226 THE FACTORS OF THE MIND

Yet, taken rigidly, this principle, if it stood alone, would
seem destructive of all factorial work, as commonly con-
ceived. And Thomson himself has recently introduced a
second qualifying assumption. He now assumes a mind that
is not entirely structureless, but divided into ' regions' or

* sub-pools' within the total' pool.' This further assump-
tion, as it seems to me, is really a special application of our
second postulate, namely (as it manifests itself in psychology)
that both mental traits and individual minds exhibit not a
regular or homogeneous distribution, but a limited variation
with a limited independence, a confluence of individual
traits into * group-factors * and of individual persons into

* types'—in short, a mottled distribution into what Broad
calls c naturally cohering sets.'*

We have noted at an earlier point (p. 93) how the elabora-
tion of the method of matrix analysis by the modern
psychologist, on the one side, and the modern physicist,
on the other, and their simultaneous and successful employ-
ment within their respective spheres, suggests that as regards
ultimate logical structure both the mental world and the
physical world must be fundamentally akin. Here, as it
seems to me, we have the ground of that kinship made
explicit. Whether this common structure in its turn has
been imposed on both worlds a priori, like a set of Kantian
categories, by the logic and the laws and forms of thought
which every human analyst is forced to use, is an episte-
mological problem into which we need not enter. The main

factors are far more conspicuous at the stage of vocational guidance than at
that of educational guidance " ([41], p. 266 ; cf. [35], pp. 63 f,).

As regards the statistical and more general arguments for the conception I
have just criticized (and these are the arguments on which Thomson and
Bartlett chiefly rely) the difference between us would seem to be mainly one
of emphasis (cf. [137], p. 88 f., and Brit. J. Educ. Psych., IX, pp. 191 f.).

1 For the philosopher, I suppose, the problem is the ancient puzzle of
reconciling what Plato called the Many and the One. Certainly it is not
peculiar to psychology. " One of the most remarkable achievements of
current quantum theory is the way-it has surmounted the difficulty of giving
to the parts of the universe a kind of self-sufficiency, which does not cut them
oil from interaction with the rest." (Eddington, loc. cit.9 p. 127.) This
achievement, we are told, has been rendered possible by a mathematical
technique, which^ as we have seen (p. 165), is closely similar to that which
has been advocated here.