16 BULLETIN 1328, U. S. DEPARTM&NT OF AGEICULTURE 
It is noticed that the bees did not respond to the cloudiness at 
2.30 p. m. ; but \Yhen the sky darkened at 2.45 p. m. the curve rep- 
resenting incoming bees immediately began its rapid ascent and the 
outgoing bees showed a decrease, the divergence of the two curves 
continuing at an equal rate until 3.15 p. m., when the greatest number 
of returning bees had entered the hive. The climax of the threaten- 
ing storm was just previous to 3.30 p. m., when the curve of outgoing 
bees reached its lowest point, and all but 13 of the bees in the field 
had entered the hive. At this point the surprising fact is revealed 
by this day's record that, although the storm was so severe as momen- 
tarily to send all the field bees back to the hive, the impulse of the 
bees to go to the field while the secretion was good was so strong that 
in the half hour in which the climax of the threatening storm 
occurred as many as 50 to 75 per cent of the bees, which were going 
out under optimum conditions for this day, were still leaving the 
hive. The quick response of the bees, this time to the gradual but 
rapidly improving conditions of the weather, is seen from 3.30 to 
4.15 p. m., when normal activity was resumed. It is interesting to 
speculate as to just what feature of the storm contributed most to 
this behavior in the flight of the colony. Was it drop in tempera- 
ture, variation in light or atmospheric pressure, difference in humid- 
ity, intensity of the wind, or general change in the electrical con- 
ditions of the atmosphere ? Unfortunately, not all of these questions 
can be answered definitely as yet. 
Of all these factors, temperature and light varied most. If tem- 
perature had been the chief cause, the flight might have been ex- 
pected to remain low after the storm, since the temperature remained 
low for the remainder of the afternoon. The response to the change 
in light, both its increase and its decrease in intensity, can be so 
closely correlated with the flight activity that this may be safely 
considered the chief cause of the variation. 
Comparing this day with May 13 (see Table 1), when the number 
of field bees was perhaps slightly less and on which there were 58,887 
returns, we find that there Avere only 54,528 on Ma}^ 15, which shows 
that the threatening storm of the latter day reduced the total flights 
by at least 7.40 per cent. On the assumption that the nectar condi- 
tions on these two days were identical, this represents a loss of 69.58 
grams, due to this clouding over. Actuall}^ 90 grams less nectar 
was gathered, which would indicate that there was little or no dif- 
ference in the nectar conditions on these two days. 
Comparing May 15 with May 16, which was bright and sunny all 
day, with 60,377 returns, it is found that the decrease in total flight 
on May 15, apparently due to this storm, was 9.69 per cent. Only 
20 grams more nectar was gathered on the 16th, showing that 
although a good day it was not quite so good a day for nectar 
secretion as either May 13 or IVIay 15. This indicates that the 
afternoon storm of May 15 reduced the day's flight by from 7.40 to 
9.69 per cent with, of course, a corresponding reduction in the day's 
gathering compared with an optimum secretion. 
The record of July 8 (fig. 5) is typical of the flight when actual 
precipitation occurs. This was a bright, sunny day until 12.30 p. m., 
when a general clouding oxer occurred. The following data show 
the progress of this storm : 
