IZ 
is usually met with in small numbers in cowdung in this coun- 
try. Bacillus cloace is sometimes extraordinarily common. 
The proportion of saccharose fermenters to non-saccharose 
fermenters is 524 to 505 (about I to 1). 
Human. Cowdung. 
Per cent. Per cent. 
Group I a Fok yee 40°3 
Group I] fe se. GA Q°4 
Group Til 7 eda Ore) 16°0 
Group IV ne et ee 35°3 
Placing the averages for cowdung and human feces side 
by side, it would appear that in human feces group No. I 
predominated, in cowdung Nos. I and IV. 
This conclusion may be interesting, but we doubt if it 
is of very much practical use to the water analyst. It would 
certainly not be safe to conclude, because the organisms in 
any sample of water (especially after exposure to the sun) 
followed the usual percentage of cowdung, or human feces, 
that therefore cowdung, or human feces, was the polluting 
agent. 
MacConkey (1906) considers that in the examination of 
milk these figures may be of considerable value, but even this 
is doubtful. 
In studying the figures (vide Tables II (a) and (0) ) of 
the various series, it is apparent that there are very consider- 
able differences in the percentage proportions at various times 
of the year. Compare, for instance, the percentages of series 
Nos. I and 7 in human feces ; again, series No. 4 human feces 
might easily be mistaken for a cowdung average; it very 
closely resembles series No. 2 cowdung. Series No. 3 cow- 
dung might easily pass for human faeces. Further, by reading 
the figures in the various groups, in Tables II (a) and (0) from 
left to right very considerable variations will be observed ; 
apparently groups 2 and 3 are subject to the greatest irregu- 
larities. 
The dividing of fecal organisms into these classes is of no 
practical value. In the first place, it is a purely laboratory 
classification, and, in the second place, we shall show conclu- 
