156 BARBOUR: ZOOGEOGRAPHY. 



Asia through Austraha into South America, can hardly be doubted. They 

 have never been especially successful in South America, and only two or three 

 species have reached to the United States. But the many prominent forms of 

 Lachesis and Crotalus, which have come from Eurasia, have passed south, and 

 have met their ancestral immigrants from Australia in South and Central America. 



One of the points which van Kampen and previous writers have failed to 

 emphasize is the marked difference between the faunas of the Ke and Aru 

 Islands. This has been spoken of in detail (p. 44—49) in the notes on the herpe- 

 tology of these two groups of islands. We can not consider them both equally 

 Papuan. The Ke Islands have an impoverished fauna, due largely to their 

 small size; for the number of species supported by any island is, other things 

 being equal, directly proportionate to that island's area; it is important to note 

 that a number of typical Papuan types found on the Ke group extend further, 

 and are met with either in the island of Ceram, or else more widely distributed 

 in the Moluccas; while, on the other hand, the Aru Islands possess a larger and 

 less differentiated number of true Papuan species, which are not found on other 

 groups of islands. The mere presence of Draco in Ke' should have suggested at 

 once that there was a fundamental difference between the biota of the two 

 islands. 



Hartert (Nov. zool., 1901, 8, p. 2), describing a collection of birds from 

 the Southeast Islands, — that is, those between Ceram Laut and the Ke group, 

 — says: "Zoologically, the Key Islands belong to the Moluccas.... Only 

 sixty miles eastward of Dobbo in the Aru group, and just as near to New Guinea 

 as the Aru Islands, the Key Islands have only very few specially Papuan bird 

 forms." He continues that the Southeast Islands in general want primeval 

 forest, and are apparently of recent, coraline nature; and that their ornis shows 

 relations to that of Ceram and Ke. 



While dealing with the Ke Islands, it is necessary to postulate land-bridges 

 which have existed in the past, and which have left behind them entirely different 

 sorts of hydrographic conditions. On the one hand, there are left distinct 

 evidences of the bridge to Ceram in the shape of submerged ridges, or chains of 

 islets connected by submarine banks; and, on the other hand, there are no sub- 

 marine traces, so far as we know, of the Ke-Aru Bridge, but rather a surprisingly 

 deep area where once the l^ridge probably existed. Soundings are, however, 

 unfortunately few in this region. Such a depth near an island, and between 

 it and the land to which it was recently joined, is not at all a rare condition. 

 Such an island can not have been raised out of the sea, but the dividing deep 



