assigned the three lots at an examination. There was considerable cor- 

 relation between the lot most frequently rated highest by the panel 

 ■and the fact that a flavor score spread of approximately 1.5 points 

 was reported at that examination. For example, in the king salmon ex- 

 aminations, the alkali process added oil lots were rated best in the 

 two instances where scores as diverse as 1,5 points were involved. For 

 the two examinations in which the range in scores most nearly approach- 

 ed 1.5 points, the best score was in favor of the alkali process added 

 oil sample. Chum salmon showed the best correlation with four instances 

 of 1.5 points range in scores. In each case the control was given the 

 highest rating. The two other examinations in which therange of scores 

 approached 1.5 points likewise showed the control receiving the best 

 score. 



Conclusions 



The flavor score rating of the tasting panel indicate that there 

 was a slight but significant preference for the salmon samples without 

 added oil. From the opinions of individual tasters at each examina- 

 tion, the reason for the preference was the presence of a slight but 

 persistent after-taste, especially in the lots containing the oil pre- 

 pared by boiling the salmon heads. Occasional comments of a similar 

 nature were made regarding samples with alkali processed oil added, 

 but confirmation by the flavor ratings was not as definite in the latter 

 case. 



The sales appeal of the cans of the salmon, as measured by the 

 opinion of the appearance of the can when opened, was invariably in 

 favor of the added oil packs. This was true even for the supposedly 

 oily species, such as king salmon. 



If the added oil was removed fran the canned salmon before the 

 taste panel tested the flavor, rating scores for any of the five spec- 

 ies were not significantly in favor of any of the three lots. 



At the first few examinations the cans of salmon to which the alka- 

 li process salmon head oil was added showed more discoloration of the 

 head space portion of the cans than was evident for either the samples 

 with the oil by boiling added or for the control samples. As the tests 

 progressed, there was a gi^diial increase in the degree of discoloration 

 of the tinplate in the control series and in the series to which the 

 oil made by boiling the heads had been added. By the sixth to eighth 

 examinations there was no consistent difference in the appearance of 

 the tin plate either on the main body of the several cans or in the 

 head space area. The discoloration was in no instance sufficiently 

 pronounced to indicate an abnormal degree of action on the cans by 

 reason of the added oil by either method of preparation. 



97 



