386 



SCIENCE 



[N. S. Vol. XXIX. No. 740 



formations, which, however, have not originated 

 independently of the vestiges of the abdominal 

 appendages. The looseness of this dependence, 

 however, is shown in certain Noctuid caterpillars, 

 in which some of the pairs of prolegs make their 

 appearance during larval life and hence at a time 

 when the abdominal appendages have completely 

 disappeared. I can not, therefore, regard the 

 pedes spurii as primary, or even as resuscitated 

 organs, but only as secondarily adapted provi- 

 sional organs. 



This statement if applied to the ontogeny 

 ■would involve the unwarrantable assump- 

 tion that the abdominal legs of the embryo 

 disappear in the larva. This they do in many 

 cases as hollow ectodermal evaginations filled 

 with mesoderm, but they persist, nevertheless, 

 as small, flat cellular areas in the ectoderm. 

 In other cases, there is abundant evidence to 

 show that they are directly transformed into 

 the prolegs of the larva (Lepidoptera) and 

 the gonapophyses of the larva or imago 

 (Orthoptera). Where they are not thus 

 transformed directly, but first flatten out, we 

 obviously have the primordia of imaginal 

 discs, and the organs would belong to Deeg- 

 ener's sixth category. Eeverting now to the 

 absence of antennae and thoracic appendages 

 in apod larvas and their presence in the pre- 

 ceding or embryonic and the succeeding or 

 imaginal instars, we see that we have a ease 

 of precisely the same nature as that of the 

 abdominal appendages, though clearer on ac- 

 count of the larger size of the cephalic and 

 thoracic structures and their imaginal discs. 

 But any such ontogenetic conclusion as 

 Deegener draws from the abdominal ap- 

 pendages of the Noctuid larva3 would here 

 land us in the absurdity of supposing that the 

 imaginal antennae and thoracic legs of such 

 insects as bees, weevils and ants are not com- 

 pletely homologous with their embryonic an- 

 tenna and thoracic legs. We are bound to 

 conclude that all insect embryos are polypod 

 and that the most ancient known Pterygo- 

 genea, the Palseodietyoptera, as Handlirsch 

 has shown, had well-developed abdominal ap- 

 pendages, which must have been ambulatory 

 in the more remote ancestors. It is, therefore, 

 simpler to suppose, even if embryology did 



not furnish a great amount of evidence in 

 support of this conclusion, that the ambula- 

 tory function has been revived in some of 

 these appendages (pedes spurii of the cater- 

 pillars of Tenthredinidffi, Lepidoptera and 

 Panorpatse, pedes scansorii of Dipteran and 

 Ooleopteran larvffi), while others have become 

 portions of the ovipositor and sting of the fe- 

 male insects^ than to suppose that these 

 various organs have come into existence de 

 novo through modification of abdominal 

 sclerites. This view, which is now fashion- 

 able in Germany, has arisen through ignoring 

 or misinterpreting the conditions in the in- 

 sect embryo, attaching undue importance to 

 supposed homologies of the sclerites of adult 

 insects and supposing that the organization 

 of the Pterygogenea is to be interpreted by 

 means of the Thysanura. It is a pleasure to 

 see that Deegener departs from the conven- 

 tional view to the extent of regarding the so- 

 called campodeiform larva in the Holometa- 

 bola as a secondary and not as a primitive 

 type. In this respect his views coincide with 

 those of Lameere, Boas and Handlirsch. 



All entomologists will probably agree with 

 Deegener that the characters peculiar to the 

 larva have " arisen during metembryonie life 

 successively in adaptation to differences in 

 the conditions of the environment." He dis- 

 cusses at some length the reasons for the 

 larval retardation in the development of the 

 wings, and in this connection gives an inter- 

 esting account of the subimago of the Ephem- 

 eridea, for the purpose of showing that an 

 insect can actually undergo ecdysis after it 

 has completely or almost completely developed 

 its wings, but he does not emphasize the obvi- 

 ous fact that the wings of insects are organs 

 primarily associated with the dissemination 

 of the species, and, therefore, correlated onto- 

 genetically with the maturation of the repro- 

 ductive organs. The few larvae that are 

 paedogenetic (Cecidomyia) and the few beetles 

 (Pissodes, Scolytidae) that become imagines 

 long before reproduction, though striking ex- 

 ceptions, can readily be explained as sec- 

 ondary adaptations. Attention is called to 

 the reduction of the number of ecdyses and 

 the manner in which pupation has become 



