DAILY RHYTHMS 499 



rhythm now shifts back — activity anticipates dawn. Aschoff and 

 Meyer-Lohmann (1955) and Ubelmesser (1954) have published 

 similar data demonstrating some dependence of phase on photoperiod 

 in mammals and the fungus Piloboliis. Finally, it is noted that con- 

 trary to strong expectations based on Fig. 2, when periodic cultures in 

 constant light are transferred to constant dark, the single light-dark 

 transition completely resets the system; after two transient cycles a new 

 steady-state phase is established with eclosion peaks occurring at n X 

 24 + 12 hr after the single light-dark step that is evidently taken, in 

 some sense, as "sunset." 



Obviously the problem of phase control in LD conditions is still far 

 from understood. This does not, however, preclude a meaningful dis- 

 cussion of phase shifts imposed on the free-running (LL or DD condi- 

 tions) rhythm by discrete light and temperature signals. It is the 

 strongly characterized pattern of such phase shifts that leads to the 

 coupled oscillator model; the further development of this model will 

 have to accommodate the problems of phase control. 



THE CELL AS OSCILLATOR; PHOTOPERIODISM, 

 THERMOPERIODISM, AND DAILY RHYTHMS 



There is no way at present of further characterizing the A and B 

 oscillators nor of localizing them to cellular parts. The coupled-oscUla- 

 tor model attributes to the A component nearly all those major clock 

 features that in the single-oscillator scheme (Fig. 1) were considered 

 inherent in what we called ESSO; the A oscillator must be self- 

 sustaining, temperature-independent, and receptive to light signals. 

 The coupling of the whole rhythmic system to temperature cycles is 

 now considered a function of the distinct B component. 



Clearly, as long as it is driven by A, we have no way of knowing 

 whether B is truly self-sustaining. It seems unHkely that the B oscilla- 

 tion will prove a single, discrete, and identical entity in the fly and the 

 hamster; or for that matter, that there should be only one B oscillator 

 even in a single organism in which several persistent rhythms run side 

 by side. A more plausible guess is that much of the whole cell's activity 

 is innately rhythmic with a period — surely both temperature-sensitive 

 and temperature-dependent — that approximates 24 hr in the optimal 



