38 COENOPTERIDACEAE [ch. 



They represent a distinct line of development of the trace designated by 

 P. Bertrand the "Inversicatenal" type. However different this type may be 

 from the rest of the Zygopterids it has this in common, that the vascular 

 tissue is decentralised, and is moulded into a curved tract which runs parallel 

 to the outer surface of the petiole, while ventilating-connection between the 

 outer cortex and the centrally-lying mass of parenchyma is maintained 

 through the gap between the recurved margins. This state is cognate with, 

 but complementaryto, the C-shaped leaf-trace so usual in monodesmic Ferns 

 of the present day (see Figs. 429, 432). In them, however, the converging 

 margins point centrally. From the physiological point of view the result is 

 the same as in Anachoropteris, though phyletically quite distinct. There is 

 in fact an underlying principle related to increase in size of monodesmic 

 petioles, which secures decentralisation while maintaining ventilation. This 

 end has been gained in three different ways in primitive Ferns: (i) by the 

 Zygopterids, (ii) by the Inversicatenal type, and (iii) by ordinary mono- 

 desmic living Ferns. Probably each is a reliable diagnostic character for its 

 distinct phylum, though the three may all have originated from a common 

 monodesmic source having a simple form of trace. 



The upper leaf, though insufficiently known in most of the Coenopteri- 

 daceae, offers still more interesting characters. It must suffice to say that 

 certain Coenopterids appear to have had pinnate leaves of an ordinary Fern- 

 like type : but the interest centres round those Zygopterids which have their 

 pinnae arranged in four rows or orthostichies. The problematical plant 

 Staiiropteris is naturally ranked with these (Figs. 325, 331). The base of this 

 very common fossil of the Coal Measures is unknown, though the details of its 

 leaf, with those of its solitary distal sporangia, have been fully made out (see 

 above, p. 30). The question is then whether this strange plant is to be held 

 as a very primitive type, perhaps stemless, with upward, radially constructed, 

 branching shoot, or whether it is to be held as a specialised derivative from 

 the Zygopterids. Till the base of the plant is known it seems useless to 

 speculate: but it is significant in its bearing on the question that its sporangia 

 are probably as primitive, in structure and in position, as any that have been 

 referred to the Filicales. 



The Zygopterideae have been divided into two groups according as the 

 pinnae are biseriate (Clepsydroideae), or quadriseriate (Dineuroideae) Sahni. 

 Scott points out that as regards antiquity there is nothing to choose between 

 them, but that the early occurrence of Staiiropteris shows that the origin of 

 the quadriseriate type was itself very early. It has been suggested that the 

 paired pinnae of the Dineuroideae resulted from a precocious dichotomy of 

 primary pinnae. If that be true, the biseriate type would have preceded the 

 quadriseriate. On the other hand, Lignier has suggested the derivation of 

 the biseriate from the quadriseriate state by fusion. Kidston and Gwynne- 



