SEX-RATIO OF THE DEER-MOUSE. 147 



amount which is less than half of its probable error. These two, 

 however, particularly the latter, differ from rubidus by amounts 

 which are possibly significant. But it is not certain that these 

 figures imply the existence of actual racial differences in respect 

 to sex ratio. One thing may be stated with confidence, however, 

 namely, that the high ratio shown by rubidus is not due to any 

 peculiarity in the seasonal distribution of its births. This has 

 been shown by the following procedure. The sex ratio proper 

 to each month, as computed for the pure races in general, has 

 been weighted by the number of rubidus born in the correspond- 

 ing month, and the product-sum divided by the total number of 

 mice of this subspecies. The mean ratio thus obtained is 92.77, 

 a figure even lower than the ratio for all mice of pure race. 



The existence of racial differences in the sex ratio have been 

 pointed out for man. According to Newcomb (1904) and Heape 

 (1909) the ratio is higher for the white race than for negroes. 

 Heape (1907) likewise gives different figures for different races 

 of dogs, these figures ranging from 96 to 136. There is little 

 analogy, perhaps, between the artificial " races " of dogs and the 

 geographic " races " of wild mammals, and the comparison with 

 man is of equally doubtful validity. 



HYBRIDIZATION. 



Of the 4,652 young of known sex, 2,930 belonged to " pure " 

 subspecies, 1,722 to subspecific hybrids. The former were not, 

 of course, pure, in the sense of having been closely inbred for 

 many generations, or of being uniform in genetic composition. 

 Their purity was relative. Among the hybrids are included, not 

 only those of the F 1; F 2 or later generations, but various back- 

 crosses with the parent stocks as well. 1 This procedure seems 

 warranted, according to any theory of the behavior of subspecific 

 characters in hybridization. There is no clear segregation of the 

 parental stocks in the F 2 generation, as would be expected if 



1 Only subspecific crosses are here included. Crosses between "mutant" 

 races within a subspecies, or between one of these and the parent stock, have 

 not been regarded as hybrids for the purposes of this study. Of these, how- 

 ever, the number is not sufficient to materially affect the ratio for the "pure" 

 stock, even though such an effect were known to occur. 



