139 
in part, ‘*atypival coli,’’ the possibilities of progress are very 
much reduced. Although it may be true that this method 
has, in skilful hands in Europe, given satisfactory results, it 
is obvious that unless each species of bacterium is isolated 
and studied separately, both as regards its position in nature 
and as regards its prevailing characteristics, we shall remain 
in our present position, talking of ** coli ’’ and “‘ atypical 
coli ’’ permanently. 
In order to demonstrate the unsuitability of the definition 
of ‘‘ coli,’’ and of the standards made use of in common 
practice in Europe and America, when applied to the Madras 
Presidency, it will be well to observe the result of applying 
them to the samples examined in our laboratory. Houston’s 
““ true coli’’ ferments lactose, ferments glucose, does not 
ferment saccharose, gives indol reaction, and the other tests 
already described. Organisms that fulfil these tests are 
indicative of fecal contamination and are more objectionable 
than organisms that ferment saccharose. ,By making use of 
MacConkey’s method it is possible to split up this ‘‘ true 
coli ’’ of Houston into about ro different bacilli (vide Table 
XII(a)), and if we add his “‘ non-typical coli’’ (that is 
to say, those that do not give the indol reaction), more are 
added. Now it must be remembered that all these organisms 
are looked upon as indicating the worst form of contamination. 
Apart from the objection of classing 1o different organisms 
under the term “‘ true coli,’’ it has never been demonstrated 
that the members of this group are egually objectionable, 
or in other words, egually represent recent and _ serious 
contamination. We have already shown that this is not 
the case. 
Savage’s definition of “‘ true coli’’ is virtually the same 
as Houston’s, but as regards the interpretation, he makes 
this statement ‘‘ all workers class both groups (saccharose 
and non-saccharose fermenters) as of equal value as excretal 
indicators.’’ Consequently all lactose fermenters except 
such as liquefy gelatine, do not give the indol reaction, and 
are non-motile (vide his previous remark on motility, indol, 
etc.) may be said to be of equal value as indicators of fecal 
contamination. Therefore, there are, according to the table, 
, 
b Lap 2 
