INSECTS OF NEW BRUNSWICK. 31 



known, the carboniferous age of the deposit. Yet we find in this devonian locality 

 not a single one of the Palaeoblattariae or anything resembling them ; and more than 

 half the known insects of the carboniferous period belong to that t3'pe. The next 

 most prevailing carboniferous type is Dictyoueura and its near allies, with their 

 reticulated wings. Gerephemera only, of all the devonian insects, shows any real and 

 close aflBinity with them ; and even here the details of the wing structure, as shown 

 above, are very different. The apical half of the wing of Xenoneura (as I have 

 siq)j)osed it to be formed) also bears a striking resemblance to the dictyoneuran wing ; 

 but the base, which is preserved, and where the more important features lie, is totally 

 different. The only other wing which shows particular resemblance to any carboniferous 

 form (we must omit Dyscritus from this consideration, as being too imperfect to 

 be of any value) is Platephemera, where we find a certain general resemblance 

 to JEj}hemerltes Hiickertl Gein., and Acridites priscus Andr., but this is simply in the 

 form of the wing and the general course of the nervules ; when we examine the details 

 of the neuration more closely we find it altogether different, and the reticulation 

 of the wing polygonal and not quadrate as in the carljoniferous types.^ In this 

 respect indeed, Platephemera differs not only from all modern Ephemeridae, but 

 also from those of other geological periods.^ Another prevailing carboniferous type, the 

 Termitina, is altogether absent from the devonian. Half a dozen wings, therefore, from 

 rocks known to be either devonian or carboniferous, would probably establish their 

 age. 



8. The devonian insects uiere of great size, had membranous wings, and were 'prohahly 

 aquatic in early life. The last statement is simply inferred from the fact that all the 

 modern types most nearly allied to them are now aquatic. As to the first, some state- 

 ments have already been made ; their expanse of wing probably varied from 40 to 175 

 mm. and averaged 107 mm. Xenoneura was much smaller than any of the others, its 

 expanse not exceeding four centimetres, while the probable expanse of all the rest was 

 generally more than a decimeter, only Homothetus falling below this figure. Indeed if 

 Xenoneura be omitted, the average expanse of wing was 121 mm., an expanse which 

 might well be compared to that of the Aeschnidae, the largest, as a group, of living 

 Odonata. There is no trace of coriaceous structure in any of the wings, nor m any are 

 there thickened and ap23roximate nervules — one stage of the approach to a coriaceous 

 texture. 



9. Some of the devonian insects are lilainly jyrec^irsors of existing forms, lohile others 

 seem to have left no trace. The best examples of the former are Platephemera, an 

 aberrant form of an existing family ; and Homothetus, which, while totally different in the 

 combination of its characters from anything known among living or fossil insects, is the 

 only palaeozoic insect possessing that peculiar arrangement of veins foiuid at the base of 

 the wings in Odonata, tyjjified by the arculus, a structure previously known only as early as 



1 Dr. 11. B. Geinitz has kindly re-examined Ephemeriles ^ The Dictyoneuiae and their allies, as may be inferred, 



Rilckerli at my request, and states th.at the reticulation is in are considered as belonging to the Palaeodictyoptera, 

 general tetragonal, but that at the extreme outer margin although their ephemeridan affinities are not disregarded, 

 the cells appear in a few places to be elliptical five- or six- 

 sided. 



