(BL, BS] under the assumption that the half-time of the reparable component was 7 days and that the irreparable injury would be equivalent to 5 per cent of any given dose, showed that such a formulation was only useful under a limited range of exposure conditions. 26, The above limited discussion of a vast amount of literature shows that short-, medium— and long-term mortality recognize different pathogenetic mechanisms whose analysis may not be performed under the same assumptions: there is probably more than one single formula that may take into account the variety of mechanisms. Also, no single recovery time or residual injury value ean define all conditions of protracted exposure. The constants applicable for acute injury may to some extent predict the results of exposures within about 100 days and 1500 rad in the mouse, but for longer times and low doses a new set of relations between injury and recovery must be established [G1]. oT. Neary [N1, N2] also proposed a model, based on the observation that the great majority of animals in a population die during a period in the last part of the life-span. The preceding period of life, in which deaths are rela- tively few, is called "induction" by Neary, who defines it as a state of intracellular changes and intercellular reactions proceeding insidiously and without marked functional impairment. When a certain level of inductive change is reached, the second stage sets in quite abruptly. This stage, called "development" would involve a different level of organization and would be sustained by physiological interactions proceeding autonomously and autocatalitically and culminating in death. The most interesting feature of this model is that once development sets in further inductive change is superfluous and therefore irradiation during development has comparatively little effect. This point was conceptually emphasized by Mole [M6] and by Mole and Thomas [M7] in the notion of "wasted radiation". Neary suggested that small radiation doses would act essentially by shortening induction but without affecting development, and also showed that there was a good correlation between the results of his experiments on CBA mice [N3] and the formal requirements of the model. Kohn and Guttman /|K6; pointed out, however, that the model, derived from experience on duration-of-life exposure, would be applicable with difficultz to acute irradiation.