sign in sign up
<<
Home , Annan River, Archer River, Austronomus, Cape York Peninsula, Coen River, Gulf Country

tRESULTS OF THE ARCHBOLD it EXPEDITIONS. NO. 66

- OF , WVITH NOTES ON THE, OCCURRENCE OF RAIN FOREST IN

G. H. H. TATE

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME9F8 ARTICLE 7 NEW YORK: 1952

MAMMALS OF CAPE YORK PENINSULA

RESULTS OF THE ARCHBOLD EXPEDITIONS. NO. 66

MAMMALS OF CAPE YORK PENINSULA, WITH NOTES ON THE OCCURRENCE OF RAIN FOREST IN QUEENSLAND

G. H. H. TATE Curator Department of Mammals

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME 98: ARTICLE 7 NEW YORK: 1952 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY Volume 98, article 7, pages 563-616, text figures 1, 2, tables 1-6

Issued March 25, 1952

Price: $.75 a copy INTRODUCTION SPECIAL INTEREST in the Cape York Penin- The number of forms now demonstrable in sula arises from the fact that it has served as the mammalian fauna of the Cape York the main highway of communication between Peninsula are approximately as follows: the faunas of and . Monotremata, two (Ornithorhynchus and Further investigation of the problems posed Tachyglossus); Marsupialia, 45; Rodentia, by the was carried out in 1948 by the native rats and native mice, 15; Chiroptera, Archbold Cape York Expedition of the 27; there is also the . American Museum of Natural History, In the course of the general studies re- New York' The specimens listed were pro- quired for preparation of this report it has cured by the Archbold expedition, unless been necessary to describe a new . otherwise stated. In addition, a general recon- The following abbreviations refer to the naissance trip through Queensland was made museums in which specimens are deposited: by the author. The results of that reconnais- sance are inserted parenthetically. Through A.M.N.H., the American Museum of Natural the kindness of the Curators of Mammals at History the Museum of Comparative Zodlogy, the C.N.H.M., Chicago Natural History Museum Chicago Natural History Museum, and the M.C.Z., Museum of Comparative Zoology United States National Museum, I have U.S.N.M., the United States National Museum been able to study and integrate into this re- port the collections made by Raven, Hoy, A map showing the localities in the Cape Darlington, and Mrs. Scott (n6e Neuhauser). York area is included (fig. 1).

PHYSIOGRAPHIC DESCRIPTION OF THE CAPE YORK AREA A short description of the general environ- the eastern side, the eastern run-off of which ment is offered before an examination of the is short and steep, the western slope very distributional patterns of the mammals of long and forming an almost imperceptible the Cape York Peninsula is made. The CApe gradient to the . The York Peninsula can be delimited from the main drainage consequently flows to the main mass of Australia by a line drawn from west, while to the east the are com- the southernmost shore line of the Gulf of paratively small and rapid. The divide at Carpentaria eastward to the Pacific coast. the northern tip of the Peninsula reaches al- This definition is arbitrary and conforms to titudes of only 200 to 300 feet, but becomes no geological or biogeographical boundary. gradually higher, in the south rising to peaks Thus the peninsula so described includes the of from 3000 to 5000 feet. This peninsular northern part of the . divide becomes increasingly complicated by The area is roughly triangular. It has two the presence of lateral ranges and spurs. long sides converging northward to Cape Many of the highest peaks stand on these York proper and stands on a shorter base- side ranges instead of on the Great Dividing the east-west line described above. Its length Range itself. from south to north is approximately 500 The climate of the Peninsula is monsoonal. miles. Its width at the base may be taken as During the period when the heat equator is 400 miles. It lies between latitudes 110 and north of the true equator (April to October) 170 S. the dry season is generally effective, though The physiography of Cape York Penin- often, as elsewhere in the , there ex- sula, sketched in broad terms, centers on a ists a rag both in the onset of the dry and in north-south divide placed relatively close to its termination. However, by May, the 1 A preliminary sketch of the travels of the party can southeast trade wind controls the climate, be found in L. J. Brass, "Camps on Cape York" (1949, dropping whatever moisture it carries when Nat. Hist., New York, vol. 58, no. 8, pp. 366-372). it reaches the mountains of the east coast 567 568 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98

say

n Ker

- - -- APPROXIMATE EDGE OF RAINFOREST

FIG. 1. Map of northeastern Australia, to show in particular the localities of the Cape York Peninsula. Details of the -Atherton Tableland area are in the inset. andicontinuing across the Peninsula as a east trade wind, modified south of the equa- dry, though sometimes cloud-bearing, wind. tor into the northwest, is felt as a series of Though the heat equator moves south of storms which herald the in the the true equator during late September, the northern parts of the Peninsula. The rains above-mentioned lag (perhaps occasioned by spread week by week farther south. Again the momentum of the southeast wind) often the afore-mentioned lag operates, so that the prolongs the dry season into October and wet may extend well into April. Actually the November. By late November, when the transition is far from uniform and may be doldrums have passed still farther to the complicated by cyclic weather phenomena, south, the effect of the rain-bearing north- including hurricanes. Generally the rivers 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 569 run full and the great western plains of the miles of it are burned over in the name of the Peninsula receive copious rain from Decem- cattle industry so that dormant grasses, ber to April, whereas from May to Novem- which in this way are given a "shot in the ber the rivers dry up and rain seldom falls arm," put forth green sprouts somewhat in west of the main divide. advance of the grass plants on unburned Temperatures are generally fairly uniform, areas. At the same time all humus is de- tropical at the northern tip. The coastal stroyed, as well as most of the older, weaker strip near the base of the Peninsula in winter trees and fallen hollow logs that may afford (July-August) becomes agreeably cool, while cover for . This ocean of open forest in the dry interior the temperature some- surrounds the islands of rain forest and iso- times falls to freezing just before sunrise. lates them from one another. Some 30 miles south of Cooktown, at Ship- Rain forest depends for its continuance ton's Flat, 800 feet above sea level, the tem- upon water, whether of aerial or seepage ori- perature at 6 A.M. in September fell almost gin. It occurs where enough water is present to 500 F. through most of the year. It comprises The geology of such an enormous area as mainly broad-leafed trees which form dense the Cape York Peninsula can scarcely be shade. In it vines and epiphytes are plenti- touched upon in this article. Very large por- ful. It grows in varied form on stony hill- tions of the region are granite; other exten- sides, well-drained flats, or old flood plains sive portions are sandstone. The mineral margining rivers. It generally forms narrow belt, chiefly in the mountainous parts, is com- belts from 10 to 25 miles in width, though as posed of greatlV deformed sediments bearing much as 100 miles in length, on the east- valuable ores of iron, tin, gold, and wolfram. facing slopes of the ranges. Only the combi- In the southwest are extensive deposits of nation of fertile soil and adequately distrib- limestone, and there also great sheets of ba- uted rainfall will allow full development of salt cover many square miles of country. By these rich tropical jungles, which must not decomposition these rocks have produced a be confused with flood-plain forests, covered variety of soils, which are further modified in the wet months by 10 to 15 feet of water, by climatic factors into sands, clays, mucks, or with forests. Both the latter, etc. though they may also receive copious rain, The geology and the climate of the Penin- obtain much of their water by seepage. Rain sula primarily control its vegetation. In the forest is sometimes found continuous with broadest sense there are but two main classes flood forest or mangrove forest, the one of vegetable cover: open forest and rain for- changing gradually or abruptly into the est. Both have numerous subdivisions. Open other. Its larger masses are important both forest, which covers most of the Peninsula, is as the nurseries of Australian timbers and found chiefly in areas where a pronounced because they act as reservoirs for the rich water shortage exists for a number of months variety of life requiring their shelter. of the year. But it is also found locally in The Cape York rain forests do not, as one areas of ample rainfall. Its constituents vary might expect, form a continuous band along widely: iron-bark forest, she-oak forest, the north-south ranges, but are parted into stringy-bark and messmate forest, blood- several distinct masses, or islands, by wood forest, wattle forest, pandanus forest, stretches of open forest. The largest and most even palm forest, according to the dominant important island is the Cooktown-Cairns- types of trees. But most of these kinds of mass at the eastern side of the open forest agree in certain features: the base of the Peninsula, which partly covers trees are widely spaced, tall or short, with the Atherton Tableland and includes the gray-green foliage throwing little shade. great mou\ntains BartJe Frere and Bellenden There is little or no undergrowth. Coarse Ker. This mass is nearly divided into two grasses grow beneath the trees, and the in- parts near Cairns, as is shown in Prescott's tense light of the sun reaches the very roots (1931) vegetation map, by invasion of the of the grasses. Open forest, unfortunately, is open forest from the west at the latitude of easily set on fire, and each year hundreds of Mt. Molloy. The second and less well-known 570 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 island is the McIlwraith (Rocky Scrub)-Iron seems also to contain scattered stands of rain Range mass of rain forest. Separated from forest. the first by 150 miles of open forest, it is Throughout the eastern side of the Penin- nearly continuous for about 100 miles north sula one observes the inter-fingering and the and south, while its width varies from 5 to mosaic patterns made by open forest and 10 miles. Both these big scrubs are the homes rain forest. These irregular pictures can per- of important and distinct of animals. haps be explained on edaphic grounds. Be- As one nears the extreme tip of the Penin- sides these two primary types of vegetation sula, scattered patches of rain forest occur, there also occur here and there heath-like isolated from one another by narrow belts areas, including the ill-defined turkey brush; of open forest. These patches shelter a de- also swamps, sand , rock outcrops, and pauperized remnant of the Iron Range fauna. . Most of those environments are The intervening area, east and southeast of strongly illuminated and without shade. the Jardine , though poorly known,

DISTRIBUTION OF RAIN FOREST IN QUEENSLAND I have been urged to set down the areas in ton on the railroad line westward to Winton. which I have observed rain forest, not only in The forest on the hills west of the Cape York Peninsula but also farther to appears to be chiefly gray-green the south (fig. 2). Prescott (1931),. in his forest, though there may be patches of ever- rain-forest map, separates a patch of rain green forest in pockets or gullies. The hills forest due west of Townsville from the Her- stand far back from the railroad, and I could bert River-Tully-Cairns patch of rain forest. not see them well. It is my belief that these are almost continu- Prescott shows large masses of rain forest ous: that the northern rain forest, crossing northwest of , lying between Gym- the Gorge, extends southeast pie and the northeast slopes of the Bunya along what on my map (fig. 2) is called the Mountains. I can personally attest to the "Seaview Range" to join the rain forests of luxuriance of the growth of the forest on the the Paluma Range and Mt. Spec, which I upper slopes of the Bunyas, where fine stands visited. The Paluma rain forest terminates of mixed hoop pines, Bunya pines, figs, nettle- at the point where the crest of the Paluma trees, and many other species are bathed al- Range, losing altitude as it approaches the most daily with cloud and rain. These condi- , drops below 2000 feet, tions are repeated among the mountains of Some 40 miles southeast of the Burdekin, the Lamington National Park area (not which traverses relatively dry country near visited by me, but well known from photo- its mouth, rain forest is again found behind graphs). Bowen. The coast from Bowen to Mackay is Thus from north to south along the east backed by the nearly isolated mountain Australian coast there are a series of gaps mass known as the Clarke Range. It upper containing open-forest or even plains condi- slopes, at least, appear to carry rain forest. tions, which separate the several patches of Still farther south, near , rain forest. The first is the Coen-Cooktown the Ransby and Berserker ranges rise above gap (150 miles); next comes the Burdekin surrounding country. These too are forest gap (125 miles); then the Mackay-Byfield covered. Rainfall is quite heavy along the gap, possibly less complete (150 miles); then first, particularly near Byfield, but it is less the Rockhampton-Miriamvale gap (north- extensive on the Berserker where periods of west of Bundaberg) where stand the town of extremely dry weather sometimes prevail, so Gladstone and the badly burned Mt. Larcom, that this forest, though giving "total cover," which possibly has a little rain scrub on the is of somewhat more xerophytic character. seaward side (100 miles); and finally the I was told that there exists a little rain forest Brisbane- gap (the extreme east- at Gogango, some 30 miles from Rockhamp- ern part of the ) separating 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 571 the Bunya scrubs from the Lamington scrubs tioned mountains: strongly cliffed horizontal (about 75 miles). sediments between 2000 and 3000 feet above It will be noted from the map (fig. 2) that sea level, which appear to carry only open- to pass around the drainage basins of the forest vegetation. In many other places, Burdekin and Fitzroy rivers the Great Di- notably at Charters Towers, the Great Di- viding Range stands some 250 miles inland viding Range is imperceptible and reduced from the coast. At the point where it again to a great expanse of plains. At Beta, on the turns east to reach the Bunya Mountains, it Winton-Rockhampton railroad it is a rugged, enlarges to become the Carnarvon Range. I intensely dissected mountain-scape, though have seen photographs of these last-men- very dry for much of the year.

FIG. 2. Distribution of rain forest in Queensland. The purely tropical phase reaches its southern limit near Byfield. Dotted lines represent the great Central Range and other divides. DISTRIBUTION OF MAMMALS Ninety forms of mammals are now re- place of origin, whether New Guinea or far- corded from the Cape York Peninsula and ther south, in Australia. They are listed in near-by areas (including the Atherton Table- table 1. land). The distribution of these mammals is It will be seen that the distributional data closely linked to their preferred environ- are incomplete in the case of the Chiroptera. ments and is, also partly governed by their Three forms are associated primarily with TABLE 1 COMPLETE LIST OF THE MAMMALS OF THE CAPE YORK AREA (Records that are extraterritorial are given in brackets.) Ornithorhynchus a. phoxinus rz aq A [Myotis australis] Tachyglossus a. aculeatus r o t A Pipistrellus papuanus p Antechinus f. adustus r s A Chalinolobus rogersi A Antechinus f. godmani r s A [Chalinolobus morio] A Sminthopsis r. lumholtsi o t A Chalinolobus picatus A Sminthopsis m. murina r t A Eptesicus p. caurinus A Satanellus h. predator O S P Scoteinus sanborni A Dasyurops m. gracilis r s A [Scoteinus greyii] Perameles n. pallescens r t A Nyctophilus bifax A Echimypera r. australis r t P Nyctophilus g. pallescens A Isoodon o. peninsulae o t A Miniopterus s. blepotis Isoodon m. torosus o t A Miniopterus australis Trichosurus v. vulpecula o a A Nyctinomus norfolcensis A Trichosurus v. eburacensis o a A Nyctinomus loriae p Trichosurus v. johnstoni r a A [Nyctinomus colonicus] A o. peninsulae r a P [Nyctinomus a. atratus] A Phalanger m. nudicaudatus r a P [Taphosous flaviventris] A t. picata r a P Taphosous nudicluniatus A Dactylopsila t. infumata r a P Taphosous mixtus Eudromicia macrura r a A P Taphozous australis 0 A a. reginae r a A [Taphosous georgianus] A Petaurus b. longicaudatus r o a A P [Taphosous troughtoni] A Acrobates p. frontalis o a A [Macroderma gigas] A p. peregrinus o a A Rhinolophus m. megaphyllus A Pseudocheirus p. incanens o a A Rhinolophus m. ignifer A Pseudocheirus p. oralis o a A Hipposideros b. albanensis [Pseudocheirus rubidus] r a A Hipposideros g. cervinus Pseudocheirus h. herbertensis r a A Hipposideros m. semoni Pseudocheirus h. cinereus r o a A Hipposideros d. reginae Pseudocheirus archeri r a P a. gouldii Pseudockeirus lemuroides r a A Pteropus conspicillatus r A Phascolarcios c. adustus o a A Pteropus scapulatus 0 A Schoinobates minor r a A Dobsonia m. magna 0 [Schoinobatks volans] o a A Nyctimene robinsoni Hypsiprymnodon moschatus r t A Macroglossus 1. nanus Bettongia penicillata o t A Canis f. dingo A Aepyprymnus rufescens o t A Hydromys c. reginae aq t A Petrogale inornata godmani o t A Hydromys c. beccarii aq t A Onychogalea unguifer o t A d. delicatula 0 t A c. eichardti o t A Zysomys argurus 0 t A Dendrolagus lumholtsi r a P g. rattoides 0 a A Dendrolagus bennettianus r a P Melomys c. capensis r s P Thylogale s. stigmatica r t A Melomys c. eboreus r sa P Thylogale s. coxenii r t A Meomys 1. australius 0 s P a. jardinei o t A Uromys c. caudimaculatus r a P Protemnodon parryi o t A Rattus a. coracius r t AP Protemnodon b. apicalis r t A Rattus 1. leucopus r t P [Protemnodon dorsalis] o t A Rattus 1. mcilwraithi r t P r. erubescens o t A Rattus 1. cooktownensis r t P [Macropus rufus] o t A Rattus lacus o t A Macropus c. canguru o t A Rattus g. apex o t A Myotis a. macropus Rattus g. conatus o t A a Explanation of symbols: P, of Papuan origin a, arboreal r, rain-forest habitat A, of Australian origin s, scandent (but not habitually arboreal) aq, aquatic t, terrestrial o, open-forest habitat 572 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 573 water courses and swamps. These are Orni- Pseudocheirus p. incanens (a) thorhynchus a. phoxinus and the greater and c. adustus (a) lesser water rats, Hydromys c. reginae and Bettongia penicillatus (t) H. c. beccarii (=longmani). Tachyglossus a. Aepyprymnus rufescens (t) acanthion occurs indifferently in open forest Onychogalea unguifer (t) and in rain forest. Lagorchestes c. Ieichardti (t) Protemnodon a. jardinei (t) Fifteen forms are apparently derived from Protemnodon parryi (t) Papuan rain forests: Macropus (M.) c. canguru (t) Echimypera r. australis (t) Leggadina d. delicatula (t) Phalanger o. peninsulae (a) Mesembriomys g. rattoides (a) Phalanger m. nudicaudatus (a) Melomys 1. australius (t) Dactylopsika t. picata (a) Rattus lacus (t) Dactylopsila t. infumata (a) Rattus g. apex (t) Eudromicia macrura (a) Rattus g. conatus (t) Pseudocheirus archeri (a) Dendrolagus bennettianus (a) Petaurus breviceps is found chiefly in open Dendrolagus lumholtsi (a) country, though in the mountains of New Meiomys cervinipes capensis (a) Guinea a rain-forest race occurs. Dactylop- Melomys c. eboreus (a) sila, though it is sometimes taken in open Uromys c. caudimaculatus (a) forest, is primarily a Papuan rain-forest ani- Rattus 1. keucopus (t) mal. Melomys lutillus climbs freely about Rattus 1. mcilwraithi (t) grasses and weeds but ascends trees only Rattus 1. cooktownensis (t) rarely. Fifteen forms are apparently indigenous to Three forms of the open forest which pre- Queensland rain forests: fer rocky hillsides are: Antechinusf. adustus (t) Petrogate i. godmani (t) Antechinusf. godmani (t) Macropus (O.) erubescens (t) Sminthopsis murina (t) argurus (t) Dasyurops m. gracilis (t) Perameles n. paUescens (t) Twenty-five species and races of of Trichosurus v. johnstoni (a) the Cape York region can be grouped for con- Petaurus a. reginae (a) venience by their resting habits. Pseudocheirus h. herbertensis (a) Six tree-roosting bats: Pseudocheirus h. cinereus (a) Pteropus a. gouldii Pseudocheirus kemuroides (a) Pteropus conspicillatus Schinobates minor (a) Pteropus scapulatus Hypsiprymnodon moschatus (t) Nyctimene robinsoni Thylogale stigmaticus (t) Syconycteris australis Protemnodon b. apicalis (t) Odontonycteris pygmaeus Rattus a. coracius (t) Trichosurus v. johnstoni of the above list is Nineteen cave (including hollow trees and apparently a rain-forest form directly modi- houses) and crevice bats: fied from its open-forest relatives. Myotis a. macropus Twenty-four forms are indigenous to open Pipistrelus papuanus forest (25 to 40 inches of rain): Chalinolobus rogersi Eptesicus p. caurinus Sminthopsis r. lumholtzi (t) Scoteinus sanborni Satanellus h. predator (t) Nyctophilus bifax Isoodon o. peninsulac (t) Miniopterus blepotis Isoodon m. torosus (t) Miniopterus australis Trichosurus v. vulpecula (a) Nyctinomus norfolcensis Trichosurus v. eburacensis (a) Nyctinomus loriae Petaurus b. longicaudatus (a) Taphozous mixtus Acrobates p. frontalis (a) Taphozous australis Pseudocheirus p. peregrinus (a) Rhinolophus m. megaphyllus 574 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 Rhinolophus m. ignifer Of the bats, almost all have wide trans- Hipposideros b. albanensis Torresian ranges. Obviously all can move Hipposideros g. cervinus about freely. The following are probably Hipposideros m. semoni fairly recent immigrants from New Guinea: Hipposideros d. reginae Nyctimene robinsoni, Pipistrellus papuanus, Dobsonia m. magna Nyctinomus loriae, Taphozous mixtus, Hippo- The majority of these bats were taken in sideros m. semoni, and Dobsonia magna. I dis- caves or mine tunnels in open-forest country. cussed the subject of chiropteran distribution I suspect that a few, Pipistrellus papuanus, recently (Tate, 1946). Nyctophilus bifax, Nyctinomus norfolcensis, Of the , six are rain-forest forms and Hipposideros diadema, prefer rain-forest derived from New Guinea. Rattus a. coracius habitats. is an Australian rain-forest form related to Eleven mammals currently found to the the Papuan rats, and seven are indigenous, south or southwest, which may in time be open-forest forms. found in the Cape York area, are: Most of the non-volant mammals found Protemnodon dorsalis (t) Rockhampton area and on the Peninsula inhabit either the open for- south est or the rain forest, locally termed scrub. A Macropus (M.) rufus (t) Hughendon and south very few, such as Tachyglossus (of Australian Chalinolobus morio Bunya Mountains origin), the aquatic Hydromys (of ancient Scoteinus greyii Pentland and west Paptan origin), and Ornithorhynchus (Aus- Nyctophilus g. paUescens Pentland and west tralian), appear to be equally at home in Nyctinomus colonicus Pentland either. Dactylopsila, the , also Taphozous flaviventer Pentland is found almost as often in open forest as in Taphozous troughtoni Pentland Macroderma gigas Rockhampton rain forest, though if one may judge by its Pteropus poliocephalus Rockhampton (winters New Guinea origin it probably belongs to the there, according to rain-forest assemblage. Ratcliff) The species of the uninterrupted open for- Rattus villosissimus Cloncurry est have extensive ranges. There appear to be few if any definable barriers limiting the Taken by orders, the species and subspe- spread of such species. Where such limiting cies of the mammals of the Cape York Penin- factors seem to occur, as in the case of the sula region comprise: northward limits of Macropus, Petrogale, SPECIES SUBSPECIES Trichosurus, Mesembriomys, and Leggadina Monotremes 2 2% 2 near Wenlock, about latitude 130 S., the con- 37 SO 45 trolling influences are certainly obscure and Chiroptera 26 33 27 possibly complex. Also there is no certainty Rodents 10 15 15 that the distributional limits of these mam- mals occur at the same place. The factors 75 100% 89 that limit any one species may have little to It can be noted that there is virtually no do with the factors limiting others. local subspeciation in the Chiroptera. The open-forest fauna of Cape Of the monotremes the spiny anteater is York Peninsula comprises 23 species, of found both in open forest and in rain forest. which seven are arboreal and hide by day in The is tied closely to the river sys- holes in trees. Of the 13 remaining, 11 are tems. definitely terrestrial, while two are good Of the marsupials, eight are arboreal, rain- climbers. With the exception of Melomys forest forms of Papuan origin, six are indig- lutillus and Satanellus hallucatus, all are posi- enous arboreal, rain-forest forms, seven are tively Australian. Melomys lutillus is an arboreal, open-forest forms, one is terrestrial open-country-inhabiting species of a mainly and of Papuan origin (Echimypera), seven New Guinea rain-forest . Satanellus are indigenous, terrestrial, rain-forest forms, hallucatus occurs in Australia chiefly north of and 10 are native, terrestrial, open-forest latitude 20° S., except in the southeast of its animals. range, where it has been taken at Yeppoon 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 575 near Rockhampton and at Clermont. The rops) archeri, Dendrolagus bennettianus, Den- nearest relative of hallucatus is S. albopunc- drolagus lumholtzi, Rattus leucopus, the arbo- tatus, a rain-forest species of New Guinea, real Melomys, and Uromys are Papuan. but Satanellus is wholly Australian from the Of the truly Australian rain-forest species, standpoint of generic relationship. Pseudocheirus peregrinus goes north to Coen. Of the seven arboreal, open-forest species, Pseudocheirus herbertensis, Pseudocheirus le- one (Phascolarctos) has not been found north muroides, and Hypsiprymnodon stop their of the latitude of Cooktown; one (Pseudo- northward spread near Cooktown, and Pro- cheirus peregrinus) reaches its northern limit temnodon bicolor reaches the tip of the Penin- near Coen; two (Trichosurus and Mesem- sula. Dasyurops m. gracilis is probably also briomys) extend little beyond Wenlock on there, but the evidence is inconclusive. The the Batavia River; one (Acrobates) attains genus Antechinus, widely spread both in Aus- the tip of Cape York; and one (Petaurus tralia and in the mountains of New Guinea, breviceps) extends across the to is represented in the McIlwraith scrubs east many parts of New Guinea. of Coen and at the Iron Range and on Ather- Seven of the 11 terrestrial, open-forest spe- ton Tableland (eastern side), but was not cies are (Petrogale inornata, discovered in the rain forests at the tip. The Aepyprymnus rufescens, Macropus canguru dormouse phalanger Eudromicia is found in and Macropus robustus, Protemnodon parryi, the Cooktown-Townsville rain forest but has Protemnodon bicolor, and Protemnodon agi- not yet been recorded farther north on the lis); two (Isoodon obesula and Isoodon macro- Cape. On the other hand it is well distributed urus) are bandicoots; one (Sminthopsis aru- in New Guinea. The genus Eudromicia is ensis) is a mouse; two (Rattus here held to be primarily Australian because gestri and Leggadina delicatula) are rodents. of the presence of the very closely related but Two of the Macropodidae (Aepyprymnus more specialized in the southern rufescens and Protemnodon parryi) reach parts of Australia and in Tasmania. Thylo- their northern limits near Cooktown, three gale stigmatica alone of the Australian rain- (Petrogale inornata, Macropus canguru, and forest species extends to the tip of the Cape Macropus robustus) near Wenlock, and the and across into the lowland scrubs of south sand (Protemnodon agilis) crosses New Guinea. into the grass country of south New Guinea. The dispersal southward of the New Sminthopsis rufigenis, Isodon macrourus, and Guinea rain-forest species, all of which are one of the rodents, Rattus gestri, have sub- truly arboreal except Echimypera, Rattus, stantially the same range in New Guinea as and Melomys lutillus, present several differ- Protemnodon agilis. The other , Leg- ing patterns. On the one hand the mainly gadina, extends north only to Wenlock. The arboreal genus Melomys, represented by ubiquitous Tachyglossus, it should be added, several local races, is found almost continu- is also at home north of the Torres Strait in ously along the humid coastal strip as far the same general area as the sand wallaby, south as northern New South Wales. The but probably tends to penetrate a short way Rattus leucopus-assimilis group displays a into the margining rain forests. generally similar pattern. The other Papuan The mammals inhabiting the Cape York invaders of Australia, if not halted farther Peninsula rain forests present markedly dif- north, appear to have been stopped in their ferent dispersal patterns. At least 21 species southward movements by the great open- occur. Of these about half can be classed as forest break at Townsville and the lower truly Australian and half as truly Papuan. Burdekin River area. Thus far southward Antechinus, Dasyurops, Perameles, Eudro- have advanced Dactylopsila, Pseudocheirus micia, Pseudocheirus peregrinus, Pseudochei- () archeri, Dendrolagus lum- rus herbertensis, Pseudocheirus lemuroides, holtzi, and Uromys. Negative evidence sug- Hypsiprymnodon, Thylogale stigmatica, and gests that Pseudochirops and Dendrolagus Protemnodon bicolor are Australian. Echimy- were originally present in the scrubs of the pera, Dactylopsila, Phalanger orientalis, Pha- tip, Iron Range, and the McIlwraith Range, langer maculatus, Pseudocheirus (Pseudochi- but have since died out in these intervening 576 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 sets of rain forest. Dactylopsila and Uromys had important repercussions on local cli- on the contrary were found in 1948 in all the mates. Increased rainfall may have partly large areas of rain forest. Bennett's tree- closed some of the gaps between present climbing , Dendrolagus bennettianus, rain-forest islands or effected the union of though at home in the northern part of the some forest-edged river systems (say, the Cooktown-Townsville rain forest, seems not Digul and other south New Guinea rivers) to have spread much to the south of the with those flowing from the Gulf of Carpen- , i.e., not to have reached the taria, which may have offered roundabout latitude of Cairns, south of which the species forested routes from New Guinea to the Dendrolagus lumhotzi is dominant. McIlwraith Atherton forests, letting certain The remainder of the New Guinea rain- mammals bypass the extreme tip. forest species, Echimypera, Phalanger orien- The extensive open forest west of the Pen- talis, and Phalanger maculatus, were stopped insular divide has apparently permitted ap- at the southern edge of the Iron Range- propriate Australian species to push towards Mcllwraith Range rain-forest mass, prob- or into New Guinea with varying degrees of ably by the great 150-mile stretch of open success. The interrupted strip of rain forest forest between there and Cooktown. Of those along the eastern slopes of the divide or its species, only Phalanger maculatus can now outliers has likewise allowed the southward be found in the forests at the tip. Echimypera spread of species haunting the rain forest of r. australis is known only from Rocky Scrub New Guinea and offered them environmental on the east slopes of the McIlwraith Range harbors. The total number of Australian in the latitude of Coen. mammals that have invaded New Guinea is The absence of so many Papuan mammals eight. Only one of these, Pekturus breviceps, from the rain forests through which they is arboreal. Only one, terrestrial Thylogake may once have passed in order to enter their stigmatica, lives in rain forest. The total num- present habitats indicates past changes of ber of New Guinea mammals that have in- those environments. The successive passages vaded Australia is 11 (actually more because of the tree to the Cooktown- I am thinking of the arboreal Melomys cer- Townsville rain forests and of the vinipes provisionally as a single unit). Of to, but not beyond, the Iron Range-Rocky these, all but two groups, Rattus keucopus Scrub forests favor the view that more than and allies, and Echimypera, are arboreal. It one wave of invasion took place and that would seem then that Australian, terrestrial, differing routes may have been followed. The open-forest mammals and New Guinea, ar- changes of environment may have been boreal, rain-forest mammals are both success- fairly extensive, for the alterations in ocean ful colonizers into essentially favorable en- level generally held by geologists to have vironments. Such invasions have probably taken place synchronously with the Pleisto- not taken place at one time but at various cene glaciations, coupled with the cooler times in late Tertiary history. world temperatures of that time, may have CAPE YORK MAMMALS, WITH NOTES ON GENERA AND SPECIES FROM OTHER PARTS OF QUEENSLAND ORNITHORHYNCHIDAE bert River and Cape York Peninsula (1878, ORNITHORHYNCHUS BLUMENBACH Jour. Linnaean Soc. London, vol. 14, p. 411). Ornithorkynchus BLUMENBACH, 1800, Gotting- "Porcupines" are apparently moderately ische Gelehrte Anz., vol. 1, p. 609. common all down the eastern side of the Pen- insula. They occur in both rain forest and Ornithorhynchus anatinus phoxinus Thomas open forest; at least they come out into open Ornithorhynchus anatinus phoxinus THOMAS, forest at night. We obtained them in both 1923, Ann. Mag. Nat. Hist., ser. 9, vol. 11, p. 176. environments. I have no data on their pres- MATERIAL: Evelyn (Neuhauser), 2. Raven also ence on the west side of the Peninsula but found specimens at Evelyn and Ravenshoe in would expect them to be present in areas not 1929. C.N.H.M.: Wongabel (Scott), 3, Ravenshoe flooded annually. They are known to occur (Scott), 1. in the . The northern limit of the platypus ap- Collett compared acanthion with lawesi, pears to be some 20 miles south of Cooktown. the Papuan race, and with aculeatus of New It was reported from "the Tableland," about South Wales. 5 miles east of Helenvale, where it was some- The following notes made by H. C. Raven, times caught in the sluices of tin miners. At though based upon the Tasmanian form Shipton's Flat at the western foot of Mt. rather than the Cape York form, are of in- Finnegan, through which a small tributary of terest: the flows, platypus were seen "Tachyglossus setosus scratches itself with occasionally by local people. We ourselves, its elongated hind toe (hallux). With this toe though we looked for them assiduously, did nail it can scratch its thigh, tail, head, belly, not observe them. Farther south, on Ather- sides. Only when frightened does it erect the ton Tableland, they are still fairly common, spines and draw them forward so as to hide though their habitat of forested streams is its head. A noise like the snapping of one's already much restricted by the deforestation finger on matting is produced when the large activities of the local dairy and lumber in- hind toe is used to scratch the back. The ani- terests. nal does not crawl along or shuffle unless TACHYGLOSSIDAE alarmed. Its usual gait is a walk, with its legs fully extended so that its belly is com- TACHYGLOSSIUS ILLIGER paratively high off the ground. It reminded Tachyglossus ILLIGER, 1811, Prodromus sys- me of the gait of a giant tortoise. When walk- tematis mammalium et avium, p. 114. ing, the hind toes are directed outwards and Tachyglossus aculeatus acanthion (Collett) backwards. Thus the hallux becomes outer- Echidna acanthion COLLETT, 1885, Forhandl. most. Vidensk.-Selsk. Christiania, for 1884, no. 13, pp. "I picked up [a young captive specimen] 1-11; 1885, Proc. Zool. Soc. London, p. 148. while it was rolled up like a ball, placed it MATERIAL: Iron Range, 7; Wenlock, Batavia gently, back down, on the palm of my hand, River, 1; Mt. Finnegan, Rossville Creek, 1; Mt. and with the forefinger of my other hand Finnegan, Top Camp, 1. A.M.N.H.: Lake Barrine gently rubbed the soles of its feet until it (Neuhauser), 2. C.N.H.M.: Ravenshoe (Scott), partly unrolled and seemed slightly less nerv- 3. Besides the foregoing, the American Museum ous. I then gently pushed a teaspoon con- has specimens from the Dawson River and Mun- taining warm milk beneath its snout. After dubbera, farther south in Queensland, referable to a moment's hesitation it began thrusting out true aculeatus. its smooth long slender pink tongue, thus Collett described the type of acanthion licking up the milk. Later, I found that if the from Gracemere, near Rockhampton, and spoon was placed so its mouth came in con- listed eight others from Coomoobooloo (80 tact with the milk, it would drink. miles farther south) collected by Lumholtz. "When the animals are alarmed they start He referred also to specimens from the Her- burrowing straight downward. It is remark- 577 578 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 able how quickly they can get beneath the present sqries of some 80 specimens has tooth surface of the ground. The earth is scratched rows ranging in length from 6.0 to 6.9 mm. loose with the powerful fore claws, and with The northern examples (Mcllwraith Range, the fore paws is thrust out, up, and just be- Iron Range) include no example of either sex hind the elbow, or is passed backward for with tooth row less than 6.3 mm., and four of the curved, scraper-like hind claws to push the examples reach 6.8 or 6.9 mm. In this out. This is done more when the animal is same group of larger size belong four speci- progressing forward as well as downward. mens from Speewah, upper Clohesy River, During the whole of this burrowing process northern part of Atherton Tableland, and the head is infolded beneath a covering of two from the Russell River lowlands, 10 sharp spines, which are continually pointed miles south of Cairns. Animals with slightly and twitched here and there towards the di- smaller teeth, which might conceivably be rection of contact, or the sound of, the pur- referred to adustus, are: Julatten, a few miles suer." west of Mossman, one female; Mt. Finnegan, 50 miles north of Mossman, three females; DASYURIIAE Ravenshoe, one male. ANTECHINUS MAcLEAY The mammary formula in this same large Antechinus MACLEAY, 1841, Ann. Mag. Nat. series of specimens is variable. In five females Hist., vol. 8, p. 242. in nursing condition the mammary area This is a large genus, extensively distri- shows a formula 5-5 = 10. But four others buted in Australia, Tasmania, and New show 4-4 = 8; and there is one example Guinea. I reviewed it a few years ago (Tate, (A.M.N.H. No. 154312) in which the formula 1947b, pp. 126-133). Only one of its species, is 4-1-4 = 9, the single unpaired mamma giv- A. flavipes, enters the Cape York area. It ing the appearance of being median and pos- seems to consist there of two races, the larger terior to the others. I think that in the cases godmani and the smaller adustus, which meet of 4-4 = 8, certain of the mammae remain un- and intermingle (or perhaps hybridize) near everted and non-functional, and that the the latitude of Cairns. Farther south still count in those should be interpreted as other races of occur. 5-5=10. All of the above formulas were - flavipes served and recorded on the fresh specimens in the field. In my paper on the Dasyuridae Antechinus flavipes Waterhouse (Tate, 1947b, p. 149) I quoted Pocock, who When Thomas descr'ibed Antechinus fla- gave the formula of flavipes as 2-2 =4. vipes adustus and Antechinus godmani from The upshot of this discussion is my con- the same locality (Dinner Creek, 9 miles clusion that godmani should no longer be southeast of Ravenshoe, Atherton Table- considered a full species but merely a part of land) there seemed to be little doubt of their theJfavipes complex. Antechinus godmani cer- specific distinctness. This seeming certainty tainly is larger than typical A. flavipes of the existed despite the fact that their only dis- southern parts of eastern Australia, and ap- tinguishing feature was one of size and also parently Thomas' small-sized race adustus despite the fact that males are always much represents virtually the northern extent (Mt. larger than females in Antechinus. The differ- Finnegan is the ultimate northern record) of ences in the dental measurements of the types such small-sized Antechinus. Just what the seemed then to require full specific separa- relationship of adustus to true flavipes is I am tion. not in a position to say, but the two are But with the additional evidence pro- surely extremely close to each other. vided by the large series collected by the re- The species occurs in suitable localities all cent Archbold expedition from Iron Range down the eastern side of Australia, from Iron to Russell River and the Atherton Table- Range in the Cape York Peninsula to the land (see beyond) doubts arise. The length of south of Australia, and reappears in the the molar tooth row, ml<, in adustus (type, southwest of Western Australia. We did not adult male) is only 5.6 mm., while that of find it in the rain forest at the extreme tip of godmani (type, adult male) is 6.6 mm. Our the Peninsula. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 579 Antechinus flavipes adustus (Thomas) Only one species, rufigenis, a member of Phascogale flavipes adusta THOMAS, 1923, Ann. the less specialized rufigenis division, was Mag. Nat. Hist., ser. 9, vol. 11, p. 175. found by us in Cape York Peninsula. A sec- MATERIAL: Mt. Finnegan, at 1500 and 3000 ond species of this division, murina, occurs in feet, 3; Julatten, 10 miles west of Mossman, 1; the rain forests on the east side of Atherton Evelyn and Ravenshoe, Atherton Tableland, 5. Tableland, and this may be expected to ex- U.S.N.M.: Nine miles south of Ravenshoe (Hoy), tend nearly to Cooktown. As a species of the 2. A large series, the skulls lost, was collected at crassicaudata division is known from Malbon, Dinner Creek, in rain forest, 9 miles southeast of near Cloncurry, it is possible that that divi- Ravenshoe, by H. C. Raven in 1922. sion will also be demonstrated sooner or later these animals were not nearly so abundant in the Cape York region. as was the race godmani farther to the north. Van Deusen, who trapped the specimen from Sinthopsis rufigenis lumboltzi Iredale high on Mt. Finnegan, found that it was at and Troughton least partly diurnal. He saw it come out of Sminthopsis lumholtzsi IREDALE AND TROUGH- a hole at the base of a tree in the rain forest TON, 1934, Mem. Australian Mus., no. 6, p. 11. after bird meat, and he trapped it a few MATERIAL: Iron Range, 7; Mt. Tozer, north minutes later. foot, 3; Camp Oven Pocket, upper Peach River, 3; Shipton's Flat (Roberts), 1. Antechinus flavipes godmani (Thomas) This avoids the Phascogale godmani THOMAS, 1923, Ann. Mag. Sminthopsis dense rain Nat. Hist., ser. 9, vol. 11, p. 174. forest in which Antechinus flavipes is at home MATERIAL: Iron Range, 24; Mt. Tozer, 11; up- and favors open rocky forest and brushy per Nesbit River, McIlwraith Range, 22; Camp places in full sunlight. There is little doubt Oven Pocket, McIlwraith Range, 4; Mossman that its nearest relative is rufigenis (=rona) Gorge, 1; Speewah, upper Clohesy River, Barron of Aru Islands and the open country of south- River, 4; Russell River, 2. U.S.N.M.: Nine miles ern New Guinea (the Oriomo and Port south of Ravenshoe (Hoy), 14. Moresby grasslands). It belongs to the Tor- The range of this large-sized form extends resian group of Sminthopsis in which the tail up through the Cape York Peninsula at never becomes incrassated. least as far as the Iron Range (we failed to It seemed not to be a common animal any- find it at Lockerbie near the tip). It was where, though with different conditions of moderately plentiful at Iron Range and Mt. trapping possibly more might have been Tozer and extremely abundant (almost as taken. It inhabited the grassy, well-illumi- plentiful as Rattus) at our camp on the upper nated environment affected by Melomys lu- Nesbit River in the McIlwraith Range. tillus, and once or twice I suspect it ate parts From Mt. Finnegan, 30 miles south of Cook- of the Melomys after the latter had been town, southward, it seems to interdigitate caught in a trap. (perhaps even hybridize) with its smaller Only one specimen was a female. It had relative adustus, which was considerably the pouch area in condition for the mammae rarer. to be counted. This female had a "circular The first specimen of godmani we took was drawstring pouch area" with the nipple seen at night running up the trunk of a rain- formula 4-4=8. According to my observa- forest tree. After that we began to find them tions made on dried skins, the formula of in rat traps set at the bases of trees in rain Papuan rufigenis is 3-3=6. However, our forest. Australian specimens were observed freshly killed in the field, so that none of the struc- SMINTHOPSIS THOMAS tures were concealed. Sminthopsis THOMAS, 1887, Ann. Mus. Civ. The type locality of virginiae Collett (not Genova, ser. 2, vol. 4, p. 503. Tarragon) (=-umholtzi Iredale amd Trough- As in the case of Antechinus, Sminthopsis ton) was given as Herbert Vale, in the upper is a quite large genus of small phascogales, middle part of the Herbert River, doubtless distinguished by several distinct characters where its headwater streams drain the (Tate, 1947b, pp. 119-125). grassy portions of the southern part of the 580 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 Atherton Tableland before cutting down the Cape York Peninsula. It is apparently into the rain forest. The distribution of the absent from the rain-forest habitat of gracilis pattern of the species in that part of Aus- but extends around that area of woods on the tralia then will correspond fairly closely to west (i.e., to Torrens Creek and Clermont) that of such other open-forest species as Iso- and south. The most southerly record I have odon macrourus, Satanellus hallucatus, and of it is at Yeppoon near Rockhampton. Melomys lutillus. In the case of Sminthopsis In the wild, the species hides by day in rufigenis, the south Papuan race has dis- fallen hollow tree trunks. Arounddeveloped tinctly darker pelage and tail and more mark- areas it often conceals itself in disused edly russet hands and feet. The Australian houses-in stoves, for example. One was race is definitely grayer. There is also the pos- trapped in an abandoned barge on the beach sibly different mammary count to be con- at Red Island Point. Another was shot from sidered. the rafters of a native schoolhouse. A third, I pointed out earlier (Tate, 1947b, p. 120) which hid by day among the girders of the that virginiae Tarragon, if ever supported by Red Island wharf, was taken by having an a type (and Paris museum types do some- overcoat thrown over it. The animals were times reappear), will probably displace rufi- rather easily trapped in meat sets prepared genis as a specific name. for dingoes. Sinthopsis murina (Waterhouse) DASYUROPS MATScHIE Phascogale murina WATERHOUSE, 1838, Proc. Dasyurops MATSCHIE, 1916, Mitt. Zool. Mus. Zool. Soc. London, p. 76. Berlin, p. 262. MATERIAL: Atherton Tableland, 1. M.C.Z.: The characters of the skull and teeth have Lake Barrine (Darlington), 1. been previously discussed (Tate, 1947b, SATANELLUS POCOCK pp. 144-145). This is primarily an animal of the rain forest. Satanellus POCOCK, 1926, Proc. Zool. Soc. Lon- It becomes necessary to amend part of my don, p. 1083. definition of Dasyurops (Tate, 1947b, p. 145). The combination of retention of hallux, The striations of the feet are not "wholly ab- full striation of the pads, and retention of the sent" but are confined to small, elongated minute p4 (in albopunctatus) seems to indi- areas at the center of each otherwise granular cate the primitiveness of this genus in com- pad. parison with related Dasyurinus, Dasyurus, The number of species must also be and Dasyurops (Tate, 1947b). changed from two to one. Dasyurops gracilis The genus contains two distinct species: is now revealed as only a weak geographical albopunctatus of New Guinea and hallucatus race of Dasyurops maculatus. It is not even of the northern parts of Australia. smaller in size; its seeming smallness was due Four races, which are barely separable to the youth of the specimen originally ex- from one another, have been named, the one inhabiting the Cape York Peninsula being amined. S. h. predator. Dasyurops maculatus gracilis (Ramsay) Satanellus hallucatus predator (Thomas) Dasyurus gracilis RAMSAY, 1888, Proc. Linnean Soc. New South Wales, ser. 2, vol. 3, p. 1296. Dasyurus hallucatus predator THOMAS, 1926, MATERIAL: The tableland near Rossville, 20 Ann. Mag. Nat. Hist., ser. 9, vol. 18, p. 543. miles south of Cooktown, 1 (no skull); Mt. MATERIAL: Red Island Point (near), 4; Mt. Spurgeon, 1; Dinner Creek, 9 miles southwest of Tozer, 3; Brown's Creek, 3; Wenlock, Batavia Townsville, 1. U.S.N.M.: Nine miles south of River, 1; , 1; Croll Creek, near Coen, Ravenshoe (Hoy), 2; 6 miles south of Ravenshoe 1; Helenvale, 2; Shipton's Flat, 1, (Roberts), 7. (Hoy), 1; Mt. Spec, 1. The type is from Mt. Bel- A.M.N.H.: (All by Neuhauser), Coen, 4; Port lenden Ker. Stewart, 3; Rocky Scrub, McIlwraith Range, 2; U.S.N.M.: Nine miles southwest of Ravenshoe Mr. Troughton arranged to have theskull (Hoy), 3. M.C.Z.: McIlwraith Range (Darling- of the type (Sydney, No. M.155) extracted ton), 12. and cleaned for me while I was in Australia This fine series indicates that the species is in 1947. It proved to be that of a very young widely distributed in the open-forest areas of animal, in which m were not yet erupted. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 581

In m1, as compared with m2, the metacone bandicoot described first from the neighbor- and metastyle are confluent, and the tooth is hood of Sydney. A weak race was proposed correspondingly narrow. The paraconid of for representative of the species from the m1 is obsolete, and the metaconid is much re- Atherton Tableland. Capture of additional duced. specimens extends the range of the species The ventral surface of the skin is even northward almost to the tip of the Cape more spotted with white than the dorsal sur- York Peninsula. face, and the spots are larger. The pads are Perameles nasuta pallescens Thomas narrow, striated; the skin bordering the stri- ated areas is granular. The tail is spotted. Perameles nasuta pallescens THOMAS, 1923, Ann. This is a form native to the rain forests of Mag. Nat. Hist., ser. 9, vol. 11, p. 173. MATERIAL: Mt. Tozer, 400 feet, 1; Helenvale, 1; . Its southern limit ap- Mt. Finnegan, Rossville Creek, 1 with 2 pouch pears to be near Townsville, beyond which young; Speewah, 1; Walter Hill Range, 2. M.C.Z.: the arid Burdekin River zone probably pre- (All by Darlington), Millaa Millaa, 1; Lake Bar- vents its extension to the south or southeast. rine, 1; Mt. Spurgeon, 1; McIlwraith Range, 2. Northward it probably extends for a short C.N.H.M.: (All by Scott), Vine Creek, Ravens- distance beyond Mt. Finnegan, and possibly hoe, 5; Wongabel, 1; Danbullan, 2; near Puzzle even northward into the McIlwraith Range. Creek, 7 miles southwest of Mt. Spec, 1. U.S.N.M.: There one night, in a creek bed, I saw the Nine miles southeast of Ravenshoe (Hoy), 2. shine of distant eyes that I thought were too Although the Archbold expedition did not large to be those of Satanellus. take this rain-forest bandicoot in the Mc- Now that I have examined the type and Ilwraith Range, two specimens had been ob- have so many extra specimens for study it tained there previously, as shown, by Dar- becomes possible to reevaluate the status of lington for the Museum of Comparative gracilis. The specimens present a growth Zoology. Southward there is every reason to series ranging from the type (a baby animal, believe that it extends at least to Townsville. though independent of the pouch) to the very The typical race of nasuta is found in New large male collected by Raven at Evelyn. South Wales and extreme southern Queens- The large size of the teeth which I noted land. earlier in the case of the young female from It now seems certain that the large Pera- Mt. Spurgeon fits perfectly when one recog- meles nasuta extends at least as far north nizes that it is a juvenal female of the north- as Iron Range and Mt. Tozer. I think it will ern race of maculatus. The specimen, un- be found eventually in the rain-forested sexed, from Mt. Spec near Townsville like- patches at the tip of the Peninsula. Our wise appears to have large teeth, though even specimens from Helenvale, Mt. Finnegan, it is subadult. The skin without skull from 20 and Speewah complete the connection with miles south of Cooktown serves only to ex- Atherton Tableland, type locality of the race tend the range that far to the north. pallescens. PERAMELIDAE ECHIMYPERA LESSON PERAMELES GEOFFROY Echimypera LESSON, 1842, Nouveau tableau du Perameles GEOFFROY, 1803, Bull. Soc. Philom., regne animal, p. 192. Paris, no. 80, p. 150. A Papuan genus distinguished from Pera- Unspecialized bandicoots with a full com- meles and Peroryctes by the absence of the plement of incisors and unspecialized audital fifth upper incisors. bullae. The genus, which is confined to Aus- Recently I recorded it for the first time tralia and Tasmania, comprises two major from Australia (Tate, 1948b, p. 334). sections, only one of which, nasuta, reaches Echimypera rufescens australis Tate northwest Queensland. Echimypera rufescens australis TATE, 1948, Bull. Perameles nasuta Geoffroy Amer. Mus. Nat. Hist., vol. 92, p. 334. Perameles nasuta GEOFFROY, 1804, Ann. Mus. MATERIAL: None. d'Hist. Nat., Paris, vol. 4, p. 62. Despite intensive efforts in the rain forests This is the large, long-snouted species of around the type locality, the upper Nesbit 582 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 River, Mcllwraith Range, made by members 4, and 1, skull only. M.C.Z.: Mt. Molloy (Darling- of the Archbold expedition, no second speci- ton), 1; Shipton's Flat (Roberts), 2. C.N.H.M.: men of this unique species was discovered. (All by Scott), Danbullan Road, Barron River, 2; In New Guinea Echimypera is a rain-forest Wongabel, 1; Cairns, 1. U.S.N.M.: Five miles genus. The same can be expected of the Aus- southwest of Ravenshoe (Hoy), 3. tralian race. Since our return from Cape Isoodon macrourus is a denizen of the open York, Darlington has described for me ex- and semi-open forest. Most of those secured actly where the type was collected, in the at Lockerbie were found in the long grass of dense rain forests of the upper Nesbit River the region. The grass at Iron Range, Portland on the east slopes of the Mcllwraith Range. Roads, Mt. Tozer, and Coen was shorter, and the broad-headed bandicoots sheltered by ISOODON DESMEST day in holes in hollow logs. In the Cairns dis- Isoodon DESMAREST, 1817, Nouveau diction- trict they were at home in sugar-cane fields, naire d'histoire naturelle, nouv. ed., vol. 16, p. and many were killed by cars on the high- 409. ways. They are sometimes vectors of the Bandicoots with somewhat broader heads mites carrying scrub typhus. than the very similar appearing Perameles The range of this large-toothed species is and readily distinguished by their greatly now known to extend south to New South enlarged, pear-shaped bullae. Wales. Besides those I listed before (Tate, There are two main types of Isoodon: the 1948b, pp. 339-340) I have a skull from large macrourus type and the small obesulus Chelmer, Brisbane, which was given me for type. Both are now found to be present in obesulus but proves to be a juvenal macrourus. open-forest country of Queensland, and both far as occur as the northern tip of the Cape This family contains three subfamilies: the York Peninsula. Phalangerinae, the Tarsipedinae, and the Isoodon obesulus peninsulae Thomas Phascolarctinae. Isoodon peninsulae THOMAS, 1922, Ann. Mag. The Phalangerinae comprise several pri- Nat. Hist., ser. 9, vol. 9, p. 679. mary groups of genera: MATERIAL: Lockerbie, 2; Newcastle Bay, 1 (in- 1. Trichosurus, Wyulda, Phalanger, and prob- side a carpet snake); Iron Range, 1. ably the fossil Wynyardia Earlier I 2. Dactylopsila and its specialized Papuan rela- (Tate, 1948b) concluded that the tive Dactylonax nearest relatives of the type of peninsulae 3. Eudromicia, Cercartetus must be the races of obesulus. The capture of 4. Petaurus additional material of the species from the 5. Distoechurus of New Guinea Peninsula, and the fact that it apparently 6. Acrobates associates freely with the larger and quite distinct macrourus, seem to indicate that the The position of the genus Gymnobelideus two species may range together wherever is not clear. It probably belongs near groups their distributional areas overlap, just as 3 and 4. they seemingly do in New South Wales and The Tarsipedinae contain only Tarsipes southern Queensland. of . One of the three specimens listed above The Phascolarctinae contain Pseudochei- (that from Iron Range) is distinctly reddish rus, the specilized Phascolarctos, and the glid- and thus suggests linkage with auralus Ram- ing Schoinobates. say from the Northern Territory and Am- PHALANGERINAE hemland. TRICHOSURUS LESSON Isoodon macrourus torosus (Ramsay) Trichosurus LESSON, 1828, Dictionnaire clas- Perameles macroura var. torosus RAMSAY, 1877. sique d'histoire naturelle, vol. 13, p. 333. Proc. Linnean Soc. New South Wales, vol. 2, p. 12, The genus Trichosurus can well be consid- MATERIAL: Lockerbie, 3; Portland Roads, 2; ered the most generalized of the Phalange- Mt. Tozer, 1; Croll Creek, 1; Mossman, 1; Cairns, rinae. It appears to be nearest to Phalanger 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 583 and may be the Australian counterpart of more pronounced. The measurements of that typically Papuan genus. Indeed the caninus are generally greater than those of short-eared Trichosurus caninus and the vulpecula. Celebes phalanger, Phalanger ursinus, are I am inclined to agree with Troughton much alike superficially. Wyulda is a direct (1948, p. 126) in regarding the Tasmanian modification of Trichosurus. fuliginosus as a full species. It is larger than I discussed a number of the characteristics either caninus or vulpecula, very much larger of the dentition of Trichosurus when writing than the latter. Its structure (the elongate on the Macropodidae (Tate, 1948a, pp. ears, inflated nasals and squamosals) is, Z42-250). It has advanced beyond that of however, more nearly in agreement with that Phalanger in having lost the lower, peg-like of the latter than of the former, and I would incisor, P2, and its upper p2 is relatively expect fuliginosus and vulpecula to be some- smaller. The condition of il in Trichosurus what closer to each other than either is to is not especially significant, since in Pha- caninus. langer ursinus it is found quite unreduced, I am not familiar with ruficollis and hype. and in P. orientalis and P. maculatus much leuca, the races from Western Australia, nor reduced. The third incisor is only slightly with arnhemensis. reduced in Trichosurus (all three species). The almost unique Wyulda from Wynd- The counterparts of the frontal and supra- ham is a Trichosurus-like form in which en- squamosal inflations of T. vulpecula and largement of the mastoid area has advanced T. fuliginosus can be observed in Phalanger so much farther that the mastoid width ap- maculatus. Frontal but not suprasquamosal proximately equals the zygomatic width. enlargements occur in P. ursinus. All in Three rather similar species occur: vulpe- all, Trichosurus and Phalanger must be cula, very widely distributed in Australia, thought very closely allied. fuliginosus of Tasmania, and caninus of the The character of the external ear, currently eastern parts of northerh New South Wales used to distinguish Trichosurus caninus from and southern Queensland. Trichosurus vulpecula (short and broad in the former, elongate and narrow in the Trichosurus vulpecula Kerr latter), is so striking that one naturally looks This somewhat variable species apparently for supporting characters in the skulls. There covers an enormous geographical range. It the distinction is far less obvious, apart is therefore not surprising that a number of from the greater size in caninus and fuligino- races have been set up. In the eastern parts sus. In the audital area of the skull the prin- of Australia there are four such: the typical cipal distinction lies in the mastoid-squamosal one from New South Wales and southern portion, particularly the superior surface of Queensland; mesurus Thomas from Inker- the squamosal inward from and behind the man, southeast of Townsville; eburacensis root of the zygoma. In the long-eared vulpecu- from the Cape York Peninsula; and the deep la and fuliginosus this area shows distinct to brown johnstonii from the rain forests of the (in some males) marked inflation. In the eastern side of the Atherton Tableland. I short-eared caninus no inflation is appre- think Troughton (1948, pp. 120-121) is right ciable. The inflation furthermore results in in suggesting that mesurus is insufficiently thE fact that the mastoid width in fuliginosus distinct to warrant subspecific separation and vulpecula is approximately equal to the from southern vulpecula. Furthermore, the width across the tips of the meatal tubes gray phase of eburacensis is in very nearly the (49:48 and 38:40), whereas it is considerably same category; the only characters distin- less in caninus (44.5:49). guishing it from vulpecula are the shorter, Elsewhere, one sees in vulpecula and fuli- rather crisper body pelage and the shortness ginosus a tendency for the base of the nasals of the black hairs of the tail, which gives the together with the anterior part of the frontals tail a much more pronounced taper than the to become elevated and even somewhat in- tail of vulpecula. flated, whereas this part in caninus remains There remains the question of the line of flat. In caninus the frontal trough is usually geographical separation between eburacensis 584 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 and vulpecula. All our material from the Darlington), Malanda, 2; Millaa Millaa, 2; Ather- latitude of Cooktown and northward to the ton Tableland, 2; Ravenshoe, 2; , Batavia River belongs unquestionably to 32 miles below Hughendon, 1; Leichardt Range, eburacensis. One collected by G. Neuhituser offshoot of the Clarke Range, near McKay, 1. near Mt. Spurgeon and two by myself at C.N.H.M.: Bluewater Creek, some 20 miles north- McLeod Creek, a tributary of the upper west of Townsville (Scott), 2. Mitchell, just west of Mt. Spurgeon, are all One of the Bluewater animals shows the referable to eburacensis. Yet the good series mixture of reddish in the under coat, de- obtained by H. C. Raven at Snubby Creek, scribed next in eburacensis, so may be annect- 6 miles southwest of Ravenshoe, which has ant. They are also possibly referable to T. v. fuller pelage and the tail much denser, should mesurus Thomas, a named race that seems perhaps be referred to true vulpeculk. Ravens- virtually characterless. hoe is less than 50 miles south-southwest Reference of this mid-Queensland material of Mt. Spurgeon. Possibly vulpecula in north- to true vulpecula is provisional and is done in ern and remains in the deference to Troughton's opinion. We have relatively better-watered strip of open forest only "zoo" specimens from New South Wales. adjoining the Atherton rain forests and far- These often have yellow under parts and ther south nearer the coast (e.g., at Towns- much denser pelage than appears in speci- ville and Inkerman), while eburacensis ex- mens from farther north, but that condition tends from Cape York Peninsula south along may either be typical or represent a response the next drier zone to the west, that is to say, to winter conditions. If mesurus TMomas were south through the Mitchell and Flinders considered a good subspecies, our middle and drainages to the western side of the upper north Queensland specimens would be refer- Burdekin River. able to it instead of to vulpecula proper. The principal distinction between the race Trichosurus vulpecula eburacensis L6nnberg vulpecula (including mesurus and eburacensis) Trichosurus vulpecula eburacensis L6NNBERG, and johnstonii is in the ears. In vulpecula the 1916, in L,5nnberg and Mj6berg, K. Svenska back of the ear is marked by a large area of Vetensk. Akad. Handl., vol. 52, no. 2, p. 9. cream-colored hairs. This area occupies the MATERIAL: Wenlock, Batavia River, 6; Coen posterior half of the pinna, the anterior half River, 1; Croll Creek, Coen, 4; Ebagoolah, south being gray like the head. There is a narrow of Coen, 1; Cooktown, 1; Atherton Tableland, 1; terminal margin of fuscous around the cream Mt. Spurgeon, 1; McLeod Creek, Mitchell River, area, which does not, however, extend down 2. M.C.Z.: (All by Darlington), Coen, 3; Mcll- the posterior margin all the way to the base. wraith Range, 2. In johnstoniui the entire ear is fairly uniformly This race is apparently absent from the brown like the head and body. In a few in- extreme tip of the Peninsula and from Port- dividuals only, there is a small patch of land Roads. West of the Sir William Thomp- yellowish, corresponding doubtless to the son Range at the Batavia River and south- large creamy area in vulpecula. ward through Coen and Ebagoolah it be- The body coloring in vulpecula is predom- comes abundant. At Cooktown it has taken inantly gray, with rufescent wool hairs in up residence in the roofs of the houses, much some examples of eburacensis. The body color as palm civets do in Malaya and Didelphis in johnstonii is strongly russet, modified in marsupialis does in Guayaquil, Ecuador. some examples by over-hairs of dark brown, It appears in two phases, one with gray- in others not so modified. based pelage and no trace of rufescent, the Trichosurus vulpecula vulpecula (Kerr) other with gray over-hairs, often russet on Didelphis vulpecula KERR, 1792, The animal the nape, and rusty wool hairs. In both, the kingdom ... of Linnaeus, p. 198. backs of the ears have a large area white or MATERIAL: Snubby Creek, 6 miles southwest of creamy. These two phases appear to be un- Ravenshoe, 10; probably west of Evelyn, 1; also connected with differences of sex, age, or from farther south, i.e., Byfield, 35 miles north- season. Both may be found in a given locality. east of Rockhampton, 1; Serpentine, 25 miles In my view eburacensis is but little removed northeast of Rockhampton, 1. M.C.Z.: (All by from southern open-forest forms of vulpecula. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 585 Its characters include short body pelage and The Archbold expedition was fortunate in short tail-hairs. securing an adequate series of the former and some few additional specimens of the latter. Trichosurus vulpecula johmstonii (Ramsay) The southern limit of Phalanger now ap- Phalangistajohnstonii RAMSAY, 1888, Proc. Lin- pears to be reached in the rain forests of the nean Soc. New South Wales, ser. 2, vol. 3, p. 1297. McIlwraith Range. Extensive grasslands, MATERIAL: Evelyn, 2; Lake Barrine, 1; Dinner which stretch southward between that area Creek, 9 miles southeast of Ravenshoe, 7; Snubby and the Cooktown-Cairns-Townsville mass of Creek, 6 miles southwest of Ravenshoe, 1. rain forest, seem to have prevented its further C.N.H.M.: Upper Barron River (Scott), 2. U.S.N.M.: (All by Hoy), 9 miles south of Ravens- spread to the south. hoe, 6; Mt. Bellenden Ker, 1. Phalanger orientalis peninsulae Tate These animals, with their long, coppery Phalanger orientalis peninsulae TATE, 1945, brown hair, give the appearance of being Amer. Mus. Novitates, no. 1283, p. 2. rain-forest dwellers. The type locality, Mt. MATERIAL: Iron Range, 3; Peach River, Mcl-, Bellenden Ker, an exceedingly wet area, sup- wraith Range, 4. M.C.Z.: Topotypical series ports this view. Perhaps the race johnstonii (Darlington), 7. has taken to the rain forest of the north just A rain-forest and gallery-forest species. I as the species caninus inhabits the southern described it originally from two specimens rain forests. collected at Peach River by Darlington. We However, it seems thatjohnstonii occasion- ourselves obtained more from the same place ally finds its way into open-forest country and others from the Iron Range farther north. by way of the tongues of green gallery woods Vernon shot one or two others for the collec- that often follow water courses. This would tions of the Queensland Museum. account for the specimen from Raven's The coloring is found to be rather variable, Snubby Creek Camp, which was otherwise as suggested by the type and paratype. in open forest and provided many examples Females are fairly uniform dark gray-brown. of the gray possums, vulpecula. Males may be dark, or quite pale ashy brown. I suspect that the form johnstonii is de- Their under parts are pale, often curiously rived from that phase of eburacensis that has mottled with a slightly darker tint. the hair bases russet. The color of the present form varies from bright coppery rufous to Phalanger maculatus nudicaudatus (Gould) rather dull brownish red. The white backs Phalangista nudicaudatus GOULD, 1850, Proc. of the ears which characterize eburacensis are Zool. Soc. London, for 1849, p. 110. almost wholly obscured by dark color in MATERIAL: Lockerbie, 4; Iron Range, 21. A johnstonii. The pelage varies from rather long very young specimen was seen in captivity at and shaggy to short. This is probably con- Coen. M.C.Z.: Mcllwraith Range (Darlington), 1. nected with season in any given individual, The Australian race of the spotted but actually both long-haired and short- of New Guinea appears to reach greatest haired examples occurred at the same time. abundance at Iron Range. The forest con- tinues nearly unbroken to the McIlwraith PEALANGER STORR Range, whence comes a specimen in the Phatanger STORR, 1780, Prodromus methodi British Museum. In the extreme north (Lock- mammalium, Ttibingen, p. 33. erbie) the animal seems to prefer the scat- I showed some years ago (Tate, 1945a) tered stands of rain forest to the open forest. that the genus Phalanger could be divided At Iron Range it inhabited the tallest rain naturally into three main sections: the orien- forest and gallery forest. talis group (typical), the maculatus group, The early growth stages (Tate, 1945a, p. 2, and the ursinus group. The maculatus group stages 1 to 5), which evince themselves so has long been known to inhabit the Cape readily in the color and pattern of the juve- York Peninsula. In the same paper, I recorded nile spotted phalangers of New Guinea, are the presence there of a member of the orien- less obvious in the spotted phalangers of talis group. the Cape York Peninsula. Also the differences 586 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 of color between the sexes, so remarkable in in the midst of more extensive areas of open such races as P. m. chrysorrhos of southern forest. One shot from a tree on the top of the New Guinea, are less apparent. In the Austra- McIlwraith Range was climbing in moder- lian race both sexes are gray, the male being ately tall, quite dense rain forest. The speci- marked with roundish white spots. In many men from Helenvale was taken in fringe instances the white marks are confluent and forest along the Annan River. so surround gray spots. The color pattern of this large coIlection of Usually these animals spend the day high picata, extending from the tip of Cape York in trees in rain forest, more or less concealed to Cooktown, serves to emphasize its very by foliage. At night they move about and close relationship to the south Papuan kataui. feed on fruits and perhaps leaves. Then ma- terial can be heard dropping to the ground. Dactylopsila trivirgata infumata Tate Sometimes, however, they descend, because Dactylopsila trivirgata infumata TATE, 1945, I found one early one morning on the ground Amer. Mus. Novitates, no. 1305, p. 4. in rather dense second growth. MATERIAL: Shipton's Flat, Mt. Finnegan, 1; Barron River, 1 (and young); Lake Barrine, 2 (the DACTYLOPSILA GRAY type and paratype); Evelyn, Atherton Tableland, Dactylopsila GRAY, 1858, Proc. Zool. Soc. Lon- 1. don, p. 109. The color distinctions between this form Two races of this primarily Papuan genus, and picata from farther north, which I indi- the striped possums, occur in Australia (Tate, cated when describing the type, are amply 1945c, pp. 4-5): D. trivirgata picata in the borne out by the additional material now in northern and central parts of the Cape York the collection. It is interesting to note that Peninsula, and D. trivirgata infumata in the this rain-forest race extends north to Mt. southern parts and the Atherton Tableland Finnegan, yet is replaced a few miles beyond and Cairns area. The most southerly record (at Helenvale in open forest) by perfectly I have is from near Townsville. Other rec- typical picata. The character I suggested for ords are Mt. Bellenden Ker and Cardwell. distinguishing the dentitions of the two races The southern spread may well be limited by (rotation of p4) has not been found valid in the almost unforested and relatively dry area practice. between the lower Burdekin River and Char- EUDROMICIA MJOBERG ters Towers. Eudromicia MJ6BERG, 1916, in Lonnberg and I have pointed out elsewhere (Tate, 1948a, Mj6berg, K. Svenska Vetensk. Akad. Handl., vol. pp. 241-243) the peculiar tusk-like character 52, no. 2, p. 13. of the first upper incisors and the special type of incisive occlusion present in this These tiny "dormouse phalangers" may be fairly common in the rain scrubs (forests) genus; also the absence of any masseteric of the , and I suspect they will canal in the mandible and the lack of inflec- eventually be demonstrated in the Iron Range tion of the angular process. These features, and McIlwraith Range rain forests. The together with the striped pattern, combine to Papuan species caudata is widely distributed give Dactylopsila and its derivative Dactylo- in New Guinea, and in Tasmania another, nax a somewhat isolated position in the lepida, is known. In the southern parts of Phalangerinae. Australia, both east and west, a second genus, Dactylopsila trivirgata picata Thomas Cercartetus, occurs, which is rather more Dactylopsila picata THOMAS, 1908, Ann. Mag. specialized (Tate, 1945c, pp. 2-4). Nat. Hist., ser. 8, vol. 1, p. 123. Eudromicia macrura Mj;berg MATERIAL: Newcastle Bay, 1; Lockerbie, 8; Coen-Rocky Scrub, McIlwraith Range area, 4. Eudromicia macrura MJOBERG, 1916, in Lonn- M.C.Z.: (Both by Darlington), 1; Helenvale, 20 berg and Mjoberg, K. Svenska Vetensk. Akad. miles south of Cooktown, 1. Handl., vol. 52, no. 2, p. 14. MATERIAL: West of Cairns (presented by Mr. Most Lockerbie material was shot at night George Brooks), 1 (A.M.N.H. No. 155090, in al- in small stands of low rain forest localized cohol). There are two from this general area in the 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 587

Queensland Museum: Brisbane No. 6571 from near and in the greater part of Australia. It seems Jordan Creek, west Palmerston area, near Innis- almost equally at home in rain forest and fail, and Brisbane No. 7011 from Mt. Spurgeon, open forest. The Papuan race tafa probably 20 miles northwest of Cairns. never meets open country but inhabits mark- Mjoberg's typical collection consisted of edly rainy tracts of mid-montane jungle four specimens found in "tropical jungle" forests. The Cape York race longicaudatus near Cedar Creek. was generally found in open forest. PETAURUS SHAw AND NODDER Petaurus breviceps longicaudatus Longman Petaurus SHAW AND NODDER, 1791, The natu- Petaurus breviceps longicaudatus LONGMAN, ralist's miscellany, vol. 2, pl. 60. 1924, Mem. Queensland Mus., vol. 8, abstr., p. ix. MATERIAL: Newcastle Bay, 2; Lockerbie, 2; Petaurus comprises two primary species Portland Roads, 1; Wenlock, Batavia River, 1; groups (Tate, 1945c, pp. 6-10): the P. aus- Coen, 2; Alderbury, 2; Seagren's Farm, 1; Ship- tralis group (large sized), and the P. breviceps ton's Flat, 1, and (Roberts), 1. (All by Neu- group (small to medium sized), which in- hauser): Somerset, 4; Rocky Scrub, 1; Mt. cludes norfolcensis. Petaurus australis is set Spurgeon, 1; Lake Barrine, 1. M.C.Z.: (All by off from the others chiefly by its very broad Darlington), Lake Barrine, 5; Ravenshoe, 1; palate and its less reduced premolars. The Townsville, 1. C.N.H.M.: Ravenshoe (Scott), 2. distributional patterns differ markedly. All our specimens were shot at night from trees, usually flowering trees, in open forest. Petaurus australis Shaw and Nodder Those from Alderbury and Seagren's Farm This quite large glider is widespread in the have the tips of the tails white. southeast of Australia and occurs as the All but two of the specimens taken by our race reginae in southern Queensland. The expedition have quite long tails. The two species remains unrecorded, so far as I know, short-tailed specimens have had the tip of northward except for the colony discovered the tail damaged. by Darlington and Neuhiiuser on Mt. Spur- We have also a specimen from Pentland, geon. 140 miles west of Townsville, which differs very slightly from longicaudatus, in the direc- Petaurus australis reginae Thomas tion of norfolcensis. It is slightly larger than Petaurus australis reginae THOMAS, 1923, Ann. longicaudatus, and the base of the tail is Mag. Nat. Hist., ser. 9, vol. 11, p. 249. fuller; 75 mm. of its tip is white. But it is MATERIAL: Mt. Spurgeon, 25 miles northwest not nearly so large as norfolcensis, nor is its of Cairns, 6. Darlington took 5 at the same place. P4 unreduced as in that species. We were informed that large flying pha- ACROBATES DESMAREST langers, living more or less gregariously, occur in the rain forests to the south of Cooktown. Acrobates DESMAREST, 1818, Nouveau diction- Those may, however, have been Schoinobates. naire d'histoire naturelle, nouv. ed., vol. 25, p. 405. The type locality of reginae is Bundaberg, This is the smallest of the gliding phalan- which is about 400 miles south-southeast of gers, resembling but not very closely related Mt. Spurgeon. The American Museum col- to Petaurus and to Schoinobates (the latter in lection includes two from the Dawson River a different subfamily). I showed (Tate, 1938; Valley, 150 miles west of Bundaberg. Though 1946, pp. 10-11) that on the basis of its immense areas of open-forest country with dentition Acrobates stands as a somewhat very little rain forest intervene, no differ- isolated member of the Phalangerinae. ence between Mt. Spurgeon and Bundaberg The genus is monotypic, comprising weakly material is appreciable. distinguishable northern and southern races. It is probably confined to eastern Australia Petaurus breviceps Waterhouse where its extends, in suitable local environ- This species, much smaller than P. austra- ments, from the extreme southern tip of Cape lis, has by far the widest range of any. It York Peninsula to Victoria. I am reasonably occurs through the whole of New Guinea sure that it does not occur in New Guinea, below 8000 feet, on many outlying islands, as I explained in 1938. 588 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 The animals seem to inhabit chiefly open forest. The only one of ours not so found PHASCOLARCTINAE came from Peach River, and there the open PSEUDOCHEIRUS OGILBY forest was only some 50 yards away from the Pseudocheirus OGILBY, 1837, Mag. Nat. Hist. river forest. (Charlesworth), new ser., vol. 1, p. 457. Acrobates pygmaeus frontalis De Vis I dealt with this genus in some detail seven Acrobates pygmaeus frontalis DE Vis, 1887, years ago (Tate, 1945b). Of the several sub- Proc. Linnean Soc. New South Wales, ser. 2, vol. genera, three, Pseudocheirus, Pseudochirops, 1, p. 1134. and Hemibelideus, are found in northeastern MATERIAL: Newcastle Bay, 1. In Brisbane: Queensland. The first has two species; the Peach River, 1. M.C.Z.: Peach River (Darling- others have one each. Hemibelideus is endemic. ton), 1. Ravenshoe, 2 in alcohol; Herberton (Raven), 1 in alcohol (A.M.N.H. No. 65653). Pseudocheirus peregrinus (Boddaert) Queensland Museum: Portland Roads, 1; Rolling- The races of this species have been dis- stone, on railroad 30 miles northwest of Towns- cussed earlier (Tate, 1945b, p. 9). The pres- ville, 1. "West Australia," 1 (M.C.Z. No. 6404). ent race is the most northerly and was the The specimen from Newcastle Bay was first of those races to be described. obtained under peculiar conditions. One night Van Deusen had shot at some tiny creature Pseudocheirus peregrinus peregrinus among the blossoms of a bloodwood tree. He (Boddaert) saw something fall in the darkness but de- Didelphis peregrinus BODDAERT, 1785, Elenchus spite protracted search with a flashlight he animalium, p. 78. failed to find it. Returning in the morn- MATERIAL: Croll Creek, Coen, 5 and 5 young. ing, he continued to search and under a (All by Neuhauser): Coen, 2; Port Stewart, 1. near-by fallen tree discovered a large carpet M.C.Z.: Coen (Darlington), 1. snake. On opening the stomach of the snake a All the Croll Creek series was collected by small bandicoot and an Acrobates were found. "shining" in the late evening in open eucalyp- It remains for X-rays to show whether or tus forest. The area appeared unpromising not the Acrobates contains lead pellets. by day, because a ground fire had passed The Peach River specimen, a male, was through it a short time before. But the taken in an ordinary rat trap placed at the Pseudocheirus, which apparently hid in holes foot of one of the giant trees standing on the in the trees, seemed to have been unaffected. flood plain of the river. Its measurements Four of those collected were females. Two of were: head and body, 80 mm., tail, 82; hind them had one, and two had two, young in foot (s.u.), 14; ear (from notch), 12. I have the pouch. Elsewhere we did not find a single trapped a flying squirrel in Virginia in just specimen of this form, the type of which was the same circumstances. The gliders appar- collected by Captain Cook at ently float down to the tree base and, very River (Cooktown). rarely, drop onto the trap. The several young permit a study of the The type locality of this northern race is order of tooth eruption in this species. The the Herbert River, and the series from near youngest are pouch twins (A.M.N.H. Nos. Ravenshoe must be virtually topotypical. 154382, 154383), in which the body hair is The additional material listed now serves still short and straight. In them i2, p3, p4, and to outline the north-south range of this little ml are fully erupted, while il, i8, c, p1, and m2 form fairly completely, though its western are partly through. In the lower jaw two limits have yet to be determined. The typical vestigial teeth (incisors, canine or premolars) race, A. p. pygmaeus, is amply represented in and ml are fully erupted, and the chisel-like southeastern Queensland. There are a few incisor, P4, and m2 are partly erupted. I have intervening localities including Clermont, never been able to detect dp4 in this genus. and Banana, 90 miles south of Rockhampton. Probably if they develop at all, they are Whether the animals from those places are minute and are shed at a stage even younger assignable to the northern or the southern than those mentioned above. Thomas (1887) race is still uncertain. reported them in Phascolarctos. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 589 In slightly older individuals, with the body are reduced to small buffy areas; and the hair already elongated and woolly, all the body fur, particularly of the head, has a previously mentioned upper anterior teeth brownish cast. There is ample white on the except the canine are functional, though tail; about two-fifths of its length is white. il-1 have not yet become elongate as in adults, This form thus approaches true laniginosus and m2 is also in place. In the lower jaw the of New South Wales in appearance, though incisor has lengthened, and p4 and m2 are the New South Wales race is even browner. both fully erupted. These older individuals were in each case closely associated with the [Pseudocheirus rubidus Troughton mother and probably still used the pouch. and Le Souef] Typical peregrinus is much darker than its Pseudocheirus rubidus TROUGHTON AND LE geographically closest subspecies, incanens, SOUEF, 1929, Rec. Australian Mus., vol. 17, p. 294. from the upper Mitchell River country (lati- MATERIAL: Crest of Bunya Mountains, 3. tude 170 S.) and southward along the Ather- This animal is sometimes killed by owls. ton Tableland west of the rain-forest edge. A number of parts of skulls and jaws were Its forearms and legs are dark gray like its found among the casts of an owl among the body. Its hands and feet are nearly black, roots of a giant Ficus. My specimens were and the extent of the white tail tip is shorter all in the rain forest which margins the open (50 to 75 mm. only). The white patch on the areas at the crest of the range. ear is reduced, and the remaining part of the ear is blackish. The under parts are dull Pseudocheirus herbertensis (Collett) whitish, with the bases of the hairs gray. This is a thoroughly distinct species be- Pseudocheirus peregrinus incanens longing in the subgenus Pseudockeirus, char- (Thomas) acterized by its elongate skull and palate Pseudocheirus laniginosus incanens THOMAS, and its brownish black or gray pelage ac- 1923, Ann. Mag. Nat. Hist., ser. 9, vol. 11, p. 249. companied by pure white under parts. Asym- MATERIAL: Junction of the McLeod and Escape metrical white markings, such as a white arm creeks, upper Mitchell River, 2. H. C. Raven se- or leg, are also quite common. cured 17 at Snubby Creek, 6 miles southwest of The species is endemic to the rain-forest Ravenshoe. Neuhauser collected 8 at Mt. Spur- mass that reaches almost continuously from geon, which is just east of, and above, McLeod Junction. M.C.Z.: Mt. Spurgeon (Darlington), 2. Townsville to Cooktown. C.N.H.M.: Ravenshoe (Scott), 1. U.S.N.M.: (All by Hoy), 5 miles southwest of Ravenshoe, 11; Mt. Pseudocheirus herbertensis herbertensis Bellenden Ker, 1. (Collett) Phalangista herbertensis COLLETT, 1884, Proc. Pseudocheirus laniginosus oralis Thomas Zool. Soc. London, p. 353. Pseudocheirus laniginosus oralis THOMAS, 1926, MATERIAL: Evelyn, 1; 9 miles southeast of Ann. Mag. Nat. Hist., ser. 9, vol. 17, p. 631. Ravenshoe (H. C. Raven), 6. (All by Neuhauser): MATERIAL: Mt. Spec, Paluma Range, 20 miles Evelyn, 7; Lake Barrine, 1; Danbullan, 1. M.C.Z.: west of Townsville, 1; Serpentine, tributary of Mt. Spurgeon (Darlington), 5; Millaa Millaa Fitzroy River, 20 miles northeast of Rockhamp- (Darlington), 4; Ravenshoe (Darlington), 3; ton, 1. Raven collected 1 at Mundubbera, 150 Babinda Creek (Schevill), 1. C.N.H.M.: Ravens- miles north-northwest of Brisbane. The type local- hoe (Scott), 3. U.S.N.M.: Nine miles south of ity is Bloomsbury, 50 miles northwest of Mackay Ravenshoe (Hoy), 4; Herberton District (Hoy), 1. and a few miles from Proserpine; it is rain-forest We did not find this species, probably because of country. the short time we spent in the Walter Hill Range. The general impression one gets of orals This blackish Pseudocheirus slightly re- is that it is a coastal race living for the most sembles a short-haired Pseudocheirus lemu- part in a relatively moist environment. Mt. roides, but has pure white instead of smoky Spec was a particularly rainy place. The color- under parts. ation is substantially like that of incanens, The form from Boar Pocket which Waite but the limbs are much more strongly colored, named colletti seems to be virtually insepa- russet rather than light tan; the ear patches rable from typical herbertensis. 590 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 Pseudocheirus herbertensis cinereus Tate Pseudocheirus (Hemibelideus) lemuroides Pseudocheirus herbertensis cinereus TATE, 1945, (Collett) Amer. Mus. Novitates, no. 1287, p. 17. Phalangista (Hemibelideus) lemuroides COLLETT, MATERIAL: Only the typical series from Mt. 1884, Proc. Zool. Soc. London, p. 385. Spurgeon. Pseudocheirus (Hemibelideus) cervinus LONG- MAN, 1915, Mem. Queensland Mus., vol. 3, p. 22. Sharply distinguished by its pale, brownish MATERIAL: Walter Hill Range, 2. (All by Neu- gray color from the blackish form of the coast hauser), Danbullan, 1; Lake Barrine, 6; Mt. Spur- rain belt. Its skull presents no distinguishing geon, 4; Evelyn, 1. M.C.Z.: (All by Darlington), features and has the characteristic elongate Mt. Spurgeon, 4; Millaa Millaa, 1; Ravenshoe, 1. narrow palate of the series. C.N.H.M.: Ravenshoe (Scott), 4. U.S.N.M.: Nine miles south of Ravenshoe (Hoy), 10; Herberton SUBGENUS PSEUDOCEIROPS MATSCHIE District (Hoy), 2. The subgenus can be recognized by its in- Both our specimens were shot at night in flated mastoid region, uninflated inferior wall dense rain forest. This locality (in the Walter of the meatus, and compressed tail hairs. Hill Range) is only some 40 miles north of There is usually a striped color pattern. The Herbertvale, Herbert River, the type locality. body hair in the single Australian species One pale brown variety from 6 miles south archeri is greenish gray. of Ravenshoe is apparently the form cervinus. Pseudocheirus (Pseudochirops) archeri (Collett) PHASCOLARCTOS BLAINVILLE Phalangista (Pseudochirus [sic]) archeri COL- Phascolarctos BLAINVILLE, 1816, Nouv. Bull. LETT, 1884, Proc. Zool. Soc. London, p. 381. Sci. Soc. Philom., Paris, p. 116. MATERIAL: Lake Eacham, 2. Lake Barrine This is a greatly specialized relative of the (Neuhiuser), 5; Evelyn (Neuhauser), 2. H. C. prehensile-tailed genus Pseudocheirus. Besides Raven collected at least 1 at Snubby Creek, 6 miles southwest of Ravenshoe. M.C.Z.: (All by the thickening of the body and shortening of Darlington), Cardwell, 1; Lake Barrine, 1; Millaa the tail to a rudiment, Phascolarctos shows Millaa, 2; Mt. Spurgeon, 1. C.N.H.M.: (All by many specializations of the skull. The pos- Scott), Danbullan Road, 2; Wongabel, 1; Ravens- terior part of the palate, with the molars, the hoe, 1. U.S.N.M.: Six miles south of Ravenshoe pterygoids, and the large, inflated bullae, is (Hoy), 1; Mt. Bellenden Ker (Hoy), 1. extended low in the skull, while, in compensa- Those taken at Lake Eacham were living tion, the articular and coronoid processes of in a remnant of dense rain forest, and prob- the mandible have been correspondingly ably that is the true habitat of this species. deepened. This deepening is not carried However, it must occasionally travel along through to the anterior part of the palate. the gallery woods bordering streams, because Instead, the facial portion of the skull has Raven collected it at Snubby Creek in an undergone a quite different modification, environment most of which is open forest. namely, extreme shortening of the nasals The species, the nearest relative of which and maxillaries in combination with great is the Papuan corinnae, is endemic in the rain broadening of the rostrum as a whole. In the forests and radiating gallery woods of the dentition the chief difference from Pseudo- Cairns-Atherton Tableland-Herbert River cheirus lies in the total elimination of the area. upper and lower rudimentary teeth between the canine (incisors in the lower series) and SUBGENUS HEMIBELIDEUS COLLETT the last premolars. This leaves upper and Hemibelideus is a misnomer, there being no lower diastemata, which are relatively short trace of a gliding membrane. owing to the above-mentioned shortness of The short broad skull comes nearest to that the rostrum. of Pseudochirops. The inferior wall of the Corresponding to the lowering of the hind meatus is inflated and cellular. The pelage palate there is developed an ample system of the body is very soft and long, and that of post-narial sinuses, and the vomer is ex- of the tail stands nearly erect. traordinarily deepened. This is accompanied There is one species, lemuroides. externally by inflation of the maxillaries be- 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 591 low the orbit and just above the third and MATERIAL: Mt. Spurgeon (Neuhauser), 6; Dan- fourth molars. The long paroccipital proc- bullan (Neuhauser), 1; Evelyn (Neuhauser), 1, esses have undergone a separate specializa- (Raven), 7; Dinner Creek, 9 miles southeast of tion and flare so widely they are far wider Ravenshoe (Raven), 16. M.C.Z.: Ravenshoe (Darlington), 2; Lake Barrine (Darlington), 1; across the tips than the greatest distance Mt. Spurgeon (Darlington), 1. C.N.H.M.: Ravens- across the bullae (57:45). In Pseudocheirus, hoe (Scott), 2. U.S.N.M.: Five miles southwest this ratio is 17:21. Finally the audital of Ravenshoe (Hoy), 6. meatus, so conspicuous in Pseudocheirus, is Collett recorded only "northern Queens- reduced to almost nothing and withdrawn land" as type area, but Lumholtz, who col- almost into the bulla. Probably most of the lected the animals, alluded twice to them in changes described can be interpreted in his book, and on each occasion he was not terms of function. far from Herbertvale. Phascolarctos is today monotypic; the As do many of the specimens of Pseudo- species is cinereus Goldfuss. cheirus, this species of Schoinobates appears Phascolarctos cinereus adustus Thomas ? to be confined to the rain forests between Phascolarctos cinereus adustus THOMAS, 1923, Cooktown and Townsville. Schoinobates volans Ann. Mag. Nat. Hist., ser. 9, vol. 11, p. 246. occurs at Rockhampton and southward to The are becoming increasingly rare. Victoria. No specimen was collected by the Archbold At Brisbane is one (No. 4738) from Kirra- expedition, but one was reported seen a few ma, Herbert River, and Nos. 362 and 363 are years earlier on the railroad line between the type series of cinereus Ramsay, Brisbane Cooktown and Laura. If correct, this record Nos. 3663 and 3664 are from 6 miles south- would constitute the northernmost for the west of Ravenshoe. The last is open country species. but is crossed by gallery woods. We have one specimen taken at Pentland, [Schoinobates volans (Kerr)] west of Townsville. Didelphis volans KERR, 1792, The animal king- SCHOINOBATES LESSON dom ... of Linnaeus, pt. 1, p. 199. Schoinobates LESSON, 1842, Nouveau tableau du MATERIAL: The Serpentine, 10 miles north of regne animal, p. 190. Rockhampton, 1. This is the only living representative of the I shot this specimen by hunting light. It Phascolarctinae possessing gliding parachutes was at the top of a very tall gum tree that substantially similar to those of Acrobates stood very close to a forest fire, and it must and Petaurus. The structure of the skull and definitely have been threatened. The eyes the selenodont dentition furnish ample evi- glowed slightly reddish. dence of the true relationships of the genus MACROPODIDAE and strongly resemble those of Pseudocheirus The kangaroos and comprise (Hemibelideus). three living subfamilies: the Hypsiprymno- Schoinobates was long held to be monotypic. dontinae, which include only the bandicoot- The northern form minor differs, however, like Hypsiprymnodon; the Potoroinae, or hare so sharply from the southern and central wallabies, comprising several genera, some volans that I think a better understanding of of which occur in the Cape York area; them can are as be had if they treated full and the , which include all species. the remaining genera, quite a number of I a pointed out number of separating char- which are found on the Atherton Tableland acters where I genus discussed the earlier and in the Cape York Territory. I have re- (Tate, 1946, p. 11). cently (Tate, 1948a) revised the entire family. Schoinobates minor (Collett) HYPSIPRYMNODONTINAE Petaurista volans var. minor COLLETT, 1887, Zool. Jahrb., vol. 2, p. 926. Hypsiprymnodon is the only living genus Petauroides cinereus RAMSAY, 1890, Rec. Aus- of this division of the kangaroos. One or two tralian Mus., vol. 1, p. 77. related fossils are known. 592 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 HYPSIPRYMNODON RAMSAY Aepyprymnus rufescens (Gray) Hypsiprymnodon RAMSAY, 1876, Proc. Linnean Bettongia rufescens GRAY, 1837, Mag. Nat. Soc. New South Wales, vol. 1, p. 33. Hist. (Charlesworth), vol. 1, p. 584. moschatus Ramsay Hypsiprymnus melanotis OGILBY, 1838, Proc. Hypsiprymnodon Zool. Soc. London, p. 62. Hypsiprymnodon moschatus RAMSAY, 1876, MATERIAL: Near Evelyn, 1; Snubby Creek, 6 Proc. Linnean Soc. New South Wales, vol. 1, miles southwest of Ravenshoe (Raven), 6. M.C.Z.: p. 34. Mt. Molloy (Darlington), 1; Yungaburra (Darling- Pleopus nudicaudatus OWEN, 1877, Ann. Mag. ton), 1; Mt. Coolon (Darlington), 1. C.N.H.M.: Nat. Hist., ser. 4, vol. 20, p. 542. Bluewater Creek, 20 miles northwest of Towns- MATERIAL: The Tableland, 6 miles east of Hel- ville (Scott), 1; Ravenshoe (Scott), 1. U.S.N.M.: envale, 1; Mt. Finnegan, Top Camp, 1; Julatten, Five miles south of Ravenshoe (Hoy), 8. west of Mossman, 1; 9 miles southeast of Ravens- hoe (Raven), 4. M.C.Z.: Lake Barrine (Darling- I saw a dead one on the road some miles ton), 3; Mt. Spurgeon (Darlington), 1. C.N.H.M.: northeast of Rockhampton. Raven collected Wongabel (Scott), 1. two at Lochaber Station, Mundabbera Dis- Our record from The Tableland near Helen- trict, in the Dawson River Valley. vale is apparently the extreme recorded limit MACROPODINAE to the north. The range seems to include the forests on the east side of the Atherton Table- Comparatively few of the genera of this land generally. subfamily appear in the Cape York fauna, I saw, but failed to shoot, several of these and some of the records are doubtful. They tiny macropods at dusk in the rain forest at include Onychogalea, Petrogale, Dendrolagus, the upper Clohesy River, a tributary of the Thylogale, Protemnodon, and Macropus, with Barron River. They looked like small blackish its subgenus Osphranter. Lagorchestes is in bandicoots as they flashed across the aban- the doubtful list. doned lumber roads. I shot the Julatten speci- Petrogale is to be sought on the scrubby, men very early in the morning in a similar rocky hills that crop out here and there on situation. At Mt. Spec I had a glimpse of the Cape. Dendrolagus is strictly confined to one individual. rain forest. We did not find it north of Cook- town, though it should be looked for still in POTOROINAE the McIlwraith and Iron ranges. Thylogale Only Bettongia and Aepyprymnus extend is a common, ground-living species in the into the Cape York area, the former probably rain-forest zone extending from the extreme not beyond the Atherton Tableland, the latter tip of Cape York to New South Wales. Pro- perhaps to some few miles south of Cooktown. temnodon is represented by agilis, bicolor, BETTONGIA GRAY and parryi only. Bettongia GRAY, 1837, Mag. Nat. Hist. (Charles- PETROGALE GRAY vol. 584. worth), 1, p. Petrogale GRAY, 1837, Mag. Nat. Hist. (Charles- Bettongia penicillata Gray worth), vol. 1, p. 583. Bettongia penicillata GRAY, 1837, Mag. Nat. Hist. (Charlesworth), vol. 1, p. 584. Petrogale inornata godmani Thomas MATERIAL: M.C.Z.: Mt. Spurgeon (Darling- Petrogale godmani THOMAS, 1923, Proc. Zool. ton), 3; U.S.N.M.: Nine miles south of Ravens- Soc London, Abstr., no. 235, p. 13; p. 177. hoe (Hoy), 3. MATERIAL: Rocky cliffs near Annan River Falls, 2 miles west of type locality, 1, female with pouch At Shipton's Flat a small wallaby, not seen young. M.C.Z.: Coen (Darlington), 4; Mt. Car- by us, was said to inhabit areas of "bladey bine (Darlington), 1. U.S.N.M.: Five miles south- grass" growing 2 to 3 feet tall in open spaces. west of Ravenshoe (Hoy), 2. This may have been Bettongia, or it may have The female from the Annan Falls is a young been Aepyprymnus. adult in which m4 are only just visible and AEPYPRYMNUS GARROD p and dp have not yet been shed (their Aepyprymnus GARROD, 1875, Proc. Zool. Soc. crown lengths are, respectively, 4.7 and 4.8 London, p. 59. mm.). 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 593 The juvenal is so young that only the ex- LAGORCHESTES GOULD treme tips of the first incisors are visible. Lagorchestes GOULD, 1841, A monograph of the There are no "milk incisors" such as I de- Macropodidae, pt. 1, pl. 12 and text. scribed in Macropus (Tate, 1947a, p. 399). The [Lagorchestes conspicillatus leichardti tympanic rings are large and widely separated Gould] from the alisphenoid and as yet give no indi- Lagorchestes leichardti GOULD, 1853, The mam- cation of the swelling which in the adult mals of Australia, vol. 2, pl. 60 and text. forms the tympanic bulla. MATERIAL: M.C.Z.: Army Downs, 35 miles Rock wallabies, presumed to be of the race north of Richmond, central north Queensland godmani, had been common 150 miles farther (Schevill), 1. north, among the sandstone scarps on the This locality must be on the somewhat hills a mile or two to the east of Wenlock on elevated basaltic formation on the north side the Batavia River. But just before our arrival of the Flinders River. I could see it from the fire had swept through the hills and the train. The type locality of leichardii is reputed animals had been driven away, at least tem- to have been between Normanton and Dar- porarily. Their droppings were still to be seen win. among the rocks. A specimen from Rockhampton collected DENDROLAGUS SCHLEGEL AND MtLLER Scott No. 64360) is clearly a Dendrolagus SCHLEGEL AND MULLER, in Tem- by (C.N.H.M. minck, 1839-1844, Verhandelingen over de Na- form of inornata. tuurlijke Geschiedenis du Nederlandsche over- [At Mt. Isa (Rifle Creek Dam in western zeesche besittingen, Zoologie, vol., 1, p. 130. Queensland) rock wallabies were fairly plenti- ful. Their eyes shone under a light, and they Dendrolagus bennettianus De Vis were caught rather easily with jump traps Dendrolagus bennettianus DE Vis, 1887, Proc. placed under overhanging rocks. They are Roy. Soc. Queensland, vol. 3, p. 13. referable to the race purpureicollis, the type MATERIAL: Mt. Finnegan, Top Camp, 1, (Rob- locality of which is Dajarra, near Cloncurry.] erts), 2. This seems to be an Australian offshoot of ONYCHOGALEA GRAY Dendrolagus dorianus of Papua. Onchogalea GRAY, in Grey, 1841, Journals of Our specimen was one of two animals found two expeditions ... in ... Australia, vol. 2, App., by two of our aboriginal "boys" in the dense p. 402. rain forest near the top of Mt. Finnegan. The Onychogalea unguifer (Gould) gray-brown skin shows a pallid area, near Macropus unguifer GOULD, 1840, Proc. Zool. straw color, at the base of the tail. The dorsal Soc. London, p. 93. surface of the extreme proximal part of the MATERIAL: M.C.Z.: Port Stewart (Darlington), tail is worn bare of hairs as is usual in this 1. genus. The under surface of the tail proxi- This specimen, a female, is colored light mally and mesially is black. brown, with a darker brown stripe extending The lengths of p4 in two adult females are from the middle of the back to the basal 10.0 mm. and 9.7, respectively; in adult fe- third of the tail. The face and head are gray- male dorianus, 10.5 mm. In a young male p3 ish, and the area beneath and behind the eyes (deciduous) is 5.7 mm. in length, compared is white. The under parts and hip stripes are with 6.1 in a young male dorianus. In an creamy white, the sides and under side of the adult female of lumholtzi (from Atherton tail buffy white. That the animal is young is Tableland) p4 iS only 9.4 mm. in length. shown by the fact that the m4 are only partly The deciduous p8 and dp4, Dendrolagus lulmholtzi Collett erupted. teeth, Dendrolagus lumholtzi COLLETT, 1884, Proc. are still in place. The slender upper canines Zool. Soc. London, p. 387. characteristic of the genus are evident. The Dendrolagus fulvus DE Vis, 1888, Proc. Roy. crown length of ps is 3.8 mm.; of dp4, 4.8. Soc. Queensland, vol. 4, p. 132. This is a very important record, since the MATERIAL: I reported an ample series of this type locality was in the northwest of Austra- species earlier (Tate, 1948a). C.N.H.M.: Won- lia. gabel (Scott), 4; Ravenshoe (Scott), 1. 594 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 TABLE 2 DENTAL DIMENSIONS OF Dendrolagus bennettianus

Sydney A.M.N.H. A.M.N.H. AMNH No. M.1818, No. 155115Q No. 155114, NoA.M1N363.0H. Juv., UV. 6 Mt. Finnegan Cooktown Mt Finneganinea Mt. Finnegan

p4, length - 9.3 - 8.8 p4, breadth (greatest) 6.7 6.0 mt, length 5.1 5.9 5.6 5.5 mt, breadth 5.0 5.7 5.7 5.4 m2, length 6.0 6.2 6.0 5.7 m2, breadth 5.4 5.9 5.8 5.5 Ms, length 6.3 6.7 - 6.4 mi, breadth 5.8 5.8 5.4 m4, length 6.5 6.9 6.9 m4, breadth 5.6 5.8 5.5 dps 4.6X3.4 - 5.5 X3.8 - dp4 4.5X2.7 5.1X4.5 -

THYLOGALE GRAY for the southern race stigmatica to be more ThylogaeGRAY, 1837, Mag. Nat. Hist. (Charles- heavily built, and this is supported by the worth), vol. 1, p. 583. fact that in it p' are decidedly heavier teeth than in coxenii. To some degree is is also Thylogale stigmatica (Gray) larger. I compare their dimensions in table 3 Our fairly ample collections from the tip below. of the Cape York Peninsula representing It is reasonably apparent from table 3 that coxenji have made possible detailed compari- specimens taken near the middle of the sons with the Cardwell-Cairns race which Peninsula (at Mt. Tozer, Iron Range, and is true stigmatica. Regarding the skins of Portland Roads) partake of the light build these two races, both vary considerably, from of the form coxenii, and that the Papuan more reddish on the sides and flanks to less form oriomo is similarly lightly built. In the reddish, and the overlap in this respect is case of the latter, however, the skins appear complete. I do not believe that they can be to be consistently darker, and provisionally distinguished on the basis of color alone. I think that race should be maintained. We But in the skulls there is a definite tendency can then say that the subspecies stigmatica TABLE 3 COMPARATIVE DIMENSIONS OF P4 Anterior Posterior Length Width Width

stigmatica from base of Cape York Peninsula 3 adult males 7.3-7.8 3.2-3.6 4.0-4.6 1 adult female 7.4 3.0 3.9 coxenii from the tip of the Peninsula 2 adult females 7.0-7.4 3.0-3.3 3.7-3.9 Specimen from the middle of Peninsula (Mt. Tozer, Portland Roads, etc.) 1 adult male 7.5 2.8 3.7 oriomo from southern New Guinea Adult male 7.4 2.9 3.8 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 595 occurs throughout the Atherton-Cairns rain It seems that the two specimens from forests to Mossman and Mt. Finnegan, some Brooklyn Station near Mt. Carbine, formerly 30 miles south of Cooktown. North of Cook- referred to agilis, should instead be placed town comes the open-forest gap, mentioned with jardinei. in the Introduction to this paper. Beyond The likeness of the present series to the that there are again rain forests in the area south Papuan race papuanus, of which we of the McIlwraith and Iron ranges and the have many species, is very great. The latter tip of the Peninsula, and this block of forest has the hind back slightly more reddish, a is seemingly the home of the small-toothed feature that may be seasonal and transitory. coxenii. I find no cranial or dental differences. In a juvenal male (A.M.N.H. No. 154501), The general color of the race jardinei is a from the Iron Range, a pair of bead-like, grayish brown, with little or no admixture of minute, deciduous incisors, each about 1 mm. reddish. The hip stripes are as usual quite in length, can be seen in the position that conspicuous. later would have been occupied by i2_i3 (see The fourth molars are fully in place only also Tate, 1947a). in exceptionally old specimens. In our col- lection they are fully developed in three males Thylogale stigmatica coxenii (Gray) (A.M.N.H. Nos. 153650, 153661, 153666) Halmaturus coxenii GRAY, 1866, Proc. Zool. and one female (A.M.N.H. No. 153654). Soc. London, p. 220. A juvenal specimen (A.M.N.H. No. 154504) Halmaturus gazella DE VIS, 1884, Proc. Roy. shows a deciduous tooth in the incisive row. Soc. Queensland, vol. 1, p. 110. This tooth, with its length 0.8 mm., is in- MATERIAL: Lockerbie and localities near by, tip forward as it is located of Cape York Peninsula, 5 with 1 pouch young; clined slightly and, Mt. Tozer, 1; Portland Roads, 3; Iron Range, 1. slightly behind the maxillo-premaxillary su- ture, may represent a canine. A similar de- As is its southern relative, true stigmatica, ciduous incisor is mentioned under the head- this race was closely confined to the rain ing Thylogale stigmata above. forest. Thylogale stigmatica stigmatica Gould Protemniodon parryi (Bennett) Macropus parryi BENNETT, 1835, Proc. Zool. Thylogale stigmatica GOULD, 1860. The mam- Soc. London, for 1834, p. 151. mals of Australia, vol. 2, pls. 33, 34 and text. MATERIAL: Mt. Carbine (Neuhauser), 2. Eve- MATERIAL: Mossman Gorge, 2; Mt. Finnegan, lyn, Kaban, Ravenshoe, all on the western, open- 2; Yungaburra, 21. M.C.Z.: (All Darlington), Mt. and Spurgeon, 1; Lake Barrine, 2; Millaa Millaa, 2; forest side of the Atherton Tableland (Raven Mt. Carbine, 1. Shipton's Flat (Roberts), 19. In Hoy), several specimens. addition, material reported before (Tate, 1948a). We did not obtain this species. Some 30 miles up the railroad west of Cooktown it was PROTEMNODON OWEN reported once common but becoming in- Protemnodon OWEN, 1874, Phil. Trans. Roy. creasingly rare. We spent two days out from Soc. London, vol. 164, p. 274. Seagren's Farm (Kilometer 10) hunting whip- For use of Protemnodon in place of Walla- tails and saw two that, from their long-tailed bia, see Tate (1948a, pp. 295-296). shape, appeared to be parryi, but we could not get near them. Protemnodon agilis jardinei (De Vis) Halmaturus jardinei DE VIs, 1884, Proc. Roy. Protemnodon bicolor apicalis (Gunther) Soc. Queensland, vol. 1, p. 109. Halmaturus apicalis GUNTHER, 1874, Proc. MATERIAL: Tip of Peninsula: Lockerbie, 4; Zool. Soc. London, p. 653. Newcastle Bay, 1; , 1 with pouch MATERIAL: Somerset (Neuhluser), 1; Albany young; Red Island Point, 1. Friday Island, Torres Island (Neuhiauser), 1; Lake Barrine (Neu- Strait, 1 (a pick-up skull). Middle of Peninsula: hauser), 1. M.C.Z.: Millaa Millaa (Darlington), Portland Roads, 4; Iron Range, 1; Wenlock, 2. 1. U.S.N.M.: Five miles southwest of Ravenshoe The Cooktown-Coen area: Keating's Lagoon, 1; (Hoy), 2; Mt. Bellenden Ker (Hoy), 1. Darling- Ebagoolah, 1. C.N.H.M.: Bluewater Creek ton also took one at the Bunya Mountains near (Scott), 2. M.C.Z.: Coen (Darlington), 2. Toowoomba in southern Queensland. 596 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 Wallabies, called by the local hunters It is this red that causes local hunters to con- "swampers," are reported rarely from the fuse them with the true , M. Cooktown area. They are said to occur in rufa. Our male and female from Coen, both swamps across the to the of which are rather younger, are predomi- northwest of Cooktown. nantly gray. In a juvenal (A.M.N.H. No. 154508), a minute milk incisor, such as I [Protemnodon dorsalis (Gray)] described earlier in Macropus (Tate, 1947a) Halmaturus dorsalis GRAY, 1837, Mag. Nat. and above in this paper in Thylogale and Hist. (Charlesworth), vol. 1, p. 583. MATERIAL: The Serpentine, 25 miles southeast Protemnodon, is present. of Rockhampton (Tate), 1. M.C.Z.: Bunya [SUBGENUS MEGALEIA GIsTEL] Mountains (Darlington), 1. Megaleia GISTEL, 1848, Naturgeschichte Thier- This is a species of south and central reichs hohere Schulen, Stuttgart, p. ix. Queensland and northern New South Wales. [Macropus (Megaleia) rufus (Desmarest)] MACROPUS SHAW AND NODDER Kangurus rufus DESMAREST, Mammalogie, Macropus SHAW AND NODDER, 1790, The suppl., vol. 2, p. 541. naturalist's miscellany, vol. 1, pl. 33 and text. I can find no evidence that the true red SUBGENUS OSP}RANTER GOULD kangaroo (Megaleia) is present on the Cape Osphranter GOULD, 1842, Proc. Zool. Soc. Lon- York Peninsula. The nearest records I have don, for 1841, p. 80. are from the Flinders River area near Rich- Macropus (Osphranter) robustus erubescens mond and Hughendon, where Darlington Sclater shot a number. Possibly the species extends a short way up the west side of the Cape to Macropus erubescens SCLATER, 1870, Proc. the lower course of the Mitchell River. The Zool. Soc. London, p. 126. of the Batavia MATERIAL: Wenlock, Batavia River, 5 c", 9; so-called "red kangaroos" River, much farther north, are in reality red Coen, 1 ci, 1 9. M.C.Z.: Coen (Darlington), 2; Mt. Carbine (Darlington), 1. I have already re- . ported (Tate, 1948a) those taken by Raven and by Neuhauser. [Rifle Creek, Mt. Isa, northwest SUBGENUS MACROPUS SHAW AND NODDER Queensland, 6 (skulls only).] Macropus (Macropus) canguru canguru Muiller The wallaroos of the Batavia River area vary in color from dusky rufescent to quite Mus canguru MtLLER, 1776, Linn6... voll- bright rufous. They seem to prefer the standiges Natursystem.. . swdlfter Lateinischen rolling Ausgabe, p 62. grassy hillsides of the Sir William Thompson MATERIAL: Sir William Thompson Range, near Range a little to the east of Wenlock. Wenlock, 1 (A.M.N.H. No. 153611); Ebagoolah, Hundreds of miles to the south, on the 30 miles south of Coen, 1; 5 miles west of Helen- cliffed sandstones near Alderbury (Kilo- vale, 2 and 1 pouch young. C.N.H.M.: (All by meter 25), west of Cooktown on the railroad Scott), Craigmore and Wongabel, Atherton Ta- to Laura, I saw three blackish brown wal- bleland, 2; Bluewater Creek, 20 miles northwest laroos. I had two under observation for nearly of Townsville, 1. [Mirrabooka, 60 miles south of 10 minutes. They were across a small canyon, Charleville, south Queensland, 1 with pouch browsing among the rocks. Once they stood young.] up face to face and "boxed." The three examples from Helenvale are the In our series there appears to be an ele- specimens referred to by Morrison-Scott ment of dichromatism such as we see in (1950) in his recent article on the identity of Trichosurus v. erubescens. Three males and a Captain Cook's kangaroo. It seems likely female which are fully adult have red mixed that Helenvale marks nearly the limit north- with their gray dorsal pelage. In specimens ward of this kangaroo on the east (lowlands) that are red-gray the backs of the ears are side of the divide, though the species may blackish; in really red specimens the backs extend a considerable way northward behind of the ears are also red. From Wenlock two the sand dunes that lie between Capes Bed- fully adult males are almost completely red. ford and Flattery in the Normandy River 19S2 TATE: MAMMALS OF CAPE YORK PENINSULA 597 country. West of the divide we have our this enormous family occur in Australia. actual records from Sir William Thompson They are the Vespertilioninae, the Nycto- Range and from Ebagoolah, and along the philinae, and the Miniopterinae. The genus motor road from Ebagoolah, south to Laura, Phoniscus, representing the Kerivoulinae, is we saw several kangaroos that we considered listed by Troughton from "Queensland," but to be great grays. may be an accidental visitor. The Mirrabooka specimens represent the In an earlier paper (Tate, 1942a) I proposed form of great gray kangaroo that occurs in certain tribal subdivisions of the Vespertilio- the western parts of southern Queensland. nidae: the Myotini, the Pipistrellini, the Nyc- The animal from Sir William Thompson ticeini, and the Lasiurini. Of these only the Range, a male weighing 58 pounds, consti- last (American and Hawaiian) is absent from tutes the most northerly record of the great Australia. The genera present in the Cape gray kangaroo. York collection can be shown synoptically, Commenting on Morrison-Scott's paper, I as follows: should like to point out that since Cook and Myotini: Myotis his men spent some months at Cooktown they Pipistrellini: Pipistrellus, Chalinolobus, Eptesi- are likely to have encountered at least four cus kinds of kangaroos: the great gray (canguru), Nycticeini: Scoteinus, with its subgenera the (robustus), the very abundant The subfamily Miniopterinae contains only sand wallaby (agilis) and the less plentiful Miniopterus; the Nyctophilinae, only Nycto- whiptail or pretty-face wallaby (parryi). philus. Sand wallabies (maximum weight in our MYOTIS KAuP records, 53 pounds) even today occasionally Myotis KAUP, 1829, Skizzirte Entwickelungs- come right into Cooktown. I concur with Geschichte und nati3rliches System der Euro- Morrison-Scott's fixation of the name can- pkiischen Thierwelt, vol. 1, p. 106. guru on the great gray kangaroo. But I sug- From the completeness of its dentition, gest that the expression "photo-lectotype" this genus can be considered one of the least should be passed upon by the International specialized of all the Vespertilioninae. I Commission on Zoological Nomenclature. At reviewed the members of the genus present such an expression is non-existent in 1941e. Only two of the five subgenera under the Rules. reach Australia and New Guinea. Those two CHIROPTERA are Leuconde and a species australis, which in Both suborders and six families of bats oc- probably belongs Selysius. cur in Australia. Of the six families, Vesper- SUBGENUS LEUCON6E BOIE tilionidae, Molossidae, , Leucon6e BoIE, 1830, Isis, p. 256. Rhinolophidae, Megadermatidae, and Pter- A heavily built, large-footed Myotis with opodidae, five are known from the Cape York headquarters of the group in southeastern Peninsula. Only the Megadermatidae remain . The subgenus reaches (dauben- so far unrecorded. Macroderma, its only tonii, capaccinii) on the one hand and North Australian genus, is probably there, as John- America (grisescens) on the other. Only one son found signs of its predatory activities in form is known in Australia-macropus. Arnhemland. The Chiroptera must have been drifting Myotis (Leuconde) adversus macropus continuously into Australia from Asia (Gould) throughout the whole of Tertiary time, yet Vespertilio macropus GOULD, 1855, The mam- I am not aware that any fossils have been mals of Australia, vol. 3, pl. 47 and text. recorded in formations associated with such MATERIAL: Cairns, Freshwater Creek, 2, and 1 extinct Pleistocene genera as Nototherium in alcohol (A.M.N.H. No. 155010). and Thylacoleo. These bats were found in crevices in and between floor joists in the house of Mr. VESPERTILIONIDAE Robert Hunter, near the "water intake" for Three (possibly four) of the subfamilies of Cairns, northern Queensland. 598 BULLETIN AMERICAN MUSSEIJM OF NATURAL HISTORY VOL. 98 Compared with a series of M. adversus Peters distinguished Chalinolobus from moluccarum from New Guinea, the present "Vesperugo" and Scotophilus by its "short, bats are somewhat darker colored. The fore- rounded ears and the lobes of skin at the arms are 36.4 and 38 mm., against 38 to 42 corners of the mouth." In most species the mm. in moluccarum. The tooth rows, c-m4, dental formula is the same as that of Pipis- are 5.9 to 6.0 mm. The large hind feet are trellus. In C. rogersi p2 is stated to be absent, characteristic of the subgenus Leuconoe. though this is not always the case. Compared with Pipistrellus, the species of Chalinolobus SUBGENUS SELYSIUS BONAPARTE have the frontal region abruptly elevated, [Myotis australis (Dobson)] and the braincase is relatively much deeper. Vespertilio australis DOBSON, 1878, Catalogue of In the species before me (rogersi) p2 is ex- the Chiroptera in the ... British Museum, p. 317. tremely minute, less than one-third the diam- The type, which is in the museum at Lei- eter of i2, and is crowded into the inner den, came from New South Wales. We did corner between c and p4. In two of the speci- not collect this form. mens p2 is absent, though it is present in A.M.N.H. No. 154661. The tooth is ap- PIPISTRELLUS KAUP parently easily dislodged and lost when Pipestrellus KAUP, 1829, Skizzirte Entwicke- skulls are cleaned. In the incisors i2 is quite lungs-Geschichte und nattirliches System der small, its crown less than one-third of the Europiiischen Thierwelt, vol. 1, p. 98. depth of il. The last molar is not shortened This large genus, distributed almost in the tooth row. In the lower jaw the incisors throughout the world, has hitherto been rep- overlap, as in Pipistrellus; P2 is almost one- resented in Australia by three recorded half as tall as P4. species: abramus, regulus of Western Austra- lia, and tasmaniensis of Tasmania and the Chalinolobus rogersi Thomas southern parts of Australia. Pipistrellus abra- Chalinolobus rogersi THOMAS, 1909, Ann. Mag. mus is characteristic of the Asiatic tropics Nat. Hist., ser. 8, vol. 3, p. 150. and reaches well into China. The other two MATERIAL: Cooktown-Laura Railroad (10 species are truly Australian, and Troughton miles), Seagren's Farm, 3; Shipton's Flat (Rob- (1948) has proposed generic rank for each of erts), 5 in alcohol. Shot at dusk. them. These bats lack the postorbital process of I have to record the presence in the Cape the zygoma, and p2 when present is exceed- York area of the small species papuanus. ingly minute, its crown area very much less Pipistrellus papuanus is the southern repre- than that of i2. Likewise p2 is a much reduced sentative of the P. tenuis group of India and tooth. Its single cusp is only half the height the Malay region. (about the cingulum) of the anterior cusp Pipistrellus papuanus (Peters and Doria) of P4. Forearms measure 34 to 36.5 mm.; c-mi3, 5.1. Groups of Pipistrellus that some- Vesperugo papuanus PETERS AND DORIA, 1881, Ann. Mus. Civ. Genova, vol. 16, p. 696. what agree with the foregoing are the kuhiii MATERIAL: Brown's Creek, 1; Wenlock, Ba- and savii groups. The extreme degree of tavia River, 1; , 1 (all males). reduction of p2 favors the latter. The higher and deeper braincase in the present species All these specimens were shot while flying seems to rule out Pipistrellus, though the at dusk. The forearms measure 30.0 to 30.5 close relationship of Chalionolobus to Pi- mm.; c-m3, 4.3 mm. in each case. pistrellus is very obvious. The three skins At Haveri, inland from Port Moresby, with frosted hair tips are identical in appear- British New Guinea, I found a colony of this ance. The retention of p2-2 in one of the skulls little species secreted in the gables of a house. merely shows that total obsolescence of these They were netted at dusk as they dived teeth, contrary to the usual understanding of downward in flight. C. rogersi, is not invariable. There is a slight CHALINOLOBUS PETERS difference in the profiles of the skulls at the Chalinolobus PETERS, 1866, Monatsber. Akad. frontal region, but this I regard as not beyond Wiss. Berlin, p. 680. the limits of individual variation. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 599 [Chalinolobus morio (Gray)] I discussed this genus rather fully a number Scolophilus morio GRAY, in Grey, 1841, Jour- of years ago (Tate, 1942a, pp. 280-283), nals of two expeditions ... in ... Australia, vol. pointing out its anomalous situation in view 2, App., pp. 400, 405. of the fact that the skull and teeth of the MATERIAL: Bunya Mountains, 4. type species, emarginatus Dobson from India, Chalinolobus picatus (Gould) have never been fully studied. I also pointed out that the Australian species are divisible Scotophilus picatus GOULD, 1852, The mammals into two groups: the large-sized ruppellii, and of Australia, vol. 3, pl. 43 and text. MATERIAL: C.N.H.M.: Craigmore (Scott), 4, in the several smaller forms. On the basis of April, 1947. this, Troughton (1948, pp. 135-359) set up the two full genera Scoteanax and Scotore- EPTESICUS RAFINESQUE pens, with types, respectively, rupellii and Eptesicus RAFINESQUE, 1820, Animals of na- orion. This action, in view of the uncertainty ture, p. 2. relating to the status of emarginatus, was The Australian group of Eptesicus, the perhaps rather premature, and, pending ulti- pumilus group, has been treated by Trough- mate clarification of the real status of Scotei- ton as a distinct genus, Vespadellus. Struc- nus, I think it better to hold the two new turally, however, it is still an Eptesicus, and, genera to subgeneric rank. if it is correct to promote this one group to The closeness of the relationship of Scotei- generic rank, then the many other groups nus to the North American Nycticeius is (Tate, 1942a) should also be so promoted. remarkable, The two dentitions, including Such a subdivision of the genus would tend loss of i2 and shifting of the enlarged, stout to make one lose sight of the essential unity il to make contact with the canine, are of the bats of the genus Eptesicus. Several almost alike. One of the chief distinctions of races are known, which depend for their the several small Australian species from distinctness chiefly upon size differences. Nycticeius lies in the greater degree of reduc- tion in size of the anterior of the two lower Eptesicus pumilus caurnus Thomas premolars. Eptesicus pumilus caurinus THOMAS, 1914, Ann. Mag. Nat. Hist., ser. 8, vol. 13, p. 439. SUBGENUS SCOTOREPENS TROUGHTON MATERIAL: Iron Range, 3; Black Mountain, 20 Scotorepens TROUGHTON, 1948, Furred animals miles south of Cooktown, 2; Alderbury, 25 miles of Australia, ed. 3, reprinted, p. 357. out on Cooktown-Laura Railway, 1. Also in alco- hol, Pink Cave, Chillagoe, 3; Submarine Mine, In this group of quite small bats there are Irvine Bank, 3; Mt. Isa, near Cloncurry, 4. seven named forms. The general distribution Glover M. Allen, when describing Eptesicus of the subgenus Scotorepens is very wide and darlingtoni, identified specimens from Millaa is shown below by the type localities. We Millaa on the Atherton Tableland as cauri- took it also in south New Guinea at Maba- nus, even though that race is typical of duane and Wassi Kussa, and we also have Kimberley, north Western Australia. The it from Pentland (Neuhauser), west of Towns- forearm of caurinus is ±30 mm., and the ville, and from Cooktown. tooth row, c-m3, ± 3.9 mm. The type localities of the six very small and We have numerous specimens of this race the seventh somewhat larger species are from Pentland and Quamby, and I myself shown: collected some in the tunnels of Mt. Isa mine sanborni East Cape, New Guinea clustering in small groups in depressions in the greyii Port Essington, Northern Territory caprenus Roebuck Bay, north Western Australia rock roof. All three of these localities are on balstoni Laverton, Western Australia the Townsville-Cloncurry railroad. The dis- aquilo Between Bowen and Rockhampton tribution then would seem to be all across orion Sydney the northern portions of Australia. infliuatus Prairie, southwest of Hughenden SCOTENUS DOBSON Our series from southern New Guinea and Scoteinus DOBSON, 1875, Proc. Zool. Soc. Lon- from Cooktown (sanborni) appear to be don, p. 371. exactly alike. In both the upper premolar 600 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 TABLE 4 COMPARATIVE MEASUREMENTS IN Scoteinus (Scotorepens)

Skull: Zygo- Inter- Breadth Palatilar Width Crowns Fore- Total matic orbital of Mastoid Length Inside of arm Length Breadth Breadth case Spine) mm c-m sanborni, type and others 32-33 13.6 3.6 7.1 8.1 4.2 - 4.8 sanborni female, A.M.N.H. No. 105062, Mabaduane 31 14 10.1 3.5 7.5 8.4 4.5 2.6 5.3 sanborni male, A.M.N.H. No. 105334, Wassi Kussa 36 13.7 9.7 3.6 7.4 8.5 4.2 2.9 5.1-5.2 sanborni female, A.M.N.H. No. 154660, Cooktown 34.5 13.4 - 6.8 - 4.1 2.8 5.1 greyii, type 34 - - 2.7 5.2 greyii female, A.M.N.H. No. 107796, Pentland 33 13.5-14 9.8 3.6 7.2 8.2 4.2 2.7 5.0 caprenus, type and others 32.5-34 14.2 3.5 6.8 8.5 4.7 5.1 aquilo, type and others 32.5-34.5 13.8 10.4 3.7 7.4 8.6 4.8 - 5.1 orion, type and others 34.5-36.3 14.9-15 10.8-11.1 4.1-4.4 7.5-7.8 9.0-9.2 5.1-5.3 5.6-5.8 influatus, type 39 16.7 11.3 3.8 7.9 9.2 5.6 5.0 has a relatively low crown, measuring 7 mm. 4.6 to 5.0 mm. These two species of Scoteinus below the edge of the cingulum. Our Pentland not only show considerable size difference, specimens (greyii) have the same tooth but they look different. The Papuan and larger, 1.1 mm., a character constantly pres- Cooktown bats are reddish tawny. ent. Gould's plate of greyii makes its pelage Of the foregoing species I have studied, appear fairly strongly rufescent-not dusty measured, and photographed the skulls of tawny or grayish tawny as in our material three: grey i, balstoni, and influatus. I have from Pentland. But Gray's plate (1875, also photographed Troughton's form capre- pl. 20, fig. 2) is more like our material. nus. None of the specimens in the Archbold Scoteinus balstoni Thomas is "bicolor,"' series is large enough to be equal to influatus. the hairs tawny at their tips, smoky at their All zygomatic widths are less than 10 mm. bases. Scoteinus orion Troughton has evident- (11.3 mm. in influatus); all tooth row lengths, ly also bicolored fur, since the author de- c-m3, are less than 5.5 mm. (6.0 in influatus); scribed the tips as "washed with cinnamon in all cases m1.3 is less than 4.0 mm. (4.5 to clay color," and the "basal fur being almost in influatus). This leaves a choice between Prout's Brown." Troughton's aquilo, which sanborni, greyii, caprenus, and aquilo. he made a subspecies of orion, was dorsally In table 4 the measurements of the types I"apparently similar to the typical race above and of representatives taken from our series [orion], but somewhat lighter tipped and are compared. therefore more bicolored below." His capre- A considerable degree of variation in size nus was "apparently a duller drab [than appears in different strains. For example, in balstoni] and more olivaceous brown through- the six specimens taken by us at Tarara, ' It must be borne in mind that in "bicolor" Thomas W,assi Kussa, the length of c-mr varies from referred to the contrasting colors of the tips and bases of 5.0 to 5.4 mm. Neuhiauser's series of four the body hairs, not, as Troughton (1948) seems to im- specimens from Pentland (greyii) varies from ply, to contrasting dorsal and ventral colors. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 601 out ... fur of underparts paler throughout, The type specimen from East Cape, Papua, therefore less bicolored ..." (Gray, 1875, is in the Australian Museum, Sydney No. p. 30). A3176. Our series from Pentland and the single MATERIAL: Mabaduane, south Papua, 1; Wassi from Malbon near Cloncurry have truly Kussa, south Papua, 5; Cooktown, 1 (A.M.N.H. bicolored (in Thomas' sense) pelage. The hair No. 154775); Seagren's Farm, 10 miles west of tips as mentioned are light dusty or grayish Cooktown, 1 (A.M.N.H., No. 154660). C.N.H.M.: tawny, but the roots of the hairs of this Port Moresby, 2. specimen are quite definitely smoky. The forearms measure from 30 to 32 mm. This NYCTOPHILINAE species of bat I have previously referred to NYCTOPHILUS LEACH as greyit. Nyctophilus Leach, 1822, Trans. Linnaean Soc. The remainder of our material (from south London, vol. 13, p. 78. New Guinea and Cooktown) has the hairs I discussed this genus briefly in an earlier of the anterior half of the back completely paper (Tate, 1941e). A classification of the self-colored to the roots, a rather bright known species had been proposed earlier tawny brown, though there is a trace of (Thomas, 1915b), using the forms of the nose darker coloring at the roots at the hind por- leaf and baculum as principal criteria. tion of the back. Ventrally the hairs of the The dentition, with but one upper in- throat tend to be unicolored, but posteriorly cisor and one upper premolar, is in approxi- a degree of bicoloration (light tips and darker mately the same stage of reduction as in the roots) can be perceived. This applies to all Nycticeini of the Vespertilioninae. our south Papuan specimens. In the case of I am now inclined to the view that the our specimen from near Cooktown the dorsal large forms, major, timoriensis, and sherrini, surface is completely unicolored, the hairs should be thrown together as races of a single being clear pale russet to their bases. The species, distributed over Timor (if this can under surface, much matted, shows traces of be demonstrated) and Australia generally. bicolored hair. It is slightly smaller than the In them the nose leaf is little specialized Papuan examples. (Thomas' first group). The species gouldi, In character of the dorsal fur of the larger which has somewhat smaller teeth, is closely form influatus was not made clear. Thomas related. merely stated that the under parts were As regards the two species daedalus and lighter than the upper parts, which he de- bifax, I find very little difference, apart from scribed as "brown," not "uniformly brown," the difference in the bacula indicated by as Troughton has it. But influatus is con- Thomas. The zygomatic width is nearly 1 siderably larger than our material from far- mm. greater in daedalus than in bifax. ther north and consequently can scarcely be We now have a number of specimens of considered exactly the same. The forearms of bifax, and the zygomatic breadth, ± 10.5 the south Papuan and Cooktown animals mm., holds very consistently. vary from 31 to 35 mm. The forearm of Bacula of bifax and geoffroyi extracted influatus is about 39 mm. Of sanborni from and cleaned by Van Deusen confirm the eastern New Guihea Troughton stated that characteristics given for them by Thomas. its fur was bicolored, though less markedly so than that of greyii. Nyctophilus bifax Thomas The measurements of the various forms Nyctophilus bifax THOMAS, 1915, Ann. Mag. compared in table 4 suggest that the northern Nat. Hist., ser. 8, vol. 15, p. 496. reddish bats of Cape York and south New MATERIAL: Shipton's Flat, 30 miles south of Guinea are a single race, and I am disposed Cooktown, 3, and (Roberts), 3 in alcohol; upper Parrot Creek, about 2 miles east of Shipton's to refer them all to the Papuan race sanborni. Flat, 1; rain forest on summit of Mcllwraith Range, upper Nesbit River-Rocky Scrub, 1. [At a Scoteinus sanborni Troughton camp 10 kilometers southwest of Bernhard Camp, Scoteinus sanborni TROUGHTON, 1937, Austra- Idenburg River, Netherlands New Guinea, at lian Zool., vol. 8, p. 280. 1500 meters (W. B. Richardson), 1, which I am un- 602 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 able to distinguish from bifax of the Cape York schreibersii (with orianae), the local member Peninsula.] of the schreibersii group, and australis, the The forearms in the Cape York series (and small species. In the future the large tristis in Thomas' original 20 specimens) range from type may be reported there as well, since I 39.5 to 43 mm. The forearm of the Idenburg found magnater in caves in New Guinea, near River specimen is 43 mm. Port Moresby. The pelage is light brown, only the tips Typically the forearm length of australis having this color; the bases are fuscous. The is 38 mm., that of s. blepotis, 46 mm., and of under parts are light grayish brown, also with magnater, 48 to 51 mm. gray bases. The ears are moderately large, At a variety, living in the about 20 mm. from the crown of the head same colony with normal specimens of M. s. when freshly caught, and are therefore pro- blepotis, has the whole of the head, nape, and portionately somewhat smaller than the ears the sides and under parts of the neck bright of the little species geoffroyi. They are dis- rufescent, in contrast to the prevailing fus- tinctly larger ears than those of the Papuan cous of the back. It thus presents a mantled species microtis. appearance like a miniature Pteropus. The At Shipton's Flat these bats used to rufescent color also appears on the under side back and forth under the roof of an aban- from the base of the wing to the base of the doned sawmill at dusk. They spent the day- hind leg. Strongly marked individuals offer light hours concealed among the rafters. The a strikingly different appearance from nor- specimen at the Mcllwraith Range was mally colored specimens. But intermediates netted in a path in the forest at a height occur, in some cases with only a trace of the of about 4 feet above the ground. reddish coloring. I do not believe that this variation indicates even a subspecific value, Nyctophilus geoffroyi pallescens Thomas for there is no indication of any geographical Nyctophilus geoffroyi pallescens THOMAS, 1913, or genetic segregation. In addition, I shot Ann. Mag. Nat. Hist., ser. 8, vol. 11, p. 79. a single one similarly colored south of Cairns MATERIAL: Pentland, 1 9 (Neuhauser); Ra- at Walter Hill Range camp. venshoe, 1 c' (H. C. Raven). The forearm length in the Pentland female Miniopterus schreibersii blepotis is 36.8 mm., that in the male from Ravenshoe, (Temminck) 37.5 mm., that of the type specimen, also Vespertilio blepotis TEMMINCK, 1841, Mono- a male, was only 33 mm. The forearms in graphies de mammalogie, vol. 2, p. 212. the types of the other races are as follows: MATERIAL: Thursday Island, 14 and 51 in alco- pacificus, 37 mm.; unicolor (Tasmania), 40 hol; Possession, 4 and 12 in alcohol; Shipton's mm.; a near topotype of true geoffroyi from Flat (Roberts), 1; Cairns, hydroelectric tunnel, 2; Walter Hill Range, 1; Chillagoe Cave, 8 in alco- King George Sound (B.M. No. 7.1.1.338), hol; Mt. Spurgeon, 27. M.C.Z.: Mcllwraith Range 35 mm. (Darlington), 2. [Glen Ferneigh, northeast New MINIOPTERINAE South Wales (Raven), 6.] MINIOPTERUS BONAPARTE The length of the forearm throughout this Miniopterus BONAPARTE, 1837, Fauna Italica, series varies only from 46 to 48 mm. This fasc. 21. includes the peculiar red-naped variety from I discussed this genus briefly in 1941f. Thursday Island. It appeared to me then that there were three main groups of forms, differing from Miniopterus australis Tomes one another chiefly by size. These were Miniopteris [sic] australis TOMES, 1858, Proc. tristis Zool. Soc. London, p. 125. the large group, the medium-sized MATERIAL: Upper Peach River, Camp Oven schreibersii group, and the small australis Pocket, 1; Shipton's Flat, 1; Tommy Burns Mine, group. Besides the difference in absolute 4 miles south of Lappa, on the Atherton-Chillagoe size, there were relative differences in the road, 9 in alcohol. M.C.Z.: Mcllwraith Range form of the sagittal crest and in the teeth. (Darlington), 17. [Pilkington's Cave, Mt. Etna, In Australia one finds only two species: 10 miles south of Rockhampton, 5 skins and skulls 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 603 and 19 in alcohol.] Both the first two specimens attached to the midline of the head, they listed were shot at dusk. extend onto the forehead and prolong them- the intermaxil- This very tiny species can at once be selves to the region of the shortness of the forearm, laries . . " recognized by a mm. not seem to offer the This description seems to imply only only 38 It does and variety in coloration that was seen in the very slight degree of union of the ears, larger blepotis. there is certainly no such deep pocket formed by them as appears in Chaerephon. In fact, MOLOSSIDAE the degree of fusion seems to be similar to In the Australian area only two genera of that in Nyctinomus australis and slightly molossid bats need to be considered. These less than in C. plicatus colonicus of Australia. are Nyctinomus and Chaerephon. I discussed The Australian species of very small size, (Tate, 1941c) the genus Chaerephon in rela- on the other hand, have their ears rather tion to Nyctinomus and . For some widely separated and show almost no con- time it has been usual to place the large, necting band. gray-brown Australian colonicus Thomas in Troughton has applied the generic terms Chaerephon, while retaining the other species Austronomus to N. australis, and Micro- in Nyctinomus. I conclude that the generic nomus to N. norfolcensis and N. planiceps. It status of colonicus should be reexamined. does not seem to me that the distinction Dobson defined Chaerephon as having between these various species belongs truly "Ears connected in front by a deep band on the generic level, and I would prefer to produced upward, and posteriorly by a sec- see the new names used to indicate "groups" ond band enclosing a hollow naked space or subgenera. The basic pattern of the skull between, on the crown of the head; tragus and teeth is the same throughout. small, quadrate." The character of the audital On almost every count we must consider area in Chaerephon is thus obviously more the small-sized Australian species of Nycti- specialized than that of Nyctinomus. I have nomus as the less specialized, the large ones been able to find no such second band or as more specialized, organisms. Thus it hap- hollow naked space in any of the Australian pens that the type species, aegyptiacus, falls species. I therefore feel obliged to keep all into the more specialized group. As between of them in Nyctinomus. In Chaerephon colonicus and australis I would treat the lat- is very well represented. ter as more extreme, partly because of the very large chest glands present, but also on NYCTINOMUS GEOFFROY account of its peculiar color pattern. Nyctinomus GEOFFROY, 1813, Description de In the large-sized Australian species of this l'Egypte, ed. 1, vol. 2, p. 114. genus the ears are united across the fore- As I have pointed out (Tate, 1941d), head by a broad, fleshy band. This is the Nyctinomus and Chaerephon were distin- case both in australis and in colonicus. In the guished originally from each other by the little species, on the contrary, the ears are absence or presence of the second band con- more separated, and there is hardly a trace necting the ears and the space between the of a connecting band. Two small species, two bands. The type species of each is, norfolcensis and loriae, in our collection have respectively, Nyctinomus aegyptiacus Geof- ears of this type. In the large-sized species froy and Chaerephon johorensis Dobson. the lower canines are closely approximated to In some species, notably in the European one another, whereas in the small-sized Nyctinomus taeniotis, there are three lower species the lower canines remain relatively incisors on each side. More commonly this far apart, as much as, or more than, the number is reduced to two. The type species thickness of one canine at its base. In con- has but two pairs as have all the Australian sequence of these conditions the two pairs forms I have seen. of lower incisors of the large species are The ears of aegyptiacus (Geoffroy, 1828) very much reduced in size and are displaced were described as "united to each other by downward and forward, well below the cin- their internal margins, at the same time gula of the canines. In the smaller, less 604 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 specialized species this reduction in size and the height of P4. Back of skull broken. displacement of the lower incisors is less The forearm was 37 mm.; the length of pronounced; they stand little if any below head and body (dry), 55; tail, 38; hind foot, the canine cingula. 7. Interorbital width, 4.0, width of brain- The forward projection of the bullae, case, 7.1; width inside ml-', 2.7; c-mi, 5.9; tending to align each with one of the ptery- ml-S, 3.6. goids in an almost continuous bony plate, is This species is apparently common in remarkably advanced in development in the eastern Australia. Iredale and Troughton large species, but only incipient in the small have indicated its presence at Sydney. It ones. The palate is domed, and the tooth is a cave bat. The five from Cairns were hid- rows form slight arcs in the small species. ing by day between the rafters and roof of Several very small species of bats of the an occupied house. The Mossman specimens genus Nyctinomus have been named from the were found in a crack in a telephone pole. Australian region. They are: FOREARM Nyctinomus loriae Thomas LENGTH Nyctinomus loriae THOMAS, 1897, Ann. Mus. norfolcensis Gray, 1839. Gray gave Civ. Genova, ser. 2, vol. 18, p. 609. no locality; corrected to Sydney 37 mm. MATERIAL: Helenvale on the Annan River, 20 planiceps Peters, 1866. Southern and miles south of Cooktown, 3; Shipton's Flat (Rob- western Australia 35 erts), 9. petersi Leche, 1884. "," on "Closely allied to Nyctinomus norfolcensis type label 36 ... size rather smaller, judged by the fore- loriae Thomas, 1897. Kemp Welch River, Papua 31-32 arm .... Fur very short, close and velvety, astrolabiensis Mayer, 1889. Bongu, only 2 to 3 mm. in length on the back. Color Astrolabe Bay, New Guinea 36 above fulvous or chestnut brown, the ex- wilcoxi Krefft, 1871 treme bases of the hairs whitish; below the hairs are yellowish or whitish buff, sometimes The ear is more acute in loriae, more tipped with brown on the sides ...." rounded in norfolcensis. To this I have added, after seeing the type, SUBGENUS MICRONOMUS TROUGHTON an adult female, "ears tall, not meeting across Micronomus TROUGHTON, 1948, Furred animals the brow." Thomas did not describe the of Australia, ed. 3, reprinted, p. 364. skull. In it the frontal depression is almost absent. A slight sagittal crest developed Nyctinomys norfolcensis Gray at the back. Upper incisors on each side Nyctinomus norfolcensis GRAY, 1837, Ann. Nat. long, simple. The anterior upper premolar Hist., vol. 4, p. 7. minute, peg-like, contained in the tooth MATERIAL: Shipton's Flat (Roberts), 8; Moss- row, the posterior premolar compressed from man, 4; Cairns, 5. front to back, molariform. Lower incisors The following redescription was drawn two on each side; i1 forked, the inner branch up from the type specimen (B.M. No. 38.10. of the fork longer than the outer. Lower 11.56), an adult, marked "Norfolk Island." premolars two, P4 slightly compressed from A small species, dark brown above and front to back and one-fourth as high again slightly paler beneath. Ears triangular, as P2. round-pointed, not obviously united across Forearm, 31 to 32 mm. the front. Hind feet with combs of hairs on The appearance of this little Papuan spe- outer faces of outer digits. cies in the Cape York Peninsula furnishes Skull of type rather low and flat. Palate one more example of the invasion by the distinctly domed. Upper incisors one on Papuan fauna of the northern parts of Aus- each side, three-fourths of the height of the tralia. The bats were shot in flight at dusk canine; p4 one-half the height of the canine; above the nearly dry bed of the Annan River. p2 quite small, with cusp, one-third of the It will be noted that the type specimen is height of p4. Lower premolars two, P4 about very slightly larger in all its measurements two-thirds of the height of c, P2 one-half of than any of our three animals from the An- 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 605 TABLE 5 COMPARATIVE MEASUREMENTS OF FEMALES OF Nyctinomus loriae A.M.N.H. A.M.N.H. A.M.N.H. No. 154756 No. 154757 No. 154758 Type Length of forearm 30 31 30 32 Length of head and body 53.5 55.0 52.0 54 Length of tail 28.5 27.0 26.0 26 Length of hind foot 6.5 6.0 6.0 7.0 Length of ear from crown 9.0 8.5 9.0 Skull Condylobasal length 14 14 13.9 14 Zygomatic width - 8.8 8.8 10.2 Palato-sinual length 5.2 5.4 5.1 5.6 Length of bulla 2.75 2.9 2.8 3.0 c-m3 5.3 5.5 5.5 5.9 m138 3.5 3.7 3.7 3.9 nan River. The most significant difference being dark brown above and gray-brown be- appears in the zygomatic width, but I do neath, without the dark hair bases present not think sufficient difference occurs to war- in our Hainan plicatus. The ears in colonicus rant the naming of a separate race. appear thinner and to a great extent lack the [Nyctinomus colonicus Thomas] internal tufts of hair which characterize the ears of plicatus. Taking all this into considera- Nyctinomus plicatus colonicus THOMAS, 1906, tion I am Proc. Zool. Soc. vol. p. 537. disposed, even though the skulls London, 2, are almost indistinguishable one from an- Originally described from Alexandria in other, to treat colonicus as a full species. the eastern part of Northern Territory, this bat is represented in our collection by a series SUBGENUS AUSTRONOMUS TROUGHTON of seven taken by Gabrielle Neuhiauser at Austronomus TROUGHTON, 1948, Furred animals Malbon near Cloncurry and one at Pentland. of Australia, ed. 3, reprinted, p. 363. Jones has suggested that it is probably a Molossus australis GRAY, 1839, Mag. Zool. Bot., full species. It has been recorded from Burke- vol.2, p. 501. town and Darwin and is found generally in The type locality of this species is New the . South Wales, according to Iredale and Molossus plicatus was described by Fischer Troughton, 1934. (1829 [1830], Synopsis mammalium, p. 91). I have carefully read Buchanan Hamilton's [Nyctinomus australis atratus Thomas] account of " Vespertilio" plicatus and exam- Nyctinomus australis atratus THOMAS, 1924, ined his plate. There are specimens from Ann. Mag. Nat. Hist., ser. 9, vol. 14, p. 456. Hainan Island, China, in our collection which MATERIAL: Birdsville, southwest Queensland apparently represent plicatus. These have (L. Macmillan), 2, in June, 1940. the "sooty" wings of the description, and Macmillan found both animals clinging to slightly ashy tips to the black upper pelage. the base of a windmill and surmised that Beneath, the brownish black hairs have well- they had been crippled by the moving vanes developed buffy tips. This form, despite of the mill. considerable similarity of the skull, is very EMBALLONURIDAE different from the Australian colonicus. In that bat, while the wings are generally fus- TAPHOZOUS GEOFFROY cous, each finger traverses a light, translucent Taphozous GEOFFROY, 1813, Description de area, and the wing tip is generally quite 1'Egypte, vol. 2, p. 113. translucent. The body fur, too, is different, Dobson (1872 and 1875) published a synop- 606 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 sis and a monograph on Taphozous. Hollister Taphozous granti THOMAS, 1911, Ann. Mag. (1913) showed differences in the bullae of Nat. Hist., ser. 8, vol. 8, pp. 378-379. different groups. Thomas (1915a, 1922) and I have seen the type series of T. (S.) Troughton (1925) wrote on the classification. nudicluniatus at the Queensland Museum Troughton (1925) also revised the genus, and and the type of granti at the British Museum. Tate (1941c) offered a few notes on it, sug- It may well be that, as Troughton says, they gesting that Saccolaimus and Liponycteris are synonymous. be held as subgenera of Taphozous. The forearm length given by De Vis was Temminck (1841) based his term Sac- 75 mm., while the forearm of A.M.N.H. No. colaimus squarely on the throat pouch, pres- 66144 from Babinda Creek, not many miles ent in the male but not in the female in that north of Cardwell, the type locality, meas- species. The presence or absence of wing ures 74 mm. pouches did not enter into his conception in The skull of our specimen has the complete any way. The genus Saccolaimus Lesson bullae of Saccolaimus. In the tooth row, was formed tautonymically. Attention was c-m3 measures 11.2 mm.; m1-3, 7.6. first called to the character of the audital In the case of granti the length of the fore- bullae (complete in Saccolaimus, incomplete arm is 69 mm., of c-mi, 10.0, and of mr1-, 5.7. in Taphozous) by Hollister (1914) with re- It is therefore a somewhat smaller bat than gard to Taphonycteris. The radio-metacarpal nudicluniatus. pouch, usually absent in Saccolaimus, was pointed out in Taphozous by Dobson (1878). Taphozous (Saccolaimus) mnixtus (Troughton) The forms in Australia and New Guinea Saccolaimus mixtus TROUGHTON, 1925, Rec. referable to the subgenus Saccolaimus are Australian Mus., vol. 14, p. 322. flaviventris (-hargravei = affinis var. insignis), MATERIAL: Brown's Creek, , 3. nudicluniatus, granti, and mixtus. Troughton Also 1 I reported in 1941 from Dogwa, Oriomo (1925) has shown with a degree of success River, southern New Guinea. that granti and nudicluniatus are alike. Of Dorsal color dark grayish brown, with a those listed above mixtus has well-developed sprinkling of whitish or white-tipped hairs. wing pouches similar to the wing pouches in Under parts creamy white to the bases of the true Taphozous, so seems to constitute an hairs. Throat pouch strongly developed in annectant species. both males, only slightly less so in the fe- males. Wing pouches are definitely present SUBGENUS SACCOLAIMUS TEmmINCK in both sexes. The entrance to Saccolaimus TEMmINCK, 1841, Monographies de the pouch is mammalogie, vol. 277. beset with buffy white hairs. The forearm 2, p. length in the two males is 62.7 mm. and [Taphozous flaviventris Peters] 63.8, and in the female, 62.5. Troughton gave Taphozous (Saccolaimus) flaviventris PETERS, 61.5 mm. for the type from Port Moresby. 1867, Proc. Zool. Soc. London, for 1866, p. 430. Our specimen from Papua has 61.0 mm. MATERIAL: Pentland, 130 miles west of Towns- In the tooth rows, c-m3 are 9.5 and 9.8 in ville, north Queensland, 6; Malbon, near Clon- the two males, 10.0 in the female, 9.0 in curry, 1. Troughton's type, and 9.7 in the Dogwa These large handsome bats have the dor- male. sum velvety brownish black, the under parts The three bats from Brown's Creek were clear yellowish white. All the six specimens shot at dusk flying over open forest near a from Pentland are females without throat gully filled with fringe forest. pouches, a fact that seems to indicate a degree of segregation of the sexes. The Malbon SUBGENUS TAPHOZOUS GEOFFROY specimen is a male and shows a distinct Taphozous australis Gould throat pouch. Taphozous australis GOULD, 1854, The mam- Taphozous (Saccolaimus) nudicluniatus mals of Australia, vol. 3, pi. 32 and text. De Vis Taphosousfumosus DE Vis, 1905, Ann. Queens- Taphozous nudicluniatus DE Vis, 1905, Ann. land Mus., no. 6, p. 37. Queensland Mus., no. 6, pp. 39-40. The situation regarding the types of aus- 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 607 tralis is rather mixed. Dobson (1878) listed takenly indicating it as the type) and when four specimens, one from New Guinea, two we consider also that Thomas' specimen, from , Cape York, and one which he merely indicated as a subspecies from King George Sound, southern Western of australis, is very old and bears a registra- Australia, the last (evidently in mistake) tion number showing that it was already marked "type." In 1905 De Vis proposed catalogued in 1844 (and both Thomas and the name fumosus for a specimen from near Dobson were eminent in the study of bats), Cardwell. Thomas (1915a) had described his we may be justified in believing that both new subspecies georgianus from King George Dobson and Thomas alluded to the same Sound. In 1937, in London, I found the two specimen. If this animal is truly a specimen Albany Island specimens labeled "cotypes" of the northern species australis, would that but could discover no trace of any "type" fact prove that australis extends into the of australis from King George Sound. extreme south of Western Australia, or imply Gould's original description reads " . . . fore- an error in labeling? In any case we are then arm, 2.5 inches. . . habitat: the maritime left without a name for the large species caves in the sandstone cliffs of Albany Island, of north central Australia. Lacking any Cape York. In great numbers in three of the synonyms that can be revived, I now propose caves. Specimens obtained October, 1948." for this species the name: This at any rate seems to settle the type locality of australis as Albany Island. It is (Taphozous troughtoni, new species] still questionable whether or not australis TYPE: A.M.N.H. No. 162708, adult fe- georgianus Thomas differs in any substantial male, from Rifle Creek, Mt. Isa, northwest way from true australis. Queensland, collected October 13, 1948, by MATERIAL: Albany Island (NeuhWuser), 3, G. H. H. Tate. topotypes; Quamby, near Cloncurry (Neuhluser), GENERAL CHARACTERS: Similar to Tapho- 2. All five formerly misidentified as georgianus. zous australis but differs by its much larger Newcastle Bay, close to Albany Island, 1, virtual size (forearm 72 to 76 mm.) and by the topotype; Possession Island, west of tip of Cape less whitish bases of the hairs. York Peninsula, 2; Wenlock, Batavia River, 6; DESCRIPTION OF TYPE SKIN: Dorsal color Portland Roads, 3; Coen, 7; Chillagoe Caves, 3; near Clove Brown (Ridgway); under parts Cairns-Mossman Road, 2. M.C.Z.: Coen (Darling- slightly grayer. ton), 3. Skull with extremely long postorbital proc- The approximate measurements of the esses, and sharply edged lambdoidal and tooth row are c-m3, 9.8; ml-', 5.2 mm. De sagittal crests. At the back of the palate the Vis gave the length of the forearm offumosus reentrant is very narrowly wedge shaped as 63.5 mm. (more open in australis). The bullae are large [Taphozous georgianus Thomas] and show the internal opening characteristic of the subgenus Taphozous. The teeth are Taphozous australis georgianus THOMAS, 1915, Nat. Hist. much larger than those of australis. Journ. Bombay Soc., vol. 24, p. 62. MEASUREMENTS OF TYPE: (Skin measured The type specimen, a female (B.M. No. in field): head and body, 93 mm.; tail, 35; 44.2.27.59), which I examined in London, hind foot (s.u.), 15; ear from crown, 19; is much smaller than are most of the animals forearm, 76. Skull: condylobasal length, referred to by Troughton (1925) under "geor- 24.1; zygomatic width, 15.0; interorbital gianus." In fact, I find no distinctively differ- width, 6.7; intertemporal width, 4.4; width of ing measurements. This brings up the ques- braincase, 11.0; width between ml-, 4.6; tion whether or not the type of georgianus length of bulla, 5.7; palate, sinus to posterior reallydiffers from australis. That there actual- notch, 7.9; c-m3, 11.0; ml-3, 5.8; ml, 2.8 by ly exists a second, much larger, species is not 2.4 (maximum); m2, 2.4 by 2.6; m3, 1.2 by to be denied. I myself collected specimens 2.3. of it near Mt. Isa. But when we consider Besides the type, I took a paratype, a that Dobson (1878) listed a specimen of male. Both were caught in small caves in "australis" from King George Sound (mis- the along the sides of the lake formed 608 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 by the dammed-up waters of Rifle Creek.The RHINOLOPHCINAE forearm of the male measured 73 mm. RHINOLOPHUS LACAPtDE I suspect that most of the large-sized ma- terial written about by Troughton (1925) Rhinolophys LACE'PDE, 1799, Tableau des divi- under the heading georgianus can be referred sions, sous-divisions, ordres et genres des mam- to the present species. A number of animals miferes, p. 15. from Pentland previously marked "georgi- A brief review of this large genus (Tate anus" belong here also. and Archbold, 1939), together with a key An only slightly smaller male found at suggesting a classification, was largely based Olsen's Cave, Mt. Etna, Rockhampton, upon the work of Andersen. seems to be in general agreement with the Of the numerous groups indicated, only foregoing. The bases of the hairs, however, the simplex group, promoted to generic are almost as white as those of australis. rank as Rhinophyllotis Troughton, reaches The length of its forearm is 72 mm. The fore- Australia. The single species megaphyllus is arm in true australis from the north is ± 64 exceedingly common there. mm. It is a remarkable thing that with all our MEGADERMATIDAE large collections from New Guinea not a single example of Rhinolophus megaphyllus is re- This family has the single representative corded. Macroderma in Australia. [MACRODA MILLERI Rhinolophus megaphyllus megaphyllus Macroderma MILLER, 1906, Proc. Biol. Soc. Gray Washington, vol. 19, p. 84. Rhinolophus megaphyllus GRAY, 1834, Proc. Zool. Soc. London, p. 52. When I discussed the Megadermatidae MATERIAL: Portland Roads, 5; Gordon and (Tate, 1941b) I had no examples of Macro- other mines, Iron Range, 15, and 43 more in alco- derma of Australia before me. hol; Brown's Creek, 1; Tunnel Gully, Wenlock, 6 in alcohol; McIlwraith Range, Camp Oven [Macroderma gigas (Dobson)] Pocket, upper Peach River, 4; Mcllwraith Range, Megaderma gigas DOBSON, 1880, Proc. Zool. Soc. crest forest, 4; Coen (Neuhauser), 8; Laura, 1; London, p. 461. Shipton's Flat, 2, and 2 in alcohol; Mt. Finnegan, MATERIAL: Johannsen's Cave, Mt. Etna, Rock- Rossville Creek, 4, and 2 in alcohol; Mt. Carbine, hampton, 4, and 2 more in alcohol. 1; Cooktown-Laura Railroad, 3 in alcohol; Julat- ten, 2; Cairns (hydroelectric tunnel), 2; Walter These great bats were nervous. Their Hill Range, 1; Perseverance Mine, Julatten-Moss- eyes shone dull reddish under the hunting man Road, 2, and 5 in alcohol; Queen Mine, Irvine light. They were shot in the cave. The type Bank, 3 in alcohol; Valetta Mine, Irvine Bank, 3 locality is "Mount Margaret, Wilson's River, in alcohol; Townsville, 2; Olsen's Cave, 10 miles central Queensland." north of Rockhampton, 2, and 2 more in alcohol. As in other species of Rhinolophus, both RHINOLOPHIDAB red and gray forms of an individual race Very extensive revisions of this family have may occur living together in single caves or been published by Dobson (1878) and by mine tunnels. At Jack Gordon's Mine, Iron Andersen (1912). In 1939 and 1941 (Tate Range, the red and gray forms were found and Archbold, 1939; Tate, 1941b) I went mixed. At Rothwell Mine, 5 miles southwest over the group once more, suggesting that of Coen, there appeared to be only the red the Rhinolophidae be treated in three sub- bats named by Allen ignifer. families: Rhinolophinae, Hipposiderinae, and The type locality is caves along the Mur- Coelopsinae (Coelopinae, by error). The rumbidgee River, southern New South Wales. Rhinolophinae and Coelopsinae are mono- It is usually this species one hears fluttering typic; the subfamily Hipposiderinae con- outside one's mosquito net at night in Aus- tains a number of genera, the characters tralia. Often they hang up under bed springs of which were reviewed in the latter paper. in darkened rooms during the day (at Laura). 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 609 R.hinolophus megaphyius ignifer G. M. Allen Though H. bicolor has an immense range Rhinolophus megaphyUus ignifer, G. M. ALLEN, through the East Indian archipelago and is 1933, Jour. Mammal., vol. 9, p. 149. common in New Guinea, it would appear to MATERIAL: Rothwell Mine, Stewart River, 5 be restricted in Australia to the northern miles southwest of Coen, 1 skin and skull, and 38 portion of the Cape York Peninsula. in alcohol. Hipposideros galeritus cervinus (Gould) that Considering the well-known fact Rhinolophus cervinus GOULD, 1854, The mam- colonies of Rhinolophus and Hipposideros and mals of Australia, vol. 3, pl. 34 and text. other genera are frequently found to contain MATERIAL: Somerset (across the strait from the both red and gray and intermediately colored type locality, i.e., 1 mile), 16, and 24 more in alco- individuals, I feel that not much value can hol; Thursday Island, 2; Newcastle Bay, a few be placed upon the red phase designated miles south of Somerset, 1; Iron Range, 7, and 6 ignifer. It is true that in the Rothwell Mine more in alcohol; Rothwell Mine, 6 miles south- a preponderance of the specimens were red, west of Coen, 8 in alcohol. but in the mines at Iron Range and Irvine As in Rhinolophus, both red and gray Bank we also took red specimens together (and also intermediately colored) examples with the usual gray ones and a number of of H. cervinus are found. In the boarded-up intermediates. dining room of the old Jardine house at HIPPOSIDERINAE Somerset we found gray, rufous gray, and brilliantly colored orange-rufous specimens HIPPOSIDEROS GRAY flying or hanging together. The Thursday Hipposideros GRAY, 1831, Zoological miscel- Island and Newcastle Bay specimens are lany, p. 37. gray. Six of the seven Iron Range specimens This genus was reviewed fairly extensively are intermediately colored reddish gray; the several years ago (Tate, 1941a). Four of seventh is gray. Several from Rothwell Mine the principal groups, the bicolor, galeritus, are strongly reddish. diadema, and muscinus groups, occur in Australia. They are readily distinguished Hipposideros muscinus semoni Matschie from one another: Hipposideros semoni MATSCHIIE, in Semon. Ears without hooked size 1903, Zoologische Forschungensreisen in Austra- 1. broadly rounded, tip; lien und dem Malayischen Archipel, vol. 5, art. 6, very small .. . . bicoor more or less hooked; size p. 774. Ears pointed, tip MATERIAL: Iron Range, 1; Mcllwraith Range, larger ...... 2 2. Ears very long, pointed, attenuate; nose leaf Rocky Scrub, 4; Coen, 1; Cooktown, 1 in alcohol. with specialized, club-like structures This peculiar bat was taken in small nets ...... muscinus placed about 4 feet above the ground in fly- Ears not very long or very pointed; nose leaf ways in the tall rain forest on the Mcllwraith normal for the genus ...... 3 Range. The one from Coen was found by day 3. Size rather small, slightly larger than bicolor; hanging in an empty clothes closet in a house. forearm, ± 45 mm...... galeritus Size very large; forearm 70 to 80 mm. Eipposideros diadema reginae Troughton ...... diadema Hipposideros diadema reginae TROUGHTON, These groups are represented in Australia 1937, Australian Zool., vol. 8, p. 275. by H. bicolor albanensis, H. galeritus cervinus, MATERIAL: Cairns, 1; Jack Gordon's Mine, H. muscinus semoni, and H. diadema reginae. Iron Range, 13, and 3 more in alcohol; Coen (NeuhWuser), 1, skull only...... Hipposideros bicolor albanensis Gray The single individual from Cairns is gray Hipposideros albanensis GRAY, 1866, Proc. Zool. and in full agreement with Troughton's de- Soc. London, p. 220. scription of the type from the Bloomfield MATERIAL: Lockerbie, 3. River. All specimens from Iron Range are These little bats were caught in the eve- tinged with russet, and three are quite nings after they had flown into the small strongly rufescent. The Cairns specimen was canvas tent used by our native helpers. shot at dusk; the bats from the Iron Range 610 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 were taken hanging from the ceiling and walls Lockerbie and Brown's Creek were shot in of the mine tunnel. the evening while feeding on the flowers of bloodwood trees. In general it can be stated PTEROPODIDAE that the preferred roosting place of this Both the principal subfamilies of fruit species is among mangroves, whence it sallies bats, the and the Macroglos- forth at night in search of fruit and flowering sinae, are represented in Australia, the first trees. If such is the case, the specimens taken by Pteropus, Dobsonia, and Nyctimene, the at Brown's Creek must have flown nearly 30 second by Odontonycteris and Syconycteris. miles inland to feed. In 1937 I collected what All but Syconycteris were obtained by the was apparently the identical race at Bugi, Archbold party. on the southern coast of New Guinea. PTEROPODINAE Pteropus conspicillatus Gould Relatively unspecialized fruit bats, in Pleropus conspicillatus GOULD, 1849, Proc. which the lower incisors are not unusually Zool. Soc. London, p. 109. proddont; the symphyseal portion of the MATERIAL: Julatten, 5; Cairns, 1; Clohesy mandible is not provided with a lingual River, 2; Peach River, 5. M.C.Z.: Mcllwraith groove, and the tongue is not Range (Darlington), 8; Shipton's Flat (Roberts), unusually 20 in alcohol. C.N.H.M.: (All by Scott), Danbul- lengthened or protrusible. lan Road, 5; Wongabel, 2; Ravenshoe, 1. PTEROPUS BRISSON At Julatten, Cairns, and the Clohesy River Pleropus BRISSON, 1762, Regnum animale, ed. specimens were shot by day from "camps" 2, pp. 13, 153-155. containing perhaps several hundred speci- Representatives of four species of Pteropus mens. Those from Peach River were shot occur in Australia. Three of these, alecto, by "shining" their eyes at night as they fed conspicillatus, and scapulatus, are common on on the flowers of bloodwood trees and other the Cape York Peninsula. Pteropus polio- kinds of eucalyptus. cephalus is probably there in migration time, In Australia the preferred roosting places when it leaves New South Wales. are in patches of rain forest within the Cook- Pteropus alecto is a species widely distrib- town-Townville mass of rain forest. We did uted in the , at least as far west not find any camps in the Coen area, although as Sumba and Celebes. Its Australian repre- the fact that we collected feeding bats im- sentative is gouldi. Pteropus conspicilatus is plied that camps were fairly near. more closely confined to the New Guinea- Pteropus scapulatus Peters Australian land mass. It is represented in the Moluccas by chrysauchen and on Ceram by Pteropus scapulatus PETERS, 1862, Monatsber. is characteristi- Akad. Wiss. Berlin, p. 574. ocularis. Pteropus scapulatus MATERIAL: Wenlock, Batavia River, 8; Peach cally Australian, with relatives only in the River, 1; Coen, 1; , 5; above Annan the Solomon Islands (woodfordi and mahaga- River gorge, 7; Shipton's Flat, 1. C.N.H.M.: Rav- nus). Pteropus poliocephalus was placed by enshoe (Scott), 6. M.C.Z.: Coen (Darlington), 3; Andersen in a group with epularius from Mcllwraith Range (Darlington), 8. south New Guinea and macrotis from Aru. Ratcliff (1931) gives the impression that I reviewed the Pteropodidae briefly in this little species mixes into the camps of 1942b. other species such as gouldii by day. No Pteropus alecto gouldii Peters doubt it does, but the Archbold expedition Pteropus gouldii PETERS, 1867, Monatsber. was so fortunate as to discover two large Akad. Wiss. Berlin, p. 703. camps containing only scapulatus. Each camp MATERIAL: Lockerbie, 1; Portland Roads, 2; was confined to the only moderately tall, Brown's Creek, 3; M.C.Z.: McIlwraith Range and often quite low, trees bordering a river. (Darlington), 3; Coen (Darlington), 9. The one at the Hann River contained several The specimens from Portland Roads were hundred specimens, or perhaps many more, collected from a "camp" in the mangroves as I could follow only a short distance up and that line the shore of the bay. Those from down stream. The camp close above the An- 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 611 nan waterfalls near Helenvale must have DOBSONIA PALMER numbered several thousands of bats. The Dobsonia PALMER, 1898, Proc. Biol. Soc. Wash- small trees margining the river were literally ington, vol. 12, p. 114. loaded with them. The examples taken at TYPE: Cephalotis peroni E. Geoffroy. Wenlock, Peach River, Coen, and Shipton's Specialized fruit bats with the hinder part Flat were shot at night as they fed on eucalyp- of the back naked and the incisors reduced tus flowers. to a single pair of upper and lower teeth in [Much farther south, at Byfield near each jaw. Rockhampton, I collected two examples of this species which were feeding on the flowers Dobsonia moluccensis magna Thomas of a bloodwood tree.] Dobsonia magna THOMAS, 1905, Ann. Mag. Nat. Hist., ser. 7, vol. 16, p. 423. A NOTE ON Pteropus brunneus DOBSON MATERIAL: Iron Range, 3; Coen, 1; Peach River, 2. M.C.Z.: Coen (Darlington), 2. During a recent visit to London I reexam- I found this to be a cave bat in New ined the unique specimen representing this Guinea, and again at Coen it was discovered bat. I have already studied all other Aus- hiding by day in a mine tunnel. Apparently tralian Pteropus: gouldii, conspicillatus, and it does not hang up by day in the open as scapulatus, from numerous freshly collected Pleropus does. The specimens from Iron specimens and their later cleaned skulls, Range and Peach River were collected at and poliocephalus from ample study speci- night while feeding on flowering bloodwood mens. The type of brunneus is positively un- trees. like any one of the four. The head and mantle NYCTIMENE BORKHAUSEN are now colored dull light brown. The fore- Nyctimene BORKHAUSEN, 1797, Deutsch Fauna, arm measures 117 mm. The back of the skull, vol. 1, p. 86. including the audital bullae, has been sheared off. The rostrum is not "long," in Andersen's TYPE: Vespertilio cephalotes Pallas. sense. The palatal length is 34.8 mm.; the Distinguished by its tubular nostrils, zygomatic width, 34; the tooth row, c-mi, broadly fused premaxillae, and obsolescence 23.3. of the lower incisors. Though I am sure that P. brunneus is not Nyctimene robinsoni Thomas truly Australian, I have not yet succeeded in Nyctimene robinsoni THOMAS, 1904, Ann. Mag. pinning it with certainty to any described Nat. Hist., ser. 7, vol. 14, p. 196. form. It seems more than likely that the Nyctimene tryoni LONGMAN, 1921, Mem. Queens- animal was a waif, carried to Percy Island land Mus., vol. 7, p. 179. by storm. MATERIAL: Portland Roads, 1. TABLE 6 COMPARISON OF THE MEASUREMENTS OF Pleropus brunneus WITH THOSE OF Pteropus hypomelanus AND Pteropus cognatus brunneus Actual Photograph h. luteus cognatus Type of Type Palatal length 34.8 38 37.5 32 Zygomatic breadth 34 36 34 33.7 c-m3 23.3 25 25 21.3 m3 5.1X3.0 6X3.2 5.0X3.1 4.8X2.9 Nasal to front of orbit 21 22.5 21 17 Width of m'3 11 11.9 12 10.8 Lacrimal foramina to premaxillary recess 13± 14.2 13.7 10.1 Length of forearm 117 - 130 120 612 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 The capture of this specimen was peculiarly marische Uebersicht der neuesten zoologischen interesting. Previously we had seen the bril- Endeckungen in Heuholland und Afrika, p. 33. liantly shining golden eyes of several as they Iredale (1947) has discovered that the flew past our hunting lights at Lockerbie. dingo was in 1792 named Canis antarticus At Portland Roads Van Deusen and I were Kerr. This will do away with the well-known hunting together in mixed eucalyptus and name dingo and calls for fiat ruling by the open grassy country. Suddenly the bat International Commission. Meanwhile I con- fluttered into the illuminated area and ac- tinue to use Canis dingo. tually alighted for a moment on my chest. Two rather recent papers on the dingo are Before I could seize it, it flapped off down of special importance, one by Jones (1921), into the grass where Van Deusen and I the other by Tichota (1937). Jones, after pounced upon it. summarizing the literature, concluded that This bat is quite unlike the much larger the dingo was positively introduced by early Nyctimene celaeno which we took abundantly Australian man and suggested that it was in the upper Fly River in 1937. Instead it derived from the "northern wolves." It con- presents a rather close agreement with N. forms, according to him, in the character of cephalotes of Celebes (Peleng Island), but the cingulum of the first upper molar, with its last upper and lower molars are consider- the wolves and the domestic dogs. ably smaller than in that species. Possibly it Tichota attempted to trace the phlogeny of should be treated as a race of cephalotes. the dingo. He found a very close resemblance to a prehistoric domestic dog from Germany MACROGLOSSINAE and Austria, Canis familiaris matris optimae Specialized fruit bats of small to medium Jeitteles (1877, p. 19), and he published size, their lower incisors markedly pro&dont, photographs comparing the two skulls, to- the symphyseal part of the mandible with a gether with comparative tables. His con- lingual groove, the tongue strongly protrusi- clusions were based upon three dingo skulls. ble. Jeitteles compared his "Hund der Bronze- MACROGLOSSUS F. CUVIER periode," which he named Canis matris Macroglossus F. CUVIER, 1824, Des dents des optimae, with a number of wild forms and mammif6res, p. 248 (referring to p. 40). (p. 55) reached the conclusion that the Indian Kiodotus BLYTH, 1840, Cuvier's animal king- wolf, C. lupus pallipes, was the most probable dom, p. 69 (footnote). ancestor. Odontonycteris JENTINK, 1902, Notes Leyden Sykes (1831) described pallipes as "dull Mus., vol. 23, p. 140. rufescent whitish; back varied with blackish I discussed the status of Odontonycteris in and ferrugineus; feet completely pallid, 1942b (p. 345). ferruginous; tail longish, pendent." He con- Macroglossus lagochilus nanus Matschie tinued: "This is the wolf of Dukhun. Its Macroglossus kagochilus nanus MATSCHIE, 1899, head is elongated, and its muzzle acuminated: Die FledermSiuse des Berliner Museums fUr a groove exists between the nostrils. Eyes Naturkunde, vol. 1, Megachiroptera, pp. 96, 98. oblique; irides yellowish bright brown. Ears MATERIAL: Seagren's Farm, 3. narrow, ovate, erect; small for the length of Two of these little long-tongued flying the head. Tail pendent, thin but bushy, foxes were collected flying about the flowers extending for the length of the head. General of century plants, Agave, at about nine o'clock colour of the fur a dirty reddish white or whit- at night by Van Deusen. A third one was ish brown. Along the back and tail very many picked up off the ground beneath the Agave, of the hairs are tipped with black, mixed with dried, stiff, and partly mummified. It had others tipped ferruginous. The tail ends in apparently been dead for a day or two. a black tip. Inner sides of limbs, throat, The same form is not uncommon at Maba- breast and belly dirty white. Legs pale .... duane on the south coast of New Guinea. Fur from occiput to insertion of tail is two to three inches long, gradually shortening as it CANIDAE approaches the sides.... Canis dingo MEYER, 1793, Systematisch-sum- "Length from tip of nose to insertion of 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 613 tail 35 to 37 inches; tail 11 to 12 inches" That jackals can be domesticated and will (a somewhat young specimen). readily hybridize with dogs has been attested We have no example of Canis paUipes. by Jeitteles (1877, pp. 42-50) and others. No one can pretend to explain how pallipes, Jeitteles was of the opinion that jackals may after being domesticated by man into C. have been ancestral to some of the smaller matris optimae, was transported to Australia. races of dogs, in particular the prehistoric Jones (1921, p. 261) probably goes near the "Torfhund," Canis familiaris palustris, of mark with his imaginative description of the Switzerland and the Rhine Valley and certain arrival of Talgai man with his wife and his dogs of ancient Egypt. dog and his dog's wife. In such case it follows that jackal blood On continental Asia there are three main may be present in the dingoes of Australia. types of canids that might conceivably have Jones (1921, p. 259) showed that in 22 been ancestral to the dingo. They are the specimens of dingoes the basicondylar length Indian red dogs, Cuon; the jackals, Thos; was 177.3 mm. and the length of the upper and the wolves, Lupus. The first can be ab- carnassial 20.0 mm. In our skulls the basi- solutely ruled out on structural grounds. condylar length is: males, 178, 179, 184; fe- Cuon has one fewer lower molars than the males, 163, 176, 195; the length of the up- dingo, and its remaining posteriorly terminal per carnassial: males, 19.0 18.6, 18.0; fe- molars in both upper and lower jaws are males, 17.6, 20.3, 19.2. The length of the relatively much smaller. There are other dis- carnassial thus is usually more than 10 per tinctive characters. Jackals, according to cent of the basicondylar length. In an exam- Jones (pp. 257-258), also differ in the char- ple of Thos, the jackal of India, these meas- acter of the cingulum of the first upper molar, urements are 147 and 16.6 mm. Such large which is quite strongly developed, while in carnassials are interpreted by Jones as an wolves and dogs it is nearly obsolete. Jones attribute of wild canids. states of a series of dingoes that they "fall A NOTE ON NEW GUINEA DOGS in line with all the other races of dogs in being of the true northern wolf type." Yet, The Papuan wild dog collected by Mac- one of our dingoes (A.M.N.H. No. 65845), Gregor on Mt. Scratchley at 7000 feet and collected by Raven at Ravenshoe, shows the described by De Vis (1911, pl. 1) was black external cingulum of ml almost as fully de- and white, the white appearing on the nape veloped as that of any jackal, and others and throat and under parts as shown in the show it in more incipient form. illustration. The absence of dew claws was mentioned. The same is true both of the MATERIAL: Lockerbie, 1 (large o", skull only, with cingulum obsolete at middle of ml); Wenlock, Papuan dogs and of the dingoes in our col- 3 (one old 9 with skin, and two pick-up skulls, lections. De Vis suggested that the Papuan their cingula as above); Ravenshoe, 1 (ci, cingu- dog might well be feral. lum almost like that of a jackal); Mundubbera Longman (1928, p. 154) mentioned two District, 1 (cingulum moderately developed). I specimens and two skeletons of Papuan dogs saw two others alive, one near Atherton, the other from Mt. Scratchley, doubtless including near Townsville. De Vis' specimen. The second skin is Russet This intermediate condition of the cingu- (Ridgway), which is a much darker brown lum of m1 leaves room for the possibility of than the normal dingo color, and thus ap- an introduction of jackal blood into the dingo proaches some of the skins from the Iden- in the remote past. The dingo's reputed in- burg River described below. tractability might support this view. That Jones (1929) studied the two skulls re- they are not wholly intractable was demon- corded by Longman, finding in them a strated by Lumholtz (1890, pp. 178-179) who "relatively large upper carnassial tooth, the showed that "dingoes" were kept tame maximum length of which is nearly ten per around the camps of natives in the Herbert cent of the condylobasal length." He was River region of Queensland. Etheridge (1916, convinced that the Papuan dogs represented p. 43) confirmed this relationship between a primitive race. natives and dingoes. Mr. W. B. Richardson of the 1939 Arch- 614 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 bold expedition obtained five Papuan dogs MURIDAB at the Idenburg River, one at 75-Meter The species of rodents foundlin the Cape Camp, the others at 850-Meter Camp. All York Peninsula and adjoining territory are were brought into camp by natives. Presum- included in Tate (1951). These are: ably they were domestic, not wild, animals. The colors are tan (the shade seen in the Hydromyinae Hydromys chrysogaster reginae Australian dingoes) and white. In two of the Hydromys chrysogaster beccarii animals the tan predominates, only the face, throat, under parts, a nape line, and the Leggadina delicatula delicatula under side and tip of the tail being white. Zyzomys argurus In the second of these two the nape mark is Mesembriomys gouldii rattoides asymmetrical and spreads to the right around Melomys cervinipes capensis the neck to join the white of the throat. In Melomys cervinipes eboreus the remaining three skins white is predomi- Melomys lutillus australius nant, the tan being isolated in large head, Uromys caudimaculatus caudimaculatus shoulder, and rump blotches. Rattus assimilis coracius Rattus leucopus leucopus The length of the upper carnassial and Rattus leucopus mcilwraithi condylobasal lengths in our Idenburg skulls Rattus leucopus cooktownensis are shown below: Rattus lacus A.M.N.H. No. Carnassial Condylobasal Rattus lacus and Sex Length Length Rattus gestri apex 151841, a' 16.0 139 Rattus gestri conatus 151839, di 17.0 155 In addition to these, I think that Conilurus 151803, 9 14.3 131 can be expected eventually to appear in the 151842, 9 14.7 143 Cape York collections, since it occurs in both In every instance the carnassial length is Arnhemland and south New Guinea. I have more than 10 per cent of the condylobasal omitted Notomys, because the type locality of length. A trace of a cingulum on ml can be N. aquilo seems to be in doubt; the center of seen in No. 151839. In the others the cingu- distribution of the genus is south central lum is absent. Australia.

LITERATURE CITED ANDERSEN, K. lection of the British Museum. London, 1912. Catalogue of the Chiroptera in the col- pp. 378-390. lection of the British Museum. London, EIFFE, 0. E. vol. 1, Megachiroptera, pp. 1-854. 1909. Kreuzung von Polarhund und Dingo. BRAZENOR, C. W. Zool. Beobacht., Frankfort on the Main, 1932. A re-examination of Gymnobelideus lead- vol. 50, pp. 312-313. beateri McCoy. Austral. Zool., vol. 7, pp. ETHERIDGE, R. 106-109. 1916. The warrigal or "dingo," introduced or DE VIS, C. W. indigenous. Mem. Geol. Surv. New 1911. A wild dog from British New Guinea. South Wales, ethnol. ser., no. 2, pp. 43- Ann. Queensland Mus., vol. 10, pp. 19- 54. 20. GEOFFROY SAINT-HILAIRE, E. DOBSON, G. E. 1828. Description de l'Egypte.... Seconde 1872. Notes on the Asiatic species of the genus Edition. Paris, vol. 23, pp. 150-154. Taphozous. Proc. Asiatic Soc. Bengal, GRAY, J. E. pp. 151-154. 1875. In Richardson, John, and J. E. Gray, 1875. A monograph of the genus Taphozous, The zoology of the voyage of H.M.S. Geoff. Proc. Zool. Soc. London, pp. 546- Erebus and Terror. London. 548. HOLLISTER, N. 1878. Catalogue of the Chiroptera in the col- 1913. Two new bats of the genus Taphozous. 1952 TATE: MAMMALS OF CAPE YORK PENINSULA 615

Proc. Biol. Soc. Washington, vol. 26, pp. Canis dingo. Proc. Linnean Soc. New 157-158. South Wales, vol. 6, p. 624. 1914. A review of the Philippine land mam- MORRISON-SCOTT, T. C. S. mals in the United States National Mu- 1950. The identity of Captain Cook's kan- seum. Proc. U. S. Natl. Mus., vol. 46, garoo. Bull. Brit. Mus., vol. 1, no. 3, pp. pp. 299-341. 45-50. IREDALE, T. PRESCOTT, J. A. 1947. The scientific name of the dingo. Proc. 1931. The soils of Australia in relation to Roy. Zool. Soc. New South Wales, pp. climate. Bull. Comm. Australian Coun- 35-36. cil for Sci. and Indus. Res., no. 52, pp. IREDALE, T., AND E. LEG. TROUGHTON 1-82. 1934. A checklist of the mammals recorded PYCRAFT, W. P. from Australia. Mem. Australian Mus., 1940. The Australian dingo. Illustrated Lon- no. 6, pp. 1-122. don News, no. 5302, vol. 197, p. 702. JEITTELES, L. H. RATCLIFF, F. N. 1877. Die Stammevater unserer Hunderassen. 1931. The flying fox (Pteropus) in Austpalia. Vienna, pp. 1-68. Bull. Australian Council for Sci. and JONES, F. W. Indus. Res., no. 53, pp. 1-61. 1921. The status of the dingo. Trans. Roy. SYKES, W. H. Soc. , vol. 45, pp. 254- 1831. Catalogue of the Mammalia of Dukhan 263. (Deccan). Proc. Zool. Soc. London, pp. 1929. The cranial characters of the Papuan 99-105. dog. Jour. Mammal., vol. 10, pp. 329- TATE, G. H. H. 33. 1938. New or little-known marsupials: a new 1931. The re-examination of the skeletal char- species of Phascogalinae, with notes acters of Wynyardia bassiana, an extinct upon Acrobates pulcheUus Rothschild. Tasmanian marsupial. Papers and Proc. Novitates Zool., vol. 51, pp. 58-60. Roy. Soc. Tasmania, for 1930, pp. 96- 1941a. Results of the Archbold expeditions. No. 115. 35. A review of the genus Hipposideros, LANGKAVEL, B. with special reference to Indo-Austra- 1892. Ueber , Pariah, und Neusealan- lian species. Bull. Amer. Mus. Nat. dische Hunde. Zool. Garten, vol. 33, pp. Hist., vol. 75, pp. 353-393. 33-38. 1941b. Results of the Archbold expeditions. No. LENDENFELD, R. 36. Remarks on some Old World leaf- 1890. Canis dingo: a popular account. Zool. nosed bats. Amer. Mus. Novitates, no. Garten, vol. 30, pp. 43-49. 1140, pp. 1-11. LESEBLE, L. 1941c. Results of the Archbold expeditions. No. 1890. Canis dingo. Bull. Soc. Acclimat., vol. 37. Notes on the Oriental Taphosous and 37, pp. 681-684. allies. Ibid., no. 1141, pp. 1-5. LONGMAN, H. 1941d. Results of the Archbold expeditions. No. 1928. Notes on the dingo, the Indian wild dog, 38. Molossid bats of the Archbold expe- and a Papuan dog. Mem. Queensland ditions. Ibid., no. 1142, pp. 1-4. Mus., vol. 9, p. 151. 1941e. Results of the Archbold expeditions. No. LUMHOLTZ, CARL 39. Review of Myotis of . Bull. 1890. Among cannibals. London, 395 pp. Amer. Mus. Nat. Hist., vol. 78, pp. 537- McGup, J. N. 565. 1921. The dingo. South Australian Nat., vol. 1941f. Results of the Archbold expeditions. No. 2, pp. 59-62. 40. Notes on vespertilionid bats of the MCINTOSH, K. S. subfamilies Miniopterinae, Murininae, 1928. Dingoes. Their destruction and control. Kerivoulinae, and Nyctophilinae. Ibid., Agric. Gaz. Sydney., vol. 39, pp. 808- vol. 78, pp. 567-597. 812. 1942a. Results of the Archbold expeditions. No. MATSCHIE, P. 47. Review of the vespertilionine bats, 1915. Der Dingo-Hund des Macdonnell-Ge- with special attention to genera and spe- bieges. Sitz.-Ber. Gesellsch. Naturf. Fr., cies of the Archbold collections. Ibid., Berlin, pp. 101-107. vol. 80, pp. 221-297. MIKLOUHO-MACLAY, N. DE 1942b. Results of the Archbold expeditions. No. 1882. On the convolutions of the brain of 48. Pteropodidae (Chiroptera) of the 616 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 98 Archbold collections. Ibid., vol. 80, pp. New Guinea. Ibid., vol. 97, pp. 183-430. 331-347. TATE, G. H. H., AND R. ARCHBOLD 1945a. Results of the Archbold expeditions. No. 1939. Results of the Archbold expeditions. No. 52. The marsupial genus Phalanger. 24. Oriental Rhinolophus, with special Amer. Mus. Novitates, no. 1283, pp. reference to material from the Archbold 1-44. collections. Amer. Mus. Novitates, no. 1945b. Results of the Archbold expeditions. No. 1036, pp. 1-12. 54. The marsupial genus Pseudocheirus TEMMINCK, C. J. and its subgenera. Ibid., no. 1287, pp. 1841. Monographies de mammalogie. Leiden, 1-29. vol. 2, p. 285. 1945c. Results of the Archbold expeditions. No. THOMAS, 0. 55. Notes on the squirrel-like and mouse- 1887. On the milk dentition of the . Proc. like possums (Marsupialia). Ibid., no. Zool. Soc. London, pp. 338-339. 1305, pp. 1-12. 1915a. Notes on Taphozous and Saccolaimus. 1946. Geographical distribution of the bats in Jour. Bombay Nat. Hist. Soc., vol. 44, the Australasian Archipelago. Ibid., no. pp. 57-63. 1323, pp. 1-21. 1915b. Notes on the genus Nyctophilus. Ann. 1947a. An example of "prelacteal incisors" in Mag. Nat. Hist., ser. 8, vol. 15, pp. 493- advanced pouch young of Macropus. 499. Jour. Mammal., vol. 28, pp. 399-400. 1922. The generic classification of the Tapho- 1947b. Results of the Archbold expeditions. No. zous group. Ibid., ser. 9, vol. 9, pp. 266- 56. On the anatomy and classification of 267. the Dasyuridae (Marsupialia). Bull. TICHOTA, J. Amer. Mus. Nat. Hist., vol. 88, pp. 97- 1937. Das Verwandtschaftverhaltnis dos Aus- 156. tralischen Dingo zu den prahistorischen 1948a. Results of the Archbold expeditions. No. typen des Hanshundes. Zool. Anz., vol. 59. Studies on the anatomy and phylog- 120, pp. 177-190. eny of the Macropodidae (Marsupialia). TROUGHTON, E. LEG. Ibid., vol. 91, pp. 233-352. 1925. A revision of the genus Taphozous and 1948b. Results of the Archbold expeditions. No. Saccolaimus in Australia and New 60. Studies in the Peramelidae (Mar- Guinea. Rec. Australian Mus., vol. 14, supialia). Ibid., vol. 92, pp. 313-346. pp. 313-339. 1951. Results of the Archbold expeditions. 1948. Furred animals of Australia. Third edi- No. 65. The rodents of Australia and tion, reprinted. Sydney, 376 pp.