which yielded the following solution y The 92. Ro = 0.531 + 0.0638 D - 1.94 107? De (8) of this fit was 0.73. It may be concluded that under the above conditions the life-shortening data for the mouse may be considered to follow a linear non-threshold rela~ tionship as a function of the x- and gamma-ray acute dose, indicating a lifeshortening efficiency of about 5 per cent/100 rad down to the smallest doses. The data may, however, also be fitted by a linerar + quadratic function, where the quadratic term is negligible and not such to give rise to an appreciably different relationship within the errors of such an analysis or to substan- tially different quantitative conclusions. 93. Analysis of the data in the various experimental series and an inspec- tion of Table 1 makes it quite obvious that in any given instance the doseeffect relationship for life-shortening may be linear or curvilinear (with upper concavity or convexity). The actual shape of any such curve depends on the interplay of the biological variables (strain, sex, age) with dose, giving rise to a different spectrum of life-shortening diseases or pathological con- ditions at the various doses. That the combination of a variety of experi- ments should produce a linear relationship cannot therefore be considered to depend on any particular biophysical law, at this stage of the analysis. It may simply reflect the fact that when all experimental conditions and all the resulting life-shortening effects are averaged over a number of different series, they combine by chance to produce an approximate linear relationship with dose. Therefore, taken as such, this observation may have no special meaning in the interpretation of the life-shortening action, but may be regarded as a very interesting observation in practice. It shows, in fact, that in a highly non-homogeneous mammalian population where all ages, sexes and strains are represented, the use of a linear function to describe the doseeffect relationship for acute exposures to x- and gamma-rays is not an unreasonable proposition. gh. The data obtained in the mouse by neutron irradiation (see Table 2) were similarly plotted on a common graph as in Figure III which includes doses up to about 500 rad and life-~span-shortening effects up to about 50 per cent