VITAMIN A 241 



is approximately proportional to the vitamin A values only when the 

 oil is used at very low concentrations, while Towle and Merrill (1928) 

 indicate that the working temperature influences the rate of the develop- 

 ment of the blue color and its change to red. At 10° C. or below the 

 rate of change is about one-twentieth of that at 25° C. Jones, Briod, 

 Arzoomanian and Christiansen (1929) reported that five samples of 

 cod-liver oil out of 13 responded to colorimetric assay within 15 per 

 cent of the value determined by the biological method, while of the 

 remaining 8, 6 exceeded the biological assays by 20 to 600 per cent. 

 Drummond and Morton (1929) attached little significance to these 

 latter results as they were not assured that sufficient vitamin D was 

 furnished to the test animals during the period in which they were 

 depleted of their vitamin A reserves. This should tend to render the 

 biological values too high rather than too low, however. They dismissed 

 the adverse criticisms of Steudel (1927, 1929a) regarding the specificity 

 of the blue color reaction for vitamin A on the grounds that the 

 antimony-chloride test is not sensitive enough to be employed with 

 preparations as poor in vitamin A as was Steudel's product. In investi- 

 gating relatively poor source materials they state that it is necessary to 

 concentrate the chromogenic substance in a non-saponifiable fraction 

 before applying the color test. Hawk (1929) found that samples of oil 

 exposed to the atmosphere gave a deeper blue color with antimony 

 chloride than did those kept in the dark to preserve their vitamin 

 potency. These results are not concordant with those of Drummond and 

 Morton, who have "not encountered a single instance in which there has 

 been disagreement between the animal tests and the intensity of the blue 

 colour showing maximum absorption near 608 ////." 



Similarity in color of the reaction products of antimony trichloride 

 with carotene and with liver oil or other vitamin A concentrates of 

 animal origin proved a stumbling block in the acceptance of the color 

 test as specific for vitamin A. Duliere, Morton, and Drummond (1929) 

 found, however, that the absorption band of the blue reaction product 

 with carotene was at 590 fijn, while that of the product obtained with 

 liver oil was at approximately 610 /nfi. Spectrographic analyses of the 

 color products, as well as of the original materials, were thus considered 

 essential in determining the specificity of the color test. In the opinion 

 of Moore (1929a), "for the present it seems safe to assume that mate- 

 rials which give no blue coloration with antimony trichloride, even after 

 the removal of saponifiable matter, must be devoid of vitamin A activity. 

 Materials of liver oil origin giving color reactions characterized by 

 absorption at 610 fi/j. may be considered active. But, on the other hand, 



