NO. I INSECT THORAX — SNODGRASS 43 



water during the Silurian age, and that the aquatic progenitors of the 

 Pterygota were themselves descendents of primitive terrestrial tra- 

 cheates of Ordovician times. This theory of the origin of insect 

 wings is pretty safe from destructive criticism, but for the same 

 reason little direct evidence can be shown in its favor. Bats and flying- 

 squirrels have not needed a baptism for the acquisition of wings, but 

 their wings do not contain tracheae. On the other hand, it is perhaps 

 not certain that, phylogenetically, the veins of insect wings were pre- 

 ceded by the tracheae. The Silurian insects, known only as the neces- 

 sary ancestors of post-Silurian insects, do not attest an adaptation to 

 life in the water in any of their descendents. If they were aquatic, 

 they have left no direct descendents ; and existing insects bear no 

 stamp of an aquatic ancestry. The abdominal gill lobes of nymphs 

 of Ephemerida, often cited as possible homologues of the hypothetical 

 thoracic gill lobes from which the wings might be supposed to have 

 developed, have been shown by Diirken (1907), from a study of their 

 musculature, to have nothing in common with the wings. The oldest 

 known insects are distinctly terrestrial, and they likewise give no 

 evidence of a recent emergence from a water environment, except as 

 adults from aquatic nymphs. For a more complete discussion of the 

 arguments that have been made in favor of a gill origin of the wings, 

 and of the facts indicating their paranotal origin, the student is re- 

 ferred to the paper by Crampton (1916) on this subject, in which 

 will be found also a long bibliography. 



If the gill theory of wing origin in insects is found eventually 

 unacceptable in any form, it should not be hard to believe that the 

 wing lobes in earlier stages enabled their possessors to glide through 

 the air from elevated situations, and that the lobes were thus organs 

 of sufficient importance to demand a considerable degree of recon- 

 struction in the pleura, or subcoxal chitinizations, of the segments 

 bearing them. There is no evidence that prothoracic lobes were ever 

 developed into wing-like or even movable appendages, and there is no 

 proof that they were present in the true ancestors of winged insects, 

 or that the wing lobes are their homologues in any case. Yet, con- 

 sidering the essential identity of structure between the prothoracic 

 pleuron and the pleura of the wing-bearing segments, it becomes evi- 

 dent that all the thoracic pleura must owe their specialized form to 

 some common guiding influence, and that this influence must have 

 determined the basic structure of the pterygote thoracic pleuron be- 

 fore the dorsal lobes of the mesothorax and the metathorax evolved 

 into true wings. A reasonable postulate, therefore, is that the deter- 



