SEX-RATIO OF THE DEER-MOUSE. 145 



particularly striking when we compare the actual and expected 

 totals for all of the broods in which all members were of the same 

 sex. The actual number of such homosexual litters, among 

 broods containing from 2 to 6 individuals inclusive, was 276. The 

 most probable number .on the assumption of purely random sex- 

 production was 274. x Thus, not only do we find approximately 

 equal numbers of males and females in the population as a whole, 

 but in single broods the distribution follows the laws of chance. 

 There is no tendency for fetuses (or germ cells) developing in 

 the same parents at the same time to give rise to organisms of 

 the same sex. 



It may seem, on first thought, that such evidence is conclusive 

 against the efficacy of any factor, except the chance meeting of 

 the gametes, which might be supposed to play a part in determin- 

 ing the sex ratio. On the assumption, for example, that seasonal 

 influences of an undetermined character may affect this ratio, 

 should we not expect undue proportions of all-male broods at 

 one time of year, and of all-female broods at another time, and 

 should not this fact result in a preponderance of homosexual 

 broods throughout the year as a whole ? We have not computed 

 the proportions of all-male and all-female broods by months, but 

 the probability of the excessive production of such broods at cer- 

 tain times of the year may be granted. We must, however, point 

 out the equal probability of an excessive production of evenly 

 balanced broods at other times of the year. During these months 

 of average sex ratios, we should presumably have not only equal 

 numbers of males and females produced /// the aggregate, but a 

 tendency toward balanced broods on the part of individual 

 mothers. The possibility of sex-determining agencies other than 

 jhance combinations of the gametes in fertilization is quite un- 

 affected by these results. 



One important conclusion seems justified, however, by this 

 utter absence of any tendency toward the preponderant produc- 

 tion of homesexual litters. This is the non-occurrence of poly- 

 embryony or true twinning, at least with sufficient frequency to 

 affect the results. 



i If we include fractions (a more exact procedure) this number becomes 

 277. Surely this is a close "fit! " 



