12 BARBOUR: ZOOGEOGRAPHY. 



Batu group; makins it ajipear probable that in times past there were three 

 great peninsulas jutting out in three directions from Sumatra, northeastward, 

 eastward, and southeastward; and that, while the origins of these three penin- 

 sulas were very close together on the Sumatra shore, still they were not cross- 

 connected with one another. This will be emphasized again when we consider 

 the fauna of these various islands. 



The same subsidence which broke up these peninsulas into islands probably 

 separated Sumatra from the mainland. 



Werner has suggested that these islands have each different Sumatran species 

 upon them because they each lie opposite the particular ranges within Sumatra 

 for various species; and that their species have thus come to them from just that 

 part of Sumatra lying nearest them. This would suggest that Werner supposes 

 the islands to have received their species by "flotsam and jetsam" methods, 

 which I do not believe. The islands are \'ery incompletely known, and this 

 probably accounts for the discrepancies in the lists of species, which came by 

 the land connections suggested above. 



There is good reason to believe, owing to the fact that all the rivers of large 

 size empty either into the straits of Malacca or into the southern part of the 

 China Sea between Sumatra and Borneo, that the island at the present time lies 

 nearer to the Peninsula, to the Riouw and Lingga Archipelagoes, and to the 

 island of Banka, than it did in the recent past. All of this eastern coast region 

 of Sumatra is composed of low, swampy alluvial lands, through which the rivers 

 flow, carrying down the detritus from the mountains in the form of silt, which 

 is constantly being deposited in the deltas of the rivers; and these deltas are 

 also being extended fast by the great fringing zone of mangrove vegetation which 

 lines the shore. The straits themselves are very shallow. This may be con- 

 sidered a recent re-approachment, which may serve to connect the lands again 

 in a short period of time. 



The reptiles and amphibians of the island show a very close relationship 

 to those of the nearest mainland. Werner (Zool. jahrb. Syst., 1900, 13, p. 

 479-508, pi. 31-35) has arranged a complete Ust of the herpetologic fauna of the 

 island, which, however, has been considerably changed by the researches of the 

 last ten years. On pages 503-508 he presented some general conclusions regard- 

 ing the geographical distribution of Sumatran reptiles and amphibians in com- 

 parison with those of Ceylon, Borneo, Java, and Celebes. The work itself is 

 useful, but contains a considerable number of statements which may be justly 

 challenged. In the first place, in 1896 Werner (Verh. Ges. Wien, 1896, 46, 



