708 
SUMMARY OF CURRENT RESEARCHES. 
is accepted as good. The inferences from the first and last of these three 
degrees are obvious, while the interest of the second lies in ascertaining 
whether and to what degree a water is rendered impure by B. coli. 
Mention has already been made of the satisfactory results to be obtained 
by Vincent’s method, but the object is more quickly attained by using 
bouillon with 5 per cent, milk-sugar. 
A number of flasks containing this medium are inoculated with 
variable quantities of the water to be examined, e. g. with 1, 10, 20 drops, 
and incubated at 35°. If coli bacteria be present, in 12-24 hours there 
will be strong fermentation, and this can be rendered more noticeable 
by shaking the flasks a little. Often water and putrefactive bacteria 
do not excite fermentation in milk-sugar. When gas is formed the 
j^resence of coli bacteria may be safely reckoned on. 
