GEOLOGICAL HISTORY OF GYMNOSPERMS 31 



mont of many students is also entitled to rank as a phylum 

 under the name of Cycadophyta as suggested by Nathorst. 

 There remain four groups commonly considered as orders and 

 probably of independent fern ancestry: (1) the Paleozoic Cor- 

 daitales, (2) the rather circumscribed Ginkgoales, which ranged 

 from the late Paleozoic to the Recent, and enjoyed a consider- 

 able adaptive radiation during the Mesozoic, (3) the Gnetales, 

 which Lignier would consider angiosperms, and which Wieland 

 would apparently derive from the cycad alliance, and which may 

 be passed over because they have no known fossil representa- 

 tives, and (4) the Coniferales. These four groups, or at least 

 the first two and the last, should be united in a larger unit of 

 phyletic rank for which the term Coniferophyta proposed by 

 Coulter may suffice. 



The oldest known floras, those of the middle and later Paleo- 

 zoic, contain two very distinct and yet not entirely unrelated 

 types of gymnosperms — ^the Pteridospermae or seed ferns and 

 the Cordaitales. 



The Pteridospermae are the more primitive of the two and 

 show every evidence of fern ancestry. Their stems were known 

 for a generation or more before the detached fronds and seeds 

 were correlated with them, and these stems because of their 

 synthetic morphological characters were long known as Cycado- 

 filicales. The Seed Ferns had foliage that was entirely fern- 

 like in form and habit, as for example in the Sphenopteris type, 

 which was highly decompound. These fern-like fronds had 

 always been considered to be those of ferns although Stur as 

 long ago as 1883 had suggested that they might be cycadaceous. 

 The seeds were large and complex but were borne on fronds that 

 were but slightly modified from the ordinary vegetative type 

 and the pollen was borne in what might well be termed synangia 

 on similarly unmodified fronds. Throughout the whole group 

 nothing approaching cones is known, and the anatomy of stem or 

 leaf, or both, was of the fern type. There was naturally great 

 diversity of detail in this large plexus of forms and some are 

 much more full}^ known than others. 



At the present time at least a dozen genera based on stem 



