APEX 0 E THE BOOT IN OSMUNDA AND TODEA. 95 
to the radial wall (x). Next may be taken the cell of the form 
shown in fig. 20. This is evidently a transition to the form of 
the truncated pyramid, the wall adjoining Segment I being 
deflected, so as to form at its lower part an oblique ending to 
the whole cell. Lastly, those examples may be cited in which 
the initial cells have their lower ends (the apices of the pyra- 
mids) decidedly truncated, as in figs. 17, 18, 21, 22. These 
figures, drawn from longitudinal sections, show a form of the 
initial cells and arrangement of the segments which conforms 
closely to the Marattiaceous type. Thus, as regards the form 
of the initial cells and the arrangement of the segments, 
Osmund a provides various intermediate stages between the 
two types of construction above described, and the transition 
may be connected, as above pointed out, with a lowering of 
the centre of construction. Intermediate forms, though 
with less gradual transitions, have been observed also in Todea 
barbara, where both pointed and truncated initial cells have 
been found. 
In respect of the arrangement of the initial cells (where 
more than one is present), as seen in the transverse section, 
Osmund a again shows intermediate characters between the 
two types. From the three-sided cell, seen in fig. 1, we may 
pass on to those arrangements with three initial cells seen in 
figs. 6 — 8. As I have pointed out above, these correspond 
closely to the arrangement seen in transverse sections of roots 
of the first type of construction at a point immediately below 
the apical cell. I would not suggest that the transition from the 
meristem with a single three-sided pyramidal apical cell to this 
with three initial cells could occur in the individual root, and 
I have no evidence that it does ; but, regarding the matter 
from a phylogenetic point of view, we may well conclude that 
in these examples we see the result of an upward continuation 
of the principal walls in radial planes, so as to divide the apical 
cell into three parts ; in fact, this may be recognised as an in- 
stance of filling up the gap in construction, accompanied, it 
must be remembered, by a depression of the centre of construc- 
tion. It is hardly necessary to state that this meristematic 
