546 



ItADIATION HIOLOflY 



curves in Figs. 1 1-1 and 2 will show that this is a justifiable procedure 

 since, because of the luiture of the incidence curves, the same relation 

 holds for any other percentage incidence. In 11 of the experiments 

 descrilx'd in Fig. 14-0 the doses were applied 5 days per week, the interval 

 being taken as 7/5 days and the curve so labeled in the figure. Above a 

 certain value of D, designated /;„„ t.i does not decrease with increase in D, 

 i.e., the curve is a horizontal line. Below D„, the points are quite well 

 fitted by the drawn curve which has the slope ->2, representing the con- 

 dition, explicit in Eq. (14-13), that ta varies inversely as the square root of 

 the dose. The reason for the flattening of the curve at D„. will be dis- 

 cussed later. Figure 14-6 also shows data for dose intervals of 1 day and 



-0.5 -0.4 -03 -0.2 



6 7 



-0.1 01 2 0.3 0.4 5 



DOSE (D), log of values 

 Fig. 14-6. Relation between dose D and development time tj for three different 

 intervals i (days). The relative accuracy of the points is indicated by the vertical 

 lines, which represent the limits which should be exceeded only once in 20 times on the 

 basis of chance alone. The symbol /)„, indicates the point beyond which td does not 

 decrease with increase in dose. (From Blirm, 1950.) 



of 7 days. The drawn curves resemble the curve for the 5-day-per-week 

 schedule. There are insufficient points on the curves for the 1- and 7-day 

 intervals to establish the shape assigned, but the agreement with the 

 curve of the 5-day-per-week schedule is obvious. The fit of the equation 

 to the data is also shown in Fig. 14-7, where the curves are plotted on 

 numerical ordinates. The graphs in Figs. 14-6 and 7 show that the 

 data support the theory within the limits of accuracy as well as c-ould 

 be expected. 



E(iuations (14-12) and (14-13) do not hold exactly for different values 

 of i, necessitating the following correction : 



1 ^'^ 



2{i — a) 



where a is a constant having the value 0.52 

 however, affect the general argument. 



Most of the tumors grow rapidly once they appear 



(14-14) 



This modification does not, 



In some cases 



