SEX-RATIO OF THE DEER-MOUSE. 149 



Nor can the differences be due to the varying proportions of 

 these two groups which were born in different years. As already 

 mentioned, the difference holds for six of the seven complete 

 years of our records, the exception being unimportant, owing to 

 the small numbers concerned. Likewise a computation analogous 

 to that described on p. 139 shows that distribution by years (on 

 the assumption of an equal sex ratio in hybrid and pure stock) 

 could account for only a small fraction (2.5) of difference actu- 

 ally found. 



Owing to the relatively small numbers, it is hardly worth 

 while to discuss at any length the separate subspecific crosses 

 which were made. The figures for the five different groups, with 

 the numbers (in parenthesis) on which they are based are as fol- 

 lows: 114.63 (176), 114.11 (698), 97.27 (360,95.65 (270) and 

 93.75 (217). Thus, all of these figures are higher than the ratio 

 for the " pure '' stock, though in the last case the difference is 

 trivial, and in only one case is it as much as three times its 

 probable error. Of the smallest of these figures it should be 

 stated, however, that it is based entirely upon mice which were 

 born from May to December, inclusive, thus missing the spring 

 maximum. 



All in all, the evidence, if not wholly decisive, points rather 

 strongly to the conclusion that in Peromyscus hybridization per 

 sc results in increasing the proportion of males which are born. 



Various previous writers have called attention to the larger 

 proportions of males resulting from hybrid matings. To mention 

 but a few of these cases, Guyer (1909) has presented evidence of 

 this sort for various bird crosses, 1 Riddle (1917) for pigeons, 

 Harrison (1919) for lepidoptera, Pearl and Pearl (1908) and 

 Little (1919) for man, King (1911) for rats find mice. In cer- 

 tain of the cases discussed by Guyer and Harrison 2 the proportion 

 of males reached 100 per cent. 



1 In some of the bird crosses discussed by Phillips (1921) the males do 

 not seem to be in excess, but the number of individuals is small, and the data 

 are not presented with a view to answering this question. 



2 Harrison's results are complicated by the different behavior of various 

 reciprocal crosses in respect to the sex ratio, by the preponderance of females 

 in some cases, and by the appearance of " intersexes," i.e., intermediates be-- 

 tween the two sexes. 



