sham, and by Erasmus Darwin. We find
Stewart, in language which reminds us of later controversy, denouncing
the 'Darwinian School'[180] for theories about instinct incompatible
with the doctrine of final causes. It might appear that a philosopher
who has re-established the objective existence of space in opposition to
Berkeley, was in danger of that materialism which had been Berkeley's
bugbear. But Stewart escapes the danger by his assertion that our
knowledge of matter is 'relative' or confined to phenomena. Materialism
is for him a variety of ontology, involving the assumption that we know
the essence of matter. To speak with Hartley of 'vibrations,' animal
spirits, and so forth, is to be led astray by a false analogy. We can
discover the laws of correspondence of mind and body, but not the
ultimate nature of either.[181] Thus he regards the 'physiological
metaphysics of the present day' as an 'idle waste of labour and
ingenuity on questions to which the human mind is altogether
incompetent.'[182] The principles found by inductive observation are as
independent of these speculations as Newton's theory of gravitation of
an ultimate mechanical cause of gravitation.

Hartley's followers, however, could drop the 'vibration' theory; and
their doctrine then became one of 'association of ideas.' To this famous
theory, which became the sheet-anchor of the empirical school, Stewart
is not altogether opposed. We find him speaking of 'indissoluble
association' in language which reminds us of the Mills.[183] Hume had
spoken of association as comparable to gravitation--the sole principle
by which our 'ideas' and 'impressions' are combined into a whole; a
theory, of course, corresponding to his doctrine of 'belief' as a mere
custom of associating. Stewart uses the principle rather as Locke had
done, as explaining fallacies due to 'casual associations.' It supposes,
as he says, the previous existence of certain principles, and cannot be
an ultimate explanation. The only question can be at what point we have
reached an 'original principle,' and are therefore bound to stop our
analysis.[184] Over this question he glides rather too lightly, as is
his custom; but from his point of view the belief, for example, in an
external world, cannot be explained by association, inasmuch as it
reveals itself as an ultimate datum.

In regard to the physical sciences, then, Stewart's position
approximates very closely to the purely 'empirical' view