CH. XXVI] A GENERAL REVIEW OF THE SIMPLICES 229 



nising these as the primitive representatives of the Class. This conclusion, 

 thus fully substantiated from the Fossil Record, emerges equally as the re- 

 sult of comparative stud}'. The broad ground of priority thus taken up is 

 the natural starting point for phyletic seriation. 



Making use of the criteria established in Volume I, it appears that the 

 Coenopteridaceae, and especially the Botryopterideae, take their place as the 

 most primitive of the Families of Ferns hitherto recognised as such. The 

 dichotomous branching and solid protostele of the stem of Botryopteris, its 

 monodesmic leaf-trace, oval in section and with a single protoxylem at its 

 departure, the hairs consisting of a simple row of cells, the leaf with narrow 

 dichotomous segments, and the Eusporangiate sporangia, marginal or distal, 

 with large spore-output and complex annulus, are characters that define an 

 extremely primitive type. When it is added that B. antiqica comes from 

 the Petticur Beds (Calciferous Sandstone), it is clear that this primitive type 

 was very ancient. It was also of small size, as were many other very early 

 fossils. It has been suggested that B. antiqua, cylindrica, racemosa, and 

 forensis form a series of progressive elaboration within the genus, as they do 

 also in horizon, up to the Permo-Carboniferous of France. 



The Zygopterideae are larger, and characteristically more elaborate than 

 the Botryopterideae, though undoubtedly related to them. Stratigraphically 

 they are recorded from the Upper Devonian, and they extended to the 

 Permian, after which they appear to have died out. It is especially in the 

 foliar development that they are remarkable, whether by vascular structure 

 or by form. Indeed the Dineuroideae provide a unique type of leaf with four 

 rows of pinnae, two rows on each side of the rachis which was probably 

 vertical. This habit is also seen in Staiiropteris, a common fossil of the 

 Coal Measures, of which the axis, if indeed such a part existed, is still 

 unknown. The stele of the stem of the Zygopterids is advanced both in 

 form and in structure, in proportion to its size, beyond that of the Botryop- 

 terids. But though their whole vegetative system appears to have been more 

 elaborate, the Eusporangiate sporangia were still distal or marginal, and the 

 spore-output high, while mechanism of distribution was rudimentary. Thus 

 notwithstanding the enhanced vegetative system, the Zygopterids are pro- 

 perly ranked with the Botryopterids as very primitive Filicales. 



Comparison, unfortunately without the support of fossil history, justifies 

 a position for the Ophioglossaceaein relation to the Coenopteridaceae. They 

 share with them many characters, such as moderate size and simple form, 

 with simple hairs as dermal appendages. The leaf is bifacial, but there is in 

 the spike of HelviintJiostacJiys a structure which offers the nearest similarity 

 to the Dineuroid or Stauropterid frond that is known among living plants. 

 The stem has a simple stelar structure, but with more advanced medullation 

 than in any Coenopterid; moreover in each of these families occasional 



