THE SMOOTH -^ ROUGH VARIATION 303 



The exact nature of the change in colony form, though it happened to give 

 the conventional name to this particular kind of variation, is clearly of quite 

 secondary importance. We may note as a point of interest that the avirulent, 

 non-specific variants of B. anthracis and Sir. ^pyogenes, which happen to give 

 smoother colonies than the normal, virulent forms, have protein instead of poly- 

 saccharide components as their dominant surface antigens when in the normal 

 virulent state. 



We can, if we wish, use some term other than Smooth — > Kough to denote these 

 particular variations, but it seems undesirable to do so, since they belong, in all 

 essentials, to the same category. Alternatively we could abandon " smooth " and 

 " rough " altogether as descriptive terms, and select some new name to describe 

 the loss of the normal surface antigen that is the essential factor concerned. It 

 seems simpler, as we have suggested in Chapter 8, to use the initial letters of the 

 terms " smooth " and " rough " to designate the variation of which the first 

 examples observed happened to be associated with smooth and rough colony 

 formation, but to dissociate " S " and " R " from a designation of colony form. 

 We should then define the R variant as differing from the normal S form in the 

 following ways : 



(1) Loss of the antigenic component characterizing the surface of the bacterial 

 cell in the normal smooth form, whether this component is normally present in 

 the form of a bacterial capsule or not. 



(2) Loss of virulence, partial or complete. 



(3) Altered sensitivity to various bacteriophages. 



(4) A change in colony form, usually, but not always, in the direction of increased 

 granularity or roughness. 



(5) A change in the hydrophobe or hydrophile properties of the cell, usually but 

 not always in the direction of a decreased affinity for water, and a consequently 

 increased sensitivity to the flocculating action of electrolytes. 



It must not be supposed that the S — > R variation represents the limit of the loss 

 of particular antigenic components that bacterial variants may display. An excellent 

 example of this progressive variation by loss is provided by the detailed studies which 

 White has carried out on members of the typhoid-paratyphoid group of bacteria (see White 

 1926, 1927, 1928, 1929o, h, 1931a, h, 1932, 1933). The polysaccharide components that 

 characterize the surface of the bacterial ceU m the normal smooth form are shared by 

 certain types, which are further differentiated from one another by the antigenic com- 

 ponents contained in the fiageUa (see Table 47, p. 713). With these flagellar antigens we 

 are not here concerned. When the normal smooth polysaccharide antigen is lost, the 

 surface of the ceU is dommated by antigenic components that are shared by all members 

 of the tjrphoid- paratyphoid group, and by some related bacteria. These include a polysac- 

 charide component that differs from that characterizing the normal smooth form, and 

 another antigen, or pair of antigens, that are apparently protein in nature and have been 

 named by White p^ and p^,. As the result of further variation the rough or R form may 

 lose its particular polysaccharide component and then give rise to a form, the antigenic 

 behaviour of which is determined entirely by the components p^ and p^. Situated stiU more 

 deeply in the bacterial cell is another antigen, which White has named t ; but it seems 

 doubtful whether this component is ever exposed at the ceU surface as the result of loss- 

 variation. 



It may be noted that the S — > R variation, and still more the progressive 

 loss-variations referred to in the preceding paragraph, are relatively irreversible. 



