298 P. M. BURNET 



which, because of their smaller number and lack of qualities for differential 

 survival, cannot be obtained for study. The process of selective survival takes 

 place at two levels, among virus particles in the stage between maturation in 

 one cell and initiation of infection in another, and in the intervening processes 

 which take place in what we have called the replicating pool of the host cell. 

 At the level of the virus particle, one can imagine that the qualities making 

 for selective survival concern mostly (1) resistance to nonspecific thermal 

 degradation, (2) the effectiveness in relation to host cell surface components of 

 the adsorptive-enzymatic structures of the virus surface. Some direct 

 evidence in favor of the second viewpoint was given by Stone (1951). 



At the intracellular level the types of competition and interaction among 

 multiple genomes can only be envisaged in speculative fashion. From our own 

 experience, we believe, as indicated earlier, that the crucial finding is that 

 homozygous virus particles free from any association with serologically 

 heterotypic virus are regularly obtainable in recombination experiments. 

 Under the conditions of Hirst and Gotlieb's experiments, however, equally 

 convincing evidence was obtained that heterozygosis is a necessary preliminary 

 to recombination. 



There seem to be three ways, not necessarily mutually exclusive, of 

 picturing what happens when multiple genomes, either from a single virus 

 particle or from more than one virus particle, enter a susceptible cell. In view 

 of the existence of heterozygotes, we shall make the initial assumption that in 

 a viable influenza virus particle there are n complete genomes (n being a small 

 number, 2 or greater, and not necessarily uniform for all virus particles in a 

 population). 



The first alternative that only a small proportion of the genomes entering 

 the cell can initiate replication has already been discussed. 



The second possibility is based on the assumption of a relation to the 

 phenomenon of incompleteness. There is still controversy about the nature of 

 incompleteness but one can feel confident that ultimately it will be related to 

 the conditions for the selective survival of multiple virus genomes in the cell. 

 The emergence of a homozygote may mean that when two dissimilar genomes 

 enter a cell and one for any reason initiates infection significantly before the 

 other, the first clone will succeed in providing genomes for the new generation 

 of fully viable virus, the second being responsible only for incomplete virus. 



The third approach is based on a particular interpretation of how the new 

 virus particles are fabricated at the free surface of the host cell (Burnet, 1956). 

 It was held that the phenomena of phenotypic mixture plus a variety of other 

 evidence indicated that the structure of the virus surface is derived rather 

 directly from the surface membrane of the host cell. At the time in question it 

 is assumed that the lipids and mucopolysaccharides of the cell surface are 

 reinforced with specific viral protein produced in the replicating pool. If more 



