meristems of Ferns as a Phylogenetic Study . 325 
whole number which I had under observation. Accordingly 
I am not prepared to state that the structure of the apex of 
the stem always accords to the type of Schwendener, though 
in some cases it undoubtedly appears to do so : it is possible 
that Holle’s observations having been made on young plants 
(1. c. p. 2,1 6), their apices may have had a more simple 
structure with a single apical cell, such as he describes. 
The apices of several strong stems of Marattia fraxinea 
were also examined, and the results are virtually the same as 
those for Angiopteris evecta ; no single initial cell has been 
found ; the meristem is however referable, in some cases at 
least, to four or five initial cells (Fig. 28), but these are not 
exactly alike in size or in shape. Whatever the details of the 
construction may be, there is obviously no single initial cell 
of definite shape, such as is found in the lower forms. 
The study of longitudinal sections through the apex 
presents greater difficulties, as is always the case where the 
segmentation is not according to a definite scheme : it is 
essential at once to recognise that the conical form does not 
necessarily mark an initial cell, for in the first place undoubted 
initial cells may have a square base, as in the root of 
Marattiaceae or the rhizophore of Selaginella 1 ; while conical 
cells are to be found at points where their character as initial 
cells is out of the question : thus in longitudinal sections 
through the leaf of Angiopteris , conical cells of considerable 
depth are to be found at points far removed from the 
actual apex. Naturally where the apex is very wide and 
flat, as it is in these plants, it is not easy to be certain which 
sections are median, and the uncertainty is increased where 
the character of the meristem is liable to variation: to the 
best of my belief the section shown in Fig. 29 is a median 
one; the wall (x,x) corresponds to that central point above 
recognised in transverse sections. It will be noted, first, that 
all the superficial cells of this region are unusually narrow and 
deep ; thus, whatever the details of structure of the meristem, 
1 Treub, Selaginella Marteusii, Plate IV. 
