PHYSIOLOGICAL PROPERTIES OF THE GONADS. 307 



it is apparent that the gonadectomized groups fall below the curves 

 of the normal groups for the greater part of the year ; but by the 

 3OOth day the spayed females have attained the weight of the 

 normal female and remain slightly above these until the close of 

 the experiment. At the end of the experiment, therefore, the 

 spayed females are heavier than normal females, whereas the cas- 

 trated males are lighter in weight than the normal males. This 

 appears to indicate that the testis does favor growth (absence 

 reduces weight), and that the ovary retards growth (normal 

 females lighter). But let us apply this evidence using as the basis 

 of our comparison four females of the same litter, Yi7, Yi8, Yi9, 

 and Y20. These four sisters, reared together in the same cage, 

 each show gradual growth throughout the year with an entire 

 absence of temporary losses in weight; two had been spayed (Yi7, 

 Yi8), while the other two remained normal females. On the 

 36oth day each of the two spayed animals weighed 800 grams, one 

 normal female (19) weighed 975 grams, and the other (Y2o) 

 weighed 745 grams. Since removal of the ovaries causes a rela- 

 tive increase in weight (when averages of entire groups were 

 compared), the spayed females should weigh more than the nor- 

 mal ones. But referring to Table I., -it becomes apparent that Yig 

 (normal) is 22 per cent, heavier than the spayed sisters, whereas 

 Y2O is 2 per cent, lighter than the operated sisters. Our supposi- 

 tion, therefore, is proven that elimination of the ovary leads to 

 relative increase in weight if the comparisons are made between 

 these operated animals and Y2O, but it is disproven if the compari- 

 sons are made with Yi9 representing the normal condition. Con- 

 sequently, if individual variations in weight are so great with 

 animals of the same litter and lead to such discrepancies in results, 

 one could not expect a more adequate basis of comparison if the 

 animals of unknown history are chosen from the stock at random. 

 One can not fail to be convinced of the magnitude of variation in 

 weight in animals of the same age, sex, and under identical con- 

 ditions if the weights on the 36oth day are compared (Table I.) ; 

 in some cases the individual variation is greater than 100 per cent. 

 If, therefore, this group of total weights demonstrates nothing 

 more (and indeed little is claimed for it), certainly it shows the 

 fallacy of comparing at random the weights of two, three, or four 



