86 CLARENCE HAMILTON KENNEDY Vol. XXII, No. 3 



off a slender branch direct to the gangHon of the segment 

 (13) the ganglionic trachea, gt. This has a tip which runs along 

 the sternum caudad from the ganglion, (14) the sternal trachea, st. 



As this hypothetical unit is derived from the tracheal 

 system as it appears in the less modified abdominal segments 

 these terms are directly applicable to the parts in Fig. 1. In 

 this figure it will be seen that the anterior and posterior dorsal 

 connective have united as they alternate along the dorsum into 

 a large dorsal trunk. In Lestes it is especially large as it functions 

 with its mate of the opposite side, as a swim bladder, while 

 along it at regular intervals are the tips of the anterior and 

 posterior dorsal connectives. 



In the front end of the abdomen the ganglions of segments 

 1 and 2 have each moved cephalad one segment. In each case 

 the ganglionic trachea has followed its ganglion cephalad. The 

 ganglion of segment 1 fuses with that of the metathorax but the 

 elements of the tracheal supply are still recognizable. See 

 Fig. 2, which is a ventral view of the thorax. 



In each segment the anterior spiracular connective has fused 

 with the leg trachea next ahead so that in the abdomen there 

 is a delicate but complete lateral trunk. In the thorax a pair 

 of ventral or sternal trunks appear by the fusion of the sternal 

 trachece with each succeeding ganglionic trachea. This fusion is 

 not completed in the abdomen. See Fig. 2. 



In the thorax but two spiracles persist, the mesothoracic 

 and the metathoracic. Of the prothoracic spiracle, some of the 

 tracheal branches are present which would be expected with it 

 but the spiracle itself has left no recognizable trace. The 

 tracheal system of the metathorax is least modified. Here the 

 spiracular trunk is still fully developed. In the mesothorax the 

 spiracle has moved dorsad until the spiracular trunk, st, has 

 been obliterated. It is also entirely lost for the prothorax. 



The modification of the thorax due to the development of 

 the wings has profoundly changed the proportions of the other 

 thoracic branches also. In a later paper the writer will trace 

 this change from the first instar where it is very similar to the 

 abdominal tracheal system through its development to the 

 extreme modification shown in Fig. 1 which is a tenth instar 

 naiad. 



The wings start as simple pleural folds no wise different 

 from those of the abdomen. By the fourth instar those of the 



