30 American Fishezies Society 



pounds of food were required to produce one pound of fish. The 

 reciprocal factor is found by dividing the gain in weight of fish by 

 the weight of food consumed and expresses the fraction of a pound 

 of fish produced by one pound of food. The direct factor is very 

 convenient for calculating the cost of the food used in producing 

 a pound of fish. It is only necessary to multiply this factor by 

 the cost of the food per pound. Thus, in the first table, in the case 

 of beef liver, the average direct factor for a period of 70 days was 

 found to be 3.16. Multiplying this by .12, the cost of beef liver 

 per pound, we find that the cost of producing a pound of fish was 

 $.379. It will be seen that the higher the direct factor the greater 

 will be the cost of the fish. In the case of the reciprocal factor, 

 the higher the factor, the lower will be the cost of the fish. 



The mortality as recorded in the table represents the number of 

 deaths occurring each week in per cent of the total number of 

 fish used in the experiment. 



Comparing the various food costs for producing a pound of 

 fish, it may be seen that the highest costs occur when liver is used 

 alone or in combination with flour — series A, B and C. As the 

 amount of beef liver is lowered so is the cost of production lowered. 

 Except in the case of chinook salmon, where the food mixture 

 contained only 15 per cent liver, we find the four lowest costs to 

 have been obtained with mixtures containing no liver — series 

 F, G, H and I. In not one of these four series, however, was 

 the experiment period longer than 36 days and hence further 

 trials may change the figures to some extent. This is particularly 

 true in the case of the meat meal-fish meal-flour mixture. Never- 

 theless the average of costs in the four cases — 10c — is low enough 

 to indicate clearly the high food value of these dried meals. 

 Outside of series F, where the production cost of 6 cents is undoubt- 

 edly too low, the most economical mixtures seem to be those 

 containing meat meal and shrimp meal with a small amount of 

 flour for binding the mass. This is indicated in the two series, 

 G and I, where the production costs are about eleven and one-half 

 cents in the former and a little under eleven in the latter. 



The trials with the meat meal-peanut meal-middlings mixture, 

 series H, were so high that it was thought best to test peanut meal 

 alone. The same fish used in series H were fed a mixture of 



Peanut Meal 90% 



Middlings 10% 



