meristems of Ferns as a Phylogenetic Study. 385 
larger are less regular in their segmentation. I do not con- 
sider that this suggestion is by any means established by my 
observations on the Ferns, but the result of the examination of 
a very considerable number of meristems makes me think 
that some such connection between nutrition and meristematic 
construction is not improbable. 
In this connection two other points deserve notice : the first 
is, that where a single initial is found in the relatively bulky 
meristems of the plants higher in the series, the segments cut 
off from it increase more rapidly in size as compared with the 
apical cell itself than in the lower terms of the series. (Com- 
pare Figs 22, 23, 24, of stems, and Figs. 35, 36, 37, 38, of leaves). 
The second is, that in the course of this growth, the sub- 
divisions of the segments are far from being regular : often 
cells which I have styled ‘ secondary initials 5 make their 
appearance in the segments, and appear to undergo sub- 
divisions which show some similarity to those of the apical 
cell itself (compare Figs. 34, 36, 38, 41, the cells marked 0 ; 
also Fig. 7). If, as above suggested, the regularity of segmen- 
tation depend upon the regularity of the nutritive supply, 
these irregularities of subdivision of the segments and the 
appearance of the c secondary initials J point to such want of 
uniformity of supply as may be expected to have its effect also 
upon the initial cell itself. With regard to these ‘secondary 
initials ’ a comparison may be drawn to the meristematic 
arrangements of Psilolum as represented by Graf. Solms 1 ; 
his account of the appearance, disappearance, and renewal of 
initial cells goes far to break down too formal ideas as to the 
importance and supremacy of the apical cell itself. 
The general results of these investigations as regards phy- 
logeny are to demonstrate, more clearly than before, that the 
Filicineae constitute a natural series, illustrating an adaptive 
sequence from plants of semi-aquatic to those of a more dry 
habit ; to show that additional evidence only tends to fix the 
Osmundaceae still more firmly than before in the position of a 
link bridging over the gap between the leptosporangiate and 
1 Ann. d. Jard. Bot. d. Buitenzorg, Vol. IV, 1884, Plate XXI, Figs. 1, etc. 
