o.^O ]\Ir. K. Andersen on the 



A'olkcnic. NeJ. Ind. (1) vii., 1883). The Leydcn specimen 

 is fully adult, the type slightly immature (but for all practical 

 purpose perfectly fuU-growu). 



1896, Cephalotes viridis, Ueude, Mem. Hist. Nat. Emp. 

 Chin. iii. p. 176, footnote, pi. v. fig. 1. — A distinct species. 

 Differential cliaraetcrs : — A well-marked antero-internal basal 

 cusp both in uii and ni' ; breadth across lower edges of 

 lachrymal foramina ir5-12'8 mm.; forearm 113-117*5. 

 Range : — Burn, Amboina, Ceram, Banda Islands, Key 

 Islands. 



The next, and thus far the latest, reviser of the genus 

 (Matschie, ' Megachiroptera dcs Bei-liner Museums,^ 1899) 

 again recognized only one species, Cephalotes paJliat us, with 

 the synonyms Cephalotes peroni, Hypoderma moluccense, 

 Cephalotes minor, and C. viridis. So far from being identical, 

 these four species represent in reality four different sections 

 of the genus, a fact which, however, can only be realized on 

 examination of their dentition. But quite apart from dental 

 differences, it would only be possible to put the above four 

 species together under one heading on supposition that the 

 individuals show a perfectly unparallelled amount of size 

 variation (forearm in Dobsonia minor 80 mm., in D. moluc- 

 censis 133'5-141). 



1905, Dobsonia magna, Thomas, Ann. &. Mag. N. H. (7) 

 xvi. p. 423. Type locality, Taraata, Mambare River, British 

 New Guinea. A distinct species, closely allied to D. moluc- 

 censis, from which it differs only in its still larger size (fore- 

 arm 146-152"5 mm.: largest species known). Range: — 

 The whole of New Guinea ; Mysol ; Waigeou. 



1905, Cephalotes pannietensis, De Vis, Ann. Queensl. jMus. 

 vi. p. 36. Type locality, Panniet Island, Louisiade Archi- 

 pelago. A distinct species : similar in dentition to D. mo~ 

 luccensis and magna, but much smaller ; forearm about 

 109-112 mm. Range: Trobriand group (Kiriwina), D'En- 

 trecasteaux group (Fergusson Island), Louisiades (Panniet). 



In 1906 Dr. Jentink gave a very elaborate table of external 

 measurements of nearly all the specimens of Dobsonia in the 

 Leyden Museum (Notes Leyd. Mus. xxviii., table facing 

 p. 168). It is necessary to draw attention to the fact, how- 

 ever, that not all of these measurements are directly com- 

 parable. Of the 42 specimens measured 18 are more or less 

 immature (specimens a, g, k, 1, n, o, p, q, v, w, x, z, P, g^, j^, 

 k^ p^, w^). 



From the above it will be noticed that all the principal 

 revisers of the genus, from Teraminck, through Peters, Gray, 



