VITAMIN A 



255 



Quantitative Determination of Relative Amounts of Vitamin A by 

 Means of Feeding Experiments. — The foundations for quantitative 

 work in this field were laid by Osborne and Mendel (1920c, 1921), by 

 Steenbock and associates (Steenbock, Boutwell and Kent, 1918; Steen- 

 bock and Gross, 1919; Steenbock, Nelson and Black, 1924) and by 

 Drummond and Coward (Drummond, 1919; Drummond and Coward, 

 1920). 



In 1919 Drummond described the technique used in his laboratory 

 to compare the vitamin A values of substances quantitatively as follows : 

 Young healthy rats selected from home-bred stock and weighing about 



120 



100 





ZO 



ZO A.0 60 80 



TIME. IN DAYS 



Fig. 10. — Showing the influence of previous feeding upon growth on diets lack- 

 ing vitamin A. Rat 4217 had received a diet relatively low in vitamin A (% 

 whole wheat and % dried whole milk). Rat 4272 had received a diet richer in 

 vitamin A (% whole wheat and % dried whole milk). Rat 4269 had received a 

 diet still richer in vitamin A (i/^ whole wheat and % dried whole milk). All were 

 placed upon a diet devoid of vitamin A at the same age, 28 days. Growth and 

 duration of life upon this diet was evidently determined by the preexperimental 

 bodily store of vitamin A, as determined by the richness of their previous diets 

 in vitamin A. (From experiments by Sherman and Boynton.) 



50 grams each, are fed a ration consisting of purified caseinogen 20, 

 purified starch 50, salt mixture 5, yeast extract 5, butterfat 15 and 

 filtered orange juice 5 parts. The animals which give evidence of a 

 normal power of growth are removed from the complete ration when 

 they have attained a body weight of 70 to 80 grams, and are given a 

 similar ration in which an equivalent amount of hardened linseed oil 

 is substituted for the butterfat. When it is definitely established that 

 growth is inhibited by this deficiency in vitamin A, the linseed oil is 

 wholly or partially replaced by the substance to be tested, and the be- 



