66 BRITISH EOCENE FLORA. 



The present range of 0. rer/alis is from Sweden to Siberia, Japan, Azores, Bombay, 

 Coehin, Plong-kong, Africa, and from Canada to Rio. Unlike O.javanica, the fronds of 

 0. regalis attain their greatest luxuriance in moist, temperate regions, such as Ireland, 

 and decrease in size towards the Tropics. Its absence from our Eocene beds (except 

 possibly from the Thanet Sands) is very significant. 



12. OsMUNDA LiGNiTUM, Giehel. (Page 49.) 



This is a species of somewhat limited vertical range common in the Middle Eocenes 

 of Central Europe, France, and England, where specimens have been found in profusion. 

 It is indistinguishable by any specific character from the existing O.javanica, and the 

 expediency of giving it a separate name is very doubtful. 



Osmunda javanica is a magnificent species inhabiting the East Coast of Asia, from 

 Kamschatka to Java and Sumatra ; more than sixty degrees. It attains in the latter 

 countries a luxuriance unknown elsewhere^ some pinnae collected by Beccari in Sumatra 

 measuring 9 inches in length, by 1| inch broad, while the pinnae of plants from farther 

 north, especially from Kamschatka, do not exceed 4 inches in length, and are only \ of an 

 inch broad. It is in the more average-sized pinnae from Formosa, latitude 24°, that 

 we meet with the most absolute identity, as far as the fragments admit of comparison, 

 with our fossil form.^ If the size of the pinnae actually decreases progressively with 

 the isotherm, as appears to be the case in the large series at Kew, this fossil, from 

 its identity with the existing species, may be of use in determining our Eocene 

 temperature. It is remarkable that the most luxuriant fronds from Java and Sumatra 

 have simple margins, while the smaller northern forms become usually more or 

 less deeply serrated. The greater luxuriance attained by it in southern habitats shows 

 it to be an essentially tropical form, while as fossil it has never been found in northern 

 latitudes : nor either recent or fossil in any part of America. It seems to be met with in 

 Europe in Middle Eocene and Oligocene times, and no case is yet known to me qi its 

 occurrence in beds of much later date, notwithstanding that Saporta places its horizon in 

 the Miocene. 



^ The longest pinna in my collection, an imperfect one, measures nearly six inches and nine-tenths in 

 length. 



