288 THE ARU ISLANDS. [chap, xxxiii. 



unique ; for I have been unable to find any other record 

 of an island of the size of Aru crossed by channels which 

 exactly resemble true rivers. How these channels origi- 

 nated were a complete puzzle to me, till, after a long consi- 

 deration of the whole of the natural phenomena presented 

 by these islands, I arrived at a conclusion which I will now 

 endeavour to explain. There are three ways in which we 

 may conceive islands which are not volcanic to have been 

 formed, or to have been reduced to their present condition, 

 — by elevation, by subsidence, or by separation from a 

 continent or larger island. The existence of coral rock, or 

 of raised beaches far inland, indicates recent elevation ; 

 lagoon coral-islands, and such as have barrier or encircling 

 reefs, have suffered subsidence ; while our own islands, 

 whose productions are entirely those of the adjacent con- 

 tinent, have been separated from it. Now the Aru Islands 

 are all coral rock, and the adjacent sea is shallow and full 

 of coral ; it is therefore evident that they have been 

 elevated from beneath the ocean at a not very distant 

 epoch. But if we suppose that elevation to be the first 

 and only cause of their present condition, we shall find 

 ourselves quite unable to explain the curious river-chan- 

 nels which divide them. Fissures during upheaval would 

 not produce the regular width, the regular depth, or the 

 winding curves which characterise them ; and the action 

 of tides and currents during their elevation might form 



