292 F. M. BURNET 



recombinants with serological character PEE, in crosses between the recent 

 filamentous strain PEE, (Persian Gulf 1/1951) and WSE (Burnet and Lind, 

 1956). 



The papers dealing with these experiments were published before the 

 findings of Briody et al. (1954) and Edney (1957) were available. These 

 indicate that some strains, including CAM, may develop virulence for mice 

 without experience of passage in the mouse lung. In Edney's work, adaptation 

 to growth in the presence of anti-CAM serum in the allantoic cavity was 

 associated with the appearance of a moderate degree of lesion-producing 

 capacity. The possibility must therefore be kept in mind that in the WSE/ 

 CAM crosses the appearance of mouse pathogenicity in CAM + recombinants 

 might have been an essentially accidental or mutational appearance in the 

 course of the mutual adaptation of virus components in the new form. On the 

 other hand, in this cross there was a significantly high yield of WS derivatives 

 with MLV far below the normal level — types that were not seen in the MEL- 

 WSE system. 



In another type of experiment CAM was adapted to high MLV by serial 

 passage in mice. The virulent strain CAM-MP was then crossed with a non- 

 pathogenic derivative of CAM obtained by growth in anti-CAM serum and 

 "marked" by serological resistance, CAM-SE. The progeny were sorted into 

 serum resistant and serum sensitive clones and each tested for MLV. None of 

 the SE clones had mouse pathogenicity but the sensitive clones showed a 

 wide range of MLV from the original high level to zero (Burnet and Lind, 

 1954a). 



Using an influenza B strain EOB adapted to mouse lung and crossing with 

 the serologically distinct strain MIL, Ledinko (1955) obtained a similar range 

 of lowered MLV in the derivatives of EOB serological type. In these experi- 

 ments mouse lung pathogenicity was the only character showing transfer 

 from one serological type to the other. 



Precisely similar results had been obtained earlier in crosses between the 

 influenza B strains LEE mouse pathogenic and MIL (nonpathogenic) 

 (Perry et al., 1954) and Table IV, modified from their paper, gives a good 

 impression of the character of the redistribution of virulence of this type. 



3. Discontinuous Mutation Involving Virulence 



Brief mention should be made of two variant strains which we have 

 recently studied in which there was a complete loss of virulence for the 

 mouse. The most striking is a derivative from the highly mouse virulent 

 strains WS — . Under appropriate allantoic passage at low dilution, reversion 

 in stages to the mouse virulent form occurs and recombination between 

 virulent and avirulent types gives some forms of intermediate virulence 

 (Lind and Burnet, 1958). 



