THE RHYTHMIC NATURE OF LIFE 291 



train them to come to a feeding station at the same time on successive days 

 but could not train them to come at different times on successive days. 

 This suggested that the bee was being trained to migrate to the feeding 

 station at some fixed time in a dail}' cycle. 



In the human, the ability that many have to awaken at the same time 

 morning after morning, the possession of such diurnally rhythmic changes 

 as those of body temperature, blood sugar, and eosinophile count, and the 

 precision of the monthly reproductive cycle lead one to believe that even 

 man possesses a physiological clock and calendar. 



The foregoing are merely a few examples of ways in which animals 

 are normally dependent upon an ability to determine rather precisely the 

 correct moment to carry out some activity. 



The question which now occurs to us is the following: Are these ac- 

 tivities set off in direct response to cyclically varying factors or signals in 

 the external physical environment, or are there quite accurate clocks and 

 calendars in all of these organisms which are capable of causing these 

 critical events to occur just when they do, without any external cues re- 

 quired? If internal clocks are involved, the organisms would be able ac- 

 tually to anticipate and prepare for the occurrence of the environmental 

 cyclic events ; if clocks are not involved, there can be no such anticipation. 



The first experimental evidence that animals possessed some means of 

 measuring off 24-hour periods independently of the daily changes in light 

 and temperature came from the work of the Polish biologist, Szymanski 

 (1918). He found that 24-hour activity patterns continued to be exhibited 

 and synchronized with the day-night cycle even when animals were kept 

 in constant darkness and temperature. During the succeeding twenty or 

 thirty years, through the work of numerous investigators, it became evident 

 that persistent daily rhythmicity was very widespread among animals and 

 plants (Welsh, 1938; Park, 1940; Caspers, 1951), being reported for ani- 

 mals ranging all the way from the simple one-celled protozoans, on the 

 one hand, to the mammals including man, on the other. Among the plants, 

 examples ranged from bacteria to flowering plants. 



My own interest in the time-measuring mechanism responsible for 

 these persistent daily cycles was aroused some years ago because of the 

 obvious implication that such a clock, if there is one, must be relatively 

 independent of temperature. Investigators appeared always to find the 

 cycles to be rather precisely 24 hours in length, irrespective of the tempera- 

 ture at which the animals were kept. Furthermore, it was easily reasoned 

 that, if a clock within the organism was really to be of any use, it must have 

 the frequency of its daily cycles independent of temperature over the range 

 to which the animals were normally subjected, in order to permit them to 

 adapt their activities to the external daily cycle. And yet, in terms of con- 



