288 
TYPHOID FEVER IN DISTRICT OF COLUMBIA. 
has also been found to be about 10 per cent higher in tap ^Yater than 
in the freshly filtered water, indicating, as might be expected, that 
some aeration takes place after the filtered water leaves the basin. 
On the other hand, the nitrites are practically the same in both, 
whereas the chlorine and volatile matter were found to be slightly 
greater in tap water than in the water of the storage basin. As 
already stated, however, none of the differences are sufiiciently great 
to be of any real significance. On the other hand, as may be seen 
from Table 25, during a period of about eleven weeks, from July 16 to 
September 28, the tap water showed a slight improvement in quality. 
An inspection of our results will show a greater or less dirninution in 
the quantities of free and albuminoid ammonia and nitrites, and 
also a decrease in the amount of oxygen consumed, and while’ on 
account of delay in securing the necessary apparatus it was not pos- 
sible to make daily readings of the turbidity of the tap water when 
this investigation was first undertaken, it is a matter of general 
observation that the water delivered from the taps and hydrants 
throughout the city during the months of August, September, Octo- 
ber, November, and part of December was considerably better, go far 
as its general appearance is concerned, than it was when the present 
investigation was begun. When this investigation was begun, the 
specimens of tap water brought into the laboratory from various 
localities throughout the city showed a distinct yellowish color and a 
slight though quite perceptible turbidity, and one specimen at least 
was markedly unattractive in both regards. On the other hand, the 
specimens collected at a later period during the investigation showed 
practically no color and no turbidity; especially was this the case 
with the specimens of tap water examined on September 28, and 
this high efiiciency of the sand filters, so far as the removal of color 
and turbidity is concerned, was steadily maintained during the 
months of September, October, November, and the greater part of 
December. 
Toward the end of December, 1906, however, the water supplied 
the city of Washington again became more turbid. Daily turbidity 
readings during the month of January, 1907, on samples of water 
collected from a number of taps and hydrants over the city show 
the tap water to have had an average turbidity of 12 as compared 
with an average turbidity of 2 to 3 during the months of July, August, 
and September, 1906, so that for some reason there was, so far as 
turbidity is concerned, a distinct deterioration in the quality of the 
water delivered to the city during the month of January. This 
increase in turbidity was nearly coincident with the onset of the very 
cold weather on or about December 24, 1906. 
By way of further comparison there is submitted in Table 26 a 
summary of the analyses of tap water for the summer period of 
