378 INDIAN CYPRINID^. Sarcohor'ma . 



the Perilamps and Leuciscs are, on the other hand, clearly enough confined to 

 Sarcohorince ; but there is a group still required to fill up the space between 

 the Opsarions and the Apalopteri7ice in order to complete the SarcohorincB, and 

 unite that group with the aberrant circle. Whether the American genus 

 Amia, or the genus Suclis, Cuv., or ErytJirinus, Gronov. which all natu- 

 ralists suppose to present near relations to Cyprinidce, may, one or all, be 

 destined to fill up this blank, is a question regarding which, without those 

 genera before me, I cannot venture an opinion. Amia, I may remark, is said 

 to be without cascal appendages to the stomach, a circumstance which 

 ought to place it with the Cyprins, rather than with the Clupeidce. 



It may be necessary to explain in this place why I have given the two 

 principal groups of Cyprinidce the rank of sub-families, rather than that 

 of mere genera. A genus appears to have been intended as the lowest deno- 

 mination of a perfect group ; and, indeed, is still so regarded that no smaller 

 groups are supposed to be comprised within it, although where the species 

 are numerous they may be conveniently separated into artificial sections, or 

 sub-genera. Should such sections be further augmented, so as to become ne- 

 cessary to separate them still farther, it is obvious that this can only be done 

 by augmenting the value of the higher group, by raising it to the rank of a 

 family, or sub-family, when the sub-genus would naturally become a genus. 

 This is what I have done ; and though a species may be so isolated, as to form 

 a distinct family of itself, by means of the numerous links that would be 

 requisite to connect it with the nearest known forms being lost, or undis- 

 covered, yet this is so unlikely now to be the case in zoology, that we may 

 regard the number of genera as the safest rule for determining the value of 

 groups ; and as genera are understood to be the lowest denomination of perfect 

 groups, and sub-genera mere artificial sections of genera, we can have no 

 vmcertainty in the nomenclature of groups ; though I am aware that for want 

 of a little reflection on this subject, the distinction between genus and sub- 

 genus is often confounded, or ill understood. 



I 



