482 Charles Brongniart — Fossil Insects of the Primary/ Boclis. 



tbeir determinations. Moreover, it has been too long thought possible 

 that the divisions created to classify living insects could be made to 

 include all fossil insects. 



It is curious to find how little these insects have changed, and to 

 see that creatures, which are reckoned among the most ancient, have 

 undergone modifications of only a secondary order, in coming down 

 from the Coal period to the present time. 



Nevertheless, there is more homogeneity among the primary Hexa- 

 pods than is presented by this group to-day, but this need not 

 astonish us. 



The insects which are found in the Palseozoic shales belong to 

 types represented at the present day by the Orthoptera, the Neurop- 

 tera and the Hemiptera. 



Some authors have announced the discovery of Coleoptera; but these 

 so-called Coleoptera are really fossil fruits or parts of Arachnida (?). 



I have, myself, considered the perforations found in some fossil 

 wood to have been the work of Coleopterous insects ; but if that 

 were so, Coleoptera, which have hard elytra, would surely have left 

 their impressions on the shales. 



The presence of Coleoptera in these Palgeozoic deposits seems to 

 me then very doubtful, and one can onl}^ assume as certain the exist- 

 ence of Hemiptera, Orthoptera and Neuropttera. 



But the limits of the two latter orders have been very differently 

 estimated by various authors, some not considering the classification 

 as a natural one, and are therefore inclined to unite them in a single 

 group. This would, in my opinion, be going too far. The existing 

 Orthoptera are well characterized as much by the general form of 

 the body as by the neuration of their wings, and by their incomplete 

 metamorphoses. 



The Physopoda, the Corrodentia, and the Amphibiotica, have, 

 with good reason, been united to the Orthoptera properly so called, 

 under the denomination of Orthoptera pseudo-Neuroptera. The 

 latter (Orthoptera pseudo-Neuroptera) would have been placed 

 a long time ago among the Neuroptera, but their incomplete meta- 

 morphoses ranged them rather with the Orthoptera. 



The true Neuroptera are then the Planipennia and the Trichoptera, 

 which have complete metamorphoses. The members of the latter 

 sub-order may serve as links to unite them with the Lepidoptera, 

 as much on account of the neuration of their wings and of the scales 

 or hairs which cover them, as for the buccal apparatus, which forms 

 a sort of proboscis by the union of the jaws and of the lower lip, 

 the mandibles being atrophied. 



Among the Hemiptera, the Homoptera or Cicadidfe are the insects 

 with incomplete metamorphoses. Some naturalists have wished to 

 divide insects into two great groups — those which have incomplete, 

 and those which have complete metamorphoses. It is easy to prove 

 the defect of this classification from the indications mentioned pre- 

 viously. In fact, it would be compulsory to sunder the Neuroptera 

 (Planipennia and Trichoptera) from the Orthoptera pseudo-Neuroptera, 

 insects which are in reality very closely allied. For analogous 



