MEATS AR TROUT FOOD 



135 



:the troughs following feeding, and the fingerlings were never seen 

 leating it. Pilchard meal, on the other iiand, being heavier and 

 icoarser, was much easier foi- the lisli to pick up and eat. It was, 

 ifurthermoi'e, less distinguisliable fiom tlie particles of liver, wiiicii in 

 the case of tiie haddock and shrimp combinations were easily selected 

 and eaten. This resulted in the more active, stronger fish securing 

 most of the meat before the smaller ones could do so. An irregular 

 lot of fingerlings in respect to size was the natural result. Systematic 

 weighing of samples from these lots revealed the irregularities in the 

 size of the fish, it being particularly evident in the growth curves. As 



.3.5 

 CO 



DCS. 



X MORTALITY 



5 10 15 2 25 30 35 40 



2.5 



-< 



o 1.5 



> 



1. 



S .5 





N4- 



N6 



N20 

 NI9 



20 10 30 19 9 29 18 7 27 

 APR.. MAY JUNE JULY AUG. SEPT 



Figure 6. — Comparative growth and total mortality of brook trout fingerlings 

 fed beef liver, pig liver, and beef melts in combination with dry prduets. 

 A'^4 = beef liver, 50; clam meal, 50. i\'6 = beet' liver, 50; clam meal, 25; 

 shrimp meal, 25. 3/^0 = pig liver, 50; clam meal, 25; shrimp meal, 25. 

 iVW=beef melts, 50; clam meal, 25; shrimp meal, 25 



1 



.the fish on the pilchard combination were more uniform in size the 

 .data secured from the systematic sampling gave a much smoother 

 growth curve. 



Our results also revealed that there was nothing gained in feeding a 

 combination of haddock and shrimp meals. A combination of 50 per 

 cent beef liver and 25 per cent each of shrimp and haddock meals 

 merely resulted in a rate of growth closer to that obtained on a diet 

 of haddock and beef liver, which has already been observed to be 

 slightly better than a shrimp combination. 



The haddock meal proved to be somewhat superior to the menhaden 

 meal ( fig. 9) , and the diflference would probably have been more marked 

 had the ordinary commercial grade of menhaden meal been used in 

 our experiments. Although the menhaden meal was flame dried, the 

 deleterious oils had been extracted by a special process developed by 



