when fed for 12 weeks at water temperatures below I|.5° F. It is doubtful 

 if these combinations would prove generally satisfactory as starting 

 diets for first-feeding fish, because their feeding consistencies are 

 difficult to control. There is, however, every evidence to indicate 

 that they would be excellent as cold-water diets after the fish were 

 feeding well. 



The second group of diets were variations of the standard meat- 

 viscera combination. Salmon eggs, either frozen or preserved, were sub- 

 stituted for salmon viscera in Diets 9 and 10 and tuna liver for beef 

 liver in Diet 11 » The salmon eggs preserved with 0,5 percent sodium 

 bisulfite had been allowed to stand for three months before being frozen. 

 These eggs differed from those of previous trials (Robinson et al. 19$lb) 

 in this respect o In the 1950 trials only the toxicity of the preserva- 

 tive was tested in that the eggs were not allowed to stand for more than 

 two or three days before being refrozen. A comparison of the growth 

 produced by Diets 9 and 10 indicates that an alteration in the growth 

 potential occurred when the eggs were preserved with 0.5 percent sodium 

 bisulfite. Partial decomposition of the eggs -was indicated by the par- 

 tial disintegration of the egg shell, heavy mold formation on the 

 exposed surfaces, and rather putrid odor. The growth response of the 

 fish that were fed the preserved eggs was equal to that of those fed 

 frozen salmon viscera but inferior to that of those fed frozen salmon 

 eggs. 



Tuna liver substituted for beef liver in the meat-viscera combina- 

 tion (Diets 8 and 11) proved adequate during the first 12 weeks of 

 feeding. The growth rate on this diet was comparable to that of the 

 meat-viscera mixture and no anemia was indicated. 



The third group in this experiment was designed to determine the 

 cause of mortality when meals were included in the diet during prolonged 

 periods of cold-water temperatures . Two hypotheses were advanced as to 

 the cause for this mortality. In the first, it was assumed that a 

 reduction in enzymatic action due to colder temperatures reduced the 

 rate of digestion of the protein. As a result, some of the more slowly 

 liberated amino acids were not made available to the fish. To test this 

 hypothesis, predigested salmon viscera meal was fed at the 10- and 20- 

 percent levels in Diets Ik and 15 for comparison with Diet 13 which 

 contained 10 percent of the standard vacuum-dried salmon viscera meal. 

 The predigested meal, prepared by Halver and Coates of the Western Fish 

 Nutrition Investigations, was enzymatically hydrolyzed by the use of 

 pancreatin and papain then vacuum dried. Digestibility as indicated by 

 formal titration was increased from 11.10 to 21.98 or 91.1 percent. In 

 addition k percent methionine and U percent lysine were substituted for 

 equal portions of the vacuum-dried salmon viscera meal on the basis of 

 100 percent meal content (Diet 18). Methionine and lysine were assumed 

 to be the two amino acids which would be the most slov/ly liberated in 

 the digestion of the meal. The substitution of 10 percent of predigested 

 meal for vacuum.~dried meal resulted in a significant increase in 



