THE FACTORS THAT DETERMINE PHAGE SPECIFICITY 339 



and jjhage sensitiveness has been recorded ; in the Flexner group of dysentery bacilli 

 (Burnet and McKie 19306, Clauberg and Marcuse 1932, Radojcic 1936, MiUer 1937, Wheeler 

 and Burgdorf 1941) ; among rough and smooth dysentery strains (Donys 1932) ; among 

 variants of a Bad. coli strain (Sertic and Boulgakov 1937) ; among rough and smooth 

 strains of V. choleroe (Asheshov et al. 1930, 19336) ; in Salm. paratyphi C {cholerce-suis) 

 (Levine and Frisch 1936, Frisch and Levine 1936) ; in Group A streptococci (Evans 1936, 

 1940, 1942, Evans and Verder 1938, Evans and Sockrider 1942) ; among certain sero- 

 logical types of C. diphthericB (Keogh, Simmons and Anderson 1938) ; in Staph, aureus 

 (Burnet and McKie 1929, Williams and Timmins 1938, Fisk 1942a, b) ; and in mucoid 

 strains of Bad. coli, Bad. aerogenes and Bad. friedldnderi (Rakieten, Eggerth and Rakieten 

 1940). 



It should be noted that these varieties of antigens and combinations of antigens associ- 

 ated with the specificity of phages are somatic antigens. Sertic and Boulgakov (19366), 

 however, record a phage which acted only on flagellated forms of Salm. typhi. 



The intimate relationship between phage sensitivity and the nature of the surface 

 bacterial antigens has been confirmed by observing the effects on phage lysis of different 

 bacterial extracts. Levine and Frisch (1934) tested the fractions obtained by alcoholic 

 precipitation of sahne extracts prepared from bacteria of the Salmonella and Shigella 

 groups, and fomid that they specifically inhibited the lysis of the homologous organisms 

 by phage. Gough and Burnet ( 1934), working with more highly purified fractions, record 

 similar results. The fractions obtained in this way were known to consist mainly of the 

 specific polysaccharide somatic antigens. Still further confirmation has been obtained by 

 the demonstration by Levine and his colleagues (Levine, Frisch and Cohen 1934, Levine 

 and Frisch 1935), that heat-killed bacteria of the Salmonella group absorb phages 

 specifically, in general accordance with their antigenic structure as revealed by agglu- 

 tination tests. 



Working with staphylococcal phage, Rakieten, Rakieten and Doff (1936) demonstrated 

 that autolysates of sensitive staphylococci inhibited phage action, but those of resistant 

 cocci did not : the same specificity was displayed by heat-kUled staphylococci. Freeman 

 (1937) found that crude preparations of the specific polysaccharide of a Staph, aureus 

 strain inactivated a phage to which it was sensitive (see also Burnet and Freeman 1937). 

 The rates of inactivation of a i^hage by culture filtrates are not compatible with the view- 

 that total inactivation foUows simply from combinations of phage and inhibitory sub- 

 stances (EUis and Spizizen 1941) ; there appear to be two competing processes, one pro- 

 ducing inactivated, the other partly inactivated phage, and full inactivation is achieved 

 only by circumstances favouring one process at the expense of the other (see below, Phage - 

 Antiphage Reaction). 



There can, then, be no doubt at all that the antigenic structure of the bacterial 

 surface is one of the main factors in determining the accessibility of the bacterial 

 cell to a given strain of phage ; and we must suppose the phage particles to be 

 so constituted that those of one strain can find attachment, or entry, only at an 

 area of bacterial surface that is characterized by a particular antigenic structure, 

 while those of another strain have wider potentialities, and are able to attack a 

 bacterium through more than one type of antigenic surface. 



It should, however, be emphasized that this is not the only factor that determines 

 phage sensitivity, or, at least, that there is not complete parallelism between 

 sensitivity and antigenic structure as revealed by the available serological tests. 

 Burnet (19296), for instance, notes that a phage-resistant strain, appearing as 

 the result of the lytic action of a particular phage on a particular sensitive bacterium, 

 may show exactly the same antigenic structure, as revealed by serological tests, 

 as the sensitive strain from which it was derived. He notes further that the resistant 

 strain not only shows no lysis, but fails to adsorb the phage ; so that some change 



