on the body size or on the size of the eyeball; the susceptibility to the induction of bone tumours, which might be equal in all mammals; and the radiation sensitivity of mammalian oocytes, which may be inversely related to the metabolic activity or to the degree of lampbrush configuration of the chromosomes. Gene- ralizations of such specific biological phenomena, rather than extrapolation of abstract matters like mortality or life-shortening should be particularly pursued. He further expanded these concepts [M3] and pointed out that in principle lifespan-shortening may be considered a meaningful parameter if and only when the spectrum of diseases in different animal populations receiving various doses is found to be the same. In the absence of this condition, life-shortening becomes simply a compounded but imprecise way of expressing differences in the incidence of pathological conditions that might be more adequately expressed otherwise. 235. An inter-species comparison of response in mice [S4, G6] and dogs con- tinuously exposed to 605, gamma-rays was reported by Norris, Tyler and Sacher (N7]. The comparison was carried out in terms of the radiation-specific death- rate [S14, S29] (see also Figure VIII) e = (1/t,.) - (1/t,) (29) being the difference of the reciprocals of survival times for the exposed and the control animals. It was found in both species that plotting the log of the radiation-specific death-rate against the log of the dose-rate (rad/day) gave rise to a dose-response with a slope of 2, indicating that the death-rate in- creased with the square of the dose-rate (Figure XVI). The phenomenon was seen over the whole range of dose rates in which damage to the haemopoietic tissues is the primary cause of death, that is above 20 rad/day in the mouse. At lower dose rates in this species the slope change to 1, indicating that injury was only a function of the total dose accumulated and independent of the rate at which it was given. Data available in the dog would suggest a similar inflec- tion taking place below 3.5 rad/day, but this suggestion will have to be proven by appropriate experimentation. The only point available for dogs below 1 rad/ day [C16] lies quite close to the curve for the mouse sensitivity of the two species at the level where injury becomes independent of dose-rate. 236. Grahn, Sacher, Lea et al. [G16] returned again to the problem of extra- polation from mouse to man on the assumptions [G6] that: (a) the Gompertz slopes for mouse and man are in the inverse ratio of their susceptibility; (bo) the previously-mentioned ratio of 30:1 for mouse : man applies; and