41 



PLATE XX. 



Pecopteris (Cyatheides) arborescens. 



BRONa. 



To this specific group, variable and elastic as it is, we 

 refer this interesting specimen, which illustrates in a 

 peculiarly obvious manner the great uncertainty which 

 must always attend any serious attempt to follow a truly 

 natural classification with regard to mere detached frag- 

 ments of fern-fronds. Here we have fortunately pre- 

 served for us a sudden modification of pinnules which 

 gives rise in the same plant to a diversity which, in less 

 happily conditioned remains, would inevitably lead to 

 the assumption that we had two species to deal with. 



With all but some very marked and constant species, 

 nothing can be done beyond reference to some group of 

 forms, with which a more or less close alliance can be 

 claimed. 



The Pecopterids are perhaps more open to remarks of 

 this kind than other Orders. " The natural classification 

 of these fossils," says Schimpek, " often so imperfect, pre- 

 sents difficulties against which all endeavours in this 

 direction have hitherto failed." (" Paleontologie 

 Vegetale," Vol. L, p. 498). 



The drawing is by Prior, and is of the size of nature. 



The specimen comes from the shale roof of the 

 Bensham coal-seam, J arrow Colliery. 



13 



