CHIRONAX. 



659 



m- lost) ; (3) the aiitero-external lobo of p' in Balioni/cteris is in 

 the present genus represented by a small but well-defined cusp. 

 ])' perhaps deciduous (present in cotype skulls " « " and "<•," 

 apparently absent in "6"). All other dental characters as in 

 Balionydens (upper canines with deep vertical groove, and Avithout 

 secondary cusps ; m\ m^, and in,^ reduced quite as in Balionifcieris, 

 ra' in section about halt' the size of p^, m^ two-thirds of p^, m^ 

 one-fourth to one-fifth of m, ; no trace of surface cusps in any 

 cheek-tooth). 



Palate-ridges. — Unknown (probably not diliering from those of 

 Bihonijcteris). 



Rvternal characters. — Differences from Balioni/cterii^ : (1) wings 

 conspicuously larger than in Cynopterns, but not nearly so large as 

 in BaUo)v/cteris :, index of third digit in Ci/nojiterus 1637, in 

 Chironax 1757, in Balwni/cteris 1847, of fourth digit respectively 

 1303, 1393, and 1485, of fifth digit 1266, 1310, and 1403 ; third, 

 fourth, and fifth metacarpals exactly as in Cijnopterus (considerably 

 lengthened in Ballon >/cter is), but phalanges of third and fourth 

 digits lengthened by 11-15, those of fifth digit by 4-8 per cent. ; 

 in consequence of these modifications, second phalanx of third 

 digit subequal to (in Cijnopterus and Balionycteris much shorter 

 than) its metacarpal, and second phalanges of fourth and fifth digits 

 slightly but distinctly longer (in Balionycteris rather shorter) than 

 rirst phalanges of same digits ; compare table below (upper row, 

 wing-indices of Chironax, based on two adult specimens, middle 

 row those of Balionycteris, lower row those of Cynopterus) : (2) 

 membranes apparently inserted on second, instead of on first, toe 

 (evidence not quite clear enough from the available material): 

 (3) ears and membranes unspotted: (4) tibia (in single species 

 known) clothed above quite or nearly to ankle. Other characters 

 as in Balionyctens : tail absent, caudal vcrtebne none (one skeleton 

 examined), interfemoral undeveloped or extremely narrow in centre, 

 unmodified along tibia, calcar present. 



Affinities. — Owing to the scarcity of specimens in collections (the 

 cotypes obtained about 90 years ago being thus far the only 

 specimens on record) the characters and affinities of the single 

 species of this genus, Temminck's Pteropus iiieJanocejjhalus, have 

 hitherto been very imperfectly known. By the large majority of 

 authors it has been considered, apparently without hesitation, a 



2 'i 



