124 FOSSIL INSECTS FROM COMMENTRY, FRANCE, 



I I'efer to Megctgnatha odonatiformis Bolton, and Sycopteroii 

 symmetrica Bolton. Bolton places the former in the family 

 Perlidce (by which the author evidently means the Ordei- Perlaria, 

 and not the family Perlido' s.str.); though, in a note added to 

 this determination, he states that Dr. A. D. Imms suggests that 

 the specimen "may possibly come nearer to the Sialidce,'' by 

 whicli, I take it, he means the Order Megaloptera, rather than 

 the actual family Sialidce s.str. As for Sycopferon, this is 

 assigned to the "family Fanorpidce" for which, again, we are 

 evidently intended to read the "Order Mecoptera," since the 

 author compares it with Orthophlehia of the Lias. If this placing 

 be correct, then we have a Holometabolous insect present in the 

 Upper Carboniferous — an occurrence whicli is, of course, not im- 

 possible, but is certainly improbable, in view of our present 

 knowledge of the Phylogeny of the Insecta. 



I have, of course, no opportunity of examining the specimens 

 themselves; but Mr. Bolton has given us such excellent photo- 

 graphic reproductions and drawings of the fossils in his paper, 

 that one would certainly scarcely desire anything better. As. 

 the determination of the affinities of these two fossils is a matter 

 of prime importance to students of Insect Phylogeny, and as I 

 have already informed Mr. Bolton by letter of the views that I 

 shall here express, I propose now to state my reasons why these 

 two insects cannot be accepted as belonging to the Orders to 

 which Mr. Bolton would assign them, and to indicate also to 

 what known Orders they may, with more probability, be con- 

 sidered to belong. 



Megagnatha oDONATiFORMis Bolton. (Text-fig. 1). 



Bolton, op. cit., pp.2-6, PL i., figs. 1-4. 



From the photograph given by Bolton in PI. i , fig.l, and from 

 the drawing in Fig. 3, it is evident, as Bolton himself admits, 

 that any reconstruction of the wing-venation of this fossil must 

 be largely a matter of conjecture. We should, therefore, attempt 

 to find other characters that may lead to a solution. The figures 

 show that: — 



(1) The wings were either absolutely, or approximately, equal 



