3G0 ZOOLOGICAL GEOGRAPHY. [part hi. 



broken up into many small volcanic islets in which a limited 

 number of Malayan types alone survived. Such a condition of 

 things will account for the very small variety of mammalia com- 

 pared with the tolerably numerous genera of birds, that now 

 characterise its fauna ; while both here and in Celebes we find 

 some of the old Malayan types preserved, which, in the extended 

 area of the Sunda Isles have been replaced by more dominant 

 forms. 



The next important change would be the separation of Java ; 

 and here also no doubt a considerable submergence occurred, 

 rendering the island an unsuitable habitation for the various 

 Malay types whose absence forms one of its conspicuous features. 

 It has since remained permanently separated from the other 

 islands, and has no doubt developed some peculiar species, while 

 it may have preserved some ancient forms which in the larger 

 area have become changed. From the fact that a number of its 

 species are confined either to the western or the eastern half of 

 the island, it is probable that it long continued as two islands, 

 which have become united at a comparatively recent period. 

 It has also been subjected to the immigration of Indo-Chinese 

 forms, as already referred to in the earlier part of tliis sketch. 



We have thus shown how the main zoological features of the 

 several sub-divisions of the Malayan sub-region may be 

 accounted for, by means of a series of suppositions as to past 

 changes which, though for the most part purely hypothetical, 

 are always in accordance with what we know both of the 

 physical geography and the zoology of the districts in question 

 and those which surround them. It may also be remarked, that 

 we know, with a degree of certainty which may be called absolute, 

 that alternate elevation and subsidence is the normal state of 

 things all over the globe ; that it M-as the rule in the earliest 

 geological epochs, and that it has continued down to the 

 historisal era. We know too, that the amount of elevation and 

 subsidence that can be proved to have occurred again and again 

 in the same area, is often much greater than is required for the 

 changes here speculatod on, — while the time required for such 

 chanues is certainlv less than tliat necessitated bv the changes 



