test of observation; and it is highly probable that the hypothesis he formed was, that the moon gravitated to the earth with a constant force, instead of a force varying inversely as the square of the distance; which was the result of another hypothesis, suggested by the elliptic orbits of the planets, and the force necessary to confine the motion of a projectile in such an orbit.

When Harvey observed the valves in the veins, he is commonly said to have made the discovery of the circulation of the blood by reasoning from final causes, or by asking of nature for what purpose such valves could be intended: but, perhaps he might have asked the question for ever, unless the resemblance between the valve of a vein and that of a pump had suggested a plausible hypothesis, in which he was confirmed by repeated experiments and observations directed to the point.

Analogy, so much slighted and over

looked, and to which such an inferior part in the advancement of science has been assigned, and that too with so much suspicious caution, appears to me to be the great instrument of generalization and invention, by which hypotheses are supplied, which are most commonly the subjects that call for the exercise of Induction. By Induction, as usually understood, we make it a rule to exclude all hypotheses: first of all, we collect the experiments, and, having obtained these, we are next to examine them and compare them; we then reject the irrelative and negative, and conclude upon the affirmatives that are left. By this means, says Lord Bacon, we question nature, and conclude upon her answers: yet I would venture to suggest, that, ninetynine times out of a hundred, the Analogy or comparison precedes the collection of the experiments: some resemblance is observed, some hypothesis is started, which is the subject that is brought to

the test of Induction. By this the hypothesis is either proved, or confuted, or more commonly limited to something less general.

I would not be understood to assert that the common inductive method is barren; for, no doubt, discoveries might be so made; but I really question whether a discovery was ever made according to its rules, which the discoverer had not, in his own mind, anticipated by Analogy as an hypothesis long before he had completed his investigation, and indeed guided his investigation by it. But, however that may be, it must be admitted, that thousands and thousands of discoveries are made and inventions brought into play, the result merely of analogy and a few experiments, or very commonly of a single experimentum crucis. By the common method proposed we take too wide a range, we embrace the whole subject at once, and require the completion of its natural history; but by

the proper use of analogy as a guide, we step cautiously but from one species to the next.

Induction has two instruments of operation; EXPERIMENT for all things within our reach, and OBSERVATION for such as are beyond us. And of these Observation is less efficient than Experiment; for in all Experiments Observation is involved but in many investigations Experiment is not attainable, and we are reduced to Observation only, because we cannot use Experiment. Now by Induction without Analogy we first ask innumerable irrelative and impertinent questions of nature, and then make use of Observation upon the experiments in hand; but by Induction with Analogy we try experiments or observe for a specific purpose, and obtain specific answers to the point.

Having thus obtained a general law or fact, or cause, for an entire genus, we may proceed in the same manner from

this genus to the next, till the whole order be included under the same or some more extensive generality: thus at length we may arrive at certain most general laws, or phenomena, or causes; beyond which it may not be within our power to proceed. To arrive inductively at a generality it is of course requisite to have in hand a most extensive collection of the facts or particulars, which must be all included under it: and Aristotle, no less specially than Bacon, insists upon this preliminary. The only question is, whether in reality we do not ascend to an extensive generality by a series of particular conclusions, by successively reducing each fact or individual under some general law, conceived a priori by Analogy from some single instance, instead of deciding at once by a formal and comprehensive survey of the whole.

The progress of science in the ascending scale consists in rising from Individuals to Generals and Universals.