METAPHYSICAL STATUS OF FACTORS 245

elusions to practical problems brings into strong relief the
helplessness of the mere factorist so long as he tries to rely
on his own unsupplemented efforts. It is as though the
surgeon were to trust to the study of gross anatomy alone,
declining all hints from physiology or cytology. No doubt,
in the history of nearly every complex science, the study of
broad relations and of observable types comes first. Fac-
torial psychology, with its correlation and classification of
persons and its correlation and analysis of traits, plays much
the same part in general psychology as the older c systematic
botany' and * morphological botany ? in general botany.
Just as the study of plant classification and plant structure
are now supplemented by c functional botany * and c plant
genetics,5 so too, as more direct experimental methods
become available, the first provisional results of factorial
psychology will have to be supplemented, and even very
largely superseded, by the functional and genetic study of
the mind. Conversely, I believe, in the other biological
sciences, many of the problems which have not yet yielded
to direct attack could be elucidated, and perhaps partly
solved, if analysed by the statistical devices of the factorist.

The Applicability of Factor-analysis in Other Sciences.—
And this brings me to the last conclusion that emerges from
my inquiry. Once we discard the notion that our c factors'
are essentially * factors in the mind,5 once we realize that
' factor-analysis,' so far from being a device adapted
exclusively to the problems of the psychologist, is simply a
quantitative refinement of common logical procedures, we
shall not only appreciate more justly its special merits and
its inevitable limitations as an instrument for studying the
mind: we shall also perceive its manifest applicability to
other fields of work.

I have already cited a research in which factor-analysis
has been successfully used to examine the existence and
nature of physical as well as mental types. Its extension to
the study of numerous other anthropological or ethnological
problems is almost equally obvious. I will mention only
one example of special interest to current psychology.
The statistical study of race-differences has proceeded on
the assumption that human races form clear-cut types,