Evolution of Plants 357 



But a marked histological difference was early established 

 in the secondary wood of these, which contrasts Tvith that 

 seen in the other vascular groups already considered. For, 

 while in the latter abundant scalaiiform or barred tracheids 

 made up the bulk of the secondary wood or xylem tissue, in 

 the group now under consideration narrower, more elongated, 

 and much lignified elements, with pore markings along their 

 walls, became a frequent, often a sole, feature of the wood. 



The precociously rapid evolution of floral or anthophytic 

 conditions through the heterosporous stage, up to that of 

 staminal and ovular production, is even more marked than 

 in the other two groups. Our knowledge also of these is now 

 fairly helpful, alike from the macroscopic and the microscopic 

 standpoints. 



The " Cordaianthus" flowers were staminate and ovulate. 

 The staminate flowers had characters that foreshadowed those 

 seen at the present day in Ginkgo, in part also in the Taxacese, 

 and in part in the genera Ephedra and Gnetum. The carpellary 

 flowers are decidedly variable, and viewed comparatively 

 suggest several evolving stages, that became more pronounced 

 in later and higher groups. Thus in some genera numerous 

 ovules in close relation to accessory leaves were arranged along 

 an elongated axis and so foreshadowed the cone of pinaceous 

 and related alliances, as well as the apocarpous polycarpellary 

 pistil of angiosperms. Others with few ovules and accessory 

 leaves resembled thus some of the Cupressinese, and the five- 

 to three-carpelled pistil of many primitive angiosperms, while 

 others bore a solitary ovule that suggests resemblances to 

 Ginkgo and the Taxacese. 



From the variety of seeds found in the carboniferous rocks, 

 that seem to have cordaital affinity, it is almost assured that 

 numerous genera will in time be identified in their vegetative 

 parts. But, as transition is made from the carboniferous to 

 the permian period, genera like Walchia, Noeggeraihia, Baiera, 

 Glossopteris, Ullmania, and others became increasingly abun- 

 dant, and indicate a splitting up of the cordaital stock into 

 one line of deviation that led, as set forth in the diagram (Fig. 13) 

 to the heterogeneous gymnospermic genera commonly known 

 as the Gnetales, and to two other gradually separating groups 

 that are there designated the Ginkgoales and Protoi)inales. 



The former or Gnetales is the one that might claim our 

 deepest interest, since from it we would trace the origin of the 



