sept. 1897.] Packard: Transformations of Hymenoptera. lit 
colony could be estimated by counting the empty cells. These were 
wanting in the upper third, or rarely the upper half, which had been 
eaten away by the bees to allow the occupants to escape ; the edges 
being rough and irregular. Some of the cells were nearly all gone, three- 
fourths of some of them having been removed; these were situated on 
the sides of or nearly beneath the bunches of small cells which sur¬ 
rounded the single female or queen cell. 
At this date there were 58 empty cells, hence the colony, if all were 
alive, was of course composed of that number of individuals; of these 
all were workers except a single male and two females. 
Larva. —The larvae are easily distinguished from those of B.fervi- 
dus and B. separates, which is the more unexpected, since the two last 
named species agree so closely after the specimens compared have been 
in alcohol. The head is considerably smaller, nearly one-fourth so, 
than m fervidus, while the transverse raised bands across each ring are 
much thicker, and the lateral raised pleural lines are much more promi¬ 
nent than in B.fervidus , thus making the under side of the body ap¬ 
pear flatter and the upper side more convex than in fervidus. The 
whole body is more lunate, compact and blunter at the extremity than 
m fervidus. Such are the differences in comparing twenty larvae with 
an equal number of those of fervidus. Whether these differences are 
constant, and have been stated correctly, future study will prove. The 
sizes of the different stages of growth correspond very exactly with 
those of the equivalent stages in fervidus. 
Eggs —The eggs of this and all the species when compared do not 
differ, and if they were all intermingled, the species could not be picked 
out. 
Pupa. —Comparing some (eight) $ semipupae with an equal num¬ 
ber of 9 semipupae of B. fervidus, there are no differences, not even 
in the tip of the abdomen. 
Compared with the male of fervidus it is very considerably smaller 
and slenderer, the abdomen being sensibly more produced towards the 
more acute tip and the limbs are throughout more slender. The head 
is shorter and broader. The second joint of the antennas is longer, 
passing beyond the eyes, where in fervidus they do pass beyond the 
lower angle and outer edge of the eyes. The maxillae and lingua are 
shorter than in fervidus , being just as described in the worker pupae of 
B. separatus, and are unitedly narrow, as in the last named species. 
The limbs are no longer, but all the joints are considerably narrower 
than in fervidus. Here, as in the other sex, the genital armature does 
