3 1 8 Bower . — The comparative examination of the 
the series of Filicineae the following conclusions may be 
drawn : — 
(1) That the apices of roots of the leptosporangiate Ferns 
are of comparatively small bulk, while those of the Osmunda- 
ceae are more bulky, and those of the Marattiaceae still 
more so. 
(2) That though the type of structure of the apex of the 
root in the leptosporangiate Ferns is stereotyped and regular, 
with one tetrahedral initial cell, in the Osmundaceae this 
regularity is lost, there being often three or even four initials. 
(3) That the initial cells of Osmundaceae and Marattiaceae 
are narrower and deeper in proportion than in the lepto- 
sporangiate Ferns, and are often not pointed but rectangular 
at the base : this implies a lower position of the centre of 
construction. 
(4) That in respect of the structure of the apex of the root 
the leptosporangiate Ferns, Osmundaceae, and Marattiaceae 
constitute a natural series, of which the first named are the 
simplest, and the last the most complex, while the Osmun- 
daceae occupy an intermediate position \ 
Apices of Stems. 
As regards the details of structure of the apex of the stem 
in leptosporangiate Ferns there has been some discrepancy of 
observation ; but there is at least coincidence on this point, 
that there is present a single initial, or apical cell, the open 
question having been as to its form. 
Hofmeister, in his ‘ Higher Cryptogamia,’ 1 2 described a two- 
sided, wedge-shaped, apical cell for Niphobolus chinensis and 
N. rupestris^ Polypodium aureum and P. punctatum , and for 
Platycerium alcicorne. He figures 3 two apices of Polypodium 
vulgare, which show that the actual form of the cell is variable, 
and an apex of Pteris aquilina in which it is of the form of 
1 For further details and figures relating to the roots of the Osmundaceae, 
reference may be made to my paper on this subject in Q. J. M. S. 1885. 
• 2 Ray. Soc. 1862, p. 248. 3 1 . c., Plate XXXIV, Figs. 5, 6 ; text, p. 248. 
