522 INVERTEBRATE MORPHOLOGY. 



lated are the Neuroptera with Thysanuriform larvae, probably to be re- 

 garded as a group which has undergone a development parallel to that of 

 the Ephemeridae and Odonata, the relationship being traceable back to an 

 ancestor common to it and the Ephemeridae. To this group may also be 

 referred the Plecoptera. 



A fourth group includes those forms in which the larvae are provided 

 with prop-legs, secondary forms in which all indications of the Thysanurid 

 ancestors have disappeared. Of such forms the Panorpata show relation- 

 ships on the one hand with the Ephemerid group, and somewhat closely 

 related are the Trichoptera, whose entire organization points to a close 

 affinity with the Microlepidoptera. From the primitive Microlepidoptera 

 two lines of descent are probably to be traced, one leading to the Macro- 

 lepidoptera and the other to the primitive Hymenoptera, the resemblance 

 between the larvae and the mouth-parts of the Tenthredinidae, and those 

 of the Microlepidoptera being very striking. 



The two remaining orders, the Coleoptera and Diptera, are very highly 

 specialized, both being holometabolic, and the temptation is to look for 

 their ancestors in forms with a similar metamorphosis. This temptation 

 may be justified in the case of the Diptera, whose larvae are the most 

 modified of all, and it is not impossible that they have descended from 

 primitive Hymenopteran ancestors, their nearest existing relatives being 

 the Tenthredinidae, whose sluglike larvae, suggest not a little the least 

 modified Dipterau maggots. With the Coleoptera, however, the case is 

 different, and it seems more probable that their holometabolism has been 

 acquired quite independently of that of the other holometabolic orders. 

 The larvae of some beetles, notably those of the Coccinellidae, are markedly 

 Thysanuriform, and prop-legs do not occur in the order. To which of the 

 groups they are to be referred it is very difficult to say, though the mouth 

 parts and the arrangement and structure of the wings in the adults point 

 to an affinity with the Orthoptera. 



Granting a descent of the Pterygota from wingless ancestors, it becomes 

 an interesting problem to discover the origin of the wings. Attempts have 

 been made to show that they are modified tracheal branchiae, a theory 

 whicli necessitates the derivation of the Pterygota from aquatic ancestors. 

 Such a derivation, however, is unsupported by any evidence at present at 

 our disposal, it being much more probable that the immediate ancestors of 

 the Pterygota were terrestrial, just as Campodea is to-day. The wings 

 arise in the embryo as dorsal outpouchings of the meso- and metathorax, 

 tracheae later pushing out into them, and transient indications of out- 

 l><mchings of the prothorax also occur in some embryos. It has been sug- 

 gested that primarily the wings were platelike outgrowths of the thoracic 

 segments which served to break the fall and increased the distance trav- 

 ersed by jumping Insects, and in support of this view the fact may be 

 mentioned that many Apterygota are saltatorial. The limitation of the 

 wings to the meso- and metathorax may stand in some relation to the 

 centre of gravity of the body. 



