MODIFYING FACTORS 349 



ture might differentially affect the rates of the relevant and of the com- 

 peting processes, thus altering h and the slope of the dose-effect curve. 

 Since we have seen above that h is also the parameter most likely to be 

 affected by change in chemical makeup of the cell, it appears that, by and 

 large, temperature is much more likely to affect the slope than the shape 

 of the dose-effect curve. Moreover, neither an increase nor a decrease 

 in slope should be very surprising. 



Another physical factor which affects dose-effect relations is the spa- 

 tial distribution of the individual energy transfers from ionizing particles 

 to molecules in the cell. I shall discuss this in terms of the rate of energy 

 transfer {RET), which is equal to —dE/dx, the instantaneous loss of 

 kinetic energy by the ionizing particle per unit length of its path. This 

 quantity is roughly proportional to the ions produced per unit, length of 

 path {specific ionization or ion density) and has been thoroughly discussed 

 by the physics panel. 



In examining the possible effects of RET, it is essential to bear in mind 

 that, at the very beginning of the action, the individual energy transfers 

 are not spaced singly at random but are grouped along ionization tracks, 

 the degree of grouping being dependent on RET. Very quickly, how- 

 ever, the ions and other relevant entities tend, because of diffusion, to be- 

 come distributed singly at random. Let us now examine two extreme 

 types of action. In the first, the decisive processes occur late, that is, 

 after the relevant entities have become distributed singly at random. 

 In this case, r — 1 = s, and therefore the shape of the dose-effect curve 

 cannot change with RET. On the other hand, it is possible for h to 

 change with RET, because of the latter's possible influence on the 

 efficiency of production of earlier relevant entities, and accordingly the 

 slope of the curve may depend on RET. In the second case, the decisive 

 processes occur early, that is, when the relevant entities involved are not 

 distributed singly at random but are still grouped substantially like the 

 sites of the individual energy transfers. (This, incidentally, is the situa- 

 tion usually discussed by writers on conventional target theory.) In 

 this case, as RET increases, r should increase, and therefore m/r should 

 decrease, with unity as the lower limit. (Likewise, but less likely, s 

 could increase and therefore n/s decrease.) Accordingly, with increase 

 in RET, the shape of the dose-effect curve should vary toward lower hit 

 number (except, of course, in cases where this is impossible because 

 m/r = 1 = n/s even when RET is minimal). The experimental litera- 

 ture contains examples of the foregoing two cases, some of which are 

 summarized in Table L Finally, we must recognize that many radio- 

 biological actions may not belong to either of the two extreme groups 

 just discussed, but have properties intermediate between the two. 



