Leaving, then, this constant error in time-appreciation, we will pass to the variable and more palpable errors in the retrospective measurement of time.  Each person’s experience will have told him that in estimating the distance of a past event by a mere retrospective sense of duration, he is liable to extraordinary fluctuations of judgment.  Sometimes when the clock strikes we are surprised at the rapidity of the hour.  At other times the timepiece seems rather to have lagged behind its usual pace.  And what is true of a short interval is still more true of longer intervals, as months and years.  The understanding of these fluctuations will be promoted by our brief glance at the constant errors in retrospective time-appreciation.

And here it is necessary to distinguish between the sense of duration which we have during any period, and the retrospective sense which survives the period, for these do not necessarily agree.  The former rests mainly on our prospective sense of time, whereas the latter must be altogether retrospective.[121]

Our estimate of time as it passes is commonly said to depend on the amount of consciousness which we are giving to the fact of its transition.  Thus, when the mind is unoccupied and suffering from ennui, we feel time to move sluggishly.  On the other hand, interesting employment, by diverting the thoughts from time, makes it appear to move at a more rapid pace.  This fact is shown in the common expressions which we employ, such as “to kill time,” and the German Langweile.  Similarly, it is said that when we are eagerly anticipating an event, as the arrival of a friend, the mere fact of dwelling on the interval makes it appear to swell out.[122]

This view is correct in the main, and is seen, indeed, to follow from the great psychological principle that what we attend to exists for us more, has more reality, and so naturally seems greater than what we do not attend to.  At the same time, this principle must be supplemented by another consideration.  Suppose that I am very desirous that time should not pass quickly.  If, for example, I am enjoying myself or indulging in idleness, and know that I have to be off to keep a not very agreeable engagement in a quarter of an hour, time will seem to pass too rapidly; and this not because my thoughts are diverted from the fact of its transition, for, on the contrary, they are reverting to it more than they usually do, but because my wish to lengthen the interval leads me to represent the unwelcome moment as further off than it actually is, in other words, to construct an ideal representation of the period in contrast with which the real duration looks miserably short.