meristems of Ferns as a Phylogenetic Study. 337 
shown In Fig. 41, In which the actual apical cell (x) is seen at 
the edge of the drawing, while from the conical cell (0) a suc- 
cession of segments has been cut off. This would support the 
view that it is a secondary initial comparable to that shown 
in Fig. 7 of the root of Osmunda. Thus in the structure of 
the apex of the leaf of Marattia fraxinea there are many 
points in common with that of the Osmundaceae, but though 
a three-sided apical cell is present in the young leaf, neither 
in its size, form, orientation, or segmentation is it so regular as 
in the latter. 
The apex of the leaf of Angiopteris evecta is more complex 
in structure than that of Marattia , and it may be stated at 
the outset that, though observations have been made on 
numerous leaves of various ages, including very young ones, 
no clear case of a single initial cell has been observed. This 
result falls in with the description given by me some years 
ago \ that ‘ the cells of the apical region may be referred, in 
some cases at least, to a group of four initial cells ; but these 
cannot always be equally well distinguished, and as seen in 
surface view from without they exceed the surrounding cells 
but little if at all in size/ I now find, on the basis of wider 
observation, that it is difficult to state the actual number of 
the initial cells, but the arrangement of the meristem seems 
in some cases to point to the number three : thus Figs. 43, 
43, which show good typical examples, have been drawn as 
accurately as possible, so as to indicate the forms of the 
cells, and their genetic relations to one another : in neither 
case is it quite clear what the actual number of the initials 
is, but comparing these with one another, with Fig. 39 of 
Marattia. and again with Figs. 3 and 4 of the root of 
Osmunda , it would appear probable that we have here further 
examples of three Initial cells (x) grouped together in the 
centre : thus it is possible, on theoretical grounds similar 
to those already explained in the case of the root, to put 
this more complicated construction of the apex in relation 
with the simpler type where one three-sided initial cell is 
1 Phil. Trans. 1884, Part II. p. 580. 
