BY R. J. TILL YARD. 



543 



of the Y-shaped leg-trachea, and enters the wing posteriorly. 

 From this, the cubital and anal trachea? are derived. Originally 

 there was no connection between the two groups of trachea?; and 

 this primitive condition is still preserved in the Perlaria and 

 certain Cockroaches. But, in most other insects, the two groups 

 are linked up by a short connecting-loop, so that a complete 

 trachea, in the form of an arch, which I have called the alar 

 trunk (18), lies at the base of the growing wing, and sends the 

 main wing-trachea? into it. 



In the more archaic of those insects in which the alar trunk 

 is complete, the median trachea may still belong to the eosto- 

 radial group. But, as specialisation proceeds, the base of the 

 median trachea tends to move along the alar trunk posteriad, 

 towards the cubito-anal group, and finally becomes incorporated 

 in that group of tracheae. 



In the Panorpoid Complex, we have, as already stated, three 

 Orders only in which the wings are holotracheate, and three 

 in which they are merotraeheate. Let us examine the holo- 

 tracheate types first." 



Text-Fig. 36. 

 Basal Tracheation of the pupal wing in the Panorpoid Complex, a, in 

 Megaloptera and Planipennia ; A, ina hindwing of Charagia splen- 

 dens Scott, (Order Lepidoptera, fam. Hepialidae), to show the 

 double origin of trachea M. Lettering on p. 535. (See also Plate 

 xxxiii., fig. 20). 



The condition to be found in both Megaloptera and Planipen- 

 nia is that in which (Text-fig. 36, a) the alar trunk trachea is 

 complete, and M stands about half-way between the costo-radial 

 and cubito-anal groups of trachea?. 



