1908.J OF SNAKES OF TUE GEXUS COHALLUS. 143 



differences between the two species of Corallus with which I am 

 concerned in the present communication ; and the difference is 

 rather remarkable in that it is closely paralleled among the true 

 Pythons, as I shall point out immediately. In Corcdlus cookii 

 the aorta gives off comparatively few intercostal branches. Yet 

 the intervertebral spaces are served by a regularly paired series 

 of arteries as in the Boidse generally. The arrangement of these 

 ai-teries is, in fact, like that of the Boine (as contrasted with the 

 Pythonine) snakes, in that the few branches arising from the 

 aorta and passing dorsally are connected together by a series of 

 secondary longitudinal trunks; from these latter arise, at the 

 requisite regular intervals, the actual branches which bury them- 

 selves in the intercostal spaces. This is precisely the arrangement 

 that would have been expected in a snake belonging to the sub- 

 family BoincB of the family Boidse according to the facts which I 

 have been able to record in various genera of that family. It is, 

 in fact, Corcdlus madagascariensis which is thus far " abnormal, "^ 

 for, as I have already recorded, in that snake the disposition of 

 the intercostal arteries is not after the Boine but the Pythonine 

 plan. This rather unexpected state of affairs to a certain extent 

 of course throws doubt upon the value of these characters from a 

 classificatoiy point of view. I have laid some stress in foi^mer 

 papers upon the value of the intercostal arteries in the classi- 

 fication of serpents; but it is at least noteworthy that the 

 difference in structure goes hand in hand with a totally different 

 range in space. The Old World Corcdlus madagascariensis agrees 

 with the Old World Pythons, while the New World Corcdlus ooohii 

 agrees with the New World Boas. The parallel instance to which 

 I have referred above is of course Python spilotes ; this snake, 

 while agreeing with Pythons generally in a number of features 

 of its organisation, has a system of intercostal arteries which is 

 not like that of the Pythons generally, but agrees with that of 

 the Boince. Here, again, the difference goes hand in hand with a 

 difference in habitat ; for Python sjnlotes belongs to the Australian 

 region, while all the other Pythons that have been examined 

 from this point of view are either African or Asiatic in rauge. 

 It would be of the greatest interest to study some of the other 

 members of the genus Python which occur in or near to the 

 Australian region, and also the other American members of 

 the genus Corallus. In the meantime I can only point out these 

 rather anomalous facts. I may also mention that in those 

 individuals where I looked for this artery each kidney had only 

 one renal artery, as in Boidse generally. 



§ Venous System. 



With reference to the venous system of Corallus I have obsei- 

 vations to make only concerning the umbilical and azygos veins. 



The umhiliccd vein is, so far as I know from my own observa- 

 tions, small and rudimentary in the Pythons, but well developed 



Peoc. Zool. Soc— 1908, No. X. 10 



