444 Scott .— The Structure of Mesoxylon multirame. 
stele, contrasting strongly with the naked elliptical shoots of M. multirame 
with their remarkable flattened stele (cf. PI. XIII, Fig. 22, with Maslen’s 
PI. XXXVI, Fig. 20). Of course this difference cannot be taken as 
a specific character; it is more likely that the distinction between the two 
kinds of axillary shoots depends on function ; the one, for example, may be 
vegetative, the other reproductive. There is some additional evidence, given 
below (p. 452), bearing on the function of the axillary shoots. 
The Wood and Phloem . 
The centripetal wood of the leaf-traces is shown in a bundle from the 
outer cortex (one of eight forming the trace) in PI. XIII, Fig. 18, and in 
a double bundle at the margin of the pith in Fig. 19 (see also the photo¬ 
graphic PI. XI, Figs. 3, 4, and 5). In the type-specimen it is perhaps rather 
less developed than in M. poroxyloides. The parenchyma accompanying 
the protoxylem is present, but in transverse section is not always easily 
distinguishable from the xylem^Pl. XIII, Fig. 19). 
The structure of the secondary wood may be said to be identical with 
that in M. poroxyloides ; in particular, we find everywhere a considerable 
thickness of spiral and scalariform elements (not less than 6-8) before the 
typical pitted tracheides of the secondary wood begin. In this respect 
M. multirame agrees with M. poroxyloides and differs from M. Sutcliffii and 
M. Lomaxii , in which the non-pitted type of wood is nearly or quite limited 
to the neighbourhood of the leaf-traces. The pitted tracheides commonly 
have two rows of alternating pits on their radial walls—sometimes there is 
a single row only. The tracheides range from 24 to 36 \x in diameter ; the 
rays are narrower—about 18-24 M- 
There is in some parts of the wood a large proportion of very low rays 
1-3 cells in height. Higher rays also occur, up to a height of 12-14 cells. 
Occasionally a ray-cell is found divided radially, making the ray biseriate at 
that point. The phloem and pericycle show no clear difference from those 
of M. poroxyloides. They are better preserved in another specimen. Imme¬ 
diately outside the pericycle a zone of periderm is already forming (PL XI, 
Fig- 4 )- 
The Cortex . 
The cortex, where completely preserved, is a broad zone 3-3*5 mm. in 
width. Between the leaf-traces the greater part of this zone is taken up by 
the mechanical Dictyoxylon tissue, which is extremely well developed ; 
only a narrow band of parenchyma intervenes between the pericycle and the 
beginning of the fibrous strands (PI. XI, Figs. 1 and 2). But here, as in other 
species, the outgoing leaf-trace is accompanied by a mass of parenchyma, 
and at these places the Dictyoxylon zone is consequently much reduced in 
width (Figs. 2 and 3). The fibrous bands of the Dictyoxylon cortex often 
form a network as seen in transverse (Figs. 2 and 3) as well as in tangential 
