was used. The amount of food fed per day was determined by reference to 

 feeding charts for the species in wi-iich the size of fish and water temper- 

 ture indicate the percentage of food required in terms of the total weight 

 of the fish. Each trough of fish was weighed at bi-weekly intervals and 

 the amount of food fed was then altered to conform to the increased weight 

 of the fish. The design of this experiment was such as to allow the use 

 of statistical methods for the determination of the significance of differences 

 in final weights between diets. The initial population of each trough was 

 retained intact, except for loss, throughout the course of the experimental 

 period. 



The diets were ground, mixed, and fed within an 8-hour period after 

 the removal of the individual components from cold storage. Salt, at the 

 rate of 2 grams per 100 grams of the mixed diet, was added where it would 

 improve the diet consistency. After the diets were prepared they were 

 held under refrigeration at a temperature approximating 29° F. until they 

 were fed. The fish were fed twice daily by means of a hand ricer. 



Seven diets were incorporated into the growth evaluation studies. 

 Their composition and their place in the scheme of the experiment will be 

 discussed individually. 



Diet 1, consisting of 100 percent beef liver, is the standard control 

 diet in fisheries nutritional studies. It was included not because it is 

 considered the acme of perfection but because it supplied a point of reference 

 by which the results of this experiment may be compared to those of other 

 workers. 



Diet 2 served as the actual control for the salmon meal evaluations. 

 It consisted of 22.2 percent each, of beef liver, hog liver, and hog spleen 

 and 33.4 percent of salmon viscera bound by the addition of salt. This 

 mixture of meals and salmon viscera in conjunction with 10 percent fish 

 meal has been tested both experimentally and in actual production diets 

 and has given excellent results. It has been found to fulfill the nutri- 

 tional requiranents of blueback salmon as measured by good growth and the 

 absence of discernable dietary deficiencies. In order to measure the con- 

 tribution of the various fish meals to the growth rate it was necessary 

 to feed the standard meat and viscera mixture alone. 



Diet 3 consisted of the standard mixture of meats and viscera (20 per- 

 cent each, of beef liver, hog liver, and hog spleen and 30 percent salmon 

 viscera) plus 10 percent of low-temperature-dried salmon viscera meal bound 

 by the addition of salt. The difference in weight between the fish fed 

 this diet and those fed Diet 2 would serve as a measure of the contribu- 

 tion of this fish meal to the growth of blueback salmon. 



Diet 4 varied from Diet 3 only in the composition of the fish meal. 

 It consisted of the standard meat and viscera mixture and 10 percent flame- 

 dried salmon offal meal bound by the addition of salt. It also was included 

 to measure the contribution of variations in fish meal preparation and 

 composition to the diet. 



55 



