Bain may be cited as one of the principal representatives of this method. I will not insist on a piece of work unworthy of such a psychologist; yet he has done it twice over, without arriving at an agreement with himself. His earlier classification gives as fundamental the emotion of relativity (surprise, astonishment), terror, tenderness, self-esteem, anger, the sense of power, of activity, of mental exercise, æsthetic emotion, moral emotion. The later includes eleven groups: love, anger, fear, the sentiment of property, the pleasure of power and its correlative pain of subjection, pride, vanity, activity (“plot-interest”), knowledge (the intellectual feeling), æsthetic emotion (beauty), moral sentiment. Three of these are “simple”—anger, love, and fear; but we find, a little later on, that love and anger are called “the giants of the group, the commanding and indispensable members of the emotional scheme;” so that fear would seem to be eliminated. The incoherence and inconsistency of this attempt are sufficiently obvious, and I need not insist on them. (It should be noted that, in both cases, the religious sentiment is omitted.) I can find only one valuable remark—viz., that “pleasures and pains are contained in every one of the classes to be described, just as the natural orders of plants may each contain food and poison, sweet aromas and nauseating stinks.”[92] I have only referred to this classification in order to show how, by its very nature, it is condemned to failure. Floating at haphazard, without fixed principle, when not contradictory, it can only be arbitrary. Footnote 92: Bain, _The Emotions and the Will_, p. 76. Herbert Spencer has criticised it in a well-known passage, which I will briefly recapitulate, since it serves as a transition to the second form of classification, and throws some light on the latter.[93] Bain has overlooked the fact that in confining his attention to the most obvious characteristics of the emotions, he is following the method of the ancient naturalists, who classed the cetacea among fishes. Every classification should be preceded by a rigorous analysis. For this purpose it would be necessary, as a preliminary, to study the ascending evolution of the emotions through the animal kingdom, to find out which of them are the first to appear, coexisting with the lowest forms of organisation and intelligence, and to note the existing differences, as regards emotion, between the higher and lower human races. Those common to all may be considered as simple, and those peculiar to the civilised races as ulterior and derivative. ----- Footnote 93: H. Spencer, _Essays_, vol. i. (Library Ed., 1891), pp. 241-264. -----