VITAMIN A 263 



might be slightly preferable to work entirely with animals of the 

 same sex. 



Probably any possible slight advantage theoretically attributable to 

 discrimination between the sexes will actually, in experiments of this 

 kind, be outweighed by the advantages of being able to make greater 

 use of litter mate controls when both sexes are employed. 



The very slight tendency toward more rapid growth on the part 

 of males than of females of the same age and nearly the same size, will 

 in practice tend to be offset by the larger size which the males will 

 usually have attained at the beginning of the test period, when the 

 animals will usually be about 2 months old. This equalizing influence 

 can readily be further availed of, at the cost of rejection of very few 

 animals, by specifying a slightly larger standard size for male than 

 for female rats at the end of the depletion and beginning of the test 

 period. Any laboratory might, for example, decide that only females 

 between 70 and 95 grams or males between 75 and 100 grams should 

 be used interchangeably as standard test animals for quantitative deter- 

 minations of vitamin A by the rat growth method. 



It should, however, be carefully noted that all the foregoing dis- 

 cussion of the relation of sex to rate of gain, refers to experiments 

 in which the level of feeding of the vitamin A-containing food or other 

 material to be tested is such as to induce a gain in weight of only about 

 3 grams per week during the test period. Much larger differences are 

 to be expected if the level of vitamin A feeding is considerably higher, 

 for in this case something approaching a normal rate of growth may 

 become possible, and, as is well known, the normal growth curve of 

 the male rat is distinctly steeper than that of the female at this age. 

 Thus when Sherman and Burtis (1928a) fed about twice the amount 

 of vitamin-A-containing food required for 3 grams gain per week, the 

 average gain during the 8-week test period was: for 117 males, 57 

 grams; for 117 females, receiving the same amount of vitamin A, 

 42 grams. 



Drummond and Coward (1920) and Zilva and Miura (1921b) 

 pointed out that a definite small gain in weight furnishes the best basis 

 for quantitative comparison. Further work has resulted in the general 

 adoption of an average gain of about 3 grams per week during an 

 experimental period of five to eight weeks, as a standard rate of gain 

 or "unit growth." A lower rate of gain has been found to give more 

 variable results, some animals dying before the completion of the test 

 period. 



"Normal growth" is undesirable as a basis of quantitative compari- 



