Illusions of memory have been touched upon on p. 186. We may remember something that never happened; we may remember something that happened, but could not have happened to us; we make all kinds of mistakes in memory; we fail to remember a great deal that has happened. These chances of error are inherent in the laws of associative tendency, and in the character of the memory-image. There is one illusion, however, that requires a word of comment: the illusion of the ‘=good old days=,’ the tendency of every man past middle age to be _laudator temporis acti se puero_. This has often been referred to the principle that we remember pleasurable experiences better than unpleasurable; we are so constituted, it is said, that the disagreeable events of our past life are forgotten and the agreeable are conserved in memory. The principle, however, has never been established, and there is some experimental evidence against it. In all probability, _the illusion is due to many contributing factors_. First of all, our nervous system takes its general set in childhood; it is then that we acquire standards of right and wrong, of social position, of daily intercourse and occupation. In so far as later experiences interfere with this set, the old order will be preferred. Secondly, our self-centredness (p. 2) leads us to idealise our past self; we think of ourselves as more important, more heroic, more dominating, more regarded, not only than we were, but also than any youngster of our sort could possibly have been; autobiographies, however truthful in intention, bring out the point with sufficient clearness. So we contrast our present struggles with the triumphs of an unreal past. Thirdly, the old days were, in one sense, really happier for us than the new; happier because we had no responsibilities, because there was a generation of adults to whom we could appeal; and we are very prone to confuse our own greater comfort with a better status of society. These are obvious considerations, but they and things like them are enough to account for the illusion. § 42. =The Pattern of Memory.=—Psychology cannot yet offer any adequate description of the pattern that mental processes display, the arrangement that they fall into, when we are remembering. Memory, as we are all aware, may occur in the state of primary attention, when we call it remembrance, or in the state of secondary attention, when we call it recollection. Something may be said under both heads; but our account must be largely figurative and conventional.