422 MESEOSEBE AND PALEOSERE. 



sented by the climax must have been many times longer than that occupied 

 by the serai development. Each period of submergence or flooding is 

 marked by a sedimentary deposit, above which lies a coal bed corresponding 

 to the hydrosere of the new swamp. Where a layer of black shale occurs in 

 a bed of coal, it indicates the limit of one stase and the beginning of the 

 nejct one. When two or more stases of this sort are found together they con- 

 stitute a costase corresponding to a cosere, lasting through one major swamp 

 period closed by extensive submergence. In the section of the Coal Measures 

 of Ohio, the black bands of coal are probably costases for the most part 

 (fig. 45)._ 



The clisere. — ^The existence of zonal differentiation and hence of a cUsere 

 during the Paleophytic seems a necessary inference from the deformation 

 sequence of the Silurian-Devonian and the Upper Mississippian mountain- 

 forming movements. This is supported by the evidences of glacial or arid 

 climates during the Devonian of South Africa and of Britain, and by the appear- 

 ance of the Glossopteris flora during the Carboniferous. As a consequence, it 

 appears probable that both of these deformations brought about a cooled or 

 glacial climate in certain centers, perhaps polar as well as cyclonic, attended 

 with more or less aridity. The first glaciation must have differentiated 

 one or more zones in the original climax mass of each glacial region. The 

 effect of refrigeration upon a tropical or subtropical vegetation should have 

 caused a marked sorting out of species, and a corresponding series of zones 

 but this was doubtless checked by the fact that the flora was still too primitive 

 to permit as many lines of evolution as later became possible. This seems to 

 be especially indicated by the the fact that Permian glaciation first destroyed 

 vegetation to a striking degree, and that the great evolution of gymnosperms 

 and the initiation of angiosperms were delayed consequences. The fact that 

 the probable centers of Devonian glaciation are in regions marked by both 

 Proterozoic and Permian glaciation makes it all the more likely that such 

 areas were more or less constantly or at least recurrently the site of zonal 

 differentiation or of movements of the clisere. The behavior of the latter 

 was probably very similar to that already sketched for the Permian, and the 

 chief r61e in the zonal shifting was probably taken by the Glossopteris flora, 

 in so far as it was evolved at this time. 



