390 j. DEWEY :

sent in the same consciousness. Their unlike significance makes them, by necessity, two distinct consciousnesses- Unity and difference, relation in short, is always a matter of significance, of content for intelligence, and not of psychi- cal existence. When we say then, as Mr. Spencer and all the later English writers do say, that a sensation is nothing until it is identified and discriminated that is, brought into relations of unity and difference it is necessary to remember that the identification and the discrimination are elements of meaning, of relation to intelligence. The sensations, or existences, never unite themselves, and never differentiate themselves. But sensations, as they exist in conscious ex- perience, are always united and differentiated. What is this but to say that intelligence is necessarily involved in every sensation as known, and, therefore, that it is impossible to derive intelligence from any combination of sensations ? Let us remember two things : first, a sensation is not knowledge until united and differentiated ; secondly, these processes have absolutely no reference to the existence of the sensa- tions, but only to their significance, to the meaning conveyed by them. Can we avoid drawing the conclusions : negatively, that relations that is, connexions of unity and difference of meaning can never be produced by sensations as psychical existences ; and, positively, that the factor of relation or ideal significance is necessarily required to make sensations elements of conscious experience ? This brings us to the fact that relations are thoroughly ideal. Lewes frequently noted that science is a process of idealisation, but he seems never to have realised either the true import of idealisation, or the fact that all knowledge, perception included, requires the ideal element. Idealisation is not a process of departure from the material presented in perception, for this material is itself ideal. The idealisation of science is simply a further development of this ideal element. It is, in short, only rendering explicit and definite the meaning, the idea, already contained in perception. In the act of perception we do not realise anything like the whole meaning of what the sensations convey ; our interpre- tation is fragmentary and inadequate. The other processes of knowledge, the so-called faculties of memory, conception, judgment, self-consciousness, &c., are only progressively fuller interpretations, as each introduces some ideal factor that is, relation neglected by the previous. Memory, for example, simply makes explicit that ideal relation of our present experience to past experiences, which is involved in every perception, and which indeed makes it what it is,