298 



MR. J. S. GA.KDKER ON THE LEAF-BEDS 



any other form. The specimen named Sassafras Ferretianum, Mass., 

 though very fragmentary, is far more like a leaf from Dulwich 

 and Newhaven than the Italian species. Fagus dentata, Unger, is 

 represented in Greenland by a few fragments, and Fagus Deucalionis, 

 Unger, by one insignificant shred of the same type as others 

 called Castanea Ungeri. Bryandra acutiloba, Brongn., ranks as a 

 Greenland species on account of another insignificant fragment, and 

 veins are introduced into the figure which heighten the resemblance, 

 but do not exist in the specimen. Rhamnus Eridani, Unger, is 

 based on a good specimen, but without distinctive character : 

 Rhamnus Gaudini, Hr., on a fragment which is alike destitute of 

 base, apex, and margins, and in which even the veins are differently 

 disposed from those of the Swiss specimen ; and Rhamnus brevifolius, 

 A. Br., from Switzerland, is quite distinct from that of Greenland. 

 Juglans Strozziana and Quercus Laharjoii, Gaudin, from Greenland, 

 are mere fragments devoid of distinctive characters ; Gornus ferox, 

 Ung., is based on fragments too imperfect for comparison ; and 

 Alnus nostratus, Ung., on specimens just sufficiently distinct to show 

 that they cannot belong to that species. Finally, the leaf deter- 

 mined as Populus scleropliylla, Sap., of Greenland, differs in every 

 detail from the same species at Armissan. 



These determinations are for the most part based upon specimens 

 which I should regard, under any circumstances, as too frag- 

 mentary to be of any value, and belonging moreover to types of 

 leaves which are so universal that they would, even if perfect, fall 

 into the undeterminable residuum of a fossil flora. It would, I 

 believe, have been just as easy to have identified them with either 

 Cretaceous, Eocene, or living species as with Miocene. It is not a 

 little significant that none of the finer and most distinctive plants of 

 the Mull, the Antrim, or the Atanekerdluk floras in question were, 

 or could be, identified with Miocene species, but were all admitted 

 to be peculiar. Among these, we have the Pterospermites or Quercus 

 platania (for these seem to be the same plant) , represented by a 

 magnificent specimen from Atanekerdluk in the British Museum, 

 and a similar one from Mull in the Glasgow Museum, recalling the 

 larger leaves of Sezanne. Then there is the very satisfactory Mag- 

 nolia Inglefieldi, and the fine species named Ilex longifolia and 

 Quercus groenlandica. 



But the best marked and most thoroughly characteristic leaves of 

 the whole are the Baphnogene Kanii, Hr., and the MacClintockia tri- 

 nervis, Hr., forms which are equally represented in the Thanet 

 flora of Gelinden and in the Antrim floras. These perhaps belong 

 to the genus Pilea of the Urticece, but they evidently existed in 

 Europe at one definite stage and no other, and afford palseontological 

 evidence that should be conclusive. With these is the Rhus bella 

 of Heer, identical with Bewalquea gelindensis of Gelinden, and a 

 few other less distinctive forms. 



The large number of fossil floras brought from Arctic regions have 

 demonstrated that there is a gradual passage from those determined 

 to be Cretaceous to those determined to be Miocene. The stratifi- 



