THE RHYTHMIC NATURE OF LIFE' 301 



been stretched out to a cycle which was 27 hours and 20 minutes in length 

 and the lunar day stretched to 28 hours and 10 minutes. The traveling 

 crabs had ignored these long cycles of all external factors and had, non- 

 chalantly, measured ofif normal 24-hour and 24.8-hour periods. There was 

 obviously, therefore, an internal clock which in darkness possessed con- 

 siderable precision. 



The crabs in California showed no tendency to shift their color-change 

 clock to California time during the six days they were watched. Using 

 their internal clock they had apparently shifted the time relationships of the 

 color-change clock to the metabolic clock and the new relationship was 

 permanent. Had the crabs been permitted to experience two or three 

 normal light cycles in California, they would undoubtedly have reset their 

 color-change clocks to the sun time of their new longitude. 



Renner (1955) has more recently demonstrated the existence of an 

 autonomous inner clock in bees. Bees were trained in Paris, France to 

 come to a feeding station at a particular time of day. Then they were sealed 

 and rapidly carried to New York. There, when the container w'as opened 

 in a room w^hich w-as a duplicate room to the Paris training room, they 

 continued to feed on their original Paris time. 



It seems highly likely in the present state of our knowledge that both 

 internal and the imposed daily and lunar clocks are necessary to provide 

 animals and plants with such precisely timed overt rhythms as those ob- 

 served for color-change, feeding, running activity, and reproduction. The 

 imposed clock would clearly be fully temperature-independent and would 

 account for the precision of the frequency of the cycles over long periods 

 in constant darkness at various temperatures. On the other hand, the im- 

 posed clock will not produce an overt rhythm in which a specific thing is 

 done by the animals or plant at, for example, every low tide or each morn- 

 ing. Variations in barometric pressure and all forces correlated with it 

 are too randomly variable from hour to hour or day to day to assure this. 

 Only on the average over a few days can these last be integrated into daily 

 cycles of the order of sharpness of the day-night light changes. On the 

 other hand, the internal clock must be a metabolic one and, therefore, 

 probably only more or less accurate and more or less temperature-inde- 

 pendent. 



Evidence being obtained currently in our laboratory is clearly indicating 

 that the organism is, in fact, integrating three to five days of fluctuation in 

 some still unknown external factor whose intensity is correlated to some 

 degree with barometric pressure, and using this product of integration 

 in determining the form of its own metabolic cycles. Such a mechanism, 

 the detailed operation of which is still far from known, provides the organ- 

 ism with an efifective pacemaking stimulus to assure the long-term pre- 



