EOCENE FERNS. 77 



tion is of a usual type. Prom its restricted vertical range at Bournemouth, and its 

 presence at Bovey, it might be inferred that the former locality marks almost its extreme 

 southern range at that time, while northward it reached to Antrim, which may also have 

 been near its limit, since it is apparently absent at Mull. It was evidently a woodland 

 and not a marsh Eern, and its exclusion from Hordwell and Hempstead on this supposition 

 is natural. The importance of the discovery of the Irish locality lies in the fact that it 

 helps to link the Bournemouth Eocenes with those of Antrim, which are themselves linked 

 to the Greenland Tertiaries. I have re-engraved (PI. XIII, fig. 7) the original specimen 

 figured by Mr, Baily in the ' British Association Reports ' for 1880, pi. 2 fig. 1. It was 

 found in an ironstone pebble among silicified wood on the shores of Lough Neagh, 

 and though undescribed, the name of Ilemitelites Fraseri is attached to it. 



All these species are obviously divisible, through the plants associated with them, into 

 marsh Eerns and woodland Eerns, while some which have the greatest range were evidently 

 not entirely excluded from either station. Examples of the latter are the Chrysodiura and 

 Osmunda, remains of which occur in the greatest profusion, mingled with palmettos and 

 other swamp plants, but which are also found with distinctly forest vegetation. Examples 

 of woodland Eerns are the Lygodium, Anemia, the two Adiantums, &c. Goniopteris Bun- 

 burii, a purely woodland Fern at Bournemouth, existed perhaps in a lower temperature with 

 a damper station at Bovey, while G. stiriaca luxuriated as a marsh Fern at Bovey, but has 

 only been met with once among forest leaves at Bournemouth. It will perhaps be safer, 

 however, to defer these and other inferences until more plants have been described. With 

 the exception of the last-named Fern and the Osmunda, which we have seen shows 

 evidence of relatively lower temperature, the higher elevation at Bovey seems to have 

 excluded the contemporary Bournemouth Ferns, or rendered them rare. 



While we apparently know much respecting the Ferns of the British Eocenes, com- 

 pared with what is known of those from the Tertiaries of other countries, several conside- 

 rations tend to show how insufficient and how incomplete is our grasp of the subject. 

 The present inability to explain the restricted range of so many Eocene Ferns, such as Phy- 

 matodes, for which there is no apparent reason, the absence of any trace of such widely 

 distributed and ancient forms as Dicksonia, Asplenium, Lomaria, and Todea, throughout 

 almost the whole Tertiary period ; the failure to detect any trace of the fertile pinnae of 

 so many and such comparatively abundant Eocene Ferns, and the vague knowledge 

 possessed of some other forms, renders the present portion of our work far from 

 satisfactory, and has shown the necessity of a more attentive consideration than it has 

 received. 



In comparing our fossils with those of other countries in Europe, we see that even 

 the Ferns may afford some clue to the relative sequence of many of the plant-bearing 

 beds, whose ages cannot be ascertained stratigraphically. The presence, in the Cretaceous 

 series of Aachen, of Jurassic types, now confined to distant isles, indicates their relative 

 antiquity. The presence of a portion of the least archaic of the Aachen types at Sezanne, 



