1918.] NATURAL SCIENCES OF PHILADELPHIA. 101 



pigmented areas between these folds could be found. The extensive 

 folds of spongy or reticulate tissue, covered on the inside by a thinner 

 epithelium, and extending from about the middle of segment IX to 

 the end of segment X, also seem destitute of tracheoles. 



In view of the above facts it seems improbable that any diffusion 

 takes place between gases in the rectal tracheoles and those in the 

 water in the rectum. This conclusion is strengthened if the position 

 of the rectal tracheal twigs in Mecistogaster is contrasted with the 

 position of the tracheal gill capillaries in Anisopterous larvae, where 

 diffusion of oxygen is known to take place. The gill capillaries 

 in Anisopterous larvae (Tillyard, 1915, p. 425, Figs. 3 and 5; also 

 1916) lie "in a thin layer of very flattened cells," the gill epithelium. 

 They lie, too, "fairly close up" to the chitinous cuticle of the gill. 

 They are very numerous, excessively fine, and each forms a complete 

 loop. In Mecistogaster the tracheal twigs are separated from the 

 chitinous lining of the gut lumen by a thick epithelium, they are 

 not excessively fine, do not form loops, and are not numerous. 

 They are absent, except for one small area (sections 60 to 80), in 

 segment VII, in which, together with segment VIII, lies the gill 

 basket of Anisopterous larvae (Tillyard, 1916, p. 139). 



Nevertheless, there must be other means of respiration in 

 Zygopterid larvae than that through the caudal gills. Calvert 

 (1911 b, p. 455) found in Mecistogaster modestus, "as in most, perhaps 

 all, Zygopterous larvae, the caudal gills may be lost without inter- 

 fering with growth, development and transformation." Mr. Bodine 

 (1918), has recently succeeded in keeping Zygopterid larvae, from 

 which the gills had been removed, alive for three and four days 

 when confined at the bottom of a Mason jar filled with stagnant 

 water. 



If we exclude respiration through the integument, the only 

 alternative seems to be some sort of rectal respiration. In support 

 of the latter we have Calvert's observations on living larvae of 

 Calopteryx maculata, Hetaerina americana, and Argia moesta putrida 

 (Calvert, 1915, pp. 437-442). He found regular rhythmic pulsa- 

 tions of the rectum, with inhalations and expirations of water, in 

 these forms may be carried on for long periods. Tillyard (1916, 

 pp. 410-413) found regular rectal pulsations in newly hatched 

 larvae of Austrolestes leda up until they were half an hour or more 

 old, when the movements became intermittent. 



If rectal respiration occurs in the larvae of Mecistogaster modestus 

 the only areas which seem morphologically suitable for an exchange 



