metaphysical assumptions as explicit as possible. We cannot totally expel them; and surely it is better for them to be openly recognised than to be lurking unseen and so giving a subtle bias to 'scientific' results.

We have now dealt with the first of the three questions indicated above; the second and third do not present so much difficulty. It has already been shown how as knowledge develops two fundamental lines of divergence appear in it-the reference to Self and the reference to Reality other than Self. Both of these are psychological facts in the sense in which we have explained the term. The reference to Self is as much of a problem for Epistemology as the objective reference; there is no absolutely self-evidencing character belonging to the former that is absent from the latter. The view of Common-Sense is, that in each of those 'duration-blocks'-as Professor James would saywhich we speak of as the Present or Now, we have an 'immediate' or 'direct' apprehension or consciousness of ourselves as Subjects of Feeling and Will as well as of Knowledge. This so-called immediate apprehension—which seems very simple but in reality is extremely complex-we have to expand and explain, and justify or criticise. I have already had occasion to indicate that it appears to combine two elements. Those conscious functions which make the 'more than knowledge' of which we have spoken, are known about, and this knowledge is reflective-i.e., belongs entirely to the presentational side; but at the same time it seems to require a basis of immediate experience.1 When, however, we take into account the successive duration-blocks, it becomes apparent that the existence of the Self through these is altogether a matter 1 These problems we shall deal with in chap. v.

of intellectual construction. With regard to this, Epistemology does not, like the historical method in Psychology, rest content with mere succession, but endeavours to show after the manner of the Kantian Deduction' of the Unity of Apperception -how such experience of succession is possible. It is possible if there is in consciousness a principle which either is permanent and "identical with itself through Time" or has a mode of existence that in some way transcends Time. Both of these possibilities are included in the idea of a principle which is present in the same sense to every term of the

succession.

The other aspect of the central problem of Epistemology lies in the reference which we have called 'trans-subjective':1 this also needs explanation and vindication, in the manner we have pointed out. It must be carefully observed that we do not "start with a self-contained subject" or assume that "at first we know nothing else" than "subjective states." Consequently we do not seek to "leap from them. [subjective states] into something absolutely different”: we start with an individual who, ideally, has "transcended his own existence." We may safely say that there never was a time in the life of any individual when he recognised his cognitive states as being his own and yet did not just as spontaneously recognise and accept their reference to realities other than his finite self. It is in fact a psychological impossibility

1 This term is employed by Volkelt, Erfahrung und Denken: with it he contrasts the reference to Self as 'intra-subjective.' But he endeavours, in the Cartesian manner, to assign superior certainty and validity to the latter. This view we shall have to regard as highly questionable.

for a man to "know only his own states":1 the supposition that it can be so arises only from the vague and unscientific character of the terminology employed. Descartes, and many others after him, regarded Thought as a kind of prehensile organ with which mind is endowed. In this case it becomes possible, and apparently intelligible, to ask whether the objects of this Thought (in knowledge) are "states of the mind itself" or "external objects." Descartes assumes the former; hence immediately arises the question, What warrant, then, have we for belief in the conscious minds of our fellow-men, or in any other trans-subjective reality? From the nature of the case, no warrant can be found. Now the proposition that "all the objects of our knowledge are ideas in our own minds" is utterly destitute of meaning, and plausible only through a careless use of language; but apart from this, if we start with mental modes unreferred save to the Self only, we cannot establish the reasonableness of their reference at a later stage to reality other than Self. These statements could only be fully justified by examining in detail some actual presentation of the opposite theory, such as that of Volkelt. In the present Introductory Studies we cannot do more than show that the theory of knowledge here outlined is thoroughly consistent with itself.

We must observe that, implicitly involved in Berkeley's theory, there is an attempt to pass from knowledge of subjective states to objective knowledge, which differs from that of Descartes, and which has its attraction for some thinkers at the present day.

1 That is, unless the term is so general that 'knowing only one's own states' means 'not being able to get outside one's thought.' But this truism was not what Descartes had in view.

The fact on which stress is laid is merely negativethe individual's want of consciousness of productive power in relation to the presentations of the senses, while he has this consciousness of productive power in relation to his thought as such, and of controlling power in relation to his mental images. It is inferred that the presence of sensation postulates a reality or realities beyond the individual's consciousness, as the originating cause of his sensation: the implication of course is that the being, or beings, must be of nature similar to his own consciousness. We must point out that the analogy breaks down unless thought is a kind of sensation or sensation a kind of thought: otherwise, we cannot infer from subjective productive power in the one case to similar but trans-subjective power in the other. Apart from this, we cannot found. such large conclusions on a mere want of explicit consciousness of production in relation to sense: why may not the individual Subject be a Leibnitian monad, and its whole experience simply the evolution of its own nature? The argument can only be made valid by presupposing an intellectual grasp of trans-subjective reality; and it then becomes simply a means of determining the constitution of that reality.

§ 5. We have spoken of the 'intra-subjective' and 'trans-subjective' references as giving rise to the central problem of Epistemology: for the critical examination of these prepares the way for a systematic criticism of the constituent elements of knowledge.

The physical sciences assume the validity of the trans-subjective reference, and in effect they seek, by hypotheses concerning the nature of the objective system referred to, to reduce the indefiniteness of the

conception, "that which does not depend for its existence upon any individual's knowledge of it." I say ' in effect,' because no science ever explicitly presented its problem to itself in this way; nevertheless the work of the sciences presupposes the fundamental reference in the general form in which we have defined it. Now the special sciences, in dealing with extra-conscious existence as thus understood, make implicit assumptions concerning its nature; or, as we may otherwise express it, each of them in its own sphere assigns a special meaning to reality, and should limit itself to dealing with reality in that special sense.

(a) Physics and Mechanics limit themselves to events that can be reduced to mechanical action—vis a tergo impressa. In using this phrase as characteristic of mechanical action, I assume the point of view of English physicists, according to which the only 'action' between bodies which is intelligible is that which occurs when they are in contact,-being, therefore, of the nature of 'stress,' that is, tension or pressure. The only form of action with which Newton and Galileo had to deal was that of visible contact or of freely falling bodies; the notion of visible contactaction (vis a tergo) was the clearest that presented itself, and was taken as the type of all action. Newton rejected with contempt the notion of 'action at a distance' or attraction between portions of matter separated by any immaterial element. In this, English physicists have followed him; hence the attempts constantly made to explain the apparently attractive forces of Gravity, Electricity, and Magnetism, as peculiar cases of action by stress.1 Continental

1 This is well illustrated in Mr MacAlister's review of Stallo's Concepts of Modern Physics, Mind, vol. viii.