HALMAHERA GROUP. 43 



them. We should expect, then, from the conditions here, that this group of 

 islands would show a much stronger preponderance of Papuan forms than the 

 Ambon group, and as a matter of fact this is the case. To sum up, we conclude 

 that Hahnahera was joined to both Celebes and Papua; while Obi had an erst- 

 while indirect connection with both by being part of the old Halmahera-Papua 

 commissure. 



Twenty-five lizards are recorded, six of which are peculiar to the island. 

 One of these is a species of Gehyra, a genus of usually wide-ranging species ; 

 while the others are all species of Lygosoma, using the generic term in its broad- 

 est sense. They are closely related to other Papuan species, and in many cases 

 have been considered identical until a very short time ago. It is strange to 

 see these Papuan species replaced in Halmahera by distinct forms, whereas in 

 the Ambon group, so far as I can discern, individuals occur which are indis- 

 tinguishable from Papuan examples, even though the island seems less closely 

 related geographically to the Papuan region than does Halmahera. 



Of all the species of lizards, only two can be said to be preeminently Malayan; 

 these are Calotes cristatellus (Kuhl) and Mahuya multifasciata (Kuhl) . The fact 

 that there are fourteen Lygosomas shows how preponderating is the Papuan 

 element among the hzards. 



Of snakes there are seventeen, not counting the sea-species; and of these, 

 six are peculiar to the island, two of them belonging to genera Styporhynchus 

 and Calamorhabdium, not occurring elsewhere. The Malayan Python reticu- 

 latus (Schn.) occurs here, along with P. amethystinus (Schn.), which does not 

 go west of Timor. The distribution of Cylindrophis rufus (Laur.), Ptyas 

 dipsas (Schlegel), Typhlops ater Schl., and Testudo forstenii (Schl. & MiilL), 

 if it exists on the island, which I now doubt very much, shows that there was 

 once direct connection between Celebes and Halmahera, perhaps through the 

 Sula Islands, or perhaps by an independent bridge to the north of that group. 

 A connection with Celebes through Obi is very improbable on hydrographic 

 grounds. The Ambon group and especially Buru seem far more intimately 

 related to the Sula Islands than these are to the islands about Halmahera. 



Two species of Natrix and a Dendrophis, all confined to the island, are, of 

 course, of Malay origin, although evidently modified by long isolation here. 

 The other snakes are Papuan. 



I have spoken already of the single land-tortoise, Testudo forstenii Schl. 

 & MiilL, a rare species, reported from Celebes and Halmahera though it is 

 probably absent from the latter; while only a single fresh- water tortoise occurs. 



