STATES OF INSECTS. 271 



themselves, that the side on which they lie may not be 

 flattened ; afterwards by far the majority merely wriggle 

 or twist their abdomen when touched, or in any way in- 

 commoded or disturbed. We learn from De Geer, that 

 the pupa of the ghost-moth (Hepialus Humuli), the co- 

 coon of which is more than twice the length of the chry- 

 salis, moves in it from one end to the other a . Bonnet 

 observed one of a moth (perhaps Lasiocampa Quercus), 

 which alternately fixed itself at the top and bottom of its 

 spacious and obliquely-fixed cocoon ; descending slowly, 

 but ascending as quickly, and almost in the same manner, 

 as a chimney-sweeper in a chimney b . The pupa of the 

 weevil of the water-hemlock (Lixus paraplecticus) will 

 move from one end of the interior of a branch to another 

 by means of its adminicular aided by the motion of its 

 abdominal segments c . But the most locomotive of pu- 

 pae of the second division are those of gnats, and many 

 Tipulidans, which pass this state in the water. These 

 will move from the bottom to the surface, and back again, 

 with great facility and velocity. I have before mentioned 

 several other motions of pupae d , which I shall not repeat 

 here, by which they extricate themselves from their seve- 

 ral places of intermediate repose, before they leave the 

 puparium : if the imago were to be disclosed in the in- 

 terior of a tree, or in the earth, its wings would be ma- 

 terially injured in forcing its way out. The object of 

 several of the above motions may be to alarm insects that 

 might attack these defenceless beings. The twirling mo- 

 tion in particular, formerly noticed e , in some species, by 



a De Geer i. 490. t. vii./. 3, 4. b CEuv. ii. 1. 



c De Geer v. 2.29. d Vol. II. 300—. 



e Vol. II. 298—. 



