W. KNIGHT, HUME. 445 of the Treatise), which is invaluable as bringing into clear light the psychological atomism of Hume and his general method of procedure. The account of the distinction between impression and idea seems unsatisfactory. Prof. Knight's statement appears to imply that Hume, guided by a clumsy metaphor, regarded an impression as something stamped upon the mind by an unknown x. To me it seems clear that the antithesis, as Hume intended it, existed only within and for consciousness. Whenever he touches on the subject, he is evidently endeavouring to express a meaning for which he could not find words. He seems to be aware how unsatisfactory it is to name the distinction as one of liveliness merely. It appears to me that the distinction he was feeling after may be expressed as follows : A mental content " gently introduced " into consciousness by a pre-existing mental content is an idea; a mental content which "makes its way into our thought without such introduction, and which may therefore be said to "enter with force and violence," is an impression. I think that, if Hume were studied with the same tender care as Kant, this would be the general meaning elicited from his statements. Where Prof. Knight confines himself purely to the exposition of Hume's treatment of special problems, he is in the main clear and correct. His defect is want of sympathy he does not enable the reader to enter into Hume's position, so as to realise the historical necessity of it. In short, he fails to make Hume credible. The critical matter is, in point of quantity, in excess of the expository, but in point of quality inferior to it. Hume's doctrine of causality is said to be " a development (and a necessary development) of the doctrine which limits our know- ledge to the realm of sense-experience ". It would have been nearer the truth if Prof. Knight had said that it was a develop- ment of the very opposite doctrine viz., the fundamental doctrine of Locke^that universal and necessary knowledge is only to be found in discoverable connexions of ideas.^ According to ]?rof. Knight, " the curious thing is that he (Hume) never seems to have seen that this link of connexion, if obtained at all, must be obtained a priori ". Now, on the contrary, this was just the point that Hume was most keenly aware of. ^Hisjwhole difficulty arose from the alleged a priori necessityjo^a^elation_nor implieif mlhe nature of the ideaSTrelated. It is~true that Hume was far from" seeing thaT^ffie causal Judgment was a priori in the Kantian sense viz., as being implied in every cognition of objective change. But this is not the sense in which Prof. Knight uses the phrase a priori. He allows that the " senses take note of phenomenal succession ". " Some impressions reach us simultaneously, i.e., we combine them in Time." Thus, in order to make the judgment of causality possible, "the intellect strikes through the phenomenal chain . . . and discerns the inner vinculum". "The judgment of causality flashes forth from the
