MEATS AS TROUT FOOD 125 



In our earlier work the experimental lots contained 500 fish which 

 were reduced to 300 later in the season w])on the compartments 

 became crowded. With these small lots of fish, each trou*;li was 

 divided by screens into two compartments of equal size with an 

 interval between. Care was taken that the ration fed in the upper 

 compartment contained no ingredient which was not likewise in- 

 cluded in the ration fed to fish in the lower compartment. Other- 

 wise, but one compartment was maintained to a trough. 



Since our experience led us to believe that more reliable results 

 could be obtained with larger lots it was decided to increase the 

 number of fingerlings in each lot to 1,000 fish. Still later this numl)er 

 was increased to 1,500 fish, which we have adopted as the standard 

 unit. Each lot is kept in a standard hatchery trough 12 feet long, 

 14 inches wide, and 8 inches deep, the average depth of water being 

 about 6 inches. The fish are held in the troughs until the experi- 

 ments are discontinued in the fall. Although the number of fish to 

 the trough early in the season is much smaller than in ordinary 

 hatchery practice, it is believed that more reliable results can be 

 obtained in this way than to attempt to divide the lots later in the 

 season, as would be necessary if the experiments were started with 

 a larger number. 



All our experiments with fingerlings at the Pittsford hatchery 

 with the exception of those in 1928, have been started with very 

 small fish which had been taking food but a short time. Conse- 

 quently, in accordance with ordinary hatchery practice, the number 

 of daily feedings early in the season was greater than later. At 

 first the fish were fed four times a day, but later this was reduced 

 to twice dail3^ 



At each feeding the amount of food given was slightly in excess of 

 that which was taken readily, so as to obviate any possibility of 

 some of the fish being unable to get all they would eat. Needless to 

 say, the troughs were carefully cleaned every day and no trouble was 

 spared to keep them in perfectly sanitary condition. 



Each season all experimental fish were hatched from the same lot 

 of eggs, and every effort was made to keep them under as nearly 

 identical conditions as possible. Consequently, all the lots of any one 

 season are directly comparable ; but owing to variations in the source 

 of the eggs and other conditions this is manifestly not true of lots in 

 different seasons, although, with few exceptions, our results with 

 similar rations have been remarkably constant from year to year. 



The exj^eriments with yearling trout have been conducted in race- 

 ways supplied with brook water which showed considerable fluctua- 

 tions in temperature. At no time, however, were the temperatures 

 high enough to injure the fish. While it is obviously impossible to 

 obtain as constant conditions in outdoor raceways as in hatchery 

 troughs, every effort has been made to keep conditions in the experi- 

 mental raceways as uniform as possible. 



In these yearling experiments, just as with the fingerlings, our 

 primary object has been to keep the fish under conditions similar to 

 those found in the average hatchery. Although this has resulted in a 

 less rigid control of environmental factors, it is believed that our 

 results are more applicable to hatchery practice than would be the 

 case had the fish been kept under less normal conditions. 



