d its number of cities very small, compared with the area included.
The conclusion may be properly drawn, therefore, that statistics confirm
the general impression that life in the country is healthier than life
in the city.

_Accuracy of death-rate records._

One factor must be considered, however, since it plays an important part
in drawing conclusions from these kinds of statistics, and that is, the
accuracy of the records. In a city in which every one must be buried in
a public cemetery, and when the physician, the undertaker, and the
sexton all have to keep records which must agree, it is not easy for any
burial to occur without the fact being recorded and later registered in
the Census Office at Washington. But in the country, a person may be
killed by accident, for example, and buried in a private lot without the
undertaker recording it at all. The result is that the total number of
deaths seems fewer and the death-rate seems smaller than the facts
warrant, so that a false idea of the healthfulness of the community
obtains. That errors of this sort have existed in the past can be seen
by examining the death-rates for New York City and those for regions
outside that city for the past ten years:--

TABLE III. DEATH-RATES IN NEW YORK CITY AND ELSEWHERE IN NEW YORK STATE,
1898-1908

=======================================
New York Outside Difference
---------------------------------------
1898 20.4 14.5 5.9
1899 19.6 14.9 4.7
1900 20.6 15.0 5.6
1901 19.9 15.1 4.8
1902 18.6 14.1 4.5
1903 17.9 15.2 2.7
1904 18.5 17.3 1.2
1905 18.3 15.8 2.5
1906 18.4 15.7 2.7
1907 18.5 16.4 2.1
1908 16.8 15.5 1.3
=======================================

The decrease in the city rate is to be expected, since with greater
knowledge of sanitary matters, more precautions against disease would
naturally be taken. But it is not likely that the country is becoming
more careless, although the tendency to concentrate population even in
rural hamlets may have an effect. It is rather more likely that the
reports are made more carefully and that the records are more complete
now than formerly. The apparent increase in the number of deaths in
rural communities is, therefore, due to greater attention in reporting
deaths rather than to any real increase in