222 OPHIDIA. 



and only the following genera are found in British India. Green is very frequently the 

 ground-colour in this group. 



Scales in 14 or 16 series Zaocijs, p. 255. 



The last maxillary tooth is the longest, in a continuous series witli the others. 



Ventrals more than 200 in number, angulated Herpetoreas, p. 257. 



GROUP OF FRESHWATER COLUBRIDES— i\MTi?/6'/iA':4. 



These snakes are generally not elongate or compressed, and the number of their ventral 

 shields is considerably less than 200. All the Indian forms have keeled scales. They freely 

 enter the water in pursuit of their food, which consists chiefly of frogs and fishes. All the 

 snakes of the preceding groups overpower their prey by throwing some coils of the body 

 round or over it, and commence to swallow it only after it has been smothered or at least 

 exhausted ; but the Natricines swallow their prey immediately after they have seized it. 

 The genus Trojndoj^Ms leads off to the true Freshwater Snakes or Ho7nalqpsides, the Indian 

 species of which are distinguished by their grooved posterior tooth. 



Scales keeled ; ventrals considerably less than 200 in number, rounded ; two 



anterior frontals ; maxillary teeth lengthened behind Tropidonotus , p. 258. 



Scales keeled ; ventrals considerably less than 200 in number, rounded ; only 



one anterior frontal Atreimm, p. 272. 



Scales keeled, in 19 rows; ventrals considerably less than 200 in number, 

 rounded ; maxillary teeth widely set, the middle a little longer than the 

 others Xenochrophis, p. 273. 



Scales keeled, in 19 rows; ventrals considerably less than 200 in number, 

 rounded ; maxillary teeth smallest behind, gradually becoming longer 

 anteriorly Prijmnomiodon, p. 274. 



Appendix. 



Among the drawings sent by Mr. Hodgson from Nepal, is one of a snake, 10^ inches long, to which 

 he gave the name of Coluber sanguiventer. Cantor, who examined this drawing (which is 

 now in the British Museum), introduced the snake into his list of Indian Serpents, with the 

 name oi Hurriah sanguiventer (Proc. Zool. Soc. 1839, p. 52). The drawing as well as the 

 diagnosis are too imperfect to show what place in the system should be assigned to this 

 snake ; but, from the characters known, it would appear that it is the type of a distinct 

 genus. We add the diagnosis given by Cantor, and a figure of the upper side of the head, 

 copied from Mr. Hodgson's drawing, and apparently of thrice the natiu-al size : — " Claret- 

 purple above, with metallic lustre ; blood-coloured beneath. Ventrals 207 ; subcaudals 99, 

 the fourteen anterior of which are entire." Valley of Nepal. 



