1857.] BUNBURY — NEUROPTERIS. 245 



certainly not very easy to prove positively that the Neuropterids 

 may not belong to that family. The leaves of Cycads have generally, 

 indeed, simple veins ; but in those of Ceratozamia Mexicana the 

 veins are occasionally, though not generally, dichotomous. In the 

 curious and anomalous Stangeria, a true Cycad, the leaves have so 

 perfectly the characters of Ferns, that the plant was originally taken 

 for a Lomaria, the veining and other characters of the leaves being 

 exactly those of that genus. If the leaves of the Stangeria had been 

 found fossil, in the imperfect state in which fossil plants generally 

 occur, it would, I believe, have been impossible to show that they did 

 not belong to a Fern. The leaves of all known Cycads, however, 

 are simply pinnated, whereas those of the Neuropterids, in all cases 

 where they are pretty completely known, are doubly pinnated. In 

 their apparent texture, moreover, in the characters of the leaf-stalks, 

 in the variations of form of the leaflets in different parts of the frond, 

 and in their whole appearance, the Neuropterids of the Coal-measures 

 are so completely Ferns, that, in the absence of any evidence to the 

 contrary, we may safely consider them as such ; and I have little 

 doubt that their fructification will in time be found to confirm this 

 conclusion. 



If we inquire to what recent Ferns the fossil genus Neuropteris 

 was most nearly allied, we shall hardly, I fear, arrive at any very 

 certain conclusion. The older writers compared the commonest Coal- 

 measure species to Osmunda regalis, but merely because of a general 

 vague resemblance. A venation nearly approaching to that charac- 

 teristic of Neuropteris is seen in the Pteris hastata, Swartz {Allosorus 

 hastatus, Presl), and some nearly allied species. These Ferns have 

 also a general resemblance in form to many species of Neuropteris, 

 and particularly in the great differences of outline and form in differ- 

 ent leaflets of the same frond. But a similar veining occurs in some 

 of the broad pinnuled species of Gymnogramme, such as G. tomentosa, 

 Ferns widely different in their fructification from Allosorus. Again, 

 while Allosorus hastatus and A. flexuosus have nearly the venation of 

 Neuropteris, Allosorus calomelanos (a very near ally of these two 

 species) would, if found fossil, be referred to Cyclopteris. The 

 veins, in fact, afford but slippery characters, although the best that 

 are generally within our reach, for the arrangement of fossil Ferns. 

 We know, moreover, that in those cases where the fructification of 

 fossil Ferns has been found in a good state, it has often indicated 

 affinities quite different from those which would have been inferred 

 from the veins or the outline. Thus, the Alethopteris aquilina has 

 a striking resemblance in form and venation to some species of Pteris, 

 yet its fructification, as shown by Geinitz, seems to place it in quite 

 a different tribe of Ferns, namely the Gleicheniacece. Again, the 

 Pecopteris exilis, Ph., with the aspect of some small Polypodium or 

 Nephrodium, has (as I have elsewhere shown) the peculiar spore-cases 

 characteristic of the Schizceacece* . I should not, therefore, be sur- 

 prised to find that the Neuropterids differ considerably in their real 



* Quart. Journ. Geol. Soc. vol. vii. p. 188. 



