364 
Transactions of the Royal Society of South Africa. 
medium in MacCoiikey's agar a mixture of pale and red colonies was 
obtained ; the pale colonies corresponded to the primary strain, the red 
colonies to the variant. 
Similarly, strains were noted which mutated as regards the fermentation 
of other sugars, e. g. dulcite, saccharose, adonite. 
The variations exhibited bj these strains and the differences between the 
original and the variant strain are shown in Table XXIX. 
It is noteworthy that the majority of these mutating strains were of the 
B sub-group. While most of these variations were in the fermentation of 
lactose, it has been shown how similar mutations may occur in the fermenta- 
tion of dulcite, saccharose and adonite. 
It may also be noted here that no variations have been met with as 
regards the fermentation of inosite. Similarly the presence or absence 
of indol formation has been found to represent stable characters. Of course 
quantitative variations in indol production have been observed, and it has 
been shown how the amount of indol produced may be increased or 
diminished under certain conditions (Peckham, Horrocks). A considerable 
proportion of all the coliform strains examined were re-tested after two 
to three months from the time of their original classification, and in all cases 
indol formation or the absence of this property proved stable. 
Variation in Cultural Characters Associated with Variation 
IN Agglutinability. 
It has been shown that variation in biochemical reactions is not associ- 
ated with any change in the serological characters of the strain (Penfold) and 
this was confirmed in the case of one of the mutating strains described above 
(p. 339). 
Variations in colony characters of certain B. coli strains produced by 
growth on an inhibitory medium {i.e. containing brilliant green) have, how- 
ever, been found to be associated with considerable quantitative variation in 
agglutinability of the variant strains as compared with the original 
(Mackie). 
It can be understood how a group of bacteria which are constantly 
developing mutants with new biochemical characters has come to represent 
in the course of time a considerable number of different cultural types as 
seen in the B. coli group ; in the same way constant variation in serological 
characters, without change in biochemical reactions, would explain the highly 
specialised serological characters of the individual strain and the absence of 
specificity of an antiserum for the cultural type. 
The process of evolution and origin of new types going on among these 
G-ram-negative bacilli is not only of the greatest biological interest, but the 
study of these variations elicits some explanation of the great diversity of 
types and also the highly specialised serological characters of coliform bacilli. 
