meris terns of Ferns as a Phylogenetic Study . 383 
and the suggestion appears ready to hand that the greater 
complexity of structure depends directly upon, or is at least 
closely connected with, the greater bulk. This point I have 
already discussed elsewhere as regards the root 1 , and the 
conclusion there arrived at for the root is equally applicable 
to stems, leaves, and wings of the leaf of the Ferns examined. 
That conclusion is this : — That greater size of a member 
does not entail greater complexity of meristem in members 
of the same individual plant, or of plants of the same 
species ; for instance, roots, etc. of smaller size may have 
a more complex meristem than larger roots ( Osmunda ) : 
but such a correlation may be traced to some degree in 
different species of the same genus (stems of Selagi- 
nelld ), and it is plainly to be seen on comparison of the 
corresponding parts of plants less closely related. In com- 
paring corresponding parts of plants not in the same genus 
or division of Ferns, a certain relation between bulk and 
complexity of meristematic structure is obviously seen, and 
this has been referred to so constantly throughout this paper 
that it need not be farther dwelt upon here. On the other 
hand, the comparison of the mode of development of the 
wings plainly shows that actual size does not directly domin- 
ate the mode of segmentation within narrow limits of affinity : 
the bulky leaf of Trichomanes reniforme (Fig. 51), notwith- 
standing its thickness, retains the simple mode of segmentation 
of its genus : the thin filmy leaf of Asplenium resectum (Fig. 
58), notwithstanding that it is actually much less bulky than 
the above, retains the more complicated segmentation of the 
Polypodiaceae. Both of these show, independently of actual 
size, a conservatism of their family type of segmentation, and 
though I would not be prepared, on the narrow ground of 
observation as yet before us, to state as a general principle 
that such conservatism or hereditary character is of common 
occurrence, such examples as these are sufficient to show that 
those who state that the arrangement of cell-walls has 
Quart. Journ. Micr. Sci. 1885, pp. 88, 98-99. 
