498 RHYTHMS IN PLANTS AND ANIMALS 



lator model we adduced initially for the peculiarities of light-induced 

 phase shifts. 



Some Qualifications: Phase Control versus Phase Shift 



Data already available on phase shifting in Drosophila, Euglena, 

 hamster, deer-mouse, and cockroach reveal complications whose full 

 discussion is postponed but which must be noted. 



Failure of a 1 2-hr light signal to give a saturation phase shift in the 

 mammals (like the hamster) is open to more than one interpretation, 

 including the likelihood that the B oscillations in this case feed back 

 strongly on A, thus conferring an inertia, so to speak, on the system 

 which a single signal cannot overcome. 



Some signals even in Drosophila and Euglena fail to give saturation 

 shifts (cf. Fig. 6, p. 97, in Pittendrigh and Bruce, 1957). There is 

 evidently some dependence of the extent of phase shift on the "magni- 

 tude" of the resetting signals; weaker signals (lower intensity, shorter 

 duration) may give smaller shifts. But there is already enough evidence 

 to dismiss a simple duration-intensity reciprocity. Signals shorter than 

 12 hr (8 hr, 4 hr, and 1/2000 sec) can all cause phase shifts if they 

 fall within the night (Biinning's scotophil) phase, but they apparently 

 fail to reset very much, or even at all, if they fall entirely in the photo- 

 phil phase. If this proves to be generally true, it would imply that the 

 efficacy of our 1 2-hr signals that begin in the photophil is due largely 

 or wholly to the tail end of the signal that extends into the scotophil. 

 The surprising fact that extremely short (1/2000 sec) high-intensity 

 flashes can reset the hamster (Burchard, 1957) and Drosophila (Bruce 

 and Pittendrigh, 1957a) rhythms makes possible a detailed analysis of 

 the resettability of the A oscillator in which the issue of time in the 

 cycle is unconfounded by duration of signal. 



Our provisional view that dawn (or any dark-light transition) is the 

 efficient phase-giver in Drosophila has strong support in spite of its 

 apparent contradiction by these indications of nonresettability in the 

 photophil. Figure 2 summarizes this support, but indicates, again, that 

 there are complications. The phase of the Drosophila eclosion rhythm 

 remains fixed relative to dawn (and independent of the light-dark 

 transition) over a wide range of photoperiods. Nevertheless this rule 

 breaks down when the photoperiod is reduced to 4 hr; the phase of the 



