314 SNAKE FAUNA OF SOUTH AFRICA. 



be judged from their external characters, these genera do not 

 form any characteristically African association, but Dendraspis 

 has the rank of a distinct section. 



VIPERINAE. 



South Africa has the following genera: Causus, Bitis, Atrac- 

 taspis and Viper a, the first three being confined to tropical and 

 South Africa, though Bitis arietans extends to Morocco and 

 Arabia. Causus and Atractaspis are each isolated and peculiar 

 genera, whilst Bitis is related to Vipera, whose headquarters are 

 in Europe, with an odd species in Mozambique and another in 

 India and Ceylon. The genera Bitis and Vipera belong to a sec- 

 tion of Vipers which includes also Pseudocerastes (of Persia), 

 Cerastes (North Africa, Palestine and Arabia), Echis (one species 

 common to Northern Africa, Arabia and India, the other species 

 occurring in Palestine and Arabia) and Athens (tropical Africa). 



Dr. Gadow has suggested that the Viperidae do not constitute 

 a natural group, and certainly the South African genera, super- 

 ficially at any rate, seem so different that it may well be that the 

 Viperine character is a result of convergent evolution. Judging 

 from the distribution of the more typical Viperine genera, Bitis 

 came from the North ; on the other hand the peculiar genera 

 Causus and Atractaspis may have developed in Africa from some 

 local opisthoglyphous snakes, but this speculation requires con- 

 firmation from anatomists. 



CONCLUSIONS. 



We are not able to fully interpret the facts of present day 

 distribution owing to our scanty knowledge of phylo-genetic rela- 

 tionship. From the facts now available there could be little 

 reason to suspect that the ophidian fauna of South Africa is of 

 dual origin, for the more specialised sections of the Colubridae 

 seem to be for the most part just as peculiar to the area as are 

 the more generalised sections, and no doubt the final differentia- 

 tion of such genera as Sepedon, Aspidelaps and Dendraspis has 

 taken place in Southern Africa ; but to what extent the ancestral 

 viperine and proteroglyphous stock is of external origin or of local 

 ancestry must remain a matter for future discovery. However, 

 as there appears to be some indications amongst the Colubridae 

 of an affinity with the Madagascar fauna, it is easv to see that 

 the known facts are in general harmony with those relating to the 

 lizard faunas of South Africa and Madagascar ; premising which, 

 the fact that both areas are inhabited by numerous snakes, which 

 we are obliged to refer to so many peculiar and distinct genera 

 merely proves that the differentiation of these Colubrine genera 

 has taken place subsequently to the separation of Madagascar 

 from Africa. 



Again, the genera Glauconia and Leptodira furnish good 

 evidence in favour of a direct land connection between S. America 

 and Africa since the period of origin of those genera, that is to 

 say probably during early tertiary times. Indeed, in view of the 



