SNAKE FAUNA OF SOUTH AFRICA. 3II 



Possibly palaeontology will eventually furnish the solution, as 

 in the case of the tortoise Podocnemis which to-day is known only 

 in South America and Madagascar, but occurs also in the eocene 

 beds of Fayoum (Egypt) and in the lower eocene of England and of 

 India ;* and the gap in the distribution of the Iguanidae may 

 perhaps be filled up by the Paliguana of Dr. Broom, t In this 

 connection it may be mentioned that a very large snake, Gigan- 

 topis, related to Python, is described by Andrews from the Eocene 

 of Fayoum, and there is a pythonid genus in the upper Eocene 

 and lower Miocene of Europe, but the relationship of these to 

 the present-day genera of Boidae is unknown. 



But to return to the Boinae. According to Beddard, who 

 has made extended and valuable investigations on the 

 anatomy of reptiles, there is reason for doubting the 

 commonly-accepted belief in the close affinity of South Ameri- 

 can and Madagascar Boidae ; after comparing the internal struc- 

 ture of Corallus Madagascariensis with the American species of 

 Corallus he suggests that they should be generically separated 

 (P. Z. S. 1908, p. 135), and in fact it appears that the former 

 snake is more in agreement with the pythons than with the boinae 

 in respect to its arterial arrangement. And Mr. Beddard now 

 thinks that in view of facts accumulated since the division of the 

 Boidae into its two sub-families it is not so desirable for the pre- 

 sent to insist upon any such sub-division. More recently (P. Z. S. 

 1909, p. 918), on comparing several American species of Boa with 

 the Madagascar species {Pelophilus Madagascariensis), he finds 

 important anatomical differences, whilst there are close points of 

 agreement in the arterial arrangementof the Madagascar genera 

 Corallus and Pelophilus. In view of these conflicting facts it is 

 obvious that the evidence of the Boidae on questions of geogra- 

 phical distribution must be reserved until we know more about the 

 inter-relationship of the various genera. 



COLUBRINAE. 



This sub-family is represented in Madagascar by ten genera, 

 of which four are monotypic and nine peculiar to the Malagasy 

 region ; the only genus also found elsewhere is Polyodontophis, 

 which occurs in the Indian and Malayan provinces and in Central 

 America. This genus appears to be one of the most primitive of 

 Colubrine genera, and on this account no special sig- 

 nificance should be attached to its extended distribution, 

 though its absence from Africa is curious. Another genus 

 Liopholidophis is closely related to, or perhaps identical 

 with, Tropidonotus. South Africa has ii genera, of which lo 

 (Ahlabophis; Lamprophis. Boodon, Lycophidium, Simocephalus, 

 Pseudaspis, Chlorophis. Philothamnus, Prosymna, Homalosoma), 

 are peculiar to tropical and South Africa ; Tropidonotus laevis- 

 sitnus (Grayia luhrica. Sclater) is the only representative in South 

 Africa of its genus, which occurs in all the continents. 



* C. W. Andrew's in " Tertiary Vertebrata of Fayum." 

 + Albany Museum Records, vol. I., p.l. 



