======================================================================
Year 1900 1901 1902 1903 1904 1905 1906 1907 1908
------------------------------------------------------------------------------
The registration
area 35.9 32.4 34.5 34.4 32.0 28.1 32.1 30.3 25.3
Registration
cities 36.5 33.9 37.5 38.2 35.2 30.1 34.2 32.9 25.8
Cities in registration
states 28.5 26.5 25.9 24.6 24.0 22.0 34.2 31.7 24.5
Rural part of
registration
states 34.6 28.8 27.0 24.7 23.8 23.0 28.6 26.0 24.3
==============================================================================

[Footnote 1: States in which full credit is given by U. S. Census Office
for Vital Statistics collected from all parts of the state.]

This table shows that, taking the United States as a whole, the
typhoid-rate in rural districts is generally less than in cities and
that in cities the rate is excessively high.

When it is remembered that by filtration of public water-supplies the
typhoid-rate may be brought down to about 15 per 100,000, and that
cities with pure water-supplies will not exceed that rate, it is plain
how serious is the danger from typhoid in such cities as Cohoes or
Oswego. The following table from statistics taken in New York State
shows the same conditions as Table VI.--

TABLE VII. SHOWING DEATH-RATES FROM TYPHOID FEVER PER 100,000 POPULATION
IN NEW YORK STATE AS INDICATED

==============================================================================
Year 1900 1901 1902 1903 1904 1905 1906 1907 1908
------------------------------------------------------------------------------
Cities average 25.4 23.9 23.4 22.6 21.6 19.1 19.0 20.7 20.1
Rural districts 32.0 27.3 23.4 22.1 21.8 21.8 20.2 19.3 20.8
Average of city
population -- 38.9 33.9 43.0 40.3 32.2 30.5 32.1 32.4
Average of rural
population -- 20.3 24.1 23.2 21.3 22.3 21.3 19.9 20.8
==============================================================================

The first line is the death-rate in cities, found by taking the ratio of
all the deaths from typhoid in cities to the population in those cities,
and the second line is a similar ratio for rural districts. If the
actual rates of the several cities be averaged, a method which has