14 BULLETIN 96, U. S. DEPARTMENT OF AGRICULTURE. 



COMPARISONS OF TEMPERATURES OF THE CENTER OF THE CLUSTER 

 AND OF THE OUTSIDE AIR. 



The curves have revealed no more striking results than the relation 

 observed between the temperature in the center of the cluster, c, 

 as compared with the temperature of the outside air, o. These 

 curves (fig. 8) at times show a peculiar inverse relation; for instance, 

 when the thermometer out of doors registered low, below zero, the 

 thermometer in the center of the cluster registered high, and vice 

 versa. It should be observed that the maximum within the cluster 

 occurs practically simultaneously with the minimum outside, and 

 vice versa. Even minor changes outside are accompanied by cor- 

 responding inverse fluctuations in the cluster. The responses of the 

 cluster to the outside temperature were shown particularly by the 

 thermometer which recorded the temperature of the center of the 

 cluster, c. 



Up to the day of the first egg laying in the spring, March 9, the 

 general courses of c and o continued relatively constant. But with the 

 commencement of egg laving c changed its trend. The temperature 

 of the brood cluster then became more and more constant, as may be 

 seen in the results of the summer observations. 



At first glance these curves might be interpreted as independent 

 of each other, that the outside atmosphere has no effect on the center 

 of the cluster, that it does not penetrate and modify the readings of c 

 as it appears to have done in the case of the temperatures in the margin 

 of the cluster. In all probability c more nearly represents the 

 activities of the bees than do the other temperatures; but there is a 

 relation of c to o. It might be supposed that the reaction registered 

 by c is deferred for a period of hours and consequently appears at a 

 time when o has changed. For instance, corresponding to the 

 minimum of o on the 4th of February, the minimum of c came nine 

 hours later. If this is due to a delay or ''lag," maxima and minima 

 in some cases are delayed for 24 hours or more. But this can not be; 

 there are many minor variations which appear on the curves, and 

 which are synchronous. Were there no relation of c to o these minor 

 variations would either not have appeared in c, or, more especially, 

 they would not have occurred simultaneously with a minor fluctua- 

 tion in the outside temperature. It is therefore impossible to explain 

 the phenomena on the ground of retardation (lag), for in that case it 

 would be far more constant than is evident. 



Related to the assumed explanation by delay or "lag," humidity or 

 condensation, convection, radiation, and conduction might be 

 assumed to be factors involved. The experimental colony furnishes 

 no data for a consideration of humidity or condensation. The factors 

 of convection, radiation, and conduction can not be conceived as slow 

 enough to retard c from 9 to 24 hours nor would it account for its minor, 



