meris terns of Ferns as a Phylogenetic Study . 333 
divided by alternate walls so as to define at once a two-sided, 
wedge-shaped apical cell, one edge of which is directed 
towards the apex of the stem ; this apical structure is main- 
tained for a considerable time : thus in its segmentation and 
the general characters of the apex there is nothing specially 
worthy of note in Amphicosmia as differing from other 
leptosporangiate Ferns. 
For purposes of comparison a few observations have been 
made on Lygodium scandens , and Aneimia hirta , with the 
result that in both the apex of the leaf was seen to be 
occupied by a two-sided apical cell, from which segments, 
and a marginal series are derived as usual : in fact, as regards 
the early development of the leaf they show no advance on 
the Polypodiaceae. Thus it appears that in the lepto- 
sporangiate Ferns there is great uniformity in the segmenta- 
tion of the young leaf, the only exceptions to the two-sided 
apical cell hitherto observed among them being those ab- 
normal cases noted by Klein : the type of construction of the 
apex is that characteristic of flattened growths. 
I have already drawn attention to the fact 1 that in the 
Osmundaceae there is a departure from the more simple type 
of structure of the apex of the leaf above referred to as 
characteristic of the Hymenophyllaceae, Polypodiaceae, 
Hydropterideae, Cyatheaceae, and Schizaeaceae : it has been 
pointed out that the three-sided type of apical cell which 
Klein showed to be occasionally found in very young leaves 
of the Polypodiaceae, is the type for Osmunda and Todea , 
though the succession and form of the segments in these 
plants is much more regular than that shown in Klein’s 
figures. I have recently made fresh observations on Osmunda 
regalis , with the following results. The position of the apical 
cell which is defined by the first divisions in the young leaf 
(Fig. 23, /), is such that one of the sides faces the apex of the 
stem, while the opposite angle is directed away from it ; this 
orientation seems to be constant for O. regalis and cinna- 
momea , and for Todea stiperba . The sub-division of the 
1 Phil. Trans. 1884, Part II, p. 575. 
