Page:Mind (Old Series) Volume 12.djvu/569

 556 J. M. EIGG : turned to rest, so also might conceivably the propagation of sound be explained. It also lay on the surface that pitch might vary as the velocity of the undulations ; and, the fact (proved by Newton) that the rate of propagation is constant for all degrees of loudness according with the fact that the rate of propagation of visible waves is constant for all degrees of amplitude or range of swing, it was inferred that loudness varies with the amplitude of the sound-wave. These essen- tial points settled, it only remained to adapt the theory to the exigencies of special cases. Even Helmholtz's analysis of the composition of musical notes is of this kind. From the familiar fact that harmony is pleasant he inferred that a musical note was probably itself a harmony, and the results obtained by magnifying sound by mechanical appliances accorded with his hypothesis. In like manner the theory that the celestial luminaries are solid bodies similar to the earth can only have been reached by a process of analogical reasoning. Anaxagoras, e.g., was clearly reasoning by analogy when he broached the theory that the sun and stars were masses of stone ignited by the force of rotation. Again, the molecular theory involves the hypothetical endowment of that which to sense is perfectly simple with a complex structure analogous to that with which we are familiar in composite bodies. It also is a mere analogy, un verifiable empirically but accepted because it enables us to assimilate the infra-sensible to the sensible world. Nay, the very conception of an object is an analogical transference of the unity and identity which we know as self to that which in itself is a mere cluster of perceptions, and the idea of force, as Mr. Spencer is fond of telling us, is derived by analogy from the experience of volition. Now if hypothesis is, as I hold, the most important part of the inductive process, it follows that the principle of induc- tion is aptly expressed in the much-criticised Newtonian canon: " Effectuum naturalium ejusdem generis eaedem sunt causse ". l On the strength of a similarity in the phenomena, which may be obvious or may be recondite, a similarity in the causes which determine them is inferred, and an attempt is then made to deduce the phenomenon from the hypothesis, which is accepted or rejected according as it does or does not admit of such deduction. Hence a subtle power of -detecting recondite resemblances, boldness in assuming essential identity of conditions on the strength thereof, and 1 Cp. Lotze's Logic, p. 317.