produced the greatest gains. The next greatest gains were produced with the 

 combination diets containing 33 percent or less of salmon viscera. The least 

 gain was produced by high-level salmon viscera diets. 



Second Experiment 



9o As in past experiments during warm water, the beef liver control 

 produced a final mean weight that was relatively low when compared with those 

 of the composite diets. 



10. Air-lift dried mackerel offal meal contributed significantly less 

 growth than did vacuura-'dried salmon viscera meal. 



11. In the tests of preserved eggs, sodium bisulphite at the 0.5 per- 

 cent level, produced the greatest growth response. 



12. Hake proved to be a satisfactory component of the meat-viscera-meal 

 combination but produced Ibss growth than either halibut sawdust or hog slpeen. 

 The diets in the hake group which tested salmon milt, salmon eggs and vacuum- 

 dried salmon viscera meal not only substantiated the compa risons made in the 

 first experiment with these variables but also demonstrated that hake was com- 

 patible with these foods. 



13. Both whole cod and halibut sawdust were satisfactory substitutes for 

 hog spleen in the raeat-viscera-meal combination, but, of these two products, 

 only halibut sawdust produced as much growth as hog spleen in this diet. In 

 the test of single components, whole cod again produced less growth than 

 halibut sawdust. 



lU. Tuna liver was a satisfactory substitute for beef liver in the meat- 

 viscera=raeal combination under the warm water conditions of this experiment. 



l5o An APF concentrate (without aureoraycin) reduced the growth rate re- 

 sulting from a diet of salmon viscera, hake, and salmon viscera meal whereas 

 an APF concentrate ( with aureoraycin) had no effect on the growth rate resulting 

 from a diet of beef liver, hog spleen, canned salmon, suid salmon viscera meal. 



l6. For present use during warm-water periods, tuna liver, halibut sawdust, 

 air-lift dried mackerel offal meal, hake, and whole cod were suggested as pos- 

 sible substitutes for certain components of the standard meat-vise era-meal 

 combination. 



ACKNOWLEDGEMENTS 



The authors wish to acknowledge the cooperation of personnel of the 

 Branch of Game Fish and Hatcheries for the use of facilities and major food 

 items necessary to conduct the experiments. 



Acknowledgement is also extended to Mr„ Maurice E, Stansby and Dr. 

 G„ Ivor Jones of the Branch of Commercial Fisheries who were in charge of the 



21 



