This chapter was first published in the _Revue Philosophique_ for October 1894. It called forth some new communications, two only of which have been added to the original text. The affirmation of a type of affective memory has, as I expected, provoked both criticism and denial. My principal opponent, Prof. Titchener, has published on this subject a somewhat extensive article in the _Philosophical Review_ (November 1895), in which he reproaches me with not having cited a single case of pure emotional memory—i.e., memory from which all sensory and ideational elements are absent, and where there is a revival of feeling _as such_. An example of this kind, which should be quite conclusive, seems to me almost impossible to produce. A pleasure, a pain, an emotion, are always associated with a sensation, a representation, or an act; revival necessarily bringing back the intellectual state which forms part of the complexus and supports it. But the real question is elsewhere: Is revival, in certain persons at least, a dry record, or a _felt_ state? In this last case—and it does occur—there is the recollection of the emotional state as such. There is another objection: Can it be said that an emotion is the reproduction of an antecedent emotion, and not a new emotion? The reproduction of an emotion can itself be nothing other than an emotion, but it bears the marks of repetition. Without returning to what has been said above, I remark that those contemporary psychologists, who study with admirable patience the mechanism of memory, neglect that of its most general conditions. Now the chief of these is that every recollection must be a _reversion_, by virtue of which, the past once more becoming a present, we live at present in the past. The recollection of an emotion as such does not escape the action of this law; it must become actual once more—must _be_ a real emotion, whether acute or obtuse. Taking account of the criticisms, and of the new material supplied to me, I may once again sum up my inquiry thus— 1. The emotional memory is _nil_ in the majority of people. 2. In others there is a half intellectual, half emotional memory, _i.e._, the emotional elements are only revived partially and with difficulty, by the help of the intellectual states associated with them. 3. Others, and these the least numerous, have a true—_i.e._, complete—emotional memory; the intellectual element being only a means of revival which is rapidly effaced. ----- CHAPTER XII. THE FEELINGS AND THE ASSOCIATION OF IDEAS.