Abstract: The 24-hr rhythms of behavior and other physiological functions are based upon an endogenous self-sustained oscillation. This is proven by experiments in which animals, kept in constant conditions, show a freerunning circadian period which deviates to a certain degree from that of the earth's rotation. The 24-hr rhythms of activity are furthermore characterized by typical patterns of which the one with two peaks in one period is most common. Under natural conditions, these two peaks of activity are often closely related to specific stimuli in the environment; e.g. to dawn and dusk. They have therefore been described as being directly caused by the environment. However, one may consider not only the circadian period but also the pattern part of an endogenous system. This hypothesis has been tested in experiments with finches. The results are: 1) Birds kept in artificial light-dark cycles show the same two-peak pattern either in cycles with interposed twilight or in cycles with rectangular changes between light and dark, 2) Under constant conditions, the second peak may disappear; but more often, it either remains or becomes more pronounced tan under conditions of a light-dark cycle. 3) With different levels of constant intensity of illumination, the birds show different circadian periods and different durations of activity; under those conditions, the two peaks keep the same proportional relationship to the activity time regardless of its duration. It is concluded that the basic two-peak pattern of locomotor activity is a persistent property of the circadian oscillating system.