XVIII] ZYGOPTERIDEAE 17 



originally suggested in 1884 (Bower, Phil. Titans. 1884, Part II, p. 565). The 

 suggestion was then based only upon analogy derived from a comparative 

 study of Fern-leaves. The results of that study have been re-stated and am- 

 plified in Vol. I (Chapter v, on Leaf-Architecture). The question was further 

 discussed in Vol. I, Chapter XVII. In Chapter IX it has been argued (p. 175) 

 that the most valuable features for comparison are found in the basal region of 

 the leaf, and in the leaf-trace itself If this principle be put in practice here, 

 the argument in favour of the dichopodial origin of leaf and axis in Ferns 

 acquires added strength. It has further been shown that the embryological 

 facts for Ferns would accord with this view; in particular it has been seen 

 that a dichotomising embryo, such as the hypothesis would demand, actually 

 exists in Tviesipteris, with that equal or unequal development of the shanks 

 that the hypothesis presupposes. In view of the cumulative effect of these 

 various facts the hypothesis may now be entertained that the shoot in the 

 genus Botryopteris, and perhaps also of Ferns generally, originated as a 

 dichopodial development from a primitively equal dichotomy, such as is 

 actually seen in the adult form of the Psilophytales. 



B. ZYGOPTERIDEAE 



This family of fossil Ferns includes a number of generic types (about 

 twelve), which are best known by the anatomical features of the leaf and some- 

 times also of the stem. But in some of them the axis has never been seen 

 {Stauropteris). The leaf-form and general habit, as well as the sporangia, 

 are only known in a few of them. Accordingly the tracing of relationships 

 between themselves, and their comparison with other families of Ferns, must 

 be held as provisional. They have been found to include some forms with 

 massive upright stems and crowded leaves {Asterochlaena, Asteropteris): in 

 others the axis appears to have been elongated and rhizomatous with isolated 

 leaves {Diplolabis, Metaclepsydropsis) : others again had a thin slightly elon- 

 gated stem, and these are believed to have had a straggling or a climbing habit 

 {Ankyropteris). In fact in these early P'ern-types those differences of habit 

 are already to be found which characterise Ferns of the present day. The 

 plants were attached to the soil by numerous adventitious roots. 



A feature which is common to them all, and distinguishes them from the 

 Botryopterideae, is the greater complexity of their vascular system whether 

 of the axis or of the petiole. This goes along with larger dimensions of both, 

 as will be shown later. It may be a question how far the increased com- 

 plexity is a direct consequence of the greater size : but whatever view may 

 be taken on this point its diagnostic value may be accepted provisionally. 



Stratigraphically the Zygopterideae were already represented in the 

 Upper Devonian, as seen in Asteropteris noveboracensis from the Portage 

 Beds of the State of New York. They extended through the Carboniferous 

 B. n 2 



