586 pPHIDIA CHAi'. 



remains belong to recent genera. There are indications that the 

 Ophidiu are a relatively young branch of Reptilia, essentially of 

 Tertiary date, but the foundations of the distribution of most 

 of the older families were laid in Miocene times. The older 

 families, notably those which still possess vestiges of hind-limbs 

 or of the pelvis, are circumtropical, e.g. Typhlopidae, Boidae. 

 The few survivors of the Glauconiidae are likewise circum- 

 tropical, with the exception of Australia. The Ilysiidae occur 

 in South-Eastern Asia and in tropical South America; their 

 offshoot the Uropeltidae are restricted to India and Ceylon. 

 The Colubridae and even many of their sub-families are cosmo- 

 politan. It is quite possible that the Opisthoglypha and Pro- 

 teroglypha are not natural groups, but that their respective 

 conditions have been developed on various occasions and in 

 different countries. The same applies more strongly to the 

 Viperidae, a further development of the Opisthoglyphous type. 

 To judge from their distribution, the Crotaline snakes were 

 possibly developed in tlie Palaearctic sub-region ; they spread all 

 over America, but they were debarred from entering either 

 Australia or Africa. The Viperidae, on the other hand, are 

 restricted entirely to the Palaeotropical region and to the 

 Palaearctic sub-region. The fact that no separating belt of 

 water existed for them between Europe and Africa, indicates 

 their being the most recently developed of poisonous snakes. 

 Madagascar is the only large country which, besides snakeless 

 New Zealand, enjoys a total absence of poisonous snakes of any 

 kind, while the Oriental is the only sub-region whicli suffers 

 from the presence of numerous species of every sub-family of 

 poisonous Elapine, Crotaline, and Viperine snakes. 



SnaJce-Poison. — Many Snakes, belonging to different families, 

 are poisonous, and unfortunately there is no external character, 

 easily ascertained, by which every poisonous snake can be 

 distinguished from a harmless kind. If the head is very broad, 

 tliis is probably due to the pair of poison-glands on the sides 

 of the head ; but many harmless snakes can flatten and broaden 

 their heads in a suspicious way, and, what is much worse, many 

 of the most poisonous snakes, for instance the Cobras, have a 

 head as smooth and as sleek -looking as the Grass- or King- 

 Snake, the most harmless of species. It so happens that, with 

 a few exceptions, for instance among the Crotalines and Vipers, no 



