18 BARBOUR: ZOOGEOGRAPHY. 



certainly not transported fortuitously. The fact that there are no species 

 strictly peculiar to the island, and that these deUcate subterranean amphibians 

 are found upon it, proves conclusively that we can not consider Nias as an 

 oceanic island, but that it has derived its fauna directly from that part of Sumatra 

 which lies directly opposite to it — that is, between the equator and the second 

 parallel of north latitude — by a land bridge, and not, as has been suggested, 

 by accidental transport. 



Mentawei Islands. 



This group consists of the little-known island of Siberut in the north, and 

 southward, Sipora and North and South Pagi. These four islands are all hilly, 

 and are separated from Sumatra by deep water; one sounding of 835 fms., 

 others of 640-350 fms., occurring on the charts published as Plate 7 of the atlas 

 of the Valdivia report. North of Siberut the water is very shallow. According 

 to the Dutch Hydrographic Bureau charts, a ridge extending to Tanahbala is 

 covered by from 6 to 26 fms. only; the Valdivia chart, however, shows one sound- 

 ing of 150 fms. Inside this ridge towards Sumatra there is a sounding of 417 

 fms., and outside, one of 206 fms. Tanahbala, Tanahmasa, and Pini Islands, 

 the largest of the Batu Archipelago, are low-lying, and are separated from each 

 other and from Sumatra by water only 12-26 fms. deep. An earlier connec- 

 tion, then, of the Mentawei Islands with one another is rendered highly 

 probable by the hydrographic data. 



We know the herpetology of but one of these islands, Sipora (also called 

 Seremu). In 1894 Boulenger (Ann. Mus. civ. Gen., 1894, ser. 2, 14, p. 613-618) 

 reported on a collection made by the Italian naturahst Modigliani. 



The same single fresh-water turtle which occurs on Nias is found also on 

 Sipora. The lizards are twelve in number; of these, two, Sphenomorphus 

 modigliani (Blgr.), and Leiolepisma vittigerum (Blgr.), have hitherto been found 

 only on this island. The fact that this island is perhaps more isolated from 

 Sumatra than is Nias may account for these two autogenous species of lizards, 

 while Nias has none. It is very possible, however, tliat they will ultimately be 

 foimd on Nias and Sumatra. 



One Uzard, Gonatodes kandianus, has a most strange distribution. It is 

 found in India and Ceylon and in all the Sumatran coast islands as yet herpeto- 

 logically explored, — Nias, Sipora, Engano, — but it has not as yet been found 

 in Sumatra. This form may or may not exist in Sumatra; it probably does, 

 though as yet imknown. 



