386 EQUISETALES 
large pith with a central cavity interrupted -by diaphragms at the 
nodes: around it is disposed a ring of vascular strands of number varying 
according to the species, or according to the rank of the axis in question. 
They are separated laterally by broad parenchymatous rays, while the 
whole is surrounded in most species by a continuous endodermis 
(Fig. 211 a, B). There is, however, a good deal of difference in the 
disposition of the endodermis in various species, and these differences are 
of such a nature as to raise questions as to the validity of the simple 
character of the stele itself. The simplest case is that above described, 
and it may be seen in the aerial shoots of Z£. arvense and palustre, where 
there is a simple endodermal sheath of sinuous outline, formed from the 
innermost layer of the cortex; in fact, the arrangement is that most usual 
in Vascular Plants. In this case the term “stele” will naturally connote 
all that lies within that sheath. A second type is that seen in the 
thizomes, but not in the aerial shoots of Z. sy/vaticum (Fig. 211 ©, D), 
in which a second endodermis is present as a sinuous layer, forming an 
inner barrier of demarcation from the inner-lying pith. A third type is 
seen in the rhizomes of £. kiemale and some others, but not in the aerial 
stems of those species: it is characterised by each single strand being 
individually surrounded by a closed endodermal sheath (Fig. 211 §, F), 
while there is no general endodermis delimiting the whole stele. Such 
individual endodermal sheaths also surround the strands in the tubers of 
£. arvense, sylvaticum, and palustre, species in which, however, a general 
endodermis is found in the ordinary axes. The inconstancy of the 
arrangements thus seen, even in the different regions of the same plant, 
indicates them as special and secondary peculiarities, which need not 
seriously affect the conception of the stem as essentially monostelic. The 
fact that the differences of the endodermis do not otherwise affect the 
anatomy confirms thiseconclusion. It may then be held that the stem 
of Eguésetum is monostelic throughout, but subject to disintegration of 
the stele. 5 
The structure of the individual vascular strands, as seen in the transverse 
section of the internode, is fairly uniform in the different species. Each 
strand shows towards its central limit a canal designated ‘“carinal,” 
because it is on the same radius as one of the keel-flanges which mark 
the fluted internode externally (Fig. 211 c). These canals indicate the 
position of the protoxylem-strands, which become obliterated as the sur- 
rounding tissues expand in development; for the primary tracheides are unable 
to keep pace in their own growth with the expansion of the surrounding 
tissues, and accordingly break down. Close on either side of the margin 
of each carinal canal the annular thickenings of one or two or more 
tracheides remain to maturity, and permanently record the position of the 
protoxylem. As we shall see later, these are directly continuous with 
the protoxylem of the leaf-trace. Further out from the centre than the 
canal, and right and left of it, two other groups of xylem arise later: 
