30 SCUDDER ON THE DEVONIAN 



3. Tliey toere all loioer Heterometahola. As wings are the only parts preserved, we 

 cannot tell from the remains themselves whether they belong to sucking or to biting 

 insects ; for, as was shown in the essay already referred to, this pomt must be considered 

 undetermined concerning many of the oldest insects until more complete remains are 

 discovered. 



They are all allied or belong to the Neuroptera, using the word in its widest sense. At 

 least two of the genera (Platephemera and Gerephemera) must be considered as having a 

 closer relationship to Pseudoneuroptera than to Neuroptera proper, and as having indeed 

 no special affinity to the true Neuroptera other than is found in Palaeodictyoptera. "^wo 

 others (Lithentomum and Xenoneura), on the contrary, ai-e plainly more neaiiy related to 

 the true Neuroptera than to the Pseudoneuroptera, and also show no special affinity to 

 true Neuroptera other than is found in Palaeodictyoptera. A fifth (Homothetus), which 

 has comparatively little in common with the Palaeodictyoptera, is perhaps more nearly 

 related to the true Neuroptera than to the Pseudoneuroptera, although its pseudo- 

 neuropterous characters are of a striking nature. Of the sixth (Dyscritus) the remains 

 are far too imperfect to judge clearly, but the choice lies rather with the Pseudoneuroptera 

 or with Homothetus. The devonian insects are then about equally divided in structural 

 features between Neuroptera proper and Pseudoneuroptera, and none exhibit any special 

 orthopterous, hemipterous or coleopterous characteristics. 



4. Nearly all are synthetic types of a comjmratively narrow rancje. This has been 

 stated in substance in the preceding paragraph, but may receive additional illustration 

 hei-e. Thus Platephemera may be looked upon as an ephemerid with an odonate retic- 

 ulation ; Homothetus might be designated as a sialid with an odonate structure of the 

 main branch of the scapular vein ; and under each of the species will be found detailed 

 accounts of any combination of characters which it possesses. 



5. Nearly all hear marks of affinity to the carboniferons PalaeodictyoiJtera, either in 

 the reticulated surface of the wing, its longitudinal neuration, or both. But besides this 

 there are some, such as Gerephemera and Xenoneura, in which the resemblance is marked. 

 Most of the species, however, even including the two mentioned, show palaeodictyopteran 

 characters only on what might be called the neuropterous side ; and their divergence 

 from the carboniferous Palaeodictyoptera is so great that they can scarcely be placed 

 directly with the mass of palaeozoic insects, where we find a very common type of wing 

 structure, into which the neuration of devonian insects only partially fits. For : 



6. On the other hand, they are often of more and not less comjilicated structure than 

 most Palaeodlctyoiitera. This is true of the three genera mentioned above with peculiar 

 neuration, but not necessarily of the others, and it especially true when they are com- 

 pared with the genus Dictyoneura and its immediate allies. There are other Palaeodicty- 

 optera in the carboniferous period with more complicated neuration than Dictyoneura, but 

 these three devonian insects apparently surpass them, as well as very nearly all other 

 carboniferous insects. Furthermore : 



7. With the exception of the general statement under the fifth head, they hear 

 little special relation to carhoniferous forms, harlnrj a distinct fades of their own. This 

 is very striking ; it would certainly not be possible to collect six wings in one locality 

 in the carboniferous rocks, which would not prove, by their affinity with those already 



