CLASSIFICATION 697 



fermenting group, finding that some strains shared a group antigen and that 

 others did not. 



All workers are now agreed that the main lines of cleavage are (1) between 

 the mannitol and the non-mannitol fermenters, and (2) in the mannitol-fermenting 

 group between the non-lactose and the late-lactose fermenters. Opinion varies 

 mainly in respect of the differentiation of antigenic types from species, and the 

 nomenclature to be employed. 



In the non-mannitol-fermenting group, Sh. schmitzi is distinguished from 

 Sh. shigcB by fermenting rhamnose, producing indole, and being antigenically 

 distinct. The Newcastle bacillus may or may not ferment mannitol ; the non- 

 mannitol-fermenting strains can be distinguished readily from Sh. shigce or Sh. 

 schmitzi by their production, on first isolation, of a bubble or two of gas in tubes 

 of glucose and dulcitol, and by their different serological behaviour. Besides 

 these species, there is a group of organisms, at present ill-defined, which resemble 

 Sh. shigcB or Sh. schmitzi biochemically, but differ from both in their antigenic 

 structure. These organisms have been described by Dudgeon and Urquhart (1919) 

 as para-Shiga bacilli. Since they vary in their fermentation of arabinose, their 

 production of indole, and their serological reactions, it is impossible to say at 

 present whether they should be classified as a single species, Sh. parashigcB, and sub- 

 divided into types, or whether more than one species will be required. We prefer 

 to leave the matter open and refer to them simply as para-Shiga bacilli. 



In the non-lactose-fermenting subdivision of the mannitol-fermenting group, 

 Sh. alkalescens differs from most strains of Sh. flexneri in fermenting dulcitol, 

 xylose, sorbitol, and, according to de Assis (1939a) most important of all, glycerol. 

 It is, moreover, susceptible to acid agglutination. The flexneri types are peculiarly 

 difiicult to classify satisfactorily at present. The old division on antigenic structure 

 into V, W, X, and Z types has been challenged by Boyd (1938, 1940), who does 

 not recognize the X type, and who proposes to replace the letters V, W, and Z 

 by the Roman numerals i, ii, iii and to include under the term Bact. dysenteries, 

 Flexner, three other types iv, v, and vi, corresponding to his strains 103, P119, 

 and 88. All of these types, according to Boyd, share a common group antigen. 

 Other strains, biochemically similar to Flexner's bacillus but not containing the 

 Flexner group antigen, he would put into a separate species, labelled by his own 

 name, and divide them on an antigenic basis into Bact. dysentericB, Boyd, Types i, ii, 

 and iii, corresponding to his strains 170, P288, and DL 



This raises serious difficulties both of classification and of nomenclature. 

 Leaving nomenclature until later, we think that there is a good precedent for 

 classifying antigenic variants of the same species into types distinguished by 

 numerals ; for this reason we would welcome the replacement of the letters V, 

 W, and Z by numerals. We think also that there is justification for classifying 

 biochemically similar, but antigenically distinct, organisms as separate species, 

 and we see no objection to the use of the term hoyd as a specific name. When, 

 however, we consider the constitution of the flexneri species, we are faced with 

 two difficulties, namely, should the X type be omitted, and should strain 88, which 

 is a mannitol-fermenting variant of the Newcastle bacillus, be included ? In his 

 1938 paper Boyd concluded provisionally that X contained a distinctive type anti- 

 gen, and it was not till 1940 that he decided that this type was merely a variant of Z. 

 Previous workers have had little difficulty in recognizing the X type, and the 

 evidence on which it has been degraded might be considered as unconvincing. 



