and quartered to give a representative sample of a few grams. The lat- 

 ter samples were analyzed by the standard St ansby -Lemon acetone-ether 

 extraction method, using a Bailey-Walker extraction apparatus. At the 

 end of the centrifuging process as much as possible of the oil was 

 flushed over with hot water. The liquor remaining in the bowl of the 

 DeLaval Separator were centrifuged in bottles at 2000 rpra. The oil 

 layer thus obtained was added to the clear oil obtained from the DeLaval 

 Separator. The total weight of oil was taken as the oil yield. 



Determination of Vitamin A in Oils 



In addition to determining the wei'^ht of oil recovered in each test, 

 the vitamin A content of the recovered oil was determined by chemical and 

 spectrophotometric methods. 



All oil samples were analyzed by the Carr-Price Blue Color method 

 as detailed in "Methods of Vitamin Assay" (3). This procedure includes 

 the "increment technique" for comparing an unknown with a known standard. 

 This technique tends to correct for the presence of substances in the 

 salmon oil which modify the intensity of the blue color. A Coleman Uni- 

 versal Spectrophotometer, Model No. 11, was used to determine the trans- 

 mi ttance at 620 mrau. of the oil solution treated with antimony trichlor- 

 ide. A vitamin A concentrate prepared by the Distillation Products, Inc. , 

 laboratories was used as the known standard. 



Twenty- five of the oil samples were also tested by the usual spec- 

 trophotometric method, determining the transraittance at 328 mmu. with a 

 Beckman Spectrophotometer. A conversion factor of 1894 was employed to 

 convert E values to US? units. 



In order to give some indication of the validity of the spectro- 

 photometric results at 328 nmu., the E values at 300 mmu. and 350 mmu. 

 were also determined on three samples. The ratios of optical densities 

 at 300 and 328 mrau. and the ratios at 350 and 328 mmu. were calculated: 



Sample No. E 300/328 E 350/328 



12 0.79 0.61 



18 0.93 0.60 



28 0.79 0.63 



The vitamin oil industry customarily accepts E value ratios 300/328 of 

 less than 0.72 and E value ratios 350/328 of 0.65 to indicate vitamin 

 A potency. Only one of these two ratios is satisfactory in the case of 

 the saLuon oil under investigation. This indicates the presence of mod- 

 erate amounts of interfering materials. It is believed the Carr-Price 

 blue color results are perhaps mor-e accurate than the spectro])hotometric 

 since they are somewhat lower. 



103 



