30 EDWARD WILBER BERRY 



carpon and Miadesmia are almost as far advanced in the seed 

 habit as the contemporaneous Pteridosperms. Among the 

 Pteridosperms we have the typical filicalian foliage, both in 

 form and structure, surviving the change from homospory 

 through heterospory to seed formation. Among the cycado- 

 phytes we find the foliage assuming the existing form in the 

 Carboniferous and continuing unchanged throughout the ages 

 during which the plants passed through the series of vascular 

 and floral changes represented by the Mesozoic Cycadeoidales 

 (Bennettitales) and Williamsoniales and the modern Cycadales. 



Are the gymnosperms in the widest sense of the term mono- 

 phyletic or polyphyletic? It seems to me that this is a question 

 of degree rather than kind. Obviousl}- if evolution is true we 

 only have to go back far enough to find common ancestors. In 

 the sense that none of the forms classed as gymnosperms are 

 genetically related except in the remotest way to the group in- 

 cluding the Lycopods and their Paleozoic relatives (which group 

 may properly be called the phylum Lepidophyta) or to the group 

 including the equiseta and their Paleozoic relatives (which group 

 may be properly be called the phylum Arthrophyta) the gymno- 

 sperms are certainly monophyletic. 



In another sense they may with propriety be considered poly- 

 phyletic and for the following reasons: The Paleozoic Pterido- 

 sperms while they grade off into little known forms of quite un- 

 suspected structural possibilities, exhibit in those forms that are 

 better known a sum of characters which precludes them from 

 forming an order in a single taxonomic group that shall be 

 termed Gymnospermae, and it seems to me that they are appro- 

 priately to be considered a separate phylum — the Pteridosper- 

 mophyta, as suggested by Ward. Their sole coniferous character 

 is the micropylar avenue of fertilization, a quite necessary stage 

 in the evolution of the seed habit, as is proved by Lepidocarpon 

 and Miadesmia among the Lepidophyta. It is true that the 

 Pteridosperms do clearly indicate the character of the ancestral 

 stock that leads to existing cycads, but they are probably not 

 directly related, and in the latter we have another group, enor- 

 mously diversified during the Mesozoic, one which in the judg- 



