THE WINGS OF LEPIDOPTERA 



335 



trachea Mi migrated alon<( this connection towards trachea R^ and became 

 separated from trachea Mo, receiving its air via trachea R^. At the same 

 time trachea r-m migrated along trachea Rs towards the base of the wing, 

 thus affording a more direct course for the air and becoming longitudinal 

 instead of transverse. This series of events, regarding which there is little 

 room for doubt, as all of the stages in the switching of vein Mi to vein R^ 

 exist in living Lepidoptera, throws light on the question of the significance 

 of trachea m-cu. The position of this trachea, which is closely analagous 

 to that of trachea r-m, indicates that it is a vestige of a secondary connec- 

 tion between the medial and cubital tracheae. 



A", 



Fig. 343. — Wings of Prionoxyslus robinicB. 



It is evident that trachea m-cu is not a vestige of a connection between 

 coalesced tracheae M3 and M4 and trachea Cui, because trachea M3 still 

 retains its primitive connection with the main stem of the medial trachea, 

 which would not be the case if at any time it had become connected directly 

 with trachea Cui. The connection must have been between tracheae M4 

 and Cui. The two intermediate branches of the medial trachea. Mo and M3 

 retain their connection with the main stem of this trachea, while on either 

 side a branch has become connected to another trachea, Mi to trachea Rg 

 and M.1 to trachea Cui. 



In each case the transferred branch has migrated along the secondary 

 connection towards the trachea with which it has become connected. In 

 the hind wing of Pier is rapes the base of trachea Mi has progressed a con- 

 siderable distance towards trachea R^, but is still distinct from it, arising 



