reader. But what is most wonderful, when we consider the actual words, is that Kant's own eventual pretensions are to "a proof," "à priori," and "from ideas"-this, at the very moment that he asserts any such achievement to have been incontrovertibly "proved " by Hume to be impossible! So far at least as words go, surely Kant defeats his own self here by a direct anticipatory suicide. It might not be difficult to explain this; but it is not worth stopping for. What we have to see at present is that Kant signalised the necessity. "The question with Hume," he says, "was only of the source of this idea, not of the indispensableness of it in use;" and he very justly censures Reid and the rest for their quite inexcusable misapprehensions.

As concerns the second question, a quotation or two will also suffice.

"Hume derived the idea of necessity from experience, namely, from custom, or a subjective necessity due to repeated association in experience, and so at last erroneously taken for objective (WW. ii. 728). What is the secret x here, on which the understanding supports itself when it believes itself to discover outside of A, and alien to A, a predicate B, that is nevertheless united to it? Experience it cannot be, for the enunciation concerned unites the one thing to the other, not only with greater generality than experience can provide, but even with the expression of necessity (ib. 23). In the proposition that every change must have a cause, the very idea of a cause so manifestly implies the idea of necessity in the union with an effect, and of a rigorous universality of rule, that it would be quite lost, did we derive it, as Hume has done, from a repeated association of what happens with what precedes, and a consequent habit (mere subjective necessity) of uniting ideas (ib. 698). How is it possible, said the discriminating Hume, that, if an idea be given me, I can go beyond or outside of it, and join with it another idea, not at all contained in it-conjoin them, in fact, so as though the one belonged of necessity to the other. Only experience can supply such unions, and all that supposed necessity is nothing but a long custom of finding something true, and so of taking a subjective necessity to be an objective one (iii. 30, ff.).”

Reflection on these passages cannot fail to prove to the reader that Kant, while perfectly recognising and understanding both the question and the answer of Hume, quite as perfectly recognised and understood the inadequacy and incompetency of the latter to the former. If I ask for an insight, an intellectual perception, into the reason of a truth, universal and common to us all, it is in vain to refer me to a mere feeling, an instinctive feeling, of my own, that has only grown up in me in consequence of my just being in the habit of, from day to day, seeing such things.

When this dropping of a particle of ink occasions a stain on the paper, I am sure that there is a reason for it that does not lie in me, but in the things themselves. You

cannot stave me off by saying the reason you want is just that you have seen the same thing before, and you have got into the habit of expecting it; the supposed necessity is but a feeling of yours; it is not an element in the facts; neither, consequently, does it admit of nor does it call forbeing intellectually seen into. To this, of course, one can only shake one's head. I feel, I know that it is not custom, it is not a habit in me, but a truth in the facts themselves that necessarily connects them, the one with the other. The union between the word cause and the meaning it suggests, is now, by habit, custom, a spontaneous, instantaneous, fixed and necessary one in my mind; but still the word and the meaning, thus inseparably united by habit, are not together the one as cause and the other as effect: they are together, after all, only by arbitrary association; and, really, any other combination of letters might have been assumed to represent the same idea. Cause, cause, to almost every Chinaman or Turk-cause, cause, to Plato and Aristotle, were simply gibberish. Here, then, we have a perfect example of the necessity of habit; but it shows at once as essentially different from the connexion of ink-drop and paper-stain. The necessity of union, conjunction, or connexion in this latter case, let its source be what it may, let it depend on what it may, let its reason, its rationale be what it may, lies manifestly, obviously, evidently-self-evidently-in the facts. themselves. It is not in me, it is not a feeling of mine; it is a thing that, as having a reason of its own, I want to see into.

Well, then, it is to the explanation of this that Kant, after Hume's failure, and in full view of it, now fairly, faithfully and confidently addresses himself. "Here is now the place, he intimates (iii. 73), "to raze Hume's doubt to the very foundation [Hier ist nun der Ort den Hume'schen Zweifel aus dem Grunde zu heben]." But before proceeding to this, intelligence for ourselves, as well as justice for Kant, demands that we should provide ourselves with the general idea of Kant in his main critical action. From the Introduction to the Prolegomena I quote as follows :—

"I shall take it upon me to say that the unprejudiced reader will not merely have doubts in regard to metaphysic as yet, but will, in the end, be completely convinced that any such science is completely impossible unless with satisfaction of the requirements here made. . . . No event has taken place which might have been more decisive as regards the destiny of this science than the foray of David Hume. He threw no light, but he struck a spark, which, &c. . . . This hint of David Hume's was what, years ago, first broke my dogmatic slumber. . . . I tried first, then, whether Hume's observation might not be made general, and soon found that the idea of the

1

connexion of cause and effect was not by any means the only one through which the understanding à priori thinks to itself connexions among things, but rather that metaphysic simply consists of such. I endeavoured to assure myself of their number, applying for that purpose a single principle; and this succeeding to my wish, I set about their deduction. . . . Thus it was my fortune to succeed in the solution of the Humian problem, not merely in a particular case, but in respect of the entire faculty of pure reason. I was able to advance slowly but surely towards an ultimate determination of the total sphere of pure reason, as well in its limits as in its contained matter: which was precisely what was wanted to secure finally, for metaphysic, on a safe and certain plan, its own entire system."

What I want to show by this extract is that Kant's whole work (and what alone led to all the others, Fichte, Schelling, Hegel) rose out of one consideration only. What was whence was that very strange and peculiar species of necessity to which Hume had drawn attention in the phenomena of cause and effect? That was the one spore, as it were the bean on the stalk of which, up there in the clouds, there rests the palace of more than one giant-perhaps in dream. In a word, to Kant metaphysic itself, to us the Kritik of Pure Reason, nay, German philosophy as a whole, has absolute foundation in the whence or why of necessary connexion. Such necessary connexion exhibited itself, in the course of the reflections of Kant, not as confined to causality alone, but as common (and, at the same time, peculiar) to all the propositions that collectively constituted what science there was of metaphysic proper. It was therefore a general problem that was concerned, and no special one. The discovery of its solution would realise at last what had been so long the dream of philosophy!

What was this one quality, then, that was at once common and peculiar to all these propositions?

A proposition is a judgment, and a judgment is a declaration of one thing of another. Now, evidently, the important point is the reason of the declaration. Why do I declare A of B-why do I declare the predicate of the subject? All bodies are extended; some bodies are heavy; the three angles in all triangles are equal to two right angles. Is the reason of the declaration-the fulcrum on which it rests in all these three propositions the same? No. Extension is implied in bodies: I cannot think a body without thinking it as extended. I can, however, think a body as light; all bodies are not heavy. Heaviness, then, is not, as extension is, implied in bodies. It is plain, therefore, that the fulcrum on which the declaration rests, is, in these two propositions, a different one. Now, to say it at once, this difference is that of analysis and synthesis. If I declare

extension of body, it is because I see by analysis that the notion or idea body already, or just of itself, contains the sub-notion or sub-idea extension. But this is not my reason for declaring heaviness of bodies. I can declare, of anything whatever, all that it means. I can declare transparency of windows, opacity of blinds, seclusion of islands, continuity of continents, necks of peninsulas, outlet and inlet of doors, ventilation of chimneys, &c., &c. To say windows are not transparent, blinds not opaque, islands not secluded, &c., &c., would be to deny of these objects their very meaningwould be self-contradictory and therefore false. What is implied-consequently, a sub-notion-is true of its notion; otherwise there would be a self-contradiction. That, then, is the rationale of the analytic judgment. Some bodies are heavy—this is a proposition that, not being analytic, must be synthetic. That is, it is by reason of a synthesis that I declare of certain bodies heaviness. Where do I find the synthesis? In experience. I have balanced certain bodies in scales or in my hand, and found them heavy. I do not find this out by analysis; I find it out by synthesis. I apply to experience, and it extends to me the synthesis. Is it true, now, that we must affirm the idea due to the synthesis under the same test as that under which we must affirm the idea due to the analysis-that it would be a self-contradiction, namely, to affirm otherwise? By no means. It is a contradiction to say, bodies are not extended; but it is no contradiction to say, all bodies are not heavy. The analytic proposition, then, is, on peril of its life (contradiction) true -necessarily true. But contradiction is not the life of the synthetic proposition. To one idea experience does add another, we find. But our only reason for the declaration is the fact of the finding. There would have been no contradiction had such and such not been found to be fact. Experience synthetically adds brittleness to glass; but it would be no contradiction were glass, as some glass might be, found infrangible. Experience adds softness to clay; but it would be no contradiction if it were found to be hard. Now these synthetic propositions of experience are called matters of fact their life is simply the finding; a fact is just found to be the fact it is. The evidence of matters of fact, consequently, is à posteriori, while that of an analytic proposition is à priori (if only relatively so). I recognise bodies to be extended―à priori. That is, remaining by the subject, the idea, and not travelling beyond it-beforehand, as it were, before experience and in independence of it-I discover the truth. I recognise certain bodies, again, to be soft-à pos

teriori. That is, turning to experience, I discover the truth behindhand, as it were; that is, after experience, and in dependence on experience. Now these two species of evidence, the à priori and the à posteriori differ in their validity: the former (even though relative) is apodictic (necessary and universal), while the latter is only contingent (probable). That depends on the test of contradiction: bodies not extended, islands not secluded, chimneys not ventilating, would be contradictions. But bodies not heavy, clay not soft, glass not weak, would be none.

As for the third proposition that relates to the angles, we have already, in our Part I., seen all that is necessary to be said in its regard. Its life, and the life of all such, is neither analysis nor experience. Its life is Relation of Ideas. It is synthetic, yet not à posteriori; necessary, yet not analytic. An à priori synthetic, then! How is it possible? By discovery of the relation of ideas present in it. For its part, again, the discovery is effected either by intuition or demonstration. The truth of the straight line being the shortest is seen by intuition, intellectual perception of the relation of ideas exhibited. The truth of the proposition in regard to the angles is perceived through demonstration. But demonstration always leads back to intuition. Mathematical à priori synthetics are thus possible through intuition of the relations of ideas.

We have now to apply these principles to the propositions in question, examples of which are these. Every change must have a cause; The matter of the universe can neither be lessened nor increased; Action and re-action are always equal. In the first place, they are, evidently matter of fact. I must have experienced causes and effects, matter, action, &c., to know anything at all about them. Without experience cause would be a blank, effect a blank. In the second place, they are synthetic. The effect is not found by analysis in the cause; it is added to it by experience. In the third place, they are of an à priori validity. The validity in cause and effect is not just found. How do we know the sea in the Straits of Dover to be shallow, or the water of the Baltic to be brackish, or Arthur's Seat to have a bare rock on the top of it? Just because we find each statement to be a fact. But it would never come into our heads to say that Arthur's Seat must have a bare rock on the top of it, or that in the Straits of Dover the sea must be shallow. Of the three propositions above, however, we use must; and yet they are all three of them matters of fact, as much matters of fact as the three contingent propositions which we have