OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
387 
zation. In Plate XXY. figs. 19, z, & 20, z, we have the two bundles belonging to 
another of the clusters of Plate XXII. fig. 1, and which are enclosed within the inner- 
most bark-layer: their relative positions in the bark were similar to those which they 
occupy upon the Plate. On the outer or cortical side of each Ave find several diverging 
lines or laminae of vessels (V), of which those nearest the space separating the tAvo bundles 
are the longer ones, whilst the others become successively shorter as we diverge from that 
space. We have here unmistakable groAvths superadded to the exterior margins of the 
tAvo bundles, similar, in all essential respect, to those added by the pseudocambium to 
the exterior of the ligneous zone, and apparently through the action of the same pseudo- 
cambial cells. Though I do not by any means find these conditions of the cortical bundles 
repeated in all the specimens of Dictyoxylon , they are frequently met Avith. Plate XXV. 
fig. 21 exhibits a yet more remarkable modification of these bundles that I have occa- 
sionally observed. In this instance the original bundle of non-radiating vessels has become 
•firmly incorporated Avith the periphery of the ligneous zone (e), and the exogenous additions 
have been made to its exterior in such a manner as to form a Avedge-shaped buttress (z'), 
running vertically doAvn the ligneous cylinder. I have dissected several of these buttresses, 
and find that when radial sections of the ligneous zone pass directly through them they 
exhibit all the ordinary appearances of, and are scarcely distinguishable from, that zone ; 
indeed, as seen in such sections, they only appear to be outward prolongations of it. In 
some of the examples, as is the case with the one figured, the central non-radiating bundle 
of vessels has its details less clear than in others. Where these are distinct, I think they 
shoAV that the bundle is in the same state as that represented in Plate XXIII. fig 7, z, 
viz. in its undivided condition. This explains why, in these wedge-shaped groAvths, Ave 
never find the twin bundle. The exogenous groAvth, which iioav binds the vessels 
together, has commenced before the separation of the cluster into tAvo parts Avas effected. 
These details are trivial, but they help to throAV light upon the nature of the objects 
in question. 
Another remarkable feature of these organisms yet remains to be examined. So far 
as outward appearances are concerned, Dictyoxylon might be declared to have been a 
simple, unbranched stem. In no one instance have I found an example which exhibits 
any obAnous division or bifurcation of the stem, such as is common amongst the 
Lejndodendra. Nevertheless several examples have come into my hands in which 
branches of some kind have existed. The first trace Avhich I met with of this character 
Avas in a tangential section of the bark made by Mr. Butterworth, and represented in 
Plate XXY. fig. 14. One of the cellular areolte of the prosen chymatous layer of the 
bark has evidently been modified and enlarged to allow the branch (y) to escape through 
it. The cellular parenchyma of the middle bark has been condensed into an oval ring, 
through Avhich the bundle has emerged. This ring is seen further enlarged in Plate XXYI. 
fig. 22, h ; within it is a space (g) which Avas, I have no doubt, originally occupied by the 
delicate parenchyma of the innermost bark, prolonged as an investing layer from the 
central cylinder of that tissue, but the cells of Avhich have now disappeared. The centre 
3 f 2 
