172 ANATO3IV. 



separated by broader membranous spaces, these he has figured as 

 round in Cetonia aurata * : in Lamia lextor he even saw small spots 

 between the windings, whereby the vessels of this insect appeared 

 punctate. When an air-vessel sends off a branch the space between 

 the two successive convolutions then widens, and the branch com- 

 mences with its own spiral filament (PL XXII. f. 11), whereas that of 

 the stem continues uninterruptedly; but if a trachea divides into two 

 equal branches, each begins with its own new spiral filament, and that 

 of the stem terminates at the point of division. These spiral filaments 

 of the tracheae may be considered as analogous to the cartilaginous 

 rings in the windpipe of the superior animals, although these are sepa- 

 rated from each other, and connected only by their softer parts. But 

 this fibrous layer of the muscular membrane in the vessels has the 

 same function, for the contraction of the spiral filament straitens the 

 tracheae, and thus helps to promote expiration, whilst its succeeding 

 expansion facilitates the inspiration by opening a larger space in the 

 vessel for the admission of air. The cartilaginous rings of the wind- 

 pipes of the superior animals fully accomplish this last purpose, and 

 they thereby distinguish themselves from the tracheae of insects. 



The innermost third membrane, which Lyonnet, Marcel de Serres, 

 and Straus-Durckheim admit, but Sprengel denies, is, according to the 

 investigations of the former, a smooth, transparent, delicate, mucous 

 membrane, and, as it were, a continuation of the exterior epidermis, 

 with which it also stands in connexion at the orifice of the spiracles. 

 The spiral filament lies closely adhesive to it, so that upon a rupture 

 of the vessel its remains hang affixed to the detached spiral thread, 

 whence Sprengel prefers considering it as a connecting membrane be- 

 tween the spiral fibres rather than as a distinct layer. But the fact of 

 this innermost membrane peeling off when caterpillars moult, or pass 

 from the larva to the pupa state, and that in place of it a new one is 

 formed beneath, speaks distinctly in favour of its being considered 

 as a peculiar and a separate one. 



This anatomical structure of the air-vessels is found precisely the 

 same in all the orders, and although their form is subject to many varia- 

 tions, yet their structure but very seldom participates in this difference. 

 This participation of the structure in the difference of form is main- 

 tained by Straus and Marcel de Serres to be found in the air bags of the 



* ConniicnUir. dc Pad. PI. II. f. 19. 



