The wings.—The imaginal buds of the wings do not participate in the larval moults. Gonin has observed, contrary to Dewitz, that their surface only forms a cuticle towards the end of the last larval stage.
The network of fine tracheæ of the wing-bud is drawn out at the time of pupation with the internal cuticle of the large tracheæ. The permanent tracheæ of the wing have already appeared at the time of the 3d moult under the form of large rectilinear trunks, the position of which corresponds afterwards to that of the veins, but they are not filled with air until the time of pupation. There are from eight to ten of these tracheæ in each wing (Fig. 159), and they give rise in the pupa to a new system of fine tracheæ (tracheoles) which replaces that of the larva. (For further details the reader is referred to pp. 126–137.)
Development of the feet and the cephalic appendages.[[114]]—In the apodous larvæ of Diptera and Hymenoptera the rudiments of the legs are, like those of the wings, developed within hypodermal sacs; at times they remain there up to the end of larval life, but sometimes they appear early at the surface. This origin of the legs, thanks to Weismann, Künckel, and Van Rees, is well known in the Diptera; in the Hymenoptera it has been proved to be the case with ants by Dewitz, and in Encyrtus by Bugnion. As for the Lepidoptera our knowledge that the legs of the imago arise from the six thoracic legs of the caterpillar, up to the date of Gonin’s paper has not been in advance of that of Malpighi and Swammerdam.
Réaumur, moreover, was supposed to have furnished the proof, having from his experiments concluded that “if the legs of the pupa appear longer and larger than those of the caterpillar wherein they were contained, it is because they were folded and squeezed.” (8e Mém., p. 365.)
This explanation of Réaumur’s has been generally accepted. Graber (Die Insekten, p. 506) accepted it, after examining microscopic sections of the legs, and Künckel averred that “Réaumur, having, in certain caterpillars, completely cut off one of the thoracic legs, had concluded that the butterfly which came from it lacked the corresponding member.” (Rech. sur l’org. et dév. des volucelles, p. 160.)
Newport, it is true, denied this disappearance of the legs, but did not wish to put himself in opposition to received ideas, and supposed that the member cut off was partly reformed in the imago.
Künckel believes that he has found a better solution in his theory of histoblasts or imaginal buds; in his opinion, “Réaumur and Newport are both right,” but “when Réaumur cut off a caterpillar’s leg, he at the same time removed the histoblast, the rudiment of the leg of the butterfly. When Newport repeated this experiment, he simply mutilated the histoblast without completely destroying it: in the first case, the adult insect was born with one leg less; in the second case, it appeared with an atrophied foot.”
“So ingenious an explanation,” says Gonin, “is not necessary.” To prove that the experiments of the two savants are not contradictory, it would have been sufficient to cite, as Künckel did not do, the exact words of Réaumur, for he having cut from a caterpillar “more than half of three of the thoracic legs on the same side,” says he found that the chrysalis had “the three limbs on one side shorter than the corresponding ones on the other side.” The same operation repeated on a somewhat younger caterpillar again showed in the chrysalis three maimed limbs, “so that they could not be said to be entirely absent. These results are like those of Newport; the interpretation only was faulty, as we shall attempt to prove.”
The real relations of the adult legs to the larval legs are thus shown by Gonin.
“If we carefully strip off the skin of a caterpillar near the time of pupation (Fig. 608), we see that the extremity only of the legs of the imago is drawn out of the larval legs; the other parts are pressed against each side of the thorax: near the ventral line a small pad represents the coxa and the trochanter; the femur and the tibia are distinctly recognizable, but soldered to each other and only separated by a slight furrow; they form by their union a very acute knee or bend. The femur is movable on the pad-like coxa, the tibia continues without precise limits with the extremity concealed in the larval legs. The three divisions of the latter do not appear to have any relation with the live joints of the perfect state. Under the microscope the rudiment appears very strongly plaited at the level of the tarsus, much less so in the other regions. A large trachea penetrates into the femur with some capillaries; reaching the knee it bends into the tibia at a sharp curve, but does not become truly sinuous in approaching the extremity. It is then the tarsus especially which is susceptible of elongation; it may, on being withdrawn, give rise to the illusion that the whole organ is disengaged from the larval leg.