394 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1959 
at the proper time for Paris. They were retrained in New York and 
flown back to Paris. In Paris they came to feed at the proper time 
for New York. This confirmed completely our crab experiments. 
These experiments, however, did not prove that a rhythm of some 
subtle external factor was not essential to the organismic 24-hour 
rhythm since there are certain important 24-hour rhythms in the 
environment that are on universal time. By universal time is meant 
that the rhythm affects all regions of the earth’s surface in the same 
ways at the same instants. Such a factor is the daily fluctuation in the 
300,000-volt potential difference which exists between the surface of 
the earth and one of the outer shells of our atmosphere, the ionosphere. 
By one of these universal-time cycles the 24-hour period could con- 
ceivably be imparted to the animals even while they are traveling. 
On the other hand, the living rhythmic system might be completely 
independent of all external cycles. 
We have seen thus far that the daily rhythms of living things in 
constant darkness, like the clocks in our homes, can measure accurately 
24-hour periods whether they are set at the normally correct time of 
day, or have been reset to register an incorrect time. As the mathe- 
matician or physicist would describe this, the rhythms in the living 
things need not have locked phase relations with any external physical 
rhythms. On the contrary, there is a freely labile relationship. This 
adjustability of the biological cycles is a very useful characteristic. 
Absence of this property would make the rhythms comparable to 
alarm clocks in which the time of day or the ringing of the alarm 
could not be altered. 
This brings us to another of the interesting properties of the 
rhythms of living things, namely, that many are known regularly 
to gain or lose a few minutes every day. Some are even known which 
regularly gain or lose two or three hours a day. Such a one as the 
last is the spontaneous activity of the field mouse, or the white rat. 
Both of these animals, when kept in constant darkness, may exhibit 
an accurate 24-hour cycle of running with about 12-hour periods of ac- 
tivity alternating with about 12-hour periods of rest, but when kept 
in an unchanging low illumination have their daily cycles of running 
longer than 24 hours. In the rat we found them to be quite regular 
and 2514 hours in length. In figure 3, starting at the top, the black- 
ened areas indicate the times of day of running during 70 days in 
constant dim light. To begin with, the rat, like any well-adjusted 
rat, ran only during the early morning and the evening hours. Since 
the rat started to run 75 minutes later every day, the daily period of 
running was occurring in the daytime in less than 2 weeks. The 
period of running can be seen to scan the day more than three times 
during the 70-day period from November 13 through January 20. 
