574 MR. F. DAT ON THE GEOGKAPHIOAL 



mouth, if transverse, is anterior in tlie various genera ; but the 

 head is most commonly compressed, and the dorsal fin armed 

 with a strong, serrated bony ray. 



Having thus briefly shown the distribution of the Indian fresh- 

 water fishes, and traced out the countries from which they have 

 been derived, we come to the question. What are the most typical 

 families having representatives in India ? 



Those most extensively distributed are the Ophiocephalidae and 

 Symbranchidae, to which I have already alluded as strictly Ori- 

 ental forms possessing an amphibious respiration : next we have 

 the Labyrinthici, also Oriental, but with some African represen- 

 tatives which, however, do not extend to India ; the Oriental 

 genus Anahas extends from Assam to the eastwards ; OspJirome- 

 nus from N.E. Bengal and Assam, also to the east ; whilst some 

 have a more local range, as Trichogaster, from the Hindustan sub- 

 region to Burma. 



Doubtless the Siluridse and Cyprinidse are the forms most pre- 

 valent in the Indian fresh waters, the former being represented 

 by 26, the latter by 35 genera. I propose first to investigate the 

 Cyprinidse, as they appear to be of a more northern (if not Palse- 

 arctic) origin than the scaleless Siluroids. Amongst the Cypri- 

 nidse the first thing that deserves attention is the absence of ad- 

 ditional means of respiration to the gills, as we see in the Acan- 

 thopterygians (as in Ophiocephalidae and amongst the Laby- 

 rinthici), an addenda which is likewise seen amongst the Siluroids 

 in the genera Glarias and Saccobranclius, and in the Symbran- 

 ehidse in the genus Amphipnoiis. 



Seeing that out o£ these four large divisions of Indian freshwater 

 fishes, the Cyprinidse is the single one not possessing any species 

 favoured with an amphibious form of respiration, we come to the 

 consideration of what are the most typical genera of Indian Carps. 

 "We know of 226 species of Carps in India, 70 of which, or nearly 

 one third, belong to the genus Sarbus, a genus which has very 

 close afl&nities with several others. If the mouth were a little 

 more transverse, it would lead us to the CirrJiina ', possessing 5 



^ CirrMna lafia takes on various modifications in accordance with the loca- 

 lities it inhabits. In hill-streams it is seen as if it attached itself by the lower 

 surface of its head to stones, as we find occurs in JDiscognatlms, and the rudiment 

 of a pad may be observed behind the lower lip. The passage of this form into 

 Biscoqnathus, having about the same number of rays and scales, would not appear 

 to be very difficult. There is likewise another curious structural change which 



