3. Hence in every inference by Induction, there is some Conception _superinduced_ upon the Facts: and we may henceforth conceive this to be the peculiar import of the term _Induction_. I am not to be understood as asserting that the term was originally or anciently employed with this notion of its meaning; for the peculiar feature just pointed out in Induction has generally been overlooked. This appears by the accounts generally given of Induction. 'Induction,' says Aristotle[6\2], 'is when by means of one extreme term[7\2] we infer the other extreme term to be true of the middle term.' Thus, (to take such exemplifications as belong to our subject,) from knowing that Mercury, Venus, Mars, describe ellipses about the Sun, we infer that all Planets describe ellipses about the Sun. In making this inference syllogistically, we assume that the evident proposition, 'Mercury, Venus, Mars, do what all Planets do,' may be taken _conversely_, 'All {75} Planets do what Mercury, Venus, Mars, do.' But we may remark that, in this passage, Aristotle (as was natural in his line of discussion) turns his attention entirely to the _evidence_ of the inference; and overlooks a step which is of far more importance to our knowledge, namely, the _invention_ of the second extreme term. In the above instance, the particular luminaries, Mercury, Venus, Mars, are one logical _Extreme_; the general designation Planets is the _Middle Term_; but having these before us, how do we come to think of _description of ellipses_, which is the other Extreme of the syllogism? When we have once invented this 'second Extreme Term,' we may, or may not, be satisfied with the evidence of the syllogism; we may, or may not, be convinced that, so far as this property goes, the extremes are co-extensive with the middle term[8\2]; but the _statement_ of the syllogism is the important step in science. We know how long Kepler laboured, how hard he fought, how many devices he tried, before he hit upon this _Term_, the Elliptical Motion. He rejected, as we know, many other 'second extreme Terms,' for example, various combinations of epicyclical constructions, because they did not represent with sufficient accuracy the special facts of observation. When he had established his premiss, that 'Mars does describe an Ellipse about the Sun,' he does not hesitate to _guess_ at least that, in this respect, he might _convert_ the other premiss, and assert that 'All the Planets do what Mars does.' But the main business was, the inventing and verifying the proposition respecting the Ellipse. The Invention of the Conception was the great step in the _discovery_; the Verification of the Proposition was the great step in the _proof_ of the discovery. If Logic consists in pointing out the conditions of proof, the Logic of Induction must consist in showing what are the conditions of proof, in such inferences as this: but this subject must be pursued in the next chapter; I now