585 



A ROUGH KEY TO THE IDENTIFICATION OF 



INDIAN OPHIDIA. 



By a. G. Cardew, i. c. s. 



{Read lefore the Bomhay Natural History Society on 



September 22, 1896.) 



A word of apology is necessary in laying before the readers of this 

 Journal the rough key to the identification of Indian snakes which is 

 printed below. It will be obvious to everyone that the key is purely 

 artificialj and brings into juxtaposition snakes' belonging to widely 

 separate families. The key, it will be seen, does not pretend to be 

 scientific, but is merely an attempt at such an arrangement as may en- 

 able unskilled observers to identify specimens of Ophidia by external 

 characters. If reference is made to the^nopsis of families which is 

 given by Mr. Boulenger on page 234 of his work on '' Repfcilia and 

 Batrachia" in the India Office series of manuals on the Fauna of 

 British India, it will be observed that-the families are there distinguished 

 by criteria drawn mainly from the bony structure of the head. In order 

 to apply this synopsis to the identification of snakes, it is necessary to 

 acquire a larger knowledge of the skeleton of a snake than is common, 

 and it therefore seems possible that an artificial key, based on characters 

 which any one can see and understand, may sometimes be useful. 



The first principle of division which has been adopted in arrangincy 

 the snakes for the purposes of the key is the number of scales round 

 the body. On page 281 of Mr. Boulenger's work will be found a dia- 

 gram showing the scaling on Dendrophis pictus, and from this will be 

 apparent the manner in which the scales should be counted. The first 

 step in applying the key to the identification of an unknown snake 

 would, therefore, be to count the number of scales round the body. 

 As soon as this has been ascertained, it will be found that, whereas 

 before there were some 260 species to anyone of which the snake 

 might belong, the range of search has now been narrowed to 30 or 10 

 species. The next step is to count the number of subcaudals and ventrals. 

 This is easily done, and with the aid of these two characters combined, 

 the area within which the snake must occur is usually made much smal- 

 ler. If the snake were Dendrophis j^ictus, for example, it would be afe 

 once apparent that the snake must be looked for in one of the genera 

 Dryophisj Dendrophis- or Zamenis, The third test to be applied is the 



