28 BULLETIN 1349, U. S. DEPARTMENT OF AGRICULTURE 
In 1920, contrary to conditions in 1921 and 1922, the nectar flow 
from black locust occurred during the last week in May, thus coming 
after the tulip-tree nectar flow, instead of before. June also fur- 
nished either nectar or honeydew, and a slight amount of nectar was 
available in September. No data are available as to pollen yields. 
In the case of each colony, brood rearing in the spring of 1920 was 
begun first in the second hive body. A check to brood rearing took 
place during the cold weather of April in both the packed and un- 
packed colonies. Inasmuch as the method of direct counting neces- 
sitated keeping the frames out of the hive for a considerable time, thus 
creating a disturbance to the colony lasting over several hours, much 
brood must have suffered undue exposure, a fact which in itself would 
account for any check at this time, even though other data are not 
available. In colony E this check is not so evident, owing to natural 
requeening during this period. In May, in each case, before the 
maximum was reached incoming nectar gave a check to brood rearing, 
while the brood nest was being maintained in the second hive body 
only, sufficient bees not being on hand to engage in brood-rearing 
activity in the lower hive body. Consumption of this fresh nectar 
soon made more room available in the second hive body, and the con- 
tinued emergence of bees so enlarged the colony population that 
possession was taken of the lower hive body for brood-rearing pur- 
poses. Only one super was given to any colony, this being provided 
on May 21. In each colony the maximum brood-rearing activity 
of the season was reached shortly afterwards. During the last week 
in August no records were taken, a fact which accounts for the gaps 
in the curves at that time. 
Colony A (fig. 19 and Table 18), after the initial check to brood 
expansion in April, proceeded at a fairly rapid rate to its maximum. 
This was not attained until after the close of the tuliptree nectar 
flow. A rapid decline then ensued, lasting through June, probably 
first brought on by incoming nectar, but later accentuated by the 
fact that the queen was confined to the lower hive body by a queen 
excluder from June 9 to June 30. She was given the freedom of both 
hive bodies during the first two weeks of July, and a slight increase 
in brood-rearing activity took place. The dearth of nectar and pollen 
during the remainder of July brought on a further decline. On 
August 11 the super was removed. A slight increase in the brood- 
rearing rate was made in September at the time of the usual fall nectar 
flow. The queen was lost during the first week of October. A virgin 
queen was then reared and mated successfully, but too late to produce 
much brood before winter. The maximum brood rearing in this 
colony took place about a month too late for ideal conditions, and 
brood rearing just before the period of final contraction was not 
sufficiently great to afford an ideal number of young bees for winter. 
Colony B (fig. 20 and Table 19), after the checks of April and May, 
attained its maximum in early June, following which there was a 
slight decline due to incoming nectar. A recovery made in the latter 
part of the month was followed by the midseasonal decline. On July 
9 the queen was confined to the lower hive body, and on August 12 
the super was removed. A somewhat pronounced response was made 
to the fall nectar flow. 
Colony C (fig. 21 and Table 20) did less in brood rearing than any 
of the other colonies. The maximum rate was reached late in May. 
The midseasonal decline is not so sharp as in the case of the other 
