Séance du 17 Novembre 1920 m 



during llie larval period as lour small prolongations of 

 Ihe posterior edges ol" the nieso- and melathorax. 



Thus the wing is eoniposed ol'lwo cliilinous laminae 

 (eutieles) joined logelheral Ihe edges and ol' two eellular 

 layei's corres[)()nding to (hose laminae acting as matrices 

 and I'orming the interstitial tissue. The outer laminae; 

 heing impermeahle in Knipusa it is necessary to make 

 a tew incisions in order to allow a stain to reach the 

 interstitial tissue. 



Seen IVom the sui'lace the interstitial tissut' appears 

 composi'd ola very large numher of starred cells, reti- 

 culated hy anastomosis. Although these cells are ofaii 

 epithelial nature, in some ways they resemhle the starri'd 

 connecli\e tissue of \'ertebrates. In the neighbourhood 

 of the \eins. the chitinous laminae suddenly separated 

 sui-round cylindrical channels enclosi'd hy cells. 

 These channels can he divided into two classes: those 

 corresponding to princii)al or longitudinal veins, arran- 

 ged I'anwise, and those corresponding to secondary, cross 

 or connecting veins. ( Mgs. Hi and 14). 



Tln' piincipal channels contain a trachea oT uniform 

 section (diameter 00-71) p.) without any dilatations, con- 

 tainiug a s|)iral thread. This thread, sometimes diliicult 

 to see, has been identilied without possibility ol' doubt 

 in the wing stump shown in lig. 7. 



The channel itself is a relatively wide (lOi)-!')!) u) 

 peritracheal sheath, built of cells and along which the 

 blood flows. This siructural detail, important in our 

 present investigation, was already observed by Ni:\\ I'ocr 

 (1(S;^9) as will be seen from the following extract (pioted 

 from Packaud (1898 p. 410): 



" As staled bv Xi:wport, the so-called veins or ner- 



