RESULTS of the ARCHBOLD EXPEDITIONS. No. 48 PTEROPODIDAE (CHIROPTERA) of the ARCHBOLD COLLECTIONS

RESULTS OF THE ARCHBOLD EXPEDITIONS. No. 48 PTEROPODIDAE (CHIROPTERA) OF THE ARCHBOLD COLLECTIONS

B. G. H. H. TATE

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOL. LXXX, ART. IX, pp. 331-347 New York !~~~~~~~~r14I-8ued December 81 1949 1") rK'(9>,,,~~j, '' 4 A "'V'42$IY %A_'.KA,.-V,4t'T0" ,~~. -1-_,;~~~~~~~k,~~~. ~~~,- -A N I~AIt,~1,- %-_IA. ","' 1 ",I I. . 1, ~~~~~~,_,-~~~- ,-,, . ,~~ ~ I~~\-,)`''i~41 I."I~~~~~~~A<' ;I. "'', ",~~.'i,ICA'X v~-1,II1 ,,_ ~ . .,'

- _1 ,~~~~-I1. ,I I 11 - ll~ _;~I lil/ t ~, , - 11 .1''N4' K ~<\ A K_ ,'rv , ,, 'A ,".,!,.-1, I . ,,-",, v, ;,.IAN/,''I-.Ir,,.~ ~ A,

~i /~,-, -4, ..- , '. , , '. ..-ttA-..-I'.(.~, AU .- ,~ , 4 .) ~ ~I , to,~,A\KA,A~ ,.. ,,

;,A K ,,'i NA4.N' S< AA-~_!./N I ,A'- A/A '- 4 1'

.- 1).,~ . ~ ~ ,, i .1,1~ I) 'A--~~~~, -,b 14 ~~~~~~N41 ,, , ~ ~ \ ,11N% -, VI"K/.>44A' N14 '~~". III ~K.~' ~ ~ ~ ~ /A j'.1I,.j.IQI..'.Itf, K ~ A > 8 "IA.,7I.t At Ib AA NI;' n'J NIA,~ .I),".'4 4 Yy '..0 -- A''~~-(U ~ :~,K. / / ~ ,- 'A~ V 4 Qx 4r,", ~ i, ', )7 A" 4NA 8,~~,~i, 'K-

", ".~~~~~~~~~~~~y "I' A \ '."' 'A2,' AA/ '.,' 'K,V ,~_11, ~ "4 ~~-;-,"~~~~,,,~~"(,N At/N.'...., :'K ~~,I,,1~~~~~~~~.K ,.~~~,I` . ,I~ ~ ~ ~-('~~r ~~,~, 7-1, , ,,~,~:.,_'K A"4,~ 11 ~ -~~,,,,,1A,K,~ ., .~ -A 'I N\AI''-N' .AI\.' 4 ' ,~ o ~- __ A's"'-,,., - ,-- -, f, 1~ N ' 7I ",A~1A A .' AV ~ AJAA1 K-., , ..1'K(-" ~ ½ ,- -4,U~ ~ 'N' A' ,;1 .KI.:, F A IK -~~~~1.,i,,,. -~~~~~~~~~~~~~4 "4 A I;.U' A' 'VA'A AA t'Nt'4 A I4 4 'A4A'A' ,I '", tt N'' "'';~ 1 ' II~-4 -I,\ ~~~~~~~-<4 ,'4 K'.1,' ' <'9' AIU N)..4 ' -\~I~~ 4;''' A .."~~~~~~~._,'~~~~,,,~~~.,- 7''I~4A<. Ai "4.~ 'IIA~A 4jA' ~

" ,'II -,~ A,,," .' ,. ,- ~. A '. / ..., .,, ,. At'l V~,, 'III''., 'IAI A~~.'.A."' 1, I ' I(A/U 4> -, , ,~ AA , '½t'N `, ,,, . ,'-4~7 1~. "421/ ~~,~~I A\KNIi " ,~~~~~''-,IV,''A A A 'N ../' -K'.I'A ''A:,'. ' A\~ ~ ~~ I ,;:, 'f~ 1~~I~ .II,~.'7I,A_ ,~ ~A ..'N',,N./VA 74,I'lit> AN~~~~~~~~IA.~-,1, .~ ,i 4,i" VKA,I/,1",,A ' 4'

'AI' .44 4-AA;">.', ,'~,,''KA / -J \ A 4 , 1'' "' ->44A 4.'.I K i"~~ ~~,r:,~~/" ,~, '--, ,~-. ,,, .,~~~.~ I,, ~~, ', #4 .; /. ~~~~ '' I~.,'Ktr--I,~K ,.~~~.I, I.,11 I~~~~~- ,.''K Kt 'IK , ~,, r,, , . 1 ,', , ,-"., t,A, -"-\,1.h,'I, tt ~; ~- A' N1~~, -,,A , N\ j K. . - ~ -".. I-''N4I t>' t(.*K't ',%/'I 'II1. tI>, Article IX.-RESULTS OF THE ARCHBOLD EXPEDITIONS. NO. 48 PTEROPODIDAE (CHIROPTERA) OF THE ARCHBOLD COLLECTIONS BY G. H. H. TATE

PAGE} INTRODUCTION ...... 331 NOTES ON INTERRELATIONSHIPS OF PTEROPODID GENERA.331 SYSTEMATIC ARRANGEMENT.333 Pteropodinae.334 Macroglossinae.343 REFERENCES ...... 347

INTRODUCTION The family Pteropodidae is represented Guinea not hitherto known to zoologists. in the collection of Mr. Richard Archbold It is to be hoped that when the existing by approximately 900 specimens, including world disturbance has ended Mr. Arch- ten of the twenty-four genera shown by bold may renew the field research which he Andersen (1912) to inhabit the oriental has already pushed so far toward its con- region. This partial representation is due clusion. to the fact that the material originates In spite of the wealth of fruit bat ma- from Sumatra, Java, Bali, Borneo, Celebes, terial, study has revealed only one new New Guinea and Australia only. The genus and species, Paranyctimene raptor collection is nevertheless a substantial one, (Tate, 1942). in many cases providing good series of For the sake of convenience the taxo- species hitherto known to science by one or nomic system set up by Andersen (1912) two specimens. Many were previously has been followed. Andersen's monograph unrepresented in the collections of represents the most up to date treatment The American Museum of Natural His- (in extensive form) of the Pteropodidae. tory. Before him, Miller (1907), Matschie Three of the expeditions (to New (1899), Dobson (1878), Peters (1867) and Guinea), which contributed to the large Gray (1866) had proposed various classi- collection of Indo-Australian mammals fications. Since the publication of Ander- now housed at this museum, operated under sen's work, only partial modifications of the personal leadership of Mr. Archbold. individual genera have appeared, Andersen The other undertakings were performed and Kloss (1915) on Cynopterus, Dietz by collectors employed by him. The use of (1916) on Rousettus and Dobsonia. Numer- airplanes permitted the Archbold Expedi- ous additional forms have, however, been tions to penetrate to areas in central New proposed.

NOTES ON INTERRELATIONSHIPS OF PTEROPODID GENERA The criteria for determin^ing natural of the skull, whereby the facial axis may affinities of the Megachiroptera are as form an angle with the basicranial axis; complex and contradictory as in the or again the two distinct types of pre- Microchiroptera. Many depend upon maxillae accompanying long or short nasals disappearing structures: the claw of d2 and the correspondingly modified anterior of the wing, certain of the teeth, the tail. portion of the palate. A third example is Others represent adaptive changes not seen in the modification of the symphysial necessarily leading to the obsolescence of region in the Macroglossinae, accompanied organs, for example, the degree of bending by adaptations of adjoining lower incisors, 331 332 Bulletin American Museum of Natural History [Vol. LXXX canines and the anterior premolar, probably the premaxillae, more or less proclivous, in relation to the extrusive function of the are unnarrowed above, include both macro- tongue. glossine bats with protrusible tongues and A combination of shortening of the modified symphyses and pteropine bats bones of the face with reduction in the size with tongues non-extensible and sym- and number of teeth has led to the short- physes not especially modified. The un- faced fruit bats, which include Penthetor, narrowed premaxillae in all cases tend to Nyctimene, Dobsonia and Cynonycteris. be thrust between the anterior part of the It is found that the genera of short-faced maxillae and the nasals, permitting pro- bats are usually those which show little or jection of the tips of the nasals over and no bending of the skull and in which inter- beyond the superior part of the pre- section of the facial and basicranial axes maxillae. Even in the aberrant Hypsi- forms an angle of minimum extent. The gnathus this condition can be observed in long-faced genera, on the contrary, have young skulls. the skulls more arched and the said angle The genera with scarcely extensile of intersection more appreciable. tongues include Rousettus, Pteropus, Epo- Intermediate states appear. In Epomo- mophorus, Hypsignathus, Styloctenium, phorus, the rostrum is elongate and the Boneia and Eidolon. The most primitive skull unarched; in Myonycteris and Epo- of these seems to be Rousettus, in which the mops the rostrum becomes shortened and skull is moderately arched; m2 and m3 the skull weakly arched. only slightly reduced, being more than To this combination of short face plus half as long (in the toothrow) as the molar unarched skull can be added the third immediately preceding each; upper in- character, premaxilla narrowed above. cisors not enlarged, aggregated at the tips Bats combining these three characteristics of proclivous premaxillae and remote from are Cynopterus, Micropterus, Dobsonia, canines; lower incisors evenly distributed Nyctimene, Penthetor, Cynonycteris, Megaer- around the symphysial arch between the ops and the otherwise much aberrant canines. Epomophorus is only slightly Scotonycteris and Casinycteris. Two short- more progressive. In it the skull is un- faced genera with incipiently arched arched, but m2 and m3 are absent, while skulls, Myonycteris and Epomops, have m2 is reduced in size. The upper incisors similarly formed premaxillae and may be are small and crowded and separated from included. No genus of this group of genera the upper canines by occlusion of the lower is provided with extrusible tongue or with canines; pl* is obsolete. modifi- In Eidolon, Pteropus (sensu lato), Sty- the accompanying symphysial loctenium and Boneia close contact is no cations. longer maintained between the dental Pronounced reduction in dentition is to shelf of the premaxilla and the anterior be noted in the Cynopterus group, notably portion of the palate. In Rousettus, Epomo- in the incisors, anterior premolars and phorus and Hypsignathus a smooth and posterior molars, though the remaining continuous surface is maintained from the teeth, canines, posterior premolars and anterior palate to the alveoli of the pre- anterior molars remain large and strongly maxillae. In the Pteropus bats and nearest functional. Loss of incisors has been allies that continuity is broken, and con- accompanied by progressive approximation tact between premaxilla and maxilla is of the canines. A series illustrating this made only above the palatal level, anterior phenomenon passes from Cynopterus to the canine. In Eidolon contact between through Penthetor and Dobsonia to Ni/cti- left and right premaxillae is interrupted. mene, in which last genus the lower in- In Pteropus (at least in certain groups) cisors are absent and the lower canines are the incisors are secondarily enlarged. In in contact. Boneia one pair of upper incisors becomes Those genera, in which the rostrum is un- * I use Andersen's terminology for the premolars, shortened, the skull is usually arched, and pl, pI, p., 1942] Tate, Results of the Archbold Expeditions. 48 333 obsolete. In Styloctenium m3 and one pair incisors show various grades of obsolescence of lower incisors are lost. beginning at the center. In Eonycteris il Pteropus may develop a postorbital is as large as i2 but placed below the al- process on the jugal, probably a secondary veolar line of i2-2; in Macroglossus ii is specialization called forth by the function- only one-half the size (diameter) of i2; in ing of the ligamentous tissue which con- Nesonycteris and Notopteris ii is obsolete, nects it and the long postorbital process of and i2 is reduced to a tiny cuspule standing the frontal. in front of the canine. The greatly specialized Hypsignathus, This progressive reduction of the lower generally regarded as a branch of Epo- incisors may be correlated with progressive mophorus, possesses, besides the raised increase in function of the tongue. It is inflated rostrum and greatly broadened usually accompanied by reduction of the interorbital region, a singularly shortened upper incisors as follows: In Eonycteris, braincase, a strong sagittal crest termi- upper incisors minute in size and widely nating abruptly some 6 mm. in front of the separated, il very slightly smaller than lambdoid crests, minute upper incisors i2; Macroglossus, il and i2 subequal, pro- widely separated from each other, a broad odont, little reduced (although the lower flat symphysis and lower incisors much incisors are more specialized, the upper reduced and widely separated. Its dental ones are less so than in Eonycteris); Neso- formula equals that of Epomophorus. nycteris and Notopteris, both with il Last come the macroglossine bats, those obsolete, i2 reduced, peg-like, and placed with greatly protrusible tongues. All have back on the outer margin of the pre- long-faced, well-arched skulls, with premaxilla. In Syconycteris, on the other maxillae narrowed above, and the coronoid hand, the upper incisors are unreduced, or process of the mandible unusually low and even enlarged. weak. In every genus the symphysial Notopteris, as can be noted from the portion of the mandible is significantly foregoing discussion, has unquestionably modified by means of a projection reaching the skull of a macroglossine, yet it retains well beyond the canines which contains a an elongate free tail, being the only mega- cup-shaped depression serving, possibly, chiropteran to do so. On the other hand for the attachment of the greatly developed the claw of d2 of the wing in Notopteris is genioglossus muscles whose contraction absent. causes the tongue to protrude. A signifi- cant accompaniment of this peculiarity is I have not seen specimens of Harpionyc- seen in the adjoining teeth, the lower ca- teris. The multicuspid molars and pecul- nines, incisors and anterior premolar (pi). iar upper incisors testify to its aberrant The lower canines and Pi-, are widely status. Anderson (1912, p. 803) regarded divergent; pi may be large and double- it as nearest to Dobsonia but followed rooted, as in the case of Notopteris. The Miller in giving it subfamily rank.

SYSTEMATIC ARRANGEMENT Dobson (1878), treating the Mega- beginning with Pteropus, Styloctenium, chiroptera, divided the single family Epomophorus, Rousettus and Boneia. Next "Pteropodidae" into two main groups, the came Harpionycteris, then Scotonycteris, Pteropi and the Macroglossi. The first followed by the short-faced bats Cyno- was characterized by "tongue moderate; pterus, Ptenochirus, Balionycteris, Gelasinus molars well developed"; the second by ( = Nyctimene), Leiponyx and Cephalotes f'tongue very long, muzzle narrow, elon- = Dobsonia) and finally the macro- gated; molars very narrow, scarcely raised glossine bats, Notopteris and Eonycteris to above the gum." Megaloglossus. Matschie (1899), after preparing an When Miller (1907) dealt with this large artificial key, dealt with twenty genera, assemblage of genera, he increased the 334 Bulletin American Museum of Natural History [Vol. LXXX number of main divisions of the Pteropod- Pteropodidae idae to four by raising Nyctimene and Pteropodinae Harpionycteris, formerly members of Dob- Rousettu8 section (including Pteropus and son's other to sub- Dob8onia) Pteropi (under names), Epomophorus section family rank. Cynonycteri8 section (including Nyctimene) Andersen (1912), attacking the problem, Macroglossinae combined in part the systems of Dobson Eonycteris section and of Miller. Shorn of generic sub- Notopteris section groups, his keys show adoption of the Harpionyeterinae ("near Dobsonia" Andersen, following arrangement: p. 803)

Pteropodinae ROUSETTUS SECTION circular crown outline of mi3. The large ROUSETTUS GRAY Javanese shortridgei agrees with leschenaulti in this particular, but stresemanni, of Rousettus GRAY, 1821, London Med. Reposi- I a tory, XV, p. 299. whose type skull have photograph, has Cynonycteris PETERS, 1852, Reise Mozamb., m3 more nearly like that of amplexicaudatus Zool., I, p. 25. in form, although in size the skull agrees GENOT1YPES.-(Rousettus), P(teropus) aegyptia- well with leschenaulti. It may be noted cus Geoffroy, from Egypt; (Cynonycteris), Ptero- incidentally that Klein compared strese- pus collari Illiger. manni not with leschanaulti and shortridgei The bats of this genus which inhabit the but with amplexicaudatus and with geo- oriental regions appear separable into two graphically remote Cingalese and African main groups: species. We may then regard stresemanni Larger forms with forearm 75-90 mm., as annectant between the le8chenaulti and leschenaulti, shortridgei, stresemanni. amplexicaudatus groups. Celebensis, of Smaller forms with forearm 68-80 mm., which I have only Andersen's description, amplexicaudatus with its races minor, appears to be a member of the amplexi- celebensis and brachyotis. caudatus group with narrower molars. In the former group the toothrow, canine to m2, measures 17 mm. i, in the Andersen recognized eleven species of latter 12 mm. R. seminudus from Ceylon, Rousettus, two of them constituting the which I have not seen, appears to occupy types of the African subgenera Stenonyc- an intermediate position. teris and Lissonydteris. The remaining In specimens from northwest Borneo, nine species were referred to Rousettus, the molar teeth are substantially wider than subgenus. Two of those were African, are the teeth of our amplexicaudatus from seven Asiatic. Java and Bali (R. a. minor). Widths of Since 1913 but one new Asiatic form ml in these forms may be expressed has been added, stresemanni from Japen 1.8: 1.3-5 mm. Island, New Guinea. Under the generic term Cynonycteris, We may then list the oriental Rousettus Peters (1873) attempted to show the as follows: differences between leschenaulti and amplexicaudatus. His conclusions were based FORM TYPE REGION upon "original examples" of those two R. leschenaulti group species, lent to him by Milne-Edwards of R. seminudus Kelaart Ceylon R. leschenaulti Desmar- Pondi.cherry, India the Paris Museum. In each type (or co- est type?) the forearm length equaled 77 mm. R. 1. shortridgei Thomas Java In Andersen's key (1912, p. 25) les- and Wroughton R. stresemanni Stein Japen Island, New chenaulti and seminudus are distinguished Guinea from the bats of the smaller amplexicauda- R. celebensis Andersen Mt. Masarang, tus group by the elliptical instead of sub- Celebes 19421 Tate, Results of the Archbold Expeditions. 48 335

FORM TYPE REGION viously he had seen only Dobson's type. R. amplexicaudatus group He concluded that the three forms am- R. amplexicaudatus T.imor were Geoffroy plexicaudatus, minor and brachyotis R. a. minor Dobson Ja ea fso" intimately interrelated as to be dis- R. a. brachyotis Dobson D u k e ofY o r k tinguishable only by average characters." Island, New Guinea Region Rousettus amplexicaudatus brachyotis Rousettus leschenaulti shortridgei (Dobson) Cynonycteri8 brachyotis DOBSON, 1877, Proc. Thomas and Wroughton Zool. Soc. London, p. 116. Rousettus shortridgei THOMAS AND WROUGH- TYPE LOCALITY.-Duke of York Island, be- TON, 1909, Proc. Zool. Soc. London, pp. 19, 374. tween New Britain and New Ireland. TYPE LOCALITY.-Kalipoetjang, Tji-Tandoei River, south Java. Our collections include but a single This Javanese form appears to differ specimen of this bat, from Cyclops Moun- from the Indian leschenaulti only by its tains, Netherlands New Guinea. greater size. It has the same type of m3 PTEROPUS BRISSON as leschenaulti, thus differing from strese- Pteropus BRISSON, 1762, Regn. Anim., 2nd ed., manni, the third member of the group. pp. 13, 153-155. R. 1. shortridgei is represented in the GENOTYPE.-"PteropUS pteropus Brisson (= Vespertilio vampyrus Linnaeus, part, = P. Archbold Collections by a series of twelve celaeno Hermann, 1804)" Palmer, 1904. Ander- specimens from Cheribon, Java, and seven- sen (1912) indicated niger G. Fischer, 1814 teen from Bali. (based upon Vesp. vampyrus niger Kerr, 1792) as type. Rousettus amplexicaudatus (Geoffroy) Matschie's (1899) treatment of Pteropus Pteropus amplexicaudatus E. GEOFFROY, 1810, Ann. Mus. Hist. Nat., Paris, XV, p. 96. may be compared advantageously with TYPE LOCALITY.-Timor. that of Andersen (1912). This form was held by Andersen to ex- Both arrangements are based upon keys tend from Timor to Cambodia and the Phil- which appear to be highly artificial. ippines. Matschie divided his Pteropus into six A series of eleven specimens from Peleng subgenera, two of which, Pteralopex and Island, east of Celebes. A second series of Acerodon, were later admitted by Andersen five, taken at Perboewa, northwest Borneo, and raised to full generic status. The re- no doubt corresponds closely to the two maining four subgenera with their typical males listed by Andersen (1912, p. 43) from species follow: Baram, Sarawak. SUBGENUS GENOTYPE Sericonycteris Not designated, but Palmer Rousettus amplexicaudatus minor (Matschie) (1904) gives Pteropus rub- (Dobson) ricollis Cynonycteris minor DoBsoN, 1873, J. Asiatic Eunycteris Gray Pteropus melanopogon Soc. Bengal, XLII, p. 203. Pteropus Brisson Pteropus celaeno TYPE REGION.-Java. Spectrum Gray Pteropus vulgaris The series in the Archbold Collection Sericonycteris includes temmincki and from Oboed, Selat and Noesa Penida in scapulatus, two groups widely separated in Bali are held to represent minor, of which Andersen's key, also molossinus of Ander- we have no topotypical specimens. sen's lombocensis group. The characters employed by Dobson to Eunycteris comprises melanopogon (but distinguish minor from amplexicaudatus not the other three species placed by (pl wedged between c and p3, the lack of Andersen with it) and the neohibernicus hairs on the backs of the tibiae) seem not group of Andersen, the latter exceptionally to be valid. specialized in regard to dentition and Andersen's detailed description (1912, dorsal pelage. pp. 811-812) was based upon a series Pteropus (subgenus) embraces the follow- collected in Java by Shortridge. Pre- ing groups of Andersen: vampyrus (typical 336 Bulletin American Museum of Natural History [Vol. LXXX includes celaeno), mariannus, melanotus, rayneri (part), samoensis (part), alecto, faced (both sections with strong posterior conspicillatus. basal ledges). His third section brings Spectrum includes Andersen's groups: together the last six groups "characterized rufus (with typical vulgaris), macrotis, by the practically complete obliteration of pselaphon, rayneri (part), samoensis (part), the posterior basal ledges of the cheek lombocensis, scapulatus, hypomelanus. teeth." The unnatural arrangement of both sys- Treatment of this last section as a unit tems is obvious when one works with speci- rather than as a miscellany is the basis of mens. Thus Andersen's two major divi- my objection. It consists of at least three sions depend upon the degree of develop- groups divergent in regard to the trends of ment of "posterior basal ledges" on the their dental specializations: (a) vampyrus, large premolars. Although his ideas on alecto, conspicillatus and macrotis; (b) this point are illustrated in Figs. 9, 10 (pp. scapulatus; (c) neohibernicus. 68-69), actually a complete series demon- In the first group (a), although the strating intermediate development of the "basal ledges" are reduced in comparison said "ledges" exists. Again, because of with such a group as the pselaphon group, the arrangement of his key, Andersen is the cusps of the posterior premolars and compelled to insert the neohibernicus group anterior molars remain fairly distinct. (giant species with rounded ears and modi- Indeed conspicillatus is so like geddiei of fied teeth and pelage) between the con- the mariannus group (remotely placed spicillatus and macrotis groups (both with under Andersen's system) that close elongate, pointed ears and in my opinion relationship may be inferred to exist. more nearly related to each other than In the second group (b), extreme narrow- either is to neohibernicus). The scapulatus ing of the molariform teeth in combination group, last in Andersen's system, is dis- with increase in length and spacing of the tinguished from all others by the wide canines is distinctive of scapulatus. spacing of its elongate, dagger-like canines, In the third group (c), the cusp pattern but this point is not brought out in the key, of the massive teeth has become so greatly although the extreme narrowing of its degenerate that even in quite young molars is. specimens the premolars and molars ap- It may be added that because the con- pear as rounded characterless stumps. cept of the geographical race was not The species of Pteropus represented in thoroughly prevalent when Andersen the Archbold Collections are reported worked, many of his "species" must be under the "groups" as arranged by Ander- reduced to representative forms. Thus it sen (1912). is reasonably obvious that Pteropus gouldi is the Australian representative of alecto of Pteropus hypomelanus Group Celebes (this opinion was advanced also Pteropus hypomelanus macassaricus by Dammerman, 1929). On the other hand Heude his treatment of the species vampyrus and Pteropus macassaricus HEUDE, 1896, Mem. hypomelanus shows that he recognized Hist. Nat. Emp. Chinois, III, p. 177 (footnote). geographical representation. A single, immature male specimen of Andersen (1912, pp. 76-79) discusses his this race from Peleng Island, east of seventeen groups, "arranged . .. according Celebes. to their probable natural affinities" and goes on rightly to question the validity of Matschie's subgenera Eunycteris, Spectrum Pteropus caniceps Group and Sericonycteris. Pteropus dobsonii Andersen He states that his seventeen groups Pteropus dobsonii ANDERSEN, 1908, Ann. Mag. "may be classed in three primary sections," Nat. Hist., (8) II, p. 370. the first including six groups, long-faced; TYPE LOCALITY.-"Celebes." the second comprising five groups, short- A series of seven specimens from Peleng 1942] Tate, Results of the Archbold Expeditions. 48 337 Island, east of Celebes, represents this (twenty from Peleng Island; nineteen from hitherto very rare species. Boemboelan, Menado, north Celebes). Pteropus argentatus Gray Pteropus alecto gouldi Peters Pteropus argentatus GRAY, 1844, Voy. "Sul- Pteropus gouldi PETERS, 1867, Monatsber. phur," I, p. 30. Akad. Wiss. Berlin, p. 703. TYPE LOCALITY.-? Amboina. TYPE LOCALITY.-Rockhampton, Queensland. A considerable series of twenty-four Four specimens from north Queensland; specimens from Boemboelan, Menado, four from Bugi, mouth of Wassi Kussa, north Celebes. This species was known to south coast of Papua. Those from New Andersen only by the type specimen. Guinea seem to constitute the first record from that island. They flew to the main- Pteropus vampyrus Group land at dusk from Boigu Island to feed. Pteropus vampyrus (Linnaeus) Pteropus conspicillatus Group Vespertilio vampyrus LINNAEUS, 1758, Syst. Pteropus conspicillatus Gould Nat., 10th ed., I, p. 31 (part). TYPE LOCALITY.- Range: Java." Pteropus conspicillatus GOULD, 1849, Proc. Zool. Soc. London, p. 109. A fine series of sixteen specimens from TYPE LoCALITY.-Fitzroy Island, northeast Java and sixteen from Bali. Queensland. Four specimens from the foot of the Pteropus vampyrus malaccensis Cape York Peninsula, Queensland. Andersen Pteropus vampyrus malaccensis ANDERSEN, Pteropus conspicillatus chrysauchen 1908, Ann. Mag. Nat. Hist., (8) II, p. 368. Peters TYPE LOCALITY.-Kuala Tembeling, Pahang, Pteropus chrysauchen PETERS, Monatsber. Malay Peninsula. Akad. Wiss. Berlin, p. 576 (footnote). Six specimens from Sumatra. TYPE LOCALITY.-Batchian Island. Three specimens (without skulls) from Pteropus vampyrus pluton Geelvinck Bay, northwest New Guinea. Temminck Pteropus pluton TEMMINCK, 1853, Esq. Zool., Pteropus neohibernicus Group p. 56. TYPE LOCALITY.-"Range: Bali; Lombok." Pteropus neohibernicus papuanus Six specimens from Bali. Peters and Doria It was Andersen's view (p. 345) that Pteropus melanopogan var. papuana PETERS non- AND DORIA, 1881, Ann. Mus. Civ. Genova, XVI, Javanese specimens of vampyrus were p. 690. melanistic, and that Balinese were melanis- TYPE LOCALITY.-Mansinaam, northwest tic. This distinction is maintained only New Guinea. imperfectly when a large series is studied. A very extensive series, ninety-eight specimens from various parts of south and Pteropus vampyrus natunae western New Guinea: the Fly River area Andersen and the Idenburg River area. Pteropus vampyrus natunae ANDERSEN, 1908, Ann. Mag. Nat. Hist., (8) II, p. 369. Pteropus macrotis Group TYPE LOCALITY.-Natuna Islands. Three specimens from Sarawak, Borneo. Pteropus epularius Ramsay Pteropus (Epomops ?) epularius RAMSAY, 1878, Proc. Linn. Soc. N. S. Wales, II, p. 8. Pteropus alecto Group TYPE LOCALITY.-Katau, south New Guinea Pteropus alecto Temminck (western division of Papua). Pteropus alecto TEMMINCK, 1837, Monogr. A good series of this small species: Mamm., II, p. 75. seven from Idenburg River area, sixteen TYPE LoCALITY.-Menado, north Celebes. from Fly River area and three from Central A large series of this bat from Celebes Division of Papua. 338 Bulletin American Museum of Natural History [Vol. LXXX Pteropus scapulatus Group Queensland, which Troughton (1941) refers to D. magna. Pteropus scapulatus Peters The species are rather poorly represented Pteropus scapulatus PETERS, 1862, Monatsber. Akad. Wiss. Berlin, p. 574. in the Archbold Collection, although those TYPE LOCALITY.-Cape York, northern present include numerous specimens. Queensland. Two specimens from Coen, Cape York, Dobsonia minor (Dobson) Queensland. Cephalotes minor DOBSON, 1878, Proc. Zool. This species is not only well distinguished Soc. London, p. 875. TYPE LOCALITY.-Amberbaki, northwest New by its narrow molars and long slender ca- Guinea. nines, but it differs from every other species in This small species was poorly repre- seen by me the pale buffy hairs along sented in collections until a few years ago. the underside of the upper arms and Andersen had but two specimens. We are the proximal parts of the forearms. now able to record seven specimens from STYLOCTENIUM MATSCHIE the Idenburg River and twenty-one from Styloctenium MATSCHIE, 1899, Fledermause the upper Fly River. Berlin. Mus. Naturk., I, Megachiroptera, p. 33. GENOTYPE.-Pteropus wallacei Gray. Dobsonia exoleta Andersen Resembling Pteropus (particularly P. Dobsonia exoleta ANDERSEN, 1909, Ann. Mag. capistratus and P. personatus, which have Nat. Hist., (8) IV, p. 531. TYPE LOCALITY.-Tomohon, Minahassa, rather similar facial color pattern) but Celebes. differing in the obsolescence of one pair of lower incisors. These may well be the This species is represented in the Arch- inner pair, as in Pterolopex, an allied genus, bold Collection by thirty-seven specimens the inner lower incisors are greatly reduced from the Togian Islands, Gorantalo Gulf, in size. east of Celebes. Styloctenium wallacei (Gray) Dobsonia moluccensis magna Thomas Pteropus wallacei GRAY, 1866, Proc. Zool. Soc. Dobsonia magna THOMAS, 1905, Ann. Mag. London, p. 65. Nat. Hist., (7) XVI, p. 423. TYPE LOCALITY.-Macassar, Celebes. TYPE LOCALITY.-Tamata, Mambar6 River, eastern New Guinea. A series of twenty-four specimens from In his main text Andersen (1912, p. 466) Malenge, Togian Islands, Gorontalo Gulf, treated magna as a full species, but in his Celebes. Andersen knew of only four "addenda" (tom. cit., p. 825) he concluded, specimens of this bat. on the basis of study of additional material, DOBSONIA PALMER that magna was merely a subspecies of Dobsonia PALMER, 1898, Proc. Biol. Soc. moluccensis Washington, XII, p. 114. In the Archbold Collection magna is well GENOTsYPE.-Cephalotes peroni E. Geoffroy. represented: Ifaer, northwest New Guinea, Specialized fruit bats with the hinder three; Cyclops Mountains, five; Hollan- part of the back naked and the incisors dia, eighteen; Idenburg River, six; reduced to a single pair of upper and lower southern Papua (Western Division), four- teeth in each jaw. teen; Kemp Welch River, Central Divi- Andersen recognized four main groups sion of Papua, fourteen. of Dobsonia of which the first, minor, de- At Daru, at the mouth of the Fly River, generate as to dentition, is the best defined. specimens of Dobsonia magna were fre- Besides the twelve forms recognized by quently shot at night. Their eyes shone Andersen, there have been described: reddish under the rays of a flashlight. remota Cabrera from Trobriand Islands; They were killed while hanging among the andersoni Thomas from Manus and Ruk. crowns of coconut trees, feeding upon the G. M. Allen (1935) records a member of young fruit and were also found hanging in the moluccensis group from Coen, northern mango trees in the daytime. Those speci- 1942] Tate, Results of the Archbold Expeditions. 48 339 mens from the Kemp Welch River were cases the outlines of the teeth, if the cusps taken in a limestone cavern at Javareri. are developed, suffice for determination of The bats hung from the roof of an outer the group to which a specimen belongs. cavern into which a little daylight filtered. Andersen admitted three full species in In inner lightless caves I found Miniop- his "Cynopterus section": sphinx, brachy- terus. The nursing young of Dobsonia otis and major. Sphinx was subdivided magna (taken in February) had the skin of into three geographical races; brachyotis body and wings almost wholly unpig- into eight. mented. One race only was allowed in the "Nia- dius section": lyoni (renaming of minor, Dobsonia peroni (Geoffroy) a homonym), a race of horsfieldi. Cephalotes peroni E. GEOFFROY, 1810, Ann. Since publication of Andersen's work, Mus. Hist. Nat., Paris, XV, p. 104. three other names have appeared: babi TYPE LOCALITY.-Timor. Lyon, apparently an insular race of This name has been applied by various brachyotis; persimilis Andersen from Sara- authors to several of the species and races wak, a close relative of C. horsfieldi lyoni; now recognized as distinct. As restricted and archipelagus Taylor from Polillo by Andersen it occurs on Timor, Alor, Island, Philippines, which appears from his Flores and Wetter Islands. The smaller description to be intermediate between form from Sumba, due south of Flores, was Cynopterus and Niadius. distinguished by Andersen as sumbana. Upon examination of a representative Our series of sixteen individuals from series of species of Cynopterus one is im- Bali extends the range to the west. The pressed by the dearth of clearly definable species is probably present also on Sum- characters. Animals of small and large bawa and Lombok, the islands lying be- size are recognizable, but considerable tween Flores and Bali. latitude in dimensions is permitted in al- CYNOPTERUS SECTION most every species, according to Andersen's tables, so that very high percentages.'of CYNOPTERUS CUVIER overlap in size appear. Within the sub- Cynopterus F. CUVIER, 1825, Dents des Mam- species, as defined by Andersen, actual size miferes, p. 80. plays an important part, but even so few Pachysoma GEOFFROY, 1828, Dict. Class. Hist. races can be so distinguished. Nat., XIV, p. 703 (homonym). Cynonycteris PETERS, 1852, Naturw. Reise In a series from one locality the color is Mossambique, Zool., I, p. 25. frequently variable. Generally adult males Niadius MILLER, 1906, Proc. Biol. Soc. are provided with a ruff of russet or yellow- Washington, XIX, p. 83. ish brown extending from beneath the ears, GENOTYPES.-Pteropus marginatus Geoffroy = Vespertilio sphinx Vahl (Cynopterus); Ptero- meeting across the throat and reaching pwU collaris Ihiger (Cynonycteris); princeps down either side of the chest to the origin Miller (Niadius). of the wings or even to the thighs. Young Andersen's revision of Cynopterus (1912) males and females usually lack this dis- resulted in his recognition of six full species, tinctive color. Occasionally an adult three of which he referred to the "Cynop- female is seen in which male coloration is terus section" and three to the "Niadius assumed, at least in part. section." The members of the latter Probably the best characters for sepa- group, proposed by Miller, were dis- rating different forms of Cynopterus are tinguished by the subrectangular outline found in the teeth. In addition to the of the cheek teeth and the strong develop- differences in the outlines of the crowns of ment in the lower jaw of surface cusps on molariform teeth and the presence or ab- P4 aiad mi. sence of the surface tubercle on P4 and mi, The surface cusp has been shown always slight but seemingly consistent differences to be present in Niadius, but it may be may be observed in p1- and in the in- present or absent in Cynopterus (Ander- cisors. The canines of males are larger sen's table, p. 590). In the majority of than those of females in all species. 340 Bulletin American Museum of Natural History [Vol. LXXX Cynopterus sphinx (Vahl) Cynopterus brachyotis (S. Muller) Vespertilio sphinx VAHL, 1797, Skr. Nat. Pachysoma brachyoti8 S. MtLLER, 1838, Tijd. Selsk., IV, Heft 1, p. 123. Nat. Gesch., V, Pt. 1, p. 146. TYPE LOCALITY.-Tranquebar, southeast TYPE LoCALITY.-River Dewei, Borneo. coast of peninsular India. The small size of these animals and the Andersen recognizes three races of shortness of the ears, combined with their sphinx, the typical race, C. s. gangeticus quite small teeth and lack of surface cusps and C. s. titthaecheilus. The typical race on P4 and ml, serve to separate the numer- is the smallest, and according to Andersen's ous races of brachyotis from sphinx and table (1912, p. 600) the measurements of major. The ear of major is, however, also the forearm (66-73.5 mm.) scarcely trans- short (see Andersen's table, 1912, p. 629). gress the range of measurements of gangeti- The species is wide, ranging from cus (73-78) and titthaecheilus (74.5-83). Assam and Siam down through Sumatra The range of true sphinx is shown to in- and Borneo to the Philippines and Peleng clude all peninsular India (Ceylon to Island. Bombay) and eastward into northern The typical form (fide Andersen) is re- Siam. The toothrow is given as 10.8-12.2 stricted to the Malay region, Borneo, (in gangeticus, 11.8-12.3), thus showing less Philippines, Celebes, Sumatra. West of distinction. brachyotis, but partly overlapping in the The American Museum has several Malay Peninsula, occurs the race angu- specimens from Calcutta which are refer- latus, present from Sumatra to Assam. It able to C. s. gangeticus, but none from is shown by Andersen as slightly larger. southern India. A specimen from Tenas- The remaining races represent insular serim, A.M.N.H. 54825, 9, has the forms: scherzeri (Nicobars), brachysoma toothrow 10.3 mm., so supports Andersen's (Andamans), minutus (Nias), babi (Sima- view that true sphinx reappears east of the lur), pagensis (Mentawi), ceylonicus (Cey- Ganges. In a large series from Hainan lon), javanicus (Java) and insularum Island, south China, the relatively small (Kangean). size of sphinx also can be noted (c-ml, A single specimen, A.M.N.H. 24147, e, 10.5). labeled "India" probably represents north- The character whereby Andersen sepa- ern angulatus (see beyond). The toothrow, rates the third race, titthaecheilus, from the c-ml, measures 8.1 mm. It is very much two first mentioned, i.e., shorter rostrum, smaller than specimens of C. s. gangeticus seems to be valid. In addition the Java- from Calcutta and lacks any trace of sur- nese bats are larger and have broader face cusp. teeth. Besides the foregoing, true brachyotis is represented in the Archbold Collection from Borneo by six specimens from Cynopterus sphinx titthaecheilus Badang; fifteen from Parit, Tjempaga, (Temminck) south Borneo; sixteen from Ngabang, Pteropus titthaecheilus TEMMINCK, 1825, Landak, northwest Borneo; and one from Monogr. Mamm., I, p. 198. Peleben, northeast Borneo. TYPE LoCALIrY.-Java. Seven specimens from Peleng Island, The Archbold Collection includes a series east of Celebes, are distinguishable from of three males and nine females from brachyotis only by their slightly more Sauver, Bali, referable to C. s. titthaecheilus; saturate color. They approach in this also a single female from Cheribon, Java. respect pagensis of the Mentawi Islands. This last is identical subspecifically with a specimen from Buitenzorg which has been Cynopterus brachyotis angulatus in the American Museum for many years. Miller In all material examined the surface cusps Cynopterus angulatus MILLER, 1898, Proc. of P4 and ml are either much reduced or Acad. Nat. Sci. Phila., p. 316. absent. TYPE LOCALITY.-Trong, lower Siam. 1942] Tate, Results of the Archbold Expeditions. 48 341 To angulatus are referred large series of A.M.N.H. 24148, from Nicobar, formerly specimens in the Archbold Collections identified as C. b. scherzeri, which has from Kalianda and Macarah Doewa, perfect, unworn molars is unmistakably a south Sumatra, also four individuals from member of the Niadius section and should Boekit Sanggoel, Benkoelen, Sumatra. probably be referred to C. h. lyoni. Cynopterus brachyotis pagensis PENTHETOR ANDERSEN Miller Penthetor ANDERSEN, 1913, Cat. Chiropt. Cynopterus pagensis MILLER, 1906, Proc. Biol. Brit. Mus., 2nd ed., I, Megachiroptera, p. 665. Soc. Washington, XIX, p. 62. GENOTPYE.-Cynopterus (Ptenochirus) lucasi TYPE LoCALITY.-North Pagi Island, Men- Dobson. tawi Islands. This specialized monotypic genus of the Our series of nineteen individuals from Cynopterus group is readily recognized by North Pagi has a generally darker color the combination of one single pair of lower tone than true angulatus. Pagensis was incisors and the extreme thinness of the synonymized by Andersen with angulatus otherwise not greatly reduced tail. from lower Siam, but the darker color Andersen (p. 645) wrote that the "true seems consistently present. characters and affinities [of Ptenochirus jagori] have been largely obscured and Cynopterus brachyotis javanicus misunderstood by confusion with those of" Andersen lucasi, which last he separated as Penthetor Cynopterus brachyotis javanicus ANDERSEN, lucasi. In Andersen's diagram (p. lxi), 1910, Ann. Mag. Nat. Hist., (8) VI, p. 624. Penthetor and Thoopterus are shown rising TYPE LOCALITY.-Buitenzorg, Java. from the Cynopterus stem by a branch A single individual, A.M.N.H. 101896, quite distinct from that leading to Cynop- 9, from Cheribon, Java, and two from terus and Ptenochirus. Penthetor and Tjerimai may be considered to represent Thoopterus were set off by their lack of javanicus. They appear to me no larger postorbital foramina. Obsolescence of i2 than true brachyotis. The forearms meas- was regarded as liable to occur in relatively ure + 61 mm. The same may be said of unrelated genera of the Cynopterus group. twenty-five specimens taken from Koeta and from Noesa Penida, Bali. Penthetor lucasi (Dobson) This brings to a conclusion the "Cynop- Cynopterus (Ptenochirus) lucasi DOBSON, 1880, terus section." Of the "Niadius section" Ann. Mag. Nat. Hist., (5) VI, p. 163. the Archbold Collections contain less TYPE REGION.- Sarawak."' material. The Archbold Collection contains eighteen specimens of this bat from Perboewa Cynopterus horsfieldi Gray (Landak), northwest Borneo. Cynopterus horsfieldi GRAY, 1843, List Mamm. Brit. Mus., p. 38. NYCTIMENE BORKHAUSEN TYPE LOCALITY.-Java. Nyctimene BORKEHAUSEN, 1797, Deutsche From Tjerimai, Java, fourteen speci- Fauna, I, p. 86. mens, also two from Peleben, northeast GENOTYPE.-Vespertilio cephalotes Pallas. Borneo. Although Andersen placed Nyctimene in his "Cynopterus section" of the Pteropinae Cynopterus horsfieldi persimilis it must not be forgotten that it is in reality Andersen a highly specialized genus, so strongly Cynopterus persimilis ANDERSEN, 1912, Ann. Mag. Nat. Hist., (8) X, p. 640. divergent indeed that Miller had earlier TYPE LoCALITY.-Sarawak, Borneo. (1907) segregated it as a subfamily dis- Three females, larger and duller colored tinguished by the high degree of obsoles- than true horsfieldi, from Perboewa, north- 1 Andersen (1912) states that the type is in the British Museum Collection (B.M.80-8-13-1). It was west Borneo, must be nearly topotypical of one of four sent to Dobson by F. A. Lucas, from Wards' Natural Science Establishment, Rochester, N. Y. persimilis. War,,s write that they no longer have any of those It may be added that a young specimen, speclmens. 342 Bulletin American Museum of Natural History [Vol. LXXX cence of the incisors, the tubular structure the dorsal stripe extends from occiput to of the nostrils (analogous to that of the tail and attains a width of 10 mm. (similar Murininae), the broad fusion of the pre- to that of aello); in major the stripe ex- maxillae, etc. tends from the scapular region to the tail The discovery of Paranyctimene with its but reaches a width of only 5 mm. In high-cusped molariform teeth (possibly a celaeno the forearm is slightly longer and secondary modification connected with the teeth are slightly smaller. The tooth- change of function) but generally Nycti- rows of celaeno converge slightly toward mene-like facies adds force to Miller's view the front of the palate; those of major that Nyctimene is strongly divergent from form an arc, such that p4-4 are as close to Cynopterus. one another as are p3-3. The four "groups" distinguished by N. aeo was founded upon a single Andersen are not entirely satisfactory, be- specimen from Milne Bay. It was sepa- cause forms described later than the twelve rated by Dobson from major by its broadt recognized by Andersen have tended to dorsal stripe and by the elongate trans- minimize distinguishing characters. Later verse diameter of its orbits. To these char- described forms are celaeno Thomas, acters Andersen (1912) added "premaxillae certans Andersen, draconilla Thomas, tryoni more proclivous than in other species." Longman. Troughton (1941) considers He treated it as the "allo group," distinct tryoni a synonym of robinsoni Thomas. from the papuanus, cyclotis and cephalotes Draconilla may be a Paranyctimene; the groups. The species celaeno was unknown teeth of the type were greatly worn. to Dobson and Andersen. Certans was stated to be closely related to Two specimens in the Archbold Collec- cyclotis. Celaeno is near aello. tion from Cyclops Mountains, northern Netherlands New Guinea, must be con- Nyctimene papuanus Andersen sidered typical of celaeno. In addition we Nyctimene papuanus ANDERSEN, 1910, Ann. have a fine series of forty-one skins and Mag. Nat. Hist., (8) VI, p. 621. TYPE LOCALITY.-Milne Bay, eastern tip of skulls taken in the head waters region of New Guinea. the Fly River which appear inseparable The range of this small species (forearm from celaeno. 54.5-59 mm.) includes the whole of New It seems probable that the bats from Guinea, and Andersen records one speci- Mysol Island listed by Andersen (1912, men from Cape York, northern Queens- p. 716) as referable to allo were in reality land. specimens of celaeno. He showed only the! A substantial series represents the species orbital diameter of the type specimen of in the Archbold Collections: forty-four aelo (11 mm., tom. cit., p. 720). The same specimens from the upper Fly River, measurements on robinsoni were 9.3-9.5. Papua, and one individual from the Iden- I have fairly good photographs of the burg River, Netherlands New Guinea. skull of the type of aelo, from which it The only observable difference is that the appears that the proclivous nature of the. Idenburg River specimen has wider molars, canines stressed by Andersen is somewhat. but even so their width comes within the exaggerated in his figure (64), but the upper range given by Andersen in his table position of the postorbital processes, (p. 721). farther back than usual, seems to be correct. Nyctimene celaeno Thomas The forms scitulus, geminus, robinsoni,. Nyctimene celaeno THOMAS, 1922, Ann. Mag. tryoni and lullulae appear, from Nat. Hist., (9) IX, p. 262. pictures, TYPE LocALITY.-Geelvinck Bay, Nether- and measurements of their skulls, to belong lands New Guinea. near the present group. Nyctimene celaeno and major, from Duke Of N. scitulus this museum possesses one of York Island, seem to be closely allied specimen, from Malapa Island (Guadal- forms but differ as follows: in N. celaeno canar), in which the characteristic V- 1942] Tate, Results of the Archbold Expeditions. 48 343 shaped posterior margin of the palate is lotes cannot be said to stand so close to any clearly to be seen. other group. The specimens from Peleng A specimen of N. geminus from Good- average very slightly smaller than is indi- enough Island (D'Entrecasteaux group) is cated by Andersen's measurements. The also represented. The toothrow is a little difference in color between the sexes is less longer, and the back of the palate is obvious. rounded, but otherwise the two, both males, appear remarkably similar, geminus PARANYCTIMENE TATE being somewhat larger. Forearms: gemi- Paranyctimene TATE, 1942, Amer. Mus. nus, 74 mm.; scitulus, 68. The foramen Novitates, No. 1204, p. 1. magnum of scitulus is unusually small. GENOTYPE.-Paranyctimene raptor. Distinguished from Nyetimene by its Nyctimene cephalotes (Pallas) high molar and premolar cusps, very long Vespertilio cephalotes PALLAS, 1767, Spicill. canines and lack of dorsal stripe. Zool., fasc., 10, p. 10. TYPE REGION.-"Moluccas." Paranyctimene raptor Tate A fine series of this species, type of the Paranyctimene raptor TATE, 1942, Amer. Mus. genus, is present in the Archbold Collec- Novitates, No. 1204, p. 1. tions, comprising twenty specimens from TYPE LOCALITY.-OroviUe Camp, Fly River, Peleng Island, east of Celebes. Papua. Andersen placed cephalotes in the same This small bat, as small as N. draconilla group as geminus, major, et al. It differs (forearm, 47 mm.), appears externally a in appreciable ways. It is much smaller. Nyctimene in which no dorsal stripe is Its teeth and skull are far more delicately developed. The nostrils are tubular as in built. The postorbital processes are Nyctimene, the wings similarly spotted with slender and weak. Nevertheless, cepha- amber color.

Macroglossinae EONYCTERIS DOBSON C. rosenbergi was based upon the type only Eonycteris DoBsoN, 1873, Proc. Asiatic Soc. (and no additional specimen seems to have Bengal, p. 148. been taken) there may be reason for in- GENO'TYPE.-Macroglossus spelaea Dobson. cluding it with Eonycteris. Miller, after This generalized macroglossine genus is studying a photograph of the type, be- remarkable because of the external re- lieved it "well characterized by its heavier semblance of some of its species to certain dentition and by the absence of the small Rousettus. Eonycteris can be distinguished Mi3." Andersen studied the type at Leyden by the absence of a claw on the index finger; and concluded that it was virtually equal to by the presence of a pair of kidney-shaped spelaea with the exception of the absent mi3. glands on either side of the anus, stated, The species of Eonycteris, excluding however, to be absent in E. robusta; by rosenbergi, are the widely spaced upper incisors, peculiar spout-shaped symphysis (a macroglossine FORM TYPE LocALITr character); the compressed form of the spelaea Dobson Moulmein, Burma molar series and the reduced size of m3. spelaea glandifera Montalban Caves, Luzon, The status of Jentink Lawrence Philippines Callinycteris (type major Andersen Mount Dulit, Borneo rosenbergi, from Celebes) seems still to be robusta Miller Montalban Caves, Luzon, doubtful. By Andersen (1912) it was Philippines treated as a synonym of Eonycteris, but other authors have recognized it as dis- The key given by Andersen (1912, p. tinct because of the absence of mi3. Be- 734) to the two species spelaea and major is cause the posterior molars of many genera rather unsatisfactory. The principal diff- of fruit bats are obsolescent, and because erence resides in the larger size of major, 344 Bulletin American Museum of Natural History [Vol. LXXX but like rosenbergi, major was based upon a specimen (A.M.N.H. 103319, 9 ) from single specimen. north Pagi in the Mentawi Islands south of *Robusta, described next by Miller, was Sumatra. considerably larger than spelaea, but its author was less certain of its distinctness Eonycteris major Andersen from major, although it had smaller teeth. Eonycteris major ANDERSEN, 1910, Ann. Mag. Taylor's (1934) species longicauda was, Nat. Hist., (8) VI, p. 625. according to Miss Lawrence a Eonycteriz robusta MILLER, 1913, Proc. Biol. (1939), Soc. Washington, XXVI, pp. 73-74. synonym of robusta Miller, but a second Eonycteris longicauda TAYLOR, 1934, Philip- form in Montalban Caves, whence ro- pine Land Mammals, pp. 131-134. busta ( = longicauda) came, was recognized TYPE LOCALITIES.-(Major), Mt. Dulit, by her and named glandifera. Borneo; (robusta and longicauda), Montalban When discussing her subspecies E. s. Caves, Luzon, Philippines. glandifera, Miss Lawrence referred (1939, The Archbold Collection includes only p. 39) to "tawny," "russet" and "ochra- A.M.N.H. 103319, 9, from Mentawi ceous tawny" colors of the pelage of the Islands. throat in males. The same condition pre- These large Eonycteris are distinguished vails in the adult males of our large series from the smaller species, E. spelaea, by a from Bali, females and young males show- number of characters: absence or ob- ing no such differentiation. On the con- solescence of the anal glands; less reduced trary males of true spelaea and those of our tail; markedly pointed ears; rather dense, series from southern Sumatra (which deep brown pelage (in typical spelaea appear entirely referable to spelaea) show mouse gray, in spelaea glandifera rufous no indication of such throat coloring. brown); by relatively massive incisors, It is suggested, following reexamination canines and anterior premolar; by the of material and of the opinions given by strongly flattened dorsal surface of the various authors, that the following may ex- combined nasals; by the squarish outline press the true situation in Eonycteris: (instead of rounded) of the suture between that but two true species exist, spelaea and premaxilla and maxilla; also by small major, which differ strongly by characters differences in the foramina of the base of to be pointed out later; that spelaea in- the skull. cludes true spelaea reaching from Tonkin In the teeth one of the outstanding (Osgood) south through Siam, Burma features is the fact that pl almost fills the (type region), Malay Peninsula (Chasen), space between the canine and p3, whereas Sumatra, east and central Java, and in true spelaea it occupies from one-fourth possibly Borneo (Moulton, 1911); that to one-third of the space. In the lower jaw spelaea includes also an eastern subspecies corresponding enlargement of pi may be characterized by rufescent throat pelage noted, and the incisors are so much in- in the males, i.e., glandifera, present in creased in size that i2 almost touches the Philippines, Bali, perhaps eastern Java, lower canine. and probably in Borneo (Moulton's series The greater depth of mandible beneath and those referred to spelaea by Hose and M3, mentioned by Miller for robusta, male Everett may be glandifera); and that (4.2 mm., instead of 3.2), measures in our rosenbergi from north Celebes, of whose female Mentawi specimen 3.8 mm. In our skull I have photographs, is also in all spelaea it is usually less than 3 mm. likelihood an anomalous specimen of Neither Andersen nor Taylor gave this glandifera (in such case glandifera would dimension for his material. Andersen's become a synonym of rosenbergi). In single specimen was a female; Taylor had regard to major, it is further suggested that ten males. Miss Lawrence, who pointed we are dealing with a rare but widely out that in longicauda Taylor had re- distributed species, namely, major from described robusta, did not give the depth Borneo, represented in the Philippines by of ramus in her several specimens of robusta ( = longicauda) and by a large robusta. 1942] Tate, Results of the Archbold Expeditions. 48 345

To prove or disprove the identity of these Kiodotusl BLYTH, 1840, Cuvier's Animal King- names conclusively it will probably dom, p. 69 (footnote). three Odontonytceris JENTINK, 1902, Notes Leyden be necessary to get the material together Mus., XXIII, p. 140. and make direct comparisons. GENOTYPES.-MacroglOSSUS, Pteropus minimus; Odontonycteris, 0. meyeri = ? Macro- Eonycteris spelaea (Dobson) glossus lagochilus. Odontonyeteris with three upper pre- Macroglossus spelaeus DoBsON, 1871, Proc. molars was separated by Jentink from Asiatic Soc. Bengal, XL, pp. 105, 261, PI. x, figs. 3, 4. Mdcroglossus which has only two upper TYPE LOCALITY.-Farm Caves, Moulmein, premolars, but Andersen (p. 765) regarded Burma. the extra tooth as a supernumerary, "an This is the classical species of Eonycteris anomaly not infrequently observed in and of the two here recognized is the more both species [minimus and lagochilus] of specialized, since the anal glands (formerly Macroglossus." believed of generic importance) are de- Macroglossus, compared with Syco- veloped in both sexes, and the incisors and nycteris, is specialized through the greater front premolars are more reduced. True degree of reduction of the teeth. In re- E. spelaea is represented in our collections taining a small external tail (about 3 mm. by a single specimen from Pegu, Burma in length) it is less specialized than Syco- (near the type area), by a series of fifteen nycteris. from Kalianda in south Sumatra and by a Andersen recognized two species, mini- unique specimen from Cheribon, Java. mus with two subspecies and lagochilus with four. Eonycteris spelaea glandifera Lawrence Macroglossus minimus (Geoffroy) Eonycteris spelaea glandifera LAWRENCE, Pteropus minimus E. GEOFFROY, 1810, Ann. 1939, Bull. Mus. Comp. Zool., LXXXVI, pp. Mus. Hist. Nat., Paris, XV, p. 97. 38-40. TYPE LOCALITY.-Java. TYPE LOCALITY.-Montalban Caves, Rizal Province, Luzon, Philippines. According to Andersen (1912, pp. 756- 757) true minimus has the rostrum rela- In the American Museum collections are tively short (orbit to nares, 7.8-8.8 mm.), two specimens of this form in alcohol from while the race sobrinus, which occurs from "Philippines" which formerly were re- Sumatra northward (but also in Java), has corded as "robusta." However, they are the rostrum considerably longer (orbit to smaller than robusta and possess well nares, 9.5-10.5). defined anal glands. The Archbold Collections contain fifteen An extensive series of twenty-two speci- specimens from Cheribon, Java, and ten mens taken in Bali, from Soka and Oboed, from Bali, all of which appear to be typical are tentatively referred to glandifera. minimus with short rostrum. Adult males possess the rufescent throat pelage referred to by Lawrence, and I have Macroglossus minimus sobrinus been unable to find any points of differ- Andersen entiation. Whether glandifera is equal to as has been Macroglossus minimus sobrinus ANDERSEN, the Celebes bernsteini, suggested, 1911, Ann. Mag. Nat. Hist., (8) VI, p. 642. or whether our Balinese bats are in reality TYPF LocALITY.-Gunong Igari, Perak, Ma- bernsteini instead of glandifera can be lay Peninsula. determined after more topotypical ma- The distribution given by Andersen for terial representing bernsteini has been this larger race includes Malay Peninsula, collected. Sumatra, Nias and various parts of Java. 1 Kiodotus Blyth, 1840, Rhynchocyon Gistel, 1848, MACROGLOSSuS F. CUVIER afld Carponycteris Lydekker, 1891, all with type Pteropus minimus Geoffroy, were successively pro- Macroglossus F. CUVIER, 1824, Dents des posed to re,place Macroglossus, tWen believed to be a Mammif6res, p. 248. homonym of Macroglossum Scopoli, 1777. 346 Bulletin American Museum of Natural History [Vol. LXXX He expected that it would occur also in Syconydteris; particularly do i2 and pi Burma, Siam and Darjeling. remain unreduced and approximated to The Archbold Collection includes a the lower canine. There is no trace of series of eleven specimens from the Men- external tail. tawi Islands (North Pagi), some of which Matschie proposed four "species" which even transgress the dimensions given for Andersen has reduced to two, crassa and sobrinus by Andersen (in some specimens australis, by making Matschie's papuana the length, orbit to nares, reaches nearly and finschi races of crassa. Andersen pro- 12 mm.). posed a third full species, naias. Macroglossus lagochilus Matschie Syconycteris crassa papuana Macroglossus lagochilus MATSCHIE, 1899, (Matschie) Fledermiiuse Berlin. Mus. Naturk., I, Mega- Macroglossus (Syconycteris) papuanus MAT- chiroptera, pp. 96-97. SCHIE, 1899, Fledermause Berlin. Mus. Naturk., TYPE LOCALITY.-Buru. I, Megachiroptera, p. 99. M. lagochilus was distinguished from Described originally as a full species, minimus by the presence of a deep vertical papuana was reduced to a subspecies of median groove on the upper lip and the crassa by Andersen (1912, p. 777). fact that the nostrils were turned half for- True S. crassa has type locality at ward instead of outward (difficult to see in Fergusson Island, D'Entrecasteaux Is- dried skins). lands; papuana occupies, according to The typical race of the four admitted by Andersen, the whole of New Guinea (see Andersen is represented in the Archbold australis). Collections by a fine series of twenty-four The Archbold Collections include three specimens from Peleng Island, east of from Balim River, central Netherlands Celebes. New Guinea and one from Weyland Moun- Macroglossus lagochilus nanus tains. A fifth specimen is from Mafulu, Matschie Central Division, Papua. Macroglossus nanus MATSCHIE, 1899, Fleder- mause Berlin. Mus. Naturk., I, Megachiroptera, Syconycteris australis (Peters) pp. 96, 98. Macroglossus australis PETERS, 1867, Monats- TYPE LOCALITY.-Lamellana, New Pomerania ber. Akad. Wiss. Berlin, p. 13. (New Britain). TYPE LOCALITY.-Rockhampton, Queensland. This race is distinguished from true Skulls of this species are readily dis- lagochilus, according to Andersen, by its tinguished from those of crassa and its narrower premolars, from M. 1. microtus by races by the much narrower molars and its longer ears, and from M. 1. pygmaeus by premolars. At the same time these bats its longer rostrum. are separable from Macroglossus by their The range includes the whole of New enlarged incisors and the total absence of Guinea, Bismarck Archipelago, Kei and the tail. Aru Islands, yet Andersen had only a few The Archbold Collection contains ten specimens, from Mysol, New Ireland and specimens from the Fly River, one from Aru. Mt. Tafa and one from Kagi. The two The Archbold Collections include five last named localities are near together on from the Fly River and one from Weyland the south side of the Central Range, Cen- Mountains. tral Division, Papua. In addition there SYCONYCTERIS MATSCHIE are six skulls (without skins) from northern Syconycteris MATSCHIE, 1899, Fledermiiuse Queensland which are to be considered Berlin. Mus. Naturk., I, Megachiroptera, pp. almost topotypical. 94, 98 (subgenus of Macroglossus). It is to be noted that in the lowlands of GENoTYPi'E.-Macroglossus australis Peters. the Western Division, Papua, I took the As pointed out under Macroglossus, the above mentioned ten specimens of australis incisors remain large and in contact in and not a single example of crassa. In 1942] Tate, Results of the Archbold Expeditions. 48 347 the Central Division behind Port Moresby, S. australis was the only species of bats on the contrary, both species of Syco- which was readily taken in bird nets at nydteris were captured. night.

REFERENCES ALLEN, G. M. Species. Proc. Zool. Soc. London, pp. 1935. A Fruit Bat, Dobsonia, New to Aus- 62-67. tralia. Australian Zool., VIII, p. 151. MATSCHIE, P. ANDERSEN, K. 1899. Die Fledermause des Berliner Mus. f. 1912. Cat. Chiropt. Col. Brit. Mus., 2nd ed., Naturk. I, Megachiroptera. MILLER, G. S., JR. ANDERSEN, K., and KLOSS, C. BODEN 1907. Families and Genera of Bats. Bull. 1915. Remarks on Some Races of Cynopterus. U. S. Nat. Mus., LVII. Jour. Fed. Malay States Mus., V, pp. PALMER, T. S. 220-222. 1904. Index Genera Mammalium. North L. Amer. Fauna, No. 23. BARBOUR, T., LAWRENCE, B., and PETERS, J. PETERS, W. 1939. Collections from the Philippine Is- 1867. Uber die Flederhunde Pteropi, und lands. Bull. Mus. Comp. Zool., Insbesondere fiber die Arten der LXXXVI, pp. 38-40. Gattung Pteropus s. s. Monatsber. DAMMERMAN, K. W. Akad. Wiss. Berlin, pp. 485-487. 1929. On the Mammals of Sumba. Treubia, 1873. Uber einige zu der Gattung Cynonyc- X, p. 303. teris gehorige Arten der Flederhunde DIETZ, P. A. und ulber Megaderma cor. 1916. Aanteekeningen over Megachiroptera. TATE, G. H. H. Leiden Zool. Med., II, pp. 152-155. 1942. A New Genus and Species of Fruit DOBSON, G. A. Bats, allied to Nyctimene. Amer. 1878. Cat. Chiropt. Brit. Mus. Mus. Novitates, No. 1204. GRAY, J. E. TAYLOR, E. H. 1866. A Revision of the Genera of Pteropine 1934. Philippine Land Mammals, p. 131. Bats (Pteropidae), and the Descrip- TROUGHTON, E. LE G. tions of some apparently Undescribed 1941. Furred Animals of Australia, p. 337.

- 4. "' ~ A -'"'.,''~ - W" (4K '7.'. -' YA YA417 ' .''i, >' N/¾.~' V 7, :4 >1~;, {'½2,iA '!' $.-I' - Ii'." !II /~~~~~~.'~~~~~~~~»4'7 11- >i4 Q'~~~~~~~~~~~~~~~-Y.~ 4, V'I ,.,7 -,: -~ ~ ~ -, -,~~~~,-Ce~ >44'>4- /4 7'.V '{ ,~l ";1.I44¼--,,, 'A ~~~,,:.~~~~~~~~~~~~,I ,-~~~~~~

~,,, ~ & 1~ ,,.,I,~~~~~~~~~~4'4. 44".1 N 4 4- ,- .-1<-1,-,, - - ' - ~ -,~ "j' ,j

T '4 . 4 > '4'' ,- j. - (>4IKj~- ''''' he" -'(A.T A--'' ,; ," 'i¾",-,. '"'" 'I-,, A, 'K / A ' -44 , -~~44~ . '4, -I~~~,.. A' A~-~ ' I I~ -1 ." -,,~.,~~ ~4 ~ `-,y ",- -~~~~~~~~~~~k~~~ IA 'A;,.)I.I~~~~~~~~~'(~~~~-l --.~.~,'~~,I~ -~~~~~lII.-~~~~~~~~ )~l I..-~~~~~~~~~~~~,"I11~,"\,-." 11'7

N 4/ 6% 4. - - .,, '7 "A'5 f, A A 7~I"f1 ~~,,---.- ~~~~, ~~... I4' > 4,7 ¼'., 4", ' f'i-,,,.~ ""~ fi X '-4 ,-~ ~ I -)/ ,, ,, . JI ~~~1-~~-;l.1.~~I,-I I~, ~~~~~~,'...~~~~(,, I ,I-I~~~~~~~~~~~i,~ 4 ,~ j4""7 a~. ''A'--'v74.NI .7'I"',.I'4.- .,1~~~~~~~.I 4~I '4II4 ,;''" , .II4'44' ,:4"Vt.. 'I4 ,,~~, 1,,~,4(',-l~ ~ 44,>,I";44' ,- "I7;l ~ ~ ~ J-~ 441.., 44. 744' ' 4 ; 4I4 II 4 ~--,,Ik.iI- , N"' K'>7-~ (/I ~ ,1( I 44,, 2I 44,t " ;''' 4,,,, ,VAI,,44 ,- 7 i444-f 1 ,~'>, 444

-~~)),;1,-1,",1.~l6 ~ ~ !, I. \ , ,I -~1.-44'~ I;4:A I. I7 /4 AI K AN ~ ~ ~ -

1, .; -~.I%V - I7>'!' 4. -744,~ ' ,1 ,PJ~,_;~- ~ ,.

/4' '. AK,,~ A-/ ">7'/'' ~ ~ 4 ~f~ ~ ~ ~ ~ ~ ~ ~ ,~~ 44I7I4II-,-., -k ,~ ¼" /,II~~~~I-,,,-~~~~~~~~~~,,- 8", ~ ~ >I 7 ~ ~ ~, I I4 4I,,; ',~;~). \ ,,-~-~4jII~ ~ \ , ~ ~ 4" >I

' -.1," .,',;' ~ ~ ~ .; 1 II -, -4' ''.I~~~¼ - -;-~~~~~~~~~~~~~~,, ~~~~~~~~~~~~~, ~~~~~~~~~: ~ ,. 1'I " >'.. ',-1 " '4 A,'- ~,,'4/ ~ A,,~ ~,, ~ `- -,,,i 422:, '4 ">' ' ,,, " ~',I7/44' -44~ ' >I4 4~ ;,.~-I" 4-~ / ,` )": ~ 4 4 > '' 4 ( 4 4 III47 , 4 4">,-,, ,, 1~~~"I.~~,-,"~744~ 4,>/,-4. i ?, ,-

4 4 .. 4 / .4 ~ ' 7< ", / t4I4 47, 444III4'N ~ .~'A,/'. I4 - 4. "' ~ A -'"'.,''~ - W" (4K '7.'. -' YA YA417 ' .''i, >' N/¾.~' V 7, :4 >1~;, {'½2,iA '!' $.-I' - Ii'." !II /~~~~~~.'~~~~~~~~»4'7 11- >i4 Q'~~~~~~~~~~~~~~~-Y.~ 4, V'I ,.,7 -,: -~ ~ ~ -, -,~~~~,-Ce~ >44'>4- /4 7'.V '{ ,~l ";1.I44¼--,,, 'A ~~~,,:.~~~~~~~~~~~~,I ,-~~~~~~

~,,, ~ & 1~ ,,.,I,~~~~~~~~~~4'4. 44".1 N 4 4- ,- .-1<-1,-,, - - ' - ~ -,~ "j' ,j

-~~)),;1,-1,",1.~l6 ~ ~ !, I. \ , ,I -~1.-44'~ I;4:A I. I7 /4 AI K AN ~ ~ ~ -

1, .; -~.I%V - I7>'!' 4. -744,~ ' ,1 ,PJ~,_;~- ~ ,.

/4' '. AK,,~ A-/ ">7'/'' ~ ~ 4 ~f~ ~ ~ ~ ~ ~ ~ ~ ,~~ 44I7I4II-,-., -k ,~ ¼" /,II~~~~I-,,,-~~~~~~~~~~,,- 8", ~ ~ >I 7 ~ ~ ~, I I4 4I,,; ',~;~). \ ,,-~-~4jII~ ~ \ , ~ ~ 4" >I

4 4 .. 4 / .4 ~ ' 7< ", / t4I4 47, 444III4'N ~ .~'A,/'. I4