OHAP. IX.] 



INDO-MAIJY ISLANDS, 



145 



existed as a distinct islund even longer than Sumatra and 

 Borneo, and tiiere are some geological and geograpliical 

 facts which render this not so improbable as it would at 

 first seem to he. Although on the map Eanca appears so 

 close to Sumatra, this does not arise from its having been 

 recently separated from it; for the adjacent district of 

 Palemhang is new land, being a great alluvial s^vamp 

 formed by torrents from the mountains a hundred milen 

 distant. Banca, on the other hand, agrees with Malacca, 

 Singapore, and the inter\'eniug island of Lingen, in being 

 fonned of granite and hitt'iite ; and these have all most 

 likely once formed an exti^nsion of the Malay peninsuhi. 

 As the rivers of Borneo and Sumatra have been for ages 

 lilling np the inter^'ening sea, wo may be sure that its 

 depth has recently been greater, and it is very probable 

 that those largo islands were never directly connected with 

 each other except through the IVfalay peninsula. At that 

 period the same species of squirrel and Pitta may hava 

 inhabited all these countries ; but when the subteiTanean 

 disturbances occuiTed which led to tlie elevation of the 

 volcanoes of Snmatm, the small island of Iknca may have 

 been separated first, and its productions being thus isolated 

 might be gradually modified before the separation of the 

 larger islands had been completed. As the southern ]>art of 

 Sumatra extended eastAvard and formed the narrow straits 

 of Banca, many birds and insects and some Mammalia 

 would cross from one to the other, and thus produce a 

 general similarity of productions, while a few of the older 

 inhabitants remained, to reveal by their distinct forms their 

 dilTerent origin. Unless we suppose some such changes 

 in physical geogrnjihy to have occurred, the presence of 

 peculiar species of birds and mammals in such an island 

 as Banca is a hopeless puzzle ; and I think I have slio^n 

 that the changes requiretl are by no means so improbable 

 as a mere glance at the map would lead us to suppose. 



For our next example let us take the great islands of 

 Sumatra and Java. These approach so closely together, 

 and the chain of volcanoes that runs through them gives 

 such an air of unity to the two, that the idea of their 

 having been recently dissevered is immediately suggested. 

 Tho natives of Java, however, go further than this; for 



h 



