EOCENE FERNS. 79 



from, northern latitudes, the knowledge we already possess of the Arctic and North 

 American fossil floras effectually takes away every ground for supposing that the strong 

 relationship between the Neo-tropical Eern flora of the present day and that of our 

 Eocene could have been due to any land communication north of latitude 50°. 



The relationship of our own Eocene Eerns to the present subtropical American flora is 

 apparent from the fact that of seventeen types all but one are known in it. A connection 

 less in degree is seen between them and the Eerns of Eastern Asia and Java, where six 

 only of the types are absent. Scarcely any of the Middle Eocene types have remained in 

 Europe, except Pteris cretica and Adiantim, for those related to Osmunda regalis and 

 Woodwardia radicans seem only to have reached Central Europe from the north during 

 the Miocene. The Eocene forms of Marattia, Anemia, Gleichenia, Phi/matodes, and 

 Hewardia were the first to disappear, and it is remarkable that these happen to diff'er 

 most widely from their living representatives. Eew of the present European Ferns can 

 be traced back to fossil ancestors, even among those of the Arctic regions, while they 

 seem so closely connected with the living Ferns of North America, that of our thirty-six 

 British species, all except three of Nephrodium and Asplenium lanceolatum and A. 

 (/ermanicum, also exist there. 



No more is sought for than to claim for such inferences and conclusions as have 

 been arrived at, that they seem warranted by the results of the work so far. 

 There are, hov.^ever, in all probability living Eerns still undiscovered, and we are as yet 

 certainly acquainted with but a fraction of the Eocene Fern world. Still, however incom- 

 plete the data for interpretation may be, the possibility that something may come to light 

 in the future to change or modify views based upon our present knowledge, does not 

 deprive them of current value. The interest attaching to these F'erns is increased by the 

 fact that while the higher orders of plants from the Tertiaries have formed the basis for 

 much speculation as to former temperatures, distribution, &c., the inferences to be drawn 

 from Ferns have hardly been taken into account. 



Amidst inference and hypothesis, the fact at least is certain, that there is no real break 

 between Eocene and Miocene floras when they are traced over many latitudes. The sup- 

 posed Miocene floras of the Arctic Circle are, we shall see, but modified temperate Eocene 

 floras which had migrated north when warm conditions set in, and which redescended 

 south, still further modified, when the decrease of warmth during the Miocene again per- 

 mitted them to do so. The results that followed the forced convergence towards the land 

 areas of the Pole of perhaps very dissimilar floras from long separated lands of different 

 longitudes, through a general increase in the Eocene temperature of the Northern hemi- 

 sphere are as yet faintly sketched and perplexing, but with patience they can be unravelled. 

 At present it can only be stated that when they redescended as Miocene floras, they were 

 greatly modified and enriched in genera and species. 



