126 RADIATION BIOLOGY 



fact that the plot of many survival curves remains straight over a large 

 range of variation of the dose indicates a very substantial homogeneity 

 of the population with respect to radiation resistance. 



It should be realized that the interpretation of the rate of action as a probability 

 implies from the beginning that the population tested is substantially homo- 

 geneous. "Probability" means the expected rate of occurrence of an event in a 

 series of trials performed under identical conditions. (The exposure of each 

 individual organism constitutes one "trial"; "identical conditions" require sub- 

 stantially identical properties of the organisms.) The experimental fact is that, 

 after identical irradiation, the "event" under consideration occurs in some indi- 

 vidual organisms and not in others. These different reactions might con- 

 ceivably reflect wide preexisting differences of radiation resistance among the 

 organisms of the population rather than a statistical variability of response of 

 identical individuals. 



Two factors support the interpretation of the exponential curves in terms of 

 some statistical variability inherent in the mechanism of radiation action. The 

 general physiological inhomogeneity within cell populations, however consider- 

 able it may be, does not appear to be as broad or as skew as would be required to 

 account for the exponential dose-action curve. Furthermore, if physiological 

 variability had a great influence on the dose-effect curves, this variability would 

 not seem likely to follow a reproducible pattern and to yield again and again the 

 same simple exponential dose-effect curve. 



The theoretical significance of exponential dose-effect curves hinges 

 most critically on accurate verification of the law for very small doses. 

 The concept of a fixed probability of radiation action implies that there is 

 a finite chance of detecting an "event" among a large population, even 

 for low doses where the probability of that event is exceedingly small. 

 If a well-defined macroscopical effect can result from a comparatively 

 minute treatment, the capricious occurrence of effects in different individ- 

 uals appears to reflect an inherent feature of the mechanism of radiation 

 action. At the same time each bit of radiation, a single particle or pho- 

 ton, appears capable of producing an observable effect. 



Very small lethal effects of radiation cannot be measured accurately 

 under ordinary conditions because the number of surviving organisms, 

 i.e., the nonoccurrence of a lethal effect, is scored rather than the effect 

 itself. Figure 1-79 show^s an example of the degree of approximation 

 with which the shape of a dose-effect survival curve for low doses may be 

 determined. 



Genetic effects can be measured accurately even at exceedingly low 

 frequencies, such as 1 in 100,000 for fruit flies or 1 in 10,000,000 for bacte- 

 ria. Here the frequency of positive events is scored directly. As long as 

 this frequency remains small, the exponential curve is practially straight, 

 and the frequency of many classes of mutations is found to increase in 

 proportion to the radiation dose (see Fig. 1-80). 



