100 
In answer to the third question, we are satisfied that, as the results 
of our studies, the bulk of the typhoid fever in the District of Columbia 
during the years 1907 and 1908 was not caused directly by water- 
borne infection. The facts do not prove, however, that some of the 
infection during those two years may not have been water borne. 
The amount of infection so conveyed may have been but an insignifi- 
cant part of the whole, }^et the amount so added may have been suf- 
ficient to have accounted for enough typhoid fever to keep the rate 
higher than it should be for a city under the sanitary and climatic 
conditions of IVashington and having pure water. 
During the summer periods of both years the water was, according 
to the present generally accepted bacteriological standards, of good 
sanitary quality, and it certainly does not seem probable that such 
water could have been directly responsible for more than a very insig- 
nificant part of the infection. Yet in these summer periods the 
typhoid-fever rate was much higher than in the other seasons of the 
two years. 
While we advocate still greater improvement of the city’s water 
supply by increased storage or the use of a coagulant, we believe that 
in order to materially reduce the prevalence of typhoid fever in the 
District of Columbia measures will have to be directed against other 
factors concerned in the transmission of the disease, especially con- 
tact and milk. 
MILK. 
The 542 cases gave the following history as to the use of milk 
during the thirty days prior to onset of illness : 
" Number 
How used: of cases. 
As a beverage 228 
On fruits or cereals, but not as a beverage 141 
In hot coffee or tea only 64 
As ice cream only . 44 
As ice cream and as condensed milk 18 
As condensed milk only 12 
None in any way 32 
Not determined 3 
Total - 542 
Of those using milk as a beverage, one used boiled milk exclu- 
sively; and of those using milk on fruits and cereals, one used pas- 
teurized milk exclusively. 
The large proportion of the cases which gave a history of having 
used raw milk during the thirty days previous to illness is striking 
when considered in connection with the data on milk obtained for the 
general population of 32 blocks of the city. (See Table No. 1, p. 15.) 
