256 THE VITAMINS 



havior of the animal is closely watched for a period of from 5 to 6 

 weeks. Absence of vitamin A is indicated by the failure to grow fol- 

 lowed by a decline in health accompanied by the characteristic eye 

 condition. That there was an individual variability in reaction to this 

 standardized technique was noted in the observation that whereas 

 experimental animals usually soon cease growing when fed this diet 

 deficient in vitamin A, "occasionally a very vigorous individual will 

 continue to grow for some weeks after the deficiency has been intro- 

 duced, and such animals should not be employed in experiments of a 

 quantitative nature." 



Certain improvements in this technique were soon reported by 

 Drummond and Coward (1920). Rice starch was substituted for the 

 wheat starch as the former in the crude state was found to be almost 

 devoid of vitamin A, and hence could be more readily purified than 

 the wheat starch; and refined fully hydrogenated vegetable oil (usually 

 cottonseed) replaced the linseed oil. On this ration it was reported 

 that young rats of 50 to 70 grams weight should show only slight 

 growth, becoming stationary after a week or two. Definite Vv'eights of 

 the substances to be examined for vitamin A were administered daily 

 before the basal ration to rats, the growth of which had been suspended 

 for from 10 to 14 days, and which did not weigh more than 80 to 120 

 grams. Later workers have confirmed Drummond's view that reasonable 

 uniformity of size of experimental animals is important to the quanti- 

 tative accuracy of the results. 



The quantitative method thus introduced by Drummond and Coward 

 has been developed by Sherman and Munsell (1925) and Sherman and 

 Burtis (1928a) into the following form: 



Feeding Method as Used in 1930 



The Basal Diet. — Much attention has been given to the develop- 

 ment of a basal diet which shall be free from vitamin A but fully 

 adequate in all other respects and which can be duplicated with cer- 

 tainty. 



As the protein component of our basal diets we have used both 

 casein and meat residue with satisfactory results. In either case, our 

 experience leads us to depend upon thorough extraction with hot alcohol 

 for removal of vitamin A rather than to attempt to free the material 

 completely from this vitamin by heating in the air. Casein is perhaps 

 preferable to meat residue as a component of a standardized basal diet 

 because it is more definite chemically. We conduct the extraction of 

 the casein essentially as described by Osborne and Mendel (1921). The 



