131 



escape from a lower uiid more, embryonic physical condition, we 

 may consider it as a reminiscent action, marking the successive 

 developmental halts in the kingdom, throngh which it is given to 

 some to pass, and at which it is lated that others shall perish. 



AVithiu the tAvo series of suborders of Hexapoda, a synthetic type 

 has been shown by Packard to exist in the Neuroptcra. In compar- 

 ing the Lepidoptera, a synthesis may be detected in the Bombycidae. 

 Thus the Lithosiinae resemble the Pyralidae, the Arctiinae the Noc- 

 tuidae, the Attacinae the Geometridae ; lower down the Cossinae the 

 Tineidae, though the actual interchange of the two latter must be 

 doubted, even since the discovery of such a form as Morpheis, per- 

 haps the most extraordinary form of the suborder. Professor Pack- 

 ard has compared the thoracic structure of Sthenopis with the 

 Neuropterous Polystichoetes. And, in a general comparison with 

 the Neuroptera, the Lepidoptera are seen to advance along a line of 

 parallel development. Indeed the development of all the suborders 

 is at least biserial, reflecting the progress of the Order. Thus the 

 Diurnals resemble the higher Neuroptera in the position of the wings, 

 while in the lower Neuroptera these are deflexed as in the Moths. In 

 considering the general progression of the Hexapoda, the Devonian 

 and earliest forms known seem to be Neuropterous, nor is there yet 

 sufficient evidence to prove that the common origin of Hexapoda 

 is to be carried back through suborders exclusively fossil. Yet that 

 the position of the Neuroptera suggests such a third series, which 

 is now no longer living, and which has given rise to the Orthoptera, 

 Ilemiptera and Coleoptera, and again to the Diptera, Lepidoptera 

 and Ilymenoptera, cannot be denied. And that the Lepidoptera 

 are the more recent, palaeontological evidence rather confirms, 

 while we should, not expect the Butterflies to be largely repre- 

 sented among the floA,y;erless forests of the Carboniferous period. 

 On general grounds Ave shall agree that the common origin of Tra- 

 cheata is to be sought in the Zoeaeform Crustacea as suggested by 

 Hacckel. Packard's objection, that Lcptus begins life on a higher 

 level than Nauplius can hardly lead us to reject the crustacean origin 

 of Hexapoda, a type which must have been evolved from a littoral 

 biregional ancestry. In studying the larval forms of Hexapoda Ave 

 folloAV Packard's exposition of larA'al types. The exceptional posi- 

 tion of the abdomen in the young Lachnostcrna recalls the usual 



