SNAKE FAUNA OF SOUTH AFRICA- 309! 



is characteristic of the East Indies, extending to Australia and 

 Papuasia, but does not occur in Africa nor in America. So far as 

 I know the question of the genetic relationship of this snake has 

 not been investigated by morphologists, and, as it is not unlikely 

 that some of the more specialised groups of Ophidia present cases 

 of convergent evolution, it would be unwise to lay great stress on 

 the fact of the existence of a solitary Homalopsine in Madagascar ;: 

 in the absence of more complete data the most obvious explana- 

 tion would represent this snake as a relic introduced at the time 

 of a land connection with the Indian region. 



The Ophidian fauna of South Africa is considerablv richer than 

 that of Madagascar, for, in addition to the four groups of more 

 primitive snakes, our region has the following : — Glauconidae, 

 Rhachiodontinae, Elapinae and Viperinae. Now, the two latter 

 families comprise the very specialised poisonous snakes, which 

 no doubt are comparatively recent developments from more 

 generalised harmless snakes ; the rhachiodont Dasypeltis is derived 

 from a simpler Colubrine type, and the Glauconias alone are more 

 or less primitive ; that is to say (the Glauconias excepted) that 

 portion of the South African Ophidian fauna which has no allies 

 in Madagascar represents the last stage of evolution, and if it can 

 be shown that the four generalised cosmopolitan families in South 

 Africa have sufficiently intimate relationship with the same fami- 

 lies in Madagascar, no one will doubt but that these generalised 

 families constituted the fauna of the island continent which united 

 Southern Africa and Madagascar, and that the other South 

 African families come from the outside or less probably were~ 

 developed in situ subsequent to the separation of Madagascar. 



GLAUCONIDAE. 



The distribution of the genus Glauconia is more restricted than 

 might have been anticipated from consideration of its primitive 

 nature. At the present day it occurs in Central and South 

 America, in Martinique and Barbadoes and throughout the whole 

 of Africa, an odd species extending as far as Sind (British Museum 

 Catalogue). 



This distribution agrees fairly well with that of the Amphis- 

 baenidae, and these two groups offer the only certain instances 

 (excluding the genus Leptodira and one or two almost ubiquitous 

 genera of geckos and skinks) amongst South African snakes and 

 lizards of a decided American relationship. West Africa, how- 

 ever, affords additional evidence, and it is commonly accepted by 

 palaeontologists and zoologists that there must have been a land 

 connection between Africa and South America during a late 

 secondary and early tertiary period.* 



At that period the connection with America was no doubt solely 

 through West Africa. 



*See Boulenger on distribution of Characinid and Cichlid fish in 

 Report British Association, 1Q05, and Schonland in Trans. S.A. Phil. 

 Soc, 18, p. 321, where the case is very clearly stated from the botanical 

 side. 



