342 THE BACTERIOPHAGE 



references to the antigenic behaviour of the phages studied, as well as to their 

 other properties. 



The studies of Burnet and his colleagues clearly show the close analogy of the 

 serological reactions of phage with those of bacteria. Bacteria which have been 

 allowed to adsorb large amounts of a particular phage, will absorb the antibody 

 from a homologous specific antiphage serum, from which all antibodies acting on 

 the bacteria themselves have been removed. The treated bacteria are agglutinated 

 by the antiphage serum (Burnet 1933c). 



Filtrates from phage preparations contain soluble substances that inhibit the 

 reaction of the phage with its antibody and will to some extent reverse a phage- 

 antibody union that has already taken place (Burnet 1933d, Burnet et al. 1937). 

 The inhibitory substance here is analogous with the specific soluble substance 

 that characterizes many species and types of bacteria. Suspensions of large- 

 particle phages free from bacteria and bacterial debris are agglutinable by antiphage 

 sera (Burnet 1933c, Merrill 1936), and the agglutinating power of the serum for 

 various related phages runs parallel with its power of phage-neutralization. 



The quantitative relations between a phage and its homologous neutrahzing serum 

 were first studied in detail by Andrewes and Elford (1933a, b), who found that, over a con- 

 siderable range of phage dilutions, a given amount of antiserum neutrahzed a constant 

 percentage of the phage present, irrespective of the number of phage particles exposed 

 to its action (see also Clifton, Mueller and Rogers 1935). This certainly indicates that 

 the particles are inactivated individually and not by aggregation ; for if aggregation was 

 effective in reducing the phage-count in the serum-phage mixture, we should expect a 

 greater percentage drop in the count as the concentration of phage, and therefore the 

 likelihood of colhsions of sensitized phage particles, increased. The results suggest also 

 that, as we have seen in other antigen-antibody systems (Chapter 7), the phage or the 

 antibody, or both, may be heterogeneous, and that in any phage preparation a certain 

 percentage of the particles either unite feebly with antibody, or are in some way insus- 

 ceptible to inactivation after union (see also Kalmanson and Bronfenbrenner 1942), The 

 validity of these interpretations is questioned by Hershey and his colleagues (1943, 1944) 

 on the grounds that there is little direct, independent evidence of heterogeneity of either 

 antibody or phage. They have observed two types of antibody effect in a coh phage. 

 Thus, with a given amount of homologous antiserum, the phage was sensitized, and its 

 infectivity thereby increased ; with double this amount the phage was neutralized and 

 became non-infective. The variations in activity displayed by phage -antiserum mixtures 

 may therefore be interpreted in terms of a homogeneous phage, varying proportions of 

 which are unaltered, increased in infectivity, or decreased in infectivity, according to the 

 degree of union with an essentiaUy homogeneous antibody. 



Though there are obvious parallels between inactivation of phage by adsorption to 

 bacterial extracts, and by antiphage sera, it does not appear that the two inactivators 

 unite with the same part of the phage particle (Burnet and Freeman 1937, Burnet et al. 

 1937), for antigenicaUy similar phages lyse antigenically uiu:elated bacteria, and phages for 

 antigenicaUy similar bacteria may themselves be antigenicaUy unrelated. Moreover, by 

 adaptation, the affinity of a phage for a bacterium may be profoundly modified without 

 any alteration in antigenic quality. Nevertheless, the active sites of union of antibody 

 and bacterial inactivator appear to be close together on the phage particle, for union with 

 antibody blocks union with bacterial inactivator, though in some cases the phage is only 

 partly neutralized by the antibody, producing lysis on the prepared plate after a delay 

 which may represent the time required for a freeing or a redistribution of the antibody 

 on the phage particle, making the bacterial receptor again accessible. The variation 

 observed in the degree of neutralization and in the rate of the phage-antibody reaction 

 with temperature (Hershey 1941) suggests that the surface of the phage particle 



