578 PHOTOPERIODISM IN INVERTEBRATES 



might well have been expected, since an endogenous rhythm quite 

 evidently has the inherent property of a natural period. Light and dark 

 periods serve, in essence, only to give information concerning phase. 

 It is therefore logical that a phase shift may be accomplished by a 

 treatment which is not accompanied by any information concerning 

 period. 



The Gonyaulax clock illustrates this principle well. It was mentioned 

 previously that with aperiodic cultures that have been grown in bright 

 light, a change in the light intensity (to dim light or dark) not only in- 

 itiates overt rhythmicity, but also determines phase. It is also apparent 

 that in the experiments shown in Fig. 6, repetitive resetting is achieved 

 by the alternating light and dark periods, and that the phase of the 

 endogenous rhythm under the subsequent constant conditions is deter- 

 mined by the last dark to light transition. 



Resetting of the Gonyaulax clock by a single perturbation has been 

 found to occur irrespective of the particular way in which the experi- 

 ment is performed. For example, the phase of the luminescent rhythm 

 in cuhures in constant dim light (which exhibit rhythmicity similar to 

 that shown in Fig. 2b) may be shifted either by an exposure to bright 

 light or by a sojourn in darkness. Alternatively, if cultures otherwise in 

 constant darkness are given an appropriate exposure to light, a phase 

 shift occurs. Some results of experiments of the latter type are shown 

 in Fig. 7. As pointed out by Pittendrigh (1958), the effectiveness of 

 phase shift in a model oscillator should be a function of the time during 

 the cycle when the exposure is made, the duration of the exposure, 

 and the intensity of the exposure. This has indeed been found to be 

 the case in Gonyaulax. 



Exposure to light, i.e., a perturbation, given within 6 hr of the maxi- 

 mum in luminescence is more effective in shifting the phase than a 

 perturbation during the other 12 hr of the cycle. Moreover the direc- 

 tion of the phase shift varies with the time of the cycle, as illustrated in 

 Fig. 7. A perturbation before the maximum in luminescence results in 

 a phase delay, whereas a similar treatment at or after the maximum 

 causes the phase to advance. Our results indicate that the initial phase 

 shift is a stable one, no transients being involved. This point must be 

 investigated with Gonyaulax in greater detail, however. 



A longer exposure to light is more effective in phase shifting than 



