of induction "per enumerationem simplicem "; of which kind of induction Bacon, as we have seen, has said that it is utterly vicious and incompetent.

The word realism may perhaps require some explanation. I mean by it the opinion, which Bacon undoubtedly entertained, that for the purpose of investigation, the objects of our thoughts may be regarded as an assemblage of abstract conceptions, so that these conceptions not only correspond to realities, which is of course necessary in order to their having any value, but may also be said adequately to represent them. In his view of the subject, ideas or conceptions (notiones) reside in some sort in the objects from which we derive them; and it is necessary, in order that the work of induction may be successfully accomplished, that the process by which they are derived should be carefully and systematically performed. But he had not perceived that which now at least can scarcely be doubted of, that the progress of science continually requires the formation of new conceptions whereby new principles of arrangement are introduced among the results which had previously been obtained, and that from the necessary imperfection of human knowledge our conceptions never, so to speak, exhaust the essence of the realities by which they are suggested. The notion of an alphabet of the universe, of which Bacon has spoken more than once, must therefore be given up; it could at best be only an alphabet of the present state of knowledge. And similarly of the analysis into abstract natures on which the process of exclusion, as we have seen, depends. No such analysis can be used in the manner which Bacon prescribes to us; for every advance in knowledge presupposes the introduction of a new conception, by which the previously existing analysis is rendered incomplete, and therefore erroneous.

We have now, I think, succeeded in tracing the cause both of the peculiarities of Bacon's method, and of its practical inutility. Some additional information may be derived from an examination of the variations with which it is presented in different parts of his writings;-less however than if we could arrange his smaller works in chronological order. Nevertheless two results, not without their value, may be thus obtained; the one, that it appears probable that Bacon came gradually to see more of the difficulties which beset the practical application of his method; and the other, that the doctrine of Forms is in reality an extraneous part of his philosophy.

(10) In the earliest work in which the new method of induction is proposed, namely, the English tract entitled Valerius Terminus, no mention is made of the necessity of correcting commonly received notions of simple natures. The inductive method is therefore presented in its simplest form, unembarrassed with that which constitutes its principal difficulty. But when we advance from Valerius Terminus to the Partis secundæ Delineatio et Argumentum, which is clearly of a later date, we find that Bacon has become aware of the necessity of having some scientific method for the due construction of abstract conceptions. It is there said that the "pars informans", that is, the description of the new method, will be divided into three parts-the ministration to the senses, the ministration to the memory, and the ministration to the reason. In the first of these, three things are to be taught; and of these three the first is how to construct and elicit from facts a duly formed abstract conception (bona notio); the second is how the sense may be assisted; and the third, how to form a satisfactory collection of facts. He then proposes to go on to the other two ministrations.

Thus the construction of conceptions would have formed the first part of the then designed Novum Organum ; and it would seem that this arrangement was not followed when the Novum Organum was actually written, because in the meantime Bacon had seen that this part of the work involved greater difficulties than he had at first supposed. For the general division into "ministrationes" is preserved in the Novum Organum 38, though it has there become less prominent than in the tract of which we have been speaking. In the ministration to the senses, as it is mentioned in the later work, nothing is expressly included but a good and sufficient natural and experimental historia; the theory of the formation of conceptions has altogether disappeared, and both this ministration and

38 Nov. Org. ii. 10.

that to the memory are postponed to the last of the three, which contains the theory of the inductive process itself. We must set out, Bacon says, from the conclusion, and proceed in a retrograde order to the other parts of the subject. He now seems to have perceived that the theory of the formation of conceptions and that of the establishment of axioms are so intertwined together, that the one cannot be presented independently of the other, although in practice his method absolutely requires these two processes to be carried on separately. His view now is, that at first axioms must be established by means of the commonly received conceptions, and that subsequently these conceptions must themselves be rectified by means of the ulterior aids to the mind, the fortiora auxilia in usum intellectûs, of which he has spoken in the nineteenth aphorism of the second book. But these fortiora auxilia were never given, so that the difficulty which Bacon had once proposed to overcome at the outset of his undertaking remained to the last unconquered. The doctrine of the Novum Organum (that we must first employ commonly received notions, and afterwards correct them) is expressly laid down in the Ďe Interpretatione Naturæ Sententiæ Duodecim 39. Of this how

ever the date is uncertain.

It is clear that while any uncertainty remains as to the value of the conceptions (notiones) employed in the process of exclusion, the claim to absolute immunity from error which Bacon has made on behalf of his general method must be more or less modified and of this he seems to have been aware when he wrote the second book of the Novum Organum 40.

(II) Thus much of the theory of the formation of conceptions. With regard to the doctrine of Forms, it is in the first place to be observed that it is not mentioned as a part of Bacon's system, either in Valerius Terminus or in the Partis secundæ Delineatio, or in the De Interpretatione Nature Sententiæ Duodecim, although in the two last named tracts the definition of science which is found at the outset of the second book of the Novum Organum is in substance repeated. This definition, as we have seen, makes the discovery of Forms the aim and end of science; but in both cases the word form is replaced by causes. It is however to be admitted that in the Advancement of Learning, published in 1605, Forms are spoken of as one of the subjects of Metaphysique. Their not being mentioned except ex obliquo in Valerius Terminus is more remarkable, because Bacon has there given a distinct name to the process which he afterwards called the discovery of the Form. He calls it the freeing of a direction, and remarks that it is not much other matter than that which in the received philosophies is termed the Form or formal cause. Forms are thus mentioned historically, but in the dogmatic statement of his own view they are not introduced at all 41

The essential character of Bacon's philosophy, namely the analysis of the concrete into the abstract, is nowhere more prominent than in Valerius Terminus. It is there said "that every particular that worketh any effect is a thing compounded more or less of diverse single natures, more manifest and more obscure, and that it appeareth not to whether (which) of the natures the effect is to be ascribed " 42. Of course the great problem is to decide this question, and the method of solving it is called "the freeing of a direction". In explanation of this name, it is to be observed that in Valerius Terminus the practical point of view predominates. Every instance in which a given nature is produced is regarded as a direction for its artificial production. If air and water are mingled together, as in snow, foam, &c., whiteness is the result. This then is a direction for the production of whiteness, since we have only to mingle air and water together in order to produce it. But whiteness may be produced in other ways, and the direction is therefore not free. We proceed gradually to free it by rejecting, by means of other instances, the circumstances of this which are unessential: a process which is the exact counterpart of the Exclusiva of the Novum Organum. The instance I have given is Bacon's, who developes it at some length.

39 Vide § viii. of this tract.

40 Nov. Org. ii. 19.

41 I refer to my preface to Valerius Terminus for an illustration of some of the difficulties of this very obscure tract.

42 Val. Ter. c. 17.

Here then we have Bacon's method treated entirely from a practical point of view. This circumstance is worthy of notice because it serves to explain why Bacon always assumes that the knowledge of Forms would greatly increase our command over nature, that it "would enfranchise the power of man unto the greatest possibility of works and effects". It has been asked what reason Bacon had for this assumption. "Whosoever knoweth any Form," he has said in the Advancement, "knoweth the utmost possibility of superinducing that nature upon any variety of nature". Beyond question, the problem of superinducing the nature is reduced to the problem of superinducing the Form; but what reason have we for supposing that the one is more easy of solution than the other? If we knew the Form of malleability, that is, the conditions which the intimate constitution of a body must fulfil in order that it may be malleable, does it follow that we could make glass so? So far as these questions admit of an answer, Valerius Terminus appears to suggest it. Bacon connected the doctrine of Forms with practical operations, because this doctrine, so to speak, represented to him his original notion of the freeing of a direction, which, as the phrase itself implies, had altogether a practical significance.

Even in the Novum Organum the definition of the form is made to correspond with the præceptum operandi, or practical direction 43. The latter is to be "certum, liberum, et disponens sive in ordine ad actionem ". Now a direction to produce the Form as a means of producing the given nature is certain, because the presence of the Form necessarily determines that of the nature. It is free, because it requires only that to be done which is necessary, since the nature can never be present unless its Form is so too. Thus far the agreement between the practical and the scientific view is satisfactory. But to the third property which the practical direction is to possess, namely its being in ordine ad actionem, or such as to facilitate the production of the proposed result, corresponds the condition that the Form is to be" the limitation of a more general nature; " that is to say, the Form presents itself as a limitation of something more general than the given nature, and as determining, not merely logically but also causatively, the existence of the latter. At this point the divergence between the practical and the scientific view becomes manifest; practical operations do not, generally speaking, present to us anything analogous to the limitation here spoken of, and there is no reason to suppose that it is easier to see how this limitation is to be introduced than to see how the original problem, the ἑξ ἀρχῆς προκείμενον, may be solved. But this divergence seems to show that the two views are in their origin heterogeneous; that the one contains the fundamental idea of Bacon's method, while the other represents the historical element of his philosophy. We shall however hereafter have occasion to suggest considerations which may seem to modify this conclusion.

(12) In a survey of Bacon's method it is not necessary to say much of the doctrine of prerogative instances, though it occupies the greater part of the second book of the Novum Organum. It belongs to the unfinished part of that work; at least it is probable that its practical utility would have been explained when Bacon came to speak of the Adminicula Inductionis.

Twenty-seven kinds of instances are enumerated, which are said to excel ordinary instances either in their practical or their theoretical usefulness. To the word instance Bacon gives a wide range of signification. It corresponds more nearly to observation than to any other which is used in modern scientific language. Of some classes of these instances collections are to be made for their own sake, and independently of any investigation into particular natures. Such, for instance, are the instantiæ conformes; Bacon's examples of which are mostly taken from comparative anatomy. One of them is the analogy between the fins of fishes, the feet of quadrupeds, and the feet and wings of birds; another the analogy of the beak of birds and the teeth of other animals, &c 44.

43 Nov. Org. ii. 4, which is the best comment on the dictum, Knowledge is power. 44 Nov Org. ii. 27. It does not seem that Bacon added much to what he found in Aristotle on the subject of these analogies.

The other classes of prerogative instances have especial reference to particular investigation, and are to be collected when individual tables of comparence are formed.

It would seem from this that the theory of prerogative instances is intended to guide us in the formation of these tables. But it is difficult to see how the circumstances which give any instance its prerogative could have been appreciated à priori. An instantia crucis 45, to take the most celebrated of all, has its distinguishing character only in so far as it is viewed with reference to two contending hypotheses. In forming at the outset of an inquiry the appropriate tables, nothing would have led the interpreter to perceive its peculiar value. This theory, whatever may be its practical utility, may supply us with new illustrations of the importance in Bacon's method of the process of exclusions. At the head of the list-and placed there, we may presume, from the importance of the end which they promote-stand the instantiæ solitaria, whose prerogative it is to accelerate the Exclusiva 46. These are instances which exhibit the given nature in subjects which have nothing in common, except that nature itself, with the other subjects which present it to us. Thus the colours shown by the prism or by crystals are a solitary instance of colour, because they have nothing in common with the fixed colours of flowers, gems, &c. Whatever therefore is not independent of the particular constitution of these bodies must be excluded from the form of colour.

Next to the instantiæ solitariæ are placed the instantiæ migrantes, which show the given nature in the act of appearing or of disappearing; as when glass, being pounded, becomes white. Of these it is said they not only accelerate and strengthen the Exclusiva, but also confine within narrow limits the Affirmative, or Form itself, by showing that it is something which is given or taken away by the observed change. A little farther on Bacon notices the danger in these cases of confounding the efficient cause with the Form, and concludes by saying "But this is easily remedied by a legitimately performed Exclusiva ".

Other remarks to the same effect might be made with reference to other classes of instances; but these are probably sufficient.

I shall now endeavour to give an account of Bacon's views on some questions of philosophy, which are not immediately connected with the reforms he proposed to introduce.

(13) It has sometimes, I believe, been supposed that Bacon had adopted the atomic theory of Democritus. This however is by no means true; but certainly he often speaks much more favourably of the systems of the earlier physicists and especially of that of Democritus, than of the philosophy of Plato and Aristotle. In doing this he may perhaps have been more or less influenced by a wish to find in antiquity something with which the doctrines he condemned might be contrasted. But setting this aside, it is certain that these systems were more akin to his own views than the doctrine of the schools of which Socrates may be called the founder. The problems which they proposed were essentially physical:given certain material first principles, to determine the origin and causes of all phenomena. They were concerned, for the most part, with that which is accessible to the senses, or which would be so if the senses were sufficiently acute. In this they altogether agree with Bacon, who, though he often speaks of the errors and shortcomings of the senses, yet had never been led to consider the question which stands at the entrance of metaphysical philosophy, namely whether the subjective character of sensation does not necessarily lead to scepticism, if no higher grounds of truth can be discovered. The scepticism of Protagoras, and Plato's refutation of it, seemed to him to be both but idle subtleties. Plato, Aristotle, and their followers, were in his opinion but a better kind of sophists. What Dionysius said to Plato, that his discourse was but dotage, might fitly be applied to them all 47.

It cannot be denied, that to Bacon all sound philosophy seemed to be included in what we now call the natural sciences; and with this view he was naturally led to prefer the atomic doctrine of Democritus to any metaphysical speculation.

45 Nov. Org. ii. 36. 46 Nov. Org. ii. 22. 47 Redargut. Phil. et Nov. Org. i. 71.

Every atomic theory is an attempt to explain some of the phenomena of matter by means of others; to explain secondary qualities by means of the primary. And this was what Bacon himself proposed to do in investigating the Forms of simple natures. Nevertheless he did not adopt the peculiar opinions of Democritus and his followers. In the Novum Organum he rejects altogether the notion of a vacuum and that of the unchangeableness of matter 48. His theory of the intimate constitution of bodies does not, he remarks, relate to atoms properly so called, but only to the actually existing ultimate particles. Bacon cannot therefore be said to be a follower of Democritus, though he has spoken of him as being, of all the Greek philosophers, the one who had the deepest insight into nature 49. But though Bacon was not an atomist, he was what has been called a mechanical physiologist. Leibnitz's remark that the restorers of philosophy 50 all held the principle that the properties of bodies are to be explained by means of magnitude, figure, and motion (a statement which envelopes every such theory of matter as that of Descartes, together with the old atomic doctrine), is certainly true of Bacon.

(14) The opinion which Bacon had formed as to the class of subjects which ought to be included in Summary Philosophy (the English phrase by which he renders the expression he sometimes uses, namely prima philosophia), is worthy of attention.

In the writings of Aristotle, the first philosophy denotes the science which since his time has been called metaphysic. It is the science of first principles, or as he has himself defined it, the science of that which is, as such. In the first book of the Metaphysics we find a proof of the necessity of having such a science, distinct from and in a manner superior to all others.

Bacon, adopting Aristotle's name, applied it differently. With him, the first philosophy is divided into two parts. Of these the first is to be a receptacle of the axioms which do not belong exclusively to particular sciences, but are common to more than one; while the second is to inquire into the external or adventitious conditions of existences-such as the much and the little, the like and the unlike, the possible and impossible, &c.

In illustration of the contents of the first part, Bacon quotes several axioms which are applicable in more than one science. Of these the first is, "If to unequals are added equals, the sums are unequal," which is a mathematical principle, but which, Bacon says, referring to the distinction laid down by Aristotle between commutative and distributive justices, obtains also in moral science; inasmuch as it is the rule by which distributive justice must be guided. The next is, "Things which agree with a third, agree with one another,"-which is also a mathematical principle, but yet, differently stated, forms the foundation of the theory of syllogism. Thus far Bacon's doctrine does not materially dissent from Aristotle's, who has taught the necessity of recognising in all sciences two kinds of principles, those which are proper to the subject of each science, and those which, connecting themselves with the doctrine of the categories, are common to all. The last are in his nomenclature axioms, though Bacon, following probably Ramus, who in his turn followed Cicero and the Stoics, gives a much more general sense to this word; and it is to be remarked that Aristotle has given as an instance of an axiom the first of the two which I have quoted from Bacon, or at any rate another which is in effect equivalent to it. But most of the instances which Bacon goes on to give are of a different nature. They are not derived from the laws of thought, but on the contrary involve an empirical element, and therefore are neither self-evident nor capable of an à priori proof. Thus the axiom that "a discord resolved into a concord improves the harmony", is, Bacon says, not only true in music, but also in ethics and the doctrine of the affections. But this axiom is in its literal sense merely a result of observation, and its application to moral subjects is clearly only analogical or tropical. Again, that "the organs of the senses are

48 Nov. Org. ii. 8. Compare Cogit. De Nat. Rerum. 49 Nov. Org. i. 51. ; also Parm. Teles. and Dem. Phil.

50 Namely, the Cartesians, Verulam, Hobbes, &c. See his letter to Thomasius, p. 48 of the edition of his philosophical works by Erdmann.