4IO MORPHOLOGY OF GYMNOSPERMS 



(called provisionally Primofilices) is as certain as phylogenetic con- 

 nections can be. The origin of the Cordaitales, therefore, presents 

 two alternatives: either they arose independently from the same 

 ancient fern stock, or they were differentiated from the Cycadofilicales 

 very early. To choose between these alternatives is not very impor- 

 tant, but the latter one seems to be the more reasonable, because the 

 Cordaitales (as we know them) are much more removed from the 

 ferns than are the Cycadofilicales. If this conclusion be accepted, 

 it follows that gymnosperms began with Cycadofilicales more ancient 

 than any yet known; that Cordaitales branched off from Cycado- 

 filicales earlier than our present records; and that the two groups 

 constituted the extensive gymnosperm flora of the Carboniferous. 



This paleozoic display of g)minosperms was succeeded by a meso- 

 zoic display, in which at least four groups are recognized. From the 

 Cycadofilicales there arose the mesozoic Bennettitales and the Cyca- 

 dales; and from the Cordaitales the mesozoic Ginkgoales and Conif- 

 erales were derived. The Bennettitales have been traced almost 

 to the Paleozoic, and their structure, as well as the habit of some of 

 the earlier forms, makes their connection with Cycadofilicales appear 

 convincing. The relation of the Beimettitales to the Cycadales is 

 not so clear; either the two groups were differentiated from a common 

 stock that arose from the Cycadofilicales and continued into the 

 Mesozoic, or the Cycadales were differentiated early from the Ben- 

 nettitales. The records indicate that the Cycadales are much younger 

 than the Bennettitales, and were much more scantily represented 

 during the Mesozoic; and therefore the latter alternative seems to 

 be the more reasonable. In any event, the only question at issue is 

 whether the gymnosperm stock which came from the Cycadofilicales 

 into the Mesozoic is to be called Bennettitales or a Bennettitales- 

 Cycadales plexus (a "common stock"); and it is altogether probable 

 that this stock has already been assigned to the Bennettitales in the 

 description of cycadophyte forms from the Triassic and Jurassic. 

 In the gymnosperm flora of today, therefore, the Cycadales, although 

 relatively a modern group, are the nearest representatives of the 

 paleozoic Cycadofilicales. 



The Ginkgoales and Coniferales have both been traced into late 

 and independent paleozoic connection with the Cordaitales, and were 



