ROACHES AND OTHER ANCIENT INSECTS 



character, though not a distinctive one, since modern 

 dragonflies (Fig. 58) and mayflies (Fig. 60) likewise 

 keep the wings extended when at rest. 



The question of how insects acquired wings is always 

 one ot special interest, since, while we know perfectly 

 well that the wing of a bird or of a bat is merely a modi- 

 fied fore limb, the nature of the primitive organ from 

 which the insect wing has been evolved is still a mystery. 

 The Paleodictyoptera, however, may throw light upon 

 the subject, for some of them had small flat lobes on the 

 lateral edges of the back plate of the prothorax, which in 

 fossil specimens look like undeveloped wings (Fig. 56). 

 The presence of these prothoracic lobes, occurring as they 

 do in some of the oldest known insects, has suggested the 



Fig. 56. Examples of the earliest known fossil insects, called the Paleodic- 

 tyoptera, having small lobes (a) projecting like wings from the prothorax 

 A, Slenodictya lobata (from Brongniart). B, Eubleptus daniehi (drawn from 

 specimen in U. S. Nat. Mus.): Ti, 72, 7j, back plates of three thoracic segments 



idea that the true wings were evolved from similar flaps 

 of the mesothorax and metathorax. If so, we must pic- 

 ture the immediate ancestors of the winged insects as 

 creatures provided with a row of three flaps on each side 

 of the body projecting stiffly outward from the edges of 

 the thoracic segments. Of course, the creatures could 

 not actually fly with wings of this sort, but probably 



[9H 



