HEAT PRODUCTION OF HONEYBEES IN WINTER. 11 
There was no apparent effect on the activity of the bees from 
variations in the carbon-dioxid content of the air in the hive, at 
least within very wide limits. One column in the table shows the 
percentage of carbon dioxid in the air at the time the residual analy- 
sis was made each day. These figures tell little of the condition of 
the air at any other period during the day; they merely show what 
it was after the air of the chamber had been passing through the 
soda lime for at least an hour; but unless the bees had been actually 
more active at the time the residual analysis was made (which, 
according to the thermocouples, did not occur in any instance) there 
must have been at least as much and probably more carbon dioxid in 
the air previous to the time of the analysis than is indicated by these 
figures. It would appear, then, that throughout the whole of the 
experiment the carbon-dioxid content of the air in the hive was 
appreciably greater than that of normal air, which is probably the 
usual condition in a hive; also there were outside variations in the 
proportion of this gas in the air, as shown by the data in the table. 
On December 21 and 22 arrangements were made to insure a consider- 
able excess of carbon dioxid in the air. During most of the time on 
these days the soda lime was removed from the train for purifying 
the circulating air and the carbon dioxid was, allowed to accumulate 
within the respiration chamber while the water vapor was removed. 
Starting with the content of nearly one-quarter of 1’ per cent on the 
20th, or almost eight times that in normal air, the increase continued 
until in the whole air system of the apparatus, which was about 170 
liters, there was included over 10 liters of carbon dioxid before the 
period ended on December 21, a proportion more than 200 times that 
in normal air. Ther®‘is no significant change in the curves on page 18 
showing the behavior of the bees, to indicate that they were materially 
affected by these abnormal conditions. The curve for thermocouple 
No. 7 cofitinued at the same level for nearly 12 hours, then began to 
rise slowly ; those for Nos. 12 and 6 fell somewhat for about 12 hours 
and then maintained a level for the remainder of the period. There 
would appear to be on the whole a quieting of the bees for this day, 
but this could be hardly attributed to the quantity of CO, present, 
for on the following day, when there was a still greater concentration, 
the activity of the bees increased. 
From the character of the curves in these two days it would be 
expected that the carbon-dioxid production on the 22d would exceed 
that of the 21st, but not necessarily by nearly 50 per cent as shown 
-in the values in the table. It is not unlikely that some of the carbon 
dioxid measured on the 22d was produced on the 21st. Uninten- 
tionally, replacement of the soda lime in the air circulating system 
was delayed until one hour before the close of the first period, and 
