148 F. B. SUMNER, M. E. MCDANIEL AND R. R. HUESTIS. 



these stocks differed by only one or by a few pairs of unit factors. 

 In consequence the F 2 and back-cross individuals are in a true 

 sense Jiybrids. 



The proportions of the sexes in the two lots are as -follows : 



The difference between these two ratios is 11.49 4.1, thus 

 being slightly less than three times its probable error. 1 These 

 figures by themselves, therefore, cannot be held to prove at all 

 conclusively the existence of a higher sex ratio among subspecific 

 hybrids in Peromyscus. But the probability of such a difference' 

 is greatly increased by several circumstances, (i) It has already 

 been shown (Fig. 4) that the "hybrid" ratio exceeds the "pure" 

 in nine months out of twelve, and that in two of the three excep- 

 tional cases the differences are trivial. (2) The hybrid ratio was 

 higher in six of the seven years, taken separately (Fig. 7), and 

 the single exception here is based upon a very small number of 

 individuals. (3) The ratios for all of the five separate groups 

 of hybrids are larger than that for the pure stock. (4) Such a 

 difference accords well with the bulk of the evidence for other 

 species of animals. 



It is quite unlikely that the difference here found is due to any 

 of the other factors with which the sex ratio of Peromyscus 

 has been found to be correlated. As regards season, it has al- 

 ready been pointed out that this difference between the hybrid 

 and pure material holds for nearly all the months of the year. 

 Likewise the difference is very nearly the same when we deal, 

 in each case, with the mean of the monthly sex ratios. 



Again, as regards the size of the brood, the mean difference, 

 in this respect, between the hybrid and pure stocks (3.26 and 

 3.20) is such that it could not exercise an appreciable influence 

 upon the sex ratio. 



1 It may be pointed out that when this probable error is computed accord- 

 ing to a method widely used (see p. 127), the difference is more than five 

 times the latter. 



