530 



RADIATION HIOLOGY 



lowing discussion is syniholi/od us /,,. Foi' piaclical reasons, Ihc appcai- 

 aiu'c of a tumor was taken as the time at which a tumor of a fi;i\'en esti- 

 mated size, approximately 100 mm' in volume and consisting of about 10" 

 cells, was present on the ear. At this time the tumors were usually 

 doubling their volume every few days, so that the error in estimation of td 

 was relatively small. Since the ears of these mice are almost hairless, 

 the skin was exposed directly to the ultraviolet radiation. No tumors 



100 



-0.2 -0.1 O.l 



TUMOR DEVELOPMENT TIME C/<f )joq dOyS 



0.2 



Fig. 14-1. Distribution of the logarithm of the tumor development time. Id. in a popu- 

 hitiou of genetically homogeneous (strain \) male miee. exposed to ultraviolet radi- 

 ation at regular intervals up to the time of tumor appearance. The data are from 12 

 experiments, a total of 676 mice. The dose and interval between doses was the same 

 for a given experiment but varied between experiments. The mean values of td 

 varied accordingly, and in the figure all the points have been corrected to a common, 

 mean value for each experiment, i.e., log Id = 0. The smallest number of animals in 

 any single experiment was 41, the largest 98. The solid dots represent the experiment 

 with the 98 animals. The drawn curve is the integral of a normal distribution, with 

 standard deviation 0.081 log days. The same distribution was found for a number of 

 additional experiments in which smaller mmibers of mice were used. 



were observed to occur on the well-furred parts of the body; rarely, a 

 tumor appeared on a paw, the snout, or the tail. Tumors of the eye 

 appeared in about 10 per cent of the mice (Lippincott and Blum, 1943). 

 Since it was important to deal with as nearly uniform an area of skin as 

 possible, only tumors of the ear were considered in the data referred to 

 subsequently. 



When the same dose of radiation was applied at regular intervals until 

 the tumor appeared, the distribution of /,/ in the population followed a 

 smooth ([uantitative relation. In Fig. 1 1-1 the cumulative percentage 

 of mice with tumors is plotted against the logai'ithin of /,;, and it is scon 



