Maslen. — The Relation of Root to Stem in Catamites . 67 
this is clear from the distinctness of the vascular bundles (x) and wide 
medullary rays (m. r.). This photograph shows only one of the two roots 
which can be seen in the section (the other one is shown in Fig. 3), but the 
centre of the one figured (r.) appears to be exactly in a line with the leaf- 
trace bundles (/./.) of the stem. This is therefore a first point of difference 
from the stem-branches, which always arise distinctly above the level of the 
leaf-traces, i.e. above the node. 
Fig. 3 is made from the same slide as Fig. 2, but it shows the other 
root which is coming from the same node, and is more highly magnified 
than the latter. This photograph shows clearly two of the stem-bundles 
(x. x.) with a wide medullary ray (m. r.) between them and a root (r.) at the 
node above. The section becomes more tangential in the upper part, and 
the root shows the connexion with the secondary xylem of the stem above 
and with the nodal wood below. The position of the root with reference 
to the bundles of the stem is quite clear. It lies midway between the two 
bundles which are coming up from below, and indeed, in this respect, 
occupies precisely the same position as the stem-branches. This will be 
clearly seen by comparing our Fig. 3 with similar sections of stem-branches 
given by Williamson and Scott. The similarity of the two is very striking 1 . 
As the vascular bundles in the stem of Catamites usually alternate in posi- 
tion in successive internodes, it follows that both branches and roots will 
usually be opposite to a bundle in the internode above the node from which 
the lateral member springs, but opposite to a medullary ray of the internode 
below. This difference may be made use of in determining whether a 
transverse or oblique section of a stem which also shows one or more roots 
is cut above or below the node from which the roots arise. It is clear, 
for example, that the sections from which Figs. 9, 10 and 11 (Plate II) 
have been made were cut below the node from which the root shown was 
developed, as the centre of the root is in a straight line with a medullary 
ray of the stem and not with one of its vascular bundles. That the 
stem-bundles of successive internodes do not always regularly alternate 
is well known, and indeed an example of it is shown near the centre of 
Fig. 2, but it is sufficiently constant to make this method of some value. 
Fig. 3 does not clearly show leaf-trace bundles quite close to the root, 
i. e. as close to the root as they are shown to the stem-branches in some of 
Williamson and Scott’s figures. Could they be seen, they would probably 
appear at the sides of the root instead of distinctly below it, as is the case 
with the stem-branches. The evidence for this statement is as follows. 
Although no leaf-traces can be made out quite close to the root shown in 
Fig. 3, they are clearly seen in other parts of the section, and on the same 
node as that from which this root arises : the centre of the latter is exactly 
1 Compare, for example, our Fig. 3 and Williamson’s figure showing a stem-branch reproduced 
in Dr. Scott’s Studies in Fossil Botany, Fig. 9, p. 30. 
