20 
Proceedings of the Royal Society of Edinburgh. [Sess. 
of certain of the larger species of Selaginella suggests that they should 
also be examined from the physiological-anatomical point of view. The 
similarity between the vascular arrangements in S. Icevigata and that in 
solenostelic ferns seems to indicate that there is some common underlying 
cause which has brought such likeness into existence. It is suggested that 
the need for physiological interchange over the surface of an enlarged stele 
has been the determining factor in both cases, though the need for a 
larger surface seems less apparent in Selaginella with its peculiar structure 
of the endodermis than elsewhere. The Medullosse with their recurrent 
“polystely” also provoke comparison. Scott {Studies, vol. ii, p. 444) 
remarks that their polystely has no relation to leaf-gaps, and he suggests 
that “ the breaking up of the original stele depended on some other cause.” 
Does it not appear probable that the cause may have been connected with 
questions of proportion of surface to bulk in the original steles of these 
large stems, before cambial increase set in ? With the knowledge of their 
primary steles and of their sheaths so incomplete as it is, one cannot do 
more than suggest that this is an aspect worthy of consideration, especially 
as it has been found to illuminate the cognate structure in ferns. 
Having thus seen how firmly established the disintegrated stelar state 
is among Pteridophytes, the question arises whether Flowering Plants 
show any similar modifications of stelar structure with enlarging size. 
Most of the Flowering Plants have met their problem of enlargement in 
other ways, as will be noted later. But the large prop-roots of certain 
palms have advanced structurally along lines that show a close analogy 
with what has been seen in the ferns. Such roots are often thick. They 
have normally a cylindrical stele delimited by an endodermis that is 
sclerotic. There is a cortical system of large intercellular spaces, and an 
intra-stelar ventilating system that is often also large. These are separated 
by the barrier of endodermis. Consequently the same problem arises with 
increasing size as in the stems of ferns. Modifications in the direction of 
disintegration, or of “ polystely ” as it is sometimes called, have been 
produced by involution and finally by disruption of the endodermis, in 
Areca catechu, Archontophcenix cunninghami, Dictyosperma aurea, 
Verschaffeltia splendida, and other palms. The departure from type 
appears in roots about 1 cm. in diameter, and it is still more marked in 
those of larger size. In roots about 1 cm. in diameter the stele takes a 
fluted form. In larger roots the involutions are deeper, and the endodermis 
and ring of vascular tissues within it are interrupted by bands of parenchyma 
with continuous intercellular spaces (fig. 18). In roots of the largest size, 
as occasionally in Areca, but more markedly in Verschaffeltia, the separate 
