48 Ml-. C. E. Narayan Rao on 



name will be Garra for the species occurring in Baluchistan, 

 India, Burma, Malayan Peninsula, and perhaps Borneo, 

 while Biscognathus, if it is established to be generically 

 distinct from Garra, may be confined to species met with 

 in N.E. Africa^ Arabia, Asia Minor, and Persia. In settling 

 all questions relating to terminology, the law of priority 

 has been relied upon usually as a safe guiding principle, and, 

 if any valid generic term conforms to the Linnean Code, 

 there is no sufficient reason why it should be suppressed or 

 the law of priority itself ignored. If the matter of accep- 

 tance or rejection of any term should, however, become 

 purely arbitrary, then, as Jordan states, there can be no 

 finality in such a case. There is therefore every justifi- 

 cation for the general adoption of Buchanan's generic 

 designation of Garra, which, as has been pointed out already, 

 has been used as such by Bleeker, Steindachner, Day, 

 Fowler, and Berg more prominently. I also agree with 

 Jordan that lamta^ is the type of the genus Garra, since 

 it has been regarded by Buchanan as the representative 

 species for his " Division Cyprinus garra^' and also being 

 the first species described by him under this genus. I 

 accordingly use the term Garra in the place of Biscognathus, 

 which I think is the correct procedure, at least so far as one 

 has to deal with forms occurring within the Indian Empire, 

 Ceylon, and Malayan Peninsula. 



Buchanan^s description of his Division Cyprinus garra is 

 too brief and bald to be of any definitive value, and, having 

 examined the somewhat rich material f in the Indian 

 Museum, collected from various localities, and my own 

 examples taken in equally interesting sources, I consider 

 that the generic definition of Garra {Discognathus, part.), 

 given by Giinther and Day, requires revision — at least, in 

 certain particulars. I proceed to append the following 

 diagnosis, which I must state is applicable strictly to forms 

 occurring within the limits prescribed above : — 



Subfamily Ctpjrininm. 



Genus Garra, Hamilton Buchanan (1822). 



1763. Gonorhynchus, Gronow (rejected). 



1H38. Flatycara, McClelland. 



1843. Discocjnntlms (part.), Hackel. 



1864. Discognathus et Lissorhynchus, Bleeker. 



1869. Mayoa,'\)&y. 



* 1917. Jordan, op. cit, p. 115, & 1868. Giinther, torn. cit. p. 68. 



tl918. Annaudale, Rec. Ind. Mus. vol. xiv. p. 45. If the specimens of 

 Discognathus, belonging to the collection of the Indian Museum now 

 held up in Budapest, were also available, our position in regard to 

 several species would have been certainly very much clearer. 



