ressions. Beyond impressions and ideas it
was unnecessary to look. Thus to look at a chessboard was to have a
number of sensations of black and white arranged in a certain order, to
listen to a piece of music was to experience a succession of loud and
soft auditory sensations, to handle a stone was to receive a group of
sensations of touch. To suppose that anything beyond these sensory units
was ever really experienced was futile fiction. Experience was a mosaic,
of which the stones were the detached sensations, and their washed-out
copies, the ideas.

If this analysis of the mind were correct--and its correctness was not
disputed for more than a hundred years, for were not the sensations
admitted to be the ultimate analysis of all that was perceived?--the
common-sense belief that knowledge revealed a world outside the thinker
was, of course, erroneous. For common sense could hardly treat 'things'
as merely 'sensations' artificially grouped together in space, each
'thing' being a complex of a number of sensations having relation to
similar complexes. It held rather that the successive appearances of
things were related in time, in such a way that they could be supposed
to reveal a single object able to endure in spite of surface changes,
and to manifest the identity of its sensory 'qualities.' Similarly, the
succession of ideas within the mind was for it supported by the inward
unity of the soul within which they arose. Moreover, Hume's analysis
made havoc of all idea, of 'causation.' If every sensation was a
separate being, how was it to be connected with any other in any regular
or necessary connection? Two events related as 'cause' and 'effect' must
be a myth.

These subversive consequences of his theory Hume did not conceal, though
he did not push his mental 'atomism' to its logical extreme. When he
defined material objects as 'coloured points disposed in a certain
order,' he was in fact admitting space as a relating factor; when he
spoke of the succession of impressions and ideas in experience, he was
tacitly assuming that what was apprehended was not a bare succession of
sensations, but _also_ the fact that they were succeeding one another,
and so allowing a sense of temporal relation. But further than this he
refused to go. The idea of a continuous self was fantastic. There was
nothing beneath the ideas to connect them. The notion of causal
connection was equally chimerical. Each sensation was distinct and
exis