NOON 6P# MIDNIGHT 6 AM- NOON 
SOCIALLY HIGH RANGING RATS 
SOCIALLY LOW RANKING RATS 
— MEAN RHYTHM 
Figure 96. — Hypothetical origin of arrhythmicity of colony. By the time the colony approached a carrying capacity density, 
all periods of the night were nearly equally utilized, and in addition there was a considerable spread of activity over 
into the daylight hours. Despite the overall appearance of nocturnal arrhythmicity, the socially higher ranking rats 
retained the biologically preferable nocturnal bimodality. On the other hand the socially lower ranking rats were 
forced to be most active before and after these two peaks. 
reveals the probable means through which space- 
time utilization became equalized during the night 
hours and how some of the activity spreads out into 
the lighter portion of the 24 hours. 
Since most of the activity always occurred be- 
tween the hours of 6 p.m. and 6 a.m. the variance 
about the mean half-hourly activity should provide 
an index of the degree of social tension existing 
among the rats of the population. The greater the 
variance, the less will be the competition over space- 
time utilization of the source of food and water. 
The variance, [(g ] 
where 
N— Total number of half-hour periods between 
6 p.m. and 6 a.m. (i.e. 24). 
n=nth half hour 
X n = number of activity units during the nth half 
hour 
7'=total number activity units during the 6 p.m. 
to 6 a.m. period 
This formulation of a 2 is in terms of the percentage 
of activity for each half-hour period with respect 
to the total activity units during the 6 p.m. to 
6 a.m. period within which it falls. The a 2 was 
calculated for each of the 6 p.m. to 6 a.m. intervals 
of the 10 days which were utilized in each of the 
major Periods 1 to 10 (fig. 81 to 90) analyzed for ac- 
tivity periodism, and the results are shown in figure 
97. The range in the variance from day to day is 
considerable, even within a single period of 10 days. 
Nevertheless, the trend, represented by the medians 
for each Period, follows a course consistent with 
the other lines of evidence. It will be noted that 
there is an overall trend of decrease in variance 
through time superimposed upon midwinter in- 
creases in variance when there is reduced sexual 
activity. If the variance in Periods 8, 9, and 10 
are compared with the variance in the comparable 
Periods 3, 4, and 5 to 7 respectively, it will be found 
that the variance (degree of rhythmicity) for the 
1948 periods is 1.7, 3.0, and 4.4 times that of the 
similar periods during 1949. If, as we suspect, the 
degree of stress is inversely proportional to the 
variance in activity, it is quite understandable that 
the observed decline in growth rate and reproduc- 
tive rate, which is documented later on in the 
paper, should have arisen. Furthermore, it is in- 
teresting that the difference in variance becomes 
125 
