II5 
As time goes on the next marked change that would be 
noted would be the tendency of lactis erogenes to disappear 
from surface and middle samples, and to be replaced by bacil- 
lus vesiculosus. Within the first four weeks, unless the 
water was run off, the surface sample would be extremely 
pure. It would probably contain no lactose fermenters 
in 60 c.c. and no glucose fermenters in Ic.c. Supposing now 
the water of the lake is used to supply a town, the clear 
surface water would be drawn off, and the whole appear- 
ance of the water would gradually become slightly hazy 
and muddy due to silt. At this period (say 3 months 
after pollution) the water will show either pure cultures of 
bacillus vesiculosus or cloace, or a mixture of the two. The 
bacilli will also be numerous, very likely be present in ‘tr c.c, 
Bacillus P will also be extremely numerous, probably will 
be present in ‘or c.c. Beyond this point practically no change 
would take place. 
This appears to be bacteriology of the self-purification 
of a lake that has been badly contaminated. It is funda- 
mentally the same in rivers, particularly if they are slow 
and shallow, but it is not easy to demonstrate these various 
stages in deep rivers containing a certain amount of suspended 
solids in the water, where a quantity of pollution is constantly 
being added along the banks. 
