bolivia 29 4,000 meters altitude. A small flock was observed on the T'atio voleano at 4,400 meters. Phoenicopterus chilensis Molina The Chilean flamingo is common in the central zone of Chile. It is found in the lagoons of the north where it is known by the name of ‘‘tococo.” The specimen taken came from the lagoons of the Ceja marshlands, situated in the interior of the upper sections of the Atacama salt marsh approximately 30 kilometers to the south of San Pedro de Atacama. It was a solitary bird. Phoenicoparrus andinus (Philippi) Bonaparte At Ojos del Rio San Pedro two immature specimens of this ““parina” were captured, locally called “hiticti.”” Birds were observed in the Ceja lagoons and in those of Carvajal, situated in the interior of the Atacama salt marsh. At night they could be heard calling, and it is at that hour that they move from lagoon to lagoon. It is difficult to distinguish them from the “parina chica” (Ph. jamesi), when they are in their habitat, since it is impossible to approach them, and their coloration and size is very similar. Phoenicoparrus jamesi (Sclater) One of the objects of this expedition was to study this rare flamingo. No one has had recent news of this species. In 1957, Dr. Francisco Behn, in company with Mr. A. W. Johnson and Mr. Guillermo Millie, had traveled in these areas with the sole object of finding this bird and making studies of a biological character (Johnson, Behn, and Millie, 1958). They succeeded in capturing a female example and obtained several nests of eggs, all in the Laguna Colorada of Bolivia in the month of February, 1957. In the same place, situated a few kilometers from the frontier, I succeeded in capturing an adult male example and two immature specimens. The population was reduced. I was only able to see 7 adults and about 20 nestlings already old enough to fly. The adults were extremely shy; not so the young ones which walked about tranquilly scarcely 30 Postilla Yale Peabody Museum No. 49 10 meters from us, as we, with the water a little bit below our knees, were exploring the lagoon in the region not affected by ice. Later on, in the lagoons at Carvajal to the east of the Atacama salt marsh and in front of the Hekar or Camar ravine, we succeeded in capturing another example. This “parina,” called “chururu” by the inhabitants, is prob- ably not as rare as had been thought previously. Many times my attention was called to the fact that the mountain lagoons were almost completely deserted by birds, according to infor- mation received from the inhabitants. However, the flamingos were abundant in the Laguna Colorada during the summer and nested there. This was proved by Dr. Behn, although his cal- culation for the species was low, 6-8% of the total of 3,000 birds of 3 species seen (tom. cit. 1958: 296). With the object of studying the possible movements of the species about the Atacama salt marsh area, I traveled among the lagoons of the interior observing an immense quantity of flamingos in a number almost impossible to calculate. But, in view of the quantity of little lagoons which exist in this vast region, I cannot make any firm estimate of this population. Unfortunately, I was not able to go as far as the Punta Negra salt marsh which is south of the Imilac railroad station. I was assured that flamingos nest there by the thousands during the months from December to February. It seems to me that the Atacama salt marsh is the nesting center and the place in which the majority of the flamingos spend the winter. These flamingos are vagrants. The mountain lagoons freeze, though not entirely, since nearly all of them receive hot springs. But the influence of hot water is only felt in a limited space. as I had observed in the Laguna Colorada (Bolivia) and the Laguna Verde. These birds migrate to salt-water lagoons of a better climate such as the Atacama salt marsh. Some speci- mens maintain themselves in the Andean lagoons, but they are always isolated. In the entire region the people who know about the “pa- rinas,” including the natives who spend the summer on the Laguna Colorada as well as the people from Toconao, Peine, and Socaire, who live for a part of the year on the eggs of these birds, distinguish four well-defined species of flamingos June 23,1961 Antofagasta Ranges of Chile and Bolivia 31 and “parinas.” They have a name for each, taken from the call which the birds emit. The first is the “tococo” (Ph. chilensis), the common species which is also found in the central region of Chile. The second is the “hiticti” (Ph. andinus) which is more reddish in color than the first one and larger in size. Its call is very similar to its name: “hi-ri-ri-ri-ri-ri.” The third one is the “chururu” (Ph. jamesi) which is the species that was thought to be the most scarce and turns out to be the most abundant. Its characteristic call is ‘‘chu-ru-ru-ru-ru-ru.”’ The fourth is a species which we have not been able to locate. In my opinion it comprises young examples of jamesi. On vari- ous occasions I was able to hear its call of “huaj-cha-ta-ta.” Upon observing the stomach contents of these birds (Ph. jamest), I found only some mud of greenish-yellow color mixed up with fine sand. Under the microscope I could see that the contents represented a compact mass of diatoms, which agrees with the observations of Dr. F. Behn. Phoenicoparrus jamesi is a common enough bird, as are all the “parinas” and flamingos of the Andes. It is very diffi- cult to observe and capture, however, since it is very shy. It nests on the Laguna Colorada of Bolivia and almost posi- tively on the Andean salt-water lagoons such as the Pujsa salt marsh and the Loyoquis salt marsh. According to the word of the people of the region, it lays eggs three times a year. A second visit to Laguna Colorada made by Dr. Roger Tory Peterson and myself from December 20-27, 1958, adds to our knowledge of the relative abundance of the three flamingo species. During the second visit the observed ratio of flamin- gos was: ELDON CEST Ses pe i ie Ce orl SO ee no less than 7,000. Re LAIN Stat ee ee ee 300. VRC RIUCTISUS eet n Tee Bake. inte td ee 100. This compares with the figures of Johnson, Behn, and Millie | g for January, 1957 (tom. cit. 1958) : JE] LO OTAR OTAGO Ss ate Sr aaa 40 to 50. Phadnis 2 8. 2. Ph. chilensis ,., eae ee 1,500 approximately. re tie, We 2 neck SN le aie eer ean 1,500 approximately. 32 Postilla Yale Peabody Museum No. 49 Chloéphaga melanoptera (Eyton) I have observed this species always in pairs at altitudes of 3,500 meters (Ojos del Rio San Pedro) to 4,250 meters on the marshlands which are the sources of the Salado River; that is, the T'atio Geysers. It was not possible to capture examples. Lophonetta specularioides alticola (Ménégaux) This is perhaps the most commen duck of the mountain marshlands which were explored. I observed it from 2,400 me- ters to 4,400 meters in Chile as well as in Bolivia. It nearly always travels in pairs and is rarely seen in flocks (the ravine of Pastos Largos, Atacama). It is known as the “pato rial” and is very much desired on account of its flesh and size. It is easily distinguished because it is much larger than the other ducks with which it lives. Anas versicolor puna Tschudi Rare in the region explored; I observed some examples only in the Ojos del Rio San Pedro (3,800 meters) and in the Laguna Colorada of Bolivia at 4,400 meters. Anas flavirostris ovyptera Meyen A common duck on the Inacaliri marshes and on the Ojos del Rio San Pedro. They were observed in flocks of from 15 to 30 examples. It was also seen on Laguna Colorada, Bolivia, at 4,400 meters; Tilopozo (2,400 meters); the Putana River; and the ravine of Pastos Largos, this last in the province of Atacama. Fulica americana peruviana Morrison Very common in the Ojos del Rio San Pedro where it is persecuted by hunters. Fulica cornuta Bonaparte This rare species is known only from the high ranges of b) the Andes. It is called “huari” or “socar” by the natives. June 23,1961 Antofagasta Ranges of Chile and Bolivia 33 Very little is known about its habits. I observed 36 examples on the Laguna Verde at 4,100 meters. They were all swimming on the eastern side of the lagoon, since the water there is always temperate. The rest of the lagoon was frozen over with a thick cap of ice. It was very amusing to see these birds emerge from the water and try to walk on the ice, while at- tempting to fly. It was impossible for them to maintain their footing on the frozen surface. They lived there in common with a large group of Podiceps occipitalis juninensis, some pairs of Lophonetta speculariotdes alticola and Anas flavirostris oxyp- tera, Three examples were captured and were weighed as fol- lows: 2.4 Ibs.= 6), 5ulbs. 12° 0z.;°6 4.5 Ibs: 5 oz: The color of the bill is yellow, with greenish-olive tints. Towards the base it is somewhat reddish and the upper part of the culmen is black. The feet are greenish-black with some yellow. The iris is red. All of the examples which were observed had fully developed ‘aruncles. On opening their stomachs, I found little food, and this was made up of aquatic grasses which abound in the temperate parts of the lagoon. Most of the stomach contents consisted of voleanic sand. In a little lagoon near the Laguna Verde (Bolivia) the remains of the nests of these F'ulica were seen. For a descrip- tion of the nests farther south see Ripley (1957 a and b), and Behn and Millie (1959). Oreopholus ruficollis (Wagler ) Flocks with numerous specimens were observed in the marsh- lands of Turi at 3,200 meters. They were extremely shy. They flew in more or less compact groups and at a distance of more than 200 meters. They are known by the name of ‘“‘chu- ehiri” or “tiutila.” Charadrius alticola (Berlepsch and Stolzmann) A relatively scarce bird which was observed only in the Ina- ‘alir1 marshes (4,000 meters). It was abundant at the Tatio voleano (4,250 meters) and was even more common in the O4 Postilla Yale Peabody Museum No. 49 Atacama salt marsh, on the Carvajal lagoons, and the marsh- lands at Ceja. This plover was seen running after food on the muddy little beaches which are formed at the shores of the salt marsh lagoons. Phegornis mitchell (Fraser) It is known as “pijlulo” by the inhabitants of the Inaca- liri region. Only two examples were observed in the Inacaliri marshes at approximately 4,000 meters. Capella paraguaiae innotata Hellmayr Only one snipe was observed in the marshes of Inacaliri. It flew off at the moment my foot was about to crush it. According to what I was told, it is quite common during the summer months. Recurvirostra andina Philippi and Landbeck ‘ ’ The “caiti,’” known in the region as “caichén,” is frequently found in the salt-water lagoons and lowlands of the Andean region. The two examples obtained are immature and come from the Laguna Colorada. At that place there were small groups of three to seven specimens resting on the warm water of the lagoon, since the rest of the water was frozen. This friendly little bird was also observed in the lagoons of the marshes of Ceja and Carvajal in the interior of the Atacama salt marsh (2,400 meters). Thinocorus rumicivorus rumicivorus Eschscholtz Two examples were taken in June while the camp was being set up in the desert region in front of the village of Domeyko in the province of Atacama. It is a bird which abounds in the region. Thinocorus orbignyianus orbignytanus I. Geoffroy St. Hilaire and Lesson The “puco-puco” or “poco-poco” is very common in all of the marshes explored; it was observed between 3,800 and 4,300 June 23,1961 Antofagasta Ranges of Chile and Bolivia 35 meters. It is always seen in little flocks of three to eight. It is elusive, hiding itself among the crevices where the water from the marshes runs, always with its head sticking up. As soon as it takes flight, it utters a characteristic cry. Larus serranus Tschudi The “gaviota” is frequently found in all weather in the lagoons of the high ranges of the Andes. An example was observed in the Laguna Colorada of Bolivia (4,400 meters) living with “caitis” and “parinas.” Others were seen on the lagoons of Carvajal and the marshes of Ceja in the interior of the Atacama salt marsh (2,400 meters). All were wearing their faded winter plumage. Zenaidura auriculata auriculata (Des Murs) This dove, so common in the central zone of the country, had not been captured previously in Antofagasta. In San Pedro de Atacama it is extremely abundant during the winter months. Several flocks were observed in Toconce as well as in the high desert zone of Paposo (the southern coast of the province of Antofagasta). An example was secured from To- conce (3,030 meters). Metriopelia melanoptera melanoptera (Molina) An example was collected in Lasana on the banks of the Loa River. Psilopsiagon aurifrons orbygnesius (Souanceé) Several flocks of ‘“‘caturros’? were observed in Inacaliri (4,100 meters), in Toconce (3,030 meters), and on the out- skirts of the Talabre ravine (3,200 meters). It is a rather common little parrot which nests in the area. Crotophaga sulcirostris sulcirostris Swainson I was surprised to find an example of this bird in Peine because it had to cross so many kilometers of desert area to 36 Postilla Yale Peabody Museum No. 49 arrive there. I observed it among the branches of a Bolivian pepper tree which is near the public square. According to what the natives told me, this bird had arrived there at the end of the month of March. It was not captured. Oreotrochilus estella @Orbigny and Lafresnaye Common in Toconce where some examples were collected. They are to be seen in cultivated areas visiting the flowers of the nettle “ortiga.” Geositta isabellina (Philippi and Landbeck) This species, so rare in collections, was observed on the outskirts of Potrerillos in the province of Atacama while I was on the road to the salt marshes of Pedernales. It has only been known previously from the province of Coquimbo further south. Geositta punensis Dabbene Very common in several places between 8,200 meters and 4,300 meters. It is found most frequently around houses and in the regions of the tolas. It is known by the name of “roilita” (Talabre). Some examples were taken on the outskirts of the Turi marshes (3,100 meters) and in the tolas of Inacaliri. Upucerthia dumetaria hallinant Chapman A rather common species which is difficult to capture because it is very shy. Examples were observed in Ayquina (3,030 meters), Guatin (8,500 meters), Peine (2,400 meters), and San Pedro de Atacama (2,400 meters). It is always seen traveling over the sown and cultivated terraces. It emits a strident and characteristic call. It is commonly known as the “lucho-lucho” (Toconce) or the “lichi-lichi” (‘Talabre). Upucerthia ruficauda (Meyen) This bird is quite rare. It frequents the tolas where a few examples were seen. It was observed in Linzor (4,100 meters ) and Inacaliri (4,000 meters). June 23,1961 Antofagasta Ranges of Chile and Bolivia 37 Asthenes modesta modesta (Kyton) Common above 3,500 meters, it was observed among the tolas. It is frequently seen on rocks and on the ground search- ing out its food. It is very easy to capture since it allows one to approach to within a few meters. The vulgar names given 5 to it in these regions are “‘pipo,” “‘lucho-lucho,” and “‘caci- que”’ (Peine). Leptasthenura aegithaloides berlepschi Hartert Known as “tijerita” or “chirivir1” (Talabre) and as ‘‘qui- ron” in Peine, it is a very common little bird between 2,400 meters (San Pedro de Atacama) and 4,100 meters (Linzor). It frequents cultivated fields and the regions of the tolas. It is always seen among the shrubs searching out food. Cinclodes fuscus albiventris (Philippi and Landbeck ) Very common in the marshes and rivers in Inacaliri (4,100 meters), the Ojos del Rio San Pedro (3,800 meters), Silala (4,200 meters), Vegas del Tatio (4,300 meters), Toconce (3,800 meters), Guatin (3,500 meters), Peine (2,400 meters), and the Laguna Colorada of Bolivia (4,400 meters). It is extremely tame; the natives call it “requete chico” (Toconce), “alcalde,” and “itirico” (Peine), and “sapero” (Talabre). It lives in some places with the congeneric C. atacamensis ata- camensis, In observing the stomach contents of some samples, we found insect larvae, possibly of Lepidoptera Heterocera, and many remains of aquatic vegetation. Cinclodes atacamensis atacamensis (Philippi) Less frequent than the previous species. Several examples were captured in Toconce (3,300 meters), Linzor (4,100 me- ters), and Silala or Siloli (4,800 meters). No examples were observed in Inacaliri (4,100 meters). In an example taken in Linzor, I found remains of coleopterous insects of the family Elmidae, so common beneath the stones submerged in the streams. There were also little stones and an uncountable num- ber of little snails of the genus Littoridina. 38 Postilla Yale Peabody Museum No. 49 Agriornis andicola albicauda (Philippi and Landbeck) The “gaucho” is one of the rare species which with luck may be observed from time to time. It has only been found in Putre in the interior of the province of Tarapaca, Department of Arica. A specimen from Linzor was collected on the out- skirts of the Ojalar River at an altitude of 4,100 meters. Agriornis montana maritima (Lafresnaye and d’Orbigny ) Examples were observed in Inacaliri (4,000 meters) and in its immediate environs. In Linzor (4,100 meters) it seems to be more common, though not so in Toconce (3,300 meters ) where it was rather scarce. The stomach contents of one of the examples was com- posed of the seeds of the prickly grass (Oxychloe andina Phil.) which grows in the swamps and marshes, the remains of a chrysidid hymenopteran, and a species of Diptera (Syrphidae, genus Volucella). In a female example captured in Linzor I found its stomach full of the same seeds, water plants from the streams, and the remains of insects which were impossible to determine. In another male example I found seeds and a “oaucho” in wing-bone of a small bird. In Peine I observed a the process of killing a small mouse. These birds frequent the tolas, the shores of streams, and marshes. Muscisaxvicola rufivertex pallidiceps Hellmayr > it is one of the most char- Known as “fraile”’ or “‘cura,’ acteristic birds of the traveled regions. It is quite common between 3,300 meters and 4,250 meters. It frequents the marsh- lands and streams in search of food. Muscisaxicola capistrata (Burmeister) As common as the previous species and with very similar habits. Examples were observed in all the places visited from 3,200 to 4,250 meters. This “fraile’ dwells in Patagonia and in Tierra del Fuego and spends the winter in the ranges to the north. All of the examples had very faded plumage. June 23,1961 Antofagasta Ranges of Chile and Bolivia 39 Muscisavicola maculirostris maculirostris Lafresnaye and d’Orbigny Called the “‘fraile chico,” it is very common in Ayquina (3,030 meters). It has been observed running on the cultivated terraces and on the shores of running waters. In San Pedro de Atacama (2,400 meters) ground-tyrants were observed in the fields cultivated with alfalfa. Lessonia rufa oreas (Sclater and Salvin) Only one example was captured at the Ojos del Rio San Pedro (3,800 meters), even though some pairs were observed in the marshes which form the streams which are the sources of the San Pedro River. Few examples were observed in the lagoons of Carvajal in the interior of the Atacama salt marsh. Anthus correndera catamarcae Hellmayr I observed this subspecies of “bailarin chico” in the marsh- lands of the Ojos del Rio San Pedro, 3,800 meters; also, in a lesser quantity, in the marshlands of Ceja, which are in the interior of the Atacama salt marsh at 2,450 meters; and in great abundance at the southern extremity of the same salt marsh at the location of the Tilopozo marshes. They are al- ways found in humid places among the tall grasses where they hide, making their capture rather difficult. It is a rather shy little bird. Troglodytes musculus atacamensis Hellmayr Common in the thickets of San Pedro de Atacama between 2.400 and 3,300 meters. Turdus chiguanco anthracinus Burmeister This species of thrush, very common in Bolivia and Argen- tina, has been sporadically collected in Chile. Its nesting had never been observed before in Chilean territory. On this trip I was able to observe numerous examples in the valleys of the Loa (Lasana) River. In San Pedro de Atacama, where it is 40 Postilla Yale Peabody Museum No. 49 known as “lachirachi,” it is extremely abundant, as it is also in the oasis of Toconao (2,400 meters). According to infor- mation given by the natives, I was assured that they nest there. Passer domesticus domesticus (Linnaeus ) The “gorrion” has now arrived at the village of San Pedro de Atacama, with its usual instinets of destruction towards the small autochthonous birds. Phrygilus atriceps (Lafresnaye and d’Orbigny ) I found this beautiful fringilline in abundance in 'Toconce at 3,300 meters. Some examples were captured on the outskirts of Linzor (4,100 meters). They also abound in Guatin, a place located to the northeast of San Pedro de Atacama at more or less 3,500 meters altitude. They always travel in small flocks. The female is much rarer than the male, and we were able to capture only one example in Toconce. It is a harmful bird since it destroys vegetables and eats corn, wheat, and oats, especially when they have not yet matured. It is vulgarly known by the name of “‘comesebo.” In Talabre it is called “chasea.” Phrygilus unicolor unicolor (Lafresnaye and d’Orbigny ) Only a few examples were observed, both in Siloli (Silala) at 4,200 meters and in Toconce (3,300 meters). An example was observed in the Laguna Colorada and another in the La- guna Verde, both in Bolivia. Phrygilus dorsalis Cabanis This rare little bird, known in Chile by only two or three examples found in these ranges, turns out to be extremely common and numerous specimens were observed and captured. The “sulte,” as they call it in some places, was observed in Inacaliri (4,100 meters), Silala or Siloli (4,800 meters), the Tatio Geysers (4,200 meters), Linzor (4,100 meters), and Laguna Colorada in Bolivia (4,400 meters). I understand that June 23,1961 Antofagasta Ranges of Chile and Bolivia 41 it abounds in Tumbre, a place to the northeast of the Laskar voleano at an approximate altitude of 4,000 meters, where I was assured that it nests. It is not a shy bird and frequently approaches to within a few meters. Numbers are seen in the sur- rounding areas of the small streams which form the marshes of the deep ravines. When chased, they fly toward the walls of the ravine, resting themselves on the rocks. From time to time they may be seen flying directly upwards where they remain immobile in the air by beating their wings. Later they let themselves fall again toward the earth. Their presence 1s revealed by a fine call which they express with a penetrat- Ine Tis, Phrygilus fruticeti fruticeti: (Kittlitz) The “yale” is known in a large part of the country as a rather destructive and damaging little bird. It is frequently seen in Toconce (3,800 meters), Guatin (3,500 meters), and Talabre (3,300 meters) where I was able to observe it in flocks of about 20 birds. Zonotrichia capensis antofagastae Chapman This is perhaps the most characteristic bird of agricultural sites situated between 2,400 meters and 3,500 meters in the entire region visited. It is known by the name of ‘“chincol” and is extremely common in the village of Toconce (3,800 meters), at Ayquina (3,030 meters), San Pedro de Atacama (2,400 meters), Tilomonte and 'Toconao (2,500 meters). It is observed in groups or alone. Spinus atratus (Lafresnaye and d’Orbigny ) The “canario” is very much desired by the natives as it has a beautiful song. After capturing them by means of traps, they raise them in cages. They abound in the region of Siloli in the summertime. We were not able to obtain ex- amples in spite of having seen some in the Toconce ravine (3,300 meters). 42 Postilla Yale Peabody Museum No. 49 Sicalis uropygialis uropygialis (Lafresnaye and d’Orbigny) I only observed two flocks of this “chirigue,” both very numerous. One of them was at the Ojos del Rio San Pedro (3,800 meters) and the other at the Vegas de Inacaliri, a place close to 4,000 meters. Lirerature Crrep Behn, F. and Millie G. 1959. Beitrag zur Kenntnis des Riisselbliisshuhns (Fulica cornuta Bonaparte). Jour. f. Orn. 100: 119-131. Johnson, A. W., Behn, F., and Millie, W. R. 1958. The South American flamingos. Condor 60: 289-299. Koford, C. B. 1957. The vieuna and the puna. Ecol. Monog. 27: 153-219. Ripley, S. D. 1957a. Notes on the horned coot, Fulica cornuta Bonaparte. Postilla no. 30: 1-8. Ripley, S. D. 1957b. Additional notes on the horned coot. Postilla no. S2eml2: June 23,1961 Antofagasta Ranges of Chile and Bolivia 43 si DIATOMS (BACILLARIOPHYCEAE) FROM THE ALIMENTARY TRACT OF PHOENICOPARRUS JAMESI (SCLATER)* Rury Parrick Acapemy or NaruraL SCIENCES, PHILADELPHIA, Pa. ~ During December, 1957, Phoenicoparrus jamest (Sclater) was collected by Sehor Luis Pena from Laguna Colorado Puna de Atacama, Bolivia, at an elevation of 4,400 meters. Walcott (1925) describes this lake as strongly alkaline with springs at the north end. On the shores were salt deposits consisting of sodium and potassium carbonate, sodium chloride, and borax. In July, 1957, Senor Luis Pena collected this species in Salt Lake, Atacama, Chile and in Lagunas de Carvajal (salt), Atacama, Chile. Both collections were taken at an elevation of 2,400 meters. An analysis of the contents of the alimentary tracts of these birds showed that they were mainly diatoms. The most common species were Navicula carvajaliana sp. nov., Amphora ataca- mana sp. nov., Navicula luisit sp. nov., and Nitzschia accedens var. chilensis var. nov. These findings confirm the prediction of Jenkins (1957), based upon the type of filters of Phoenicoparrus jamesi, that this bird would feed on algae or diatoms. It is interesting to note that the more common diatoms in the alimentary tract have a similar size range. At all three locations the commoner diatoms were the same. This is probably due to the fact that the areas in the lake which were favorable feeding grounds for these birds supplied ecologically similar habitats for diatoms. Also, since these strongly alkaline salt-water lakes represent a very specialized *'The author wishes to express her appreciation to an anonymous donor who helped to make this publication possible. She also wishes to thank Miss Helen Wu who made the drawings. 44 Postilla Yale Peabody Museum No. 49 habitat for diatoms, the numbers of kinds of species which can grow under these conditions are relatively limited, and therefore the diatom floras of these lakes would be more similar than is generally the case in the more usual fresh or brackish water lakes. The two lakes in Chile are quite close together, but it is unlikely that the birds captured in Bolivia had pre- viously fed in the Chilean lakes. The fact that flamingos may feed on algae has been re- ported by various workers. Ridley et al. (1955) and Jenkins (1957) state that in East African Lakes, Phoeniconaias minor (Geoffroy) feeds almost exclusively on blue-green algae and on small diatoms. Usually diatoms are mixed with the blue- green algae and sometimes diatoms without the blue-green algae seem to be the main source of food. According to a letter written by Dr. Robert P. Allen to Mr. Ridley (Jenkins, 1957), Phoentcopterus ruber Linné in the Bahamas feeds on mud rich in bacteria, blue-green algae and, to a lesser extent, diatoms. A systematic list of the species identified in this study 1s given below. The slides on which these identifications are based are in the diatom herbarium of the Academy of Natural Sei- ences of Philadelphia. The type specimens are ringed on the slides. Suborder MONORAPHIDINEAE Family ACHNANTHACEAE Genus Achnanthes Bory Achnanthes brevipes var. intermedia (Kiitz.) Cl. Achnanthes brevipes var. intermedia (Kitz) Cl. K. Svenska Vet.-Akad. Handl., ser. 2, 27 (3): 1938, 1895. Disrripution: Chile, Atacama, Salt Lake, alt. 4,400 m, coll. Luis Pena, July, 1957 (A-G.C. 26098). Achnanthes hauchkiana var. rostrata Schultz Achnanthes hauckiana var. rostrata Schultz, Bot. Arch., 13: 191, fig. 39, 1926. Disrrisutrion: Chile, Atacama, Salt Lake, alt. 4,400 m, coll. Luis Pena, July, 1957 (A-G.C. 26098a) ; Lagunas de Carva- jal, alt. 2,400 m, coll. Luis Pefia, July, 1957. (A-G.C. 26100). June 23,1961 Antofagasta Ranges of Chile and Bolivia 45 Family NavicuLacEAE Genus Navicula Bory Navicula atacamana sp. nov. Pi Big. 0 Valva lineari-lanceolata apicibus acutis apiculatis leniter. Area axiali angusta. Area media rectangularis latus distendens ad margines valvae. Striis lineatis, parallelis in media parte valvae et convenientibus ad apices. Striis, 10-11 in 10”; longi- tudo, 49-50; latitudo, 6-7. Valve linear-lanceolate with acute, slightly apiculate ends. Axial area narrow. Central area a broad rectangle extending to the margins of the valve. Striae lineate, parallel in’ the center of the valve and slightly convergent toward the apices. Striae, 10-11 in 10p; length, 49-50; breadth, 6-7. This species is similar in shape, striae structure and angle, and narrowness of axial area to Navicula directa (W. Sm.) Cl. It differs in the presence of a broad rectangular central area and in its smaller size. Tyre tocariry: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pena, December, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26098a, Holotype. Disrrisution: Known only from the type locality. Navicula carvajaliana var. carvajaliana sp. nov. Bie i abigss 1, 2513 Valva lineari ad linearem-lanceolatam. Apicibus mutabilibus in forma rotundis, cuneatis aut rostratis. Pseudosepta brevi distende supra apices. Area axiali distincta semper ferme minus quam uno quadrante latitudinis valvae. Area media mutabili, transversa formante fasciam saltem in uno latere valvae, altero latere saepe cum uno aut duobus stris positis late. Stris radiatis leniter in media parte valvae, parallelis aut conventis leniter ad apices. Angulo striarum inique facto valve concava leniter ad aream axialem. In formis angustis strus paene pa- rallelis et valva non concava ad aream axialem. Strus fractis inaequaliter, punctis obscurissimis, s1 adsunt. Strius, 14 in 10 in media parte valvae ad 18 in 10 ad apices valvae; longitudo, 36-70; latitudo, 8-15p. 46 Postilla Yale Peabody Museum No. 49 Valve linear to linear-lanceolate. Apices variable in shape, rounded, wedge-shaped or rostrate. A short pseudoseptum ex- tending over each of the apices. Axial area distinct, usually less than one-fourth the width of the valve. Central area vari- able, transverse, forming a fascia at least on one side of the valve, the other side often with one or two widely placed striae. Striae slightly radiate in center of valve, parallel or slightly convergent at the apices. Angle of striae partially caused by the valve being slightly concave toward the axial area. In nar- row forms the striae almost parallel and valve not concave toward axial area. Striae irreguiarly broken. Puncta if present very indistinct. Striae, 14 in 10 in center of valve to 18 in 10p at ends of valve; length, 36-70”; breadth, 8-15. This taxon is highly variable as to the shape of the valve, and one would recognize some of the variations as separate subspecies or varieties if they did not intergrade into each other. On plate 1, figs. 1, 2, 3 are illustrations showing some of the extremes of variation of these intergrading populations. Some specimens have been found, that are not illustrated, in which the apices of the valve are not drawn out, but the valve simply narrows to a rounded end. This species is a member of that group of taxa which are intermediate between Stauroneis and Navicula, that is, it seems to be closely related to S. thermicola (Peters.) Lund (New Phytol., 45(1): 61, figs. 3 K-AA, 1946) and Navicula incom- posita Hagelstein (New York Acad. Sci., Sci. Surv. Porto Rico & Virgin Isl., §(3) : 286, pl. 7, fig. 2, 1939). This taxon resem- bles Stauroneis in that the specimens have a transverse hyaline area which in girdle view seems to be slightly thickened, par- ticularly at the central nodule of the valve. However, it does not have striae which are radiate at the ends of the valve and resolvable into puncta, which characters are typically asso- ciated with species belonging to the genus Stauroneis. This species has a pseudoseptum at the ends of the valve which is found in species belonging to the genera Stauroneis and Navi- cula. Since I cannot be sure that the central area is a true stauros and since the striae, although breaking into irregular pieces, do not resolve into puncta and are slightly convergent at the ends, it seems wiser to place this species in the genus Navicula. June 23,1961 Antofagasta Ranges of Chile and Bolivia 47 This taxon is also related to Navicula allorgei Manguin (Algues Guadeloupe, p. 58, pl. 3, fig. 51, 1952) but differs in the angle of the striae and the usual presence of a distinct fascia at least on one side of the valve. It is also larger and the shape is more variable. Tyre Ltocauiry: Bolivia, Puna de Atacama, Laguna Colo- “ado, alt. 4,400 m, coll. Luis Pena, December, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26098a, Holotype, fig. 1; Cotypes, figs. 2, 3. Distrisution: In addition to the type locality: Chile, Ata- ‘ama, Carvajal Lake, alt. 2,400 m, coll, Luis Pena, July, 1957 (A-G.C. 26100) ; Atacama, Salt Lake, alt. 2,400 m, coll. Luis Pefia, July, 1957 (A-G.C. 26099a, b). Navicula carvajaliana yar. attenuata var. nov. Pl. 1, Fig. 12 Valva lineari lanceolata, apicibus rotundis attenuatis. Pseu- doseptis praesentibus. Area axiali angusta, distincta. Area media fascia transversa interdum stria praesenti in ea in uno latere valvae. Striis radiatis leniter in media parte valvae, plus aut minus convenientibus ad apices. Stris intermissis inaequa- liter sed non punctatis. Striis, 13 in 10 in media parte valvae usque 16 in 10h ad apices; longitudo, 44-601 ; latitudo, 6-11. Valve linear-lanceolate with attenuated, rounded apices. Pseudosepta present. Axial area narrow, distinct. Central area a transverse fascia sometimes with a stria present in it on one side of the valve. Striae slightly radiate in the center of the valve, more or less convergent at the ends. Striae irregularly interrupted but not punctate. Striae, 13 in 10» in middle part of valve to 16 in 10p at apices; length, 44-60»; breadth, 6-11. This taxon is distinguished from the nominate variety by the narrow attenuated ends of the valve. No intergrades were found between this form and the other forms which are highly variable. Typr Locauiry: Chile, Atacama, Salt Lake, alt. 2,400 m, coll. Luis Pefia, July, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26099a, Holotype. Disrripurion: Known only from the type locality. 48 Postilla Yale Peabody Museum No. 49 Navicula luisii sp. nov. Pl. 1, Fig. 4 Valva lineari lanceolata, apicibus rotundis acutis. Pseudo- septo ad apices. Area axiali distincta circa unum quadrantem latitudinis valvae. Area media fascia transversa. Stris attin- gentibus aream crassioribus quam alterae striae. Strus radiatis leniter per magnam partem valvae, parallelis ad apices. Strus, 14-17 in 104; longitudo, 63-89; latitudo, 9-14y. Valve linear-lanceolate with acute, rounded ends. Pseudo- septum present at apices. Axial area distinct, about one-fourth the width of the valve. Central area a transverse fascia. Striae bordering the central area a little thicker than the other striae. Striae slightly radiate throughout most of the valve, parallel at the apices. Striae, 14-17 in 10/; length, 63-89; breadth, 9-L4pu. This species is distinguished from Navicula carvajaliana by the striae bordering the fascia being distinctly thicker than the other striae and by its larger size. It is near N. in- composita Hagelstein (New York Acad. Sci., Sci. Surv. Porto Rico and Virgin Isl., 8(3) : 386, pl. 7, fig. 2, 1989) but differs in the regularity of the striae which are a little coarser than in N. incomposita. The shape of the valve is also different and the axial area is broader. This species is named for Senor Luis Pena who collected the flamingos which were examined for this study. Type tocauiry: Chile, Atacama, Salt Lake, alt. 2,400 m, coll. Luis Pena, July, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26099a, Holotype. Disrrisutrion: Known only from the type locality. Navicula oppugnata Hust. Navicula oppugnata Hust., Arch. Hydrobiol., 40(4) : 925, pl. 42, fig. 1, 1945. Our specimens differ from those illustrated by Hustedt in that the central area is a little larger and the striae in the middle of the valve are slightly curved. The central area of these specimens is similar to those illustrated by Foged (Folia Limnol. Scandinavica, no. 6, pl. 2, figs. 12-14, 1954). June 23,1961 Antofagasta Ranges of Chile and Bolivia 49 This species was fairly common in Laguna Colorado. Disrrisution: Bolivia, Puna de Atacama, Laguna Colorado, fo) alt. 4,400 m, coll. Luis Petia, December, 1957 (A-G.C. 260982). Navicula pseudosepta sp. nov. Pl. 15; Big. 5 Valva lanceolata, apicibus rostratis ad rostratis-capitatis. Pseudoseptis praesentibus. Area axiali angusta, distincta. Area media non dissimili areae axiali. Striis radiatis in media parte ralvae et parallelis ad convenientibus leniter ad apices. Inter- dum striis in uno latere nodulis mediae crassioribus quam in latere altero. Striis non decernunt in puncta. Strus, 13-15 in 10 ad mediam partem valvae usque 18 in 10 ad apices; longitudo, 51-68»; latitudo, 11-13. Valve lanceolate with rostrate to rostrate-capitate ends. Pseudosepta present. Axial area narrow, distinct. Central area not differentiated from axial area. Striae radiate in the center of the valve and parallel to slightly convergent at the apices. Sometimes striae of one side of the central nodule coarser than on the other side. Striae do not resolve into puncta. Striae, 13-15 in 10 at center of valve to 18 in 10 at apices; length, 51-68; breadth, 11-18p. This species is most closely related to N. carvajaliana which is described in this paper. It differs in the lack of a central area which is a fascia on one or both sides of the valve. Also, the striae are not irregularly broken. It resembles in shape of valve and structure of axial and central areas N. cuspidata var. ambigua (Ehr.) Cl. It differs in the formation of the striae which do not resolve into puncta forming longitudinal lines. Nor are pseudosepta present in N. cuspidata var. ambi gua (Ehr.) Cl. Tyrer tocatiry: Chile, Atacama, Lagunas de Carvajal, alt. 2.400 m, coll. Luis Pena, July, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26100a, Holotype. Disrripurion: Known only from the type locality. 50 Postilla Yale Peabody Museum No. 49 Navicula salinicola yar. boliviana yar. nov. Pl i> Bie sir Valva lineari attenuata ad apices acutos. Area axiali an- gusta, distincta. Fissuris terminalibus distinctis. Area media vix dissimili areae axiali. Striis lineatis parallelis in media parte valvae et conventis leniter ad apices. Striis, 10-12 in 10u; longitudo, 16-30; latitudo, 4-6. Valve linear, narrowed toward the acute apices. Axial area narrow, distinct. Terminal fissures distinct. Central area scarcely differentiated from axial area. Striae lineate, parallel in center of valve and slightly convergent at the apices. Striae, 10-12 in 10/; length, 16-30; breadth, 4-6p. This taxon is very similar to the nominate variety N. salini- cola Hust. (Abhandl. Naturwiss. Verein zu Bremen, 3/: 638, figs. 61-69, 1939) in size, shape, structure of axial and central areas, and type and angle of striature. It differs in the number of striae, which are much coarser in this taxon. Tyrer Locauiry: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pena, December, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26098a, Holotype. Disrrisution: Known only from the type locality. Family CyMBELLACEAE Genus Amphora Ehr. emend. Kitz Amphora atacamana sp. nov. PI 1, Figs Margine dorso valvae convexo valide, margine ventrali con- cavo leniter. Raphe propiore ad marginem dorsum quam ad marginem ventralem. Area axiali angusta, distincta. Area media parva. Strus punctatis subtilissime. Striis, 26-28 in 102; longitudo, 31-52; latitudo, 6-9p. Dorsal margin of valve strongly convex, ventral margin shghtly concave. Raphe nearer to dorsal margin than to ven- tral margin. Axial area narrow, distinct. Central area small. Striac very finely punctate. Striae, 26-28 in 104; length, 31-524; breadth, 6-9n. The valves of the frustules are very convex and therefore the shape varies greatly according to the angle at which they June 23,1961 Antofagasta Ranges of Chile and Bolivia pil lie on the slide. In some specimens the raphe appears very close to the dorsal margin. Sometimes, as in the illustration, a hyaline area is present near the ventral margin. One rarely finds this diatom in girdle view, but in such specimens as I have seen the intercalary zone is complex. This species is distinguished by the convexity of the valve, the raphe which is fairly near the dorsal margin, and the very fine striae. The portion of the valve ventral to the raphe lies in a distinctly different plane from the portion dorsal to the raphe. Type Locauity: Bolivia, Puna de Atacama, Laguna Colo- cado, alt. 4,400 m, coll. Luis Pena, December, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26098a, Holotype. Distrisution: In addition to the type locality: Chile, Ata- cama, Salt Lake, alt. 2,400 m, coll. Luis Pefia, July, 1957 (A-G.C. 26099a); Lagunas de Carvajal, alt. 2,400 m, coll. Luis Pefia, July, 1957 (A-G.C. 26100). Amphora boliviana sp. noy. Pl lakes 9 Margine dorso valvae convexo valide, margine ventrali con- cavo leniter. Area axiali angusta. Area media lanceolata in forma in latere ventrali raphis, obscuro in latere dorso. Striis punctatis crasse in latere dorso raphis, punctatis obscure in latere ventrali. Stris 18 in 10» in latere dorso raphis praeter ad apices ubi sint 20 in 104. Striis in margine ventrali 22-2: in 10h, punctatis obscure; longitudo, 57-58; latitudo, 7-8p. Dorsal margin of valve strongly convex, ventral margin slightly concave. Axial area narrow. Central area lanceolate in shape on ventral side of raphe, indistinct on dorsal side. Striae coarsely punctate on dorsal side of raphe, indistinctly punctate on ventral side. Striae 18 in 10 on dorsal side of raphe except at the apices where they may be 20 in 10. Striae on ventral margin 22-23 in 10, indistinctly punctate; length, 57-58; breadth, 7-8. This species is characterized by the striae which are coarse and distinctly punctate on the dorsal side and fine and in- distinctly punctate on the ventral side. The central area is 52 Postilla Yale Peabody Museum No. 49 absent on the dorsal side of the valve and lanceolate in shape on the ventral side. This species does not seem to be very closely related to any species which I have seen. It belongs in the general group of species comprising A. coffeaeformis Ag. and A. acutiuscula Kiitz. Tyre Locauity: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pena, December, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26098a, Holotype. DisrrisutTion: Known only from the type locality. Amphora carvajaliana sp. nov. Pl. 1, Rien Valva marginibus ventralibus concavis et margine dorso con- vexo valide. Apicibus valvae rostratis-capitatis. Raphe prope marginem ventralem valvae. Area axiali distincta, nulla area media in latere dorso valvae. Striis absentibus in latere ventrali valvae; in latere dorso punctatis crasse, radiatis leniter. Strus, 18 in 10¢; longitudo, 10-28; latitudo, 4-5. Valve with ventral margin concave and dorsal margin strongly convex. Apices of valve rostrate-capitate. Raphe near ventral margin of valve. Axial area distinct, no central area on dorsal side of valve. Striae absent on ventral side of valve; on dorsal side coarsely punctate, slightly radiate. Striae, 18 in 104; length, 10-28; breadth, 4-5y. This species is characterized by the lack of striae on the ventral margin of the valve and the coarsely punctate striae on the dorsal side of the valve. This species, in size and shape, resembles A. banyaiana Greguss and Weber ? (Botanikai Kozlemenyek, 35: 287, pl. 3, fig. 55, 1988) but differs in that it does not have a broadening of the axial area into a recog- nizable central area on the dorsal side of the valve, and the striae are distinctly and coarsely punctate. It also resembles A, turgida Greg. (Trans. Roy. Soc. Edinb., 2/(4) : 510, pl. 12, fig. 68, 1857) but differs in the striae which are coarsely punctate and finer. Tyre tocaritry: Chile, Atacama, Lagunas de Carvajal, alt. 2,400 m, coll. Luis Pena, July, 1957. June 23,1961 Antofagasta Ranges of Chile and Bolivia 53 SPECIMEN ILLUSTRATED: A-G.C. 26100a, Holotype. Disrrinution: In addition to the type locality: Chile, Ata- cama, Salt Lake, alt. 2,400 m, coll. Luis Pena, July, 1957 (A-G.C. 260992). Family NirzscHIAcEAE Genus Nitzschia Hass Nitzsehia aecedens yar. chilensis yar. nov. Pl, 1, Eig. 6 Valve linear with rounded ends. Keel puncta short, distinct, distinctis, 12-13 in 10; striis distinctis, non decretis facile in puncta, 26-28 in 10; longitudo, 56-90; latitudo, 5-6z. Valve linear with rounded end. Keel puncta short, distinct, 12-18 in 10. Striae distinct, not easily resolved into puncta, 26-28 in 10; length, 56-90»; breadth, 5-6p. This variety differs from the nominate variety (Hust., Abh. Naturw. Ver. Bremen, 3/: 663, fig. 115, 1939) in the more rounded apices, size of the valve, and the fine keel puncta. Type Locauiry: Chile, Atacama, Salt Lake, alt. 2,400 m, coll. Luis Pena, July, 1957. SPECIMEN ILLUSTRATED: A-G.C. 26099a, Holotype. Disrrinution: In addition to the type locality: Chile, Ata- ‘ama, Lagunas de Carvajal, alt. 2,400 m, coll. Luis Pena, July, 1957 (A-G.C. 26100). Nitzschia amphibia Grun. Nitzschia amphibia Grun., Verh. Zool.-Bot. Ges. Wien, 12: 574, pl. 18, figs. 23 a-c, 1862. This species was common in the collection. Disrrinution: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pefia, December, 1957 (A-G.C. 26098a). 54 Postilla Yale Peabody Museum No. 49 Nitzschia epithemoides Grun. in Cl. and Grun. Nitzschia epithemoides Grun. in Cl. and Grun., K. Svenska Vet.-Akad. Handl., ser. 2, 17(2): 82, 1880. This species was fairly common in the collections. It is a brackish to marine species. Disrrisution: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pefia, December, 1957 (A-G.C. 26098a). Chile, Atacama, Salt Lake, alt, 2,400 m, coll. Luis Pefia, July, 1957 (A-G.C. 26099a); Lagunas de Carvayjal, alt. 2,400 m, coll. Luis Pefia, July, 1957 (A-G.C. 25100). Nitzschia hungarica Grun. Nitzschia hungarica Grun., Verh. Zool.-Bot. Ges. Wien, 1/2: 568, pl. 18, figs. 31 a-b, 1862. This is a brackish to fresh-water species. It was fairly fre- quent in the one collection. Disrrisuttion: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pena, December, 1957 (A-G.C. 26098a). Nitzschia palea (Kiitz.) W. Sm. Nitzschia palea (Kitz.) W. Sm., Syn. British Diat., 2: 89, 1856. This species can stand a great variety of water conditions and often develops in large masses in polluted water. Disrrisution: Bolivia, Puna de Atacama, Laguna Colo- rado, alt. 4,400 m, coll. Luis Pena, December, 1957 (A-G.C. 26098a). June 23,1961 Antofagasta Ranges of Chile and Bolivia DD LaireraAtTuRe CIvrepD Jenkins, P. M. 1957. ‘The filter-feeding and food of flamingoes (Phoeni- copteri). Philos. Trans. Roy. Soc. London, Ser. B., (no. 674) 240: 401-495. Ridley, M. W., B. L. Moss, and R. C. Lord Percy. 1955. The food of flamingoes in Kenya Colony. Jour. East Africa Nat. Hist. Soc., 22, no. 5 (97): 147-159. Walcott, F. C. 1925. An expedition to the Laguna Colorado, Southern Bolivia. Geogr. Rev., 15: 345-366. 56 Postilla Yale Peabody Museum No. 49 PLATE 1 Fig. 1, 2, 3 Navicula curvajaliana sp. noy. Fig. 4 Navicula luisii sp. nov. Fig. 5 Navicula pseudosepta sp. nov. Fig. 6 Nitzschia accedens var. chilensis var. nov. Fig. 7 Amphora carvajaliana sp. nov. Fig. 8 Amphora atacamana sp. noy. ice Amphora boliviana sp. noy. Fig. 10 Navicula atacamana sp. noy. Fig. 11 Navicula salinicola var. boliviana yar. nov. Fig. 12 Navicula carvajaliana var. attenuata var. nov. \ W\ AH tH fi il suf 2 aque Hi vit 12 ~~ & - i i ny rite iv. ‘ Le ine a ee Sed — Hee i bhi i _ | boy fd Ls | YALE PEABODY MUSEUM or NaTurRAL History Number 50 June 26, 1961 New Haven, Conn. => THE AVIFAUNA OF MOUNT KATANGLAD S. Ditton RieLey AND D. S. Razor In April-May and again in the last week of December 1960, the Peabody Museum of Yale and Silliman University jointly sponsored a small trip by Professor Rabor’s assistant, Mr. R. B. Gonzales, and a group of Silliman University students to Mount Katanglad, Bukidnon Province, central Mindanao Island. This area had been visited once before by ornithologists of the Danish Philippine Expedition (Salo- monsen, 1953). Mount Katanglad is most interesting as it reaches an altitude of over 7,800 feet above sea level and lies in a range of hills in central Mindanao, somewhat isolated from the high massif of Malindang and Dapiak to the west in the eastern edge of the Zamboanga Peninsula. Likewise, it is separated from the Mount Apo hills of the southeast by the drainage valley of the Mindanao River and from the Diuatan Mountains of the northeast by the similar but narrower valley of the Kalgasan River. Of the endemic subspecies of montane birds of Mindanao found on Mount Katanglad, four prove to belong to more western Malindang forms and six to southeastern Mount Apo or Mount McKinley forms. Twelve others are shared in com- mon with both areas of mountains, while five forms are found to be endemic to this central mountain massif alone. ae 2 Postilla Yale Peabody Museum No. 50 The discovery of two new species, a finch and a fire-tail finch, illustrated herein by Robert Verity Clem, is particularly noteworthy. Accipiter trivirgatus extimus Mayr A pair of crested goshawks from Mount Katanglad have prompted us to re-examine the Mindanao population. We have, therefore, borrowed material kindly loaned by Dr. A. L. Rand of the Chicago Natural History Museum and Mr. R. M. de Schauensee of the Philadelphia Academy of Natural Sci- ences. Mayr (1949) reviewed the races of this species and pointed out the differences between the male and female plumages in the adult. Adult females of extimus are very dark tawny rufous on the breast with a variable amount of rufous sometimes form- ing an almost solid rufous breast-shield as in the male. The flanks are variably barred more lightly or more heavily. When heavily barred they are very close to trivirgatus and in fact would be very difficult to separate except for size, being much smaller. When the underparts are lightly barred, the appear- ance is close to the continental forms except again for size. The Palawan race palawanus Mayr is far more distinct from its neighbors, the female’s pattern of droplets of blackish rufous on the breast being close only to layardi of Ceylon. Males of ewxtimus are far more distinct interalia than females, their pale light-rufous underparts setting them apart from palawanus or trivirgatus with minor variation. The single male from Katanglad is astonishingly dark, approaching javanicus. Two Negros males are very light in color with some- what reduced barring, although they can be matched by a Davao male. A Negros immature tends to be very darkly spotted on the underparts. Altogether, this population shows considerable variation in color which, were it not for its small size, would make it difficult to identify with certainty. Writing of Ceylon and South India birds, Mayr (tom. cit.: 8) questions Whistler’s diagnosis of the difference between females from the two areas. Comparing a female of layardi June 26, 1961 Avifauna of Mount Katanglad 3 from Ceylon with one of peninsulae from Kerala, the differences cited by Whistler are shown to be correct. The Cevlon female has smaller and darker markings on the lowerparts, exactly as pointed out by Whistler. Trichoglossus johnstoniae johnstoniae Hartert A large series from Mount Katanglad belongs to the Mount Apo form rather than pistra Rand and Rabor from Mount Malindang. Prioniturus montanus waterstradti Rothschild Six specimens from Mount Katanglad seem as bright about the head as waterstradti from Mount McKinley and Apo and also similar to birds from Mount Malindang which have been kept separate as malindangensis Mearns by Rand and Rabor (1960). We feel that these Katanglad birds span the slight differences enumerated by Rand and Rabor and that Salo- monsen was right in combining the Mindanao populations. Collocalia esculenta bagobo Hachisuka A male was taken May 1 above 4,200 feet. Cuculus saturatus horsfieldi Moore A female from Katanglad enlarges our collection of this migrant cuckoo from the Philippines to include Luzon, Min- doro, Samar, and Mindanao. Found from sea level to 5,000 feet; all specimens taken in April and early May. Weight: 66 92-5, 106.5 ¢, 2 2.80, 80) zc. Otus bakkamoena everetti (Tweeddale) A pair taken on Mount Katanglad at 4,200 feet are in the rufous phase of this small owl as listed by Delacour and Mayr (1946), agreeing well in size and color with two additional specimens from Davao in the Hachisuka collection. A male from Bohol, boholensis auctorum, agrees in size with the Mindanao birds but is in the gray phase. The race nigrorum 4 Postilla Yale Peabody Museum No. 50 Rand (1950) is a striking one. An adult male in ‘the Yale collection from Cuernos de Negros has a wing measurement of 148 mm; tail, 75; culmen, 20. Mimizuku gurneyi (Tweeddale) A female taken on Mount Katanglad at 4,300 feet repre- sents perhaps the sixth known specimen of this rare species. It measures: wing, 274; tail, 139; culmen (from cere), 26. The description of this genus in Hachisuka (1934) empha- sizes the cere which is indeed tumid and in which the external nares are large and prominent. The bill is heavy, the maxillary tomia are buttressed with a cutting tooth-like point before the downward sweep begins towards the tip, a feature totally unlike the smooth bills of Otus species. The upper surface has a softly mottled appearance, not heavily vermiculate as in Otus, the scapulars ornamented with patches of whitish buff, tipped black, the patches on both inner and outer webs. The back is heavily streaked with black as in the head, the nuchal collar is very broad, the feathers only tipped with black, and the primaries and secondaries have a much reduced, barred pattern. Until more is known of the habits and behavior of this rare owl, we would hope that its distinct appearance and huge size would entitle it to remain as a monotypic genus. Batrachostomus septimus septimus 'Tweeddale A juvenal sexed as a female was collected May 6. Presum- ably just out of the nest, this bird has well-formed wings and tail feathers capable of flight, but the feathers about the head and face are still downy. Overall, the plumage can be charac- terized as juvenal. Some downs, barb downs, carried on the ends of barbs of the juvenal feather tips, may be seen, especi- ally on the throat and undersurface. On the back this bird is indistinctly barred with wavy bars of blackish brown. Some feathers are much darker than others, perhaps indicating a replacement stage. On the undersurface the throat and breast are barred, the belly whitish, the lower tail coverts pale buff. above Serinus mindanensis below Erythrura coloria 6 Postilla Yale Peabody Museum No. 50 Lanius validirostris hachisuka Ripley Two females of this species have wing measurements of 86.5, 90; culmen (from skull), 20, 20. In size, therefore, they agree with both hachisuka Ripley from Mount Apo and quartus Rand and Rabor from Mount Malindang. The-type and one other specimen of hachisuka measure: wing, 87.5, 87.5; culmen (from skull), 20, 19.5. (In the original description [1949 ] I measured the culmen from the beginning of the feathers of the forehead.) The type and unique quartus (1958) measure: wing, 93.5; culmen, 22. Rand and Rabor separated quartus on size and color. The breast and abdomen are whiter, the under tail coverts white, and the flanks richer and deeper rufous. The Katanglad speci- mens are mixed in color. One resembles hachisuka in having a rich rufous wash on the underparts. The other has this con- fined to the flanks. Dr. Rand has kindly examined these and states (in litt.): “One specimen [pale-breasted with rufous flanks] is very similar to the type of quartus, differing only in the 2 mm shorter wing, the slightly more pale gray in the forehead (in quartus the forehead is almost like the back), and the faint rufous wash on the breast.” These wholly unexpected specimens representing 40% of the known specimens of Strong-billed Shrikes from Mindanao, occurring as they do on an isolated mountain in central Mindanao separating eastern Mount Apo from western Mount Malindang, are surprising in bridging exactly the essential color differences which allowed quartus to be described from the west and hachisuka from the east. It appears likely that additional Mindanao material would show that these shrikes are oversplit. Coracina mcgregort Mearns We have examined 39 specimens from Mount Katanglad and 15 from Mount Malindang, and we find individual differences in the shade of the color of the flanks are great enough in each population to prevent assigning them to a geographical local- June 26, 1961 Avifauna of Mount Katanglad a ity. There is no appreciable difference in the color of the breast. In size we note the following: Mount Katanglad wing adult ¢ ¢ 103-108 (104.6), 2 2 99-104 (100.8). Mount Malindang wing adult ¢ ¢ 104-109 (107.2), 2 2 101, 103; These wing measurements imply an overlap of all but 1 millimeter which seems far too small to be significant. Re-ex- amination of Salomonsen’s description and comparison of these specimens forces us to the conclusion that peterseni Salomonsen (1953) should be regarded as a synonym of mcgregort. Katanglad Mt. Soy = > 2) ran) y 4 ys MINDANAO ISLAND 8 Postilla Yale Peabody Museum No. 50 Pericrocotus flammeus gonzalesi subsp.n. Type: 6 ad. (Y.P.M. No. 58896), collected May 10, 1960, by R. B. Gonzales on Mount Katanglad, Malaybalay, Bukid- non Province, Mindanao Island, Philippines. Diacnosis: From johnstoniae of Mount Apo, this form differs in the male by being more richly orange-yellow on undersurface, wing edgings, and tail. Two females appear to bear this color difference out, although in one specimen the difference is only readily observable in the color of the tail. Compared to novus of Luzon, males are paler, far less ver- milion, especially on the rump and tail. Similarly, gonzalesi is far less vermilion than is leytensis. In general then, gonzalesi is a very well-marked intermediate, especially in the males, representing a discontinuous cline in a stage from the lemon- vellow or egg-yellow populations, johnstoniae and marchesae of Sulu, and the rich vermilion orange of novus and leytensis. MEASUREMENTS: Wing, ¢ ¢ 78, 83, 2 2 78, 80; tail, ¢ 2 76, 83, 2 80. Rance: Mount Katanglad, central Mindanao from 4,000 to 5,000 feet. Turdus poliocephalus hatanglad Salomonsen Salomonsen’s description (1953) brings out very well the striking characters of this well-marked subspecies. Zoothera andromedae 'Temminck Another record for Mindanao, taken in May and December. Ptilocicichla mindanensis mindanensis (Blasius) A single female of this little-known form was taken December 26, between 4,500 and 5,200 feet. Macronus striaticeps mearnsi Deignan A small series of this species confirms the dark-rufous tone of the montane forms of striaticeps from Mindanao. It appears as if the Katanglad birds are even darker than those from Mount Apo and Mount Malindang. June 26, 1961 Avifauna of Mount Katanglad 9 Bradypterus caudatus unicolor (Hartert) Two females were taken on Mount Katanglad April 24 and May 9. They measure: wing, 57, 61; tail, 67, 77; culmen, 15, 14.5. These two specimens are the first of this rare species received at Yale. The bird with smaller measurements is pre- sumably subadult. In any case, the throat is buffy with reduced white patch, no blackish spotting, and the gray of the breast is patchy. It compares favorably with the type of wnicolor which is also in immature plumage according to Dr. Amadon (in litt.) who has kindly compared our specimens in New York. The adult bird appears indistinguishable from malindangensis (Mearns) according to the plate in Hachisuka (19385). Under the circumstances, it seems wiser to combine the Mindanao populations under the oldest name wnicolor for that island pending securing additional material from the Philippines. Megalurus palustris forbesi Bangs Taken at 4,000 feet. Phylloscopus trivirgatus flavostriatus Salomonsen A series of this form exhibits the distinctive character cited by Salomonsen, darker, more olive crown, more buffy super- ciliary and pale, washed-out looking underparts. Collected from 5,800 to 7,400 feet above sea level. Phylloscopus olivaceus olivaceus Moseley Apparently common from 4,200 to 5,500 feet. Phylloscopus borealis borealis (Blasius ) Found from 4,200 to 5,200 feet. Orthotomus cucullatus heterolaemus (Mearns) Apparently common in the forest from 4,200 to 6,200 feet. Rhinomyias gularis goodfellowi Ogilvie-Grant Two males and a female of this little-known form were taken on Mount Katanglad in April at 6,200 feet. The males with wing measurements of 95, 96 are larger than the female (wing 10 Postilla Yale Peabody Museum No. 50 93.5), but otherwise indistinguishable, being perhaps only a trace darker, more slaty on the back. Muscicapa hyperythra montigena (Mearns) Three males and five females belong to the rufous-tailed Mount Apo form of this thicket flycatcher. There appears to be no difference in size or color between these birds and the latter race. Muscicapa panayensis nigriloris (Hartert) Apparently common above +,200 feet on Mount Katanglad. Muscicapa mugimahi Temminck A female from Mount Katanglad taken December 22 repre- sents our second specimen of this migrant flycatcher from the Philippines, the first being a male from Cuernos de Negros, Negros I., taken December 25, 1952, by Rabor. Rhipidura nigrocinnamomea hutchinsoni Mearns This is a somewhat intermediate population, as might be expected, between hutchinsoni of northwest Mindanao, Mount Bliss, and Mount Malindang, and southeast Mindanao, nigro- cinnamomea from Mount Apo. In a series of 22 specimens, all have a more or less broad band of white across the forehead, although the makeup of the skins is often poor in this region. However, 2 of the 22 show traces of white on the upper breast. One of these has the white area as well-developed as typical nigrocinnamomea. The other is paler only on the upper breast. Sitta frontalis apo Hachisuka A series of 25 specimens from Katanglad is nearer the south- eastern Mindanao apo, of which the type is in the Ripley collection at Yale. One Katanglad bird, a male, is as dark below and washed with lilac as Rand and Rabor’s zamboanga (1957). Three specimens of the latter from Mount Malindang show a considerable range of color from very dark washed with lilac below to closely similar to our Katanglad and Mount Apo birds. Evidently, this is a somewhat variable population. June 26, 1961 Avifauna of Mount Katanglad ie Rhabdornis inornatus zamboanga Rand and Rabor The smaller creeper from Mount Katanglad appears to match closely zamboanga from Mount Malindang, although here we lack the race alaris from Mount McKinley. Dicaeum anthonyt kampalili Manuel and Gilliard As Salomonsen has pointed out (1960), the Katanglad population agrees with southeastern Mindanao hampalili. Our 4 specimens are smaller than those measured by Salomonsen and are, therefore, closer in size to the race named by Manuel and Gilliard. Measurements: wing, $ 55, 57.5; 2 56, 57. At present there appear to be 10 known specimens of this species. We are grateful to Dr. Gilliard for comparing our male with the American Museum specimen of ham palili. Dicaeum ignipectus apo Hartert Five males and two females of this rare form were collected at the 4,200 foot level. Nectarinia jugularis jugularis (Linnaeus ) A pair were collected at 4,200 feet, and agree well with lowland specimens. Aethopyga primigena primigena (Hachisuka) A series of 33 specimens from Mount Katanglad agrees perfectly with the type and another male in the Ripley collec- tion at Yale from Mount Apo. These birds were found up to 6,000 feet. An immature specimen taken April 3 has a paler throat with more pronounced yellow on the longitudinal central streak and more noticeable yellow breast spot. The vellow color on the flanks and underparts is paler, also more citrine than in adult examples. Aethopyga boltoni malindangensis Rand and Rabor These birds are closer to the west Mindanao race than to that of Mount Apo. They are somewhat darker below, the feathers of the breasts of males and females having pronounced greenish-olive centers, but above they are indistinguishable, sharing with malindangensis the tendency to iridescence on We Postilla Yale Peabody Museum No. 50 the head which is nearly absent in typical boltoni. It seems best, then, to keep this population combined with that of Malindang. Arachnothera clarae malindangensis Rand and Rabor This newly described form (1957) proves to be represented on Mount Katanglad from whence Gonzales has sent one male taken March 20 at 4,200 feet in breeding condition, and a further 5 specimens in December. Zosterops montana montana Bonaparte Apparently very common from 4,000 to 6,200 feet. We can- not distinguish the slight differences between these birds and those of Mount Apo cited by Salomonsen (1953). Apoia goodfellowt goodfellowi (Hartert) This mountain whiteye belongs to the Mount Apo subspecies and was found from 4,200 to 7,400 feet. Hypocryptadius cinnamomeus Hartert A series of 34 specimens from Mount Katanglad proves to be intermediate when compared with 7 specimens from Mount Apo (topotypical cinnamomeus) and 17 specimens from Mount Malindang (malindangensis Rand and Rabor). The latter form was described (1957) as being “like citnnamomeus of Mount Apo but upperparts brighter cinnamon rufous; breast tinged with brighter cinnamon; abdomen and under tail coverts whiter (less grayish).” One individual from Mount Apo is as bright and whitish below as any specimen from Mount Malindang. Katanglad birds are in general rather somberly colored below, more grayish throughout, but on the upperparts they are indistinguishable from those of Mount Malindang. These differences may be described below: Upperparts Lowerparts AMOS cricnrae rsp oousearieirs © tele, Fee darker? darker (86% ) Ka tan pla de jemicutnecoccroratciey lighter darker Malindanga ae aceecria eae : lighter lighter June 26, 1961 Avifauna of Mount Katanglad 13 If, then, these characters are to be taken at face value, the Katanglad birds should be combined with the Malindang pop- ulation on the basis of the color of the upperparts, and with the Apo population on the basis of the lowerparts. There is no difference in size. As the differences at best seem relatively slight, perhaps indicative only, faced with this solomonian choice it seems wisest to us to include all the Mindanao birds under a single name and revert to a monotypic species. Birds were collected from 4,200 to 6,200 feet. Serinus mindanensis, sp.n. Tyre: 6 ad. (Y.P.M. No. 58898), collected Aprjl 19, 1960, by R. B. Gonzales at Malaybalay, Mount Katanglad, Bukidnon Province, Mindanao Island, Philippines. Diacnosis: Upperparts blackish-brown, the margins of the feathers indistinctly and widely edged with greenish olive; forecrown extending very nearly to the base of the bill, golden vellow, reaching below to the cheeks, throat, and breast; no white ring around the eye; greater median and lesser wing coverts, rump and upper tail coverts broadly edged with golden yellow; underparts including under tail coverts dull white, the flank feathers narrowly streaked with a central streak of blackish brown; primaries and rectrices black, a faint trace of a yellowish edging on the outer margin of the median portion of the sixth and seventh primary. Bill appar- ently olive horn-colored, feet dark-brown. The bill of this species is remarkably stout and arched, far more so in proportion than in estheraec. The maxillary tomia are angled and thickened to produce a dentate bulge midway from the angle to the tip. This gives a pronounced cutting mechanism to the mid-point of the comissural line. MeasurEMEnTs: Wing, 70; tail, 49; culmen, 9 mm. Rance: Known only from Mount Katangiad, central Min- danao, southern Philippines. 14 Postilla Yale Peabody Museum No. 50 Remarks: Serinus mindanensis, while obviously close te estherae, is best kept in a superspecies. The species estherae divides into 3 subspecies as follows: (a) vanderbilti (de Schaunensee), of which ripley: (Chasen) is a synonym, Mount Loser area, Atjeh, north Sumatra, 7,000 feet above sea level: (b) estherae (Finsch), west Java on Mount Pangrango and Poentjakpas near Bogor 4,300 to 6,000 feet and perhaps Mount Telemojo south of Semarang in central Java; and (c) orientalis (Chasen, 1940) east Java on Mount Ajekajek, Tengger mountains over 7,000 feet. All these races are very similar to each other, differing only in size, tone of color, and size of the white ring around the eye. These populations are roughly equivalent to each other; the sub-specific category is unequivocal. In contrast, mindan- ensis is distinctly different in pattern and color, not equivalent, and to merge it with the races of estherae would be decidedly ambiguous. This specimen has prompted us to look again at the relation- ships of these isolated montane relict forms. The most recent note on the taxonomy of estherae is that of Delacour (1946) who remarked simply: ‘The species estherae is certainly not referable to the genus Serinus, but to Carduelis, its nearest relative being C. (=Hypacanthis) monguilloti from the moun- tains of southern Annam.” Hypacanthis (type, spinoides) had been combined earlier with Carduelis by Mayr, among others, who remarked (1941): “‘As far as the genera Hypacanthis and Spinus are concerned, nobody has yet brought forward any valid reason why they should not be united with Carduelis, as Hartert proposed more than thirty years ago.” Hartert (1910) indeed proposed that the Goldfinch be united with the Siskins, Linnet, Twite, and American Gold- finches on the basis of bill shape, color pattern, similarity of wing and tail, and stout feet. This suggestion was followed in the British Handbook (1938 et seq.) in which the genus Carduelis includes, besides the Goldfinch, the Siskin, the Twite, the Linnet, and the Redpoll. These latter birds, in which the wing pattern differs, browns and pinkish reds predominate in June 26, 1961 Avifauna of Mount Katanglad 15 the plumage, and in which the pale-edged forked tail appears longer in proportion to the wing, have been kept separate in Acanthis more recently by Vaurie (1959). All of these species have a distinctive bill, almost conical in shape with a tendency to a thickened, swollen base. In contrast Serimus, the Serin, the Citril, and the Gold-Fronted Finch, §. pusillus, all have short, stubby, thick bills, with the culmen distinctly curved, not straight. After looking over these species of the Palearctic, it seems impossible to align the species from Java, Sumatra, and Mindanao with Carduelis monguilloti or its relatives, ambigua and spinotdes. Some of the differences may be expressed below: Longest Wing Tail Primary Patch Patch | Top of Head Bill Carduelis first; not | present on| present | dark, only sharp, superspecies always primaries yellow conical superciliary estherae, second and | absent absent | yellow swollen, mindanensis third bullfinch-like Once a careful examination of the specimens is made, the island birds with bills as tumid and curved as in the genus Carpodacus or Pyrrhula, with an entirely different wing pat- tern, with areas of yellow round the head and breast in a very different arrangement, with the yellow edgings of the rump feathers carried right out throughout the upper tail coverts, it can be seen that they are strikingly different. These species are as different from Carduelis in their own right as is Rhynchostruthus of the Somali arid zone of northeastern Africa and southwest Arabia. If these birds do not fit in Carduelis, could they then fit in some other Cardueline genus? The first description of estherae (Finsch, 1902) placed it in Crithagra. The type of this genus is sulphuratus of South Africa, and the genus is now considered synonymous with Serinus. In the arrangement of the pri- maries, second and third longest, they fit in better with Serinus than with Carduelis. Serinus, found in the Palearctic and in Africa, is characterized as having a very short, thick, tumid bill, culmen distinctly curved, tail deeply emarginated. The 16 Postilla Yale Peabody Museum No. 50 wing is long, longer than the tail, the wing/tail index varying from 75-82%. Serinus estherae, mindanensis small, tumid more rounded, more tumid, larger in proportion tendency to super- ciliary ; crown patch in two species Wing Covert Edgings .. Outer Web of streaked forked, wing/tail index 75-82% yellow, usually not on upper tail coverts normally match color of the back edged with yellow no eye stripe; crown patch plain barely forked 67-10% yellow extending through upper tail coverts distinct from back almost totally reduced Primaries or absent The primary arrangement of Serinus, with the exception of flaviventris in which the first primary is longest, inclines us to feel that the tropical island species are nearer to Serinus than to Carduelis contra Delacour. The bill shape is far more similar, especially to some of the African members of the Serinus throng. All these are less strongly hooked or rounded, or indeed as swollen in the mid-section of the tomium of the maxilla. In some ways these tropical island species strongly resemble Carpodacus. The bill, although more bullfinch-like, resembles that of the common Asian species erythrinus. The plumage patterns, if yellows are substituted for reds, are not too dissimilar. In the same way Rhynchostruthus bears a certain re- semblance to Rhodopechys. Callacanthis burtoni of the Hima- layas bears an even closer resemblance to Carduelis carduelis, and in this case it is possible to hazard a guess that burtoni may be a Palearctic relict which was evolved from an ances- tral goldfinch. No such guess seems readily apparent or in order for Rhynchostruthus on the other hand. June 26, 1961 Avifauna of Mount Katanglad 17 Our opinion, then, is that these birds belong with the ex- panded genus Serinus and that the Philippine bird, though sympatric, differs as significantly from the three known closely allied races of estherae as do the three sympatric forms of Carduelis, spinoides, ambigua, and monguilloti, kept by modern workers, Mayr (tom. cit. supra) and Vaurie (1949), as separate species in a superspecies. Pyrrhula leucogenys coriaria, subsp.n. iiver=6 ad. (Y.r-M: No. 58899), collected April 11, 1960, by R. B. Gonzales at Malaybalay, Mount Katanglad, Bukidnon Province, Mindanao Island, Philippines. Diacnosis: From apo Hachisuka (1941), of which the type and one other specimen are in the Ripley collection, this form differs by being darker, more suffused with olive, “mummy brown” rather than “prouts brown” both above and below. From steerei Mearns this form differs by being much darker, the darker smoky brown being of the same grayish tones rather than in the more tawny tone of apo. All these three populations are smaller than leucogenys of Luzon and have all-black bills. MeasurEMENTS: Wing, é 6 74-79, 2 2 76, 78; tail, ¢ 4 64-66.5, 2 2 61.5, 64; exposed culmen, ¢ 6 10-11, ? ? 10, a5. Remarks: It is interesting that this race situated in a central position geographically should be much darker than the two populations it separates, that of Mount Malindang to the west and Mount Apo to the southeast. Taken together running from west to east or vice versa, these populations represent a sharply discontinuous geographical cline. Lonchura malacca jagori (Martens) As Parkes (1958) has pointed out, this variable black- headed population had best be left under the catch-all jagori rather than further split as Salomonsen (1953) has suggested. 18 Postilla Yale Peabody Museum No. 50 Erythrura coloria, sp.n. Tyer: ¢ ad. (Y.P.M. No. 58897), collected March~ 26; 1960, by R. B. Gonzales on Mount Katanglad, Malaybalay, Bukidnon Province, Mindanao Island, Philippines. Diacnosts: This species resembles Erythrura trichroa of the Moluccas and Melanesian areas, thus differing completely from hyperythra and viridifacies, the other known parrot finches of the Philippines. From trichroa it differs in its richer, more intense emerald-green coloration and in the pres- ence of a bright scarlet patch lying behind the blue cheeks and extending from the postocular area down to the sides of the throat. The blue patch covers the forehead and forecrown, the cheeks, and an area immediately behind the eyes. MeasvurEMENTs: 446 6 wing, 51-56, 2? 54.5; tail, 35-38, é 33; culmen (from skull), 11-13, 2 11 mm. Rance: Known only from Mount Katanglad, Bukidnon Province, central Mindanao, Philippines. Remarks: This species was found in small clearings or openings in the forest from 4,200-4,500 feet, often perching on grass close to the ground. The birds were quiet and moved singly or in pairs; when disturbed, flying to the nearest dense growth, often near small streams, perching on branches close to the ground. ADDITIONAL Species TAKEN ON Mount KaTANGLAD Pernis philorhynchus philippensis Mayr Butastur indicus (Gmelin) Accipiter virgatus confusus Hartert Hieraaétus hieneri formosus Stresemann Spilornis cheela holospilus (Vigors) Falco severus Horsfield Gallus gallus Linnaeus Ptilinopus leucotis brevirostris ('Tweeddale) Ptilinopus amethystina mindanensis (Manuel) June 26, 1961 Avifauna of Mount Katanglad Ptilinopus occipitalis Gray Ducula carola mindanensis (Ogilvie-Grant) 19 Columba vitiensis griseogularis Walden and Layard Macropygia phasianella tenuirostris Bonaparte Loriculus philippensis apicalis Souancé Cuculus fugax pectoralis Cabanis and Heine Cacomantis variolosus sepulcralis (Miiller) Surniculus lugubris velutinus Sharpe Centropus viridis viridis (Scopoli) Eurostopodus macrotis macrotis (Vigors) Hemiprocne comata comata (Temminck) Harpactes ardens ardens (Temminck) Halcyon smyrnensis gularis (Kuhl) Merops viridis americanus Miiller Eurystomus orientalis cyanicollis Vieillot Penelopides panini affinis Tweeddale Aceros leucocephalus leucocephalus (Vieillot) Megalaima haemacephala haemacephala (Miller) Dryocopus javensis multilunatus (McGregor) Dendrocopus maculatus fulvifasciatus (Hargitt) Chrysocolaptes lucidus lucidus (Scopoli) Lanius cristatus lucionensis Linnaeus Lanius schach nasutus Scopoli Oriolus chinensis suluensis Sharpe Oriolus xanthonotus samarensis Steere Dicrurus hottentottus striatus Tweeddale Artamus leucorhynchus leucorhynchus (Linnaeus) Aplonis minor todayensis (Mearns) Sarcops calvus melanonotus Ogilvie-Grant Basilorns miranda (Hartert) Corvus macrorhynchus philippinus Bonaparte Coracina striata kochii (Kutter) Pycnonotus goiavier suluensis Mearns Hypsipetes philippinus saturatior (Hartert) Muscicapa griseisticta griseisticta (Swinhoe) Muscicapa westermanni westermanni (Sharpe) Muscicapa panayensis nigriloris (Hartert) Culicicapa helianthea panayensis (Sharpe) 20 Postilla Yale Peabody Museum No. 50 Monarcha azurea azurea (Boddaert) Pachycephala philippensis apoensis (Mearns) Orthotomus atrogularis frontalis Sharpe Turdus obscurus Gmelin Dicaeum hypoleucum hypoleucum Sharpe Dicaeum nigrilore nigrilore Hartert Dicaeum bicolor bicolor (Bourns and Worcester ) Dicaeum pygmaeum davao Mearns Lonchura leucogastra manueli Parkes LireraturE CITED Chasen, F. N. 1940. Notes on some Javan Birds. Treubia 17: 263-4. Delacour, J. 1946. Notes on the Taxonomy of the Birds of Malaysia. Zoologica 31: 4. and Mayr, E. 1946. Birds of the Philippines. Macmillan, New York. Finsch, O. 1902. Notes Leyden Museum 23: 151. Hachisuka, M. 1934-1935. The Birds of the Philippine Islands (3, 4): 50, 402. Witherby, London. 1941. Further Contributions to the Ornithology of the Philippine Islands. Tori 11 (51-52): 88. Hartert, E. 1910. Die Vogel der paliktischen Fauna 1: 65-6. Mayr, E. (with Stanford, J. K.) 1941. The Vernay-Cutting Expedition to Northern Burma. Ibis: 361. 1949. Geographical Variation in Accipiter trivirgatus, Amer. Mus. Novit. no. 1415: 1-12. Parkes, K. C. 1958. Taxonomy and nomenclature of three species of Lonchura (Aves: Estrildinae). Proc. U. S. Nat. Mus. 108: 291. Rand, A. L. 1950. A new race of owl, Otus bakkanoena, from Negros, Philippine Islands. Nat. Hist. Misc. 72: 1-5. and Rabor, D. S. 1957. New Birds from the Philippines. Fieldiana: Zool. 42: 18. 1958. The races of the Shrike, Lanius validirostris. Fieldiana: Zool. 39: 85. 1960. Birds of the Philippine Islands: Siquijor, Mount Malin- dang, Bohol, and Samar. Fieldiana: Zool. 35: 260-308. Ripley, S. Dillon. 1949. A new race of Shrike from the Philippines. Bull. Brit. Orn. Cl. 69: 121. Salomonsen, F. 1953. Miscellgneous notes on Philippine Birds. Vidensk. Medd. fra Dansk naturl. Foren. 115: 272-281. 1960. Notes on Flowerpeckers (Aves, Dicaeidae) 2. The Primitive Species of the Genus Dicaeum. Amer. Mus. Novit. no. 1991: 22-3. ee C. 1949. Notes on some Asiatic Finches. Amer. Mus. Novit. no. 1424: 8. 1959. The Birds of the Palearctic Fauna: 611. London. Witherby, H. F., Jourdain, F. C. R., Ticehurst, N. F., and Tucker, B. W. 1938 et. seq. The Handbook of British Birds 1: 57. a i he (C3 pe : Harvard MCZ Libra ~ IOI Many 3 2044 066 305 228 ee rai are eee ae ett ee ae tee peste Ty “3 barat uae POW eer aed suas ape ana pare ae Te att ~ ea 15S fi ee one a ed oats