were both dependent upon a decrease of population. But if the population should decrease enough to lessen the pressure of competition, would not that result in a retardation of progress? No,' say some, 'for progress no longer depends on competition and the destruction of the incompetent, so much as upon education and conscious effort at improvement.' But that thought hardly reaches the bottom of the matter: for what makes people resort to education and self-improvement, what but the fear of competition? How many would be at the painsirksome and bitter it is to them-to educate their children, or themselves if they were not convinced that it is their only hope of success? Thus the apparent displacement of Natural Selection by direct adaptation really comes to no more than this, that the forces of Natural Selection have reflected themselves in almost everybody's mind. So far, then, as the increase of happiness depends upon the development of individuals, it depends also upon the maintenance of competition; so far as it depends upon the increase and diffusion of wealth and leisure, it requires the decrease of competition. From these data, we cannot expect happiness to increase as fast as the species develops; and a process of development must be slow which depends upon the impulse of forces (such as the love of Superiority) that at the same time retard it. So much longer must the world endure to enable the future to make amends for the past.

And even then how unfair it must seem; though the dead do not feel it, nor shall we when our turn comes to be as they are. How unjust of Nature that nothing but the joys of men unborn should recompense their fathers' sorrows! That yet unrealised happiness is something to us who foresee it-far off its coming shines; but what has it been to them who did not foresee it, but prepared its way-like hordes of slaves doing a work whose purpose is hidden from themdriven by despotic instincts, arbitrary passions, and every sort of uninterpreted illusion? For whence but from their accumulated afflictions could the feelings which we call our noblest have sprung? The bitterness that sweetens so much æsthetic ecstacy is the salt stain of innumerable tears. What hope and folly, what disappointment, what yearning and remorse must have commingled and distilled in human hearts before the first notes of Lohengrin could awaken there such an exquisite response! And they who prepared that cup never tasted it, but were only sickened with its crude ingredients. I believe the recollection of such things will sadden mankind for ever. Unless they can feel that the past also was for its own sake not in vain, they must dwell in the shadow of an inexpiable wrong. And when the ignorant admiration and hollow mimicry which now serve instead of reverence for the past, have been outgrown and abandoned, men will not forget the debt they owe it; but will exhort one another to bear it in mind, will appoint days of commemoration, will desire even passionately to have shared those sufferings, and will pay with fasting and sacrifice just homage to the dead.

In conclusion, I may observe that Mr Sully has not shown so fully as he might have done the importance of his subject, especially at the

present hour it is, of course, clear to himself, but he has neglected to impress it upon the reader. I suppose it is not too much to say that there are now in Europe more people than ever before who do not expect another life; to whom, therefore, for the sake of both themselves and their descendants, the worth of this life is all in all. Upon this supremely interesting subject few books exist that can pretend to be impartial; and the present work is, as far as I know, by much the best of them.

CARVETH READ.

Prof. Jevons on Mill's Experimental Methods.-Professor Jevons's review of Mill's theory of Induction (Contemporary Review, April, 1878) seems to me to omit one or two considerations which are indispensable to a full statement of the doctrine, and consequently to involve some misapprehension of Mill's meaning. The gist of the article is given in the following sentences:-"These methods (the Experimental Methods) are the only means of proving the connexion of cause and effect; yet the methods depend for their validity upon our assurance of the certainty and universality of that connexion" (p. 89). "The Experimental Methods are of no validity, until we have proved a most general, in fact an universal, law, which can only be proved by these methods" (p. 91). The first of these sentences everyone will recognise as familiar in the System of Logic; the second must come, I think, with a shock of surprise to most students of Mill, for they are very well aware that according to him this universal law is never exactly proved, and does not stand to the methods in the relation of proof to the thing proved. On the strength of the apparent contradiction, however, Prof. Jevons rejects Mill's theory of Induction as being inherently inconsistent.

Confining attention solely to the question of consistency, I have in the first place to point out that Prof. Jevons has not taken into due account the fact that according to Mill the belief in Universal Causation is a slow growth. It is an integral part of the theory that originally generalisations from experience were determined solely by psychological motives-by belief grounded on association; and that such belief did not involve universality of causation, but merely uniformity in the class of objects observed. Only after such generalisations had been made to a considerable extent, and had been verified by experience, could the principles on which they rested be applied more widely. Not until the belief in uniformity had attained the dimensions of an assumption that all natural phenomena were subject to law, could a philosophy of induction be constructed. The methods of scientific induction might very well be applied in particular cases, and with merely particular import, before the assumption of general uniformity was made. Difference, e.g., is the most familiar mode in which the more obvious and palpable connexions of cause and effect would be discovered, but it could not be generalised into a

method, applicable to all phenomena, unless the assumption were made that all phenomena were subject to law. This is Mill's persistent contention. He repeatedly points out that inductive generalisations not involving the universal law are essential preliminaries to any statement of inductive methods which involve that law. I would refer in support of this to Logic, Vol. I., 345 n., 355, Vol. II., 99 n., 101 n., 104-5 (7th ed.), passages which completely dispose of the argument on pp. 96-7 of Prof. Jevons's article.

It seems to me, in the second place, that Prof. Jevons, in supposing that we must have proved the Law of Causation before the Methods are valid, misapprehends Mill's theory. The peculiar relation between the universal law and the methods may, I think, be put in the following way. The methods are canons or rules of evidence, specialised statements of the signs of causal connexion. If our evidence exhibits certain signs, or satisfies the requirements of the methods, then we assume that causal connexion obtains among the phenomena, for this reason, that in such a case either causation is proved or the general law of causation is disproved. We do not say that causation is proved in this particular instance because causation is universally true, but we show that the evidence either warrants causation or disproves the universal law. In other words, our inductive reasoning exemplifies the special relation between the major premiss and conclusion of any reasoning.

It may be asked, why do we assume one alternative rather than the other? The answer to this will, I think, bring out a certain ambiguity in the word proof, which seems to have misled Prof. Jevons. The only reason is that the evidence for universal causation is incommensurably greater than the evidence against it. But the only evidence for an ultimate law of experience is conformity with fact; and to say that the evidence for universal causation is exceedingly great-so great as to be practically conclusive-is merely to say that mankind have so steadily found their inductive assumptions verified by experience that, in any instance where law is not at once apparent, the hypothesis of absence of law is not even momentarily admitted. Proof of all subordinate laws is given by comparison of the evidence in favour of them with the universal law, while the establishment of such laws lends additional strength to the belief in general conformity to rule. It is evident, then, that to Mill proof of the law of causation can never be in one sense absolute, for we have not exhausted the universe of facts (see close of Bk. III., ch. xxi.), but that the certainty with which it is held grows with experience, and has become so strong as to be equivalent in its effects to the certainty of a demonstrated doctrine.

I cannot think that Prof. Jevons has given due weight to this relation between the Universal Law and the Methods. Neither the note to p. 94 of his article, nor his reply (Academy, 4th May, 1878) to a critic who had correctly but in an objectionable manner called attention to the point, can be regarded as dealing satisfactorily with a question which is fundamental. Much of what Prof. Jevons rather rashly throws out with regard to the possible growth of the theory of

induction in Mill's mind might have been spared had he fairly weighed such a passage as the following:-" Neither would it be correct to say that every induction by which we infer any truth, implies the general fact of uniformity as foreknown, even in reference to the kind of phenomena concerned. It implies, either that this general fact is already known, or that we may now know it: as the conclusion, The Duke of Wellington is mortal, drawn from the instances A, B, and C, implies either that we have already concluded all men to be mortal, or that we are now entitled to do so from the same evidence. A vast amount of confusion and paralogism respecting the grounds of induction would be dispelled by keeping in view these simple considerations" (Vol. I. 345 n.).

It would require an article fully as long as that of Prof. Jevons, if one were to follow him into the minor points raised. But I should like to say that the absurdity detected by him in the passage quoted from Mill (p. 90, C.R.) seems due to some rather arbitrary interpretation of the words 'general law'; that to base scientific induction on the unscientific is exactly the same process which has produced the doctrine of Probabilities,-both are but good sense reduced to rule; and that, despite the awful fate predicted in his last sentences for all who base induction on causation, I should maintain not only that every inductive generalisation involves the assumption of Uniformity, but that the Inverse Doctrine of Probabilities is in precisely the same

case.

ROBERT ADAMSON.

Necessary Connexion and Inductive Reasoning.—Were the question asked-What it is in the attitude of logicians that seems to me to make the inquiries advanced in this paper indispensable to the progress of logical theory? I should be disposed to return the following answer: The weightier matters of the law are not receiving sufficient attention. While the theory of Evolution, that era-creating discovery of the present century, is being so largely verified in physical science and even in that department of physiology embraced by psychological inquiry, in Logic, the Scientia Scientiarum though it has been called, it has not yet been successfully shown to be the law regulating all intellectual processes. It is true, the à posteriori school of logicians, guided by this luminous principle, has met with considerable success in prosecuting its inquiries; nevertheless, the opposite school, I venture to assert, still retains hold of enough of the truth to justify its position. The Lualaba of so-called transcendental truth has not, as yet, been identified with the Congo of generalisation from experience. This identification, I need not say, would, at any time, be a consummation devoutly to be wished, but more especially so when, as now, a crisis is impending, and when the extravagant procedure of certain imaginative votaries of science has called forth even from so advanced and fearless an inquirer as Prof. Virchow a warning to keep within the fortified lines of objective truth; and when, therefore, the

guiding light of the true Scientia Scientiarum seems to be so much needed. What leaves to logicians of the à priori school a cause still to uphold is, I believe, the fact that such explanations as their opponents have been able to give of inductive knowing fail to satisfy the implicit convictions of the mind. While, for ages, the domain of reasoning has been largely explored in connexion with deduction and general truths, it is only in modern times that induction and the concrete or the individual have had much attention yielded to them. The differentiating processes, the working-classes of the human intellect, which, in the order of evolution, seem to be prior to the generalising operations, still await their full explication. This incompleteness in the fundamental truths of Logic produces obscuring effects upon the whole science, and causes logicians to be divided in opinion. This, at a period when the civilised world threatens to separate into two hostile camps-authority versus free inquiry, is by all real lovers of truth to be deplored.

Truths are usually divided into necessary and contingent—which are here called related terms, contingent being regarded as equivalent to non-necessary, and non-necessary to contingent. Some, however, contend that there is no sufficient reason for dividing truths into necessary and contingent. Any truth, regarded as such, is, they hold, necessarily true. To say that a truth is contingently true is to imply that it is open to doubt. This, however, is not what is meant by a contingent truth. Contingency as applied to truth is not usually understood as a synonym for probability, because many a contingent truth is true beyond all question, is, indeed, necessarily true. For instance, it is as true that a £5 note remains in my purse as long as I can manage to keep it there, as it is true that a whole is greater than its part. What, therefore, is the precise meaning to be attached to the terms necessary and contingent as applied to propositions? By the former, I understand a necessary connexion between one thing and another; by the latter, a contingent connexion. The question here to be discussed, then, is, How is necessary connexion perceived ?

Two facts being observed as merely joined together, we have but an indefinite notion, perhaps, of the nature of the union that subsists between them, unless it happens to be previously known to us. This prior knowledge forms the mental nexus by means of which we determine the nature of such union, as regards necessity or contingency. How is this nexus obtained?*

This application of a mental nexus is deduction; the simplest form of which, it appears, is reading out, as we have occasion, what a universal or a general proposition declares as to each case to which it is judged to apply. It rains, some one informs me. Rain constitutes one of the terms of the proposition (implicitly contained in the mind for the most part) "All ram wets," therefore, I conclude, this rain wets, and I never think of going out to ascertain the fact- I feel certain of it.

It is commonly held that deduction involves syllogising. This I fail to perceive. When a chain like Judæa, Samaria, Galilee is given, I perceive that Judæa is mediately joined to Galilee; when I do this I syllogise, but