32S INSECT ARCHITECTURE. 



thus in some measure impenetrable from without, is 

 readily opened from within; and when the moth 

 issues from its pupa case, it easily makes its way 

 out without either the acid or eye-files ascribed to the 

 silk-worm. The elastic silk gives way upon being 

 pushed from whhhi, and when the insect is fairly 

 out, it shuts again of its own accord, like a door 

 with spring hinges, — a circumstance which at first 

 puzzled Roesel not a little when he saw a fine large 

 moth in his box, and the cocoon apparently in the 

 •same state as when he had put it there. Another 

 naturalist conjectures that the converging threads are 

 intended to compress the body of the moth as it 

 emerges in order to force the fluids into the nervures 

 of the wings; for when he took the chrysalis pre- 

 yiously out of the cocoon, the wings of the moth never 

 expanded properly.* Had he been much conversant 

 with breeding insects, he would rather, we think, 

 have imputed this to some injury which the chrysalis 

 had received. We have witnessed the shrivelling of 

 the wings which he alludes to, in many instances, and 

 not unfrequently in butterflies which spin no cocoon. 

 The shrivelling, indeed, frequently arises from the 

 want of a sufficient supply of food to the caterpillar 

 in its last stage, occasioning a deficiency in the fluids. 

 The elasticity of the cocoon is not peculiar to the 

 emperor-moth, A much smaller insect, the green 

 cream-border-moth {Torlrix chlorana) before men- 

 tioned (page 170), for its ingenuity in bundling up 

 the expanding leaves of the willow, also spins an 

 elastic shroud for its chrysalis, of the singular shape 

 of a boat with the keel uppermost. Like the cater- 

 pillar of Pip-alis sirigulalis (page 198), whose build- 

 ing, though of different materials, is exactly of the 

 same form, — it first spins two approximating walls 



* 3Iejnecken, quoted by Kirby and Fpence, iii. 280, 



