568 Bower.—On Medullation in the Pteridophyta. 
forward evidence to show that in the Cyatheaceae the erect habit is secondary, 
and that the family are derived from a rhizomatous ancestry, probably of 
near kinship with the creeping Gleicheniaceae. They appear to have 
become erect subsequently to the formation of a pith by intrusion through 
the leaf-gaps. The same is probably the case for the Dicksonieae and for 
Plagiogyria , and my own comparative studies already indicate that a like 
probability will emerge in the case of other upright Leptosporangiate Ferns. 
The point is that in such cases the extrastelar tissue had already penetrated 
through the foliar gaps into the stele of the rhizomatous ancestry before the 
erect habit was assumed, and that that type of medullation, once initiated, 
was perpetuated with minor modifications when the race later assumed the 
upright habit. It is this conservatism more than anything else which has 
concealed the factors that have been at work. Examining Ferns at large, 
without any clear phyletic ideas regarding them, we find the pith extra¬ 
stelar in many upright as well as creeping forms. This apparent incon¬ 
sistency has obscured the relation which seems to have subsisted in the 
first instance between the prostrate position and the initiation of an extra¬ 
stelar pith. 
Two leading factors have been mentioned in the preceding paragraph 
as influencing the medullation of Vascular Plants, viz. the position of the 
axis at the time when the medullation was initiated, and the disturbing 
influence of the insertion of the appendages. Other factors not yet 
recognized may also have had their effect. Pending their discovery, it may 
now be inquired what is the balance of influence of these two factors in 
cases where both may have been effective, and whether that balance is 
constant or variable. It is in such plants as the Osmundaceae and Ophio- 
glossaceae, where the two factors are, in a sense, pitted against one another, 
that the striking of the balance may be witnessed in its structural results. 
For in them the axis is upright, which might be expected to encourage the 
formation of an intrastelar pith, while the relatively large leaves would 
encourage the intrusion of foliar pockets, leading to an extrastelar pith. 
A further circumstance, which evidently had its effect (and may indeed 
be regarded as a mechanical consequence of the creeping or upright habit of 
the shoot), is the relative size of the axis compared with the area of the leaf 
insertion. Megaphylly is merely a term indicating vaguely a large size of 
the leaf. But the proportion of leaf to axis may vary even in types accepted 
as megaphyllous. And this fact is also illustrated by a comparison of the 
Osmundaceae and Ophioglossaceae. In the stratigraphically older forms of 
the Osmundaceae, which had a very bulky axis, but no leaf-pockets (e.g. 
Thamnopteris ),the large xylem-core already shows differentiation at its centre 
in the direction of a pith formation. Leaf-pockets appear in the later forms 
of the phylum (Osmundites Kolbei ), but being small they never seem to 
have influenced the general structure so greatly as did the intrastelar pith, 
