TYPHOID FEVEE IX DISTRICT OF COLUMBIA. 
33 
in river water has perhaps been exaggerated. It must be borne in 
mind that many colon bacilli in a surface water, such as the Potomac, 
derive their origin from the intestinal tracts of animals other than 
man. 
The filtered river waters of Lavuence, Albany, Philadelphia, etc., 
also contain a few colon bacilli, and the total number of bacteria here 
compare favorabl}^ with the number found in the effluents of the 
above-mentioned cities. Sand filtration of the above river waters, 
giving a somewhat similar bacterial effluent, as judged by the total 
number of bacteria and the colon bacilli, has been followed by a 
reduction in the tj^hoid death rate. 
According to the researches of Kinyoun and Sprague in 1897, the 
tap water contained the greatest number of colon bacilli during 
September and October, whereas the disease during that year pre- 
vailed to the greatest extent during the preceding months. 
It seems that, so far as the incomplete records are obtainable, 
typhoid fever prevails to a great extent at the same season in Wash- 
ington and over the entire watershed. This indicates that the causes, 
whatever the}" may be, act simultaneously along the entire course of 
the river. 
While it is true that the Potomac receives the sewage directly of 
many thousands of persons, much of this pollution is distant and 
greatly diluted. Self-purification takes place to such an extent that 
the w"ater at the intake averages comparatively few bacteria per 
cubic centimeter the year round compared with other grossly polluted 
streams and may be considered of fair sanitary quality compared 
with many such streams used for drinking purposes. 
It is probably true that in former years some of the t}"phoid fever 
in Washington was due to infection in the river water; but in view 
of the improvement of the water by storage, sedimentation, and 
filtration, it would seem reasonable to suppose that the danger of 
infection from the water may have been largely eliminated. 
Typhoid fever in Washington is mainly a summer disease. There 
is a great and sudden increase in the number of cases with the advent 
of hot weather. This seasonal prevalence repeats itself year after 
year with great regularity. If the bulk of the infection were water- 
borne, it would seem that we would have more marked irregularities 
of the curve from year to year, as is seen in other communities where 
much of the disease is plainly due to highly infected water. 
The seasonal prevalence of typhoid in Washington corresponds to 
that of Boston, New York, Baltimore, and other cities in which the 
water is believed not to play much part. 
If we attribute the bulk of infection to imported cases, contacts 
and milk, then it would not be necessary to consider that some un- 
29643—07 3 
