GENETIC INTERACTIONS BETWEEN ANIMAL VIRUSES 297 



the genotype of the virulent parent will have had to make rearrangements 

 which do not allow such smooth functioning as the parent. Hence, a lower 

 virulence is likely except in the circumstances already mentioned of the 

 MEL/WSE system, where mutual adaptation seems to be particularly easy. 



With a recombinant of low virulence, i.e., by hypothesis, one in which the 

 genotype is lacking in mutual adaptation of its components, is passed in the 

 allantoic cavity at limit dilution, the almost invariable finding is a further 

 loss of mouse virulence. Adaptation by minor mutation or any other mechan- 

 ism is to the conditions of multiplication in the allantoic cavity, not in the 

 mouse lung. Any change will, therefore, tend to deviate the genetic structure 

 away from mouse virulence. 



c. Genetic loci. The above discussion of virulence also has some bearing on 

 the number of genetic loci that may be postulated. The very conservative 

 view from the MEL/WSE system that there are only two loci could well 

 depend on the greater predominance of recombinants of the types found. 

 Even if small numbers of every other possible recombination of our six 

 marker characteristics arose, they could not be detected by current techniques. 

 The CAM results have other possible interpretations but they are com- 

 patible with a genetic structure involving a large number of allele differences 

 but very limited capacity to produce viable recombinations of any sort. Those 

 that were produced and isolated may well have depended on minor preliminary 

 mutation in one or the other component; the chief difficulty in interpreting 

 the results of any recombination experiment giving very small yields of 

 progeny combining parent characteristics is to exclude the possibility that the 

 changes are due to simple mutation and a complex environment where 

 interference or other conditions in doubly infected cells may provide unusual 

 criteria for selective survival. 



The crux of the matter is that no detailed genetic interpretation is possible 

 without extensive numerical data which, with current techniques, could only 

 be provided for the dominant recombinants in a few specially favorable 

 systems. Where rare types have to be obtained by the use of selective environ- 

 ments, it is hardly possible, even in principle, to obtain more than qualitative 

 information. Such information, however, could be valuable in providing 

 evidence at least of the number of loci that can be demonstrated and should 

 be sought by the use of a wider range of selective environments. 



2. The Bearing of Genetic Findings on the Nature of Influenza Virus Multi- 

 plication 



It is a truism that between the inoculation of a single infective particle and 

 the harvesting of the virus yield from an infected organ or cavity, a dozen 

 virus-cell generations and an intense struggle for survival have taken place. 

 The harvest undoubtedly contains mutant and incomplete forms, most of 



