however, did not superimpose except at doses below 100 rad. Thus, not all of the observed life-shortening, particularly at the high supralethal doses, can be explained by the induction of leukaemia, as in male RF animals. A possibi- lity does remain (but it was not examined in a most recent publication, |C15}) that ovarian tumours occurring in 50 per cent or more of the animals might account for the extra life-shortening remaining after subtraction of leukaemia. Th. In his 1975 review Walburg [W1] refers to data in the male RFM mouse ex- posed when 5 - 6 weeks old to a single acute treatment of 300 R of 300-kVp x rays. Routine histopathology allowed the assessment of causes of death with reasonable accuracy and the data were corrected for competing probabilities of death. The cumulative mortality curves for all causes showed significant life- shortening; when death attributable to leukaemia were excluded the cumulative mortality curves of the control and of the irradiated mice became superimposable, suggesting that radiation did not significantly induce or accelerate under these conditions other non-specific causes of death. 75. Ainsworth et al. [AY] irradiated male and female B6CF1 mice with single doses of gamma rays (90 to 788 rad) and found that life-shortening had a reasonably linear dose-response. In the case of fission-spectrum neutron irra- diation, however, the shape of the dose-response curve (20 to 240 rad) appeared to be convex upward. Possible explanations for this shape of the relationship were suggested: they will, however, remain unclear until the causes of death will be completely worked out. 76. Data of life-span-shortening induced by single acute exposures of 250 kVp x rays (100 to 900 R) were obtained by Maisin et al. periments on the effects of chemical protectors. [M8] in the course of ex- Data refer to male Balb/c mice (4-12 weeks old) and to male C57B1 mice (1 to 3 months old, 350 and 650 R). An essentially linear decrease of the life-span with dose was found for the two sets of data: TT. ted by the numerical values are given in Table 1. Pathological observations in the course of these experiments were evalua- the method of competing risks (M10) with a classification of the causes of death comprising various forms of leukaemia and solid tumours, glomeruloscle- rosis, non-neoplastic lung lesions and others. In the non-irradiated Balb/c animals tumours were mainly responsible for deaths, while in the normal C57B1