74 CAUDAL GILLS Of ZYGOPTERIt) LARV.E, 



tvpes, there is nothing but a weakly-formed meshwork, sup- 

 ported by occasional secondary laminte, and a few branch trachea^ 

 of very small size. 



The tracheal plan of a saccoid gill is always somewhat 

 irregular, the main longitudinal tracheae being free to take a 

 tortuous course in the alveolar meshwork. As in all caudal 

 gills, two main tracheae penetrate the base of the median gill, 

 while one only passes through the base of each lateral gill. But, 

 in the median gills belonging to the Epallagince, it is usual to 

 find each main trachea giving off a strong ventral longitudinal 

 branch, so that, in nearly all cross-sections of the gill, there 

 appear to be four longitudinal tracheae in the gill. In the lateral 

 gills, likewise, the single main trachea divides into two soon after 

 entering the gill, and, fairly frequently, one of these branches 

 may again divide, so as to give three longitudinal tracheae in this 

 gill. In the median saccoid gills belonging to the Protoneurince 

 (Neosticta), one sometimes finds only two main tracheae in all 

 three gills, yet I have frequently noted four in the median, and 

 sometimes three in the laterals. These gills are so frequently 

 cast off and regenerated, that it is not easy to determine whether 

 any of this variation is due to such a cause, or whether the 

 number of main tracheie is essentially variable. 



In all cases examined by me, there are four longitudinal nerves 

 in the median gill, and two in the laterals, as shown in Text-fig. 

 11 {dn, vn). 



The two following sub-types have been studied : — 



(i.) I'he Simple Saccus (Text-figs. 12, 13). This is found in 

 the genera Pseudophcea and Diphlehia both belonging to the 

 Epallayince, and in the genus Cora (Thoririce). I have only 

 studied the gills of Diphlehia. Ris' work on Pseudophcea {2S) 

 shows that the gills in this genus are closely similar to those of 

 Diphlehia, differing principally in being even more tensely 

 swollen, and much less pointed at the tips. The gills of Cora, 

 as studied by Calvert (3), differ from those of the other two 

 genera in being sharply truncated posteriorly. 



The gills of Diphlehia (Text-figs. 12, 13) are large, cumbersome 

 sacs, which clearly impede the movements of the larva. These 



