50 Scott and Brebner . — On the Secondary 
It will be seen that this outer secondary zone shows a 
general agreement with the corresponding tissues of a 
Dracaena. 
The inner zone has a more remarkable structure. The 
phloem-groups stand out plainly enough, but the outer limits 
of the bundles are often impossible to trace. The whole 
appearance rather suggests some anomalous Dicotyledon 
with ‘phloem-islands’ imbedded in a continuous mass of 
wood. The bundles are in fact to a great extent confluent, 
and are only here and there separated by a radial or tan- 
gential row of parenchymatous cells. The great bulk of 
the tissue in this zone consists of the tracheides of the 
crowded bundles. Consequently it is not surprising to find 
that no regular radial arrangement is evident. This is also 
partly accounted for by the mode of development, which 
will be explained below. The whole structure is the ex- 
pression of a network of bundles, with thick strands and 
nodes, and very small meshes (occupied by parenchyma) 
between them. 
The limit between the outer and inner secondary zones 
is a fairly sharp one. So also is the boundary of the primary 
cylinder, which is easily distinguished by its scattered bundles, 
circular in transverse section, imbedded in sclerotic ground- 
tissue. 
This will perhaps be the best place to say a few words 
as to the periderm and the secondary cortex. 
The first periderm forms at about the time when secondary 
thickening begins. The seat of its formation at the sides 
of the flattened stem is the hypodermal layer. Towards 
the narrow edges of the stem the next inner layer of the 
cortex takes up the division, then a more internal layer, 
and so on. Thus opposite the prominent edges of the stem 
the phellogen is deep-seated, and gives rise from the first 
to internal periderm. The thickness of the cortical layer 
within the periderm is consequently about the same all 
round the stem. 
The phellogen usually forms two layers of phelloderm on 
