304 
PEOFESSOE W. C. WILLIAMSON ON TILE OEGANIZATION 
by successive centripetal growths, these vessels have so increased in number as to cause 
the two bundles to meet in the middle of the procambial cylinder, through the longer 
axis of the transverse section of which they now form a continuous line. The central 
vessels of this linear series are now the largest. A further distinction appears in the 
circumstance that the central vessels are generally more perfectly scalariform, whilst the 
outer vessels are spiral ones. Whilst these changes have been going on, corresponding 
ones have connected the remaining procambial cells into an interrupted ring of prosen- 
chyma with somewhat thickened walls, or, in other words, into a ring of limiting tissue, 
making the whole axis a closed bundle. The cortical layers appear to be composed of 
meristem ; that is, they do not grow from any true cambial structure, but by the division 
of the preexisting cells of all the parts, until the normal dimensions of each stem are 
reached. The central vascular bundle of the axis thus represents the Xylem, and its 
investing prosenchyma the Phloem, whilst the bark derives its existence from an inde- 
pendent source, originating in the primitive cellular tissue. In the plant quoted most 
of the vessels of the fibro-vascular bundles appear to be derived from the leaves. The 
outer vessels of these bundles are smaller in size as well as more spiral in structure 
than the inner ones. These facts have an important bearing upon the interpretation of 
our fossil forms. Nageli has argued that the fibro-vascular bundles belong to the stem 
and not to the leaves, because he finds such bundles in Psilotim, in which the leaves are 
deprived of them ; but the Psilotum is altogether so exceptional a form that it can 
scarcely outweigh the evidence afforded by the Lycopodia and Selaginellce. 
Guided by these examples, I think we can ascertain the homologies of the fossil stems 
so far as their tissues are represented in the living types. It is clear that the central 
bundle of Plate XLI. fig. 2 corresponds substantially with a young state of the central 
fibro-vascular bundle of Selaginella Martensii , only here some of the central primitive 
tissue has remained to form the basis of a future pith which has no existence in the 
living forms. In the latter we have no central axis preserved which can, hy the utmost 
stretch of the imagination, be identified with a pith ; their primary central axis of pro- 
cambium is wholly converted into or replaced by the central vessels (Xylem) and the 
investing zones of prosenchyma (Phloem). Parenchyma has no longer an existence in 
this part of the plant ; hence we must conclude that the preservation of a central portion 
of primitive parenchyma, capable of very considerable increase by cell-division, is peculiar 
to the fossil types. 
We have seen that the number of the vessels in the central vascular bundles of living 
types increases, up to a certain point, with age, and also that each foliar vascular bundle 
unites with those of the central axis, at least where the first two come in contact and for 
some distance down the stem, at the external surface of the central bundle. The fossil 
and recent forms agree in this point; but we now face a difficulty. The number of the 
vessels in such a cylinder as I have represented in Plate XLI. fig. 8 represents, doubtless, 
the aggregation of the bundles of a yet larger number of leaves than there are vessels ; 
and if each leaf of the upper part of the stem added its quota to the whole externally 
