GENETIC INTERACTIONS BETWEEN ANIMAL VIRUSES 293 



The second example is a mutant of M + in which, in addition to loss of 

 virulence for mouse and chick embryo, two other marker characters were 

 changed. The strain could be passed indefinitely at limit dilution but rapidly 

 reverted to normal M + on passage at low dilution. Recombinant experi- 

 ments suggested that the types obtained were all derived by interaction of 

 revertant M + with the other parent, WS — (Burnet and Lind, 1957a). 



TABLE IV 



Distribution of Mouse Lung Virulence in Clones from the 

 Influenza B Strains LEE and MIL and from Fluids 

 Resulting from Their Interaction in De-embryonated Eggs 



a Percentage of clones falling in each grade is shown. 



G. Interpretation of Influenza Virus Recombination 



Several attempts have been made to provide a general interpretation of the 

 phenomena encountered in the field of influenza virus genetics (Burnet, 1957) 

 but none has been satisfactory and it is conceivable that the difficulties may 

 persist indefinitely. At this level of organization it is almost impossible to 

 avoid confusion between phenotypic characters and conceptual genetic units 

 or to be clear where chemical, physiological, or genetic concepts are needed. 

 The difficulty extends to every aspect of the virus-host cell system and it is 

 impossible to deal effectively with the genetic aspect unless an acceptable 

 interpretation of the process of virus synthesis is available. This, of course, is 

 not the case. If no comprehensive discussion is possible, the alternative is to 

 look for the type of limited generalization that may be of value in the present 

 stage of developing knowledge. 



(a) It is clearly desirable to have available rules by which applications of 

 virus genetics to matters of practical importance could be implemented. It 

 has been suggested, for instance, that if a new serological type of influenza A 

 should appear with exceptional human virulence of the 1918 type, it might be 

 necessary to graft its serological character to another strain of high infectivity 

 for human beings but low pathogenicity. Such a recombinant might allow the 

 use of firebreak tactics to deal with a grave emergency. For purposes like this 



