448 Scott .— The Structure of Mesoxylon multirame. 
verse sections is shown in PL XIV, Fig, 23. It is a small strand, only about 
0*7 mm. in diameter; the phloem and cambium are well preserved ; the 
wood is very parenchymatous, but tracheides extend to the centre of the 
stele. It is quite possible that this was a rudimentary or abortive stele, for 
those shown in the longitudinal sections are considerably larger, and have 
a definite pith. 
The tangential sections show the changes in the axillary stele very 
clearly—one of these steles can be recognized in the phloem (section 2773); 
but it is in the wood that the preservation is best. In the outer zone of the 
wood (section 2774) an axillary stele (about 1 mm. x 1*4 mm. in diameter) 
is cut almost transversely; the pith is elliptical, with the major axis hori¬ 
zontal. The wood of the branch, mostly secondary, is chiefly developed on 
the lower side, when it passes over directly into the longitudinal strands of 
the stem-wood : the same is the case laterally, but on the upper side the 
transition is much more abrupt, the tracheides here running horizontally, 
often with upward bends. The subtending trace is not visible, though the 
wood is shown for about 4 mm. below the stele. 
In the next inner section (2775 ; PI. XII, Fig. 12) an interesting change 
appears ; the axillary steel is deeply bilobed, a wedge-shaped incision, filled 
with stem-wood, extending almost half-way through it from the upper side. 
The lower and lateral connexions with the stem-wood are essentially the 
same as before ; on the upper side the stem-tracheides converge, to enter 
the incision. About 4 mm. below the stele the two strands of the subtending 
leaf-trace are present, though partly cut away ; they are not included in the 
photograph. 
The next section (2776) passes in the middle part, through the persistent 
zone of the pith, showing the wood, in oblique section, on either side. In 
the wood, to the left, an axillary stele (not the same as in the preceding 
sections) is seen, quite in the inner part of its course (PI. XII, Fig. 13). 
The stele is here completely severed into two, the longitudinal stem-wood 
extending uninterruptedly between the two halves. The width of the gap, 
narrowest in the middle, varies from about 0*4 to 0*25 mm. The half-steles 
here consist largely of pith. The wood is mostly limited to the upper and 
outer sides, and is everywhere in continuity with that of the main stem, the 
lateral connexions being the most extensive. 
Thus the occurrence of a division of the axillary stele into two, where 
it approaches the primary wood of the stem, is demonstrated. The strand 
of stem-wood separating the two halves of the divided stele is no doubt 
continuous with the median strand observed in the transverse sections. 
The division of the axillary stele is not, however, constant, as is shown 
by the transverse sections; of five clear cases observed in the series, the 
stele was divided in three and undivided in two. One of the latter is 
represented in PI. XII, P"ig. 14. The stele abuts on a pith-bay and is clearly 
