4.80 INSTINCT OF INSECTS. 
While, therefore, they divide the major part of their in- 
ner side from the leaf, they artfully leave them attached 
to it by one of the large nerves at each end: and these 
supports they do not cut asunder until the mtermediate 
space has been sewed up, and they are ready to step, 
with their house on their back, upon the ¢erra firma of 
the disk of the leaf. In this instance, therefore, the 
larvse do not wholly separate’ their case from the leaf, 
until it is sewed. But when the same larvee cut out their 
materials from the middle of the leaf, where, though com- 
pletely cutround, they are retained in their situation secure 
from all danger of falling by the serratures of the incisions 
made by the jaws of the larvee, these little tailors vary 
their mode, and entirely detach the pieces from the surrond- 
ing leaf, before they proceed to set a stitch into them?. 
In the preceding instances the variation of instinct 
takes place in the same individual, but Bonnet mentions 
a very curious fact in which it occurs in differént genera- 
tions of the same species. There are annually, he informs 
us, two generations of the Angoumois moth, an insect 
which has been before mentioned’, as destructive to 
wheat: the first appear in May and June, and lay their 
egos upon the ears of wheat in the fields; the second ap- 
pear at the end of the summer or in autumn, and these 
lay their eggs upon wheat in the granaries. These last 
pass the winter in the state oflarvee, from which proceeds 
the first generation of moths. But what is extremely 
singular. as a variation of instinct, those moths which are 
disclosed in May and June in the granaries, quit them 
with a rapid flight at sun-set, and betake themselves to 
4 Reaum. in. 112-119. b Vor. b 4th Ed. 171, 
