86 
It should be observed that coli communis was only once 
isolated in this period, in spite of the fact that a constant 
stream of this bacterium was coming in. Note the great 
number of varieties present, and the truly fecal nature of 
the mixture of organisms shown in each analysis. 
Sample No. 5 taken on November 29th shows that the 
action of sunlight has already commenced ; for the surface 
layers of the lake are considerably purer than the bottom 
and middle. This improvement goes on steadily till sample 
No. 8 (a) taken on December 18th, which shows that the 
surface water is extremely pure. Note that with the improve- 
ment lactis erogenes becomes common. 
On December 5th a heavy downpour brought in to the 
lake a further rush of very highly polluted water, though it 
does not appear to have raised the total number of lactose 
fermenting organisms very much. As this new pollution 
was chiefly shown by the presence of coli communis, im- 
provement in quality of the water occurred very rapidly. 
After the rain on December 5th, note the disappearance of 
lactis erogenes, and its preponderance again a few weeks later. 
The sample taken in January 6th is very unusual. In 
this case the surface water is distinctly inferior in quality to 
that in the lower depths. The reason for this is, that when the 
sample was taken heavy rain was falling, probably a layer of — 
warm highly polluted surface water was lying on the surface of 
the lake, and thorough mixing between the purer water and this 
contaminated layer had not yet taken place. The sample taken 
a week later shows that mixing of the whole volume of water 
had occurred, for the lake has risen nearly six inches, and 
surface, middle, and lower layers are equally contaminated. 
Samples 11 a, 6, c, taken on the 28th January, are quite 
inexplicable and probably'there is an error somewhere. 
As has already been pointed out, the samples taken week 
by week do not show a progressive purification. There is a 
considerable variation, particularly between samples 15 a, 0, ¢, 
and 16 and17a,b,c. The explanation of this irregularity 
is due to an in-rush of irrigation water from paddy fields down- 
one or other of the inlet channels. In spite of this, the surface 
water remains always very pure. 
