8. APPLICATIONS AND PREDICTIONS 137 



to about 7 h. If the subharmonic is still of order i, then the clock will now show 

 a period of 28 h. At this point the subharmonic of order ^ may become un- 

 stable, and the system may suddenly shift to one of order -J. The clock period 

 will then decrease to 21 h. Then as continues to increase under the pulsed 

 stimuli, rising finally to about 16, say, the fundamental oscillations will 

 increase in period to about 8 h, so that the subharmonic or order -^ will once 

 again generate a daily rhythm with a period of about 24 h. 



The behaviour of such an idealized system can hardly be expected to provide 

 an adequate quantitative prediction of the possible response of a real circadian 

 system to the procedure outlined above. It does serve, however, to suggest the 

 major features of the response which is predicted by our theory, providing of 

 course that changes in 6 are reflected by changes in the clock without too many 

 complicating factors. Obviously one of the difficulties with the theory is that 

 its major macroscopic parameter, d, is not directly observable by any obvious 

 means short of measuring directly the actual sizes of the oscillations in certain 

 macromolecular species in living cells by some very refined microspectro- 

 photometric or similar optical technique. However, the accessibility of 

 circadian rhythms to observation immediately suggests them as a key to the 

 analysis of G-levels and to the whole structure of the present theory. It is to 

 be hoped that a close enough causal connection exists between the sizes of 

 the fundamental generating oscillations and circadian and other rhythmic 

 behaviour to allow the one to serve as observables for the other. Such an 

 assumption seems at present to be a reasonable one. 



The experimental procedure outlined above could also be turned upside 

 down, so to speak, so that it is designed to cause 6 to decrease rather than 

 increase. One way of doing this might be to pulse an inhibitor of protein 

 synthesis, such as puromycin, into a stationary (or near-stationary), rhythmic 

 culture of cells periodically so that the cells are exposed to a small concen- 

 tration of the antibiotic for, say 20 min every 2 h. The antibiotic would have 

 to be washed out of the cultures after the 20-min exposure, since it is not 

 degraded. Karakashian and Hastings (1962) have shown that puromycin will 

 rapidly eliminate the luminescence rhythm in Gonyaulax, although it is not 

 established that the effect is solely on protein synthesis. The effect of each pulse 

 should then be to slow up protein synthesis briefly, thus tending to draw the 

 trajectories closer to the steady state axis and thus decrease 6. The prediction is 

 that such a treatment, continued over a day or two, should lead to a speeding 

 up of the clock. Once again it is rather a crucial question whether or not such a 

 change of clock period, if it occurs, is stable in the sense that the clock continues 

 to run faster after the experimental treatment ceases. If the period immediately 

 returns to its original value when the pulsing stops, then it would indicate that 

 there are no stable ^-levels other than the natural one and the concept of 

 talandic energy in analogy with physical energy could not be sustained. 



A similar treatment with actinomycin should have the same effect, this 

 time the primary variable influenced being mRNA. Transiently reduced 

 mRNA levels should lower protein levels also, so that we might anticipate a 

 stronger effect from actinomycin than from puromycin. Again the work of 



