BACTERIAL VARIATION AS A GUIDE TO ANTIGENIC STRUCTURE 277 



suffering from typhus fever. The proteus bacillus, in its normal, flagellated form, 

 grows on the surface of nutrient agar as a thin spreading film which resembles the 

 mist produced by breathing on glass, and was named the Hauch form — the " breath " 

 form, by Weil and Felix. A non-flagellated variant grew in isolated colonies with 

 no thin, spreading growth between them ; this was the Ohne Hauch form — the 

 form without an exhalation. As is the way in laboratory shorthand these soon 

 became the " H " and " " forms ; and so the thermolabile, flagellar antigens are 

 now the H antigens, and the thermostable, somatic antigens are the antigens. 



Arkwright (1920, 1921, 1924), working with bacteria of the typhoid-paratyphoid- 

 dysentery group, described a type of variation that has proved particularly instruc- 

 tive from this point of view. He noted that a particular type of variant was of 

 relatively common occurrence among these bacilli ; and this variant differed from 

 the parent form in several characteristic ways. Thus, the normal parent form gave 

 smooth colonies on a solid medium and a diffuse growth in broth, and was not 

 auto-agglutinable in normal saline (0-85 per cent.). The variant form gave rough 

 or granular colonies on solid media and a granular growth in broth, and was auto- 

 agglutinable in normal saline, though a stable suspension could usually be prepared 

 in distilled water, or in a saline solution with a greatly decreased salt-content 

 (0 2-04 per cent.). These differences in colony form, growth in broth, and salt- 

 sensitiveness were associated with a profound change in antigenic structure. The 

 normal parent form reacted specifically as regards its agglutinability by immune 

 sera ; the variant was agglutinated not only by its own antiserum but by antisera 

 prepared against many other bacteria in the rough state, some of which were only 

 distantly related to the parent form according to the ordinarily accepted bacterio- 

 logical criteria. By the usual transition — through shorthand to symbols — the 

 normal form, giving smooth colonies, became the Smooth (or S) form ; the variant, 

 giving rough colonies, became the Rough (or R) form. And here again we have 

 become entangled in the web of our words, for we have never defined exactly the 

 criteria of roughness and smoothness. The implicit meaning is certainly not the 

 meaning we want ; since, as we shall see, the correlation between colonial form and 

 antigenic structure is by no means constant. If we give Smooth and Rough their 

 obvious or commonsense meaning, then we shall want new terms for the underlying 

 changes that really interest us. If we employ the legitimate licence of scientific 

 terminology, and say that by Smooth we mean one sort of antigenic structure and 

 by Rough another, then we must at least define quite clearly what we mean. Up 

 to the present we have evaded our difficulties, and so we find that the rough type 

 of one species is recorded as having the more important characters of the smooth 

 type of another species, as in the case of the anthrax bacillus (see Preisz 1904, 

 Eisenberg 1912, Bordet and Renaux 1930, Tomcsik and Szongott 1932, 1933) ; 

 or it becomes necessary to invent new descriptive terms, such as matt and glossy, 

 as in the case of the hsemolytic streptococcus (see Todd 1928a, b, Todd and Lancefield 

 1928). Probably the most convenient convention is to use the symbols S and R 

 to distinguish the antigenic variation, and retain, as far as is necessary, the words 

 Smooth and Rough as descriptions of colonial variants. There will thus be no 

 contradiction in asserting that the S form of B. anthracis grows as a rough 

 colony. 



In most of the cases with which we are familiar the change from S form to 

 R form (the S — >■ R variation) is associated with a loss of virulence — an associa- 

 tion of obvious importance to the immunologist — as well as with a particular kind 



