78 INTRODUCTION. 



this the animal pushes its hinder pair of pro-lefgs, 

 which immediately get entangled among the threads 

 by the small hooks which project from a part of the 

 outer eA]ge of the foot. (Plate II. fig. 2.) The ante- 

 rior part of the body is then permitted to fall down, 

 and it therefore hangs in a vertical position, with the 

 head lowest. Soon after, it begins to bend the head 

 upwards, in such a manner that the convexity of the 

 curve is formed by the back : it retains this position 

 for a considerable time, then allows the head to fall 

 down, and again bends itself, always taking care to 

 do so in one direction. After this process has beeu 

 continued for some time, occasionally not less than 

 twenty-four hours, and in one instance mentioned by 

 Reaumur, it lasted for two whole days, the skin rends 

 in the hack, and a portion of the chrysalis projects 

 from the opening. The latter is gradually enlarged 

 by the pressure and swelling of the chrysalis, which 

 acts as a wedge, both in splitting the skin and in 

 pushing it upwards to the tail. By alternate contrac- 

 tions and expansions, the head, or lower portion of the 

 chrysalis, becomes wholly disengaged ; and the skin of 

 the caterpillar, now dry and shrivelled, is pressed toge- 

 ther into a small bundle which surrounds the tail of 

 the chrysalis. This, however, is still the only means of 

 support, and the difficult task which yet remains for 

 the chrysalis to perform, is to extricate itself from the 

 skin, and suspend itself to the silken mooring, which is 

 now considerably above it. In order to accomplish 

 this, which seems to require an effort beyond the 



