The operations of the intellect may furnish us with an illustration. Ideas are symbols of sensations. The idea of a horse is an abstraction easily traceable to concrete sensations, yet as an abstraction is so different a state of feeling that we only identify it with its concretes by a careful study of its stages of evolution, namely, sensation, image, reproduced images resembling yet differing from the original sensation, a coalescence of their resemblances, and finally the substitution of a verbal symbol for these images. With this symbol the intellect operates, and sometimes operates so exclusively with it that not the faintest trace of image or sensation is appreciable--the word horse takes the place of the image in the sequence of sensorial processes, just as the image takes the place of the sensation. It does this as a neural equivalent. In the same way we substitute verbal symbols for a bag of sovereigns when we pay a creditor with a check; he pays the check away to another; and this monetary equivalent passes from hand to hand without a single coin making its appearance. Does the transaction cease to be commercial, monetary, in this substitution of signs? No; nor does a process cease to be psychical when an image is substituted for a sensation, and a verbal symbol for an image. This every one will admit. Must we not go further, and extend the admission to automatic actions which originally were voluntary, and have now lost all trace of ideal prevision, and almost all traces of accompanying consciousness? _The motor mechanism has its symbols also_; in this sense, that whereas the action which at first needed complex sensorial processes to set it going and keep it going, is now determined by a single one of those processes taking the place of their resultant. When a practised accountant runs his eye up a column of figures, he does not pause to realize the values of those figures by decomposing the symbols into their numerical units, he simply groups one symbol with another according to their intuited relations, and the final result is reached with a certainty not less, and a rapidity far greater, than if it had been reached by step-by-step verification. It is thus with the pianoforte-player. It is thus with all automatic performances, except those dependent on the connate adjustments of the mechanism. CHAPTER VI. THE PROBLEM STATED.