RESPIRATION OF MnSOGOMPHUS- 



427 



my North American work studies of several Somh African 

 ( )donata. I there discussed a Gomjjhine nymph, which I took to 

 ])e Mcsogoiuphiis. Later work has shown me that this identifica- 

 lion was not correct, and that the nymph in question was most 

 prohably Podogompluis. Having, liowever. material which is 

 almost surely Mesogomphus, I studied it, and have embodied the 

 results in this paper. 



The nymphs of all the Anisoptera, or dragonflies proj^er, 

 breathe by a means of a highly specialized respiratorv rectum. 

 The rectum is provided with tracheal gills, from which the 

 oxygen dissolved in the water passes by diffusion into the 

 tracheae of the insect. The gills are curious in that they are not 

 l)lood-gills as in most aquatic animals, but i)urely tracheal gills. 

 The exchange is not of dissolved gases between water and blood, 



Fig. 2. — Dorsal View. 



but uf oxyiicn inwards and carbon dioxide outwards. l)etween 

 tlie water and the air in the tracheae. 



Figure i shows the nymph in question. 



h'igure 2 shows the appearance of the caudal part of the 

 abdomen with the dorsal wall removed. The large tracheal 

 trunks. DD, correspond to the dorsal tracheae of most insects, 

 but are much enlarged. As in many nymphs of Odonata, the 

 fat-body is in the specimens which I studied, closely adherent to 

 the tracheal trunks. From these trunks there pass to the rectum 

 the many small branches, BBB. These divide and re-divide as 

 the\' ])ass to the rectum. The dorsal trunks loop back into a 

 similar pair of ventral trunks, likewise an enlarged form of the 

 usual trachere otf the insect body. At the point where they loop 



