718 SALMONELLA 



The flagellar antigens are unafifected. This type of variation seems to occur rather more 

 frequently in nature, is common in strains maintained through many generations on 

 ordinary laboratory media, and can readUy be induced by several different procedures — 

 the prolonged incubation of a broth culture, growth in the presence of antibodies acting 

 on the smooth somatic antigens, subjection to the action of an anti-smooth phage, and 

 so on. To prevent its occurrence, strains should be dried from the frozen state in vacuo, 

 or, if this is not practicable, kept on a dry Dorset egg medium in the ice-chest and sub- 

 cultured as infrequently as possible. Moist sohd media and sugar-containing media 

 should be avoided. 



It is of interest to note that the rough polysaccharide antigen appears to be limited 

 to members of the Salmonella group. It is possessed by all those strains that have been 

 examined, but not by other species of bacteria, such as coliform or dysentery bacilli 

 (White 1929a). 



Many of the observations recorded by workers in this field indicate quite clearly that 

 the S ^ R variation is not an " all-or-none " process so far as the culture as a whole is 

 concerned. It is gradual or step-hke. A smooth strain may lose some of its normal 

 somatic antigen and uncover some of its rough antigen, or develop it to replace the smooth, 

 so that it will respond to both an anti-smooth and an anti-rough serum. As has already 

 been noted (see Chapters 8 and 9), variation by loss may proceed further than the R form, 

 giving rise to the p forms of White (1932), in which the R polysaccharide is lost. 



Variation of the mucoid or M antigen.— Mention has already been made (p. 705) of 

 the formation by some members of the Salmonella group, notably Salm. paratyphi B, 

 of mucoid colonies containing a nitrogen-free polysaccharide (Birch-Hirschfeld 1935). 

 This so-caUed mucoid or M antigen (Kauffmann 1935o, 1936a) is formed best at room 

 temperature, is resistant to heating at 60° C. for 1 hour and to formalin, but is destroyed 

 partly or completely by heating at 100° C. for 2 hoiu-s, and by treatment with 96 per cent, 

 alcohol or normal HCl at 37° C. for 20 hours. The presence of the antigen is best demon- 

 strated by the use of living suspensions and an anti-M serum. The mixtures should be 

 incubated for 2 hours at 37° C. and overnight at room temperature. Sera usually have 

 a low titre, seldom exceeding 1/160. Only one M antigen has yet been met with. 



Variation of the X antigen. — The X antigen was first observed by Topley and Ayrton 

 (1924) in suspensions of Salm. typhi-murium, Salm. newport, and Salm. enteritidis which 

 had been grown in broth for some days at 37° C, but has since been found to develop 

 in almost all the salmonellse studied. Like the M antigen, it is non-specific. It is 

 thermostable, resisting heat at 100° C. for 30 minutes, and is unaffected by 0-25 per cent, 

 formalin. It occurs in both rough and smooth strains. Its appearance is most common 

 in old broth cultures, though it may be present in broth cultures grown for the norma 

 length of time, or even in suspensions from agar, especially if the surface is moist. Agglu- 

 tination resembles the granular somatic type, but the flakes are as a rule sUghtly coarser 

 and tend to lie along the sides of the agglutination tube. They become apparent in 

 2 to 4 hours, but continue to increase for some time. High-titre serum is easy to prepare 

 by the inoculation of rabbits with a culture rich in X agglutinogen. The importance 

 of the X antigen Lies in the confusion it may cause in routine serum agglutination tests 

 should any of the suspensions contain this antigen or any of the sera contain anti-X 

 agglutinins. In practice, few workers seem to have met with the X antigen, but in one 

 or two laboratories its presence has caused considerable trouble. It is always wise to 

 make certain that standard agglutinable suspensions are free from it (see Cruickshank 

 1939). 



Fermentative Varieties of Antigenic Strains, — Another point that perhaps needs 

 some comment is the recognition of fermentative varieties that are antigenically 

 identical. As a systematic procedure this is a reversal of the common finding, in 

 which different antigenic types can be recognized in a species that is homogeneous 

 as regards its enzymic activities. Many of the fermentative differences that have 



