576 MR. F. DAT ON THE &EOQEAPHICAL 



time, we sliould expect to see tlicm breeding there; but we find 

 that Barbels in India breed either in the waters of the plains or 

 in those of hilly regions, whilst it is an invariable rule that the 

 larger forms choose the latter place. Thus we see the Mahseer 

 and its allies residents of rivers which take their origins in moun- 

 tains, that during the cold months of the year, when the mountain- 

 streams are at their minimum size, these fishes descend to the 

 waters of the plains, but reascend the hill-waters with the first 

 burst of the monsoon in order to deposit their ova in cooler loca- 

 lities. Although this is most easily observed on the Himalayas, 

 I have likewise found the Mahseer in Sind, and that it ascends 

 the Beloochistan hills to breed ; that the same phenomenon occurs 

 in Malabar, where the rivers descend from the western Ghauts ; 

 and I have likewise been fortunate enough to be a witness to the 

 fact that the larger species of Barbel (Sarhodes) in Southern 

 India ascend the Neilgherry streams for the same purpose. 



Some species of Barhodes oi a moderate size breed in the waters 

 of the plains of India ; but these have generally a serrated bony 

 ray to the dorsal fin. As a rule, all Barbels breeding in the plains 

 are of a moderate or small size ; whilst of the subgenus destitute 

 of any barbels {Puntius) all are small, some minute, whilst none 

 normally breed in hill-waters. This leads me to believe that 

 Puntius is a degenerate Barhodes, due perhaps to constant resi- 

 dence in the plain ; that such deterioration is shown in their dimi- 

 nutive size and want of barbels. 



If such is the case, we ought perhaps to be able to show spe- 

 cies in which this modification is even now going on ; and such I 

 believe I have found in Southern India, Barhus maliicola, 

 Cuv. & Val., has two barbels, but is otherwise similar to B.fila- 

 mentosus, Cuv. & Yah, which possesses none. Should a number 

 of examples be exam.ined, it is seen that these appendages in 

 some are very minute, being, as a rule, smallest in specimens ob- 

 tained furthest from the hills. In South Canara, the Wynaad, 

 and base of the Neilgherries, where the finest examples are met 

 with, B. maliicola abounds ; towards Cochin and up the Coro- 

 mandel coast as far as Madras the barbels, when present, are small, 

 and the B. filament osus is the common type. 



I may also here record a curious change which has Occurred in 

 one species of Barhus, the B. conchonius, H. B., which has been 

 transported, within the memory of man, from the plains into the 

 JN'ainee-tal lake on the Himalayas. It is evidently losing the ser- 



