272 NATURAL HISTORY. [CH. xvi. 



^af-rollers, settle upon the oak : some of its leaves 

 may be observed rolled from the apex half-way down 

 towards the stalk (fig. 1) ; others from one side to- 

 wards another (fg* 2). Some of theso tubes are 

 composed of two leaves rolled tog-ether (fig> 3), while 

 others are constructed by bringing 1 each of the edges 

 of the leaf towards the principle nervure (fig- 4). 

 The fingers of the most dexterous seamstress could 

 not effect with greater precision, and in a more art- 

 ful manner, that which these caterpillars execute 

 without fingers, needles, or scissors. Their inge- 

 nious proceedings are well worth our attention. 



However flat the surface of a leaf may at first sight 

 appear, still on nearer inspection some portion of it 

 will he found more or less approaching to a curve ; as 

 soon as this is discovered by the caterpillar it hastens 

 to avail itself of this advantage, and, with great art 

 and dexterity, proceeds to increase the natural cur- 

 vature. Its head may be seen vibrating like a pen- 

 dulum, from the edge towards the principal nervure 

 of the leaf. After two or three hundred of these 

 movements, the part on which the insect has been 

 operating may be observed to be rolled up. The 

 mechanical means by which this effect has been 

 produced appears in the shape of a silken cord (fig. 

 5). Two or three of these cords, fixed at a mode- 



rate distance from one another, will bend the leaf 

 considerably beyond the points to which they are 

 attached. Although Reaumur watched a caterpillar 

 while at work, and saw the leaf roll up under his 

 eyes, still he was unable to satisfy himself as to the 

 causa which brought about this effect It can 



