The variation in the number of eggs shed by the two ovaries of mice has been examined by statistical analyses of 697 egg counts and 390 corpora lutea counts, made on mice from a variety of outbred strains, both after natural oestrus and after oestrus induced in adults by pregnant mares' serum (pms) and human chorionic gonadotrophin (HCG). The numbers of eggs or corpora lutea were distributed between sides approximately at random, the variation conforming fairly closely to a binomial distribution. This was true even after superovulation. There was, however, a slight but significant excess of variation between sides over the random amount in the egg counts, particularly after natural ovulation. Corpora lutea counts differed from egg counts in showing a slight but significant reduction of the variation below the random amount. Several possible reasons for these small deviations from a random distribution are discussed.
The correlation between the numbers of eggs shed by the two ovaries was negative after natural ovulation but positive after superovulation. This difference can be fully accounted for by the random distribution between sides together with the differences of mean and variance between natural ovulation and superovulation. The variation of total egg number was proportional to the mean egg number after natural ovulation. The variation after superovulation was much higher than after natural ovulation, even when the difference of mean was taken into account, and the greater variation of total egg number caused the correlation between sides to be positive after superovulation.