experiential inference from that one Ego, I ascribe the same reality to other Egoes, or Minds.

Having thus, as I hope, more clearly defined my position in regard to the reality of the Ego, considered as a question of Ontology, I return to my first starting point, the Relativity of human knowledge, and affirm (being here in entire accordance with Sir W. Hamilton) that whatever be the nature of the real existence we are compelled to acknowledge in Mind, the Mind is only known to itself phenomenally, as the series of its feelings or consciousnesses. We are forced to apprehend every part of the series as linked with the other parts by something in common, which is not the feelings themselves, any more than the succession of the feelings is the feelings themselves: and as that which is the same in the first as in the second, in the second as in the third, in the third as in the fourth, and so on, must be the same in the first and in the fiftieth, this common element is a permanent element. But beyond this, we can affirm nothing of it except the states of consciousness themselves. The feelings or consciousnesses which belong or have belonged to it, and its possibilities of having more, are the only facts there are to be asserted of Self-the only positive attributes, except permanence, which we can ascribe to it. In consequence of this, I occasionally use the words "mind" and "thread of consciousness" interchangeably, and treat Mind as existing, and Mind as known to itself, as convertible: but this is only for brevity, and the explanations which I have now given must always be taken as implied.*

Dr. M'Cosh has renewed his attack upon the doctrine of Permanent Possibilities. But I cannot find in his later remarks, so far as they are to the purpose, much more than a repetition of his earlier. On some minor points he does present some novelties. He is severe upon me for hesitating to decide whether the attribute of succession as between our sensations is given in the sensations themselves, or annexed to them by a law of the mind. The first supposition he characterises as a mere verbal generalisation like those which I have laid to the charge of Condillac ; forgetting the opinion held by some acute metaphysicians, and which is no mere verbal generalisation, that to have sensations in succession is only the same thing as having more sensations than one. The other supposition, that the attribute of succession is annexed to our sensations by a law of the mind, he says is giving to the mind the "power of generating "in the course of its exercise a totally new idea," an opinion, he says, utterly inconsistent with my "empirical theory;" he does not say with what theory. In any scheme of human knowledge that I am able to form, the resemblances and the successions and coexistences of our sensations are real facts, and objects of direct apprehension. Whether we are said to apprehend them by our senses or by our minds (which is the real meaning of the alternative I have left open) affects no theory of mine, and is to me a matter of indifference.

The most curious part of Dr. M'Cosh's reply is that he thinks, accord

ing to my "theory" there is no difference between sensations and thoughts. According to him, if I am right, the facts of external nature being only possibilities of sensation, ought to succeed one another according to "mental laws, say the laws of association." The reader will scarcely believe that I am not misrepresenting Dr. M'Cosh; but I refer him to the article, pp. 345 and 346.

Dr. M'Cosh still maintains that the action of bodies on one another cannot be accounted for on the hypothesis of Immateriality, takes credit for having, on this point, detected me in an oversight, and seems to consider the answer I was obliged" to give him as an afterthought of my own. This only proves that Dr. M'Cosh has forgotten, if he ever knew, the very elements of the Berkeleian controversy. Whoever knows anything of that, has got far beyond the stage of thought at which Dr. M'Cosh remains. Berkeley would indeed have been easily answered if his doctrine could give no account of the greater part of all the phenomena of physical nature.

CHAPTER XIII.

THE PSYCHOLOGICAL THEORY OF THE PRIMARY QUALITIES

*

OF MATTER.

FOR the reasons which have been set forth, I conceive Sir W. Hamilton to be wrong in his statement that a Self and a Not-self are immediately apprehended in our primitive consciousness. We have, in all probability, no notion of Not-self, until after considerable experience of the recurrence of sensations according to fixed laws, and in groups. Nor is it credible that the first sensation. which we experience awakens in us any notion of an Ego or Self. To refer it to an Ego is to consider it as part of a series of states of consciousness, some portion of which is already past. The identification of a present state with a remembered state cognised as past, is what, to my thinking, constitutes the cognition that it is I who feel it. "I" means he who saw, touched, or felt something yesterday or the day before. No single sensation can suggest personal identity: this requires a series of sensations, thought of as forming a line of succession, and summed up in thought into a Unity.

But (however this may be) throughout the whole of our sensitive life except its first beginnings, we unquestionably refer our sensations to a me and a not-me. As soon as I have formed, on the one hand, the notion of Permanent

* In the first edition I said: "But without the notion of not-self, we "cannot have that of self, which is contrasted with it." In saying this I overlooked the fact, that my own sensations and other feelings, as distinguished from what I call Myself, are a sufficient Not-self to make the Self apprehensible. The contrast necessary to all cognition is sufficiently provided for by the antithesis between the Ego and particular modifications of the Ego.

Possibilities of Sensation, and on the other, of that continued series of feelings which I call my life, both these notions are, by an irresistible association, recalled by every sensation I have. They represent two things, with both of which the sensation of the moment, be it what it may, stands in relation, and I cannot be conscious of the sensation without being conscious of it as related to these two things. They have accordingly received relative namės, expressive of the double relation in question. The thread of consciousness which I apprehend the sensation as a part of, is the subject of the sensation. The group of Permanent Possibilities of Sensation to which I refer it, and which is partially realised and actualised in it, is the object of the sensation. The sensation itself ought to have a correlative name; or rather, ought to have two such names, one denoting the sensation as opposed to its Subject, the other denoting it as opposed to its Object. But it is a remarkable fact, that this necessity has not been felt, and that the need of a correlative name to every relative one has been considered to be satisfied by the terms Object and Subject themselves; the object and the subject not being attended to in the relation which they respectively bear to the sensation, but being regarded as directly correlated with one another. It is true that they are related to one another, but only through the sensation: their relation to each other consists in the peculiar and different relation in which they severally stand to the sensation. We have no conception of either Subject or Object, either Mind or Matter, except as something to which we refer our sensations, and whatever other feelings we are conscious of. The very existence of them both, so far as cognisable by us, consists only in the relation they respectively bear to our states of feeling. Their relation to each other is only the relation between those two relations. The immediate correlatives are not the pair, Object, Subject, but the two pairs, Object, Sensation objectively considered; Subject, Sensation subjectively considered. The reason why this is over

looked, might easily be shown, and would furnish a good illustration of that important part of the Laws of Association which may be termed the Laws of Obliviscence.

I have next to speak of a psychological fact, also a consequence of the Laws of Association, and without a full appreciation of which the idea of Matter can only be understood in its original groundwork, but not in the superstructure which the laws of our actual experience have raised upon it. There are certain of our sensations which we are accustomed principally to consider subjectively, and others which we are principally accustomed to consider objectively. In the case of the first, the relation in which we most frequently, most habitually, and therefore most easily consider them, is their relation to the series of feelings of which they form a part, and which, consolidated by thought into a single conception, is termed the Subject. In the case of the second, the relation in which we by preference contemplate them is their relation to some group, or some kind of group, of Permanent Possibilities of Sensation, the present existence of which is certified to us by the sensation we are at the moment feeling-and which is termed the Object. The difference between these two classes of our sensations, answers to the distinction made by the majority of philosophers between the Primary and the Secondary Qualities of Matter.

We can, of course, think of all or any of our sensations in relation to their Objects, that is, to the permanent groups of possibilities of sensation to which we mentally refer them. This is the main distinction between our sensations, and what we regard as our purely mental feelings. These we do not refer to any groups of Permanent Possibilities; and in regard to them the distinction of Subject and Object is merely nominal. These feelings have no Objects, except by metaphor. There is nothing but the feeling and its Subject. Metaphysicians are obliged to call the feeling itself the object. Our sensations, on the contrary, have all of