84 
F. 0. BOWER. 
I 
of its segments, together with the nearly central position 
which are so characteristic of the roots described by Naegeli 
and Leitgeb. Taking as the first examples (figs. 11, 12) two 
of those which approach most nearly to that type, the pyra- 
midal apical cells (x) can be readily recognised in each case ; 
but a comparison with the figures of Naegeli and Leitgeb 
shows certain points of difference. In the first place there is 
a difference in the form of the cell itself; it is in Osmund a 
proportionately narrower and deeper, that is, more elongated in 
a longitudinal direction, and consequently the principal walls 
by which it is bounded laterally are less inclined to the longi- 
tudinal axis. Secondly, the arrangement of the cells surrounding 
the pyramidal cell does not show any definite regularity, and 
it is thus difficult to ascertain their genetic connections. This 
irregularity is found not only in the tissue adjoining the sides 
of the pyramidal cell, but also in those which have been derived 
from its base, and will go to form tissues of the root-cap. 
Comparing these figures with Naegeli’s, though the similarity 
is obvious in its main points, the regularity of detail so charac- 
teristic of the plants he described is absent even in these, 
which are the most regular examples of the apex of the root 
of Osmunda which I have observed. 
A further point to be noted in figs. 11 and 12, but which is 
much more prominent in figs. 13 and 14, is that in point of 
size the wedge-shaped cell is smaller in comparison with the 
adjoining cells, than is the case in other Ferns; whereas in 
NaegelFs figures the area of the apical cell exceeds that of any 
of the surrounding cells, in Osmunda it is as a rule smaller 
than they. It may be observed not unfrequently that one 
cell of a group, obviously of sister cells, takes the lead in point 
of size, the form of such a cell being usually a truncated 
pyramid, as shown in fig. ]5 (x). I think it probable that in 
such cells we may trace an ascendancy in more than mere 
size; in fact they may have, in part at least, the function of 
initial cells. This would appear to be more clearly the case in 
fig. 13. In the apex represented in fig. 14, the pyramidal cell 
is clearly seen ; here, though there is no marked ascendancy of 
