THE COMMON INDIAN SNAKES. 245 



about PeshaAvar, viz., " chamcha-mar " (spoonsnake). The compari- 

 son to a flattened spoon is qnite a good one, and has been noticed by 

 natives in other parts, "mywe howk,"' the nsnal name for the 

 cobra in Bnrma implies "hissing snake." In the Chin Hills, 

 Venning* tells ns, the vernacular name is " tlna-kan. " Flowerf 

 writing of Siamese snakes says, the cobra in Siam is called " ngu 

 how.'' The same authority mentions "toodong sta" as the name 

 for it among the Malays at Kedah. Cantor, however, gives the Malaj?- 

 name for it as " ular mata -ari '"' (sun snake), and Annandale and 

 Robinson;]: "ular tedong sendok kunyit" meaning " turmeric ladle 

 hood snake." 



Varieties. — I think it wisest in this paper to adhere to the arrange- 

 ment of the varieties contained in Mr. Boulenger's catalogue (Vol. 

 Ill, p. 380). Though my notes are voluminous and I have examined 

 many hundreds of cobras I do not consider the material I haye 

 worked upon adequate to justif}^ the introduction of another sj'stem 

 3^et. I have appealed through this journal to our readers to assist 

 our Society in procuring specimens from ever}^ part of India, and this 

 may be possible hereafter^!. 



* Bomb. N. H. Jourl., Vol., XX, p. 335. 



t P. Z. S. 1899, p. 690. 



% Fasci., Malay, 1903, p. 168. 



•^ From the available literature at my disposal, and the much larger materia 

 contained in my note-books I reah'se that the classification of the cobra {N. tripudi- 

 ans) is a very complicated matter. Stejneger (Herp., Japan, p. 394) suggests that a 

 method based on lepidosis, i.e., scale rows, ventrals and subcaudals is more scientific, 

 and might be productive of a better grouping of the varieties than the scheme laid 

 down by Boulenger which gives primary attention to colour, and hood markings. 

 My own views accord completely with those of Stejneger, and I think that most 

 herpetologists would unhesitatingly allow that lepidosis is of far greater importance 

 than coloration. A scheme based primarily on lepidosis however does not pan out 

 as satisfactorily as Stejneger anticipated, still complicated though it remains it is 

 less so than Boulenger's scheme, and it groups the specimens much more satis- 

 factorily in accordance with geographical areas. 



I find that m many localities the variations in colour and hood marks are very 

 numerous, whereas there is a very close agreement in scale characters. According 

 to Boulenger's method a specimen from South India (where the scale rows in 

 midbody are invariably 23 to 2.5) which has no mark on the hood, is grouped with 

 a specimen from the Punjab without a hood mark, though in the Punjab all the 

 cobras have but 21 scales in midbody !. Into the same group falls the oxiana of 

 Eichwald, a variety with a range of ventrals and subcaudals distinct from the 

 cobras of the East of Asia, and with a distinct geographical distribution restricted to 

 the North-West«rn part of our Indian Empire, and beyond. An anocellate cobra 

 from Burma, Java, or even the Philippines falls into the same group, viz., variety B. 

 coeca. 



It seems to me this method is as unscientific as it is unsatisfactorj'. The fact 

 is the colour and hood marks of individuals in many localities are too varied to 

 permit of their being grouped satisfactorily on these characters. As an example, 

 take the cobra referred to usually as the black coeca. A typical specimen 

 is black, and bears a hood mark. In Fj^zabad, U. P., a uniform black 

 or blackish cobra was the commonest form. Many of these had a perfect 

 " spectacle " mark on the hood, many had a modified or disintegrated " spectacle," 

 some had one or more spots which however were placed in the situation usually 

 occupied by the " spectacle " and some rare individuals had no indication of even 

 a spot. The marks in many of these black specimens are obscure, often so much so 



