138 BACTERIA IN WATER 
from a score or more per C.c. in very pure waters to a few hundreds 
in a moderately pure reservoir; and from this number to many 
thousands in streams which are badly contaminated with sewage. 
I. THE PURITY OF DRINKING-WATERS 
In determining the purity of drinking-water we are not so much 
concerned with the number of bacteria it contains as with the 
kinds. Very large numbers may be present and yet the water may 
be perfectly wholesome, while, on the other hand, with only a small 
number present the water may be deadly. As we shall see in a 
later chapter, the bacteria in milk may be reckoned by the millions 
per c.c., and yet the milk may be perfectly healthful; and at the 
same time bacteriologists regard with suspicion water that con- 
tains them only in thousands. The reason for this difference is 
simple. When milk contains these large numbers they are almost 
sure to be harmless types; but if water contains even a few thou- 
sands, the typhoid bacillus is likely to be among them. It is im- 
possible to condemn any sample of water simply from the number 
of bacteria which it contains; still the number serves as a useful 
measure of purity for the following reason: Water that is fairly 
pure and contains only the bacteria liable to come from ordinary 
sources seldom contains more than a few hundreds of bacteria per 
c.c.; 1t is only water that is receiving contamination from sewage 
or some other source of decaying filth that contains large numbers 
of bacteria. Hence, the finding of large numbers of bacteria in a 
water-supply suggests sewage contamination and the water at 
once becomes suspicious. 
SEWAGE CONTAMINATION 
Water may receive hosts of bacteria from various sources, but 
the one great and almost only source of real danger is from sewage 
contamination. Most of the types of bacteria found in nature 
and in natural water are perfectly harmless, so that it makes little 
difference whether they are abundant or few in the water we drink. 
