164 J. Starhie Gardner — Development of Dicotyledon!^. 



cene. Enough has, however, been said to show that no scientifio 

 A'ahie can attach in a general way to these determinations, and it 

 will suffice to select the Greenland floras as an example, and to in- 

 vestigate the evidence upon which their supposed Miocene age was 

 originally established. This was stated in 1868 to rest on an assured 

 basis, when of 70 described species, 20 were alleged to occur in 

 Miocene beds of Central Europe. . 



Of these two are Ferns, and appear respectively allied to Osmvnda 

 regalis and Pteris aquilina. They are ancient and northern forms 

 which do not appear in any profusion to the south until Miocene 

 times, though the Osmunda occurs in the old Eocene of Gelinden. 

 Of Coniferse two are referred to Sequoia ; the first is erroneously 

 united to S. Gouttsice of Bovey, and the second, S. Langsdorfii, is a form 

 known in the Cretaceous, and to which every fragment of distichous 

 Coniferous foliage from the Tertiaries, that could not be placed in the 

 tbird form, Taxodium duhium, has been at one time or another referred. 

 The latter genus is found in the Middle Eocene of England. The 

 sixth plant is merely a rush-like fragment referred to Phragmites. 

 Of the thirteen Dicotyledons, ten are the most ordinary and wide- 

 spread forms of leaf, such as occur in every flora, and of the three 

 that are at all distinctive forms, two, the Platanus and Populus 

 ZaddacM, are characteristic of the Lower Eocene of Heading, and the 

 third has only been found elsewhere in the Oligocene of Armissan. 

 The age of the Greenland formations can hardly be considered settled 

 on such evidence as this, even were the determinations trustworthy, 

 which is far from the case. Of course enormous quantities of plant 

 remains have been described from Arctic formations since then, but 

 they do not help to justify their reference to the Miocene, and the 

 significant fact remains, that the most distinctive species belong to 

 the Eocene floras of Eeading and of Gelinden. 



But supposing the identity of the floras were satisfactorily estab- 

 lished, should this suffice to establish contemporaneous growth? 

 Could floras of forest trees so closely related to existing species have 

 grown at so recent a geological period synchronously with 24° of 

 latitude intervening? This, it must be remembered, is the distance 

 between Lombardy and Iceland, or Florida and Newfoundland. 

 There is no reason why Eocene floras shpuld be unrepresented 

 round the Pole, and all the Eocene formations existing there be 

 unfossiliferous. A series of sedimentary rocks destitute of fossils, 

 and sometimes a thousand feet thick, is said to occur between the 

 uppermost Cretaceous and lowest supposed Miocene, representing 

 probably the temperate Lower Eocene period, when forests of leafy 

 trees could no more have grown within the Arctic Circle than they 

 M'ould at the present day. As the temperature increased towards 

 the Middle Eocene period, the temperate floras of the Lower Eocene 

 would have migrated north, and we accordingly find, not fancied 

 resemblances, but identity established by the presence of such pecuHar 

 and typical species as the Reading Populus ZaddacM, P. BicJiardsoiii, 

 Platanus, and the Gelinden McClinlockia and Cocculus Kanii. 



The Middle Eocene temperature was moreover exactly that re- 



