of induction, what has here been advanced; but before this is done, it is perhaps desirable to notice that Liebig mentions "a particular case in which iron, in a damp atmosphere, may get rusty even in the absence of carbonic acid; that is, when the air contains ammonia, but then the rusting ceases as soon as all the ammonia has been absorbed."

First, then, as to the composition of common rust, the chemist discovers that it contains iron, oxygen, and water, that without these it does not exist, and that therefore it is necessarily composed of these. Secondly, as to the causes of rust, it is found that atmospheric air and watery vapour in contact with iron are followed by rust, that without these in contact with iron no rust follows, and that therefore these in contact with iron must be necessary to the formation of rust.

In these two instances the reasoning is inductive. From particular premisses, necessary conjunction is inferred, which amounts to the same thing as universal conjunction.

The substances concerned in the production of rust are, however, compound ones. For instance, atmospheric air contains nitrogen, oxygen, and a small quantity of carbonic acid; yet rust has no nitrogen and carbonic acid in its composition; it is therefore inferred that very probably they are not essential to its production.

Here the reasoning is deductive, but from a hypothetical datum, namely, that what is not discovered in the composition of a body is not demanded to generate the same; that therefore nitrogen and carbonic acid are not necessary to the formation of rust. This datum being, however, hypothetical merely, the conclusion which is drawn has to be verified by induction. Here take note, that deductive reasoning is conclusive only when the data which it employs have been vigorously established by inductive argument. But now for the next step.

It has been demonstrated that air and watery vapour in contact with iron are necessarily followed by rust. But it is presumed that two of the component parts of air, namely, nitrogen and carbonic acid, are unessential concomitants. An experiment is therefore tried which involves the elimination of these, namely, oxygen alone with watery vapour are brought into connection with iron, but the result is that no rust is formed.

Here the hypothesis-that of the elements composing air, oxygen alone was necessary to create rust, has been falsified by induction; and the conclusion is, that either nitrogen, or carbonic acid, or both, are a sine qua non in its production. The next step is to discover, by having recourse once more to induction, which of these three alternatives is true.

It can be proved inductively that nitrogen is an unessential concomitant, and therefore that carbonic acid is a sine qua non. For when it is known that oxygen is indispensable, but is not sufficient alone to produce the effect, and also that nitrogen is not essential, then we are bound to conclude that carbonic acid is essential. This is induction according to the Method of Residues. But we may also ascertain by direct inductive reasoning that carbonic acid is necessary, and then, by the same means, that nitrogen is not.

"But," continues Liebig, "rust does not contain any carbonic acid; and so the question arises, What share the acid has in that process? Another known fact is now sufficient to complete the explanation; viz., the properties of the carbonate of protoxide of iron. In a damp atmosphere this carbonate absorbs oxygen, and the protoxide is converted into the higher oxide, which does not combine with carbonic acid. During the conversion of the metal into rust at first, the lower oxide is generated, binding carbonic acid; but the latter is freed again when the protoxide becomes sesquioxide, and so the carbonic acid can resume its original action on the remaining metal for the second and hundredth time, until gradually the whole piece is thoroughly converted into rust."

Here the reasoning is deductive. The chemist is able to explain the absence of carbonic acid from rust, by knowing that this acid is concerned in generating carbonate of protoxide of iron, which in a damp atmosphere absorbs oxygen, and liberates the carbonic acid, thus forming rust.

A fitting conclusion to this inquiry will be, to make a few observations on the position which Mr. Mill holds in relation to the philosopher whose tenets he has the credit-with men of his own school-of having so cleverly and effectually demolished in his late work. There can be little room for doubt that Sir William Hamilton laid himself open to refutation in more instances than one. His Law of the Conditioned, based upon the Antinomies of Kant, admitted of being overthrown with no very tremendous effort by a thinker of Mr. Mill's calibre; and his theory of reasoning failed to converge to a focus all the light which was already gained on the subject; it was behind, rather than in advance, of what was already discovered in this field of thought. Hamilton lived more in the world of books than in that of living men. Great in learning, he was not equally great as a thinker, and an analyst of the mental world. But he had, nevertheless, what we hold to be, strong points. He had undoubting faith in the spontaneous and universal convictions of the human race; and his merit as a philosopher mainly consists in the strong, able, and lucid manner in which he vindicated, both as a scholar and as a thinker, the trustworthiness of these

convictions. We do not maintain that Common Sense (voûs), -that is, Reason operating we know not how, the completed result only being revealed to us; this being held by many to be primary, and therefore inexplicable, and self-evident—we do not maintain that Common Sense, thus understood, is the final appeal in philosophy; but we fully believe that it ought to be the point from which we start. Till a better is discovered, no more philosophic course can be followed than to begin with the speculative presumption that our spontaneous and universal convictions are to be trusted. We make use advisedly of the words, speculative presumption. Practical convictions we shall have whether, speculatively, we place confidence in them or not. Let what may then, in the second place, be thought of these convictions, that is, after a reflective examination of them, they must, in the first place, and always practically,-by sensationalist, idealist, transcendalist, and realist alike, be attended with implicit reliance in their truthfulness. Now the obvious impression derivable from this fact is, that these deliverances of consciousness cannot be mendacious. Is it indeed credible that the God of Truth, unless the humiliating fact were proved, would keep us all our days in bondage to a lie?

To renounce Common Sense entirely as a guide is quite as unphilosophic a procedure as the opposite course, namely, making it the court of final appeal in philosophy. Reid, and perhaps Hamilton, estimated its claims at too high a value; but if they did this, it was by that law of mind which causes a recoil from the extravagancies of sensationalists, idealists, and sceptics. Now, if we take a judicial view of the question in dispute, we must bear in mind that before knowledge of any subject can reach a demonstrative, scientific, or final stage, it must previously pass through at least two stages,-a spontaneous or birth stage, and a transitional or growth stage; and that many truths of simple character, such as the axioms and definitions of geometry, and of a fundamental and obtrusive character, such as the independent existence of the external world, come full grown to man in their spontaneous stage; full grown, we mean, as practical convictions. To the reflective inquirer, they may or may not, according to one's peculiar turn of mind, appear to be exactly such as Common Sense declares them to be. This variety of opinion exists, because the evidence demanded by reason in its reflective or philosophic research is not complete. When this evidence becomes conclusive, the mental peculiarities of philosophers will be controlled, and unanimity will prevail. Now to notice some of these peculiarities, the demonstrative inquirer, whether the philosopher with à posteriori bias, like Mr. Mill;

NEW SERIES.-VOL. IX., NO. XVII.

D

or with à priori bias, like Ferrier, too frequently seeks to limit knowledge to what is within his own clear but yet unextensive view. Thousand to one, says Lessing, the goal of your philosophy will be the spot where you become weary of thinking any further. He who insists too much for proof in all things seems to overlook the fact, that before knowledge can come to him in an acceptable shape, it must pass through certain preliminary processes. He is the finisher of other men's work, yet he would. have the iron ready to work into rails, or into Armstrong guns, before he admits its existence, and is loth to acknowledge the necessity of those preparatory processes by which it is divided from the ore. Although we must, whenever it is practicable, have recourse to this stringent method, yet it is possible, by too bigoted an adherence to it, to have our mental horizon sadly circumscribed, to have a cold shade of negation chilling spontaneous thought, and frowning down the uprising of a wise and pleasant faith. The part of true wisdom is to accept, not merely what is demonstrated, but to open wide the portals of the mind to those spontaneous and transitional thoughts and theories which invariably precede a final or scientific stage of knowledge. A grand and powerful inducement for cultivating this temper of mind is, that the superior truths are those which enter last upon their final growth, and are consequently, at this present moment, demanding from us the exercise of a large-hearted faith.

Now we humbly opine that every one who addresses himself to the lofty task of discovering philosophic truth, ought to start with a reverent trust that the faculties which God has given him are not leading him astray, even in their spontaneous deliverances, and more especially if these are the universal assurances of mankind; and that it is certainly more in accordance with the spirit of true philosophy, to seek to establish these deliverances than to overthrow them. In the inquiry which we have made into the nature of necessary conjunction in this essay, we started with the complete adoption of this principle, and we trust that the result is not of such a nature as to lead any candid and unprejudiced judge to remark, that we started with a wrong method, and that "philosophy," to quote the words of Ferrier, "exists only to correct the inadvertencies of man's ordinary thinking;" that "philosophy assumes, and must assume, that man does not naturally think aright, but must be taught to do so; that truth does not come to him spontaneously, but must be brought to him by his own exertions." We may here mention, by the way, that hard as Ferrier hits at Common Sense,

Institutes. Introduction.

his view of necessary truth exactly ratifies the declaration of our spontaneous intelligence as to the existence of such truth. But because Common Sense, forsooth, declares the independent existence of the external world, and according to his merciless logic there is no such contradiction, O then! Common Sense is not to be trusted! It is quite true that the world, as known, cannot exist out of synthesis with the knower. The known world is (subject plus object). Now this synthesis cannot be what it is, and at the same time be less than what it is. You cannot, at one and the same time, predicate contradictory attributes of one and the same subject. This is Ferrier's position; and, so far as it extends, it is unassailable. But there seems to be a sphere which does not come within its extension, namely, the external world before it becomes known.

Reason infers, and cannot avoid doing so, that in the order of nature, the cognition of an external object presupposes the prior existence of that object, that is, cannot avoid concluding that the object must have existed before it could possibly become known. A man looks at St. Paul's Cathedral. Did it exist before he looked at it, and does it exist after he has ceased to do so? Yes, from his perception of it, he is forced to conclude that it must have existed before he cast his eyes upon it, and consequently that it must exist when he has taken his eyes from it. This, we believe, is the root of the assurance, that the external world exists independently of the knower; and if this be correct, Common Sense is veracious again.

Mr. Mill's position as a philosopher seems to us to be something like the following:-He is more lucid and less assailable than Hamilton, because he is, when tried by Common Sense, of a sceptical turn of mind, and does not feel disposed to admit the veracity of our primary convictions, unless he has presented to him by others the most conclusive proof of it. He performs a negative function rather than a positive one. He seems, for example, to be an idealist, not from the force of his argumentation, like Ferrier, but because he does not deem the evidence which is cited in support of the realistic doctrine to be satisfactory. Similarly, it is for lack of evidence that he appears to deny the existence of necessary truth. The function which he performs as a psychologist may be compared perhaps to that which the pylorus fulfils in the digestive economy; and Common Sense counts with him, as with Ferrier, for nothing. Mr. Mill, in short, is great at constructing a system; he tests much and well, but he originates little.

W. G. D.