8 FOSSIL INSECTS OF THE BRITISH COAL MEASURES. 



may have dwelt in damp eartli rather than in water, and that " the formation of 

 the larval tracheal system undoubtedly proves that this at one time was the case. 

 It may well have been so in Carboniferous times." 



Tillyard's views are well worth quoting in full, especially as they support in 

 some measure those of Pruvost : " We may picture to ourselves the giant insects 

 of Commentry as inhabiting the shores of a large, shallow, nearly stagnant 

 lake. In the muddy ooze around its borders grew forests of the Giant Mare's 

 Tail, while further back on the sandy slopes the graceful Cycads and other extra- 

 ordinary plants formed a more diversified medley. There, amidst rotting vegetation, 

 these insects lived and bred. In such almost amphibious conditions it may well be 

 that the larvae of Protephemeroidea and Protodonata first began that series of 

 adaptive changes which finally led them to adopt a purely aquatic mode of life." 



The larvae of Brodta and of other forms whose wings I describe under the 

 name of "Pteronepionites " must have lived under conditions fitted for their gradual 

 metamorphosis. The body was long, well segmented, and bore rudimentary 

 wings, which were carried well up over the thorax in an erect or semi-erect position. 

 Though rudimentary, the wings possess features which may have determined to a 

 large extent the mode of life. They are attached by broad, strong bases to the 

 thorax, and are very muscular, as shown by the stout ridges proceeding from the 

 point of attachment into the wings, and the anterior margins are also thickened. 

 The bodies with their lateral expansions of the terga are very suggestive of those 

 of the Diplopoda, and like them would offer no serious obstacle to progression 

 through rank and rotting vegetation. That these larval insects would also 

 penetrate soft muds, if necessary, in search of food is possible, since the 

 soft-bodied caterpillars of the Hawk-moths of to-day are able to enter hard soil 

 before pupation takes place. 



The stout wing-bases and the strengthened margins of the wings would 

 prevent damage to these structures as the larvae crawled about, or sought to bury 

 themselves in the soil or muds. They were essentially adapted for a ground 

 habit. Whether they were capable of an aquatic or semi-aquatic habit can only 

 be settled by a knowledge of the mode of respiration. 



Lubbock, Gegenbaur and others have adduced strong reasons in favour of an 

 aquatic origin of the insects, and in the Carboniferous types we should naturally 

 expect that the original habits had not had time to undergo any great modification. 

 Larval wings of the " Pteronepionites " type must have been living structures in 

 which metabolism was active, and very unlike the dried membranous sac-like 

 expansions of the adult insect. The growth of the larval wings was continued 

 throughout metamorphosis, and during this period their delicate nature, broad 

 expanse, and the thinness of the integument may have enabled them to assist in 

 the respiratory function. 



The researches of Comstock and Needham show that larval wings of recent 



