ACERODON. 415 



femur and Blong outer side of forearm. Leugth and distribution 

 of lur esseutially alike in all species. Individual hairs on back and 

 rump often finely crinkled, producing (on not too close examination) 

 the impression of a narrow aunulatiou of the hairs. 



Se.vual differentiatio7i. — Canines in all species slightly heavier in 

 males than in females. Males with a tuft of rigid unctuous hairs 

 on each side of neck, more or less concealed by surrounding f ur • 

 neck-tufts entirely wanting or only indistinctly dilierentiated in 

 females. 



liange. — Celebes group (Celebes, Salayer, Sula Mangoli) ; Timor 

 group (Timor, Flores, Alor, Sumba) ; Talaut Islands ; Philippines. 



Affinitus of r/enus. — Acerodon is closely allied to Pteropvs, from 

 which it differs only in the more highly specialized dentition *. 

 But even these differences in dentition are by no means great when 

 viewed in the light of certain modifications of the teeth occurring in 

 various species of Fteropits. To the broad and sharply defined inner 

 basal ledge in p^, m,, and m.^ of Acerodon there is a perfect analogy 

 in one single species of Ftero/ms, Pt. anetianus (allied to Ft. sa- 

 moemls), the dentition of which is, however, in other respects 

 unmodified Pteropine ; a faint initial stage towards the development 

 of a similar inner basal ledge is seen in some (not all) individuals of 

 Ft. samoensis (see above, p. 2S6), pilosus (p. SOti), and ornatus, 

 species which belong to three different groups of the genus. No 

 known species of tteropus possesses a well differentiated antero- 

 internal basal cusp in ]>* and m\ but tendencies to the development 

 of a similar cusp in p^ and p,, p^ and p^ are not quite unknown in 

 that genus (see above, p. 7U, under the heading " traces of antero- 

 internal basal tubercles in p', p% p^, and p^ "'). It is evident, there- 

 fore, that the peculiar characters of the Acerodon dentition are 

 scarcely more than further developments of tendencies already 

 present, more or less clearly, in diifereut species of Fteroiyus. The 

 claim of Acerodon to be considered a distinct genus must therefore 

 rest on the following facts and considerations only, viz., that it 

 exhibits a higher degree of specialization of the dentition than any 

 Fteropns, in so far as no form of Fteropus possesses an inner basal 

 ledge in the lower molariforra teeth combined ivith a well-developed 

 antero-internal cusp in p* and m' ; that the six species here referred 

 to Acerodon form a perfectly natural group, the members of which 

 arc undoubtedly more closely related to each other than to any 

 living species of Fteropus, and therefore may be safely presumed to 

 have a common origin; and that they cover a well-defined area 

 of the Indian .\rchipelago, extending from Celebes southward to 

 the Timor group and northward, through the Talaut Islands, to the 

 Philippines. 



* AI30 the somewhat larger size of in- in Acerodon is probably a secondarv 

 character, due to the general specialization of the dentition. The structure of 

 this looth is, allowing for its niiich smaller size, quite similar to tiiat of rn' • 

 even a small antero-intern.al basal cusp is sometimes developed ; and iii A. cele- 

 hensis, which in dentition is the most primitive species of the genusj m- is not 

 larger than in many Pieropus. 



