chap, ix.] INDO-MALAT ISLANDS. 231 



sequent on the activity of the Javanese volcanoes along 

 the southern extremity of the land, and leading to the 

 complete separation of that island. As the volcanic belt 

 of Java and Sumatra increased in activity, more and more 

 of the land was submerged, till first Borneo, and after- 

 wards Sumatra, became entirely severed. Since the epoch of 

 the first disturbance, several distinct elevations and depres- 

 sions may have taken place, and the islands may have been 

 more than once joined with each other or with the main 

 land, and again separated. Successive waves of immigra- 

 tion may thus have modified their animal productions, 

 and led to those anomalies in distribution which are so 

 difficult to account for by any single operation of elevation 

 or submergence. The form of Borneo, consisting of radiat- 

 ing mountain chains with intervening broad alluvial 

 • valleys, suggests the idea that it has once been much more 

 submerged than it is at present (when it would have 

 somewhat resembled Celebes or Gilolo in outline), and has 

 been increased to its present dimensions by the filling up 

 of its gulfs with sedimentary matter, assisted by gradual 

 elevation of the land. Sumatra has also been evidently 

 much increased in size by the formation of alluvial plains 

 along its north-eastern coasts. 



There is one peculiarity in the productions of Java that 

 is very puzzling — the occurrence of several species or 

 groups characteristic of the Siamese countries or of India, 



