digestion, and 2. by boiling at atmospheric pressure. During a stor- 

 age period of ei^t months, samples from these packs have been com- 

 pared at monthly intervals with samples from a control pack to which 

 no oil was added. A tasting panel was used to determine whether the 

 added salmon oil imparted any undesirable flavors to the canned salmon. 

 Flavor ratings have been expressed in terms of numbers with the best 

 possible rating scored as 10 and the lowest possible rating scored as 

 zero. 



The flavor scores for each species of salmon indicate several gen- 

 eral trends. The pack of king salmon with added alkali process oil 

 and of coho salmon with added boiling process oil were rated highest 

 in these respective species. The control samples were rated highest 

 in the packs of red, pink and chum salmon. In those instances where 

 there was a degree of difference between second and third choice, the 

 lots containing the alkali process oil were slightly favored over those 

 containing the boiling process oil. 



It is very possible that had equipnent been available to prepare the 

 oil by pressure cooking as originally planned, instead of bj' the process 

 of boiling at atmospheric pressure, an oil more coaparable to that now 

 used cajmercially would have resulted. For this reason any inferences 

 drawn from these tests should be rechecked by the storage and examina- 

 tion of salmon packs in which the vacuum cooked salmon head oil is com- 

 pared with that prepared by the alkali digestion process. 



There was a consistent trend in that the combined average flavor 

 scores for the three lots at each examination period declined as the 

 tests proceeded. There was some difference in the rate of flavor 

 score decline with species. King and coho salmon lead with the highest in- 

 itial scores of 9.1 and 8.9 but declined 2,1 points each to approximately 

 the final average of 6.8 for the five species. Pink salmon ratings 

 dropped only 1.1 points from the initial 7.9. Chum salmon was rated 

 lowest at the first examination and at the last, but changed only 0.9 

 points. Red salmon samples decreased by only 0.5 points frcm the 

 first rating of 7.7. 



On the basis of the flavor preference of the seven tasters who rat- 

 ed the three lots for the five species of saLnon, the control samples 

 were best in 19 of the 35 tests. The samples with alkali process oil 

 added were best in 8 tests, the samples with oil by boiling added were 

 best in 5 instances. The balance of three tests were inconclusive with 

 respect to a definite first choice. 



The difficulties in obtaining an unbiased judgment based entirely on 

 flavor differences by the use of a tasting panel are recognized. The 

 apparent preference scores were, therefore, not considered to be sig- 

 nificant unless there was a difference of 1.3 points between the ratings 



96 



