492 GEOLOGY. 



to restore the original plant by combining the dissevered parts of 

 different individuals. Until these are found associated, it is usually 

 necessary to describe and name the parts separately, and hence the 

 possibility that different species and even genera will be founded on 

 the separated parts of the same plant, a possibility that has been 

 realized not infrequently. As discovery proceeds, these are detected 

 and the nomenclature revised. 



Though occasional relics of land plants have been found in the 

 Silurian and earlier systems, it was not until the Devonian that land 

 vegetation was fairly well preserved, but, for the reasons just 

 adduced, it is by no means safe to assume that prolific land vege- 

 tation had not long clothed the surface, and that, perhaps, of a 

 somewhat advanced order. Physical reasons for assuming a vegetal 

 protection of the surface as far back as Proterozoic times have 

 already been presented. The study of the Devonian and Carboni- 

 ferous plants shows that they were widely differentiated, so much so 

 that their ancestral relations are uncertain, which is best explained 

 by a long previous history. The alternative is to suppose a very 

 rapid evolution attended by an equally rapid extinction of the inter- 

 mediate forms and of the residual structures that usually linger long 

 to mark the course of evolution. This alternative is, however, pre- 

 ferred by some authors. 



The larger part of the Devonian plants preserved were buried 

 in inland basins, sometimes mingled with the fossils of fresh-water 

 animals. Of the upland vegetation little or nothing is known. The 

 lowland forms may have spread over the uplands in greater or less 

 degree, or a very different and quite unknown vegetation may have 

 prevailed there. Comparatively few forms have been found in the 

 earlier Devonian formations and the apparent evolution during the 

 period is very marked, but it is doubtful whether this represents much 

 more than improved preservation, for the types do not converge rapidly 

 when traced backwards as though they had originated within the 

 period, nor do they diverge rapidly as traced on to the Carboniferous 

 period. The Devonian types were nearly the same as those of the 

 Carboniferous, and as the latter are much better represented in the 

 Coal flora, their special description and illustration may be deferred 

 with advantage until the discussion of that period. 



The dominant Devonian forms were pteridophytes (ferns and their 



