Reproductive activity might have been measured in a number of ways. Three indicators were tested: the percentage of females gravid or lactating, the percentage of juveniles in the month following the sampling period, and the percentage of females with a vaginal orifice in the sampling period. The condition of vagina proved to be most useful. Whether or not there is a vaginal cycle in _Microtus_ is uncertain. Bodenheimer and Sulman (1946:255-256) found no evidence of such a cycle, nor did I in my work with laboratory animals at Lawrence. How much the artificial environment of the laboratory affected these findings is unknown. The presence of an orifice seemed to indicate sexual activity (Hamilton, 1941:9). The percentage of gravid females in the population could not be determined accurately by a live-trapping study and was not useful in this investigation. The percentage of juveniles trapped in the month following the sampling period tended to follow the curve of the percentage of adult females with a vaginal orifice. The ratio of trapped juveniles to adults trapped was a poor indicator of reproductive activity. Juveniles were caught in relatively small numbers because of their restricted movements, and no way to determine prenatal and juvenal mortality was available. Reproductive activity continues throughout the year. Within the thirty-month period for which data were obtained, December and January showed the lowest percentages of females with vaginal orifices (Fig. 8). The other months all showed higher levels of reproductive activity with a slight peak in the August-September-October period in both 1950 and 1951. In the species of _Microtus_ that are found in the United States, such summer peaks of breeding seem to be the rule (Blair, 1940:151; Gunderson, 1950:17; Hamilton, 1937b:785). Jameson (1947:147) worked in the same county where my field study was made and found that the high point of reproduction was in March, although his samples were too small to be reliable. The peak of reproductive activity slightly preceded the highest level of population density in each year (Fig. 8). [Illustration: FIG. 8. Variations in density and reproductive rate of voles, with variation in monthly precipitation. Abnormally low rainfall in 1952 caused a decrease in breeding activity and eventually in the numbers of voles. The solid line indicates the number of voles per acre, the broken line the percentage of females with a vaginal orifice and the dotted line the inches of rainfall.]