36 BULLETIN 1328, U. S. DEPARTMENT OP AGRICULTURE 
tical application would perhaps be in a study of nectar secretion, as 
it affords a measure of the variations in flight and in the duration 
and number of trips which the bees make, as well as their average 
daily load, facts which are to a large degree a reflection of the nectar 
conditions in the field. 
This study demonstrates the feasibility of obtaining data on prob- 
lems pertaining to the flight of bees by means of an automatic re- 
cording mechanism. As far as can be ascertained, the mechanical 
principles on which the design was made are in general correct. 
Much work, however, on further details is desirable. 
Owing to factors incidental to any new and untried mechanism, 
a variable experimental error was introduced which as far as can be 
determined for any day varied from 0.08 per cent to as much as 27.81 
per cent. The magnitude of this error, though prohibitive for cer- 
tain phases of the study, was not entirely so for others, and by a 
selection of those days on which it was obviously at a minimum, in- 
formation may be obtained which is practically as valuable as if no 
error were incurred. 
A survey of the total daily exits and returns for the period of the 
observations shows that a. factor or group of factors can reduce the 
total number of possible exits by an amount varying from total pro- 
hibition of flight to a fraction of 1 per cent. A threatening storm, 
for instance, of but one hour's duration, reduced the possible flight 
on one day in the honey flow by 7.41 to 9.67 per cent. 
Comparatively few data have been obtained on the effect of wind on 
the flights. On one day, however, a wind velocity of 16 to 21 miles 
per hour during the hours 9 a. m. to 6 p. m. reduced the possible 
maximum flight by 28.53 per cent. 
Under a particular set of conditions the temperature at which the 
day's flight commences is uniformly near a certain definite tempera- 
ture, but this definite temperature is not always the same. In April 
it was from 12° to 14° C. and in May from 16° to 18° C. On dull 
days this temperature was usually 2° higher. The internal conditions 
of the colony govern this temperature somewhat, a strong colony 
commencing flight at a lower temperature than does a weak one. 
There is a considerable variation in the hour and temperature at 
which the peak of the flight in the honey flow occurs. No conclusive 
evidence has been obtained that under similar conditions a good 
honey flow induces the bees to go out in large numbers at a lower 
temperature than they would if no nectar were available. The tem- 
perature at which the flights in the evening begin to slacken was 
without exception higher by from 1° to 9° C. than the temperature at 
which flight began in the morning. Days Avhich appear to be simi- 
lar in every respect but which show a variation of as much as from 10 
to 25 per cent in their total flights are found to differ on account of 
a lower temperature in the early part of the day. 
Under honey-flow conditions the total exits proved to be three to 
four times as great as they were at any other time of the investiga- 
tion. 
A typical flight, where all conditions are fairly uniform and fa- 
vorable throughout the day, is a gradual to rapid increase in the 
successive numbers of bees wliich set out from the hive, to be fol- 
lowed by a condition of equilibrium in which the outgoing numbers 
