540 FILICALES 



structure certain fossils attributed to the Botryopterideae have pronounced 

 Osmundaceous characters : in view of the sequence of fossils above quoted, 

 it seems probable that the Osmundaceous structure is referable in origin, 

 with upward differentiation of the stele, to some type of the nature of the 

 Botryopterideae (Kidston and Gwynne-Vaughan). It thus appears that a 

 study of the related fossils in their stratigraphical sequence lends no serious 

 support to a theory of reduction of the stele from an " amphiphloic siphono- 

 stele " : it indicates rather an upward development from a protostelic state. 

 Taken with the comparative considerations already advanced, the evidence 

 against Jeffrey's view appears very strong indeed. 



Zenetti l had already regarded the stele of Osmunda as being in the 

 up-grade of development, and had compared it with the structure seen in 

 certain of the Lycopodiales. With these a very interesting parallel may 

 be drawn, and especially with that series illustrating a progressive elaboration 

 of the stele, and its disruption into separate strands, which Kidston has 

 recently demonstrated by a stratigraphical sequence of fossils as cogent as 

 this in the Osmundaceae (see above, pp. 230, 337). The fact that such 

 parallels have been shown to exist in distinct phyla is in itself a support of 

 the views above indicated. 



EMBRYOLOGY. 



The primary embryology of the Osmundaceae being on the same 

 plan as that of the Leptosporangiate Ferns as a whole, it need not be 

 described in detail. The Leptosporangiates all differ from the Marattiaceae 

 in the position of the basal wall : in the latter it is transverse to the axis 

 of the archegonium, in the former it is parallel with it : in relation to this 

 the epibasal half, which gives origin to the axis and leaf, is here directed 

 laterally, and ' the cotyledon originates from its lower quadrant. The 

 consequence is that, as in all the other Leptosporangiates, the cotyledon 

 of the Osmundaceae emerges on the lower side of the prothallus, not from 

 the upper as in the Marattiaceae. Comparing the embryo itself with that 

 of other Leptosporangiate Ferns, it will suffice to remark that in the 

 segmentation there is somewhat less regularity in the later divisons, and \ 

 that the external differentiation of the members appears later, the embryo 

 retaining longer than in them its spherical form. These are but minor | 

 differences; they indicate, however, for the Osmundaceae an intermediate 

 place between the typical Leptosporangiates and the Eusporangiate Ferns. 2 



A similar intermediate character comes out also from comparison of 

 the meristems of the Osmundaceae with those of the Marattiaceae on the 

 one hand, and of the typical Leptosporangiate Ferns on the other. I 

 have shown at length elsewhere, 3 that in respect of the apices of root, stem, 



1 Bot. Zeit.^ 1895, pp. 72-76. ' 2 For details see Campbell, Mosses and Ferns, p. 356. 



3 Quart. Journ. Micr. Sci., vol. xxv., 1885, p. 75, etc. ; Phil. Trans., 1884, part ii, p. 

 565; Ann. of Bot., vol. iii., p. 305. This- matter will be taken up again later, when 

 the general comparison of Ferns is made, and also in Part iii. 



