HEAT PRODUCTION OF HONEYBEES IN WINTER. 13 
cumulate in the air of the system in which there was a deficiency 
of oxygen. The effect on the activity of the bees was soon apparent; 
the temperature curves, which for some reason had begun to. rise, 
very shortly turned in the opposite direction and continued to fall 
for about 12 hours. The proportion of oxygen was then 12 per cent 
and it was thought that the bees had probably been suffocated. Eight 
hours before the time at which the period would regularly end the 
air of the system was again putin circulation and water vapor and 
carbon dioxid removed, oxygen being also removed at the same time. 
This was continued until the close of the period (which was also 
the end of the experiment) in order that the air of the system might 
be quite thoroughly freed of carbon dioxid. After the circulation 
of air was resumed the bees again indicated that they were living, 
and during the time that the air-purifying system was operating 
their activity increased until by the end of the experiment the tem- 
perature curve had reached as high a point as at any time during 
the course of the experiment, even though the proportion of oxygen 
in the air was low. Analysis of the sample taken at the end of the 
period showed only 7.3 per cent of oxygen. 
If the decrease in the activity of the bees in this instance was due 
to atmospheric conditions in the hive, the cause was probably excess 
of carbon dioxid and water vapor rather than deficiency of oxygen. 
Though the proportion of oxygen in the air was decreased from 18 
to 12 per cent in 16 hours, it is doubtful if this alone would have 
an appreciable effect upon the physiological activity of the bees. In 
experiments with men in atmospheres about as deficient in oxygen as 
this, there was no noticeable effect upon their metabolism. In these 
experiments, however, there was no such excess of carbon dioxid and 
water vapor as in the experiment with the bees. 
It is possible, as mtimated on page 6, that the reason for the 
increase in activity of the bees after the circulation of air was re- 
sumed may.have been physical disturbance. Since it was thought 
that the bees were dying, movement about the laboratory was some- 
what less restricted when the air-circulating device was started, al- 
though care was still taken to avoid jarring the calorimeter. The 
circulation of air through the calorimeter could hardly have caused 
any disturbance of the bees, because the low rate, while sufficient to 
keep the air in motion, could not produce any current that would stir 
the hive. It is also possible that, since the removal of oxygen from 
the air was continued during this period, the proportion of oxygen 
in the air eventually became so low that the bees had to respire more 
rapidly to obtain a sufficient quantity of this gas. It would be ex- 
pected, however, that this effect would be manifested somewhat later 
in the period than the time at which activity was renewed. 
