[PENHALLOW] OSMUNDITES SKIDEGATENSIS 13 
Radial extensions of the pith project into the stele between the arms 
ot the curved xylem bundles. These consist of thin-walled elements with 
prominent nuclei but no starch, all along the region of contact with the 
xylem; but in the larger of these projections, which then enlarge to- 
wards the free end, the central cells become thicker-walled and con- 
tain an abundance of starch exactly as in the fossil. 
The Medullary Rays.—The rays consist of thin-walled elements with 
prominent nuclei but no starch, except where unusually broad. Then 
the central cells become thick-walled and are filled with starch. Out- 
wardly they spread right and left along the outer face of the xylem, lying 
between it and the sieve cells, and thus connecting with the inner 
phloem. (Plate VI., fig. 11.) 
The Stele-—The stele has an average thickness of 0°75 mm. and an 
extreme external diameter of 2°5 mm.* In O. cinnamomea these dimen- 
sions are somewhat larger, the external diameter being 3 mm. In the 
fossil it has been found that the stele has an average thickness of 3 mm. 
and an external diameter of 19 mm., from which it appears that it has a 
diameter 7°06 times greater than in O. regalis, and an average thickness 
about four times greater. From this an inference may be drawn as to 
the relative dimensions of the two plants. The observation that the 
rhizome of Todea barbara, with an external diameter of 11°5 cm., and 
therefore more than twice the diameter of O. regalis, nevertheless has a 
stele only slightly larger, and of the same size (3 mm.) as in O. cinna- 
momea, appeared at first to indicate that relative dimensions cannot be 
established by comparison of the steles. But when it is recalled that 
the much greater angle at which the stipes are given off in Todea offers 
a complete explanation of the greater external dimensions, and that both 
O. cinnamomea and the fossil are essentially the same with respect to the 
divergence of the leaves from the central axis, as well as in the dimen- 
sions of the stele, it will be found that the latter offers a fairly safe 
basis upon which to estimate differences in size of the plant as a whole. 
From this, therefore, it is fair to assume that the fossil under considera- 
tion was about seven times larger than O. regalis, and this view is also 
borne out in part by the supposed foliage. 
The xylem bundles have the horseshoe shape so well known in this 
genus, and are about 11 in number. The elements are much narrower 
than in the fossil and somewhat broader than in O. cinnamomea. 
(Plate V., fig. 10.) In both O. regalis and O. cinnamomea, they also 
show the structural features described for the fossil, and this very close 
resemblance is fully borne out in longitudinal section. 
1 This does not agree with De Bary’s statement that the stele is 6 mm. 
thick. (Comp. Anat. of Phan. & Ferns, p. 279.) 
