320 Transactions of the Royal Society of South Africa. 
Kligler emphasised the importance of salicin fermentation for differential 
purposes, and elaborated a comprehensive system of classification of dextrose- 
fermenting intestinal bacilli ; he subdivided these organisms first according 
to their action on lactose, and the further classification depended mainly on 
the reactions in dulcite, saccharose and salicin; he included among the 
various sub-groups B. proteus and B. cloacae (gelatin liquefiers). 
An older system of classification which is of interest to consider briefly is 
that of Jensen, who, on the basis of certain simple fermentative reactions, 
arbitrarily divided the whole coli-typhoid group into several main sub-groups ; 
the criteria were the reactions in glucose, lactose, saccharose and maltose, 
and in this way the following organisms were differentiated : B. faecalis 
alJcaligenes, B. metacoli/'' B. typhosus, B. paracolon, B. coli anaerogenes, B. 
coli (two types) and B. " pseudocoli."" This system was later amplified for the 
differentiation of coliform bacilli by Wulff, who employed certain additional 
tests — galactose, glycerin, adonite, mannite, dulcite and xylose fermentation. 
This system of classification cannot be considered as sufficiently complete; 
thus no cognisance was taken of indol production and the fermentation 
of inosite. As will be shown later (v. p. 347), these reactions are important 
and characteristic features of certain B. coli sub-groups. 
Houston, in connection with the bacteriological examination of water, has 
classified coliform bacilli into " typical " and " atypical " varieties (v. supra), 
and further divided the typical organisms according to the fermentation of 
dulcite and saccharose. 
While different systems of classification have thus been adopted which in 
the hands of various workers have served a practical purpose in enabling 
them to recognise typical varieties as apart from those which are less 
characteristic, it can hardly be claimed that these organisms have yet been 
completely classified, nor that the significance of different characters has 
been accurately assessed. 
General Outline of the Investigation. 
In the observations to be recorded a further study has been made of the 
" B. coW group, and the characters of a large number of strains of Gram- 
negative, aerobic, iion-sporing, glucose-fermenting, non-liquefying bacilli 
(excluding the specific pathogenic organisms of this class) have been 
investigated. 
The question of the biological classification of these organisms has also 
been studied from the serological aspect. 
In addition certain observations have been made with regard to varia- 
tions among the coliform bacilli and their biological significance. 
The various strains were isolated from pathological specimens of urine, 
faeces, appendix, abscesses, etc. 
