258 THE VITAMINS 



eluded in the basal diet, as our animals (containing considerable body 

 fat) seem to thrive as well without it (for the time covered by these 

 tests, at least) and its omission avoids any possible uncertainty as to its 

 freedom from vitamin A. 



The basal vitamin A-free diet which we now use consists of purified 

 casein 18, salt mixture 4, dried brewers' yeast 10, sodium chloride 1 

 and cornstarch 67 parts with the addition of 0.500 grams irradiated 

 commercial cholesterol per kilogram of the basal ration or some equiva- 

 lent form of irradiated ergosterol. 



Selection and Standardization of Experimental Animals. — The ex- 

 perimental animals are young albino rats reared under the supervision 

 and control of the experimenter. Sherman and Kramer (1924) found 

 that animals reared on food mixtures of dry whole milk and ground 

 whole wheat with increasing proportions of milk and therefore of 

 vitamin A, possessed correspondingly increased body stores of vitamin 

 A, even at four weeks of age, as was apparent from the increasing 

 average survival periods, when placed on a vitamin A-free diet. Along 

 with the difference in survival period, the dififering store of vitamin A 

 in the body resulted also in markedly different growth curves on the 

 same vitamin A-free diet. Hume and Smith (1928) suggest that vita- 

 min A potency can best be evaluated by prolonged tests, dispensing 

 with a depletion period. But unless the factor of varying storage in 

 the body of the young rat is very carefully safeguarded it is likely to 

 be an extremely important source of error in studies of vitamin A and 

 especially in attempts to make such studies quantitative. 



Because of the large differences shown by rats of the same age 

 when placed on the same diet, partly individual and partly due to 

 previous diet or family history, we follow the practice of Drummond 

 and his coworkers in beginning each experiment by keeping the rat 

 upon a vitamin-A-free diet until his stored surplus of vitamin A is 

 exhausted as indicated by cessation of growth, before beginning the 

 feeding of the food to be tested. 



Too large a bodily store of vitamin A is inconvenient, for during a 

 long preliminary period the animals become too large to be the most 

 sensitive reagents to measure growth. On the other hand, a previous 

 diet so restricted as to result in a very small bodily store also appears 

 to be undesirable; Sherman and Burtis (1928) have shown that the 

 young reared to 28 days by mothers receiving a ration of % whole 

 milk powder and % ground whole wheat are more susceptible to infec- 

 tion during the vitamin A test period than those reared by mothers 

 receiving twice as much milk, even though the variations in the surplus 



