NATURAL SELECTION IN MAN 187 
best regulated asylums the death rate ‘‘is hardly less than 7 per 
cent, even under favorable conditions,”’ which is about four times 
as great as should exist in well-regulated municipalities of the 
ordinary population. If, however, we take out certain forms of 
insanity, such as paresis and organic dementia, we have the ratio 
somewhat reduced. In any case, however, it will decidedly exceed 
that amongst the general population. The death rate in asylums 
is less than that of the insane outside of these institutions. Barr 
(Mental Defectives, p. 131) states that out of 625 cases of mental 
defectives of whose deaths he had records, ‘‘ the largest number of 
deaths occurred between 10 and 20 years; but comparatively 
few passed the 25th year and exceptional cases appeared from 30 
to 40 years.” According to Clark and Stowell in the New York 
City Children’s Hospitals and Schools the mortality among the 
feeble-minded is double that of other children, and the mortality 
of the lowest grades, idiots and imbeciles, is four times as great as 
among the feeble-minded. With the higher grades of the feeble- 
minded the expectation of life is much greater, but among these 
natural selection takes a relatively heavy toll as is evinced by 
their high infant mortality. 
It is a fair inference that natural selection causes a higher 
mortality among those who, while not feeble-minded, are below 
the general average of intelligence. Not only is their station in 
life apt to be such as to raise their death rate, but through igno- 
rance or lack of the ability to afford the proper surroundings for 
their children they have a high infant mortality which tends to 
offset, in a measure, their greater fecundity. 
Contrasted with the rather high general death rate of inferior 
stocks is the relatively low death rate of the classes with excep- 
tional intelligence. Sir Francis Galton has noted that English 
men of science as a class are long lived, and Cattell finds that the 
death rate and especially the infant mortality in the families of 
American men of science is unusually low. The death rate is 
relatively low in professional classes in general and among others 
who have achieved a noteworthy success in other fields. If it is 
said that their reduced death rate is due to better environment 
