A STUDY OF INFLUENCES WHICH MAY AFFECT THE SEX-RATIO OF THE DEER-MOUSE (PEROMYSCUS)

Data have been presented, based upon over 4,600 deer-mice of known sex, which were born and reared in captivity, under temperature conditions not far different from those existing in nature. The following results seem to be of most importance.

1. In size the broods ranged from 1 to 9, the mean of the 1,567 broods being 3.22.

2. The sex ratio for the entire lot was 97.37 ± 1.93. When we include only those broods in which no deaths are known to have occurred (nearly nine tenths of the whole) the figure becomes 98.01 ± 2.07. Broods known to have been incomplete give a ratio of 93.08 ± 5.25. In order to eliminate the effect of seasonal differences, the mean of the monthly means has been computed. This is 95.65. None of these figures can be regarded as differing significantly.

3. The probable errors employed throughout this paper are based upon a formula different from that which has been used by various previous writers. As a result, the errors here given are about twice as great as would formerly have been computed, and the probabilities claimed by us are correspondingly lower. According to this safer criterion, some of the most interesting of our differences are not decisively "significant."

4. Considering our aggregate material, there seems to be a definite seasonal cycle in the proportion of males and females born. The sex ratio presents two annual maxima, in March-April and August-October, respectively; and two annual minima, in winter and summer, respectively. The lowest figure is 78.79 (± 6.93), in November and the highest 113.04 (± 8.04) in April. This difference is about 3 times its probable error, and would ordinarily be regarded as "significant." A well-marked biennial rhythm is shown by both pure and hybrid stock, taken separately, though the fall maximum occurs in different months in the two cases. On the other hand, when our material is subdivided in certain other ways, the results are in some cases highly contradictory. We cannot, therefore, regard the existence of a seasonal cycle in the sex ratio of Peromyscus as being proved conclusively by our data.

In any case, the position of these annual maxima and minima, as found by us, does not correspond with those which have been reported for the white rat or for man. Indeed, the conditions in Peromyscus are very nearly the reverse of those described by certain other authors.

5. Subspecific hybrids (1,722) give a mean ratio of 104.76 ± 3.41. Mice of "pure" race (2,930) give a mean ratio of 93.27 ± 2.32. Considering these figures alone, the difference is barely significant statistically, but its reality is borne out by various other considerations. Furthermore, such an effect of hybridization upon the sex ratio was to be expected in view of the findings of various other biologists, working upon widely different organisms. It is possible, also, that the several subspecies of Peromyscus differ inter se to some extent.

6. A positive correlation exists in our material between the size of the brood and the sex ratio. Broods containing one to three individuals give a mean sex ratio of 94.85 ± 2.94, ones containing four to ine individuals give a mean sex ratio of 102.42 ± 3.01. The magnitude of these probable errors raises the suspicion, however, that these differences are accidental, and our doubts are further increased by the lack of a uniform gradation when the broods are grouped according to size.

7. When the number of each possible combination of males and females, in broods of each size, is compared with the number expected according to chance, the conformity is found to be, on the whole, very close. For example, the number of all-male and all-female broods (excluding broods of one) was 276, the "expected" number being 274. There is thus no preponderant tendency toward the production of homosexual litters, and thus no likelihood that polyembryony or true twinning is at all common in these animals.

8. In our material, the sex ratio is lower for the earlier broods (91.7) than for later broods of the same mother (103.3). The numbers are so small, however, that the difference is probably accidental.

9. Likewise, inbreeding and outbreeding seem to have had no influence upon the relative numbers of males and females, within the limited material available for this comparison.

10. Similarly negative results were obtained from a comparison of the offspring of meat-fed individuals with the offspring of those whose diet was strictly vegetarian. Here again, the numbers were too limited to permit us to regard this experiment as decisive.

11. The most significant result of all, statistically speaking, and one which is at present utterly inexplicable, is the fact that the sex ratios for the seven different years included in our records show a wide range of variation. The extreme figures are those for 1916 and 1917, the ratios being 125.36 ± 7.82 and 70.56 ± 4.70, respectively. These figures are based upon 471 and 423 individuals, respectively. The difference is 54.80 ± 9.1 (i.e., difference = 6 x P. E.). The likelihood of obtaining such a result by "accident" is less than one in 40,000. We have furthermore determined that this difference is not due either to the seasonal distribution of births, to the preponderance of hybrid births in one year, as compared with another, or to the operation of any of the other factors previously considered.

Mention was likewise made of the occurrence of similar annual differences in the sex ratio of man as revealed by the tables of Pearl and Pearl (1908). Some of these differences we have shown to be much more highly "significant"—according to accepted standards—than the differences between the pure and hybrid races with which these authors were concerned.