508 F. H. KNOWLTON EVOLUTION OF GEOLOGIC CLIMATES 



their leaves being either spinoid or very small, and delicately thin. The latter 

 vt^ere probably dorsally rolled at first during the intervals of exposure to the 

 air. . . . The expansion of a proper leaf and the production of an aerial 

 system of transpiration were presumably gradually evolved as the plant be- 

 came vreaned from its subaqueous habitat, and accustomed to gain its food 

 from the atmosphere. However this may be, it is fairly clear that the early 

 representatives of the dominant Devonian types were of limited foliar ex- 

 panse. ... It also appears that to support their weight in air a reinforced 

 cuticle, later developed as a very thick and complicated cortex, was made to 

 serve until a woody axis, and, eventually, secondary wood should be fully 

 produced by their descendants. From the characters of some of the fossils it 

 seems probable that, unable to stand alone, they sprawled or clambered about 

 on the ground or on other plants." 



The known land flora of the early Devonian is very scanty indeed. 

 The vegetation of the time had but recently emerged from an aqueous 

 habitat, and, it is to be presumed, was of thin-leaved, hence weak, flaccid 

 type, diflering but little from the water-inhabiting ancestors. The ma- 

 trix in which it should be found is coarse-grained, and, moreover, the 

 vicissitudes to which the rocks were subsequently subjected have made it 

 almost impossible to recover recognizable plant remains, even if such 

 were present. 



But evolutionary development was rapid and by Middle Devonian 

 time a considerable number had assumed an upright position, thougli 

 of very strange and forbidding aspect. Perhaps the most characteristic 

 types are Psilopliyton, Arthrostigma, Rachiopteris, and the peculiar fern- 

 like plant known as Archwopteris. According to best information avail- 

 able Archceopteris is not a true fern, but is a pteridosperm, showing that 

 it had already developed the seed-bearing habit which would seemingly 

 imply a very considerable developmental history. The huge fern trunks 

 of Caulopteris show that the erect habit had gained marked headway. 



As regards the place of origin of the Middle Devonian flora, Whi{;3 

 says : "Though eastern America has contributed most to our knowledge 

 of this flora, it is probable that either the estuaries of northwestern 

 Europe or the Arctic regions offered the conditions most favorable to its 

 development. It extends both east and west in a high degree of unity." 

 It has not, however, been found in the Southern Hemisphere. 



On coming to the Upper Devonian we find the land flora well estab- 

 lished and in full swing. It shows little evidence of climatic change 

 from the Middle Devonian, but exhibits a number of new types such as 

 Pseudohornia, a probable protocalimarian ancestor, several ferns, a 

 probable ancestor of the Lycopodiales, and a Lepidodendron-\ike form 

 as well as Archceopteris and trunks of Cordaites {Dadoxylon) . In at 



