1871.] F.Day — Monograph of Indian Cyprinidce. D9 



moveable articulation may exist there, or there may be lateral 

 prominences on the mandibles. 



The lips may be exceedingly, moderately, or but slightly deve- 

 loped, sometimes absent from one of the jaws, closely investing 

 both, or reflected from off one, or either. There may be an un- 

 interrupted labial fold across the mandible, or portions of the lip 

 may be much developed, fringed, or crenulated. Occasionally, in 

 some genera, a horny covering to one or both lips is invariably or 

 generally present. 



The existence, number or absence of barbels has been thought by 

 some authors to be a reason for constituting genera, but such is 

 not generally held to be valid. These appendages in some genera, — 

 more especially when the fish are not kept in a state of domes- 

 tication, or confinement, but left in their natural situations, — 

 appear to be pretty constant, and though not in themselves cause 

 enough for defining a genus are frequently sufficiently well marked 

 for the purpose of forming sub-genera, good examples of which 

 may be seen in the genus Barbus, or Barilius. On the other hand 

 in some species, these appendages may be abnormally absent as in 

 the Danio, consequently sub-divisions of the genus founded solely 

 on this character would lead to erroneous results. 



The position of the fins indicates sub-divisions which might 

 be used in the primary groups, and have for convenience sake 

 been adopted as follows : — 



a. Dorsal fin commencing nearly opposite the ventrals, the anal 

 being short. 



b. Dorsal fin commencing very distinctly posterior to the ven- 

 trals, but not extending to above the anal, ivhich last is short, or of 

 moderate length. 



c. Dorsal fin commencing in the interspace between the ventral 

 and anal, or over the latter, and generally extending to above it, 

 whilst the anal is of moderate length or elongated. 



The shape of the dorsal fin is likewise important, but its 

 length alone, or rather number of its rays, appears insufficient for 

 the purpose of defining a genus, much less a sub-division of a 

 group. When an uninterrupted series extends from a few rays to 

 a large number, and no other sufficient difference exists, such divi- 



