INTERFERENCE BETWEEN ANIMAL VIRUS 173 



may explain in part seemingly conflicting reports (e.g., Isaacs and Edney, 

 1950a; Burnet and Fraser, 1952). At any rate, the overwhelming independent 

 evidence pointing to the cellular mechanism of interference would be incom- 

 patible with the idea that protection of cells or organ systems at a distance 

 from those directly exposed to the interfering agent might be due to 

 interference. 



B. Interference among Members of the Myxovirus Group in the Allantoic 

 Membrane of Chick Embryos (CAM) 



Since the original demonstration of reciprocal interference between type A 

 and type B strains of influenza virus (Henle and Henle, 1943; Ziegler et al., 

 1944), the system, with various modifications, has revealed the complex way 

 in which interference and genetic interactions between viral particles are 

 interwoven. This is true particularly for experiments with antigenically 

 related strains of virus. 



1. Heterology and Homology Among Myxoviruses 



The six members of the myxovirus group (Myxovirus influenzae A, influenzae 

 B, influenzae C, multiforme, pestis galli, parotitidis) are considered as heterol- 

 ogous with respect to one another ( Andre wes et al., 1955). At the other extreme, 

 two laboratory-bred variants of a single strain, e.g., egg-adapted and neuro- 

 tropic WS, are obviously homologous. A special problem arises with different 

 strains of, say, influenza type A, which, though sharing antigenic constituents, 

 differ by strain-specific antigens. Are these strains less different from one 

 another than, for example, the different members of group B arborviruses? 

 At least we can infer evolutionary relatedness of type A influenza strains 

 from the historical evidence of a continuous "immunological drift" (Burnet, 

 1955) which imprints new antigenic specificities on newly arising strains 

 (Hirst, 1952). This homology is borne out by results of recombination 

 experiments suggesting that exchange of genetic materials occurs not only 

 with laboratory-induced variants of a single strain (e.g., CAM: Burnet and 

 Lind, 1954d) but also with different strains, be they historically close (WS 

 and MEL: Burnet and Edney, 1951; Burnet and Fraser, 1952; Burnet and 

 Lind, 1951a,b, 1952, 1954a,b,c; Hirst and Gotlieb, 1953a,b, 1955; Lind and 

 Burnet, 1953-1957; Gotlieb and Hirst, 1954, 1956) or far apart (WS and FM1, 

 Hirst and Gotlieb, 1953a; WS and CAM, Burnet and Lind, 1955; WS and 

 various strains, Burnet and Lind, 1956). Assuming that mixed infection with 

 all these pairs may give rise to diploid or heterozygous particles comparable 

 to those postulated for the WS-MEL system (Gotlieb and Hirst, 1954) or to 

 truly homozygous recombinants (Lind and Burnet, 1957a), and that, 

 contrariwise, progeny from mixed type A-type B infection or from influenza- 

 NDV infection contains "phenotypically mixed," but not genetically mixed, 



