OF THE FOSSIL PLANTS OF THE COAL-MEASUEES. 
219 
been prosenchymatous in its general character, but of a very corky form. This is not 
only shown by the large amount of parenchyma which enters into it, but also in its 
extreme liability to compression. In one of my specimens this compression from without 
has given the transverse section an appearance of numerous concentric bands arranged 
parallel with the surface of the ligneous zone ; an appearance which puzzled me the 
more, since this was the first example in which I found the entire bark. Other examples 
subsequently coming into my hand threw light upon the perplexing arrangement. 
Numerous little apertures exist in this bark, through which, I doubt not, the vascular 
bundles passed to reach the rootlets. 
The portion of the bark which is most frequently preserved is the epidermal layer, 
the structure of which is interesting because of its relation to that of the rootlets long 
since described by Mr. Binney. Plate XXX. fig. 51 represents a transverse section of 
this epiderm with the bases of three of the rootlets implanted in it, whilst Plate XXXI. 
fig. 52 represents the same tissue yet further enlarged, from another specimen lent to 
me by Mr. Boyd Dawkins. As will be seen from the latter figure, the epiderm con- 
sists of a very regular form of thick-walled parenchyma (/), the cells of which become 
very much smaller and more dense as they approach the outer surface. Internally 
this parenchymatous layer is continuous with a more delicate one (/d) of the same 
character but with thinner cell-walls, and the cells of which soon become vertically 
elongated and arrange themselves in straight lines radiating inwards, as in Sig ill aria. 
This is doubtless the radiating cylinder seen by Mr. Binney in Mr. Russell’s specimens 
previously referred to, only instead of being arranged in wedge-shaped masses, as repre- 
sented by Mr. Binney, a result of the imperfection of his specimen, it is a perfectly 
continuous layer, and doubtless represents the exterior of the bast-layer of the Lepido* 
dendroid type. The thickness of this outer bark is unequal, in consequence of the depres- 
sions (Plate XXXI. fig. 53, p) that receive the proximal extremities of the rootlets, 
whose bottle-like bases, when perfect, as in fig. 51, o', are implanted in concave depres- 
sions of unequal depths, displacing both parenchyma and prosenchyma. The external 
cellular cylinder of the rootlet (o) is merely an extension of the thick-walled outer epiderm 
k, and is not in any way articulated to the prosenchymatous tissue. Within this is a space 
(Plate XXX. fig. 51, o', & Plate XXXI. 52, o') in which I have never seen the tissue 
preserved in rootlets which could be proved to be Stigmarian. Myriads of rootlets exist 
in the calcareous nodules from which our Lancashire specimens are obtained, of which 
all the cellular structures are preserved from their central vascular bundle to their peri- 
phery ; but these, I am convinced, belong to other plants than that under consideration. 
In the centre of the vacant space we have the vascular bundle (n), which, though always 
intersected in some part of the section, can rarely be traced far in one plane because of 
the flexures of the rootlets. The central axis of about twenty barred vessels is always 
surrounded by a thin cylinder of delicate cellular tissue. Wherever we see these vascular 
bundles in the bark between the woody cylinder of Stigmarise and the epiderm, we inva 
riably find this cellular ring surrounding the vessels, as at fig. 52, n' ; it is continuous with 
