316 RADIATION BIOLOGY 



actually is not known. Moreover, the term "specific ionization" empha- 

 sizes the importance of the ions to the exclusion of the excited molecules, 

 which, so far as currently known, may also play significant roles in radio- 

 biological mechanisms. In view of these uncertainties, it appears pref- 

 erable to discuss the distribution of physical events along the tracks in 

 terms of energy transferred, rather than of ion pairs formed, per unit 

 length of track. In this paper, this quantity will be designated the linear 

 energy transfer,^ be abbreviated to LET, and be expressed as thousands of 

 electron volts per micron (kev//i) of tissue. 



Within the last three decades, numerous radiobiological investigations 

 have been carried out with two or more radiations differing in average 

 LET, and in a large fraction of these experiments it has been found that 

 the total dose (energy absorbed per unit mass of tissue) required to pro- 

 duce a standard degree of a given biological effect is not constant but may 

 vary considerably with LET. 



All these experiments obviously involve two physical requirements: 

 (1) practical sources of at least two radiations which produce tracks 

 differing in LET; (2) methods of estimating, in some common unit, the 

 doses delivered wdth each type of radiation. The usual means of meeting 

 these requirements and the degree of success attained will be indicated 

 briefly in the next two sections. 



EXPERIMENTAL VARIATION OF LINEAR ENERGY TRANSFER 



The linear energy transfer along a particle's track varies as the square 

 of its charge; for instance, the LET from an a particle is four times that 

 from a proton, deuteron, or electron of the same instantaneous velocity. 

 LET also increases as the particle's speed decreases; it therefore varies 

 greatly along any individual track and is maximal near the end. 



There are two general ways of obtaining a variation of LET for exper- 

 imentation. The first, basically the simpler, is to use a beam of particles 

 that have straight tracks and uniform initial speed and allow them to 

 traverse various samples of the biological object with selected portions of 

 their tracks to which the desired values of LET pertain (Zirkle, 1935, 

 1940; Zirkle and Tobias, 1953). This method is obviously restricted to 

 biological objects whose dimensions are small compared to the range of 

 the particles in tissue. Within this limitation, it is extremely satisfactory 

 because each experimental value of LET has a very small spread. This 

 may conveniently be termed the track-segment method. It has not yet 

 been widely used because, only with very high-energy accelerators such 

 as the synchrocyclotron, can a wide range of values of LET be obtained. 



1 Numerically identical expressions are linear energy absorption (Zirkle, 1940) and 

 linear rate of energy loss, each directing attention to only one aspect of the linear 

 energy transfer. 



