214 i\Ir. K. Andersen on the Characters an-l 



nearly all species known, and arrived at the conclusion that 

 Desmalopex cannot be separated from Pteropus. It is fair to 

 add, however, that two or three years ago, when Miller was 

 studying the British Museum collection of bat-skulls for 

 his revision of the genera of bats, the series of Pteropus 

 skulls was much less complete than now; since then all the 

 species of Pteropus have been worked out for the ' Catalogue 

 of Chiroptera,^ and the number of skulls now available for 

 examination is more than four times as large. Miller's 

 opportunities for verifying the supposed differential characters 

 of Desmalopex were therefore much less favourable than 

 they would be now. 



The characters of Z)e5ma/o/?e.?- given by Miller (here printed 

 between inverted commas), and my comments thereon^ are 

 these : — 



(1) "Like Pteropus ; but skull with broader rostrum and 

 palate." — In no small number of species of Pteropus (e. g. 

 Ft. nmvaiensis, samo'eiis'is, anetianus, pselaphon *, tuberculatus, 

 pilosus,insularis,ph(Eoceptiahis, lomhucensis ; among the larger 

 forms Pt. mclanojjogon) the rostrum is quite as broad as or 

 even broader than in Pt. leucopterus. 



(2) "Orbits directed slightly more upward.'' — The orbits 

 of Pt. leucopterus are not directed more upward than iu 

 certain other species of Pteropus, e. g. Pt. vampyrus and 

 gxganteus f. 



* A comparison of the measurements of P^. leucopterus and pselaphon is 

 perfectly fair, since the skulls of these two species are precisely of the same 

 size (and indeed so similar also in other respects as to differ only in trivial 

 details). SliuUs measured, Pt. leucopterus, B.M. 6:^.1.14.3, and (in 

 parentheses) ft. pselaphon, B.M. 94.7.3.2 : gnathion to back of zygomatic 

 process of squamosal 54 mm. (54) ; zvgomatic breadth 38 (37) ; across 

 postero-external coruers of alveoli of m'-m^ 19 (19j ; across alveolar 

 borders between p^ and p* 15'6 (15'8) ; breadth of palate between inner 

 sides of m^-m* V2b (13) ; between postero-internal corners of p'^-p* lO'S 

 (11-2) ; between p'-p' 9-8 (97). 



t Measurements, by goniometer, of angle formed by alveolar margin 

 (front of p^ to back of m'j and tangent to upper and lower edges of 

 orbit:— Pf. pselaphon (B.M. 94.7.3.2) 4-5=, Pteralopex atrata (88.1.5.9; 

 type) 32-33°, Pt. leucopterus (62.1.14.3) 28°, Pt. giga7iteus leucocephalus 

 (45.1.8.274; Nepal) 27^. — Miller writes {op. cit. p". 61) that the orbits 

 of Pteralopex are " strongly upturned," i. e. more so than in " Desma- 

 lopex," and la^'s some stress on the supposed three stages of tlie po.'^itiun 

 of the orbits as marked by Pteropus (orbits less upturned), Desmalopex 

 (more upturned), and Pteralopex (struugly upturned), this being one of 

 his arguments for the alleged intermediate position of Desmalopex 

 between Pteropus and Pteralopex. The true facts are those shown by 

 the measurements given above and verified by an instrument still finer 

 than a goniometer, nameh', the eye, that Pt. leucopterus does not differ 

 in this respect from I't. giijanteus, and that the orbits of Pteralopex are 

 even sliglitlv less upturned. Millar's mistake is, however, perfectly 



