HABITATIONS OF INSECTS. 457 



enters backward, dragging after it a leaf of Banksia 

 serrata, which it holds by the footstalk*. 



A third description of larvas, chiefly of the two lepi- 

 dopterous genera TortrLv and Tinea, form into conve- 

 nient habitations the leaves of the plants on which they 

 feed. Some of these merely connect together with a 

 few silken threads several leaves so as to form an irre- 

 gular packet, in the centre of which the little hermit 

 lives. Others confine themselves to a single leaf, of 

 which they simply fold one part over the other. A 

 third description form and inhabit a sort of roll, by 

 some species made cylindrical, by others conical, re- 

 sembling the papers into which grocers put their sugar, 

 and as accurately constructed, only there is an open- 

 ing left at the smaller extremity for the egress of the 

 insect in case of need. If you were to see one of these 

 rolls, you would immediately ask by what mechanism 

 it could possibly be made — how an insect without fin- 

 gers could contrive to bend a leaf into a roll, and to 

 keep it in that form until fastened with the silk which 

 holds it together? The following is the operation. 

 The little caterpillar first fixes a series of silken cables 

 from one side of the leaf to the other. She next pulls 

 at these cables with her feet ; and when she has forced 

 the sides to approach, she fastens them together with 

 shorter threads of silk. If the insect finds that one of 

 the larger nerves of the leaf is so strong as to resist 

 her effbrts, she weakens it by gnawing it here and 

 there half through. What engineer could act more 

 sagaciously ? — To form one of the conical or horn- 

 ghaped rolls, which are not composed of a whole leaf, 



* Le\vin"s Prodromvs Entomology (sic !) p. 8, 



