10 
Proceedings of the Royal Society of Edinburgh. [Sess. 
as a place of storage for water, but also for plastic substances. In the 
living Osmundacese it contains large quantities of starch. As usual in 
storage tissues, intercellular spaces are present, though these are absent 
from the simplest steles. Here they form an internal ventilating system, 
quite separate from that of the cortex, excepting that in some species there 
is communication at the point of dichotomy of the stem. A careful 
examination of Osmunda regalis and of Todea barbara shows no con- 
nection at the xylic gaps between the outer and inner ventilating systems. 
The endodermis surrounds the stele completely as well as each leaf-trace. 
Consequently the inner ventilating system is here as isolated and self- 
contained as is the intercellular system of a submerged plant. Where the 
Fig. 7. — Traces of the actual sizes of steles of living and fossil Osmimdacese, 
all to the same scale, i.e. approx, nat. size. 
i. Todea barbara (3 mm.). iv. Thamnopteris schlechtendalii (13 mm.). 
ii. Osmunda cinnamomea ( 4 mm.). v. Osmundites skidegatensis { 25 mm.). 
iii. Osmunda regalis (5 mm.). vi. Osmundites Carnieri (35 mm.). 
size is small this condition is possible. It is so in T. barbara with a stele 
3 mm. in diameter, or in 0. regalis (5 mm.). In such instances the propor- 
tion of surface to bulk of the small stele is relatively high (fig. 7, i, ii, iii); 
but the case is different for the large steles of Osmundites skidegatensis 
(25 mm. in diam.), or of 0. Carnieri (33 mm. in diam.), and in them the 
problem is solved by breaking down the barrier. In the former fossil each 
leaf-trace at its departure interrupts the whole vascular ring, and the pith 
is continuous with the cortex through each leaf -gap. Moreover, no layer 
resembling an endodermis can be distinguished, so that it is practically 
impossible to set a definite limit to the stele. In 0. Carnieri , though a 
line of delimitation appears which is believed to be endodermis, it is 
discontinuous at irregular intervals, and the ventilated cortex is directly 
related to the greatly distended pith (fig. 8). Thus in both of these 
large fossil stems a concomitant, and it probably was even a necessary 
condition of their large size, was this interruption of the endodermal 
