ZOOGEOGRAPHY. 159 



can now offer no explanation for, the fact that Mysol, as has been pointed out, 

 seems to bear a definite relation to Ceram as regards its reptiles. The birds of 

 Mysol are strictly Papuan, and the relationship of the island to Papua is well 

 established by the fact that it is separated by only the shallowest water. Roths- 

 child and Hartert (Nov. zooL, 1901, 8, p. 56) remark, after reporting on the 

 ornithology of the island: "Mysol belongs faunistically entirely to New Guinea; 

 and this is easily explained by the number of small islets stretching in a northern 

 and northeastern direction from Mysol towards Salwatty and New Guinea, and 

 the shallow sea surrounding Mysol and New Guinea; while deep sea, of more 

 than 200 fathoms, separates Mysol from the Moluccan Islands. Most of the 

 birds are quite similar to those of New Guinea; others, subspecifically allied to 

 the latter. Moluccan influence is scarcely perceptible." Nevertheless, the 

 strong Malayan element in the herpetology can only be explained by supposing 

 either that there has been a land-bridge to Ceram where deep water now exists, 

 or that we have to do with a number of erroneous records. We have not seen 

 reports of any recent collections made in Mysol, but it liardly seems possible that 

 so many records are invalid. 



The relation of Halmahera to New Guinea has already been spoken of; and 

 there remains now to mention the fact that Timor has a strong Australian, and 

 perhaps a Papuasian, tinge to the fauna, which has probably not come through 

 Timor-Laut, which lies but a little off the great shallow bank that extends out 

 beyond the Aru Islands, from there down to Melville Island, and thence far to 

 the southward. The fauna of Timor-Laut is different from that of Timor; 

 therefore it seems entirely possible that while Timor-Laut may have been 

 connected formerly with the Ke Islands, or with Papuasian land about Aru, 

 and so indirectly with the Moluccas, Timor must have been connected directly 

 with some old extension of Papuan land reaching westward. A connection 

 which allowed such forms as Liasis and Chelodina to reach Timor must have been 

 fairly recent, or perhaps have lasted for a very long time. 



This whole series of suggestions may seem absurdly simple, for there can be 

 no possible doubt that conditions were far more complicated than these con- 

 clusions would seem to presuppose. There is strong evidence of a two-fold fauna 

 in Java. This would make it far from unprobable that Java was for a long 

 time in reality two islands, since the west Javanese fauna is in many respects 

 strikingly different from that of east Java. Moreover cUmatic conditions are 

 very different in the two ends of the island, west Java being far more heavily 

 forested, and having a much greater annual rainfall, than east Java. Again, 



