THE FLIGHT ACTIVITIES OF THE HONEYBEE 
29 
condition previously described, vrhen the outgoing bees equal the 
returning bees in number. Again, as flight closes, a considerable 
shortening of the duration of the trips is shown. On May 15 (fig. 
11), at 3.30 p. m., the two curves, representing the accumulated 
totals of bees emerging and bees returning, come together, to di- 
verge immediately to a resumption of the previous condition of 
parallelism. This convergence with an immediate divergence was 
occasioned by a threatening storm, illustrated in a previous curve 
(fig. 4) , which sent all the bees in the field back to the hive and neces- 
<'(?*'* 
K 
P.- 
/C 
c 
/I 
/ ^ 
/ - 
\\ 
/ 
1 "" 
/ 
, . 
11 
I 
/, 
/ 
/ 
- 
— — 
^/%^ 
//V-^ 
' 
I 
1 
/ 
f 
AXOC 
// 
// 
1 
/ 
1 
1 
1 
1 
( 
/ 
/ 
V 
/ 
/ 
/ 
1 
I 
// 
// 
/, 
' 
% 
// 
1 
// 
1 
z& 
V 
// 
/ 
/ 
av ' 
/ 
I 
/7 
/, 
/ 
^ 1 
/ 
f 
1 
// 
// 
L 11 
: 
' 
I 
■xx> / 
A-'^ 
x_/ •"^-.' ' 
& 
,^ 
1 
r 
\ 
Fig. 
11- 
-Gr 
aph 
9 i 
repi 
? A 
•esec 
7 / 
iting 
flig 
' / 
ht da 
ta for May 15, 1922, a day of good 
flow 
of tuliptree honey 
sitated an abrupt shortening of the duration of the voyage. At the 
end of the day the incoming curve, instead of ending below the out- 
going curve, crosses it, the vertical distance between the ends of the 
two curves representing the accumulated experimental error for this 
day, when 467 more returns than exits were registered. This error 
has naturally tended to reduce the distance between the two curves, 
making the apparent duration of the voyages shorter than the real 
duration. To obtain some idea of the magnitude of the error pro- 
duced in the determination of the duration of the trips in the period 
