such a method, even if successful, would be defective and inadequate. Even at its best, even if it could be shown that reality could be known only as a coherent system of thought-relations, it would not necessarily follow that reality was nothing more, and he would not necessarily have proved anything but the impotence of his thought to grasp reality, by reducing his symbolical expressions for reality to absurdity and contradiction. Thus his proofs cannot prove what he desires, and his refutations only recoil upon his method.
But it may be shown also that his criterion is not valid. He suggests (p. 267) a triple test of rationality, a triple basis for the metaphysical assertion that reality is thought, (1) "The agreement between the inferences drawn from the experience of our different senses; (2) the agreement between the judgments of different persons; (3) the harmony of present experience with the results of their and our previous experience, constitute between them the test of reality." It is to be feared that "between them" they fall very far short of giving a reliable test of reality.
(1) The first is open to objection as a matter of fact. It is doubtful how far the testimonies of the various senses really corroborate one another, and how far they are not rather incommensurable and referred to the same 'thing' for reasons of practical convenience. As Dr. Schwarz (Das Wahrnehmungs problem) has asked, are after-images and overtones, which regularly accompany sights and sounds, to be esteemed unreal because we generally find it convenient to neglect them? And yet it is hard to say to what data of touch they correspond. Again, what can this criterion make of cases of hypersesthesia of one sense, or of an occasional activity of some special sensitiveness ? Are they to be rejected because they necessarily lie beyond confirmation by the other senses? As far as this criterion goes, there is nothing to prevent a real thing from contravening it, and an 'unreal' thing from conforming to it. Is "Pepper's ghost" unreal because it cannot be touched? Or is a hallucination affecting several senses to be esteemed real?
(2) The second criterion is no better than the first. So Mr. Ritchie 'smells a rat,' in the case of his hypothetical mouse (p. 267), and limits its value by stipulating that B, C, D, and E (who do not see it) should have good eyesight. But how is it to be established that A (who does see it) does not considerably surpass them in the delicacy of his senses? In this difficulty, Mr. Ritchie proposes to call in expert opinion in the shape of "a hungry cat." (What scorn he would pour on such an appeal to the lower animals if it were a question of establishing the objectivity of an apparition!) Very good. But how if the cat side with the minority? It is to be hoped that Mr. Ritchie will prefer science to democracy, and the authoritative judgment of Athanasius and the cat
