22 BULLETIN 1328, U. S. DEPARTMENT OF AGRICULTURE 
2. On dull, overcast days the bees usually do not begin their flight 
until the temperature has risen at least 2° C. (3.6° F.) higher than 
the usual temperature for beginning flight for that period. This 
was very marked in April, May, and the early part of June, but 
was not so obvious during the latter part of June and the month of 
July. 
3. The internal condition of the colony is important in determin- 
ing the temperature at which flight will begin. Colonies in the 
apiary that were stronger than the one under experimentation al- 
ways commenced their flight at somewhat lower temperatures. The 
lowest temperature at which the flights from this colony began was 
on two days in April, when the temperature was as low as 10° C. 
(50° F.). 
4. No marked difference in flight temperature between honey flow 
and dearth conditions in the same period can be found, and no con- 
clusive evidence has been observed from these data that a heavy 
honey flow induces the bees to go out in large numbers at a tempera- 
ture lower than they would if no nectar were available. 
5. If from a lower temperature the thermometer rises rapidly to a 
condition suitable for flight, the temperature at which flight com- 
mences is usually 2° G. or more higher than if the temperature had 
risen slowly. In the first case, the flight curve usually shows a more 
gradual ascent to its peak ; in the latter, a more rapid rise is observed, 
indicating, as might be expected, that the hive as a whole must ab- 
sorb a certain amount of heat before the bees become active enough 
for a general flight. 
6. A comparison of the temperature at which the flight begins its 
upward trend and that at which it begins its descent in the after- 
noon shows that, without exception, the temperature at which the 
bees begin to slacken their flight activities in the afternoon is higher 
than that at which they become active in the morning, the difference 
ranging from 1° to 9° C. This indicates that it is the waning light 
which accompanies the approach of sunset, rather than a fall in 
temperature at that time, that causes the decrease in flight. 
7. A study of the seasonal flight curve (fig. 2) shows that a low 
morning temperature, by postponing the time at which flight com- 
mences, may reduce the total possible daily exits by as much as 50 
or even 75 per cent. Days which appear to be similar in every other 
respect, but on which a variation in flight of from 10 to 25 per cent 
is noted, furnish records which show that this difference may be 
traced to lower temperatures in the early part of the day. 
8. On some excessively hot days the flight curves remained low. 
Whether this behavior is due to the high temperature or to the 
dearth existing it is impossible to say. 
THE EFFECT OF LIGHT 
Since temperature and light intensity are so closely correlated, it 
is exceedingly difficult to differentiate the behavior produced by 
variations in either of these factors. In the study of the effect of 
light, a record was made of the degree of cloudiness each quarter 
of an hour while flight was occurring. Because of variations in the 
density, altitude, and position of the clouds, to say nothing of the 
