THE EVOLUTION OF THE PLANT-WORLD 9 



the ocean basins. The facts of the earth's form are 

 shown to be still more in accordance with this theory 

 in comparison with a form, intermediate between the 

 sphere and the tetrahedron, known in geometry as 

 the hexakistetrahedron, or six-faced tetrahedron, which 

 may be described as a tetrahedron with a low pyramid 

 of six triangular faces on each of its four sides. 



The present distribution of land and water agrees 

 very closely with the requirements of this hypothesis; 

 but it is more important that the primitive distribution 

 should be shown to do so. Most of the present land 

 surface can be shown, by its rocks and their contained 

 fossil remains, to have been at one period or another 

 beneath the sea, though not all at one time. It may, 

 however, be true that most of the sedimentary rocks 

 are of shallow water origin, and have not originated 

 in such abyssal depths as those of our present ocean- 

 bed. From this it will follow that, though subject to 

 repeated oscillations, so that one tract after another 

 has disappeared and reappeared from beneath the sea, 

 the continents, though constantly varying in shape and 

 size, have in the main retained their individuality, and 

 the existing ocean basins have been, perhaps from the 

 very beginning, great trough-like depressions of the 

 earth's crust. Such a belief in the permanence of at 

 least the skeletons of our continents and of the deeper 

 parts of the ocean as land and water respectively is 

 compatible with the admission of great changes linking 

 land-masses now separate, or vice versa. Thus, there is 

 considerable evidence of former extensions of land, at 

 different periods, between Scandinavia, the Highlands 

 of Scotland, and Donegal; between the mainland of 

 Manchuria, Saghalien, and Japan; and between the 

 south-west of South America and the adjacent islands. 

 Some evidence, however, points to the former existence 

 of land-masses which can hardly be in any way con- 

 sidered as extensions of our present continents. Thus 

 there is much geological evidence of the connection in 

 a remote geological period of the Scandinavian area 

 mentioned above with North America and Greenland 

 in a continental mass named Arctis; and at a later date, 

 lasting from the Coal-Measure period at least to that 

 known to geologists as Permian, of the union of Brazil, 

 Argentina, and the Falkland Islands with Africa south 



