354 Bower. — The comparative examination of the 
similar to Osmunda , the wings being massive, and consisting 
of about nine layers of cells. 
Now the question is how are these massive wings pro- 
duced ? If the margins of young wings be examined ex- 
ternally, no marginal series of cells is seen ; they appear to 
be composed of a complicated net of relatively small cells, 
which are arranged more or less definitely in transverse rows 
(Fig. 59). Examining these rows individually 1 there is no 
single cell in each which can readily be fixed upon as a single 
initial for the whole row ; in fact no marginal series of cells is 
present here, such as is found in the simpler Ferns. This 
conclusion is borne out by transverse sections through the 
young wing (Fig. 60), in which no single cell is obviously 
larger than the rest, and the segmentation is not by inclined 
and alternate walls, but there is a T division, the bases of the 
individual cells being not pointed, but approximately rect- 
angular. It may frequently be seen, as in Fig. 60, that a 
clearly defined wall (x, x) divides the wing into two almost 
equal halves : in such a case the idea of a single marginal cell 
is clearly out of the question, and it appears probable that at 
least two initials are present in each transverse layer ; if this 
were regularly the case the margin of the wing would be 
occupied by two longitudinal rows of initials placed side by 
side, which would take up the function performed in the 
leptosporangiate Ferns by a single row. It is clear enough 
that the origin of the tissues in Fig. 60 might be traced to the 
two cells marked (x, x) : but though such a definite arrange- 
ment of the meristem may be found in some sections, it does 
not appear to be constant, nor does the surface view from 
without (Fig. 59) clearly point to such a conclusion. Accord- 
ingly, without actually defining the exact order of segmenta- 
tion of the cells at the margin of the wing of Todea barbara , 
or even assuming that a fixed and definite rule exists, we may 
conclude (1) that no single series of initial cells is present ; 
(2) that the superficial cells of the wing are of oblong form, 
1 It can hardly be doubted that these rows of cells correspond to one marginal 
cell together with its segments, in the Polypodiaceous type. 
