302 INSECT TRANSFORMATIONS. 



issuing by jets, and a few smaller bubbles over the 

 body of the chrysalis, probably from not taking the 

 precaution afterwards suggested by Bonnet, of 

 moistening it before immersion.* 



Upon exauihiing the structure of the spiracles, 

 M. Reaumur farther discovered that their mouths are 

 furnished with ciliary valves, which are shut when 

 the pupa is plunged into- water, but opened again 

 when it is taken out. This circumstance accounts 

 for the swelling of the body under the exhausted 

 receiver of the air-pump, for it may be supposed the 

 animal would closely shut the valves when it felt the 

 air forcibly extracted from its body. It is remarkable, 

 however, that tliough the shutting of the valves pre- 

 vents the entrance of water, it has not the power of 

 excluding oil ; because, as Reaumur conjectures, 

 it is not in the ordinary course of its nature exposed 

 to such an accident, and therefore Providence made 

 no provision for it.'j" It may be recollected that the 

 larva? of the cheese-fly and of blow-flies are provided 

 with a sort of valve with which they can cover such 

 of their spiracles as become immersed in any greasy 

 matter.J 



a ■ b 



Spiracles of pupae, a, the valve open ; i, tlic same shut. 



In the aquatic pupa of the ringed China-mark- 

 moth {Hydvocampa stratiotata, Stephens) De Geer 



* Bonnet, CEuvres, vol- iii, p. 39, &c. t Mem. i, 407, Ac. 

 t See page 265. 



