1871.] F.Day — Monograph of Indian Cy-prinida. 315 



33. Barbtts (Capoetd) parrah, pi. xxi, fig. 4. 



Puntius parrah, Day, Proc. Zool. Soc. 1865, p. 301, and Malabar Fishes, p. 211, 

 pi. 7, f. 3 ; Giinther, Catal. vii, p. 142, (passim). 



Parrah perlee, Mai. ; Katcha-Jcarawa, Hind. 



B. III. D. 3/8, P. 15, Y. 9, A. 3/5, C. 19, L. 1. 25, L. tr. 5/4. 



Length of head 1/5, of caudal 1/5, height of body 1/4 of the total 

 length. 



Eyes. Diameter nearly 1/3 of length of head, 1 diameter 

 from end of snout, 1 J diameters apart. 



Dorsal profile slightly more convex than that of the abdomen. 

 Height of body 4/1 1 of its length, excluding the caudal fin. Barbels 

 equal to two-thirds the length of the orbit. 



Fins. Dorsal commences midway between the end of the snout 

 and the base of the caudal fin, its last undivided ray is osseous, 

 weak in the young but strengthening with age, and as long as the 

 head without the snout. 



Scales. At the fourth scale on the row above the lateral line two 

 rows commence. 



Lateral line — complete. 



Colours. Back greenish, divided from a silvery abdomen by a 

 dark bluish line. Cheeks golden red. Pectoral, ventral and anal 

 tinged with yellow ; dorsal and caudal dusky. A diffused black 

 spot on the lateral line extending from the twentieth to the twenty- 

 second scale. Eyes golden. 



Mob. Malabar, Mysore and Madras, growing to 6 inches in 

 length. 



Regarding some specimens of B. parrah and B. perlee presented by 

 me to the British Museum, Dr. Giinther observes in his Catalogue 

 " the fish given us as P. perlee agrees much more with the descrip- 

 tion and figure of P. parrah, than with that of the species to 

 which the specimen is said to belong." Having re-examined the 

 two fish, both 3J inches long, I find they have been transposed, an 

 accident which might easily occur, but which could have been as 

 easily rectified by any Ichthyologist, had he felt inclined to do so, 

 when examining the specimens. As only one bottle (e) appears to have 



