5 66 
Bower.—On Medullation in the Pteridophyta. 
apparently to encroachment of the cortex on the stele,, projects convexly 
into the pith, and constitutes that internal endodermis recognized by Van 
Tieghem and Poirault. Thus the medullation in Botrychium corresponds 
to the theoretical anticipation, being both intrastelar and extra-stelar in 
origin. It arises from three distinct sources : (a) intraxylic parenchyma, 
(b) conjunctive parenchyma (both of which are intrastelar), and (c) intrusive 
parenchyma of the foliar pockets (extrastelar). 
The details as regards the origin of the pith in Helminthostachys are not 
yet to hand. It should afford an interesting example, being megaphyllous, 
but having a vertical position of its axis only in the earlier state of the 
individual. In its older stages the rhizome becomes prostrate, a condition 
which, on our hypothesis, should lead to the formation of large pockets in 
a plant with leaves of so large a size. 
The Osmundaceae and Ophioglossaceae are thus seen to accord 
structurally, though in varying degree, with the theoretical anticipation. 
Both of them show an intrastelar pith in relation to their primitively 
vertical axis, and leaf-pockets may also be formed of varying size in 
relation to their megaphyllous character. The converse must now be 
examined. We shall inquire whether in any microphyllous shoots of 
rhizomatous nature there is evidence of a pith of extrastelar origin. 
It may be said at once that the large majority of creeping micro¬ 
phyllous forms do not illustrate the anticipated structure. But this might 
be expected, for, judging from other experience, it is only in cases where the 
axis had already the creeping habit at the time when the medullation was 
phyletically initiated that the effect would be likely to appear. In cases 
where an intrastelar medullation preceded the adoption of the creeping 
habit of the axis, the course of medullation primarily initiated appears to 
have been continued, even though the creeping habit might subsequently 
be assumed. This may be illustrated by the Equisetales. In them the 
embryogeny is vertical and the primary shoot is upright. We may pre¬ 
sume that an intrastelar pith was phyletically initiated in the primary 
axis, as it is structurally seen in the primary axis of the embryo to-day. 
The creeping habit being only assumed subsequently by branches of a higher 
order, these maintained, with the conservatism already recognized, their 
primitive type of medullation, that is the intrastelar. But in other cases 
medullation may be entirely absent, as in the creeping axes of the more 
advanced types of Lycopodium. This is also the case for the straggling 
species of Selaginella. But within that genus there is the well-known case 
of 5 . laevigata , var. Lyallii, which illustrates the point in question, though 
with details which plainly indicate homoplasy as compared with the Filicales. 
Considering the constancy of the vertical embryogeny in the microphyllous 
forms, and the fact that the upright radial axis is their primitive type, the 
rarity of the reaction in them cannot be held as a valid objection. 
