NOVEMBER, 1905. SOME OBSERVATIONS ON PARASITIC BEES. 
159 
During the early part of their life the larvae of this species live 
quietly side by side, but they gradually become less tolerant, and 
from the age of 5 or 6 days on they often grab each other with 
their mandibles when in close contact. On several occasions I 
have seen one of them secure a firm hold on another's body, and 
toss it around, without, however, puncturing the skin, or making 
any attempts at feeding on the one attacked. Under these circum- 
stances the larvae become very restless, do not feed normally, 
and die before reaching their full development. 
Through the fact that the attack of the Alcidamea-larva, does 
not produce any visible injury, while the bite of the parasitic 
larva of SieMs sexmaculata punctures the skin of the host-larva 
I have been led to examine the mandibles of the two larvae. In 
the parasite the mandible is simple and sharp-pointed, (Fig. 3), in 
Alcidarnea the apex of the mandible is broad and cleft (Fig. 2), 
and this difference in structure explains the difference in the 
result of the bite. The sudden attack of the parasite has evident- 
ly also much to do with the outcome. 
An examination of the mandibles of several other species of 
host-bees gave the following information. Osmia atriventris, 
Cress., Megachile infragilis Cress, and Cemtina dupla Say. 
have larvae with cleft (bidentate) mandibles, while in Calliopsis 
andreniformis Sm., and Prosopis pygmcea Cress, the mandibles 
are simple. 
There is a difference in the degree of tolerance as displayed 
by the larva of the host-bee Ceratina on one side and that of the 
host-bee Alcidarnea on the other, but even in the latter case the 
actions of the larva have hardly anything in common with those 
of the parasitic larva of Stelis sexmaculata. The AlcidameaAarva. 
is not hostile during the first days of its existence, and later on it 
attacks only when stimulated by close contact with a neighbor- 
ing larva. Furthermore its bite is ineffective, and is never accom- 
panied by any sucking movements. In Stelis sexmaculata the 
biting instinct is in evidence from the beginning - , and the larva 
is able to give up its original position on the bee-bread, and ad- 
vance towards its victim. As a result of the bite the skin of the 
latter is pierced, a firm hold is secured, and sucking is begun at 
once. 
The egg of Stelis sexmaculata (Fig, lb), is mostly deposited 
near the base of the bee-bread, but occasionally higher up, and 
in one instance it was lying on top alongside of the host's egg 
(Fig. la). It is only half as long, and a trifle more than half as 
broad as that of the host, and accordingly the newly hatched 
