XX V] ANKYKOPTERIS 461 



diflScult to explain the occurrence of tyloses in tracheae not 

 immediately in contact with living parenchyma. It may be 

 that the pits in the tracheae of Ankyropteris were open spaces 

 as in the xylem of recent ferns described by G^vynne-Vallghan, 

 and if so this would facilitate the invasion of the conducting 

 elements by growing cells. A comparison is made by Weiss 

 between certain cell-groups found by him in the tracheae of 

 Ankyropteris and by Miss Jordan' in the vessels of the re- 

 cent dicotyledon Ciicumis sativa. In a more recent paper on 

 tyloses Miss McNicol'' expresses the opinion that pseudo- 

 parenchyma in the tracheae of the fossil petioles owes its origin 

 to fungal hyphae. 



Williamson compared the petiole bundles of Ankyropteris 

 corrugata with those of recent Osmundaceae, a comparison 

 based on the structure of the leaf-trace before its separation 

 from the stem and its assumption of the H-form. It is 

 noteworthy, however, that this comparison has acquired a 

 greater significance as the result of recent work. The stele 

 of Ankyropteris bears a fairly close resemblance to that of 

 Zalesskya described by Kidston and Gwynne-Vaughan ; in 

 both types the xylem is represented by two kinds of tracheal 

 tissue. In the Permian Osmundaceous genus the centre of 

 the stele consists of short storage tracheids, while in Anky- 

 ropteris we may regard the central parenchyma and scattered 

 tracheae as derivatives of the solid xylem core of some ancestral 

 type. Moreover, the appearance and arrangement of the phloem 

 and the tangentially elongated elements external to it (fig. 314) 

 remind one of the extra-xylem zone in recent Osmundaceae. 

 That the Osmundaceae and Zygoptereae are closely related 

 groups there can be little doubt ; of this affinity and common 

 origin' Ankyropteris corrugata affords striking evidence. 



The difference between the steles of Ankyropteris Grayi 

 and A. scandens (figs. 310, D; 311) and that of Ankyropteris 

 corrugata is comparatively small. In the two former species 

 the cylindrical form has become stellate owing to the radial 

 extension of the xylem arms. It may be that this more 

 elaborate style of vascular construction is connected with the 

 1 Jordan (03). = McNicol (08). ' Gwynne-Vaughan (09). 



