GROWTH AND FEEDING OF HONEYBEE LARVAE. 11 
NATURE AND COMPOSITION OF BROOD FOOD. 
The foregoing observations have a bearing on the much discussed 
question of the source and nature of brood food at different times 
during the feeding period. The relation of the composition and 
amount of brood food to the rate of development may assist in explain- 
ing the variations observed above. 
To obtain the weight of food unconsumed in the cells, as an indi- 
cator of the amount of food consumed in relation to the weight of 
larvae of approximate known age, the following method was used : 
A frame of brood was chosen from a normal colony, containing in 
an area of 2 or 3 square inches several larvae of uniform size. The age 
of these larvae was approximately determined by inspection and com- 
parison in size with the drawings of larvae of definite age (figs. 6 to 
11). A piece of this comb was then cut out containing about the 
desired number of larvae. The edges of this piece of comb were 
trimmed clean with a sharp scalpel and the cells of the comb were cut 
down so as to reduce the depth of the cells to about half. The edges 
and all cells containing larvae of abnormal size not desired for weigh- 
ing were cleaned and dried with swabs of absorbent cotton on the 
ends of matches. This was done as rapidly as possible so as to prevent 
loss by evaporation from the remaining cells. The piece of comb was 
then carefully weighed. The larvae were then removed with fine 
forceps, care being taken to remove as little adhering food material 
as possible, and then the piece was weighed again. The difference 
between these two weights gives the approximate weight of the larvae 
(Table 3) . To obtain the weight of the remaining food, each cell was 
cleaned with a swab of absorbent cotton and the cell was then dried 
out thoroughly with a second swab to insure complete removal of all 
the food material. The weight of the empty comb was taken, by 
which the weight of the food material is determined. The average 
age of each lot was more approximately determined by comparison 
and interpolation with the weights of larvae of known age, given in 
Table 2. 
Unfortunately, owing to lack of time and the small number of de- 
terminations made, the data given in Table 3 are only approximately 
accurate. The data are suggestive, however, and considerable inter- 
esting information may be obtained by correlation with the rate of 
growth, the time of change in food composition, and the time spent 
by nurse bees in caring for the larvae of various ages. 
It is noticeable that the small, young larvae are always surrounded 
by or even are floating on an excessive amount of food material, 
which has a uniformly grayish- white, pastelike consistency. From 
the data observed (Table 3, figs. 3 and 4), it was found that up to 
between the second and third clays this excess of food, as indicated 
by the amount unconsumed in the cell, is considerably greater than 
the weight of the larva itself. There has been much controversy 
as to whether the food of the young larva is a secretion from glands 
or a regurgitation from the ventriculus [Snodgrass (#)]. Its con- 
sistent lack of pollen grains and uniform appearance suggest, secre- 
tion rather than regurgitation, since if regurgitated it would of 
necessity contain undigested pollen grains, as these are always found 
in the ventriculus. If a secretion, the food of the younger larva 1 
would be affected much less bv variations in external conditions, such 
