“The study of the tracheæ is closely connected with that of the veins (nervures). It is well to guard against the error of Verson, who mistakes for these last the large tracheal branches of the wing. This confusion is easily explained; it proves that Verson had, with us, recognized that the secondary system is, in the larva, exempt from all respiratory function. Landois thought that the pupal period was the time of formation of the veins. It seems to me probable that they are derived from the sheath of the peritracheal spaces.” (Gonin, pp. 30–33.) [Illustration: FIG. 147.—Left anterior wing of a larva 3 days before pupation. The posterior part is rolled up: _st_, prothoracic stigma; _tr. i._, internal tracheal trunk; _tr. e._, _tr. e.′_, external tracheal trunk; _p_, cavity of a thoracic leg, with the imaginal bud _b_.—After Gonin. ] The appearance of the wing-germs in the fully grown caterpillar, as revealed by simple dissection, is shown at Fig. 146; Fig. 147 represents a wing of a larva three days before pupation, with the germ of a thoracic leg. [Illustration: FIG. 148.—Graber’s diagrams for explaining the origin and primary invagination of the hypodermis to form the germs of the leg (_b_), and wings (_f_, _A-C_), and afterwards their evagination _D_, so that they lie on the outside of the body. _E_, stage _B_, showing the hypodermal cavities (_f_) and stalks connecting the germs with the hypodermis (_z_).—After Graber. ] [Illustration: FIG. 149.—Section lengthwise through the left wing of mature larva in _Pieris rapæ_: _t_, trachea; _hyp_, hypodermis; _c_, cuticula.—After Mayer. ] A. G. Mayer has examined the late development of the wings in _Pieris rapæ_. Fig. 149 represents a frontal section through the left wing of a mature larva and shows the rudiment of the wing, lying in its hypodermal pocket or peripodal cavity. How the trachea passes into the rudimentary wing, and eventually becomes divided into the branches, around which the main veins afterwards form, is seen in Figs. 144, 147, 159. The histological condition of the wing at this time is represented by Fig. 151, the spindle-like hypodermal cells forming the two walls being separated by the ground-membrane of Semper.