THE PHYSIOLOGY OF RESPIRATION IN INSECTS. 373 



In other preparations the pupa-sheath is seen to be con- 

 nected with the base of the inter-segmental spiracular cornu. 

 I am inclined to believe that the pupa-sheath of Weismann is 

 the result of a second ecdysis ; and, in parts, at least, the pupa- 

 sheath is distinctly formed of several layers closely pressed 

 together. It is probable, I think, that two or more ecdyses 

 occur, which are either partial or complete ; but owing to the 

 extreme thinness of the shed layers it is extremely difficult to 

 trace them. It appears to me probable that the first ecdysis 

 is only partial, and is limited by the extent of the discs at the 

 time of its occurrence, as there is no appearance of a con- 

 tinuous membrane enclosing the whole of the limbs in a single 

 sheath, a condition which could hardly fail to exist if a com- 

 plete ecdysis occurred before their evolution. The only indica- 

 tions of more than one ecdysis in the pupa are the early 

 separation of the prothoracic stigmatic cornua, and the lami- 

 nated structure of the pupa-sheath. 



The Posterior Thoracic and Abdominal Spiracles of the imago 

 are all new formations in the nymph, which had no pre- 

 decessors in either the larva or pronymph ; but they are 

 developed in precisely the same manner as the larval spiracles 

 and the anterior thoracic spiracles of the nymph ; that is, by 

 the invagination of the cellular integument, which forms the 

 spiracle and vestibule, in relation with a newly-formed tracheal 

 vessel, which is developed from one of the tracheal branches 

 distributed to the integument. 



4. THE PHYSIOLOGY OF RESPIRATION IN INSECTS. 



It has been asserted that the respiration of insects, when 

 in full activity, has the same energy as in the higher verte- 

 brates ; this statement is, however, misleading, and cannot be 

 accepted without modification. 



From the observations of Regnault and Reiset [157], it 

 appears that insects absorb a weight of oxygen equal to 

 unr oth of their body weight per hour. This estimate does not 

 differ materially from that obtained amongst moderate sized 



