Element in the Fauna of India. 281 



over, on several isolated hill groups in Southern India, the 

 number of representative forms apparently diminishing gra- 

 dually to the northward. The best-known of these groups is 

 that of the Shevroy hills, near Salem. The plains of the 

 Carnatic from the Krishna (Kistna) river to Cape Coniorin 

 are included in this region by Mr. Wallace ; but in this he is 

 certainly in error; and he has, I think, been misled by incorrect 

 localities for some typical forms, such as the Uropeltidaj *. 



4. The Eastern-Bengal province. This is limited on the 

 west by a line drawn northwards from the head of the Bay 

 of Bengal. Calcutta is just on the edge, and perhaps rather 

 within than without it. It belongs to Mr. Wallace's Indo- 

 Chinese subregion, the limit of which I should be inclined to 

 draw a little further to the westward than he does. This, 

 however, is a trifling detail. 



I further subdivided the Indian province into subprovinces, 

 as below : — 



a. Gangetic subprovince or Hindustan f, extending south 

 as far as the Nerbudda, in its eastern portion comprising 

 only the valley of the Son and the Gangetic plain as far east 

 as Benares. 



h. Deccan subprovince — from the Nerbudda to the Krishna, 

 bounded on the west by a line drawn a little east of the crest 

 of the Western Ghats or Syahddri range, and on the east by a 

 line drawn nearly north and south a little east of Nagpur. 



c. Bengal subprovince — bounded by the last on the west, 

 and extending as far south as the Godilvari. 



d. Madras subprovince — all the peninsula south of the 

 Krishna river and to the eastward south of the God^vari, 

 and east of the Nilgiri and other hills belonging to the range 

 of the Western Ghats. The upper portions of some small 

 isolated hill-ranges, however, such as the Shevroys and 

 Kolamullies, have a Malabar fauna. This Madras sub- 

 province also includes Northern Ceylon. 



My provinces correspond to Mr. Wallace's subregions. The 

 accompanying small map (p. 282) shows the approximate limits 

 of the provinces and subprovinces. It is as well, since I have 

 evidently been misunderstood, to say that the subdivision 

 proposed refers solely to the Indian peninsula. The Eastern- 



* This family of snakes is entirely confined to the province or sub- 

 region I have defined. Colonel Beddome, by far the best authority on 

 South-Indian Reptilia, has pointed out that no Uropeltidse are ever found 

 in the plains of the INIadras Presidency, except on the west coast. 



t I pointed out that by natives of India this term is applied to the 

 Upper Gangetic plain, and not, as it is by European geographers, to the 

 whole of India. 



