radiate from the definition of knowledge as a centre" (p. lvii). The author thinks that there is some ground for Stewart's suggestion that the Fourth Book was the part of the 'Essay' which was first written and which led to the inquiries of the other books. Whatever may have been the order of composition, it is essential to bear in mind that, although the 'Essay' contains much matter that is psychological, logical, and metaphysical, its main purpose is epistemological. But it is also necessary to remember, in reading Book II, "On Ideas," that ideas are not for Locke equivalent to knowledge. It is true that 'where we have no ideas we can have no knowledge,' and it was this fact—that ideas form a necessary element in our knowledge—which led Locke to undertake an account of their origin. But in Book II he is well aware that he is treating ideas in abstraction from the concrete system of knowledge, where they form not the whole, but simply one element. There are, indeed, many passages in that book in which Locke restores to the idea its epistemological reference, but the main problem is to give the psychological analysis and genesis of states of consciousness, quite apart from their actual employment in the structure of knowledge. This inquiry had grown out of Locke's problem about knowledge, and was necessary in order to answer it completely; but it would be a mistake to confuse Locke's psychological account of ideas with the epistemological conclusions of the Fourth Book. From this point of view, then, we can see that much of Green's criticism of Locke in the Introduction to Hume's Treatise ignores both Locke's account of knowledge, and the psychological nature of the Second Book.
Professor Fraser's interpretation of the "Essay" is already well known and pretty generally accepted. One of the chief merits of that interpretation, it seems to me, consists in his refusal to apply to his author distinctions and categories which have only arisen as a result of subsequent reflection. He shows that although the problem of analysing out the mind's contribution to knowledge was foreign to Locke's thought, yet in his candid account of ideas he was constantly obliged to recognize that contribution. Yet there is no drawing of hard and fast lines between 'sensationalism' and 'intellectualism' in order to force Locke into one school or the other.
While Professor Fraser does full justice to the general significance of Locke's polemic against abstract ideas, he yet maintains that so far as Descartes is concerned the attack was not justified. A letter from Descartes to Regius is quoted to show that the former did
