8 BULLETIN 1222, U. S. DEPARTMENT OF AGRICULTURE. 
grooves one-fourth of an inch in diameter in the surface of a flat 
block of wood which had been impregnated with melted wax. These 
cells were closed at the ends with wax and at the top with cover slips. 
The larvae were then placed in an incubator and kept at 37.5° C. All 
of these larva? spun more or less perfect cocoons, but only two suc- 
ceeded in forming normal pupae. From these two pupae perfect adult 
bees emerged. The weight of these in milligrams, as compared with 
their larvae and the weights of normal larvae and pupae, are given 
below : 
Larva. Adult. 
No. 1 115.0 68.1 
No. 2 144. 1 94. 2 
Normal 158.3 112.0 
Since adult bees from fully fed larvae reared under similar artificial 
conditions are only slightly under normal weight, it is possible to 
produce undersized bees from larvae that are not fully fed. Under- 
sized or dwarf bees are familiar to most beekeepers, but whether 
they are due to underfeeding of the larvae is not definitely known. 
METHODS, LOT 8. 
In order to counteract variations observed in the earlier experi- 
ments as much as possible, an attempt was made during the work of 
1922 to perfect the method suggested under Lot 7, which seemed to 
give promise of the best results. A space in the middle of a hive wide 
enough for two brood frames was partitioned off with queen-excluding 
zinc at either side and at the entrance. This was to prevent the queen 
from escaping in any way after the cover was placed on the hive, but 
the workers could come and go at will. A fairly strong colony of bees 
was transferred to this hive, placing one frame nearly filled with 
brood in this partitioned-off space, with a second frame either empty 
or with only a small amount of sealed brood in it. The queen was 
then confined in this chamber. The nearness of the well-filled brood 
comb to the empty comb generally induced the queen to lay a fairly 
large number of eggs in a period of 12 hours or less. 
To obtain average live weights of larvae of known age at definite 
age periods frames of eggs were removed from this chamber after 
a definite length of time. The frame removed was replaced by 
another empty one for the next series. The removed frame of eggs 
then was kept with several other frames of brood covered with nurse 
bees in a super over a strong colony, with the queen confined below 
by a queen excluder. After 3 days' incubation as eggs, at intervals 
of 24 hours each, larvae were removed by means of small forceps for 
weighing. So far as possible, the time of day selected for weigh- 
ing the larvae corresponded to the middle of the egg-laying period 
for that series. A method of weighing similar to that previously 
employed was used, except that, to reduce the probable error, five 
larvae were taken for each weighing instead of individual larvae. 
Larvae of average maximum size for the age period, and as nearly 
alike in size as possible when selecting from gross appearances, were 
chosen for each group of five. These were carefully removed, 
cleaned of food material, placed in a watch crystal, and weighed 
immediately by means of a chemical balance. Even with this 
method more or less variation in weight for each age period was 
