220 
PROFESSOR W. C. WILLIAMSON ON THE ORGANIZATION 
the innermost bark and with the cells of the primary medullary ray through which the 
bundle emerged. It appears at fig. 51, n", where the bundle is entering the rootlet, 
and it equally reappears if we intersect the latter at its extreme tip. The epidermal 
tissues immediately subjacent to each rootlet are always dense, consisting of small paren- 
chymatous cells, which show a tendency to arrange themselves in radial lines (fig. 51,#). 
When the vascular bundles are intersected between the woody zone and the epidermal 
layer, their outline is usually that of a triangle with convex sides, as seen at fig. 52, n'. 
Plate XXXI. fig. 53 is a restored diagram, exhibiting what I believe to have been the 
structure and form of Stigmarici in its integrity. The several parts of this diagram 
will be easily identified with the details of the preceding description, because the same 
letters have been employed in both to represent corresponding tissues. 
I have frequently found in the Lower Coal-measures at Oldham fragments of a very 
curious bark that long perplexed me, because I was unable to discover it in association 
with any woody axis. In one example it appeared partially to surround a Dijgloxylon ; 
but as the portion on one side of the ligneous cylinder appeared to be in a reversed 
position to that on the other side, I both hesitated to connect them and was unable to 
decide which was the inner and which the outer surface of the bark. On cutting verti- 
cally through a part of the specimen represented in Plate XXVIII. fig. 33, and which 
is essentially a Diploxylon, I again found the anomalous bark associated with this type 
of ligneous cylinder, and under such circumstances as left me little room to doubt that 
it belonged to the same plant. The general appearance of these fragments, when cut 
transversely, is shown in Plate XXXI. fig. 54. Spaces of a lenticular form (Ji) radiate 
towards the periphery*; these are filled with cells, whose parallel sides cross the short 
axis of each space. The long axis of each cell is often as much as -0075 in length and 
•005 in the opposite direction. At one extremity these lenticular masses show a disposi- 
tion to converge at irregular projecting points ; at the other they gradually pass into 
regularly disposed lines of narrow prosenchyma (V). The lenticular masses of cells exhibit 
nearly the same appearance in the radial section that they do in the transverse one ; but 
a tangential section exhibits the cells in fasciculi (fig. 5G), where small clusters of them 
are seen enclosed within dark and strongly defined areas of a doubtful nature. In the 
regularly arranged prosenchymatous portions the tangential and radial sections are very 
different from the transverse ones, and, indeed, they vary in different specimens. Fig. 55 
is a radial section of a strongly marked type, in which the prosenchymatous cells appear 
to be of nearly equal lengths and with square ends, so much so, indeed, as to resemble 
some varieties of mural tissue common amongst the Calamites ; but in the tangential 
section (fig. 56) we see that they are prosenchymatous, but of a very sharply defined 
geometric type, with straight walls and very distinct angles. But in many other speci- 
* I have just mot with an example of Stigmarian bark, with its characteristic rootlet attached, in which the 
structure represented in fig. 54, h, occurs, intermediate in position between the outermost parenchyma and the 
more internal radiating prosenchymatous layer, blending the two : whether it is merely a variety, or belongs to 
the Stigmarian root of some distinct species of Lepidodendroid plant, has yet to he ascertained. — Aug. 6th, 1872. 
