Corrrlation and Variation in the Toad 607 



namely, that such organs rarely function to the limit of their 

 capacity. It is well known that large portions of many of the inter- 

 nal organs may be removed without causing any serious or some- 

 times even any visible disturbance of the physiology of the organ- 

 ism. The entire spleen, nearly the entire thyroid, or ovary, one 

 entire kidney, and even large parts of the brain may thus be 

 removed without visible effect. This it seems can only mean that 

 ordinarily such organs are functioning only in small part, that 

 they are working with a large margin of reserve; that their func- 

 tional value is not determined by their size. 



J Correlation 

 a Comparative Degrees of Correlation in the Sexes 



In the discussion of this subject we are again limited practically 

 to human data. The relation here is quite similar to that of 

 variability, /. e., there is no uniform difference between the sexes 

 but in general the female is perhaps slightly more perfectly corre- 

 lated than the male. Of 29 pairs of coefficients among individuals 

 of the same societies collected from various sources the females 

 show higher correlation coefficients in but 14 and the average 

 degree in the female is .442 as compared with .439 in the male — 

 probably a non-significant difference. In swine (Davenport and 

 BuUard '96) the coefficients between the numbers of Miillerian 

 glands on the right and left sides of the body are .783 — female 

 and .772 — male, a barely significant difference. Among the 

 fishes the difference is somewhat more marked. Duncker (Ver- 

 non '03) states that in 40 pairs of coefficients 17 were unaffected by 

 sex while in 1 1 the males, and in 6 the females were the more highly 

 correlated. In the toad however, in correlation as in variability, 

 there is both a decided and a uniform difference between the sexes. 

 In 63 of the 66 pairs of coefficients (in two of the three exceptions 

 the difference is less than the probable error) the female coeffi- 

 cients are higher than those of the male, the average degrees being 

 .779 and .727, respectively. 



We see then that in the toad the females are about 10 per cent 

 better correlated than the males. Here then we have a relation 



