FOSSORIAL HYMENOPTERA 
455 
retreat — the fly * knowing ’ the sound of her wings, and retiring 
before she actually alights. Sometimes, however, the careless 
Bernbex has failed to seal the entrance, and Idia has its chance, 
and the Bernbex larva is perhaps supplanted in its birthright. 
The Bembecidce, as has been said, usually feed their larvae 
on flies, very often selecting especially the blood-suckers. 1 I 
have found in the nest of Bernbex capensis remains of Tabanus 
thoracinus and Chrysops brucei, and seen it carry into its 
burrow a full-fed Glossina. I have watched Bernbex jorcipata 
carry down one Tsetse after another, so that in three and a 
quarter hours twenty-nine Tsetse and two other flies had been 
taken into the burrow, which, when opened, was found to con- 
tain thirty-one Tsetse. Another burrow was opened up and the 
larva removed and put into a pill-box, when it ate nine Tsetse 
flies in twenty-one hours. This was more than half grown : 
it would seem that if it only requires a week to become full 
fed, it must devour some forty or fifty flies. When full fed 
it spins a neat pear-shaped cocoon of silk, grains of sand being 
firmly embedded in the outer covering, in which it changes 
into the quiescent pupa, from which the perfect insect emerges 
in a few months. 
It may have been noticed in the accounts given that 
Bernbex never puts her prey down, but holds it always pressed 
close to her body, so that her enemies have not the chances 
given to the enemies of two other families — Pompilidce and the 
burrowing members of Sphegidce. Sphegidce differ consider- 
ably from Bembecidce in appearance : instead of rather a broad 
black-and-yellow abdomen, they have a long narrow abdomen, 
with often an extremely elongated, slender pedicle, and their 
colouring is mainly black, with an admixture of a little dull 
red, white, or yellow ( vide Frontispiece, Fig. 7, Sphex 
luteipennis). They are of varied habits, but all agree in 
supplying their young with paralysed prey, of which an 
ample supply is provided once, for good and all ; the young 
being left entirely to itself. 
I noticed one species, whose name I forget, making use 
of a dead hollow cane as a nursery, so that it saved itself all 
the trouble of either burrowing or building; this habit, how- 
1 Sleeping Sickness Commission of the Royal Society, Report No. XII, 1912. 
