THE NEW AET OF BEEEDINQ FISH. 69 



to time, by means of a glass pipe, the bottom of the 

 stream or vessel of the deposit formed of particles of 

 animal matter which the fish have not eaten. 



I have thought that this inconvenience, which 

 without great care is followed by serious results, 

 might be easily avoided, if, in place of dead food, 

 the young salmon and trout could be furnished with 

 living prey. Although the results I have thus far 

 arrived at in experimenting with this view are not 

 yet confirmed by long practice, they nevertheless ap- 

 pear conclusive enough to be reported. 



The spawn of the frog, so extolled by the fisher- 

 men of Bresse, was the subject of my first experi- 

 ment. The eggs of this animal were placed in ponds 

 in which were hatching boxes containing young sal- 

 mon and trout, and were there developed and hatch- 

 ed ; but neither the tadpoles nor the albumen that 

 enveloped them were sought by the young salmon 

 and trout. I do not mean to say that such prey was 

 not to their taste, but only that its size was too 

 great for such little fish. _ If tadpoles are suitable 

 food for fish a year or two old, I am convinced they 

 cannot be fit for little ones which have just lost the 

 umbilical bladder. I found it necessarj', therefore, 

 to abandon the attempt of feeding fish of that age 

 with such food. 



I had recourse to another expedient and with 

 happier results. Pike's eggs, artificially fecundated, 

 being put to hatch in reservoirs in which were very 



