OCT. 1906. HABITS AND LIFE-HISTORY OF LEUCOSPIS AFFINIS SAY. 
155 
its body around in such a manner as to bring the tip of the abdomen 
just over the point palpated by the tips of the antennae. There- 
upon the ovipositor is brought into action. This has its origin on 
the lower surface of the abdomen, but is bent up over the back of 
the latter. Standing high up on its legs the insect bends the tip 
of its abdomen forward under its body, and by boring movements 
forces the ovipositor down through the wood. This was done by 
the insect under observation in three different places correspond- 
ing to the three cells of the bee's nest, and in each instance it took 
about ten minutes to complete the work. After it had got through 
with all of the cells it spent a considerable time resting on the 
branch. 
The European Leucospis gigas has to drive its ovipositor 
through the hard cement comprising the walls of the nest of the 
mason-bees Chalicodoma muraria and C. sicula, and according to 
Fabre it spends at times three hours before reaching the cavity of 
the cell. 
Egg. The slightly curved egg of Leucospis afhnis (Fig. i) 
shows a small knob-shaped tip on the rounded cephalic pole, while 
the other end is drawn out into a long, slender process. It reaches 
a length of about iy 2 mm., and shows under the microscope an 
elegant sculpture, due to numerous small papillae covering the 
egg-membrane. It differs in appearance from the egg of Leu- 
cospis gigas (Fig. 2) with its hook-shaped slender process as 
illustrated by Fabre. 
Larva. In the larva of Leucospis we are dealing with a 
marked degree of hypermetamorphosis and the conditions met with 
in the European species L. gigas have been considered very' thor- 
oughly by Fabre. If the size of the head of the larva before the 
first moulting, the so-called "primary larva" is correctly repre- 
sented in Fabre's illustration of the primary larva of L. gigas 
(Fig. 4) then we have in our species L. afflnis (Fig. 3) a more 
advanced degree of hypermetamorphosis since the head of the 
latter is comparatively much larger. In other respects the larvae 
of the two species seem to agree pretty closely. 
Of the three eggs of Leucospis afUnis found in the cells of 
Osmia atriventris not one was in contact with the Osmia-larva, 
every one of them was attached to the wall of the cell. 
At 6 a. m. on July 14th a newly hatched primary larva was 
