712 SALMONELLA 



One of the most interesting features of the Salmonella group is the way in 

 which the same antigenic components recur in different combinations. Thus, 

 different antigens are found in combination with the same H antigens. The 

 same set of Phase 2 antigens are replaced by different Phase 1 antigens, when 

 different diphasic species change from the group to the specific or from the ^ to 

 the a phase. The same set of antigenic components recurs in different combina- 

 tions in Phase 2 of different species, and so on. The implications of these facts 

 in relation to the evolution of the Sahnonella group are discussed by White (1926). 



The Kauffmann-White Diagnostic Scheme. — It is clear that the observed dis- 

 tribution of these antigenic components forms a basis for a natural scheme of 

 classification, and this has been adopted in the Kauffmann-White scheme, proposed 

 for international adoption by the Salmonella Sub-committee (Keport 1934). It 

 seems reasonable to regard the antigenic structure of the bacterial cell as more 

 fundamental than the antigenic structure of the appended flagella. The Sahnonella 

 group, as a whole, has therefore been divided into sub-groups, each of which 

 shares a common somatic antigen. Where more than one somatic antigen is 

 present, one of these antigens is regarded as determining the sub-group to which 

 the species concerned shall be allocated. Thus, Group B consists of those organisms 

 that possess the antigen IV, or the antigens IV or V. Group D consists of 

 those organisms that possess the antigen IX, and so on. The group letter, it 

 should be noted, forms no constituent part of the name of any species or of its anti- 

 genic formula. The groups display the natural relations of the different salmonellse, 

 but the antigenic components that determine those relationships are labelled 

 according to the Kauffmann convention. It should, perhaps, be recalled that 

 there is a fundamental difference between this scheme of classification and that 

 adopted in describing the antigenic structure of other bacterial groups. A 

 " Group B haemolytic streptococcus ", for instance, means a haemolytic strepto- 

 coccus that possesses the Group B antigen, and that antigen has no other label. 



It must be made clear that the Kauffmann-White scheme is essentially a 

 scheme for the differentiation in practice of the various species. It is not, nor 

 does it pretend to be, a record of the complete antigenic structure of each organism. 

 The antigenic constitution of most organisms is far more complex than is suggested 

 by the formulae given in the table. Only the major antigens or those antigens 

 that are of differential importance are recorded. It follows that, owing to the 

 possession of minor somatic or flagellar antigens, cross-agglutination may occur 

 between organisms which, judged by the table, possess no common factor. It 

 follows, too, that as further members of the Salmonella group are discovered altera- 

 tions will have to be made in some of the present formulae if the diagnostic value 

 of the table is to be preserved. 



Another warning must be issued. Many of the antigenic components identified 

 by a single numeral or letter are themselves complex, consisting of two or more 

 fractions. For instance, the V and VI somatic antigens each comprise two portions 

 labelled Vi and Vj and VI^ and Vlg. The XII somatic antigen contains three 

 portions, XII^, XII2, and XII3. The d flagellar antigen contains five partial 

 antigens, d, dj, dg, dg, and A^. The fact, therefore, that two species are repre- 

 sented in the table as containing, for example, the d antigen, does not mean that 

 their d antigens are necessarily identical. One of the species may contain one 

 pair of the d fractions and the other species another pair, so that though both 

 species will be agglutinated by an anti-d serum prepared against all the fractions, 



