DAILY RHYTHMS 485 



new steady state is reached. We call these lengthened or shortened 

 intervals transients. 



Some Generalizations about Phase Stiifting 



In earlier papers (Pittendrigh and Bruce, 1957; Pittendrigh, 1958) 

 full descriptions have been given of the phase-shift experiments in 

 Drosophila that have stimulated the present study. Figure 4 gives histo- 

 grams illustrating the effects on the eclosion rhythms of a 12-hr light 

 signal applied to 23 cultures of Drosophila pseudoobscura. Each cul- 

 ture receives a light signal beginning at the time marked by the line di. 

 Three features are outstanding. 



1. After 72 hr, at line ds, all cultures have achieved a phase shift 

 that is virtually complete; that is, the phase of the overt rhythm now 

 bears the same relation to the di, do, and ds coordinates as it originally 

 bore to those marked Do, Di, Dj • ■ • Ds. It is clear that the particular 

 phase ultimately attained by each culture separately within the interval 

 D4-D.-, is, in some sense, fully determined by the single signal given 

 three cycles earlier. 



2. Attainment of new phase is not instantaneous; it proceeds gradu- 

 ally as the system moves through several transient cycles in which the 

 "period" differs from the natural period of the steady-state system, 

 which in this case is indistinguishably close to 24 hr and exemplified 

 by culture in Fig. 4. 



3. The character of the transients is dependent on the time in the 

 cycle when the phase-shifting signal is given. Signals beginning within 

 the first 1 5 hr of the cycle generate transients whose "period" is greater 

 than the natural period; those falling in the last 9 hr of the cycle gener- 

 ate transients with "periods" less than the natural period. This is well 

 shown by the behavior of cultures 8 and 18, Fig. 5. In cultures 12, 13, 



Fig. 3. Typical data of diurnal (Eutamias) and nocturnal {Zapiis and 

 Peromyscus) mammals. Circled points show the time of onset of running- 

 wheel activity on successive days, at first in LD (light-dark) conditions and 

 later in DD (constant dark) conditions. (Left) A typical long period in DD 

 for the diurnal chipmunk {Eutamias) and short period for the nocturnal jump- 

 ing-mouse (Zapus). (Right) Data for eight nocturnal deer mice (Peromyscus) 

 numbered 2, 8, 15 ... , etc. The time scale of each individual mouse is dis- 

 placed 2 hr to the right; hr-20 is indicated for each mouse. Nocturnal mammals 

 in general have periods less than 24 hr in DD, but it is evident that there are 

 individual exceptions. 



