=Pieris.=—The rudiment is concealed in a sac attached to the hypodermis by a short pedicel. Destruction of the partition and its replacement by a part of the thoracic wall by means of the imaginal epithelium. =Musca.=—The pedicel is represented by a cord of variable length, whose cavity may be obliterated (Van Rees). The imaginal hypodermis is substituted for the larval hypodermis, which has completely disappeared, either by desquamation (Viallanes), or by histolytic resorption (Van Rees). =Extension of the wing; drawing out of the tracheoles.=—When it is disengaged from the cavity, the wing greatly elongates and the creases on its surface are smoothed out; the blood penetrates between the two walls, and the cellular fibres, before relaxed and sinuous, are now firmly extended. Of the two tracheal systems, the large branches are sinuous, and they are rendered more distinct by the presence of a spiral membrane; but the two tunics are not separated as in the other tracheæ of the thorax; moreover, the mouth choked up with débris does not yet communicate with that of the principal trunk. The bundles of tracheoles on their part form straight lines, as if the folds of the organ had had no influence on them. As they have remained bound together, apart from the chitinous membrane of the tracheal trunk, they become drawn out with this membrane, at the time of exuviation, _i.e._ of pupation, and are drawn out of the neighboring spiracle. [Illustration: FIG. 146.—Full-grown larva of _Pieris brassicæ_, opened along the dorsal line: _d_, digestive canal; _s_, silk-gland; _g_, brain; _st I_, prothoracic stigma; _st IV_, 1st abdominal stigma; _a_, _a′_, germs (buds) of fore and hind wings; _p_, bud of prothoracic segment;—those of the third pair are concealed under the silk-glands; _I–III_, thoracic rings.—After Gonin. ] “This is a very curious phenomenon, which can be verified experimentally: if we cut off the wing, while sparing the larval integument around the thoracic spiracles, we preserve the two tracheal systems; the same operation performed after complete removal of the larval skin does not give the secondary tracheal system.” (Gonin.) Deceived by the appearance of the tracheoles while still undeveloped, Landois and Pancritius, who have not mentioned the drawing out of the capillaries of the larva, affirm that they are destroyed by resorption in the chrysalis.