1. The synchronizing effect of four sinusoidal temperature cycles of different amplitude on the circadian rhythm of locomotor activity in lizards (Lacerta sicula) was measured. The results were compared with the influence of forcing oscillations of different strength on single self-sustained oscillations in the technical sense. 2. The number of animals synchronized with the temperature cycle decreased with decreasing amplitude of this cycle (Table 2). Especially those animals were not entrained, in temperature cycles with smaller amplitude, whose free running period deviated strongest from that of the Zeitgeber (Fig. 4). This behaviour corresponds closely to that of self-sustained oscillations which can be entrained by forcing oscillations within limited ranges of entrainment only, the width of which depends on the strength of the external excitation. 3. The phase angle difference between circadian periodicity and temperature cycle depended on a) the free-running frequency of the circadian oscillation, and b) the amplitude of the synchronizing temperature cycle (Kg. 5). This behaviour coincides with that of technical coupled oscillators (in this case with unipolar coupling), in which phase angle difference depends upon a) the difference of natural frequencies, b) the strength of coupling. 4. The time necessary to synchronize the circadian rhythm after an inversion of the temperature oscillation increased with decreasing amplitude of the latter (Fig. 6). This corresponds to the dependence of transient time on strength of excitation in technical oscillations. 5. The direction of phase shift of the circadian oscillation after inversion of the Zeitgeber cycle depended on the free running period of the circadian rhythm (Fig. 7). 6. Phase angle difference before and after a phase jump of the Zeitgeber oscillation showed greater differences in temperature cycles with small amplitude (Fig. 8). These findings indicate that phase control is less rigid with weak Zeitgebers. 7. The results demonstrate that the circadian oscillation can normally be described as a single self-sustained oscillation (though a number of other findings has shown that it probably consists of a multitude of oscillators). Models based on such an assumption are therefore adequate to describe the gross behaviour of the circadian oscillation. It is stressed that the features of the circadian periodicity shown here (point 2 to 4) are valid for all single self-sustained oscillations under the influence of a forcing oscillation, regardless of their special mechanism of excitation. The findings can therefore not be considered as supporting a particular single oscillator model. © 1969 Springer-Verlag.
