OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
291 
numerous straight lines of transverse cell-walls to traverse the section horizontally from 
within outwards, as shown in Plate XLIII. fig. 17, i. There is no doubt that the walls of 
the more tubular of these elongated cells are thickened by internal depositions of lignine, 
and that they thus assume the character of bast-tissues. I have already described thin- 
walled cells arranged in regular rows which, in outward form, closely resemble those of 
Plate XLIII. fig. 17, but occurring in the primary and secondary medullary rays of 
Calamites. The tissue is a very peculiar one. I have not succeeded in discovering any 
structure absolutely identical with it elsewhere than amongst these Carboniferous plants. 
I have already referred, both in my preceding memoir (Part II.) and in the present one, 
to the fact that transverse sections of this prosenchymatous layer of the hark exhibit the 
cells arranged in regular radiating lines proceeding from within outward, as in the wood 
of the Conifer®. On seeing such sections, it is difficult to resist the impression that we 
are looking at true vascular tissues. 
The subepidermal layer differs in no material respect from that of the young twigs, 
being composed of ordinary parenchyma. The same remark applies to the structure 
of the persistent petioles, except that in transverse sections of the latter we find the 
position of the central vascular bundle very distinctly marked, as in the scars of the 
ordinary Lepidodendra. It will he remembered that this was not the case with the 
leaflets of the smallest twigs. Plate XLIII. fig. 18 represents part of a tangential section 
of a cluster of these petioles made close to the subepidermal layer of bark. In their 
disposition and general aspect they remind us vividly of a similar section of Corda’s 
Lomatophloios crassicaule , figured by him in his ‘Flora der Vorwelt ’ *. 
Having thus completed our review of the ordinary structure of these stems, I would 
next direct attention to some peculiarities connected with their growth. 
In preparing my sections, on one or two occasions I met with small, detached, medul- 
lary cylinders corresponding in all respects with those of the young twigs, only instead 
of being perfect rings of vessels, they were interrupted on one side, giving the transverse 
section of each the form of a horseshoe. I was long before I succeeded in discovering 
what this meant. It was obviously a medullary cylinder, and I at length obtained 
specimens which explained its nature. When one of the stems is about to dichotomize, 
the central vascular cylinder first becomes elongated laterally in the plane of the 
approaching bifurcation ; it then splits into two halves, each of which is, of course, open 
at its inner side. Plate XLIII. fig. 19 represents the centre of one of these specimens, 
belonging to a twig of about the same size as Plate XLI. fig. 1. What takes place 
subsequently is uncertain; hut there is reason to believe that the opening thus made 
into the interior of the medullary cylinder, bringing the medullary and cortical tissues 
into direct contact, never closes through any growth of new medullar// vessels. I am 
confirmed in this opinion by the fine section shown in Plate XLIII. fig. 20, which 
reveals similar conditions, only in this example the plant has attained to the Diploxylon 
stage of growth, having developed an ample exogenous cylinder externally to the medul- 
* Taf. 1. fig. 1. 
2 R 
MDCCCLXXII. 
