LEAF IN THE VASCULAR CRYPTOGAMS AND GYMNOSPERMS 
5 77 
the base to the upper regions of the leaf, where they are interrupted by the insertions 
of the pinnte. Thus the phyllopodium of Osmunda is a winged structure throughout 
its length, though the mode of the development of the wings is different at different 
parts of it, being more massive near the base, thus forming the sheath, and being less 
bulky, though still traceable, in the upper part. 
Osmunda cinnamomea, L. 
In this species also, a three-sided, conical, apical cell was found at the apex of the 
phyllopodium, and having a similar orientation to that in O. regalis. It may still be 
recognised after a considerable number of pinnae have been formed. In all important 
points the leaf of this plant corresponds to that of 0. regalis (Plate 37, fig. 6). In a 
leaf which had not yet formed any pinnae, I was able to see clearly the relation of the 
massive wings at the base of the leaf to the apical cell; the limits between the rows of 
the segments could be traced as zig-zag lines down to the wings, thus showing that they 
are derived partly from the ventral, partly from the lateral segments (Plate 37, fig. 6). 
In this species also the apical part of certain of the leaves becomes abortive, the 
basal part remaining as a permanent scale, as described by Peantl for 0 . regalis. 
\ 
Todea superba. 
It was in this plant that the three-sided apical cell in the leaf was first found : since 
the Osmundcicece differ from the other Leptosporangiate Ferns in various characters both 
external and anatomical, it was thought that an examination of the development of 
the leaf might bring interesting facts to light. The three-sided apical cell has the 
same orientation as in the two species of Osmunda described above, but its position 
does not appear to be maintained so exactly as in these plants, nor are the walls of 
segmentation so accurately parallel to the sides of the apical cell. Further, the sub¬ 
divisions of the segments present more frequent irregularities (Plate 37, fig. 7). In 
the absence of distinct marginal series of cells in the early stages of the first formed 
pinnae, and in the phyllopodium itself, while such marginal series are to be found 
in the later formed pinnae, and in later stages of the earlier formed pinnae, Todea 
superba corresponds to the above species of Osmunda. A peculiarity of structure of 
the wings on the higher ramifications of the leaf in this species cannot be passed over 
without remark. Transverse sections of the wings in these parts show that they 
consist of two, or at the margin of but one layer of cells ; the corresponding parts of 
T. barbara consist of about nine layers. Thus between the species of a single genus, 
the affinity of which is indubitable, a very marked difference may exist in the bulk of 
the winged developments, and this should be borne in mind in considering the similar 
and simple structures in the Hymenophyllacese. 
Attention must be paid to the basal portion of the leaf of Todea superba. Here, as 
MDCCCLXXXIV. 4 E 
