Once more we may contemplate this dominant idea of likeness as exhibited in art.  All art, civilised as well as savage, consists almost wholly in the making of objects like other objects; either as found in Nature, or as produced by previous art.  If we trace back the varied art-products now existing, we find that at each stage the divergence from previous patterns is but small when compared with the agreement; and in the earliest art the persistency of imitation is yet more conspicuous.  The old forms and ornaments and symbols were held sacred, and perpetually copied.  Indeed, the strong imitative tendency notoriously displayed by the lowest human races, ensures among them a constant reproducing of likeness of things, forms, signs, sounds, actions, and whatever else is imitable; and we may even suspect that this aboriginal peculiarity is in some way connected with the culture and development of this general conception, which we have found so deep and widespread in its applications.

And now let us go on to consider how, by a further unfolding of this same fundamental notion, there is a gradual formation of the first germs of science.  This idea of likeness which underlies classification, nomenclature, language spoken and written, reasoning, and art; and which plays so important a part because all acts of intelligence are made possible only by distinguishing among surrounding things, or grouping them into like and unlike;—­this idea we shall find to be the one of which science is the especial product.  Already during the stage we have been describing, there has existed qualitative prevision in respect to the commoner phenomena with which savage life is familiar; and we have now to inquire how the elements of quantitative prevision are evolved.  We shall find that they originate by the perfecting of this same idea of likeness; that they have their rise in that conception of complete likeness which, as we have seen, necessarily results from the continued process of classification.

For when the process of classification has been carried as far as it is possible for the uncivilised to carry it—­when the animal kingdom has been grouped not merely into quadrupeds, birds, fishes, and insects, but each of these divided into kinds—­when there come to be sub-classes, in each of which the members differ only as individuals, and not specifically; it is clear that there must occur a frequent observation of objects which differ so little as to be indistinguishable.  Among several creatures which the savage has killed and carried home, it must often happen that some one, which he wished to identify, is so exactly like another that he cannot tell which is which.  Thus, then, there originates the notion of equality.  The things which among ourselves are called equal—­whether lines, angles, weights, temperatures, sounds or colours—­are things which produce in us sensations that cannot be distinguished from each other.  It is true we now