544 r. M. BURNET 



serum. At the other end, the susceptibiHty of the cells to virus hemagglutina- 

 tion was lower while higher antiserum titer was indicated and there w^as much 

 inhibition by normal serimi. To obtain specific results m antiserum titrations 

 with recently isolated viruses, only cells of the first type were suitable. The 

 analogy of this model to the results obtained with neutrahzation tests in vivo 

 when hosts of different susceptibihty are used is obvious. 



Isaacs and Stone (1949) found that when mixtures of serologically identical 

 and D virus were used to titrate homologous antiserum, the results differed 

 greatly according to whether fowl (agglutinated only by D) or guinea pig 

 cells (agglutinated by both and D phases) were used. The result obtained 

 corresponded to the amount of virus capable of agglutinating the test red cells 

 used. With fowl cells, the presence of virus had no influence on the titer 

 obtained. In a basically similar experiment, Burnet (1955) found that virus 

 inactivated at the lowest temperature necessary to destroy hemagglutinin 

 had no influence on the apparent titer of a homologous serum when it was 

 added to active virus. 



(3) An interesting indication of the reversibility of both virus-cell and 

 virus-antibody unions can be obtained by setting up a uniform series of 

 serum dilutions S with standard amounts of virus V and red cells C in three 

 diiferent ways, by putting two components together for an hour and then 

 adding the third and reshaking all tubes, i.e., (a) /S + V — C,{h) S -\- C — V, 

 (c) V -\- C — S. At the end of the first hour (c) will, of course, show uniform 

 agglutination. On resettling, all tliree will give similar end points, which will 

 come closer stiU if the tubes are again reshaken and allowed to settle. 



The results of hemagglutination inhibition experiments indicate that in 

 the reaction mixtures where the end-point reaction of partial agglutination 

 is shown, there is always present considerable amounts of free antibody and 

 that, in the early stages at least, virus-antibody union is highly reversible. 

 There is a good deal to suggest that there is a slowly progressive development 

 of irreversible union in aU virus-antibody reactions. Many workers have 

 shown that antihemagglutinin can be absorbed out of an immune serum 

 by the addition of an excess of virus, followed by centrifugation at a speed 

 sufiicient to sediment the virus particles (Hirst, 1952). 



5. Neutralization of Influenza Virus in the Allantoic Cavity 



This reaction has quahties which give it some general interest. The char- 

 acteristic finding (Burnet, 1943; Walker and HorsfaU, 1950) is that the effect- 

 iveness of antiserum falls off sharply with dUution. When plotted logarith- 

 mically, the neutrahzation curve is an approximate straight line with a slope 

 between 3 and 5, i.e., if a certain concentration of serum S reduces the titer 

 of virus to 10~^ of the original, 8/2 may reduce it to 10~^-^ and ^/lO to 10"^. 



