Mesoxylon Lomaxii and M. poroxyloides. 1019 
middle into a number of thin discs (PL XC, F'ig. 22; PL LXXXVIII, 
Fig. 1 3). Thus the rupture of the pith appears to have taken place in two 
stages. It is possible that this may have only happened locally. 
The external, persistent zone of pith is about o-8 mm. thick, and 
consists for the most part of thin-walled, fairly isodiametric cells. In the 
inner layers, the cells appear empty or have light-brown contents ; those 
next the central cavity are sometimes flattened (PL LXXXVIII, Fig. 10). 
In the outer half of the zone the cells are a trifle smaller, and many of them 
have almost black contents. In contact with the wood a few narrow, ver- 
tically elongated, delicate-looking cells may be distinguished (PL LXXXVIII, 
P"igs. 10, 11, and 14). 
The diaphragms are composed of larger cells than those of the 
persistent zone ; they appear round or somewhat hexagonal in transverse 
section, and when seen in longitudinal section are found to be much 
flattened (Fig. 13). The tissue of the thin central diaphragms has collapsed ; 
connecting shreds of membrane show that the diaphragms have been torn 
apart by the growth of the stem (Fig. 13). The excellent preservation of 
the persistent zone is a clear proof that the partial destruction of the central 
tissue was a natural process occurring during life. It will be seen that the 
structure of the pith in M. poroxyloides agrees very closely with that of 
M. Satcliffii , described by Mr. Maslen (Maslen, ’ll, p. 391). 
The Leaf-traces. 
In the zone of wood, as shown in a complete section, PL LXXXVIII, 
Fig. 8, three pairs of bundles are seen passing in, and seven more leaf-traces 
can be recognized at the margin of the pith, either in course of fusion or 
already fused. In all these strands the centripetal xylem is well developed. 
As the twin xylem-strands come in through the secondary wood they 
converge (PL LXXXVIII, P'ig. 9), and almost immediately on reaching the 
border of the pith they prepare to fuse. At this level the two centripetal 
xylem groups open out towards each other (see the right-hand bundle in 
Fig. 9), the xylem-parenchyma of the two strands becoming continuous 
across the medullary ray between them. At a slightly lower level the two 
centripetal arcs of xylem unite, so that the joint strand now has a single 
primary xylem-mass. The photograph in PL LXXXVIII, Fig. 10, shows 
a double strand in the act of fusion ; the two centripetal groups of xylem 
have just become continuous. The drawing (PL XC, Fig. 23) shows a 
similar strand lower down in its course, where union is complete. That 
this is the actual course of the changes, as traced from above downwards, 
is shown on comparing the same strands in successive sections. Other 
strands around the pith have reduced remains of centripetal xylem, or the 
smallest elements are quite on the inner edge, showing that the structure 
has become endarch. 
3x2 
