172 THE PHYSICS OF VIRUSES 



which will presumably break more readily at certain points. One 

 possible study is, therefore, of virus length. This has been done 

 by Oster (1947) and Newton (1951) and will shortly be described. 

 In addition to this rather easily visualized process, the rapid 

 development of a cavity will bend and distort the virus to some 

 degree. This may easily cause it to become noninfective. 



If the above picture of sonic action is accepted, it will be ex- 

 pected that viruses should vary greatly in susceptibility to sonic 

 action. This is found to be so, and animal viruses seem to be 

 much harder to inactivate than rod-shaped plant viruses. This 

 variability in sonic action is shown in experiments of Anderson, 

 Boggs, and Winters (1948). The relative loss of ability to form 

 plaques after exposure to a magnetostriction oscillator in a water 

 cooled cell is shown in Fig. 7.1. The seven T-series E. coli bac- 

 terial viruses were studied, and it can be seen that T-2, T-4, T-5, 

 and T-6 lose their infectivity rapidly and in much the same way, 

 whereas T-1, T-7, and T-3 are much less sensitive. T-3 appears to 

 be inactivated in a nonlogarithmic way, whereas T-1 and T-7 

 follow a logarithmic relation. It is interesting that both T-3 and 

 T-7 are very much alike in appearance in the electron micro- 

 scope, being both spherical and about 450 A in diameter, but are 

 definitely different in sonic sensitivity. This was found to be true, 

 although in a less definite way, for deuteron bombardment by 

 Pollard and Forro (1951). A similar study of five megaterium 

 phages was made by Friedman (1952) who also verified Ander- 

 son's work for T-1 and T-5 phages. Since a similar irradiation 

 arrangement made by the same manufacturers was used, the 

 results are comparable. They show logarithmic inactivation in 

 all cases. The relation holding is, therefore, the familiar 

 expression 



In (n/??o) = —^%t 



where n/no is the survival ratio, kg is a rate constant, and / is 

 time. If we tabulate kg from Anderson's and Friedman's data 

 (using common measurements on T-1 and T-5 as mutual cali- 

 bration), together with electron-micrograph dimensions, it can 

 be seen that ks is larger for the larger viruses. The rate constants 



