WALLACE'S LINE. 311 



book on the geographical distribution of animals, repeatedly 

 points out that many species encroach on the limits of the 

 neighbouring province in a very singular and incomprehensible 

 manner; and he justly infers that there must be some peculiar 

 means of dispersal as yet unknown to us, by which these species 

 are enabled to overstep the limits apparently assigned to them 

 by nature. Moreover, he speaks of the fauna of the island of 

 Celebes, included in the Australian region, as exhibiting so re- 

 markable a mixture of animal types that it might just as well 

 be included in the Indian region. But irrespective of these 

 forms, which prove that Wallace's line does not indicate an im- 

 passable frontier, there yet remains so vast a number of ex- 

 tremely different species, peculiar to each of thess regions and 

 belonging exclusively to one, that the sharp distinction so long 

 recognised between them appears fully justified. 



The question now is whether this distinctive contrast can 

 be explained, and more particularly how it happens that two 

 islands lying so close to each other as Bali and Lombok should 

 by Wallace's line be placed in two different regions of animal 

 distribution. Wallace himself who, so far as I know, was 

 the first to attempt to explain this phenomenon does so as 

 follows : 



He assumes that at some former period the Indian continent 

 and the Indian islands as far as Java, Borneo, and the Philip- 

 pines, were connected ; and that, in the same way, Australia 

 with New Guinea, the Moluccas, and Celebes, were in connec- 

 tion, and that only the group of islands from Timor as far as 

 Lombok were perhaps excluded. This last must probably have 

 been divided from Java by a deep sea ; and it was not till a later 

 period that Bali by the side of Java and, the smaller islands 

 as far as Lombok by the side of Timor, were raised from the 

 bed of the sea. The differences of the fauna which nevertheless 

 occur on the individual islands of each region he endeavours to 

 account for by their separation at different periods from the 

 Indian or Australian continents. 



Now, it certainly cannot be disputed that very many circum- 

 stances in the distribution of the land animals on these islands 

 argue in favour of this view ; thus, for instance, the fact, un- 



