Ebbinghaus maintains, correctly enough, that not one of these explanations is universally satisfactory, but it must be granted that now one, now another is useful in controlling this or that particular case. The processes of the destruction of an idea, may be as various as those of the destruction and restoration of a building. If a building is destroyed by fire, I certainly can not explain the image given by merely assuming that it was the victim of the hunger of time. A building which has suffered because of the sinking of the earth I shall have to image by quite other means than those I would use if it had been destroyed by water. For the same reason when, in court, somebody asserts a sudden “occurrence,” or when we want to help him and something occurs to him, we shall have to proceed in different fashion and determine our action empirically by the conditions of the moment. We shall have to go back, with the help of the witness, to the beginning of the appearance of the idea in question and study its development as far as the material permits us. In a similar manner we must make use of every possibility of explanation when we are studying the disappearance of ideas. At one point or another we shall find certain connections. One chief mistake in such reconstructive work lies in overlooking the fact that no individual is merely passive when he receives sensations; he is bound to make use of a certain degree of activity. Locke and Bonnet have already mentioned this fact, and anybody may verify it by comparing his experiments of trying to avoid seeing or hearing, and trying actively to see or to hear. For this reason it is foolish to ask anybody how it happened that he perceived less than another, because both have equally good senses and were able to perceive as much. On the other hand, the grade of activity each has made use of in perception is rarely inquired into, and this is the more unfortunate because memory is often proportionate to activity. If, then, we are to explain how various statements concerning contemporaneous matters, observed a long time ago, are to be combined, it will not be enough to compare the memory, sensory acuteness, and intelligence of the witnesses. The chief point of attention should be the activity which has been put in motion during the sense-perception in question. Section 53. (b) The Forms of Reproduction. Kant analyzes memory: 1. As apprehending something in memory. 2. As retaining it for a long time. 3. As immediately recalling it.