146 INSECT ARCHITECTURE. 



pro-legs as a point of support, he exerted the whole muscles 

 of his body to shorten his threads, and pull down the edge 

 of the leaf. When he had drawn the threads as tight as he 

 could, he held them till he spun fresh ones of sufficient 

 strength to retain the leaf in the bent position into which 

 he had pulled it. He then left the first series to hang loose 

 while he shortened the fresh-spun ones as before. This 

 process was continued till he had worked down about an 

 inch and a half of the leaf, as much as he deemed sufficient 

 for his habitation. This was the first part of the architec- 

 ture. 



By the time he had ^worked to the end of the fold, he 

 had brought the edge of the leaf to touch the mid -rib ; 

 but it was only held in this position by a few of the last- 

 spun threads, for all the first-spun ones hung loose within. 

 Apparently aware of this, the insect protruded more than 

 half of its body through the small aperture left at the end, 

 and spun several bundles of threads on the outside pre- 

 cisely similar to those ropes of a tent which extend 

 beyond the canvas, and are i^^gged into the ground. 

 Unwilling to trust the exposure of his whole body on 

 the outside, lest he should be seized by the first sand-wasp 

 (odymrus) or sparrow which might descry him, he now 

 Avithdrew to complete the internal portion of his dwelling, 

 where the threads were hanging loose and disorderly. 

 F-or this purpose he turned his head about, and proceeded 

 precisely as he had done at the beginning of his task, but 

 taking care to spin his new threads so as to leave the loose 

 ones on the outside, and make his apartment smooth and 

 neat. When he again reached the opposite end, he con- 

 sti-ucted there also a similar series of cables on the out- 

 side, and then witlfdrew to give some final touches to the 

 interior. 



It is said by Kirby and Spence,* that when these leaf- 

 rolling insects find that the larger nervures of the leaves 

 are so strong as to prevent them from bending, they 

 "weaken it by gnawing it here and there half through." 

 A\'e have never observed the circumstance, though we 

 * Introd. vol. i. p. 457. 



