SUSPENSION OF PUP^. 279 



after being removed from their suspensories, also roll 

 about and manifest great uneasiness.* 



But this is only one mode by which chrysalides 

 are suspended ; for nature, rich in variety, has taught 

 others to employ a different mechanism, and consi- 

 derably more complicated, not only fixing themselves 

 by the tail, but throwing around their body a girdle 

 of silk, which binds it firmly to the spot selected, and 

 frequently in a horizontal position. Amongst those 

 the caterpillars of the pretty butterflies called by col- 

 lectors hairstreaks (TheclcF, Fabr.), are remarkable 

 both for their resemblance in shape to the common 

 woodlouse (Oniscus), and for their singular proceed- 

 ings. In order to construct a silken cincture around the 

 middle of its body, after it has secured itself at the tail, 

 the Thecia draws back its head, and pushing out its 

 spinneret on one side, forms an arched thread by pass- 

 ing it over to the other side. It then insinuates its 

 head under this thread, and pressing the fore part of 

 its body down as closely as possible, it contrives to 

 place the girth over its middle. This circumstance 

 is the more remarkable when it is considered that the 

 silk is so fine as scarcely to be distinguishable to the 

 eye, and that the back of the caterpillar over which it 

 has to pass is thickly bestudded with spines. The 

 caterpillars, indeed, of this whole family (Lyccenceidce, 

 Leach), which includes our splendid blue and copper 

 butterflies, seem to follow the same process, repeat- 

 ing it from thirty to fifty times, in order to strengthen 

 the band. As the caterpillars of the family just men- 

 tioned are but seldom found, those who are desirous 

 of observing the formation of the cincture of a chry- 

 salis may readily gratify their curiosity by watching 

 a brood of any of the native white butterflies, as 

 those of the cabbage or of the hawthorn {Pieiis 

 CratcBgi, Stephens). A caterpillar of this kind, 

 ■•' Bonnet, CEuvies, vol. ii. p. 109. 



r2 



