( i ) 



Tli cy are to be found in rivers, irrigation or other fchoolsj 

 canals, lakes, tanks muljhils, and are of various degrees of im- 

 portance, not only as regards the amount of fish residing in 

 them, hut likewise in accordance with the character of the 

 contiguous people, as to whether they are fish-eaters or 

 reject this article of food ; also as to the sparsity or the 

 reverse of the population. 



Rivers and their Expansions. 



V. The rivers of India and British Burma possess 

 . ,, _ .. „ . certain peculiarities which tell, to a 



Hi vers or the Indian hmpire. l n , ~ , 



greater or lesser degree, upon the fishes 

 that inhabit them — some of these being due to the season of 

 the year, others to the amount of rain-fall. There are those 

 descending from the hilly regions, as the Irrawadi in Burma, 

 and the Indus in the Panjab and Sincl. Besides which, 

 there are certain differences to be observed when the rivers 

 are in the hilly districts from what obtains in the plains, and 

 these re-act upon the fishes which inhabit them. 



VI. The hill rivers of India, or, more properly speaking, 

 h;ii rivers of India : those those which take their rise in the hill 



having Alpine origins. ranges, may be divided into (1) those 



which have, or (2) have not, Alpine sources. Generally speak- 

 ing, the rivers which have Alpine origins, as those which de- 

 scend from the Himalayas, have for their sources of replenish- 

 ment (exclusive of springs) two most prominent ones. During 

 the hot months, that derived from melted ice and snow is 

 abundant, and a daily rise and fall in the amount contaiued 

 in them may be observed at certain hours, corresponding 

 to the distance from their snowy sources, and which is due to 

 solar influence. Throughout the monsoon season, doubtless, 

 the rains also assist in the melting of the snows : exclusive of 

 this, however, they are sufficient to fill the rivers in what 

 may be termed a spasmodic manner. Thus, in the commence- 

 ment of March, the snow-floods begin in the Indus, the 

 inundations of this river being more entirely due to the 

 influence of the melting snows than to that of the rains, 

 as in the Ganges and Jumna, owing to the rain-fall being 

 greater in the upper regions of the latter rivers than in 

 those of the former and its affluents. These hill rivers con- 

 sequently form torrents, rising rapidly and as rapidly sub- 

 siding, more especially during the rains, whilst, having no 

 contiguous tanks into which the fish could retire, their 



