The reason is plain; the impression was very unfamiliar, and affected the beholder, both as regards the loss of one who was dear to him, and as reminding him with more than common force that he will one day die himself. Moreover the impression was a simple one, not involving much subordinate detail; we have in this case, therefore, an example of the most lasting kind of impression that can be made by a single unrepeated event. But if we examine ourselves closely, we shall find that after a lapse of years we do not remember as much as we think we do, even in such a case as this; and that beyond the incidents above mentioned, and the expression upon the face of the dead person, we remember little of what we can so consciously and vividly recall. II. Deep impressions are also made by the repetition, more or less often, of a feeble impression which, if unrepeated, would have soon passed out of our minds. We observe, therefore, that we remember best what we have done least often—any unfamiliar deviation, that is to say, from our ordinary method of procedure—and what we have done most often, with which, therefore, we are most familiar; our memory being mainly affected by the force of novelty and the force of routine—the most unfamiliar, and the most familiar, incidents or objects. But we remember impressions which have been made upon us by force of routine, in a very different way to that in which we remember a single deep impression. As regards this second class, which comprises far the most numerous and important of the impressions with which our memory is stored, it is often only by the fact of our performance itself that we are able to recognise or show to others that we remember at all. We often do not remember how, or when, or where we acquired our knowledge. All we remember is, that we did learn, and that at one time and another we have done this or that very often. As regards this second class of impressions we may observe:— 1. That as a general rule we remember only the individual features of the last few repetitions of the act—if, indeed, we remember this much. The influence of preceding ones is to be found only in the general average of the procedure, which is modified by them, but unconsciously to ourselves. Take, for example, some celebrated singer, or pianoforte player, who has sung the same air, or performed the same sonata several hundreds or, it may be, thousands of times: of the details of individual performances, he can probably call to mind none but those of the last few days, yet there can be no question that his present performance is affected by, and modified by, all his previous ones; the care he has bestowed on these being the secret of his present proficiency.