Diet 1 (beef liver) v;as retained as a control for Diets 2 (beef lung), 

 6 (hog liver and salmon viscera), and 7 (hoc liver - hog spleen - salmon 

 viscera). All of the p-roun testing; variations of the raeat-viscera 

 combination (Diets 8 through 12) were retained but altered to include 

 10 percent of vacuum-dried salmon viscera meal during the second 12 

 weeks . The results for the 2l4.-week peri od are summarized in Table 3 • 



The growth rate of the fish fed beef lung (Diet 2) declined sharply 

 during the second 12-week period when conp.ared with the beef-liver 

 control (Diet 1). At the conclusion of the first 12-week period a 

 significant difference in favor of the beef lung diet existed between 

 these lots of fish. After 2k weeks the final w eights of the two lots 

 were comparable. Hemoglobin determinations indicated no anemic condi- 

 tion in the beef lung lot. The hog liver plus salmon viscera (Diet 6) 

 and the hog liver - hog spleen - salmon viscera (Diet 7) combinations 

 showed similar gains which were much greater than the beef liver control. 

 A possible anemic tendency in Diet 6 v/as indicated by the hemoglobin 

 content of the blood taken at the conclusion of the experiment. This 

 result may have been due to a sampling variation or to an inferior feed- 

 ing consistency. Diet 7 which differed from Diet 6 only by the substi- 

 tution of 10 percent of hog spleen for hog liver shovred no anemia. A 

 comparable ration in the second experiment altered only by the addition 

 of 10 percent meal (Diet 29) did not show the anemic tendency indicated 

 in Diet 6. That the ration itself was adequate for the anti-anemic 

 factors is indicated from the hemoglobin content of the samples taken 

 after the first 12 weeks of feeding. As the fish increased in size and 

 the water temperature rose, the bound quality of the diet may have been 

 inadequate and a loss of a portion of the water soluble components may 

 have been the result. In both Diets 7 and 29 the bound quality of the 

 diet was enhanced either by the addition of a more adequate binding 

 agent — hog spleen in Diet 7 — or an absorbing agent — meal in Diet 

 29. The chances of the procurement of an aberrant sample appear to be 

 remote. In no other group of 10 fish composing a sample did the 

 hemoglobin content of the blood reach the low level of 7" 8 grams per 

 100 milliliters that was found in 3 of the fish from Diet 6. The 

 evidence indicates an anemic tendency in the fish fed the hog liver 

 plus salmon viscera mixture caused very probably by an inadequate feed- 

 ing consistency durijng the second 12Hweek period. 



T'ne meat-viscera combinations were altered after 12 weeks by the 

 addition of 10 percent of vacuum-dried salmon viscera meal and a pro- 

 portionate reduction in the other diet components. Frozen salmon eggs 

 (Diet 9) were superior to salmon viscera when substituted for viscera 

 in the meat-viscera-meal mixture. Preserved salmon eggs (Diet 10) were 

 inferior to frozen eggs and, when the stored eggs were used, vt^ere 

 inferior to salmon viscera. The supply of stored eggs was exhuasted 

 after I6 vieeks of feeding and preserved eggs from the 19^0 trials were 

 substituted. A marked impetus to the grovifth rate of the fish was noted 

 when this substitituion was nade, sufficient to close the gap between 

 the viscera- and egg-fed fish at the close of the experiment. Tuna liver 



10 



