TROUT BREEDING. 69 



should be made from milk which has turned in the pan ; 

 it is not so hard, but lighter and more digestible than that 

 made over the fire. I have fed fifty thousand fry with 

 curd for more than two months, the time occupied at each 

 feeding, not exceeding a half-hour. For feeding so large a 

 number two or three quart bowls should be provided, and 

 a lump of curd as large as one's forefinger dropped into 

 each, half-full of water. The curd in each bowl is then 

 triturated successively, and the milky water poured off 

 after the particles settle to the bottom. Two or three tri- 

 turations reduce it to atoms sufficiently small. In feeding 

 it to the fry, the rim of the bowl is placed beneath the 

 surface j the influx of the water suspends and whirls the 

 light particles around, when the bowl is canted and a por- 

 tion distributed in different parts of each nest. Four lumps 

 of curd each as large as one's forefinger, fed in this way, suf- 

 fices for forty thousand fry when they commence feeding ; 

 as they grow, the quantity should be gradually increased ; 

 but double this quantity is enough as long as they remain 

 in the hatching-troughs. When they are let into the 

 nursery the quantity may be again increased, but not 

 enough to foul the bottom. An hour or two after feeding, 

 if the gravel is stirred lightly, the particles of food that 

 have settled to the bottom are set adrift, when the fish will 

 take it again. If too much of this food is given it will 

 make a mouldy covering over the gravel, and emit an un- 

 pleasant odor. As often as this occurs, the gravel in the 

 troughs should be washed as already directed. 



A strip four or five inches wide should now be placed 



