GREEN'S METAPHYSICS OF KNOWLEDGE. 87 If I might hazard a conjecture, I should say that Green was misled in this matter by an error similar to that which mars his theory of reality. As in dealing with that subject he seems to have supposed that all those who did not hold "with him the opinion that reality consists in relations, necessarily accepted Locke's theory that reality consists in simple sensation ; so in dealing with the nature of the con- scious self he_sains to have thought that the only alter- native to his view that mind is a unifying principle not in tune, IsT Hume's theory that mind is nothing more than a stream of mental phenomena. "We thus avoid," he says (p. 63), "the necessity of facing the question how an object determined by relation can have its being in a consciousness which consists of a series of occurrences. Even ' knowledge,' though we often mean by it a system of known facts or laws, is apt to lose this sense when we speak of it as a form of consciousness. It then becomes merely the mental event of arriving at an apprehension of related facts. It does not represent the relation of the facts in consciousness. That there must be such a relation of them in consciousness, and that a onuctotuiMM consisting of events cannot contain such a relation, is a conclusion which we avoid by eviscerating knowledge of its content and transferring this content from consciousness to ' external things '." Let it be observed in passing that, however absurd may be the view that consciousness, i.e., a conscious self, consists of events, the reasoning by which in this passage it is refuted, though common enough in transcendental writings, is open to grave suspicion. It is surely dangerous to talk of that transcendental "combining agency" the self as an in- ventor talks of a machine : ' constructed in this way it cannot perform this ; constructed in the other way it may perform that '. Reasoning of this kind presupposes a. general knowledge of jnecbanics^j^pplied to the ^criticism of a particular machine. But we have no such general ^knowledge ofThe mechanics of consciousness. Ve know perh aps "something about our own, what it is and how it acts. But we cannot infer from this how a consciousness .of an_entirely different kind would act ; we cannot make a working modp.l nf itj we cannot even in the vaguest "manner imagine what it would be like. Though it may therefore be correct to say that, since " a consciousness con- sisting of events" is a meaningless expression, no assertion about it can be significant, it cannot be correct to suggest, as Green certainly does suggest, that by contemplating the structure of this hypothetical consciousness we can a priori discern its incapacity for knowledge. By direct intuition we may assure ourselves that we are something more than a succession of mental phenomena ; it is superfluous therefore, even if it be possible, to deduce from a general theorem re-
