86 Bulletin de la Société Entomologique d'Egypte 



arises from the Iwo thoracic Iruul^s \vliicli continue the 

 longitudinal trunks of the ahdomen towards the head (1). 

 Brought closer together on entering the thorax, the 

 longitudinal trunks give rise in the region of the mela — 

 and mesothorax, to cross hranches which can he traced 

 as far as the veins of the wings for which they are 

 intended. The hlood which flows to the wings at the 

 time of unfolding very proha!)1y follows the same path. 

 The Iracheae are, ap[)arenlly, surrounded hy lacunae, 

 ill-defined in the tiioracic region, which on entering the 

 wings ac(|uire walls com})osed of cells and continuous 

 with Ihose of the peritracheal system already descrihed. 



VH. The Causes of the Unfolding of the Wing. 



How can we explain Ihal the wing slumps, already 

 formed in the larva, go through Ihe larval stages without 

 change, and suddenly unfold at the imaginai moult ? 



There are. we think, several factors to he consi- 

 dered. 



]. 11 is lirsl necessary thai Ihe winglets should he 

 coipi)lele. H is indeed cleai' thai, in oi-der to he ahle 

 to unfold in less than an houi-, the wiiig at the last 

 moult must he already comiiletely formed and possess 

 its final superficies. In this conneclii/.i let us (piote 

 the following experiment of Ui:.u Miii, (I, 'Jnd Part, 

 1737 [).352). It concerns a huiterlly which had pisl 

 left the chrysalis. Having pulled off one of the wing 

 stumps and stretched it with his lingers longitudinally 

 and laterali V. Rkaumuu saw the slump spread out and 



1. The longiliuiinal Iriniks in the abdomen of Maniidac 

 arc peculiarly arranged, giving rise lo a series of very 

 rcLHiIar lozenges. 



