INSTINCT OF INSECTS. 573 



directs it to malce its cocoon, substituted pieces of paper that were given 

 to it, tied them together with silk, and constructed a very passable cocoon 

 with them. In another instance the same naturalist having opened 

 several cocoons of a moth (^Cucullia Verhasci), which are composed of a 

 mixture of grains of earth and silk, just after being finished, the larvae did 

 not repair the injury in the same manner. Some employed both earth 

 and silk ; others contented themselves with spinning a silken veil before 

 the opening.^ 



The larva of the cabbage-butterfly (JPoniia Brassicicn), when about to 

 assume the pupa state, commonly fixes itself to the under side of the 

 coping of a wall or some similar projection ; but the ends of the slender 

 thread which serves for its girth would not adhere firmly to stone or brick, 

 or even wood. In such situations, therefore, it previously covers a space 

 of about an inch long and half an inch broad with a web of silk, and to 

 this extensive base its girth can be securely fastened. That this proceed- 

 ing, however, is not the result of a blind unaccommodating instinct seems 

 proved by a fact which has come" under my own observation. Having 

 fed some of these larvae in a box covered by a piece of muslin, they 

 attached themselves to this covering ; but as its texture afforded a firm 

 hold to their girth, they span no preparatory web. 



Bombus^ Miiscorum, and some other species of humble-bees, cover 

 their nests with a roof of moss. M. P. Huber having placed a nest of 

 the former under a bell-glass, he stuffed the interstices between its bottom 

 and the irregular surface on which it rested with a linen cloth. This 

 cloth, the bees, finding themselves in a situation where no moss was to be 

 had, tore thread from thread, carded it with their feet into a felted mass, 

 and applied it to the same purpose as moss, for which it was nearly as 

 well adapted. Some other humble-bees tore the cover of a book with 

 which he had closed the top of the box that contained them, and made 

 use of the detached morsels in covering their nest.^ 



The larva of Cossus ligniperda, which feeds in the interior of trees, 

 previously to fabricating a cocoon and assuming the pupa state, forms for 

 the egress of the future moth a cylindrical orifice, except when it finds a 

 suitable hole ready made. When the moth is about to appear, the chry- 

 salis with its anterior end forces an opening in the cocoon. If the orifice 

 in the tree has been formed by itself, in which case it exactly fits its body, 

 it entirely quits the cocoon, and pushes itself half way out of the hole, 

 where it remains secure from falling until the moth is disclosed. But if 

 the orifice, having been adopted, be larger than it ought to have been, 

 and thus not capable of supporting the pupa in this position, the provident 

 insect pushes itself only half ivay out of the cocoon, which thus serves 

 for the support which in the former case the wood itself afforded.^ 



The variations in the procedures of the larva of a little moth described 

 by Reaumur, whose habitation has been before noticed — one of those 

 which constantly reside in a sub-cylindrical case — are still more remarka- 

 ble. This little caterpillar feeds upon the elm, the leaves of which serve 

 it at once for food and clothing. It eats the parenchyma or inner pulp, 

 burrowing between the upper and under membranes ; of portions of which 



' CEuvres, ii. 238, See above, p. 4^)1. » Apis. **. e. 2. K. ^ Linn. Trans, vi. 254. 

 * Lyonet, Traiti Anatomique, &cc. 16. 



