576 F. O. BOWER. 
even to be found embedded in the secondary phloem, in which 
case they must have originated from cells of the bast parenchyma, 
or directly from the cambium. The longitudinal axis of the 
spicular cells is usually, but not always, nearly parallel to the 
axis of the root. Sclerenchyma fibres are also to be found in 
considerable numbers in the old root: they lie scattered singly in 
the parenchyma, chiefly around the periphery of the vascular 
bundles ; they take a sinuous, longitudinal course. Gum pas- 
sages are also to be found in the roots: they appear irregular in 
their distribution (fig. 1v). 
The parenchyma in which all these several tissues are em- 
bedded deserves special attention, since many of the abnormalities 
of structure of the root and also of the stem of Welwitschia are 
the outcome of its peculiar properties. The most remarkable of 
these is its constant capability of cell division (Halb-meristeme- 
tisch, De Bary, ‘ Vergl. Anat.,’ p. 634). The parenchyma seems 
in no living part of the plants which I have examined to have 
passed over to the permanent condition ; it seems normally to be 
subject during its whole life to occasional division. It is true 
that the activity of this division is not uniform, but that it is 
specially localised at certain points, such as (a), the base and sides 
of the leaf groove (fig. x) ; (4) the point of origin of new vascular 
bundles (fig. v) ; (¢) the phellogen layer; but that does not alter 
the fact that cell division may and usually does occur (at least 
in moderately young plants) more or less actively wherever there 
is in stem or root a parenchymatous ground tissue. We shall 
now be able better to realise what occurs when new peripheral 
bundles are formed, as in fig. v. An active division takes place 
at the point where the new bundle is to be formed; the surround- 
ing tissue keeps pace with the radial extension thus produced, 
but the activity is merely local; there is no tangentially conti- 
nuous ring of activity dividing tissue, such as that in the mono- 
cotyledonous stems with secondary thickening. 
Though this activity of division is well marked in the paren- 
chyma of the root, it is in the stem hereafter to be described that 
it is most striking; and, indeed, the swelling of the stock, which 
starts from the apex and proceeds downwards is mainly due to 
this activity. 
A further peculiarity of the parenchyma is this—that appa- 
rently any cell may develop into a spicular cell or sclerenchyma 
fibre, and that this development is not restricted to any particular 
zone, but seems to be another proof of the general activity of 
the parenchyma. We find spicular cells in very different stages 
of development, occurring side by side, both in root and stem 
(fig. xx). (For the development of the spicular cells, see below, 
p. 590). 
