e up
almost wholly of chance, non-selective, unavoidable deaths. As the
number of children from marriages, both parties to which reached
85 years of age, is so small as to render any safe conclusions
impossible, our only recourse is to take the children of the
85-year-old fathers and the children of the 85-year-old mothers,
add them together, and strike an average. But we must recognize
that the minimum so obtained is nevertheless still too large,
because among the consorts of the long-lived fathers and mothers,
some died early with the result of increasing the infant mortality.
The infant mortality with the 85-year-old fathers and mothers is
found to be 11.2%-15.4%, average about 13%. The total
child-mortality reaches 31-32%, of which the 13% make about 40%.
Accordingly at least 60%, and considering the above mentioned
sources of error we may say two-thirds, of the child mortality is
selective in character. That accords reasonably well with the
55-74% which Pearson found for the extent of selective deaths in
his study.

In general, then, one may believe that more than a half of the persons
who die nowadays, die because they were not fit by by nature (i. e.,
heredity) to survive under the conditions into which they were born.
They are the victims of lethal natural selection, nearly always of the
non-sustentative type. As Karl Pearson says, "Every man who has lived
through a hard winter, every man who has examined a mortality table,
every man who has studied the history of nations has probably seen
natural selection at work."

There is still another graphic way of seeing natural selection at work,
by an examination of the infant mortality alone. Imagine a thousand
babies coming into the world on a given day. It is known that under
average American conditions more than one-tenth of them will die during
the first year of life. Now if those who die at this time are the
inherently weaker, then the death-rate among survivors ought to be
correspondingly less during succeeding years, for many will have been
cut down at once, who might otherwise have lingered for several years,
although doomed to die before maturity. On the other hand, if only a few
die during the first year, one might expect a proportionately greater
number to die in succeeding years. If it is actually found that a high
death-rate in the first year of life is associated with