of the Aru Islands. 479 



a portion of such continents, and geological evidence shows that 

 the separation had taken place at no distant period. We must, 

 therefore, suppose Aru to have once formed a part of ^aw 

 (iuinea, in order to account for its peculiar I'auna, and this view 

 is supported by the physical geography of the islands ; for, while 

 the fathomless Molucca sea extends to within a few miles of them 

 on the west, the whole space eastward to New Guinea, and south- 

 ward to Australia, is occupied by a bank of soundings at a 

 uniform depth of about 30 or 40 fathoms. But there is another 

 circumstance still more strongly proving this connexion : the 

 great island of Aru, 80 miles in length from north to south, is 

 traversed by three winding channels of such uniform width and 

 depth, though passing througli an irregular, undulating, rocky 

 country, that they seem portions of true rivers, though now^ 

 occupied by salt watei*, and open at each end to the entrance of 

 the tides. This phasnomenon is unique, and we can account for 

 their formation in no other way than by supposing them to have 

 been once true rivers, having their source in the mountains of 

 New Guinea, and reduced to their present condition by the sub- 

 sidence of the intervening land. 



This view of the origin of the Aru fauna is further confirmed 

 by considering what it is not, as well as what it is ; its defici- 

 encies teach as much as what it possesses. There are certain 

 families of birds highly characteristic of the Indian Archipelago 

 in its western and better-known portion. In the Peninsula of 

 Malacca, Sumatra, Java, Borneo and the Philippine Islands, the 

 following families are abundant in species and in individuals. 

 They are everywhere common birds. They are the Buceridce, 

 Picidce, Bucconidts, Trogonidce, Meropidce, and Ewi/Iaimidce ; but 

 not one species of all these families is found in Aru, nor, w'ith 

 two doubtful exceptions, in New Guinea. The whole are also 

 absent from Australia. To complete our view of the subject, it 

 is necessary also to consider the ^lammalia, which present pecu- 

 liarities and deficiencies even yet more striking. Not one species 

 found in the great islands westward inhabits Aru or New Guinea. 

 With the exception only of pigs and bats, not a genus, not a 

 family, not even an order of mammals is found in common. No 

 Quadrumana, no Sciuridre, no Carnivora, Rodeutia, or Ungu- 

 lata inhabit these depopulated forests. With the two exceptions 

 above mentioned, all the mammalia are Marsupials ; in the great 

 \vestern islands there is not a single marsupial ! A kangaroo 

 inhabits Aru (and several New Guinea), and this, with three or 

 four species of Cuscus, two or three little rat-like marsupials, a 

 wild pig and several bats, are all the mammalia I have been able 

 either to obtain or hear of. 



It is to the full development of such interesting details that 



