units for chum salmon. E value ratios at 300/328 millimicrons for the 

 viscera oils were within the acceptable range: chum salmon viscera oil, 

 0.623, and pink salmon viscera oil, 0.698, 



The concentrations of vitamin A reported here should not be con- 

 sidered necessarily typical or representative of those to be obtained in 

 a coinnercial operation. It is Impossible to generalize from the tests made 

 on a few samples of the raw material taken at one cannery to the actual 

 potential production from the several species over all of Alaska for the 

 entire fishing season. These data do give some conception of the range 

 that might be expected for the Seldovia district. Very probably if an 

 average value were computed from these data, additional tests at other 

 canneries or at a different time in the fishing season would result in oils 

 even more divergent than those reported here. Material as variable as fish 

 waste is very difficult to sample accurately even on a considerably larger 

 scale of operations. The data reported is presented merely as the best 

 criterion available for the evaluation of this potential source of vitamin 

 A oils. 



The removal of the testes from the waste eliminated the viscous stage 

 in the digestion. Two percent of sodiura hydroxide was sufficient to pro- 

 cess the material in ij hours at 200° F. The oil recovered from the 

 viscera less testes did not show any consistent difference in oil yield or 

 in vitamin A concentration over that from the total viscera. 



The combination of heads and viscera less gonads, was adequately 

 processed within 2 hours at 200° F. using 1.5 percent of sodium hydroxide. 

 No emulsion difficulties were encountered and the appearance, odor, color, 

 and flavor of the oils was excellent. Oil yields ranged from 5.3 percent 

 for king salmon to 9.5 percent for coho salmon. The vitamin A content 

 varied from 14,690 units for the king salmon oil to 2,126 tinits for the 

 coho salmon oil. 



With raw material that varies from lot to lot and from day to day as 

 much as fish does, it is difficult to account for the specific proportion 

 of oil and of vitamin contributed by each of several component parts with- 

 out elaborate control experiments. One method used to give an approximation 

 of the relative contribution from the salmon heads was the substitution 

 for the heads of a definite amount of salmon head oil previously prepared 

 from material as nearly comparable as possible. The vitamin A concentration 

 of this head oil was known and the volum»of oil added was known. The total 

 vitamin A from the head oil could, therefore, be calculated. The total 

 vitamin A in the oil derived from the digestion of the viscera and the added 

 head oil minus the total vitamin A supplied by the head oil would give the 

 amount of vitamin A contributed exclusively by the viscera. In the experi- 

 ments on red salmon waste. Lot 35 was on heads yielding an oil containing 

 154,000 U. S. P. units of vitamin A per pound of oil. Lot 36, prepared 

 from heads and viscera less gonads, yielded an oil containing 2,369,000 

 units per pound of oil. The heads therefore supplied at least 6.5 percent 

 of the total vitamin A recovered. Lot 42, for which 5 pounds of head oil 



42 



