20 Nelson Annaudale — The Lizards oj the Andama^is, [Supplt. 



in the Malay Archipelago, but has not been recorded from the Penin- 

 sula ; 0. ophiomachus is only known from Ceylon and S. India. Several 

 Skitiks probably occur in the Nicobars which are absent from the Anda- 

 maris, and the same may be true of Dibamiis novse-guinese — the sole 

 representative of a family closely allied to the Skinks and once 

 regarded as peculiar to the Nicobars and Papuasia, but now known to 

 exist both in the Malay Peninsula and in several of the islands of the 

 Malay Archipelago. 



What is perhaps a true relationship between the Andamans and 

 Cejlon consists in tlie absence from both of the genus Draco, which 

 occurs within the Malabar province in S. India (also in Malaya, Burma 

 and Assam) and consists of forms too striking to escape notice readily. 



Prain (12, 13) has shown that the flora of the Cocos group and that 

 of Narcondam differ considerably from that of tlie southern Andamans. 

 The geological separation between the different islands, and especially 

 between Narcondam and the rest of the archipelago, is well illustrated 

 by the marine depths marked on the maps recently published by Alcock 

 (13) and Kloss (14). Narcondam is distinguished zoologically by the 

 possession of an isolated Hornbill ( Rhylidoceros narcondami ). I have 

 very little information about the lizards of the Cocos, except that Calotes 

 versicolor and Gonyocephalus subcristatus occur ; on Narcondam Major 

 Anderson has taken Gymnodactylus rubidus, Gonatodes andersonii and 

 Phelsuma andamanense, as well as Mabuia multifasciata and another 

 Skink. The three Geckos characteristic of the Andamans therefore 

 occur on this island. Of three specimens of G rubidus irom. Narcondam 

 in the Museum, two are considerably larger than any in a large series 

 from other parts of the Andaman archipelago ; otherwise no difference 

 can be detected. The only adult specimen of G. andersonii,^ is from Nar- 

 condam, but, except in point of size, it agrees with two young individuals 

 collected by Wood-Mason somewhere in the Andamans (exact locality 

 not specified) and confused by him with the young of G. o'ubidus, to 

 which they bear a close superficial resemblance. Specimens of P. 

 andamanense agree in every respect with those from the other islands. 



I do not propose to generalize as to geography on the basis of the 

 lizards. The study of a single sub-order somewhat poorly represented 

 ( or perhaps rather imperfectly known) does not permit wide generaliza- 

 tions as to the whole fauna, much less the geology and geography, 

 even of a small group of islands. It has been my object to show, 

 in the first place, that the vertebrates of the Andamans are not devoid 



i Since this was written two more liave been received from Narcondam, col- 

 ieottod and presented by Mr. 0. G, Rogers. 



