'We have, beside these, the palætiological sciences, which proceed mainly on the conception of Historical causation.'-Vol. i. p. 77.

It is impossible for us with our limited space, on which we have already made large demands, to follow our author in his discussion of all these subjects. Amidst much that is ingenious in argument, and frequently beautiful and impressive in the illustrations which his extensive knowledge of science supplies, amidst occasional vigor and felicity of style, we too often fail to discover clearness of thought in the more metaphysical parts, and have still more frequently to complain of awkwardness and obscurity of expression. We must not, however, pass by this portion of the work without saying that in the chapters on his favorite sciences of Statics and Dynamics, he has, in his well known desire to assimilate these sciences as far as possible to the purely mathematical, and to dispense as far as he can with the necessity of experiment, gone at least to the full extent of his tether. Sometimes, indeed, there is almost a whimsical struggle between his wish to assert absolute necessity and self-evidence of as many of his principles as he can, and his desire, as an experimental philosopher, to do justice to experiment. Thus, for example, while admitting that the first law of motion was 'discovered his'torically speaking, by means of experiment,' and after combating, as it appears to us, very acutely and successfully, the hypothesis of those philosophers who would contend that it may be established by pure reasoning, he seems almost inclined, if not to retract, yet to modify his admission, and half to regret, as it were, that he cannot claim the principle as self-evident. 'Thus,' he tells us, though the discovery of the first law of 'motion was made, historically speaking, by means of experiment, we have now attained a point of view in which we see 'that it might have been certainly known to be true indepen'dently of experience. This law in its ultimate form, when completely simplified and steadily contemplated, assumes the 'character of a self-evident truth.' Now we believe, in point of fact, that there never was a proposition (certain as it is) which carried less appearance of being self-evident with it, which nine-tenths of mankind would more promptly or summarily reject at a first hearing, or which has to struggle with greater prejudice before it is fully and finally admitted. When a proposition is not evident at all till it has been discovered by

That a body once set in motion, and not checked by any external force, would continue to move on for ever with an uniform velocity.

experiment, and is admitted to be true at last only by experiment, it is surely too late in the day to talk about its being self-evident; it may be evident enough, nay, be proved to be more reasonable than its opposite, but can hardly with propriety be called self-evident. Similar remarks apply to Professor Whewell's observations on the other laws of motion. Nor must we omit to mention here that not a few of our author's scientific axioms, which are not derived from experience, are merely logical deductions from our definition of objects, the true idea of which is gathered from experience; they are enveloped in the definition itself, and are but expansions of it. Thus the axiom that Fluids press equally in all directions,' and which he is very anxious to rescue from the suspicion of being derived from experience, he himself asserts, and justly, is 'involved in our idea of a fluid, which is considered as matter, and as matter which has its parts perfectly moveable among one another.' Now that the very existence of matter in general, and of such a modification of it in particular, are the result of experience, we suppose few will be disposed to deny.

But we must not pause longer upon this portion of our author's work. We proceed, therefore, to give some account of the application of the preceding reasonings to the Theory of 'Induction.'

Having, in Part I. of the work, and which occupies the whole of the first and a portion of the second volume, treated of Ideas,' our author proceeds in Part II., to treat of 'Know'ledge.' He shows that there are two processes involved in the construction of science, the 'explication of con'ceptions' (as modifications of fundamental ideas), and the 'colligation of facts;' that these conceptions gradually work clear by the controversies and discussions of scientific men; that for the purposes of science, they must be appropriate as well as clear, by which our author means that they must be modifications of that Fundamental Idea' by which the phenomena can be really interpreted; as for example, that a man must not apply to the interpretation of the phenomena of morals the fundamental idea of space or number, or any of the conceptions thereon dependent; that this maxim may warn us from error, though it may not lead to discovery, a proposition few will be inclined to dispute; that discovery results from the previous cultivation or natural clearness of the appropriate idea, and that, therefore, no discovery is the fruit of chance. After illustrating these and some related topics at considerable length, our author proceeds to observe, that science begins with common observation, to be succeeded by scientific observation and experiment; that the conceptions by which the 'Colligation of Facts' is effected, are

the result of the sagacity of discoverers; that the 'Colligation ' of Facts' cannot be taught; that it is generally effected by repeated conjecture, or as our author expresses it, by framing several tentative hypotheses and selecting the right one; that a series of these appropriate hypotheses cannot be constructed by rule or in the absence of inventive genius; that a great part of the genuine philosophic character consists in rigidly testing these hypotheses by facts, and rejecting them if they are found wanting, however ingenious and beautiful; and, lastly, that hypotheses may have their use, which are in some degree superfluous and not free from errors, seeing they may suggest the true harmonizing conception, and be purged both of their superfluities and their errors. He then proceeds to show that the name by which we designate the process of a true 'Colli'gation of Facts,' is Induction; he defines the Consilience of Inductions, to be when an Induction obtained by one class of facts coincides with an Induction obtained by another class of facts. The Logic of Induction consists in stating the Facts and the Inference in such a manner that the evidence of the Inference is obvious; as the Logic of Deduction, in stating the Premises and Conclusion so that the evidence of the Conclusion is obvious. On this point, the author closely follows the admirable view of Whately, in the remarks on Induction contained in his work on Logic. He has added, however, some observations on the analogies between the Logic of Induction as the criterion of Truth inferred from facts, and the Logic of Deduction as the criterion of Truth deduced from necessary principles. Into these we have no space to enter, and a few sentences could give no more than a very obscure exhibition of our author's views, which, indeed, are not always expressed by him with the perspicuity that could be wished. Nothing, however, which Professor Whewell has said, contravenes, of course, the great truth which Whately has so fully illustrated, that the logical process is everywhere and always essentially the same, the phrases Logic of Induction' and 'Logic of Deduction,' being employed to discriminate not any essential difference in the reasoning process itself, but in the sorts of propositions which are made to supply the premises and the conclusion.

Our author further goes on to show that Inductive Truths may be divided into two great classes; namely, Laws of Phenomena and Theories of Causes; that it is necessary every science to commence with the former, but that it is impossible that we should be satisfied to stop short of the latter. On the differences between Art and Science (which are illustrated with a somewhat undue copiousness) there is nothing at variance with the common views.

This book concludes with a short chapter on the classification of the sciences. Though it exhibits some marks of the haste with which the whole work has been composed, and though the expression is often far more obscure and diffuse than it might have been, yet we think it very superior to those portions which treat of Fundamental Ideas' and 'Ideal Conceptions.' The author seems to be more at home, and to have wrought his mind free from much of the cloudiness which envelopes the earlier portions of the first volume.

This book is followed by one entitled 'Review of Opinions on the Nature of Knowledge and the Methods of Seeking it.' It is purely historical, and is written with judgment. We have already presented our readers with a brief extract from it. We doubt, as may be gathered from what we have already said in a previous page, whether full justice has been done to Locke; that is, whether he is answerable for the errors which the 'sen'sational school,' as Mr. Whewell calls it (though we cannot say we much like this adjective any more than some others from the same mint), grafted upon his doctrine. At the same time, our author is pleased to say-and it is but candid to cite itthat 'Locke himself did not assert the exclusive authority of 'the senses in the extreme unmitigated manner in which some 'who have called themselves his disciples have done.' But then in the very next page he says, reclaiming with the one hand what he had given with the other,'-' We need not spend 'much time in pointing out the inconsistencies into which 'Locke fell; as all must fall into inconsistencies who recognize 6 no source of knowledge except the senses.' Locke has fallen into inconsistencies, but (as already shown) he has recognized another source of knowledge besides the senses.

The last book contains an account of the Methods employed ' in the Construction of Science,' but the topics discussed are both so numerous and so important, that it is impossible we can give an analysis of them. The slight notice which our remaining space would permit us to give would hardly be intelligible. We may mention, however, with particular approbation, the Introductory chapter; the second chapter, entitled 'Methods ' of Observation; chapter fifth, on the 'Analysis of the Process ' of Induction;' and chapter seventh, on 'Special Methods of 'Induction applicable to Quantity.'

It will be seen from our general remarks, that the commendations, with which we are disposed to receive this work, must be taken with many abatements. In particular, with regard to the more metaphysical parts of it (though there is not probably much with which, when duly explained, we should be disposed to quarrel), we often regret the want of a more thorough analysis, and cannot help fancying ourselves put off

VOL. IX.

2 x

with sonorous phrases and prolix repetitions instead of clear and definite thoughts. Still more frequently have we to complain of obscurity, diffuseness, and the employment of superfluous technicalities in statement, even where we do not quarrel with the matter of it. Indeed, throughout there is great heaviness and want of vigor and elasticity in the style, which, we once more repeat, might have been wholly or in great measure obviated by a more severe revision.

While we feel ourselves bound in honesty to make these strictures, we are equally bound in honesty to say, that there are many detached portions of the work which we think truly admirable, and with which we have been much delighted. Even many of the metaphysical portions are treated with great ability. We would particularly mention the strictures on Condorcet's theory that Ideas' are nothing but 'Transformed Sensa'tions'-the remarks on Nominalism and Realism-on Contingent and Necessary Truths, on the Acquired Perceptions of Sight-on the mode by which we obtain the Idea of Resistance; on all which subjects, though our author does not advance any other doctrines than those of the best metaphysicians of modern times -more especially Brown-he has ably sustained and illustrated them. On the last point, in particular, he has a very interesting passage, in which he shows the coincidence of the views of Brown with those of the great modern physiologist, Sir Charles Bell, both of whom reached the same conclusion by different roads. We are glad to see that our author opposes Brown on the question of Visible Figure,' though, while we agree with him in the conclusion (as we believe most recent speculators do), we are not sure that the commentator is not in one or two sentences almost as obscure as the original. His observations on the several 'Paradoxes of Vision,' as, for example, how we see a single object though there is a double picture on the retina -how we see objects upright, though the picture on the retina is inverted, are full of interest, though not to us wholly satisfactory. Brown's views on these two paradoxes always appeared to us most untenable and extravagant. His notion on the latter point is, that we do actually see objects inverted, and learn to interpret them as upright only by the associations derived from the experience of other senses, and that these associations are so strong that we now actually believe that we see objects upright though in reality we do not. Truly this is a hard saying, and few can receive it.

Professor Whewell's observations on the curious paradox that it is not the original prerogative of the eye to inform us of the distance of objects, or of magnitude in three dimensions (first established by Bishop Berkeley), and which never fails to fill the young metaphysician, when first propounded, with