THE FOSSIL OSMUNDACE^:. 467 



the establishment of a new order to receive them. At the same time, when the great 

 range in structure of the vascular system found in the stems of such orders as the 

 Polypodiacese, Gleicheniacese,, Schizteacew, etc., is taken into account, there is absolutely 

 nothing in the anatomy of Zalasskya and Thamnopteris that forbids their acceptance 

 as true Osmundacese* There is, in fact, a considerable amount of positive anatomical 

 evidence in favour of this conclusion. In particular, we lay stress upon such a small, 

 but we believe important, point as the occurrence of secretory sacs in the pericycle of 

 the petiolar meristeles of Thamnopteris (Part III., PL XLVI. figs. 36 and 37) — a feature 

 which has hitherto only been recorded in the leaf- trace of the modern Osmundacese. 

 The geological antiquity of the two genera presents no serious obstacle to their inclusioo 

 in the order ; for sporangia, both isolated and attached to fronds, which might well be 

 regarded as Osmundaceous, are not unfrequently met with even so far down as the 

 Carboniferous formation, f 



In relation to this subject the question also arises as to whether all the species at 

 present placed in Osmundites ought really to be included in the same genus. In the 

 first place, some of them, such as Osmundites Dunlopi, and perhaps O. Kolbei, are so 

 distinctive as regards their anatomy that if all parts of the plant were known they 

 might perhaps require a genus or genera of their own. If even so little as the anatomy 

 of the stem and petiole were completely known, such a step might become advisable. 

 The porose layers might perhaps serve as a useful criterion for this purpose. These 

 layers are known to be wanting in Zalesskya and Thamnopteris, but unfortunately the 

 structure of the soft peripheral tissues of the stele in Osmundites Dunlopi and O. Kolbei 

 is as yet unknown. It would, therefore, be premature to alter their generic names in 

 the present state of our knowledge. 



The Medullation of the Osmundaceous Stele. 



To return to the question of the medullation of the stele in the Osmundacese. The 

 first step towards this result was the transformation of the tracheae in the central region 

 of the xylem into short, comparatively thin-walled elements, which, however, were still 

 tracheal in nature.^ From this point there are two paths by which the final advance 

 to a purely parenchymatous pith might proceed : (1) by the gradual but simultaneous 

 conversion of all the central tracheae into parenchyma ; (2) by the early conversion of 

 some of these elements while others still retained for a time their tracheal characters. 



* The striking similarity between these fossils and the stocks of the modern Osmundacece, even when judged 

 superficially by the external appearance alone, has been pointed out by Mr Carruthers, although he does not go so far 

 as to include them in the order (Carruthers, Quart. Journ. Geol. Soc, vol. xxvi. p. 352, 1870). 



t Scott, Studies in Fossil Botany, 2nd edit., p. 291, 1908 ; Renault, Cours de botanique fossile, pi. xxi. figs. 12-14, 

 1883 ; Zeiller, (1) "Etude sur la flore fossile du Bassin Houllier d'Heraclee," Mem. Soc. G4ol. de France, Mem. 21, 

 p. 21, 1899 : (2) Elements de pale'bot., p. 65, 1900 ; Bower, " Is the Leptosporangiate or the Eusporangiate the more 

 primitive type of Ferns?" Ann. BoL, vol. v. p. 126, 1891. 



I The idea of a direct transformation of the central xylem of a protostele into parenchyma and other tissues has 

 already been put forward by L. A. Boodle in the Schizeeacem : 1, "On the Anatomy of the Schizxacem," Ann. Bot., 

 vol. xiv. p. 410, 1901 ; 2, "Further Observations on Schizeea," Ann. Bot., vol. xvii. p. 530, 1903. 



