of a general nature, is itself, in this respect, comprehended as a species of Contiguity: whereas he should have extended his view, so as to have embraced all the three primary principles which, as it were, connect and regulate our trains of thought: upon the relations of every one of which, I conceive, that, probable reasonings and analogical* arguments are continually based.

ANALOGY can hardly be called Resemblance or any species of Resemblance as it is ranked by Brown; but it is a kind of argument, a probable argument founded chiefly upon that relation. And it is also manifest, that the very

* As the word Analogy is not necessarily, though it is commonly and even anciently, connected with Resemblance alone, I see no reason why a probable argument, founded upon either of the other relations of Contiguity and Contrast, should not be termed Analogical, especially as all the three principles are probably reducible to one and the same, viz. that of Contiguity.

common restriction of the word hypothesis, to imaginary causes only, is incorrect, and springs from the same error which would confine probable or moral reasoning to the relations of Causation.

The great ABUSE OF ANALOGY is resting in its hypotheses, without bringing them to the test, building systems upon such hypotheses, and bending the facts to their support. Of this, the Timæus of Plato exhibits a curious example. He professes in that discourse to philosophize by λόγοι εἰκότες, which be translated analogical reasonings, alone and this dialogue, no less than many of the physical treatises of Aristotle, is a proof how little the ancients. were in the habit of attending to the excellent rules of investigation, which they had themselves laid down.

may

An ascertained fact, or law, or cause, is like a centre or a stem, from which many hypotheses branch forth on all

sides, varying according to the different resemblances, which the different objects or genera bear to one another. One only of these hypotheses can be the truth: but each of them, if assumed, becomes another centre or stem, from which as many more hypotheses proceed: and such is the case ad infinitum. The chances of error, therefore, increase in a very high geometrical proportion, whose ratio of course must be indefinitely variable, so as to include all the possibilities. A single hypothetical step into the regions of discovery we may, probably, make aright, but it is not to be depended upon, till it is proved: and the chances are very much against it. Nevertheless, we may, possibly, even make a second step into these unknown regions without proving the first: but the chances against it are prodigiously multiplied. And any attempt at a second step, before the first is proved, is a false method of proceeding. It is like false heraldry,

D

laying colour upon colour. The colour can only be blazoned on a metal, nor ought an hypothesis to be laid but upon the sterling basis of a truth. When we make a step into these regions of imagination, a new field is opened to us, presenting us with explanations of phenomena and objects of attention, which we had never before conceived, each inviting us to make another step; and so delighted are we with the wide and fair prospects before us, that we are apt to forget that we are but in the fields of imagination, and that it signifies but little, which hypothesis we had assumed, which path we followed, as all of them open to us prospects as delightful, as they are visionary. It is a fault and an abuse, almost equally to be imputed to modern as to ancient philosophers.

Another abuse of Analogy is arguing from individuals to genera, or from genus to genus, when these genera are too remote; which is skipping to generalities

instead of cautiously proceeding from species to species: not that it is of very material consequence, save only, that, the difficulty of the proof being increased, an hypothesis, so obtained, is apt to be received upon insufficient grounds.

But the most dangerous abuse of all is arguing from Matter to Mind, of which, as well as of the kind of similarity that exists between them, I shall have occasion presently to speak more at large.

I would observe, also, the great laxity in the significations of the word THEORY. It is sometimes used for a general law or principle obtained by Induction, and as something almost synonymous with hypothesis in this view it might be looked upon as a proved hypothesis. In its other and more general signification it implies the whole system or chain of reasoning from general laws and principles, and sometimes the result of such