NO. 2 DRAGONFLY LARVA SNODGRASS 29 



locomotion may in part result from an intensification of the abdominal 

 expiratory movement of respiration, but the conspicuous abdominal 

 movement seen as the larva darts forward in the water is the length- 

 wise contraction of the abdominal segments from behind. The prin- 

 cipal locomotor force exerted on the inflated respiratory chamber, 

 therefore, evidently results from contraction of the bands of dorsal 

 and ventral intersegmental muscles. These muscles, as already noted, 

 are strongly developed in the larva, but they are totally lost after the 

 transformation to the adult, showing that they are of functional im- 

 portance only in the larval stage. It is possible, however, that the 

 transverse muscle and perhaps the diaphragm assist in the expulsive 

 contraction of the abdomen. 



Inasmuch as the modern anisopterous and zygopterous larvae are 

 mainly differentiated by their manner of breathing, the question natu- 

 rally arises as to how the primary odonate larva accomplished respira- 

 tion when it first took to the water. The fact that the labial mecha- 

 nism is the same in both groups of larva and is a device for catching 

 aquatic prey shows that the larvae were differeniated in their modern 

 respiratory methods after the labium was perfected as a grasping or- 

 gan. Yet, some means of breathing in the water must have been the 

 first adaptation necessary for an insect on its adoption of an aquatic 

 life. The answer to the question may possibly be furnished by those 

 modern zygopterous larvae that have tracheated gill appendages along 

 the sides of the abdomen, such as the species of Cora (fig. ii D) de- 

 scribed by Calvert (1911), or Euphaea variagata described by Ris 

 (1912), who concludes these appendages are clearly respiratory or- 

 gans since each is filled with a rich tracheal arborization. Calvert says, 

 "We have good grounds for looking on Cora and its allies as being 

 in many respects the most primitive of living Odonata." It is not im- 

 possible or even improbable, therefore, that the primary odonate larva 

 developed tracheal gills along the sides of the abdomen, and that in 

 this respect it resembled the larvae of modern mayflies. We can 

 further imagine that these gills of the early larvae of both the 

 Ephemeroptera and the Odonata were primarily abdominal styli in- 

 herited from terrestrial ancestors, such as those still preserved on 

 the terrestrial Thysanura. 



The caudal appendages of young zygopterous larvae are slender 

 processes, which, according to Tillyard (1917), become triquetral in 

 the second or third instar, and in subsequent instars are flattened to 

 form the lamellar gills (fig. ii A). In some forms, however, the ap- 

 pendages remain as slender, horny processes (B) that certainly can 

 have no respiratory function. In those larvae mentioned above hav- 



