22 
Proceedings of the Royal Society of Edinburgh. [Sess. 
is due to a progressive change in the mode of differentiation of the 
apical meristem. It is in fact here, as it is also in ferns, a question of 
procambial destination. But how it is, and why it is, that the meristem 
of a part beyond a certain size should in fern-stems and in palm-roots 
differentiate its procambium so as to produce a disintegrated stele, and 
the same root below that size so as to produce an integral stele: and 
why the latter type of meristem should pass into the former, as in the 
ontogeny of the Leptosporangiate Ferns: or the former into the latter, 
as in certain individual roots of palms, must for the present remain an 
open question. It is in fact as great a mystery as is the power of the 
root-tip to receive the stimulus of gravity, and to pass it on to the 
point of reaction. The behaviour of meristems remains now, as it has 
always been, the greatest enigma of the plant-body, and not the least of 
the questions which it raises is this : How does the meristem forecast in its 
embryonic tissues those proportions of surface to bulk which will be 
necessary when the tissues still embryonic shall have matured to their full 
size ? In this the biologist of the present day may see a new application 
for the old word “ Prolepsis.” 
It may be asked, if certain plants show these modifications of structure 
according to size, why are they not more common among the higher plants, 
many of which seem to be of such size as to justify, or even to demand 
them ? The reply is that in the stems and roots of very many seed-bearing 
plants, the difficulty of a delimiting endodermis does not arise, except in 
the young state, and while they are relatively small. In many small 
Monocotyledon-stems, such as the creeping rhizomes of Convallaria 
majalis. a well-marked endodermis may exist. Even in this plant 
Schwendener has recorded a “ perforation v similar in nature to those 
seen in Ferns. But in plants of larger size where the stele is dilated so 
as to take up almost the whole of the transverse section, as is the case 
in the palm-type generally, the endodermis is inconspicuous, and does not 
appear as an efficient barrier. These plants have in fact adopted a con- 
dition which for practical purposes is like that of the Marattiaceae, though 
it has been reached by a different developmental method. Isolated vascular 
strands traverse the parenchyma of the distended stele, and neither they 
nor the stele have an efficient endodermal barrier. 
Those plants which possess secondary thickening, such as the Dico- 
tyledons and Gymnosperms, stand in a different position. Frequently they 
show in their young stems and roots a well-marked endodermis. But 
before the stem or root enlarges to that size which appears to be critical 
in so many plants (that is about 1 cm. in diameter), the secondary cambial 
