Isabel Soar 
290 
stele to the cortex. The passage of water through these radial walls 
could also be made impossible if the inner tangential walls possessed 
an unbroken suberised layer on their outer surfaces, thus preventing 
the passage of water from the pericyclic tissue into the endodermal 
cell. This was never found to be the case, suberisation when present 
occurring on the inner layer of the tangential wall. Thus the water 
stream from the stele to the cortex flowing in response to the pull due 
to transpiration has two channels through which it may pass. The 
first is the narrow core of lignocellulose in the centre of the radial 
walls. The second path is through the tangential walls and the 
protoplasm of the endodermal cell. It is probable that by far the 
greater part of the water must travel by the latter path as the 
suberisation of the radial and the transverse walls is considerable, 
especially in older leaves. As in the case of roots, control of the rate 
of the flow of water to the cortex doubtless depends on osmotic 
phenomena. 
Passage of water is possible from one endodermal cell to the next 
in a tangential direction through the pits in the radial walls, but is 
impossible in a longitudinal direction as the transverse walls are 
unpitted. 
The relative impermeability of the endodermal walls to water 
travelling from the conducting tissue to the cortex must have some 
effect in retarding the transpiration current, and in this lies the 
importance of the role of the endodermis as contributor to the 
xeromorphy of the leaf. 
The suberisation of the pericyclic tissue which is found in the 
more xerophytic types, such as Finns sylvestris and Picea excels a, 
also helps to reduce the rate of flow of water outwards through the 
leaf. 
An interesting phenomenon is the presence of supplementary 
checks to the passage of water in those regions of the leaf where the 
endodermis is incomplete, or where its walls have not yet developed 
their characteristic thickenings. One example of this is the relatively 
great development of lignified hypodermal tissue at the base of the 
leaf in Picea, where 6-7 layers of this tissue are found and where a 
complete endodermal sheath is wanting, whereas, further up the leaf, 
with the evolution of a continuous endodermis with suberised radial 
walls, we find the hypodermal tissue is reduced to two layers of cells 
only. A similar state of affairs can be observed, but in a less marked 
degree, in Abies. Further, in the case of Pinus, the base of the leaf, 
where intercalary growth is proceeding, and where the endodermal 
