DISCUSSION OF MORRISON'S AND FANO'S PAPERS 21 



quantitatively. And, although some relevant experimental data are available, 

 they are meager indeed and not entirely concordant. However, it does appear 

 that W, the mean energy required to produce an ion pair, starts to rise for a 

 particle velocity somewhat lower than vq, and thereafter increases steadily as the 

 velocity declines (cf. the work of Madsen*). The functional dependence of W 

 on the particle velocity, in this velocity domain, is still largely unknown. The 

 critical velocity, namely that for which W starts to rise appreciably, appears to 

 be considerably smaller than vq. 



That the overall ionization by slower ions is not as inefficient as might be 

 anticipated if the energy loss were simply a competition between nuclear col- 

 lisions, which lead to virtually no ionization, and familiar excitation and ion- 

 ization, which are often assumed to correspond to the same value of W as for 

 higher velocities (an unjustified extrapolation), might possibly arise from the 

 extremely important contribution to the energy loss, in the velocity domain 

 under consideration, of capture and loss of electrons by the positive ion — a proc- 

 ess often erroneously ignored in discussions of this problem. This process might 

 quite possibly prove to have a value of W smaller than that for excitation and 

 ionization by high-speed particles, and thus tend to compensate to some extent 

 for the energy lost in nuclear collisions and therefore wasted as far as ionization 

 is concerned. 



These matters are also discussed briefly in a later contribution to this volume 

 ("On the Primary Processes in Radiation Chemistry and Biology," p. 97), and 

 a detailed study of the problem by the writer is now in progress. 



Morgan : 



My question is directed to Fano. In the case of a fast neutron (with an energy 

 of, say, 2 mev) colhding with an oxygen atom, the most probable energy given to 

 the oxygen atom is approximately 0.2 mev. This corresponds to the energy of an 

 electron of about 8 ev if the electron has the same velocity as the 0.2-mev oxygen 

 atom. In other words, a 0.2-mev oxygen atom would not be expected to produce 

 much, if any, ionization in tissue. It is my understanding from Fano's discussion 

 that he advises including this energy loss of such heaiy ions (to energy exchanges 

 other than ionization) in the calculation of the maximum permissible flux for 

 fast neutrons. 



Failla: 



It depends on the energy of the neutrons whether or not this effect is going to 

 be an appreciable fraction of the total. I would say that for permissible limits 

 the figures currently at hand have such a large factor of uncertainty that, in 

 general, this would not be a very important matter. 



Morgan: 



It is true that most (or greater than 95 per cent) of the fast neutron energy is 

 lost to hydrogen in the proton production, and from this point of ^^ew the energy 



*B. Madsen, Kgl. Danske Videnskab. Selskab, Mat.-fys. Medd., 23: No. 8, 1945. 



