The first duty of radical empiricism, taking given conjunctions at their face-value, is to class some of them as more intimate and some as more external. When two terms are _similar_, their very natures enter into the relation. Being _what_ they are, no matter where or when, the likeness never can be denied, if asserted. It continues predicable as long as the terms continue. Other relations, the _where_ and the _when_, for example, seem adventitious. The sheet of paper may be 'off' or 'on' the table, for example; and in either case the relation involves only the outside of its terms. Having an outside, both of them, they contribute by it to the relation. It is external: the term's inner nature is irrelevant to it. Any [Footnote 1: Particularly so by Andrew Seth Pringle-Pattison, in his _Man and the Cosmos_; by L.T. Hobhouse, in chapter xii (the Validity of Judgment) of his _Theory of Knowledge_; and by F.C.S. Schiller, in his _Humanism_, Essay XI. Other fatal reviews (in my opinion) are Hodder's, in the _Psychological Review_, vol. i, 307; Stout's, in the _Proceedings of the Aristotelian Society_, 1901-02, p. 1; and MacLennan's, in the _Journal of Philosophy_, etc., vol. i, 403.] book, any table, may fall into the relation, which is created _pro hac vice_, not by their existence, but by their casual situation. It is just because so many of the conjunctions of experience seem so external that a philosophy of pure experience must tend to pluralism in its ontology. So far as things have space-relations, for example, we are free to imagine them with different origins even. If they could get to _be_, and get into space at all, then they may have done so separately. Once there, however, they are _additives_ to one another, and, with no prejudice to their natures, all sorts of space-relations may supervene between them. The question of how things could come to be, anyhow, is wholly different from the question what their relations, once the being accomplished, may consist in. Mr. Bradley now affirms that such external relations as the space-relations which we here talk of must hold of entirely different subjects from those of which the absence of such relations might a moment previously have been plausibly asserted. Not only is the _situation_ different when the book is on the table, but the _book itself_ is different as a book, from what it was when it was off the table. He admits that 'such external relations