242 W. T. Blanford — Description of a new Varanus, with, Sfc. [No. 4, 



The number of ventrals in the three specimens is 203, 205, and 207, there 

 are 72 pairs of subcaudals in two and 74 in the third, the smallest indivi- 

 dual, in which the number of ventral shields is 207. It is manifest tbit 

 this variety is more elongate than the forms usually found in India and 

 the countries to the eastward. All of the varieties of cobra described 

 in Gunther's ' .Reptiles of British India' have less than 200 ventrals. The 

 colouration too is peculiar, especially in the young specimen, and remark- 

 ably similar to that in the Central Asiatic type described by Eichwald 

 under the name of Tomyris oxiana* This snake has since been shewn by 

 Strauch,f who re-examined the original specimen, to be a true Naja closely 

 allied to N. tripudians, but distinguished by the form of the head and the 

 characters of the temporal shields. JV. oxiana is founded on a young 

 specimen, the only one hitherto recorded, 40 centimetres (15f inches) long, 

 and having 202 ventrals and 66 pairs of subcaudals, so that it resembles 

 the Gilgit snake in the great number of the former. The essential differ- 

 ences, however, are considered by Strauch to consist in the facts that in 

 JV. oxiana the length of the head is thrice the height and nearly twice 

 the breadth in the temporal region (whereas in JV. tripudians the length is 

 only twice the height and one and a half times the breadth), and that 

 there are two larger temporals in front in contact with the postoculars, 

 the posterior temporals being numerous and scarcely larger than the back 

 scales, whilst in the Indian cobra the temporals consist of 4 or 5 subequal 

 shields. 



Now in the Gilgit snakes the temporals vary in form. The two anterior 

 are generally the largest, but the lower posterior temporal is nearly, some- 

 times quite, as large. The anterior lower temporal is in contact with 3 

 others. In each of the larger specimens 3 temporals are in contact with 

 each occipital, in the smaller 4. I find precisely similar temporal shields 

 in many Indian cobras in the British Museum. The head too in the Gilgit 

 snakes is precisely similar in form to that of N. tripudians. 



If, therefore, Strauch is correct in considering the distinctions he has 

 pointed out in the type of JV". oxiana of specific value, it is clear that the 

 Gilgit snakes cannot be referred to that species. The temporals are so 

 variable in form and number in cobras that, were there no other distinc- 

 tion, I should doubt whether the Central Asiatic form really deserved separa- 

 tion, but the shape of the head is a more important character. At the 

 same time it is very interesting to find in Gilgit, where several birds and 

 mammals belonging to Central Asiatic types occur, a cobra presenting so re- 

 markable a similarity in colour and structure to the form described from 

 Central Asia. 



* Zool. Spec, III, p. 171 : Fauna Casp. Cauc. p. 130, pi. xx, 

 t Bull. Acad. Sci. St. Pet. 1868, xiii, p. 81. 



