IxXXViii GEOGIIAPIIICAL DISTIUIiUTION. 



confined to New Guinea is : (1) occurring both west (Amboiua group) 

 and east of New Guinea (Solomon Islands), Hoiisetius hrachi/otis : 

 (2) common with the Gilolo group, Pteropus chrysauchen: 

 (^3) common with the Key and Aru Islands, hut not extending to 

 Australia, Macrofjlossus hujochilKS nanus: (4) common with Aus- 

 tralia and also with either Key or Aru Islands (if not both), JSycti- 

 mene papuanus and Suconiicteris crassa papuana : (5) direct invader 

 from New Guinea through the Torres Straits islands into Australia, 

 Pteropus conspicillaius : (6) common with the Bismarck Archipelago 

 only, Melonycleris melanops. A faunistic contrast between the 

 North-west and South-east of New Guinea is, so far, only indicated 

 by Pieropus chri/saachtn of the former area (extending to the Gilolo 

 group) being replaced in the latter by its geminate species. Pi. con- 

 spiciUatits (extending to N. Australia) ; but more definite suggestions 

 for a subdivision of New Guinea into faunistic districts may be 

 derived from the distribution of the species of Nyctimene when 

 better known. 



Key Islands. — Four genera have been recorded, Pteropus, Doh- 

 sonia, Nyctimene, and Syconycteris, each (so far as known with 

 certainty) represented by one form ; in addition to these, Macro- 

 glossiis probably occurs in the islands. Of the four forms three are 

 ncculiar, viz. Pteropus keyensis, a species allied to Pt. melano- 

 pogoii from the Amboina group, Banda Islands, and Timor Laut, 

 but with no close relative in New Guinea; Dohsonia viridis viridis, 

 also with Amboiuan affinities (no ecjuivalent in New Guinea); and 

 Syconycteris crcssa Jceyensis, which is more intimately connected 

 with the New Guinean than with the Australian race of the species. 

 The fourth form, JSyciiinene p)apuanus,\s common to the Key Islands 

 and New Guinea. As evident from this, the Fruit-bat fauna shows 

 both New Guinean and Amboinan affinities, and the former are in so 

 far more intimate than the latter, as the two Amboinan forms have 

 differentiated the one into a distinct species, the other into a sub- 

 species, whereas of the two New Guinean forms the one is unchanged, 

 the other distinguishable as a subspecies. 



Arw Islands. — Nyctimene has not been recorded from these 

 islands, otherwise the genera are the same as in the Key Islands, 

 but the relative distinctness of the Aru fauna is shown by the fact 

 that no form is known with certainty to be common to both groups 

 (Dohsonia inoluccensis and Macroylossus layochilus nanus may occur 

 in both). The affinities point partly toward the Key Islands and the 

 Amboina group (Pteropus aruensis, a peculiar species, related to 

 Pt. keyensis and to the Amboinan Pt. melanopogon ; and Dohsonia 

 violuccensis, indistinguishable from the Amboinan form), partly and 

 more decidedlj- toward New Guinea (two common forms, Macro- 

 ylossus lagochilus nanus and Syconycteris craesa papuana ; one 

 distinct species, Pteropus macrotis, closely allied to the New 

 Guinean Pt. ('pularius). 



