INDIAN CYPRINID^. 



245 



44. The consideration of the third or aberrant group in Cyprinidce 

 involves some points which will not be so clear to those who have not 

 studied the principles of natural classification. This group should possess 

 three types, and these should be so related as to form a circle of affinities 

 among themselves. This property has suggested the following proposition 

 to Mr. Swainson, which tends to reconcile some diversity of opinion that 

 formerly prevailed as to the number of primary types — " The primary 

 circular divisions of every group are three actually, or five apparently." The 

 three aberrant types are named by Mr. Vigors (Lin. Transac. vol. 14) from the 

 corresponding groups in ornithology, natatorial, suctorial, and rasorial.* The 

 third type I find to be represented by the Loaches, but before we enter into 

 a consideration of that part of the subject, it is necessary to point out the two 

 first types which I have already only alluded to ; this I must do by entering 

 into more particulars than may seem to be necessary. 



45. Buchanan in defining his ninth division of the old genus Cyprinus 

 which is composed of Gonorhynchs, as I have already pointed out, gave 

 them the barbarous name of Garra, and compares their habits to those 

 of the Loaches, and observes that they are called Balitora, or sand-dig- 

 gers, by the natives ; a name I may observe which in Assam, and I 

 presume also in Bengal, is applied to Loaches only. Indeed the Gonor- 

 hynchs, or Garra of Buchanan are peculiar to mountains, from whence they 

 are driven down during floods, and do not extend beyond the rapids that 

 skirt their base, so that they can scarcely be said to be entitled to any Bengal 

 name. In this group JMr. Gray detected a new genus, to which I wish he 

 had given a more appropriate name than Balitora ; for independently of the 



* As these terms liave been applied conditionally by Mr. Swainson to Manniialt'a, I need 

 make no apology for following the e.xampic of so good an autiiority in applying them to 

 lisiics. Indeed I have bt'cii guided entirely l)y tlic \iews of iMr. INIacleay, as exhibited in the 

 Linna'an Transactions, and the works of Mr. Swainson from beginning to end of this paper. 



I 



