MUSEUM OF COMPARATIVE ZOOLOGY. 281 



Dyscritus vetustus. 



A very siricall fragment, said to belong to the bind margin of a wing, with two 

 series of eight square-shaped cells between three veins, one of them branched 

 at the base, is all that is preserved. It can belong to Orthoptera, to Pseudo- 

 neuroptera, or to Neuroptera, but it is too insignificant to be identified. Simi- 

 lar cells are found in Isophlebiu. 



The conclusions to be made from the results of my examination of the 

 Devonian Insects are the following : — 



1. The known fragments belong to five species. 



Two are Odonata, belonging to the Pseudoneuroptera. The very im- 

 perfect fragments do not permit us to say more than that some charac- 

 ters are similarly represented in the gigantic species of the Solenhofen 

 state, in Stenophlebia and in Isophlebia. These characters are the sud- 

 denly narrowed second cubital space in Platepliemera, and the straight 

 sector trigonuli inferior in Gerephemera, neither mentioned by the 

 author. 



The three other fragments belong to the Neuroptera, and probably all 

 to the Sialina. One of them is more related to the Corydalis type, the 

 two others to the Chauliodes type. 



2. None of them have any relation to the Ephemcrida?, as is asserted 

 by the author. 



3. None of the Devonian Insects are of a synthetic type. Besides 

 that such a type could hardly be derived from the wing only of living 

 species, these specimens are too fragmentary for such conclusions. 



4. The previous stages of all were probably aquatic. 



5. No related species is known from the North American carbonifer- 

 ous strata. Probably all insects known from them are terrestrial. Till 

 a more complete account is given of Euephemeritcs, it cannot be consid- 

 ered to be an insect wing. 



6. Concerning the four famiJies proposed by the author, one, the 

 Atocina, is out of the question, because belonging to the Odonata. Tha 

 other three are based upon extremely vague characters, which are not 

 justly to be considered family characters at all. 



The stud}'- of fossil insects, and especially the study of fragments of 

 fossil insects, is doubtless extremel}' difficult. The most detailed knowl- 

 edge of the living fauna is indispensable, and, as the specialist will in- 

 evitably find, the actual literature is entirely insufficient for the details 

 needed for comparison, and a very complete collection, such as does not 

 yet exist here for any group, is necessary in order to avoid grave errors. 



