KE ISLANDS. 45 



a few miles wide; Nuhu Roa, some 20 miles long and ten miles wide; as well as 

 a large number of islets of varying size, on one of wliich the town of Tual is 

 situated, the main settlement of the group, the island being of the same name. 



The islands are, one might say, almost connected with Ceram; for the few 

 soundings are of depths of only about a hundred fathoms, while many islands 

 lie in a direct line and form a chain with many submerged banks and reefs just 

 awash between them. Between the Ke Islands and the Ai-u Islands, and be- 

 tween the Ke Islands and New Guinea, the depth of water seems to be much 

 greater. There are a number of soundings off towards the .\i-u Islands reaching 

 the depth of almost 2,000 fathoms; and towards the Lsland of Adi, off the New 

 Guinea coast, almost as great a depth has been found. 



These conditions at once associate the islands rather with the Moluccas 

 than with the true Papuan groups; although, as we find on studying the records 

 for the fauna, the Papuan element seems predominant. Our hydrographic 

 knowledge of the region is too limited to say surely that there are no sub- 

 marine evidences of connection with either New Guinea or the Aru Islands. 

 If such evidences do not exist, we have simply another added case which shows 

 how dangerous it is to rely exclusively upon data of this sort for the recon- 

 struction of past land connections. That this connecting area was one of 

 comparatively recent date, geologically speaking, is evidently proved by the 

 intimate faunal relationship which the Ke Islands have with true Papuasia. 

 It has been customary in the past to speak of the Ke and Aru groups as both 

 being purely Papuasian in their fauna. This is perfectly true of the Aru Islands, 

 but the Ke Archipelago shows strongly the influence of its direct relationship 

 with Ceram. 



There has just appeared an exhaustive study of the herpetology of these 

 two groups from the pen of Dr. Jean Roux (Abh. Senck. nat. ges., 1910, 33, p. 

 211-247, taf. 13-14). Dr. Roux's data are especially valuable inasmuch as he 

 adds considerably to the list which we previously had of species from both these 

 groups. The following notes are made up largely from his list. 



The snakes are nine in number, three of which are species of Typhlops; 

 one, T. kraalii Doria, is pecuUar to the island, and another, T. muUilineatus 

 Schl., is found both at Tual and in Dutch New Guinea. The other species are 

 Python amethystinus (Schn.), Stegonoius modestus (Schl.), S. cucullatus (Dum. & 

 Bibr.), Dendrophis calUgasier Gthr., Boiga irregularis (Bechsf.), and A canthophis 

 aniarcticus (Shaw). These, as may be seen at a glance, are forms of Papuan 

 origin; yet nearly all of them occur in the Moluccas as well. The distribution 



