Willcox: Fewer Births and Deaths 
in 1890 the difference was 56.5%. 
These earlier rates are not strictly 
comparable with each other or with 
those for 1900 and 1910, partly because 
they speak for different areas, neither 
of which agrees with the registration 
States of 1°00, and partly because in 
them no attempt has been made to 
allow for differences in the age and sex 
composition of the two races or for 
changes in these respects. The general 
conclusions, however, that the death 
rate of each race has fallen by approxi- 
mately the same amount and that the 
ratio by which the Negro rate exceeds 
the white has been rising, seem to be 
supported by evidence enough to make 
them deserve acceptance. 
MARRIAGE AND LONG LIFE 
Still another influence upon mortality 
which has been recently investigated is 
that of marital condition. European 
figures indicated long ago that the 
mortality of husbands is much less than 
that of bachelors or widowers of the 
same age and the conclusion is now 
corroborated by figures for New York 
State. No doubt this difference is 
largely due to the selective process by 
which the men who marry are, on the 
average, at the time of marriage more 
healthy and vigorous than men of the 
same age who do not marry. Evidence 
that direct benefits to health accrue 
from married life is found in the death 
rate of widowers, which is much higher 
than that of husbands; in the death 
rate of Catholic celibate clergy, which is 
higher than that of Protestant clergy, 
most of whom are married; and in the 
death rate of all males during the years 
at which marriage is most common. 
The normal and usual course of mor- 
tality in each sex is for a slow but steady 
increase to begin just after the minimum 
is reached in the early teens, and to 
continue with accelerating rapidity to 
the end of life. To this course there is 
no significant exception among women, 
but among men in several countries 
the increase is checked or even turned 
into a slight decrease for a longer or 
shorter period between the ages of 
20 and 35 when marriage is most 
common. The probable explanation is 
123 
that the normal increase of mortality 
during these years is nearly or quite 
neutralized by the steady transfer of 
many men from the less healthy bachelor 
state to the more healthy married state. 
The explanation finds some support in 
the fact that during these years the 
death rate of bachelors and of husbands 
treated separately rises steadily with 
age. On the whole, the evidence justi- 
fies the inference that marriage exerts a 
directly beneficial influence on men’s 
health and prospects of longevity. For 
wives the facts are less favorable. In 
New York the death rate of wives 
between 20 and 29 is greater than that 
of single women at the same age period. 
The difference is slight and part of it 
no doubt is due to the fact that wives 
20 to 29 years of age are, on the average, 
more than two years older than spinsters 
belonging to the same age group. But 
as less than half the difference can be 
thus explained, it seems clear that in 
New York, as in most other areas in 
which the facts have been determined, 
during the early years of married life 
wives have a higher death rate than 
spinsters and probably that the differ- 
ence is largely due to the dangers attend- 
ing childbirth, especially among those 
bearing a first child. 
THE BIRTH RATES 
There are no American birth rates 
which extend over a long series of years 
and may be trusted as accurate. In 
default of evidence from this country I 
have used in the diagram the birth rates 
for England and Wales since 1838 and 
for Berlin since the beginning of the 
last century. (See Fig. 11.) 
The diagram shows that the birth 
rate was probably at its maximum about 
1875 and has decreased almost steadily 
since that date, that the decrease in 
England has been about one-third and 
in Berlin about one-half of the maximum 
amount, that the variations both from 
year to year and through longer periods 
were greater in the city than in the entire 
state, that the decrease since 1876 has 
been almost uninterrupted, that there 
was a marked depression in both areas 
in 1890, and that Berlin had a much 
more notable depression in 1871. If in 
