274 F - B - SUMNER. 



ever, the period during which the male had access to the female, 

 and the date of discovery o.f the brood. During the times 

 when frequent births were expected, the nests were examined 

 daily. Hence a given birth must commonly have occurred 

 between two known observations. It has been my invariable 

 practice to refer the birth to the day when the brood was dis- 

 covered, 1 except where circumstances were such as to make it 

 probable that the young were bprn a day or two earlier. In such 

 cases, this probability is indicated in the records. 



Since, in a large series of experiments, it is probable that 

 successful insemination would occur in some cases pretty soon 

 after the males were introduced into the cages, it seems likely 

 that the minimum interval between the introduction of a male 

 and the birth of a brood represents the actual period of gestation, 

 at least for the individuals in question. This minimum was 

 found to be 22 days. In all but five cases, however, out of a 

 total of 206 broods, for which I have accurate records, the figure 

 was greater than this. As a control, I have determined, for each 

 series, the maximum interval between the removal of the male 

 and the discovery of a brood of young. This maximum figure 

 is also a fraction over 22 days. In the single instance recorded, 

 a brood was known to be born as late as some time during the 

 23d day after the last opportunity for copulation. 



From these data, therefore, we know that the period of gesta- 

 tion of Peromyscus does not, in some cases at least, exceed 22 

 days. We also know that it may be as much as 22 days. Ac- 

 cordingly, I have provisionally adopted this figure as representing 

 the normal term for the species which I am engaged in breeding. 

 It is likely, however, that this term is subject to some variation. 



Tables I. and II. comprise two different classes of cases. In 

 the former, the interval between the birth of the first and second 

 litters ranges from 13 to 39 days. In all of these cases the youth 

 of the first brood of young, at the time of the second conception, 

 precludes the possibility that a male of the former lot was the 

 father of the second. Even when the interval was as great as 

 39 days, conception must have occurred when the first lot was 



1 Mice (at least white mice) are said by Long and Mark (1911) to be more 

 commonly born in the early morning than at any other time of day. 



