712 MR W. T. GORDON ON 



Dr Kidston and Professor Gwynne-Vaughan have worked out a scheme of evolution 

 among the Osmundaceae founded upon the hypothesis that the original representatives 

 of that order had solid woody axes, and this is receiving much support from a com- 

 parison with other groups. The specimen described below has also a very distinct 

 bearing on this question. 



All the above-mentioned recent research has been from an evolutionary and 

 taxonomic standpoint, with the intention of clearing up many of the difficulties 

 continually met with in studying a flora which is at best fragmentary. 



The classification of palaeozoic ferns has been attempted from two standpoints. M. 

 Paul Bertrand (2), who has just published an exhaustive treatise on the Zygopteridese, has 

 taken the form of the petiole-trace as his basis of classification, and has demonstrated 

 five distinct series among the members of that order. He does not, however, lay much 

 stress on the anatomy of the stem. As few zygopterid stems are known at present, this 

 classification founded on the petiolar structure is quite a feasible one. As our knowledge 

 of the stems of this order increases, modifications may have to be made in his scheme. 



In their work on the fossil Osmundacese, Dr Kidston and Professor Gwynne- 

 Vaughan (6) have called attention, on the other hand, to the nature of the stem 

 structure. They show that the stems of the older members of that order exhibit 

 characters which may be regarded as primitive, such as the absence of any pith in the 

 centre of the stele and the protostelic departure of the leaf traces. The fossil Osmun- 

 dacese can thus be arranged in an ascending series, and their occurrence in geological 

 time corresponds with their position in this series. A parallel development can be 

 demonstrated among the Lepidodendreze. In the zygopterid alliance, however, no such 

 series has been established ; but, as these authors remark,* " if the zygopterid line of 

 descent is at all close to the Osmundaceous, we must be prepared for the existence 

 of a Zygopteris with a solid xylem like that of Thamnopteris, the central elements 

 consisting entirely of short tracheae without any parenchyma. Such a discovery, in 

 fact, we hopefully anticipate." 



In this paper it is proposed to describe a zygopterid stem which satisfies the 

 requirements of the above prediction. Some specimens of a portion of the petiole 

 belonging to this fern have been recorded previously, but these fragments have not 

 been sufficiently long to give an adequate idea of the complete structure even of the 

 petiole itself. Of course, specimens of fossil plants showing all their various members 

 in organic continuity are of the rarest occurrence, but I have been fortunate enough in 

 this case to obtain specimens exhibiting stems, petioles, and roots in actual connection 

 with one another. 



Zygopterid stems and petioles have been obtained from several different horizons, 

 but the specimens described here were derived from the lowest of the great Carboni- 

 ferous subdivisions — the Calciferous Sandstone Series (Culm) — to which the rocks at 

 Pettycur belong. The petrifactions occurring at Pettycur are either found lying loose 



* K. and G.-V., Trans. Roy. Soc, Edinburgh, vol. xlvi. pt. iii., pp. 663, 664. 



