322 INSECT TRANSFORMATIONS. 



sion. When the insect, therefore, becomes desirous 

 of escaping from its prison, it blows out the extensile 

 part of its head like a bladder, alternately pusliiiig it 

 forward in the form of a muzzle, and swelling it out 

 at the sides in the form of a ball, till it succeeds in 

 rupturing the pupa case. As this envelope is too 

 opaque to see the process distinctly on the outside, it 

 is necessary to open the pupa just before its transform- 

 ation, when the movements become obvious. The 

 same mechanism occurs in the pupa; of some of the 

 fibrous gall flies {Tephritcs), for the purpose of dis- 

 severing the woody fibres wliich imprison the insects. 

 In the instance of the thistle gall fly ( Tephritis Canlui)^ 

 Reaumur found that those kept in his study often be- 

 came too rigid for the insects to force their passage, 

 and after making repeated efforts they gave up the 

 task in despair and died.* In the open air this acci- 

 dent is prevented by the rain moistening the galls. We 

 have more than once had occasion to make the same 

 remark in the woody galls, such as the bedeguar of 

 the rose, in which the flies have to gnaw for them- 

 selves a passage, but which they cannot always effect 

 when the galls are kept through the winter in a dry 

 room.! 



In another genus of flies, the pupa does not make 

 use of its head, but turns round and employs its tail to 

 force a passage. This may be observed in the pupa 

 formed from the rat-tailed maggot of the commou- 

 sewer fly {Eristalis /e?irt.r, Fabuicius), which was ob- 

 served by R'aumur to push off the lid of its pupa case 

 by means of its tail. 



The caterpillar of the clear-wing hawk moth 

 {^gcria asiliformis, Stephens), before going into 

 pupa, gnaws away the wood of the poplar tree, 

 where it is lodged, till it leaves only a plate of it as 



♦ Reaumur, iv, Mem. 8. t J. R. 



