Severin and Sever in—Habits of the Saw-fly . 71 
difficulty; but when the under surface of the leaf is examined, 
one can readily observe a slight elevation of the epidermis. 
Within a few days the elevation is increased, and, as has been 
jjreviously noted by Comstock (1) and Harris (3), in some 
strange way, perhaps by the absorption of moisture from the 
leaf, the egg increases in size until it becomes twice or three 
times its former size. The receptacle itself becomes enlarged, 
and a few days before hatching the outline of the future larva 
is easily seen through the lower epidermis, bent upon itself 
around one pole of the egg. 
On July 14, six specimens were allowed to deposit eggs on 
a peach-leaved willow and their development was observed 
from day to day. On July 22, eight days after the deposition, 
all the eggs with the exception of a few hatched. We had the 
opportunity of observing the manner in which the larva escapes 
from its receptacle. Whether the larva first eats its vray 
through the thin membrane which encloses it, or whether this 
thin shell is broken by the growth or movements of the larva, 
we are not certain. As soon as the head is free from the egg, 
one can easily see a pair of brown mandibles cut a semi-cir¬ 
cular lid through the thin epidermis of the receptacle (Fig. 
2, a ). The larva, without stopping to eat the membrane which 
encloses it, pushes up the lid and makes its escape. 
THE EARVA. 
One would naturally suppose that the larva would escape 
from the mouth of the receptacle, but if this opening is ex¬ 
amined rather closely one is struck with surprise to find that 
the epidermis around the mouth of the receptacle has grown 
together with the parenchymatous tissue. The question which 
now arises is, Why does not the rest of the epidermal layer cf 
the receptacle, not occupied by the egg, grow together with 
the parenchymatous tissue? We can give only one possible 
explanation in answer to this question. As already mentioned, 
after the female has deposited an egg, it injects a certain 
amount of liquid from the collateral sac into the receptacle. 
The use of this fluid, perhaps, prevents adhesion between the 
epidermis and the parenchymatous tissue. 
