720 SALMONELLA 



As already mentioned, however, the method is subject to limitations. In 

 the first place, not all strains of typhoid bacilli possess a Vi antigen or possess 

 it in sufficient quantity to enable the bacteriophage to act satisfactorily. For 

 example, in Craigie and Yen's series of 706 strains, 42 belonged to the W form 

 and 72 to the VW form. The remaining 592 strains belonged to the V form, 

 possessing a fully developed Vi antigen ; of these, 98-6 per cent, were successfully 

 typed. The proportion of W and VW forms was higher in this series, which 

 contained many strains that had been isolated months or years previously, than 

 is commonly found in routine laboratories handling freshly isolated strains, of 

 which usually only about 5 per cent, are found to be lacking in Vi antigen (Craigie 

 and Brandon 19366, Kauss 19396). Another limitation is that only species of 

 Salmonella containing the Vi antigen can be typed. This is a very serious limita- 

 tion since, according to our present knowledge, the distribution of this antigen 

 is restricted to Sahn. typhi, Salm. paratyphi C, Salm. ballerup, and Salm. hormcBchei. 



More recent observations, however, suggest that some other species possess 

 either a Vi antigen or another type of antigen which is susceptible to differential 

 attack by the bacteriophage. Felix and Callow (1943), for example, have been 

 able to develop a series of bacteriophages acting on Salm. paratyphi B, and have 

 by this means succeeded in dividing strains of this organism into four bacteriophage 

 types, which correspond closely with the epidemiological types. 



The technical methods used in the bacteriophage typing of typhoid and para- 

 typhoid bacilli cannot be described here. Table 48, however, illustrates the 

 susceptibility of typhoid bacilli to the different types of bacteriophage. 



It will be noticed that Type A strains are sensitive to all bacteriophage type 

 strains, but that the remaining types possess a fairly high degree of specificity. 

 Further types and sub-types are constantly being described. Types L and M 

 and sub-types B3, B4, D3, Eg, and Fg have since been added ; and Felix (1943) 

 in this country has reported three new sub-types D4, D5 and Lj, and a new type 

 labelled provisionally No. 91. As more strains of typhoid bacilli are studied 

 from different environmental conditions, it is probable that even further bacterio- 

 phage types will be discovered. Sufficient, however, are already known to render 

 this method of study profoundly valuable in the field of epidemiological inquiry. 

 More recently, Craigie (1942) has modified his original scheme, but the principles 

 of his method remain unaltered. 



Chemical Structure of the Salmonella antigens. — We may add a brief note on 

 our knowledge, such as it is, of the chemical constitution of the antigens on which 

 the present classification of the Salmonella group is based. Of the nature of the 

 flagellar antigens we as yet know nothing, beyond the hints conveyed by heat 

 lability and sensitiveness to extraction with alcohol. 



Of the somatic antigens we know rather more. Furth and Landsteiner (1928, 

 1929) isolated several different chemical components from bacilli of the Salmonella 

 group. Some of these were polysaccharides. The studies of White (1929a, 6, 

 1931) make it clear that the somatic antigens — those labelled I, II, III and so 

 on in our antigenic formulae — are polysaccharides, or have a polysaccharide com- 

 ponent ; and further evidence on similar lines has been recorded by subsequent 

 observers (Casper 1928-29, Combiesco et al. 1930, Basilewsky and Remgild 1935). 

 More recent studies by Freeman, Challinor, and Wilson (1940) and Morgan and 

 Partridge (1942) on Salm. typhi suggest that the antigen of this organism is a 

 polymolecular complex formed of a specific polysaccharide, a protein, and a 



