25 
During the three months July, August, and September there were 
in 1906, 541 cases; in 1907, 376 cases; in 1908, 326 cases. In the pre- 
ceding month of June, 1906, 64 cases; 1907, 42 cases; 1908, 64 cases. 
The rate increased quite sharply in the latter half of June, and 
reached its height early in July, continued high until the early part 
of September, when there was a marked decrease in the rate. 
The sharp rise in the curve in October was due largely to the occur- 
rence of a pronounced milk outbreak, which was responsible for about 
45 of these cases. 
Chart Xo. 1 shows by curves the rate of prevalence of the disease 
for corresponding periods of the three years 1906, 1907, and 1908. 
Chart Xo. 2 shows the cases by dates of onset and by attributed 
causation. 
It is interesting to note that the pronounced summer rise in the 
curve has become decidedly less each year. The curves for 1907 
and 1908 differ in that the rise in 1908 occurred earlier and the high 
rate continued more uniformly and later — until the early part of 
September. 
The marked rise from about July 20 to August 20 which occurred 
in 1907 and 1906 was absent in 1908. (See Chart Xo. 1.) 
In the summer months — June, July, and August — of 1908 there 
occurred 300 cases, as against 310 in the same peroid of 1907 and 
505 in the same period in 1906. 
For what may be termed the typhoid months here — July, August, 
and September — there developed in 1908 326 case; in 1907, 376 cases; 
and in 1906, 541 cases. It is evident that for the summer and early 
fall the disease was decidedly less prevalent in 1908 and in 1907 than 
in 1906. In 1908 the prevalence was slightly less than in 1907. 
It will be noted (Chart Xo. 1) that, though the typhoid-fever curve 
has changed considerably in height for the different years, it has 
shown in each of the three years a marked summer rise. In other 
words, the difference has been more in degree than in character. 
Chart Xo. 3 shows by curves the monthly death rate from typhoid 
fever since 1888 in TTashington, Baltimore, and Boston. 
Aliile rather marked differences occurred in the prevalence of 
the disease in Washington for the warm season of the three years, 
the rates for the winter, spring, and late fall seasons were remark- 
ably parallel. (See table, p. 38.) Thus the total cases reported 
during January, February, March, April, May, October, Xovember, 
and December of the three years were as follows: In 1908, 449 cases; 
1907, 451 cases; 1906, 444 cases. 
It is evident, therefore, that the difference in rate for the three 
years has been due to the difference in extent of operation of some 
