trols in the age interval from 1 day to 3 months; it was then gradually decreased to 32, tively. 14 and 7 per cent of controls at 6, 12 and 18 months of age, respec- Thus, while the acute response to radiation appeared to increase, the life-shortening response decreased as a function of age. This effect cannot be traced to the incidence of leukeamia or of ovarian tumours which were evaluated separately in these experiments. be different. In other strains the situation may, however, In the C57BL mouse, for example, age is a factor which alters the sensitivity to lymphoma induction. Kaplan [K15!/ showed in fact on animals at 2 weeks and at 1, 2, 3 and 4 months of age that lymphoma incidence is higher and the appearance time earlier at the young ages than at the older ones. 266. Upton, Kimball, Furth et al. [U5] considered in detail the variable "age at exposure" in relation to the life-shortening effects in both sexes on 9-12 week old mice. No differences were found of any significance for either sex and this observation was at variance with what was seen in the same animals in regard to early mortality, which was highest among young animals. Moos [M1] was also unable to find differences between the longevity of young (40-50 days) or o1d (140-150 days) mice within the daily range of doses of 8 to 128 R/day. How- ever, such differences were seen at exposure rates of 2, 4 and 256 R/day, and the old mice were less resistant under these conditions. The greater variability of the radiation response in the young mice and the increased frequency of natural death in the older animals could account for the observed differences. 267. In a series conducted on about four thousand RF/J female mice Storer [519] examined the age-dependent changes in radiation sensitivity in normal and previously-irradiated animals. He obtained life-table data on all these mice, about one half of which received at the age of 90 days LOO rad of 250 kVp x rays. The rate of mortality from all causes in the irradiated mice showed marked departures from the Gompertz equation and this dose of radiation shortened the median lifespan to 63 per cent of the control animals. Mortality rates for all causes other than leukaemia gave reasonably good Gompertz fits to the control and irradiated populations: life-shortening amounted to 24 days/100 R under the assumption of linearity between dose and effect. No latent period was found between exposure and the time when the mortality increase became detectable. The mortality rate of irradiated mice was at all times higher than in control animals. The 1D 59/30 of the control and irradiated mice tested between 120 and 560 days of age declined linearly with age without effect attributable to the previous hOO R exposure. The mean after-survival following exposure to 100 R/day also declined