WARD] DISCUSSION OF DIAQEAMS. 447 



Pterophyllum from the coal measures of China, and there is probably a 

 second from Europe. Fourteen species occur in the Permian, includ- 

 ing the typical genera Bioonites and Glathraria. It is not, however, 

 until the Keuper is reached that this type of vegetation assumes a 

 leading part, and throughout the Jurassic it continues to be the most 

 abundant form of plant life. In the Lias it forms 43 per cent of the 

 iiora of that formation, though this may be accidentally exaggerated. 

 It was 28 per cent of the Oolitic flora and more than 35 per cent of 

 that of the Wealden. From this point, however, its decline was rapid 

 and uninterrupted until in the living flora only 75 species of cycadace- 

 ous plants are known to botanists. Of these North America can claim 

 but a single one, the sago-palm (Zamia angustifolia) of our extreme 

 Southeastern States. 



Passing to the Coniferse, we find the Gordaites Bobbii of Dawson 

 from the Devonian of Canada recurring in the Upper Silurian of H6- 

 rault. This genus was formerly supposed to be the prototype of the 

 Cycadacese, but, as already remarked, this opinion is now abandoned by 

 the best authorities, and the genus referred to the Coniferse. The evi- 

 dence upon which this change rests cannot be presented here, but it is 

 proper to say that the savants who have marshaled it have done so in 

 such a manner as to render their conclusion akin to irresistible. But 

 its adoption has carried with it a train of consequences which cannot be 

 escaped. Not Gordaites alone, or with its spore-bearing parts {Cordai- 

 anthus) and its fruit [Gordaioarpus), but Nmggerathia, Trigonocarpus^ 

 Gardiocarptis, Bhabdocarpus, Sternbergia, Artisia, etc., must all follow 

 in its wake and be gathered, one and all, into the great family of the 

 Coniferae. It is thus, as shown by our table and diagram, that this 

 type assumes such a commanding position far back in Paleozoic time, 

 forming about one-fourth of the vegetation of the Permo-carboniferous 

 epoch. Doubtless this effect is exaggerated by duplications caused 

 by giving different names to separate parts of the same plant, but this 

 occurs throughout the series only to a less obvious degree. 



The true Coniferae, which have some representatives in the Paleozoic, 

 replace the Gordaitem entirely in the lower Trias and thereafter vie 

 with the Oycadacese for supremacy, which they do not fairly attain 

 until the lower Cretaceous is reached. Being of a higher type of struct- 

 ure than the latter by reason of their exogenous mode of growth and 

 other peculiarities, they refuse to succumb in competition with the now 

 rising Angiosperms and continue to hold their own through much of 

 the Tertiary. At the present time the number of known species (300) 

 would denote a great decline, but this is in large part made up by the 

 wonderful predominance and territorial expansion of these persistent 

 forms. Although from the point of view of the number of species 

 alone, the present Coniferae would form but one-fourth of 1 per cent 

 of the vegetation of the globe, we in fact find vast tracts of country 

 covered with pine, fir, and spruce forests, excluding almost completely 



