328 CHAELES MERCIER: This group is then a higher group than any of the preced- ing, and includes feelings of a different class, excluding of course the feelings belonging to the groups below. But the description that Mr. Spencer gives of the constitution of these feelings is exactly applicable to some at least of the feelings in the inferior classes. To guard against any possi- bility of misrepresenting Mr. Spencer, I will take his own example of the Preservative-representative class the emotion of Terror and ask if his description of a Ke-representative emotion is not strictly applicable to it. When a child enters a dark room alone, is it not a fact that the terror which it feels is " awakened not by the presence of any special object but by terrifying objects at large " ? Is it not true that "it is not from the mere presence of such objects, but from a certain ideal relation [of accessibility] to them that " the terror arises ? Does not the terror consist, not of the represented injury inflicted by this or that, but of the represented feelings of injury in general ; is it not made up, not of certain concrete represen- tations, but of the abstracts of many concrete representa- tions ; and is it not so re-representative ? If this is a true description of the constitution of the feeling of terror, and it is difficult to perceive a flaw in it, what becomes of the distinction drawn by Mr. Spencer between feelings of the Preseritative-representative class, and those which are Re- representative ? And if, on the other hand, it is not a correct description, yet it must be acknowledged to be so close an approximation, that a classification whose primary groups are divided by such subtle differences is for practical purposes unworkable. While, therefore, as already admitted, the principle of representation, in common with the quality of pleasure or pain, does no doubt indicate an actual element which pervades every class of feeling and varies in each, yet the classification founded on it fails at every point at which it is brought to the test ; and this collapse of the classification is sufficient proof of the inadequacy of the principle on which it is founded. The impracticable character' of Mr. Spencer's classification of feelings seems to me, however, to be the most cogent evidence of the truth and of the value of his system of psychology ; for it demonstrates that when that system is departed from, not even Mr. Spencer himself can succeed in constructing a stable edifice on any other foundation. The classification that he has raised with such care is built upon sand, and falls to pieces before the first gust of criticism. It is true that his classification of feelings, unlike his classi-
