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 existed in its own right.  It could therefore occur alone.  There was nothing to link together the distinct impressions.  Hence necessary connection in events could not be more than a fiction of the mind based on expectation of customary sequences; how the mind he had described as non-existent could form an expectation or observe a sequence was calmly left a mystery.

Hume, then, seemed to leave to his successors in philosophy a task of synthesis.  He had tumbled the soul off her high watch-tower, but how to combine her shattered fragments again into a working unity he declined to say.  He saw the sceptical implications of his analysis, but professed himself unable to suggest a remedy.

He had, however, made the embarrassments of the theory of knowledge sufficiently clear for Kant, his most important successor, to hit upon the most obvious palliative, and in the Critique of Pure Reason Kant set himself to patch up Hume’s analysis.  Experience as it came through the channels of sense, he admitted Hume had analysed correctly; it was ‘a manifold,’ a whirl of separate sensations.  But these per se could not yield knowledge.  They must be made to cohere, and the way to do this he had found.  The mind on to which they fell was equipped with a complicated apparatus of faculties which could organize the chaotic manifold of sense and turn it into the connected world which common sense and science recognize.  First it views the data of sense in the light of its own ‘pure intuitions,’ and, lo! they are seen to be in Space and Time; then it solidifies them with its own ‘categories,’ which turn them into ‘substances’ and ‘causes’ and endow them with all the attributes required to sustain that status; finally it refers them all to a Transcendental Ego, which is not, indeed, a soul, but sufficiently like one to provide something that can admire the creative synthesis of ‘mind as such.’

Had Hume lived to read Kant’s Critique, he would probably have jeered at the vain complications of Kant’s transcendental machinery, and made it clear that between the primary manifold of sensation and the first constructions of the intellect there still yawns a gulf which Kant’s laboured explanations nowhere bridge.