UPPER PALEOZOIC FLORAS 155 



changes in the western American areas, which are more pronounced 

 than in the Appalachian trough, may be due either to changed physical 

 conditions consequent to nearer and greater orogenic movements, or 

 possibly to position nearer to the region of Gondwana glaciation. In 

 our trans-Mississippi region the contrast between the Permian and 

 Stephanian floras does not seem so sharp as that between the Ural 

 flora and those, for example, of the fresh-water basins of France or 

 England. 



Greater contrasts in eastern Europe. — In France and in the Appa- 

 lachian trough the early Permian floras are to a large extent identical, 

 thus indicating continued freedom of intercourse, x^lso the woods 

 present obscure rings or none at all. But in Kansas and Texas annual 

 rings while still but slight are increasing in distinctness though many 

 of the associated plants have a European distribution. 



The plant fragments from the Permian of Nova Zembla are quite 

 too insufficient to support the hypothesis of a far-northern route of 

 intercontinental migration, or even to show that the climate of 

 western Europe extended so far north. 



Ural Mountains on western edge of a great climatic zone. — Indeed, 

 on the contrary, the great scarcity of true Neuropterid and Pecop- 

 terid elements, the great development of Psygmophyllum, and the 

 remarkable phases of Callipteris observed even in the Artinsk of the 

 Ural region indicate, in my judgment, either temporary isolation or 

 an appreciable difference in climatic conditions. At the same time 

 the somewhat unique forms of Odontopteris are not without repre- 

 sentation both on the central plateau of France and in the Wichita 

 of Texas. This testimony of migration is supported by the supposed 

 land relations, which, if Lapparent is correct, should have been 

 particularly favorable for such a migration during portions of Per- 

 mian time. 



Extension 0} Gondwana flora into northern Asia. — Eastward of the 

 Ural Mountains the floras of the Altai and headwaters of the Yenesei 

 and in northern Mongolia, though provisionally referred to the Per- 

 mian on account of the presence of Rhipidopsis, mingled with other 

 types, including some of Gondwana facies, are possibly wholly lack- 

 ing in specific types characteristic of the Cosmopolitan Western 

 Permian. Though the stage of these plant beds is not yet fixed on 



