848 Transactions.— Zoology. 
forming a wonderfully close imitation of the natural bark of the tree. 
Beneath this covering is seen a cavity, the depth of which varies from a 
quarter to half an inch, and rather above the centre of the cavity is the 
entrance to the burrow. The cavity around the entrance of the burrow is 
used by the larva to turn itself in. The burrow at first takes a course 
inwards and upwards for one, two, or more inches ; this upward inclination 
preventing the entranee of water. Then the burrow turns downward ina 
nearly vertical direction. This vertical portion of the burrow varies in 
length according to the age of the larva. That of a full-grown larva is about 
four or five inches long. I have seen them eight inches and proportionally 
wide. I have ascertained positively that the larva frequently inhabits the 
same burrow for more than two years, and I am of opinion that they 
generally keep to the same burrow during the whole larval stage. But I 
have occasionally found larve in terminal shoots which would not admit 
of their attaining full growth, and consequently they must in these cases 
change their ground. 
When the larva has attained its full size, it spins, at the top of the 
vertical portion of its burrow, a contrivance very much resembling that of 
the Trap-door Spider, as an additional security against its foes during the 
pupa state. 
The imago emerges in October and.November. The best time to obtain 
the pupa is in September and the early part of October. It is easy to 
ascertain if the insect in a burrow is in the larval or pupa stage, for, if the 
exterior web is torn off, the larva, if inside, will replace it by the next day. 
The larva is flesh-coloured, tinted with purple; head dark brown, with 
a few strong bristles ; spiracles black; segment next the head darker than 
the rest, horny, with a large black mark on each side, just above the spiracle. 
The pupa is flesh-coloured, inclining to brick red; head and thorax deep 
chestnut brown. The semi-transparent wing-cases show the markings of the 
future imago. 
I have examined these larve and pups in their different stages up to 
the time when by stripping off the pupa case, just before the insect was 
ready to emerge, the easily recognised Charagia virescens was disclosed. 
I do not think that the larva of C. virescens is the larva which is 
attacked by the fungus Cordiceps robertsii. Not only do the two laive differ 
in the size of the head and shape of the body—the larva cf C. virescens 
being more cylindrical and with proportionately a larger head— but I do not 
see how C. virescens could get into the ground, which is the position in 
which the larva, which is attacked by the fungus, always is found. I tLink 
that the fungus-attacked larva is probably a Purina, 
