and doubtful it appears requisite that some things should be interposed;* partly for the purpose of instruction, partly for present use. Of these the first is, that some examples be offered of investigation and discovery according to our system and method.... We speak now of such examples only as may be of the nature of types and models, placing as it were before our eyes the whole process of the mind, and the continuous frame and order of discovery in particular subjects, and they various and of note. . . To examples of this kind, therefore, we devote the Fourth Part of our work; which in fact is nothing else than a particular and expanded application of the Second Part.

"The Fifth Part is introduced only for a temporary purpose, until what remains can be finished. . . . It is made up of whatsoever things we have ourselves either found out, or proved, or added; and that not exclusively by the proper methods and rules of interpretation, but simply by that same exercise of the understanding which other men are accustomed to use in investigation and discovery.

"Finally, the Sixth Part of our work, to which all the other parts are subservient and ministerial, at length discloses and propounds that philosophy which is educed and constituted out of that legitimate, chaste, and severe inquisition, which we have previously taught and prepared. But to accomplish and bring to a termination this last part is a thing both beyond our strength and beyond our hopes. We hope indeed to furnish no contemptible beginning of it; the fortune of the human race will supply the end; which will be such perhaps as, in the present state of things and of men's minds, the imagination cannot easily comprehend or take measure of."

The panoramic view of his vast design which Bacon spreads out before us in this preliminary discourse, is for

* The meaning is not, as Mr. Wood gives it, "a few reflections must necessarily be here inserted." The "quaedam interponenda" are the subjects of the Fourth Part of the work, the Scala Intellectus.

the greater part as luminous and distinct as it is sweeping and magnificent. It will convey a complete conception to whoever will study it attentively of the general nature and object at least of the three first parts of the Instauratio Magna; the latter portion of the work, upon the actual composition of which the author cannot be said to have ever properly entered, seems to have floated somewhat vaguely before his own eye, and it may be said to form a distant back-ground in the picture he has here sketched. In our abstract, we have omitted much of the mere eloquénce and illustration, with many ingenious, penetrating, and most felicitously expressed remarks; but we have preserved all the substance of the state

ment.

Bacon's adoption of the designation of a new logic, or dialectics, for his proposed method of investigating nature, and his comparison of the method with the vulgar or common logic, are sufficiently accounted for by the use that had come to be made of logical formulæ in the discussion of scientific questions. It is true that the syllogism is the universal form of reasoning, that all demonstration when fully developed and expressed must fall into one or other of the varieties of that form. The defect of the scientific reasoning of the schools, therefore, did not consist in its addictedness to syllogistic forms. The most perfect reasoning in the world, that of Euclid's Elements of Geometry, is every where a series of syllogisms. The error of the philosophy, both physical and moral, which formerly prevailed, and against which Bacon directs his attacks, lay in the employment of the syllogism for a purpose for which it was wholly incompetent, which was altogether beside its function and out of its province. A syllogism can establish no absolute truth. Its conclusion may be absolutely true: but all that the syllogism makes out, or professes to establish, is, that it is true provided the premisses are true. A syllogism is only a conditional affirmation. It is a statement that, given certain things, a certain other thing will follow. And one of the advantages which the syllogisms of geometry have is, that their premisses are all pure sup

positions, mere conceptions which the mind forms without having to look beyond itself. We are not denying that the conceptions or suppositions are true. They have in fact the peculiar character of being such that it is impossible for the mind not to believe them to be true. But, for that matter, the Arabian Nights' Entertainments might be delivered in a series of syllogisms. Given, it might be said, so many genies, giants, and enchanters, and such and such effects will follow. The one proposition would be as true as the other; the conclusion would be true if the premisses were true; and that is all that logic can make out in any case. The old writers on science were wont to employ it as if they thought it could do a great deal more. Its proper and only function is the exposition of an argument; they seemed often to think that a correctly constructed syllogism was the sufficient explanation of a phenomenon.

At the same time Bacon is not justified in making this matter of charge against the common logic. There is usually no fault to be found with the mere logic of the old scientific writers. Their conclusions are legitimately deduced from their premisses; and that is all that can be required on the score of logic. The single respect in which their demonstrations are objectionable is, that they often set out from false or insufficiently established premisses; but with the establishment of premisses, as such, logic has nothing to do; its sole office is the deduction of conclusions. Its premisses are assigned to it, or may be assumed at pleasure.

It is true that a false proposition which is adopted as one of the premisses of a syllogism has often been previously obtained as the conclusion of another syllogism. But, although false as a premiss, it may have been true as a conclusion; that is to say, it may have been quite legitimately deduced from other premisses. In that case the fault of the demonstration will still be, as before, that some one or other of the premisses has been false.

The greatest amount of misconception and confusion of thought, however, in regard to these subjects, has been occasioned by Bacon's describing the method he proposes

for the investigation of natural phenomena and processes as a new logic, and designating it by the term induction. It has become common to distinguish it as the Inductive Logic.

Whatever else may be new in the Baconian method, there most certainly neither is, nor can be, any novelty in its logic. If there were, it would only be an illogical, that is, an unreasonable or absurd method. For nobody has ever pretended that the old logic is false; the worst charge that has been brought against it is that it is useless or inefficient. To talk of a new logic, differing in its principles from the old, is tantamount to talking of a new geometry, or a new species of square or circle.

But what Bacon understands by Induction is not a logic at all, or anything of the nature of a logic. Induction is the name given by the logicians to that kind of syllogism in which a universal conclusion is obtained from premisses relating to particulars, instead of a particular conclusion being derived from a universal proposition, as is more commonly the case. But the enumeration of particulars in such an induction is complete; and the conclusion, therefore, is as necessary as in the common syllogism. Thus, John, Thomas, and Henry, are each dark-haired; John, Thomas, and Henry make up all the family of the Smiths; therefore the Smiths are all dark-haired; is an example of logical induction. Bacon's induction is altogether different. In that, from a number of particular instances, examined by means of observation and experiment, and sifted by the proper rejections or exclusions, we infer, not by the necessary laws of thought (with which alone logic concerns itself), but on our experience of the uniformity of the operations of nature, on grounds of analogy, or on other such considerations, that a certain thing is probably universally true. This is not such a process as comes within the domain of logic, which, as already explained, undertakes to teach nothing more than how two propositions having a certain relation combine to generate a third, and in so teaching is entirely indifferent as to whether the generating propositions be true or false. A logical induction does not, any more than a logical deduc

tion, look beyond the mind itself: logic is the science of a certain mental process, not the science or art of the collection and examination of material facts. Its conclusions are, in all cases, necessary and irresistible, the premisses being admitted; and depend for their reception by the mind in no degree upon its knowledge or experience of any kind, or even upon the degree of its judgment, or capacity of weighing evidence. There is no evidence to be weighed or balanced in a syllogism, whether deductive or inductive all the evidence is upon one side.

It is true that so much of the Baconian Induction as consists in drawing the conclusion may be resolved into a logical form, by introducing, or assuming that there is always present to the mind, as one of the premisses, a proposition asserting the uniformity of the operations of nature. In this way the major proposition will be, What is found in examined instances will be found in all instances; the minor, A certain thing is what is found in examined instances; the conclusion, Therefore the same thing will be found in all instances. The middle term, (that by which the two premisses are connected so long as they continue distinct, and which like a bridge becomes unnecessary, and is removed, when they are in the conclusion brought together into one affirmation) will be, What is found in examined instances. But this only proves that, in so far as the Baconian Induction is a logical process, its logic is merely the common logic. As the term is used by Bacon, however, it includes also, and that, we may say, as its principal part, another process, the collection and examination of the instances, which, as we have seen, is not a logical process at all.