Tur BioLoGy OF THE CHRYSOPIDAE 1319 
larva from view. This last act is carried out very slowly. Spinning 
usually requires from twenty-four to forty-eight hours, though some 
larvae apparently finish in a shorter time. 
Special effort was made to see whether the lid, by way of which the 
pupa leaves the cocoon, was spun into the cocoon. But to date this has 
not been seen, nor have any constant characteristics in the spinning 
process been observed which would warrant the conclusion that it was 
spun into the cocoon by the larva. 
If, while a larva is spinning, the cocoon be torn or the outside be 
pressed in with a dissecting needle, interesting reactions follow. The 
larva tries to defend itself and may plunge its jaws through the unfin- 
ished cocoon in this effort. Spinning stops for the time being. After 
a short period of waiting, the jaws are withdrawn and the spinning 
proceeds. If the cocoon be cut or torn, the opening is patched so that 
at the end the tear can scarcely be found. It is made of the same thick- 
ness as the remainder. of the cocoon. 
If a spining larva is disturbed, it does one of two things: it 
either walks around for a while and begins to spin at another place, or 
goes to the bottom of the vial, coils up, and spends its pupal life outside 
the cocoon. Disturbed larvae have been observed to come back to the 
cocoon first begun and spin another beside it. If the first cocoon is 
well started, the chances are that the larva will not spin further, since, 
as Lurie (1898) pointed out, the amount of silk secretion is undoubt- 
edly limited, and if too much silk has been wasted in an unsuccessful 
attempt to form a cocoon the larva cannot secrete enough to complete 
another. It may, however, spin feebly for about twenty-four hours, 
makine a mat of silk around the tail. A very few larvae appear to 
make no attempt at all to spin, but pupate in the open. 
The cocoon 
The cocoon is spherical or slightly elongated in shape(Plates LXX XVII 
and LXXXVITI), and in all cases pure white in color. The cocoon 
proper is very thin. It appears like paper, but the original framework 
gives it more or less of a shaggy appearance. The silky layer is thin, 
hard, and difficult to tear. It was found, by submerging cocoons for 
different periods of time, that they could be submerged for several hours 
without being permeated by water ; after a longer time, however, the water 
did enter. In some eases the cocoon has one or more ringlike bulges. The 
size of the cocoons varies somewhat in the species, very probably with the 
Size of the larvae that spin them. One cannot with any certainty dis- 
tinguish the species by the cocoon. The packet-carrying larvae use 
