52 



if the animals lived ouly one day and were replaced by 

 others. This is not the fact, and how long they may 

 live I cannot sa}^, bnt if each individnal lived a week 

 the amonnt of food prodnced would be less than i lb. 

 in the entire season, as the calculation is for a daily 

 renewal of all life. Two hundred baby trout could 

 have lived there during the first week of their lives 

 and fed well ; after that time, when their appetites 

 began to get sharper, say in a fortnight, all the food to 

 be found would be just what came in the water suppl}^ 

 and that would not have fed half a dozen when two 

 months old. If I had been skeptical of the practical 

 utility of this scheme before this experiment there has 

 been nothing to convince me of error ; still, if other 

 trials under other circumstances show that it is practi- 

 cable to raise enough natural food to rear 20,000, or 

 even 10,000, to be six months old, I must try the plan 

 which has proved to be successful. While writing 

 this I do not know that any other men but Mr. Clark 

 and myself have worked in this field in America, still 

 it is to be hoped that they have done so and that they 

 will publish their experience. Such work is very inter- 

 esting to one who has a taste for it, as most fish cultur- 

 ists have, and this paper may stimulate others to similar 

 trials. I think one plan was to have a number of sep- 

 arate ponds in which to breed the food and to tap them 

 in succession, and allow each one to furnish food to the 

 fish, which wete not to be driven to the pasture, but to 

 remain in one pond and get the food supply from dif- 

 ferent sources at different times. This is certainly the 

 best plan, as any trout breeder will certif}^ because it 

 is a difficult matter to get the last dozen trout from a 

 pool containing vegetation or hiding places of any 

 kind. At present writing I have less faith in the 

 scheme than when I began to experiment with it. 



