found it uniformly to be white, we conclude, that the milk of all animals is so. In like manner, having witnessed the effect of fire on several pieces of gold, iron, lead, and so forth, we affirm, that all metals are fusible. In this way, beginning with individuals, we ascend to species; and thence proceed from less general to more general conclusions, till we arrive at those abstract propositions, which are called axioms or general truths.

98. This method of induction is recommended by Lord Bacon, as the first and most important instrument of reason, in its search after truth. We employ it not only in the investigation of general truths, relating to things in actual existence; but in gaining those practical rules and maxims, by which the common business of life is carried on.

99. The use of induction, in learning the signification of words, is thus happily explained by Mr. Stewart: "A familiar illustration of "this process presents itself in the expedient, "which a reader naturally employs for deciphering the meaning of an unknown word, in "a foreign language, when he happens not to "have a dictionary at hand. The first sen

66

66 tence, where the word occurs, affords, it is "probable, sufficient foundation for a vague conjecture concerning the notion, annexed "to it by the author; some idea or other being "necessarily substituted in its place, in order "to make the passage at all intelligible. The "next sentence, where it is involved, renders "this conjecture a little more definite; a third "sentence contracts the field of doubt within "still narrower limits; till at length a more "extensive induction fixes completely the sig"nification we are in quest of. There cannot

"be a doubt, I apprehend, that it is in some "such way as this, that children slowly and "imperceptibly enter into the abstract and com"plex notions, annexed to numberless words in "their mother tongue, of which we should find "it difficult, or impossible, to convey the sense "by formal definitions."*

100. In another place, Mr. Stewart has described the manner of using induction, in tracing an event to its physical cause: "As "we can, in no instance, perceive the link, by "which two successive events are connected, so

* Philosophical Essays, essay v. ch. 1.

"as to deduce, by reasoning a priori, the one "from the other, as a consequence or effect, "it follows that, when we see an event take

66

66

place, which has been preceded by a com"bination of different circumstances, it is impossible for human sagacity to ascertain, "whether the effect is connected with all the "circumstances, or only with a part of them; "and, on the latter supposition, which of the "circumstances is essential to the result, and “which are merely accidental accessories or 'concomitants. The only way, in such a case, "of coming at the truth, is to repeat over the "experiment again and again, leaving out all "the different circumstances successively, and "observing with what particular combinations "of them the effect is conjoined.

"When, by thus comparing a number of "cases, agreeing in some circumstances, but differing in others, and all attended with the "same result, a philosopher connects, as a

66

general law of nature, the event with its

physical cause, he is said to proceed according "to the method of induction."*

• Elements of the Philosophy of the Mind, vol. ii. ch. 4. sect. 1.

101. Inductive conclusions will amount to moral certainty, whenever our experience has been uniform, and the number of cases examin

ed sufficiently numerous. But this reasoning is liable to be fallacious through impatience in the investigation, by which judgments are hastily formed, without a sufficient accumulation of facts. The number of instances, required to justify a general conclusion, must be increased in proportion as the facts, from which we reason, are more irregular in their appearance. In judging concerning the properties of inanimate matter, a general inference may sometimes be drawn from a small number of particular cases. If, for example, aqua fortis has been known to dissolve silver in one instance, the presumption is very strong, that it will do so in all. But the success, which may happen to attend a medicine in a single instance, furnishes but a slight presumption with regard to its general operation on the human body.

102. When our experience has not been uniform, the conclusions we make will fall short of moral certainty. An equal number of favourable and unfavourable instances leaves the mind in a state of suspense, without exciting

the smallest expectation on either side. As the ratio, which the instances on the two sides bear to each other, may vary indefinitely, so must the judgments, founded on them, vary in a like degree from the neighbourhood of certainty, down to that of entire improbability.*

CHAPTER FOURTH.

ANALOGY.

103. Analogy is the foundation of another species of moral reasoning, similar in most respects to analytical induction. They both proceed on the same general principle, that nature is consistent and uniform in her operations; so that from similar circumstances similar effects may be expected; and in proportion as the resemblance between two cases diminishes, the less confidence must be placed in the conclusions, made from the one to the other. The word analogy is used with much vagueness. Sometimes it denotes only a slight and distant

Campbell, Phil. Rhet. vol. i. Truth, part i. ch. 2. sect. 6.

* Bacon, Novum Organum, lib. i. ch. 5. sect 2. Beattie, Essay on Tatham, Chart and Scale of Truth, vol. i. ch. 4. sect. 1. Stewart, Elem. vol. ii. ch. 4. Gambier, Mor. Evidence, ch. 2. Scott, Intel. Phil. Appendix, ch. 2.