Discussion 141 



etc., can be computed in turn. My chief concern was to validate a 

 theory of radiation mortahty and ageing. However, empirical 

 analysis and theoretical analysis should have the same goals of 

 parsimonious description. Thus the Gompertz function, which is a 

 theoretically meaningful one, should also be best for empirical 

 analysis. 



Verzd?': Could you describe the same thing with survival curves, 

 Mr. Sacher ? 



Sacher: An animal following a single radiation dose acts at a given 

 age like a control animal at somewhat greater age. The irradiated 

 population tends to show shallower survivorship curves which can 

 be translated and scaled so that they can be superimposed on a con- 

 trol population of a later starting age (Fig. 2). This can be accom- 

 plished without changing the time scale, and corresponds to the fact 

 that single X-ray doses displace the Gompertz function parallel to 

 itself without change of slope. If we give daily doses of irradiation, 

 the effect is not as if we had set the clock forward but rather as if we 

 had changed the regulator, so that the clock runs faster. Thus, con- 

 comitant with the decrease in survival time there is a steepening of 

 the survival curve in the daily dose condition. That corresponds to 

 the fan of lines on the log rate of mortality plot. 



Maynard Smith: In comparing life-tables based on wild popula- 

 tions and on laboratory populations, I think what both Dr. Comfort 

 and Mr. Sacher have had in mind here is that what such distributions 

 are most likely to have in common is the maximum lifespan; the 

 oldest individuals in wild populations may correspond roughly in 

 age to the oldest individuals in laboratory populations, but the two 

 distributions have little else in common. 



Sacher: I can agree with that. It is not yet possible to reach a 

 meaningful correlation between life-tables in the field, and life- 

 tables in controlled environments. These conditions are so far apart 

 that we cannot discuss the respective life-tables in terms of common 

 parameters. It would seem that there have to be intermediate 

 grades of environment between the wild and the laboratory. 



Lindop: Is there any one method of investigation in which, 

 instead of going through the changes gradually, we could correlate 

 them more rapidly ? For instance, one might take the causes of 

 death in wild animals and the causes of death in laboratory animals, 

 exclude from each group the causes which are not in common, and 

 see how the survival curves fitted for the causes of death which are 

 in common. 



Sacher: That certainly could be done if they had enough causes of 

 death in common. 



