34 
The records for 89 days, which have been chosen as representative 
of the data obtained, show that of the bees which left the hive (2,434,- 
666), 3.16 per cent did not return. If the assumption is made that 
the errors in the counts of exits and entrances balance each other, 
this would mean, since every exit represents a trip by a bee, that one 
bee dies after every 31.65 exits have occurred, or that each bee makes 
31.65 trips before death overtakes it in the field. 
It is generall}^ accepted that the mortality of honeybees is in pro- 
portion to the amount of work which they do. It is also evident that 
during the night, when flights do not occur, all deaths will take place 
within the hive, and during the active season there is usually a period 
of 10 hours daily when there are no flights. If, therefore, about 98 
per cent of the deaths from the colony occur in the field, this indicates 
that most of the deaths occur during the approximately. 14 hours of 
flight. These data suggest that the energy expended by the bees 
in flight and in activities outside the hive greatly exceeds that ex- 
pended within the hive. This conclusion is to be anticipated from 
the enormous amount of energy which must be expended in flight. 
Assuming (1) that a field bee is away from the hive one-half of the 
flying hours, (2) that a death rate of 2 per cent occurs within the 
hive, and (3) that there is a 14-hour period of flight for the day, 
the chance of a bee dying outside the hive during any given hour 
of the period of flight is about 120 times as great as the chance that 
death shall occur within the hive. This at least suggests that for 
any given instant the expenditure of energy on flight or away from 
the hive is about 120 times as great as that expended by a single 
bee within the hive. 
THE BEHAVIOR OF THE BEES TO THE INSTRUMENTS 
In the initial stages of the experiment, although about 2,000 bees 
passed satisfactorily through the experimental model, the reaction 
of the colony as a whole to the complete set of instruments was en- 
tirely problematical. The success or failure of the experiment, 
therefore, depended upon this reaction. The plan of placing the 
outgoing gates above the incoming ones, so that the actual entrance 
apertures were only about 2 inches below the exit apertures, proved 
very satisfactory, only an occasional bee attempting to enter the hive 
by an outgoing gate. 
Considering the tunnel of the contact device as a hole in the wall 
of the hive which drops a distance equal to twice the height of a 
bee while the latter passes through it, the delay produced by these 
instruments is negligible and the only abnormality introduced is the 
jar of the tunnel meeting the lower stop. Owing to the fact that 
any hesitancy of a bee in walking out of the tunnel prevents another 
from using this aperture, a net delay on all the apertures ensues. 
The return movement of the tunnel to receive tlie next bee is a minor 
source of delay when compared with that produced by the hesitancy 
shown by some of the bees in selecting the aperture by which they 
finally enter. Probably this hesitancy on the outgoing gates is not 
so marked. It has been impossible to determine the actual delay 
produced. On seeing the rapidity shown on certain occasions by 
some of the bees when leaving the outgoing gates, one would not 
hesitate to state that these bees left the hive as rapidly as they do 
