Palmén adds that most dipterous larvæ are amphipneustic; Cecidomyia, the Mycetophilidæ, Bibionidæ, and Stratiomys are typically peripneustic. (p. 92.)
Moreover, a single insect, as Sialis, may be apneustic as a larva, peripneustic as a pupa, and holopneustic in the imago stage.
[74]. Mr. J. W. Folsom, who has made the accompanying sketch of the nymph of Euphæa splendens in the Cambridge Museum, finds only seven pairs of gills, there being no traces of them on segments 1, 9, and 10. A stout trachea, he writes us, enters the base of each gill, and subdivides into several long branches, which course along the periphery. Hagen in his original account said there were eight pairs on segments 1–8 respectively.
[75]. Harris, Correspondence, p. 226, Pl. III., Fig. 7.
[76]. Nusbaum’s view has been questioned by Heymons, who, from his studies on the embryology of the cockroach (Periplaneta and Phyllodromia), Forficula, and Gryllus, concludes that the ectodermal ends of the sexual outlets owe their origin to an unpaired median hypodermal invagination, and that it is quite doubtful whether the ectodermal portions of the sexual passages of insects were ever paired (p. 104). On the other hand he appears, even throwing out the case of Ephemera, to have overlooked Nassonow’s discovery of paired outlets in the young of Lepisma.
[77]. Acta Acad. German., xxxiii, 1867, No. 2, p. 81. Quoted by Dr. Sharp, Insecta, p. 142.
[78]. Journ. Morph., iii, Boston, pp. 299, 300.
[79]. Proc. Boston Soc. Nat. Hist., xi, pp. 88, 89.
[80]. In the following general account of the embryology of insects, I have closely followed the admirable arrangement and description of Korschelt and Heider, in their Lehrbuch der vergleichenden Entwicklungsgeschichte der wirbellosen Thiere, pp. 764–846, often translating their text literally, though not omitting to state the results of other writers.
[81]. Korschelt and Heider state that no cellular embryonal membranes are present in Synaptera, Uljanin finding none in the Podurids. In the embryo of Isotoma walkerii we, however, observed a membrane which we compared to the larval skin of many Crustacea, and both Sommer and Lemoine have detected in eggs of the same group a cuticular larval skin which is provided with spines for rupturing the chorion. The amnion is also wanting in Proctotrupids (Ayers), and is rudimental in Muscidæ (Kowalevsky, Graber), in viviparous Cecidomyidæ, according to Metschnikoff, who also states that in certain ants of Madeira the envelopes are represented only by a small mass of cells in the dorsal region.