OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
305 
to all previously formed ones, it is difficult to understand why we do not find the lower 
ones traversing tangential sections of the medullary cylinder, as we do in corresponding 
sections of the exogenous zone (Plate XLII. fig. 12). I think there can be no doubt 
that the large inner vessels of the vascular medullary cylinder belong directly or indirectly 
to leaves located at points of the stem inferior to those smaller ones belonging to the 
periphery of the circle. Yet in radial sections we witness the anomalous arrangement 
represented in Plate XLII. fig. 12, where the leaf-bundle (m) joins the cylinder (c) at a 
point external to the larger vessels of c, but which latter is connected with leaves higher 
up the stem than that supplied by the vessels m. At m! we still find the foliar bundle 
retaining its position external to the cylinder. I can only conclude that as they descend 
into the stem the vessels of each foliar bundle pass inwards, but do so obliquely and slowly, 
thus preventing their altered direction from being conspicuous in tangential sections of 
the cylinder. This peculiar difference in the arrangement of the upper and lower extre- 
mities of the foliar vessels may explain the sinuous course which those of the medullary 
cylinder pursue. They never exhibit the mutual parallelism seen in those of the ligneous 
zone, but twist about, so that they rarely preserve such parallelism, for any distance, either 
with each other or with the plane of the section*. 
But supposing this peculiarity in their arrangement to be explained by what I have 
stated above, there yet remains another problem to be solved. We have seen that, 
in the first instance, these medullary vessels are few in number, and exhibit scarcely any 
central medullary area, whilst at later periods of growth opposite conditions prevail in 
both these respects. The pith becomes larger as the branch increases in size, involving a 
corresponding enlargement of the vascular ring composing its peripheral boundary. This 
could only be accomplished either through the pressure of the growing pith causing dis- 
placement and rearrangement of the surrounding vessels, or by producing absorption of 
the inner ones, the loss of which must, in that case, have been antagonized by a constant 
addition of new ones at the periphery. But after what I have seen of the displacement 
of older vessels through the pressure occasioned by the growth of newer ones, I have no 
hesitation in adopting the former of these explanations ; the more so, since I have 
not in any instance seen such ragged irregularity in the vessels in contact with the 
medulla as continuous absorption would produce. Plate XLII. fig. 15 demonstrates that 
the real condition of things is precisely the reverse of this, the cells of the pith and the 
vessels of the cylinder adapting themselves to one another with geometric regularity. 
After the development of the foliar bundles and their aggregate product the vascular 
medullary cylinder went on for some time, an altogether new set of vessels began to be 
formed laying the foundations of my exogenous growths. These differ from those of 
the cylinder in almost every respect, whether of origin, structure, or function. 1st, as to 
origin : they are not, directly or indirectly, associated with the leaves ; hence the foliar 
bundles have had nothing to do with their production. They have been formed in 
unequal concentric rings, in immediate contact with the inner layer of the bark. It 
* These peculiar arrangements are represented in the diagram, Plate NLY. fig. 36. 
