Séance du 17 Novembre 1920 87 



Halten. Ho wrote : " Tlic wing reaclu'd between my 

 lingers the lull size it would have aecpiired il it had 

 developed in its natural position ". 



It is probable that this expeiinieni would give the 

 same result il" tried on an Kmi)usa. 



'2. The seeond I'aetor is the lieeing of the slumps. 

 The winglets, whieh until Ihe moult had been eom- 

 j)ressed in tlu ir sheaths, are immediately alter free and 

 loose. It is owing to the withdrawal of the compressing 

 shealh that the blood can How into the wings. It is, 

 however, important to note that the Irecdom gained 

 at t-ach of the larval moults produces no visible 

 change in the apj)earance of the stumps or sheaths. 



'A. in my opinion, the new jjosilion taken up by 

 the stumps is a thiiil factor. Hefoic the last moulh 

 these organs lay along the body directed backwards, 

 but after the moult are turned outwards, oblitjuely^ 

 (see tigs. .') and 4). It is possible that this change of 

 diieclion facilitates the opening of the peritracheal 

 sheaths in the neighbourhood of the blood channels 

 into which they open. 



4. Finally there are, as deteiininating causes, Ih*' 

 blood ))ressure and the circumstances which produce it. 



When observed under a lens after its complete 

 liberation, the Kmpusa is seen to be ^ubjected to 

 regidar contractions owing to which the body walls 

 alternately close up and dilate throughout the abdo- 

 men and thorax. In the specimen under observation, 

 these contractions occur at the rale ol about 90 per 

 iuinule. The wing slumj)s Ihemselves are raised by 

 rythmical beats whicli, incidentally, make focussing 

 very ditTicull. (Comparable with the inspiratory motions 



