INFLUENCE OF LINEAK ENERGY TRANSFER 317 



The second method may be termed the track-average method. Here, 

 instead of allowing only a selected linear portion of each track to traverse 

 the object, either the entire track is spent in the object or random hnear 

 portions of different tracks traverse it. This results in a spread, usually 

 great, of the value of LET. If the particles are initially uniform in 

 energy, the spread extends from the maximal value at the ends of the 

 tracks to some lower value, which depends on the nature and energy of 

 the particles, at the beginnings of the tracks. This heterogeneity in most 

 cases is increased by nonuniformity of initial energies of the individual 

 particles, so that LET extends over a wide and sometimes complex spec- 

 trum. For brevity some average value must be used ; this is customarily 

 the mean, i.e., the total energy transferred along a large number of tracks 

 divided by the total length of these tracks. Different mean values of 

 LET are obtained by selecting radiations with suitably different spectra 

 of LET. 



With currently available equipment LET can be varied continuously 

 from about 0.25 kev//i (a singly charged particle traveling with a velocity 

 approaching that of light) to about 260 kev/^i (the end of an a-particle 

 track). There is no chance of substantially extending this range toward 

 lower values because the theoretical minimum is about 0.2 kev//x (Gray, 

 1947). Above this range, Tobias (personal communication) has recently 

 used a beam of accelerated carbon ions whose LET is about four times 

 that of a particles. Fission fragments (LET « 6000 kev/fi) are also 

 available for certain types of experiment but have not yet been used in 

 investigations pertinent to this paper. Theoretically higher values are 

 attainable, the maximum being the LET of a low-velocity nucleus of 

 greatest available atomic number (currently berkelium, Z = 98). Avail- 

 ability of such values of LET appears unlikely in the foreseeable future. 



In the investigations reviewed here, the following radiations have been 

 used : 



A. Fast electrons or radiations which produce them. 



1. Hard 13 rays (mean energy, 0.4 Mev or higher; mean LET, 0.28 

 kev/ju or less). 



2. Hard y rays (e.g., those of Ta^^^ or of RaB + RaC with 0.5 mm 

 platinum filter) which ionize almost entirely by Compton recoil 

 electrons (mean LET, about 0.35 kev/n, depending on filtration 

 and other factors). 



3. X rays of various wave lengths. With the usual tube voltages, 

 average LET of the ejected electrons is considerably less than 

 that of those ejected by y rays. For very hard X rays (0.05 A 

 or less), which produce practically 100 per cent Compton recoil 

 electrons, the mean LET increases with wave length. The same 

 is true for soft X rays (0.4 A or greater), which produce practi- 



