RADIATION AND VIRUSES 330 



The first objective of the analysis has been to obtain vahies of k for dif- 

 ferent viruses, to compare them with the physical dimensions of the virus 

 particles, and to obtain estimates of P. A major difficulty is the unc^er- 

 tainty as to what should be considered as the elementary act of absorp- 

 tion, i.e., the "hit." For example, if the elementary act is considered to 

 be the production of one single ionization, a different value is obtained for 

 k than if the elementary act is taken to be the production of a cluster of 

 ionizations supposedly occurring so closely together as to produce only one 

 hit. Moreover, an appreciable fraction of the energy absorbed from 

 ionizing radiation will be dissipated in the form of excitations without 

 ionization. Ultraviolet irradiation studies (see Sect. 2-2) have shown, 

 however, that an excitation has a very low probability of producing virus 

 inactivation. The part of energy dissipated in this way probably makes a 

 minor contribution to the biological effect of ionizing radiation. In the 

 case of radiations that produce dense columns of ionizations, k could be 

 measured as the cross section for collision between the target and the 

 ionization column. There is some uncertainty as to the size of the latter 

 and even as to its approximate reducibility to a cylinder. It should be 

 emphasized that these uncertainties are due to the limited amount of 

 information available concerning the distribution of the energy released 

 by ionizing radiation in liquids. 



In spite of this, values of A- for inactivation of several viruses are avail- 

 able for comparison with the known physical dimensions of the virus par- 

 ticles. A representative group of such data is presented in Table 9-1, 

 from which the following facts emerge: 



1. For a given virus, when the inactivation dose Dye for different radia- 

 tion energies is measured in comparable units, based on ionizations per 

 unit \'olume, there is a clear decrease of effectiveness per unit dose as the 

 ionization density increases, as required by the target theory. Densely 

 ionizing radiation should often produce more than one ionization (or 

 cluster) within each target, with a resulting waste of ionizations. Lea 

 (1946) proposed an elaborate method (the "associated volume method") 

 to calculate the "true" size of the target (supposed to be spherical) from 

 the dependence of k on ionization density. Target sizes obtained by this 

 method are included in Table 9-1. For densely ionizing particles, Lea's 

 method is approximately equivalent to measuring the cross section for 

 collision between target and ionization column. 



2. There is a certain degree of parallelism between particle size and 

 radiation sensitivity for different viruses and for a given radiation, at 

 least qualitatively and with a few exceptions. This is particularly evi- 

 dent for groups of agents presumablj^ similar in kind, such as bacterio- 

 phages. This parallelism justifies the use of radiation data to estimate 

 particle size by interpolation (WoUman and Lacassagne, 1940), a method 

 that may still be of use when a virus cannot be purified enough for electron 



