314 THE CLASSIFICATION OF BACTERIA 



which he intends to regard as genera or species ; but he cannot ignore small 

 differences, and he needs a vocabulary which will allow him to talk or write about 

 the bacterial types which interest him. 



This leads to the consideration of another difficulty, which has grown acute 

 during recent years, and will clearly increase rather than diminish in the absence 

 of some agreed international ruling. There is no agreement at all as to the end from 

 which a bacteriological classification should start. Are we to begin by an intensive 

 study of one or another relatively small group, seeking to differentiate within it all 

 the identifiable and stable types, and giving names to these ? Or are we first to 

 differentiate the larger groups, and only when these have been adequately demar- 

 cated seek to divide them into their constituent species, varieties or types 1 As 

 a method of mapping out the ground either approach will serve ; but they lead, 

 unfortunately, to quite incompatible nomenclatures. The method of the intensive 

 study of a small, or relatively small, bacterial group has been adopted by several 

 groups of workers within recent years, and has resulted in such conspicuously 

 successful systematic descriptions as those of the Salmonella group (see Chapter 

 30), or of the haemolytic streptococci and the pneumococci (see Chapter 25). 

 In each of these instances the final differentiation has depended, entirely or almost 

 entirely, on an analysis of antigenic structure. The workers who have been engaged 

 in the study of the Salmonella group have, however, adopted, at the extreme end 

 of the scale, differentiable types listed as Salmonella typhi, Salmonella dublin, 

 Salmonella eastbourne, and so on. Those who have studied the pneumococci and 

 hsemolytic streptococci have, we think more wisely, labelled their recognizable types 

 with numbers, or letters, or letters and numbers combined ; though, except in the 

 case of the pneumococci, there is as yet no general agreement as to the lettering 

 or the numbering. 



We do not, ourselves, think that the agreed definition of species and genera, 

 which has still to be achieved, should be prejudged by the results obtained by 

 antigenic analysis, particularly in the light of recent work on the sharing of antigens 

 among species generally regarded as distinct (see Chapter 8). In the coli-typhoid- 

 dysentery group, however, there are major distinctions, both antigenic and other- 

 wise, which serve to divide the salmonella and the dysentery bacilli from the 

 colon group of organisms in the genus Bacterium, and in the light of the growing 

 number, we have accordingly assigned the enteric and food-poisoning bacilli to 

 the genus Salmonella and the dysentery to the genus Shigella. 



While remaining convinced that, in naming genera and species, weight should 

 be given to other criteria in addition to antigenic relationships, we should wish to 

 record our entire agreement with those who, like White (1937), hold that antigenic 

 analysis affords the best available method of differentiating the ultimate types 

 or varieties into which bacteria are divided, that the antigenic similarities and 

 differences provide a most valuable clue to the natural relationships of these types 

 and the lines along which they have probably been evolved, and that each type 

 or variety so differentiated should be given a distinctive label. We may perhaps 

 add that this distinctive labelling is of particular importance to the medical bacterio- 

 logist, since it is the antigenic make-up of a bacterium that determines all its 

 immunological reactions, in the body as well as in the test-tube. We must, how- 

 ever, try to be consistent. We ought not to use letters and numbers for one set 

 of labels, specific names for another, though current bacteriological usage compels 

 us to do both. 



