360 
Transactions of the Royal Society of South Africa. 
Variation in Biochemical Characters among the Coliform Bacilli. 
Such variations have been observed occurring spontaneously in culture 
medium, and aft'ord some indication of the process of evolution going on 
among these organisms under natural conditions. The marked diversity of 
types in this group as regards cultural reactions has been well shown by the 
work of all those who have studied these organisms, and in the serological 
observations already recorded the highly specialised characters of individual 
strains have been alluded to ; it may be assumed, therefore, that new types 
or species are constantly being developed from pre-existing varieties, and 
that these bacilli tend to acquire new characters which are not spontaneously 
lost, i. e. that the tendency is a progressive one and represents the origin of 
new species. 
In 1907 Massini* described a non-lactose-fermenting Gram -negative 
bacillus coU mutabilis) which on Endo-agar developed lactose-fermenting 
mutants represented by red papillae on the pale colonies of the original strain, 
and after further subculture as red colonies. This was corroborated by Burk 
and later by Miiller in the case of the fermentation of other carbo- 
hydrates. 
In studying the fermentative reactions of coliform bacilli it has been 
noted that certain strains may not show any obvious change in a particular 
sugar, e, g. lactose, until the lapse of several days ; organisms of this type were 
investigated by Penfold, who showed that these slowly fermenting strains 
were primarily non-1 actose-fermenters which in fluid media threw oft* lactose- 
fermenting variants, the variant producing rapid fermentation of the sugar. 
Thus the obvious difference between the variant and the original strain lay 
in the rate of fermentation of lactose. The strains studied were characterised 
by the formation on differential medium (such as MacConkey's neutral red 
lactose agar or Endo-agar) of pale colonies, which developed after a few days 
red papillae as in the case of Massini's strain. 
A number of strains were met with in the course of my own study of the 
B. coli group which, though primarily non-lactose-fermenters, in fluid medium 
containing lactose apparently threw off ]actose-fermentiug variants. All 
these organisms formed pale colonies on MacConkey's agar; certain of them 
corresponded to the type described by Penfold in developing red papillae. 
By subculturing the papillae on another plate of MacConkey's medium, red 
colonies were grown representing a variant which differed from the original 
strain in producing rapid fermentation of lactose in fluid medium, whereas 
the sugar was only fermented after several days (and often without gas pro- 
duction) by the original strain. 
Some of these organisms, on the other hand, showed no evidence of 
mutation on solid media containing lactose (e. g. MacConkey's agar), i. e. red 
papillae were not observed, but on subinoculating from the fluid lactose 
* Eeported also by Neisser (1906). 
