716 SALMONELLA 



absorption of such a serum with one species will not remove the agglutinins for 

 the other species. The g and t antigens are likewise complex (Edwards, Bruner, 

 and Hinshaw 1940). In the descriptions of the individual species we have recorded 

 many of the minor and the partial antigens that do not appear in the Kauffmann- 

 White scheme. 



This is not the place to describe the practical identification of Salmonella species. 

 For information on this subject readers are referred to publications by Kauffmann 

 (19396, 1941) and Edwards and Bruner (1942c). 



Antigenic Variation. 



The variations that occur within this group have been studied by many 

 observers ; and these studies have been particularly fruitful in adding to our 

 knowledge of the mechanisms of bacterial variation as a general phenomenon. 

 For this reason they have been discussed in some detail in Chapters 8 and 9. There 

 is no need to repeat that discussion here, but for the sake of clarity and complete- 

 ness, it will be well to give a brief resume of the antigenic variations that may 

 be met with in the Salmonella group. 



Phase variation affecting the H antigens. — In specific-grouj) phase variation, which 

 was described by Andrewes (1922, 1925), the flageUa may assume two alternative antigenic 

 forms. The antigenic components of the specific phase are either peculiar to the particular 

 species concerned or are shared by only a few other species. On the other hand, the 

 group phase contains antigenic components that are shared by many other species. Any 

 given culture of a particular diphasic stram may consist entirely of bacilU in the specific 

 phase, or entirely of bacilU in the group phase, or may contain representatives of both 

 phases. A bacillus in one phase, though always capable of giving rise to descendants 

 in the alternative phase, usually maintains a constant phase over a number of generations. 

 If, therefore, we prepare plate cultures of a diphasic organism and make numerous sub- 

 cultures, each from a single colony, we should as a rule obtain some suspensions in the 

 specific, others in the group phase. If these are killed by the addition of formaUn after 

 18-24 hours' growth, there will usually not have been time for any change in phase to 

 occur. 



The IX- ^ phase variation, which was described by Kauflfmann and Mitsui (1930), consists 

 in a variation that does not differ in principle from specific-group variation. Instead, 

 however, of Phase 2 containing the antigenic components 1, 2, 3, etc., it contains the 

 antigenic components e, n, etc. It will be noted that the antigens in Phase 1 are similar 

 in both types of variation, being of the a, b, c . . . type, but that in Phase 2 they differ. 



A special type of a-^ phase variation has been described by Edwards and Bruner 

 (1938), which may perhaps be regarded as a third type. It was found that, in Salm. 

 worthington. Phase 2, instead of containing the 1, 2, 3 ... or the e, n . . . type of 

 antigens, contained an antigenic component of the a, b, c . . . series, which had previously 

 been met with only in Phase 1. In other words, organisms of this type have what, for 

 practical purposes, may be regarded as two alternative first phases. 



It must be reahzed that, in organisms showing diphasic variation, strains may occur 

 which remain persistently in one phase — usually Phase 2 — and in which the alternative 

 phase can be demonstrated only by growth in the presence of immune serum containing 

 agglutinins to the Phase 2 antigens. Salm. thompson, for example, is almost invariably 

 in the non-specific phase on first isolation and the k antigen of Phase 1 may not be 

 demonstrable till after two or three subcultures in the presence of an anti-group serum. 



Artificial phase variation is a phenomenon that appears at present to be more of academic 

 than of practical importance, though its significance must remain doubtful tiU further 

 work has been carried out. It was observed by Kauifmann (1936() that, if a certain 

 strain of Salm. typhi was cultured in broth containing agglutinating serum to the normal 



