12 
Proceedings of the Royal Society of Edinburgh. [Sess. 
which communicates directly with the cortex through very wide leaf-gaps. 
The extreme condition is seen in the tuberous stocks of 0. palmatum, 
which may be as much as 2 cm. in diameter. This seemingly reckless 
discarding of the protective endodermis goes along with a leathery foliage. 
Such plants have only a sluggish circulation of fluids, and the protection 
of their conducting tracts seems less vital for them than the establishment 
of free gaseous and other interchange between the tissues of their sappy 
stocks. 
The parallel condition seen in the Marattiacese indicates the truth of 
this. The massive stocks of these ferns are also soft and sappy, and their 
leaves are as a rule leathery and thick. They dispense early in their 
ontogeny with the endodermis, and the stele at once breaks up into small 
parts which are widely scattered through the transverse section. The stem 
grows to a large size, with no limit between the distended pith and the 
cortex. Consequently after the brief juvenile stage is past no question of 
proportion of surface to bulk arises. But, on the other hand, by discarding 
the endodermis the conducting tracts have lost that protoplasmic control 
which the endodermis gives. This state may serve for semi-xerophytic 
plants, such as the Ophioglossaceae and Marattiacese, with sappy stocks and 
leathery leaves, and sluggish fluid-transit. But it would not serve for 
plants where fluid-transit requires to be rapid, and in particular for those 
with delicate leaf-structure. 
The Leptosporangiatse, which are mostly delicate hygrophytes, com- 
prise the vast majority of living species of ferns. They have taken a 
quite different course of structural development, in which the endodermal 
barrier is strictly maintained in its complete form, while intercellular 
spaces are as a rule absent from their vascular tracts. They show in their 
peculiar vascular structure to what shifts a plant is put as it increases in 
size by primary and not by cambial activity, maintaining meanwhile its 
vascular system under complete protoplasmic control. All of them start 
from the protostelic state. It appears from comparison along phyletic 
lines parallel but yet distinct, that a transition has taken place from the 
protostele to a disintegrated stelar structure. The successive steps of this 
may be seen with varying degrees of clearness of detail in the successive 
stages of the individual life. These steps appear as the stele enlarges. 
According to the reasoning already brought forward, it is on enlargement 
that the problem of proportion of surface to bulk of the stele becomes 
insistent. The modifications of form of the stele seen in the advanced 
Leptosporangiate Ferns may be held as the means of its solution. The 
critical point in the individual development of the solenostelic type is 
