710 SALMONELLA 



whether or not the pH has been lowered to the point at which this indicator gives its 

 characteristic red colour. Instead of the Uquid medium, a citrate agar medium introduced 

 by Simmons (1926) may be employed. This has been extensively used by Hohn and 

 Herrmann (19366, c). According to KaufFmann (1941), less reliance can be placed on 

 the ammonium reaction than on the fermentation of organic acids. The ammonium 

 reaction and the reaction in Stern's medium are, however, useful at times, as, for instance, 

 in the diGFerentiation of the fermentative types of Salm. enteritidis (KaufFmann 19356). 



In litmus milk tte great majority of species produce transitory acidity followed 

 by permanent alkalinity in 3 days, but with a few species, like Salm. paratyphi A, 

 Salm. abortus-ovis, Salm. typhi-suis, Salm. typhi, Salm. sendai, and Salm. pullorum, 

 the reaction in 3 days is neutral or slightly acid. On the whole little information 

 of differential value is gleaned from a study of the reaction in litmus milk, and Kauff- 

 mann (1941) has now ceased to include this medium in his series of routine tests. 



With the exception of variant colonies of Salm. eastbourne, and very occa- 

 sional variants of Salm. enteritidis and Salm. panama (see Seligmann and Saphra 

 1943), indole production is uniformly negative. The majority of strains form 

 H2S, but Salm. typhi-suis, Salm. sendai, Salm. gallinarum var. duisburg, and Salm. 

 senftenberg var. newcastle fail to do so. HgS formation is variable with Salm. 

 typhi and Salm. paratyphi A, some strains being positive and others negative. 

 Gelatin liquefaction is very uncommon, but Salm. abortus-bovis and Salm. schleiss- 

 heim liquefy it rapidly and Salm. dar-es-salaam and Salm. arizona very slowly. 

 Urea is not decomposed. 



Antigenic Structure. — -As has already been noted, the classification of the 

 Salmonella group is now firmly based on the antigenic structure of the species 

 of which it is composed. The arrangement of the various antigenic components 

 in the cells and flagella of the salmonellee has already been described in Chapter 8 

 in relation to the general problem of antigenic structure, and in Chapter 9 in 

 relation to the study of bacterial variation. Before considering the different 

 species that have been differentiated by this method of analysis, we may, however, 

 recall the relevant facts and consider some of them in greater detail. 



In the Salmonella group of bacilli we are dealing, in the main, with flagellated 

 organisms. We therefore have to consider both the H (flagellar) and (somatic) 

 antigens. By appropriate methods we can test for these components separately. 

 A formolized broth culture of a flagellated siDCcies readily agglutinates in the 

 presence of the homologous H agglutinins, but responds very poorly, if at all, 

 to the homologous agglutinins. The reason for this behaviour is, perhaps, 

 that the formalin fixes the flagella over the bacterial surface in such a way that 

 the somatic antigens are no longer exposed to the action of the agglutinins. 

 To test for the antigens we can employ a bacterial suspension that has been 

 treated with hot alcohol, thus removing or inactivating the H antigens. 



In respect of their flagellar antigens many of the species with which we are 

 concerned are diphasic ; that is to say, their flagella may assume two alternative 

 forms. Originally the two phases were referred to as specific and group ; but 

 later, when another type of phase variation was discovered, the so-called on-^ varia- 

 tion, this nomenclature became no longer generally applicable. It is now usual 

 to call the two phases 1 and 2, and to restrict the terms specific and group to 

 organisms in which Phase 1 contains a, b, c, etc., antigens and Phase 2, 1, 2, 3, etc., 

 antigens. Sj)eaking generally. Phase 1 tends to contain more specific, and Phase 2 

 less specific, antigens ; but it will be convenient to reserve a full description of 



