In connection with this last vital point, which is decisive, call to mind a celebrated page of Sainte-Beuve where he defines his method: "Enter into your author, make yourself at home in him, produce him under his different aspects, make him live, move, and speak as he must have done; follow him to his fireside and in his domestic habits, as closely as you can... "Study him, turn him round and round, ask him questions at your leisure; place him before you...Every feature will appear in its turn, and take the place of the man himself in this expression... "An individual reality will gradually blend with and become incarnate in the vague, abstract, and general type...There is our man..." Yes, that is exactly what we want: it could not be better put. Transpose this page from the literary to the metaphysical order, and you have intuition, as defined by Mr Bergson. You have the return to immediacy. But a new problem then arises: Is not our intuition of immediacy in danger of remaining inexpressible? For our language has been formed in view of practical life, not of pure knowledge. IV. The immediate perception of reality is not all; we have still to translate this perception into intelligible language, into a connected chain of concepts; failing which, it would seem, we should not have knowledge in the strict sense of the word, we should not have truth. Without language, intuition, supposing it came to birth, would remain intransmissible and incommunicable, and would perish in a solitary cry. By language alone are we enabled to submit it to a positive test: the letter is the ballast of the mind, the body which allows it to act, and in acting to scatter the unreal delusions of dream. The act of pure intuition demands so great an inner tension from thought that it can only be very rare and very fugitive: a few rapid gleams here and there; and these dawning glimpses must be sustained, and afterwards united, and that again is the work of language. But while language is thus necessary, no less necessary is a criticism of ordinary language, and of the methods familiar to the understanding. These forms of reflected knowledge, these processes of analysis really convey secretly all the postulates of practical action. But it is imperative that language should translate, not betray; that the body of formulae should not stifle the soul of intuition. We shall see in what the work of reform and conversion imposed on the philosopher precisely consists.