of physical the latter resolving ultimately into our instinctive expectation of the laws of nature, and consequently, never amounting to that demonstrative certainty which excludes the possibility of anomalous exceptions.

I have been led to this train of thinking by a remark which La Place appears to me to have stated in terms much too unqualified; "Que la marche de Newton, dans la découverte de la gravitation universelle, a été exactement la même, que dans celle de la formule du binome." When it is recollected, that, in the one case, Newton's conclusion related to a contingent, and in the other to a necessary truth, it seems difficult to conceive, how the logical procedure which conducted him to both should have been exactly the same. In one of his queries he has, in perfect conformity to the principles of Bacon's logic, admitted the possibility, that "God may vary the laws of nature, and make worlds of several sorts, in several parts of the universe." "At least," he adds, "I see nothing of contradiction in all this." (Query 31.) Would Newton have expressed himself with equal skepticism concerning the universality of his binomial theorem; or admitted the possibility of a single exception to it, in the indefinite progress of actual involution? In short, did there exist the slightest shade of difference between the degree of his assent to this inductive result, and that extorted from him by a demonstration of Euclid?

Although, therefore, the mathematician, as well as the natural philosopher, may, without any blamable latitude of expression, be said to reason by induction, when he draws an inference from such cases, he rests his conclusions on the grounds essentially distinct from those which form the basis of experimental science.

The word analogy, too, as well as induction, is common to physics and to pure mathematics. It is thus we speak of the analogy running through the general properties of the different conic sections, with no less propriety than of the analogy running through the anatomical structure of different tribes of animals. In some instances, these mathematical analogies are collected by a species of induction; in others, they are inferred as consequences from more general truths, in which they are included as particular cases. Thus, in the curves which have just been mentioned, while we content ourselves, as many elementary writers have done, with deducing their properties from mechanical descriptions on a plane, we rise experimentally from a comparison of the propositions which have been separately demonstrated with respect to each curve, to more comprehensive theorems, applicable to all of them; whereas, when we begin with considering them in their common origin, we have it in our power to trace from the source, both their generic properties, and their specific peculiarities. The satisfaction arising from this last view of the subject can be conceived by these alone

• L'Hospital, Simson, &c.

who have experienced it; although I am somewhat doubtful whether it be not felt in the greatest degree by such as, after having risen from the contemplation of particular truths to other truths more general, have been at last conducted to some commanding station, where the mutual connexions and affinities of the whole system are brought, at once, under the range of the eye. Even, however, before we have reached this vantage-ground, the contemplation of the analogy, considered merely as a fact, is pleasing to the mind; partly, from the mysterious wonder it excites, and partly from the convenient generalization of knowledge it affords. To the experienced mathematician this pleasure is farther enhanced, by the assurance which the analogy conveys of the existence of yet undiscovered theorems, far more extensive and luminous than those which have led him, by a process so indirect, so tedious, and comparatively so unsatisfactory, to his general conclusions.

In this last respect, the pleasure derived from analogy in mathematics, resolves into the same principle with that which seems to have the chief share in rendering the analogies among the different departments of nature so interesting a subject of speculation. In both cases, a powerful and agreeable stimulus is applied to the curiosity, by the encouragement given to the exercise of the inventive faculties, and by the hope of future discovery, which is awakened and cherished. As the analogous properties, for instance, of the conic sections, point to some general theorems, of which they are corollaries; so the analogy between the phenomena of electricity and those of galvanism irresistibly suggests a confident, though vague anticipation of some general physical law comprehending the phenomena of both, but differently modified in its sensible results by a diversity of circumstances.* Indeed, it is by no means impossible, that the pleasure we receive even from those analogies which are the foundation of poetical metaphor and simile, may be found resolvable, in part, into the satisfaction connected with the supposed discovery of truth, or the supposed acquisition of knowledge; the faculty of imagination giving to these illusions a momentary ascendant over the sober conclusions of experience; and gratifying the understanding with a flattering consciousness of its own force, or at least with a consolatory forgetfulness of its own weakness.

For the analogies between Galvanism and Electricity, see Traité Elémentaire de Physique, par M. L'Abbé Hauy, sec. 717.-The passage concludes with the following remark, which may be regarded as an additional proof, that even when analogical conjectures appear to depart the most widely from the evidence of experience, it is from experience that they derive their whole authority over the belief. "Partout le fluide électrique semble se multiplier par la diversité des phénomènes; et il nous avait tellement accoutumeés à ses métamorphoses, que la nouveauté même de la forme sous laquelle il s'offrait dans le Galvanisme naissant, semblait être une raison de plus pour le reconnaître."

SECTION V.

Of certain misapplications of the words Experience and Induction, in the phraseology of modern science. Illustrations from medicine and from political economy.

In the first section of this chapter, I endeavored to point out the characteristical peculiarities by which the inductive philosophy of the Newtonians is distinguished from the hypothetical systems of their predecessors; and which entitle us to indulge hopes with respect to the permanent stability of their doctrines, which might be regarded as chimerical, if, in anticipating the future history of science, we were to be guided merely by the analogy of its revolutions in the ages that are past.

In order, however, to do complete justice to this argument, as well as to prevent an undue extension of the foregoing conclusions, it is necessary to guard the reader against a vague application of the appropriate terms of inductive science to inquiries which have not been rigorously conducted according to the rules of the inductive logic. From a want of attention to this consideration, there is a danger, on the one hand, of lending to sophistry or to ignorance the authority of those illustrious names whose steps they profess to follow; and, on the other, of bringing discredit on that method of investigation, of which the language and other technical arrangements have been thus perverted.

Among the distinguishing features of the new logic, when considered in contrast with that of the schoolmen, the most prominent is the regard which it professes to pay to experience, as the only solid foundation of human knowledge. It may be worth while, therefore to consider, how far the notion commonly annexed to this word is definite and precise: and whether there may not sometimes be a possibility of its being employed in a sense more general and loose, than the authors who are looked up to as the great models of inductive investigation understood it to convey."

* As the reflections which follow are entirely of a practical nature, I shall express myself, as far as is consistent with a due regard to precision, agreeably to the modes of speaking in common use; without affecting a scrupulous attention to some speculative distinctions, which, however curious and interesting, when considered in connexion with the theory of the mind, do not lead to any logical conclusions of essential importance in the conduct of the understanding. In such sciences for example, as astronomy, natural philosophy, and chemistry, which rest upon phenomena open to the scrutiny of every inquirer, it would obviously be puerile in the extreme to attempt drawing the line between facts which have been ascertained by our own personal observation, and those which we have implicitly adopted upon our faith in the universal consent of the scientific world. The evidence, in both cases, may be equally irresistible; and sometimes the most cautious reasoners may justly be disposed to consider that of testimony as the least fallible of the two.

By far the greater part, indeed, of what is commonly called experimental knowledge, will be found, when traced to its origin, to resolve entirely into our

In the course of the abstract speculations contained in the preceding section, I have remarked, that although the difference between the two sorts of evidence, which are commonly referred to the separate heads of experience and of analogy, be rather a difference in degree than in kind, yet that it is useful to keep these terms in view, in order to mark the contrast between cases which are separated from each other by a very wide and palpable interval; more especially to mark the difference between an argument from individual to individual of the same species, and an argument from species to species of the same genus. As this distinction, however, when accurately examined, turns out to be of a more vague and popular nature than at first sight appears, it is not surprising that instances should occasionally present themselves, in which it is difficult to say, of the evidence before us, to which of these descriptions it ought to be referred. Nor does this doubt lead merely to a question concerning phrascology: it produces a hesitation which must have some effect even on the judgment of a philosopher; the maxims to which we have been accustomed, in the course of our early studies, leading us to magnify the evidence of experience as the sole test of truth; and to depreciate that of analogy, as one of the most fertile sources of error. As these maxims proceed on the supposition, that the respective provinces of both are very precisely defined, it is evident, that, admitting them to be perfectly just in themselves, much danger may still be conceivable from their injudicious application. I shall endeavor to illustrate this remark by some familiar instances; which I trust, will be sufficient to recommend it to the farther consideration of future logicians. To treat of the subject with that minuteness of detail which is suited to its importance, is incompatible with the subordinate place which belongs to it in my general design.

confidence in the judgment and the veracity of our fellow creatures; nor, in the sciences already mentioned, has this identification of the evidence of testimony with that of experience, the slightest tendency to affect the legitimacy of our inductive conclusions.

In some other branches of knowledge, (more particularly in those political doctrines which assume as incontrovertible data the details of ancient history,) the authority of testimony is, for obvious reasons, much more questionable; and to dignify it, in these, with the imposing character of experience, is to strengthen one of the chief bulwarks of popular prejudices. This view of the subject, however, although well entitled to the attention of the logician, has no immediate connexion with my present argument; and accordingly I shall make no scruple, in the sequel, to comprehend, under the name of experience, the grounds of our assent to all the facts on which our reasonings proceed, provided only that the certainty of these facts be, on either supposition equally indisputable.

The logical errors which it is the aim of this section to correct, turn upon a still more dangerous latitude in the use of this word in consequence of which, the authority of experience comes insensibly to be extended to innumerable opinions resting solely on the supposed analogies; while, not unfrequently, the language of Bacon is quoted in bar of any theoretical argument on the other side of the question.

I have added this note, partly to obviate some criticisms, to which my own phraseology may, at first sight, appear liable: and partly to point out the connexion between the following discussion, and some of the foregoing speculations.

It is observed by Dr. Reid, that, "in medicine, physicians must, for the most part, be directed in their prescriptions by analogy. The constitution of one human body is so like that of another, that it is reasonable to think, that what is the cause of health or sickness to one, may have the same effect on another. And this," he adds, "is generally found true, though not without some exceptions."-(Essays on the Intellect. Powers, p. 53.)

I am doubtful if this observation be justified by the common use of language; which, as far as I am able to judge, uniformly refers the evidence on which a cautious physician proceeds, not to analogy, but to experience. The German monk, who, according to the popular tradition, having observed the salutary effects of antimony upon some of the lower animals, ventured to prescribe the use of it to his own fraternity, might be justly said to reason analogically; inasmuch as his experience related to one species, and his inference to another. But if, after having thus poisoned all the monks of his own convent, he had persevered in recommending the same mineral to the monks of another, the example of our most correct writers would have authorized us to say, (how far justly, is a different question,) that he proceeded in direct opposition to the evidence of experience.

In offering this slight criticism on Dr. Reid, I would be very far from being understood to say, that the common phraseology is more unexceptionable than his. I would only remark, that his phraseology on this occasion is almost peculiar to himself: and that the prevailing opinions, both of philosophers and of the multitude, incline them to rank the grounds of our reasoning in the medical art, at a much higher point in the scale of evidence than what is marked by the word analogy. Indeed, I should be glad to know, if their be any one branch of human knowledge, in which men are, in general more disposed to boast of the lights of experience than in the practice of medicine.

It would, perhaps, have been better for the world, if the general habits of thinking and speaking had, in this instance, been more agreeable than they seem to be in fact, to Dr. Reid's ideas; or, at least, if some qualifying epithet had been invariably added to the word experience, to show with how very great latitude it is to be understood, when applied to the evidence on which the physician proceeds in the exercise of his art. The truth is, that, even on the most favorable supposition, this evidence, so far as it rests on experience, is weakened or destroyed by the uncertain conditions of every new case to which his former results are to be applied; and that, without a peculiar sagacity and discrimination in marking, not only the resembling, but the characteristical features of disorders, classed under the same technical name, his practice can not, with propriety, be said to be guided by any one rational principle of decision, but merely by blind and random conjecture. The more successfully this sagacity and discrimination are exercised, the more