496 RHYTHMS IN PLANTS AND ANIMALS 



cycle entrains the B oscillator. Only when the phase relations of light 

 and temperature are those found in nature (low point of the tempera- 

 ture curve near dawn) can we expect the B oscillator, and hence the 

 fly's overt rhythm, to exhibit its typical phase. Displacement of the 

 temperature curve relative to dawn should lead to conflicting entrain- 

 ment forces on the B oscillator: the light regime will pull in one direc- 

 tion via its control oi A, which in turn entrains B; and the temperature 

 regime will pull in the other direction via its direct entrainment of B. 

 Figure 1 1 shows that the phase of the fly rhythm is indeed some 

 compromise between the light and temperature cycles until the peak of 

 activity is moved to 15 hr after dawn; there is then a clear phase jump 

 over to the next dawn. (This phase jump from hour 15 to hour 24 is 

 a striking phenomenon we are not yet ready to discuss; it implies a 

 zone of forbidden phase relations between B and its driving oscillator 

 A that is evident in other unpublished data. ) 



The interpretation of these effects in terms of the coupled oscillator 

 scheme can be to some extent tested in the experiments summarized 

 by Fig. 1 2 which involve replications of conditions applying to cultures 

 5,6,7, and 8 in Fig. 11. These are the cultures subjected to the most 

 drastic disturbance of normal light and temperature phase relations. 

 In Fig, 12A the system, initially in steady state, dictated by the con- 

 flicting light and temperature signals, is transferred to constant dark 

 and a steady 21°C. Within 3 days all the cultures revert to a phase 

 corresponding with that of the previous light cycle, thus confirming 

 the conclusion that A drives B with very little feedback. Figure 12B, 

 however, demonstrates that some feedback is surely present. When the 

 cultures are released from control by the external light cycle, but main- 

 tained under the temperature cycle, they all assume a phase determined 

 wholly by the temperature. The phase of B, still being forced by the 

 temperature oscillation, ultimately entrains A , which is no longer held 

 by a light cycle. It is unnecessary to emphasize that the experiments 

 summarized in Figs. 10, 11, and 12 reveal strong details we have not 

 attempted to explain. It is simply our current opinion that the com- 

 plex interaction of light and temperature in Drosophila and mixture of 

 temperature-dependent and temperature-independent features are all 

 best approached and will ultimately be explained by the coupled oscil- 



