PALEOBOTANY BERRY. 353 



groups are to be regarded as divergent or convergent and not directly 

 filiated. 



Although very probable, the view that the Coniferophyta is a 

 monoplryletic group can not be said to be established. It has been 

 demonstrated beyond reasonable doubt that the Cordaitales, Gink- 

 goales, and Cycadophytes are alike related to the Pteridosperms and 

 Ferns. In the case of the Coniferales and Araucariales this is not 

 so certain, and some students (Renault, Campbell, Potonie, Seward) 

 consider the latter groups as of Lepidophytic origin. There is a 

 plausible amount of evidence for such an origin for the Araucariales, 

 although such a derivation is not generally accepted. There are, 

 however, insuperable difficulties in the way of including the Conifer- 

 ales in such a stock, even if this theory be accepted for the Araucariales. 

 The most interesting order from an evolutionary standpoint, the 

 Gnetales, since they seem to form a connecting link with the Amentif- 

 erous Dicotyledonae, have an ancestry completely unknown, although 

 they show certain structural and developmental points of contact 

 with the Cordaitales, Ginkgoales, and Cycadophytes. 



In the present treatment the coniferophytes are grouped in six 

 orders: Cordaitales, Ginkgoales, Taxales, Araucariales, Coniferales 

 (Pinales), and Gnetales, which are best discussed separately. The 

 oldest of these orders, long since extinct, is the Cordaitales — a knowl- 

 edge of which is due primarily to the work of Grand 'Eury. The 

 Cordaitales were exceedingly abundant and presumably varied at 

 several horizons in the Paleozoic and the synthesis of a study of im- 

 pressions of their foliage and fructification, pith casts, and petrifac- 

 tions enables us to draw a satisfactory picture of their general ap- 

 pearance and habit, although as yet it is usually impossible to corre- 

 late specific foliar impressions with petrified stems and seeds. 



They were tall and relatively slender trees with trunks that were 

 frequently over 100' feet in height and unbranched except at the 

 crown, where their spirally arranged foliage of simple and often 

 large parallel veined leaves was displayed. Leaf form has been 

 used as a basis for the three forms genera : Eucordaites, with spatu- 

 late blunt tipped leaves often several inches in width and 2 or 3 

 feet in length ; Dorycordaites, with pointed leaves approaching those 

 of Eucordaites in size; and Poacordaites, with narrow linear leaves. 

 The parallel venation, unbranched in Poacordaites but repeatedly 

 forked iii Dory- and Eucordaites, suggest monocotyledon foliage 

 and the early writers consequently considered Cordaites as a mono- 

 cotyledon. Both the wood structure, the floral organs, and the seeds 

 were known long before their true nature was appreciated. 



The stem has the general features of a modern conifer, except for 

 the much larger pith, sometimes as much as 10 cm;* in diameter, 



