toward it or opinion concerning it. It is 'hypothetical,' as Kant would say, but valid. Moral laws define acts as vicious or virtuous according to their bearing on the field of interests which they affect. Typical moral experiences contain both a judicial and an affective exponent. This complexity is reduced to a pseudo-simplicity by such terms as 'appreciation' and 'evaluation,' or by such notions as Westermarck's 'emotions of approval.' But this is simply to exploit the equivocation which their dual nature makes possible. Most moralists are now agreed that goodness is relative to desire. This does not mean that the good is what is judged to be good. Value lies in the desire relation, which is prior to all judgment about it. There is a logical or strictly ethical question concerning the meaning of moral obligation which cannot be answered by any account of the genesis, history, or psychological structure of the sentiment. And it is impossible to make a rational appeal to anything but reason. Arguments do not compel feeling or action. An agent is always free to ignore the truth.
Two problems arise in connection with association by similarity: (1) May similarity be reduced to contiguity or vice versa? (2) Is similarity an associative force? Peter's recent results show that similarity is an associative force and that there are individual differences in the degree to which it is exercised. The present study is based on two sets of experiments. In the first, numbers were used, these being arranged in couplets according to four types, of which three showed similarity. There were eight couplets under each type and these were distributed over four series, two to each series in all possible orders. The method consisted of presenting successively to the observers the couplets of a single series. The degree of the impression was then tested; the first members of the series were given in succession and the observers attempted to recall the second members. This was repeated until the entire series had been impressed. In the second set of experiments nonsense words were used. Here, the method was similar to that of the first set; but composite and homogeneous series were used and the exposures regulated by a special apparatus. The experiments show (1) that resemblance between thought has no associative value; (2) that contiguity alone forms associative connections; (3) that similarity can act only in a secondary and indirect manner. There is the possible objection that association by similarity may be brought about through physiological intermediaries. The reply would be that we are here dealing with the appearance of images in consciousness, and that association is insufficient to explain such facts. Associative connections are only one of the many forces which determine the appearance of images.
