1922.] S. L. Hora : Fish of Mountain Torrents. 59 



accidentally carried into them, and secondly, that there has been 

 a step by step colonisation of the hill-torrents from the sluggish- 

 streams that flow in level country. As regards the first possibility 

 it may safely be inferred that forms which have not previously 

 acquired adaptive characters, cannot live in rapid waters because 

 at every move they are liable to be swept away down stream. 

 The second view is more probable and the following are some of 

 the main points in its favour : — 



(1) As a hill-stream changes into a sluggish stream in almost 

 level country its fauna changes accordingly and intermediate 

 forms, like Crossochilus latia, between the typical hill-stream fishes 

 and the fishes of the slow streams are always met with in the 

 intermediate regions where the water is neither flowing very fast 

 nor very slowly. 



(2) The very fact that the members of certain genera such as 

 Glyptothorax and Garra can be arranged in a series according to 

 the degree of modification they exhibit in response to the strength 

 of the current of stream, shows that there has been a gradual 

 colonisation of rapid streams. 



(3) The hill- stream fishes in the course of their development 

 pass through many different stages which clearly show, at any 

 rate in the case oi Garra, 1 that the evolution of such forms is from 

 those that live in sluggish streams. Not only is this shown by 

 the form of the body but also in the modifications of such organs 

 as the air-bladder, the position and form of the mouth and the 

 eyes, and in the reduction of the branchiostegal membrane and 

 rays. In short, the developmental series of Garra as given in a 

 previous paper recapitulates the history of the evolution of the 

 genus. 



Having subscribed to the view that the hill-stream fauna has 

 originated by the process of gradual colonisation from the slow 

 streams, it will not be out of place to discuss the causes that might 

 have led to the migration of these forms. There are two chief 

 factors which might compel such a migration : — food and safety. 

 In the hill streams there is always plenty of food in the form of 

 algal slime on the exposed surface of rocks; but only those 

 animals can make use of it which have their jaws specially 

 adapted for rasping it off the stones on which it grows. As 

 regards safety, it may be said that there is very little competition 

 in the hill-streams among the fishes themselves. Moreover, they 

 are practically safe from the ravages wrought by birds and large 

 predaceous fishes, crocodiles, etc. Certain fishes like Nemachilus 

 can find shelter underneath stones and the readiness with which 

 they hide themselves is marvellous. Probably the fishes inhabit- 

 ing slow- running waters originally ascended the hill -streams step 

 by step in search of food and gradually acquired certain char- 

 acters which made them specially suitable for living in the newly 

 chosen environment. 



' Hora. Rec. Ind. Mtis. XXII, p. 639 (1921). 



