RESULTS Al^TD RECOMMENDATIONS 



Resxilts of feeding the various diets of the first experiment and second 

 experiment were judged almost entirely by significant differences in mean lot 

 weights using analysis of variance for paired experiments as described by 

 Snedecor (I9U6). The mortalities, being low and comparable, offered no assist- 

 ance in judging the diets except in one or two instances. Conversion factors 

 merely served to confirm the weight differences since, with minor exceptions, 

 the greater the mean lot weight the more efficient (or lower) the conversion 

 rate. Blood counts and hemoglobin determinations showed that none of the diets 

 in the first experiment produced a tendency toward anemia, but that two of the 

 diets in the second experiment created an anemic condition. The blood counts 

 and hemoglobin determinations proved almost valueless for the establishment of 

 clear-cut differences by analysis of variance because of variation within lots. 



RESULTS OF FIRST EXPERIiyiEMT 



Beef Liver Control 



In the first experiment the beef liver control, represented by Diet 1 

 (Table 1), repeated the results attained in previous trials. The relationship 

 of the mean lot wei^^ht produced by the beef liver control and the mean lot 

 weights produced by the other diets changed however, during the cold-x-7ater period. 

 Previously, beef liver had been one of the better cold-water diets. In this 

 trial the growth produced by beef liver was not only exceeded at the 5 percent 

 confidence level by all but tiiio diets, but also there was, with two exceptions, 

 no significant difference in mean mortalities. These results at the end of 

 the cold-water period indicated that the diets in this first experiment worthy 

 of consideration as production diets during the most difficult time of linger- 

 ling rearing: the cold-water period. During the warm-water period, beef liver 

 produced the same relatively low growth rate encountered in previous experiments. 



Effect of the Deletion of Hog Spleen from the Meat-Viscera Mixture 



This experiment compares the meat-viscera mixture with a similar diet 

 containing no hog spleen. The purpose for this deletion was to form a diet with 

 a higher vitamin and protein content. The higher vitamin and protein level was 

 attained by replacing hog spleen with increased amounts of liver. 



The meat-viscera mixture (represented by Diets 2 and 3) consisted of 



22.2 percent each of beef liver, hog liver, t-nd hog spleen, and 33 .Ii percent of 

 salmon viscera during the cold-water period, "During the' warm- Water period, but a 

 salmon waste meal was added to Diet 3 to form a mixture of 20 percent each of 

 beef liver, hog liver, and hog spleen, 30 percent salmon viscera, and 10 per- 

 cent vacuum-dried salmon viscera meal. 



The diet without spleen (represented by Diets 8 and 9) consisted of 



33.3 percent beef liver, 33.3 percent hog liver, and 33. U percent salmon viscera 

 during the cold-water period. During the warm-water period Diet 8 remained the 

 same but salmon waste meal was added to Diet 9 to form a mixture of 30 percent 

 each of beef liver, hog liver, and 10 percent of yacuum-driegl salmon viscera meal, 



