16 SECONDARY ELECTRONS 



of the atoms from which each electron was ejected. These activations 

 are said to form a "cluster" (2). Some of the activations in a cluster are 

 produced by the secondary electron which originates the cluster, but 

 some are produced by other electrons ejected with sufficient energy as 

 a result of ionizing collisions within the same cluster. 



Electrons that have spent nearly all their energy usually wander 

 around in a kind of diffusion path, undergoing numerous elastic col- 

 lisions, until finally they are captured by some neutral atom or molecule 

 to form a negative ion. Cloud-chamber pictures, which provide much of 

 the scanty observational evidence on clusters, show the position of 

 negative as well as of positive ions. A large proportion of the clusters 

 appear to consist of a single pair of ions, corresponding to energy-poor 

 secondary electrons. 



The larger the initial energy of a secondaiy electron, the longer and 

 the more nearly straight is its path. The transition from cluster forma- 

 tion to an arrangement of activations along a clear track takes place 

 gradually, of course, as the energy increases from about 100 to about 

 500-1000 ev. Secondary electrons whose energy amounts to at least 

 several hundred volts are loosely called delta rays. 



We can now form the following picture of the action of primary fast 

 charged particles. The tracks of fast electrons are marked by a series 

 of variously spaced clusters of various sizes and by a few delta rays. 

 Occasional delta-ray tracks of unusually high energy may fork out from 

 the main track. 



Heavy particles of energy up to about 10 mev undergo collisions so 

 frequently that the clusters of activations produced by their secondary 

 electrons merge and blend to form a sort of "column." 



Thus the mapping of the energy distribution by secondary electrons 

 appears to be qualitatively understood. Nevertheless a detailed 

 quantitative picture of this process is still missing. To produce a 

 detailed mapping would constitute a fairly laborious task. [This spatial 

 distribution of ions represents a quantity which is under rough control. 

 By varying type and energy of incident particles the mean spacing can 

 be varied over very wide limits. If the effects of individual ionizations, 

 and accompanying excitations, turn out to be independent, the spatial 

 size of the structures involved must be large compared to the mean ion 

 spacing. If the biological structures are not large compared to the ion 

 spacing, correlations will be found. The high-densit}^ columns of 

 ionization will almost always be expected to show correlation effects. 

 This type of analysis is, of course, greatly oversimplified. It is in particu- 



