190r.] OF CERTAIN SPECIES OF SQUAMATA. 51 



a paper communicated to this Society, It is impossible to be 

 certain of the exact position of the actual umbilicus in Lachesis 

 lanceolatus for the purposes of comparison between these two 

 types, i. e., which of the four or five broken scales correspond to 

 the two scales in Bitis nasicornis which are actually divided by 

 the foetal blood-vessels. Assuming, however, that they ai'e even 

 the last two, there still remains a substantial difference in position 

 between the umbilicus of the two species. For in Bitis nasicornis 

 the actual numbers of scales intervening between the umbilicus 

 and the anal scale are respectively in the five examples stvxdied 9, 

 11, 11, 12, 14. There is thus exactly the same amount of variation 

 as in Lachesis lanceolatus, but round a different mean. 



The position of the umbilicus in Russell's Viper {Vipera 

 russellii) is again difl'erent and relatively more fixed than in either 

 of the species hitherto considered. As in Bitis nasicornis, the 

 actual umbilicus consists of two scales only, which do not meet 

 ventrally, and between which the plug of tissue bearing the 

 umbilical vein &c. passes into the interior of the body. The 

 young Vipers in question were a very few days old, but all 

 external traces of the yolk-sac had disappeared. In front of 

 these two scales either two or three scales were divided by a suture 

 in the middle ventral line, and posteriorly to the two " umbilical 

 scales " either one or two scales were similarly split by a ventral 

 sutvire. Between the last of the two completely divided scales 

 and the anal scale there intervene in the five examples examined 

 respectively 16, 16, 16, 17, 17 scales. The position of the umbi- 

 licus is therefore different in this Viper, and its fluctuations of 

 position are less than in the two species to which I have already 

 called attention. It is perhaps permissible to call attention to 

 the fact that Vijyera and Lachesis agree with one another more 

 nearly than either does with Bitis. This is, of course, not in 

 accord with generally received views upon the classification of 

 Vipers. 



I have examined several Vipers of more mature age, and in 

 two specimens, at any rate, I find what appear to be obvious traces 

 of the umbilicus. In a not fully-grown example of Bitis arietans 

 measuring 30 inches from the snout to the cloaca, four scales 

 showed a line of division in the ventral median line. The second 

 of these had the most strongly marked groove, and possibly 

 therefore represents one of the two scales already described in 

 the young as immediately surrounding the stalk of the yolk-sac. 

 Between the last of the grooved scales and the anal scale 9 scales 

 intervened. The species evidently therefore comes nearest to the 

 other species of Bitis which has been described above. In a second 

 specimen measuring 32 inches there were 12 scales between the 

 last of four grooved scales and the anal scale. I have also seen 

 similar traces in a large adult example of Bitis gahonica. Here 

 there were also four scales showing traces of the umbilicus ; but 

 instead of being grooved they were merely nicked posteriorly. 

 Between the last of them and the anal scale 8 scales intervened 



