236 F. FENNER AND J. CAIRNS 



heading as if the exact limitations imposed by each situation were precisely- 

 known. 



1. Pock Variants of the Poxviruses 



Infection of the chorion by the poxviruses results in a mixture of cellular 

 proliferation, infiltration, and necrosis, involving both ectodermal and 

 mesodermal cells (Burnet, 1938). The appearance of the pocks is therefore 

 an index not only of the response of the infected cells but also the extent of 

 spread of the infection and the reaction of the host animal — more complex 

 properties than those that control, say, the morphology of the plaques 

 produced by poliovirus in tissue culture. However, like plaques in tissue 

 culture (Dulbecco and Vogt, 1954), each pock must be the result of infection 

 by a single virus particle, since the number formed is proportional to the 

 inoculum size (Burnet, 1936; Keogh, 1936; and many others since). For this 

 reason the pock-producing viruses make a suitable group in which to study 

 virus variation. 



Cowpox virus (Downie and Haddock, 1952) and neurovaccinia (Fenner, 

 1958a) both give rise to white pock variants. These arise with a frequency 

 of 10 -2 to 10 -4 , are stable on further passage, and presumably owe their 

 difference from the normal hemorrhagic necrotic pock to a decreased tendency 

 to invade underlying blood vessels and an increased tendency to cause cell 

 infiltration. In both cases, attempts to demonstrate significant fluctuation in 

 the frequency of white variants in sister populations of virus particles have 

 failed, but it is not certain whether the observed absence of significant 

 fluctuation is itself significant evidence that the change is not due to random 

 mutation (Fenner, unpublished). Neither change appears to be host-induced 

 in that the frequency of white variants is independent of the source of virus 

 (pock on chorioallantoic membrane or nodule on rabbit skin) (Fenner, 

 unpublished). In view of the exceptionally high frequency of the change 

 (up to 10 -2 for some strains) compared to all examples of mutation in other 

 viruses, the possibility must be considered that, as in the case of the r 

 mutation in the T-even bacteriophages, it may be the common phenotypic 

 expression of a large number of different genotypes. Whether the wild type 

 can be obtained from crosses between different white variants has yet to be 

 determined. 



Associated with this change in pock morphology there is a change in 

 virulence for other hosts (Downie and Haddock, 1952; van Tongeren, 1952). 

 The white variants produce less severe reactions in both rabbits and mice. 

 However, the correlation between the pock morphology of the poxviruses and 

 their virulence is not complete, for some strains of vaccinia and cowpox which 

 produce hemorrhagic pocks have low neuropathogenicity for mice and 

 rabbits, and some which produce white pocks are highly pathogenic. 



