DAILY RHYTHMS 503 



24-hr temperature cycle obviates the damage otherwise induced by 

 constant Hght is strong evidence for the interpretation offered here. In 

 terms of the coupled-oscillator model synchrony is achieved in Hill- 

 man's case through direct temperature entrainment of the cells' B 

 oscillations. Went (1957) has already suggested that the beneficial 

 effect of a thermoperiod on a growing shoot is due to its synchroniza- 

 tion of cell division which he considers essential for normal organo- 

 genesis. We would only add that the need for synchronization probably 

 goes well beyond the geometrical consequences of cell division to the 

 coordination of the whole biochemical activity of adjacent cells. 



The coupled oscillator model appears to offer a unique explanation 

 of Went's (1957; and this symposium) description of the interactive 

 effects of light and temperature on the African violet. These plants die 

 when grown on a 24-hr light cycle at 10°C but not 23 °C; on the 

 other hand they grow well at 10°C when the light cycle is lengthened 

 to a period of 32 hr. This implies to us that at 10°C the period of the 

 temperature-dependent B oscillations have lengthened to such an ex- 

 tent that they are out of the range within which the A oscillator (itself 

 entrained to 24 hr by the light cycle) can entrain and thus synchro- 

 nize them. When, however, the A oscillator is entrained by the light 

 regime to a 32-hr period, it is able, once again, to entrain the long- 

 period B oscillations and thus establish the essential synchrony of cell 

 processes. In discussing the entrainment relations of A and B it is 

 pertinent to note the peculiarities of temperature independence de- 

 scribed by Bruce and Pittendrigh (1956) for the Euglena phototaxis 

 rhythm. Temperature independence is good between 30 °C and 18°C, 

 but it breaks down quite sharply at lower temperatures. This relatively 

 abrupt transition to temperature dependence in the overt rhythm sug- 

 gests that the B oscillation's free-running period had extended beyond 

 the range where the A oscillator (with a 23% -hr period) could hold 

 it by entrainment. 



Damage induced by cycles with a period too far from 24 hr can 

 also be interpreted as due to failure of entrainment and, hence, cell 

 asynchrony. But an intriguing possibility, also suggested by the oscilla- 

 tor model, is not excluded, and merits the experimental tests to which 

 it is open. It seems plausible to us that the efficiency of the cell, the 

 overall activity of which pursues an oscillatory course, should fall when 



