500 MR. F. E, BEDDARD ON THE VASCULAR AND [May 1 , 



bifurcate as in Python, &c., and thus supply both sides of the 

 vertebral column. Interspersed among these are a few arteries 

 which, as in Colubrines generally, plunge into the dorsal parietes 

 on one side or the other of the middle line. There is in these 

 arteries no question of a bifurcation. In this anterior section 

 of the body from the junction of the two aoi-tse I counted 16 

 bifurcated intercostals, and only three which were single arteries 

 throughout supplying only one side of the body. It must 

 be pointed out, however, that there is not here, as there is 

 in Python, an artery to each intercostal space. There are 

 interspaces of several vertebrae between consecutive arteries. 

 Rarely there are arteries following each other immediately. For 

 a considerable region of the body, beginning towards the end of 

 the liver, the intercostals are mostly single trunks, and therefore 

 entering the body- wall to the right or to the left of the dorsal 

 median line as the case may be. Further back the arteries again 

 become prevalently double. It is clear therefore that there 

 are some grounds for comparing the intercostal arteries of this 

 genus with the Pythons on the one hand and with the Colubrines 

 on the other. The irregularity of those arteries in the Colubrines 

 generally (though it must be remembered that after all our 

 knowledge is at present very deficient) is shown in Erythrolamprus, 

 and coupled with this the bifurcation in the middle line befo:6"e 

 entering the body-wall of some of those arteries, which is a 

 Pythonine characteristic. We may perhaps also see in this latter 

 character a point of likeness to the Viperidfe. In these Snakes 

 there is up to the present no exception to the rule that the inter- 

 costal arteries arise irregularly, but enter the middle line of the 

 dorsal parietes instead of to the right or to the left as in the Colu- 

 brines. The division of these vessels therefore takes place within 

 the thickness of the parietes, instead of outside as in Python and 

 its allies. It seems therefore that, starting from the conditions 

 observable in the Boid^ — and there is now much evidence for the 

 reasonableness of the assumption that this family lies nearest to 

 the base of the Ophidian series — we can trace the modifications of 

 the intercostal arteries in at any rate two directions. The usual 

 Colubrine arrangement may be derived, as I have already 

 suggested *, by an obliteration, now on one side and now on the 

 other, of one of each of the paired intercostals, the usual gaps being 

 already indicated among the Boids by the secondary longitudinal 

 intercostal trunks which are connected only at intervals with the 

 aorta. The second path of development is completed in the 

 Vipers, where in one way the Boid arrangement may be looked 

 upon as more obviously preserved. It appears to me that 

 Erythrolamprus may be looked upon as a stage in this metamor- 

 phosis. The Boid character has been largely retained and the 

 Colubrine character correspondingly feebly developed. The dis- 

 appearance of the latter and a slight change (already referi-ed to) 

 in the former would give the Yiperme character. It is iiote- 

 * P. Z. S. 1904, vol. ii. p. 108. 



