I5I 
these organisms present in this quantity of water. In the latter 
3 groups it is, of course, impossible to keep the total quan- 
tity of water used the same. Therefore, for purposes of rec- 
ording’ results, we consider that there are feecal bacilli in rc.c., 
if 3 out of the 5 tubes give the reaction. Precisely the same 
applies to the ‘I c.c. and ‘oI c.c. groups. This number of tubes 
of bile salt broth has given very satisfactory results. It is 
economical on the whole, and it certainly gives sufficient detail 
of the quantity of fecal bacilli present in any given sample. 
The total number of tube per sample used is 22, and we do not 
consider that it would be advisable to cut this down to any 
great extent. If, however, saving of media is of great im- 
portance, 3 tubes to each of which 5 c.c., I¢.c., ‘Ic.c., and ‘orc.c. 
of water are added, might be used ; and if 2 out of 3 in a group 
gave the reaction, fecal bacilli might be considered as present 
in that group. Only tubes that give acid and gas are counted 
as giving the reaction. The bile salt broth tubes are incubated 
at 37°C; 42°C is not satisfactory as some attenuated speci- 
mens of bacteria fail to ferment a carbohydrate at this 
temperature, but succeed at 37°C. 
The next step is the separation of the various bacilli that 
have grown in the bile salt broth, and their identification by 
passing them through the sugars. The method of doing this 
is as follows, and it is necessary to carry out these details with 
considerable care, otherwise great variety in the results will be 
obtained, owing to the possibility of causing too great acidity 
in the bile salt medium. Practically all of the bile salt broth 
tubes that have had water added, will show acid and gas after 
about 18 or 24 hours’ incubation. For sub-culturing the 
bacilli present a tube containing a large quantity of the water 
is taken. In actual practice we take one tube containing 
always 20 c.c. of the water, 18 hours after it has been inoculated. 
If tubes which have received smaller quantities of water are 
taken for this purpose, there is a danger that perhaps a single 
fecal bacilli may have been added to the tube, and in that 
case a pure culture of this bacillus is obtained, this cannot 
represent the bacteriological flora of the water as a whole, 
whereas the growth from 20 c.c. is much more likely to do so. 
Further, if the tube selected for sub-culture is allowed to 
