22 DR. J. ANDERSON ON MAMMALS, REPTILES, [Jan. 5, 



I have also given tvpo drawings of its head (Plate I. figs. 6, 7). 

 If these are compared with Guichenot's figure of V. mauritaniea\ 

 which Mr. Boulenger informs me is a good representation of the 

 above-mentioned specimen No. 4017, the differences between this 

 variety and the typical form will be seen to be considerable. 



To summarize the features of this variety, the most striking is 

 certainly the entire absence of any approach to keeling on the scales 

 on the upper surface of the head from the parietal region forward to 

 the rostral. As already stated, in all the hitherto recorded speci- 

 mens of V. lebetina, the scales of the head over the foregoing area 

 and over the upper surface of the head generally are keeled in various 

 degrees of intensity, while at the same time they preserve to a great 

 extent the general form and character of the scales on the body. In 

 this variety, on the other hand, the head-scales, besides being 

 perfectly smooth in the region specified, are somewhat unhke those 

 of the body in that they are rounded at their margins and are very 

 fiat. The entire absence of the canthus rostralis is another feature 

 by which this variety is distinguished from the typical form, in which 

 it is always defined although less marked in some individuals than 

 in others. 



The rostral shield is higher than broad, but in an example of this 

 species from the Lake of Galilee it is as high as broad, a modifica- 

 tion which connects this variety with the other specimens mentioned 

 in the foregoing table in which the rostral is decidedly broader than 

 high. The supraoculars in the first six specimens enumerated in the 

 table (p. 21) are not markedly differentiated from the ordinary head- 

 scales as regards their size : occasionally one or more scales may be 

 larger than the' others, either on both sides of the head or on one only, 

 but in none of them are two scales so developed as in this variety. 

 On the other hand, the specimens from Persia and Afghanistan can be 

 distinguished at once by the presence of a large supraocular occupying 

 nearly the entire upper border of the eye. It will be observed that, 

 in the labials and the scales between the eyes, the variations among 

 these specimens are very insignificant. It is worthy of attention 

 that the four specimens from the mainland of South-western Asia 

 present a marked increase in the number of their ventrals as com- 

 pared with the typical form from Algeria and Cyprus, and that the 

 two groups are connected with each other by var. desertifrom. Eastern 

 Tunisia. 



In V. lebetina, var. deserti, the colour is pale yellowish brown 

 above, with very faint indications of the dorsal and lateral dark spots 

 distinctive of Algerian and Cyprian examples of the typical form ; 

 and the under surface is pale yellow, almost immaculate anteriorly, 

 the finely powdered aspect of the species being only feebly indicated 

 posteriorly. The specimen from thcj Lake of Galilee very much 

 resembles the var. deserti in colour and general appearance, but it 

 has keeled scales on the head and a canthus rostralis. 



I have selected the term deserti to designate this variety, because 

 reptiles received from Duirat, the locality from which this Viper 

 ^ Explor. Sc. Alg., Eept. pi. iii. 



