87° C. V8. 20° C. COUNT 291 
100-1000 cells per cubic centimeter...... .......... pure 
1000—10,000 cells per cubic centimeter.... .. ... . mediocre 
10,000—100,000 cells per cubic centimeter... ..... .. Impure 
Over 100,000. .......... cece ccc eee nee very impure 
Such a standard would be followed with difficulty to-day. Obvi- 
ously, according to our present information, such a scheme for grading 
waters is untenable. Probably Koch’s old standard of not over 100 
bacteria per cubic centimeter on gelatin is as accurate as any. It is 
evident that a very few cells of a certain variety would be far more sig- 
nificant than thousands of another variety. In natural waters, however, 
it has become established that a high content of organic matter usually 
means a high content of bacteria. High organic content usually means 
pollution and, in this way, the count itself may mean something. Pos- 
sibly a water with a large number of bacteria, according to the law of 
probability, might contain some objectionable forms. 
Relation of the 37° C. Count to the 20° C. Count. This has re- 
ceived the attention of a number of bacteriologists and their opinions 
sum up our knowledge with regard to the two counts. 
Whipple (1913) made the following comment: 
“ That the 37° count more nearly represents the sanitary quality of 
water than the 20° count is probably true in a general way, but it is not 
necessarily so.” He cites counts made at the Watertown, N. Y., fil- 
tration plant and states: 
“ The difference between the two counts is striking. The 20° counts 
are very high in the winter and spring but are low during the summer, 
whereas the 37° counts are low during the winter and spring and high 
during the summer, at which season the sewage which enters the stream 
is less diluted.” He calls attention to the fact that this seasonable 
difference gives a varying ratio, and suggests that it would seem advis- 
able to procure more comparative data before making the recommended 
change. Caird (1918) thinks it advisable to continue the use of gelatin 
to determine the efficiency of water-treatment plants, and he appends 
to his paper discussions by various waterworks men, whose opinions 
seem about evenly divided. Baton (1914) also protests against the 
recommended change. He concludes that if only one medium is to be 
used it should be gelatin as it is more reliable and serviceable. 
Gaub (1913) at the Washington, D. C., filtration plant found a vari- 
able ratio between counts on agar and on gelatin plated from the un- 
filtered water of the Potomac River. He also found a seasonal ratio 
which seemed to remain constant for several months at a time and then 
