1906.] KESPIRATORY SYSTEMS IN THE OPHIDIA. 531 



that Lioheterodon and Heterodon are distinct. The lung con- 

 firms Boulenger's view of the separateness of these genera. I have 

 already referred to the long headwai'd extension of the lung in 

 Heterodon, which nearly reaches the head. In LioJieterodon 

 Tnadagascarwnsis this diverticulum of lung is only | of an inch in 

 length, and is anangious at its free extremity. Moreover, in 

 Heterodon, the interannular membrane is very narrow ; in Lio- 

 heterodon on the other hand, as I have already partly indicated 

 in another communication upon the Ophidia *, the membrane in 

 question is much wider than are the tracheal semirings. Near to 

 the heart are two folds across the membrane which result in the 

 formation of a pouch. This is not unsuggestive of the pouches 

 in the Hamadryad Snake t. In any case the differences in lung- 

 structure between the two genera will be obvious. 



§ Resume of Principal Facts. 



From the foregoing account of new facts in the structure of the 

 Ophidia I may select the following as embodying the principal 

 results : — 



(1) The umbilical vein largely persists in Boa, as in Eunectes, 

 as a functional vessel. In Python there are less conspicuous 

 traces of it ; but it is distinctly recognisable in P. molurus and 

 P. regius as well a.s in P. sehm. ISTo blood-holding diverticulum 

 of the post-caval has been yet met with in Oolubrine Snakes, the 

 only vestige of the embryonic umbilical being a membranous seam 

 on the liver of Coluber corais. 



(2) The additional facts recorded in the present communication 

 support the probability that a forward prolongation of the renal 

 afferent vein to the region of the gonads, which is to be regarded 

 as part of a persistent post-cardinal, is characteristic of the Boidse 

 and at least not characteristic (though rudiments may occur) of 

 the Oolubrine Snakes. So far the Boidse lie at a lower level than 

 other Snakes. 



(3) In some, but not in all, Boidse the parietal tributaries of 

 the renal afferent vein are connected at their emergence from the 

 body- wall by a continuous longitudinal trunk which extends from 

 behind to in front of the kidney. In the region of the kidney 

 this longitudinal vessel gives off no branches to the kidney itself. 

 In the Oolubrine Snakes, on the contrary, there is no such longi- 

 tudinal dorsal parietal vessel ; but intercostal veins may eraerge 

 from the parietes in the region of the kidney to cross it and enter 

 the afferent renal vein. But there is no direct connection of 

 these intercostals with the capillary network of the kidney such 

 as has been found to occur in the snake-like Lizards Ophisaurus 

 and Amphishcena. 



(4) The generality of the occurrence in the Boidse of but a 



* P. Z. S. 1906, vol. i. p. 12. 



t Beddard, P. Z. S. 1903, vol. ii. p. 319. 



