SEX-RATIO OF THE DEER-MOUSE. 1 29 



January (185) 88.78 + 8.92 



February (287) 90.07 + 7.04 



March (722) 106.29 + 5.14 



April (343)- 113.04 + 8.24 



May (506). 94.62 + 5.63 



June (405) ., 87.50 + 5.80 



July (347). 91.71+6.57 



August (565) 103.97 + 5-84 



September (445 ) 96.04 + 6.13 



October (406) 107.14 + 7.17 



November (236) 78*79 + 6.93 



December (205) 89.81 + 8.37 



These figures and the accompanying graph (Fig. i) reveal the 

 existence of two annual maxima, one occurring in March and 

 April, the other extending from August to October. Alternating 

 with these are summer and winter minima. 



It will be seen that the differences between the successive 

 months, taken by themselves, are of very doubtful significance, 

 while even the difference between the highest and lowest months 

 of the year (April and November) is only a little more than three 

 times its probable error. With the exception of September, how- 

 ever, the seasonal cycle presents a perfectly consistent picture, 

 there being an almost uninterrupted rise and fall twice annually. 



The dotted line in the figure is based upon the date of concep- 

 tion, instead of the date of birth. The date of conception has 

 'been computed, for each brood, by subtracting 22 days from the 

 date of birth, this being the usual period of gestation. 1 Such a 

 procedure has inevitably resulted in the transference of part, 

 though not all of the broods to the month immediately preceding. 

 Were the period of gestation exactly one month, the one graph 

 would be an exact duplication of the other. As it is, they present 

 some obvious differences of form. The chief of these is the 

 division of the fall ''maximum'' into two entirely distinct peaks. 



Since, for various reasons, it does not seem probable that the 

 differences between the consecutive months should be regarded as 

 significant, we may profitably combine our monthly birth records 

 into four seasons of three months each. Reference to the table 

 of monthly ratios shows that we may distinguish two high periods 



1 Sumner, 1916. 



