54 
[August, 
though of much greater size than the prothorax alone would account 
for. On the posterior slope of this protuberance may be seen the 
thoracic square, while in front the head seems bent down and in 
towards the sternum. Pressure is thus brought to bear on this region 
of the shell. The abdomen appears attenuated posteriorly, leaving 
empty spaces between it and the shell, as if to make up for the 
increased thickness in the thoracic region. The vermicular movements 
continue ; the anal pro-leg, at first dorsally recurved, moves forward 
in the shell. A process of the prothorax (which is somewhat more 
translucent than the other segments of the thorax) is protruded 
through the opening. It is wedge-shaped, the prothoracic spiracle at 
its posterior margin — the dorsum broadened, and the prosternum 
running out in the thin edge of the wedge. In front of it, the larva 
is nipped in, as it were, at the neck, abruptly to the head, whilst 
behind, the meso- and metathorax slope off towards the abdomen more 
gradually. After awhile the head is drawn out, vertex first. The 
hairs on the head point forwards, those on the other segments back- 
wards, and so oppose a return. Head and shoulders out, it seems to 
rest awhile. The remainder of the body follows more easily, taking 
15 or 20 minutes ; but the tail of the larva remains much longer 
within the now somewhat rounded opening in the clear transparent 
shell. 
The hatching having begun in one or two, others follow, and not 
in the order in which they were laid, after a longer or shorter interval ; 
but it may be several hours before all are hatched out, although laid 
within half an hour or a little more. And this extension of the time 
for a whole batch is still further increased in the successive following 
metamorphoses. In the meantime, the first hatched larvae will often 
attack and destroy the eggs still unhatched, even when not compelled 
thereto by want of other food. In nearly every batch there will be 
found a few “residual” eggs — undeveloped or arrested — 2 or 3 in 
general, or half a dozen. Sometimes many more, especially in the 
batches first laid by a $ , and towards the end of her career, when, in 
the latter case, there will generally be found a number of eggs, ab- 
normal in shape and size, smaller and sometimes spherical — all of 
which never come to anything. 
The larva feeds according to the season and the weather, from 12 
to 18 days, during which time it moults twice, so dividing the period 
into three pretty equal stages. After it is fed up it is quiescent for 
2 or 3 days before pupation. At the time of the first moult it is from 
2 to 21 mm. long; and 4 to 4| mm. when it undergoes the second 
