OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
289 
shall afterwards see, from the way in which these spaces are formed, that they do not 
differ in any essential respect, except in their size and in the number of their cells, 
from such true medullary rays. The vascular bundles (m) are of course divided trans- 
versely in these tangential sections, in which they exhibit a diameter of from -005 to 
•0075. Each bundle consists of a large number (rarely less than 100) of minute barred 
vessels, varying from *0005 to -0008. The origin of these bundles amongst the minute 
vessels which abound at the point of junction of the medullary cylinder and the ligneous 
zone has already been shown. In the Diploxylon originally described by Corda (Flora 
der Vorwelt, tab. x. fig. 3, and tab. ii. fig. 1) these bundles are represented as ascending 
obliquely upwards and outwards ; but in the plant before us such is not the case ; they 
wend their way outwards through the ligneous zone, as do also the medullary rays, in a 
perfectly horizontal plane. The second of Corda’s figures also represents them as 
originating abruptly at the external surface of the medullary cylinder. Their real 
origin has been already shown in Plate XLII. fig. 10. Corda further describes his 
plant as having no medullary rays. This, as I have pointed out in my previous memoir, 
is also an error, and has arisen from the circumstances there indicated, viz. that in some 
species of Diploxylon the cells of the medullary rays are barred, which caused Corda to 
mistake them for true vessels. 
In other specimens of Diploxylon which I possess I find some variations from that just 
described, as well as some points which are more fully elucidated by them. In several 
examples the medullary cylinder is very much thicker than in others, in proportion 
to the diameter of the medulla. In some its thickness is as much as T2. One remark- 
ably fine example exhibits the true structure of the medulla ; a vertical section of the 
medulla and medullary cylinder of this specimen is given in Plate XLII. fig. 14. The 
space between the letters a a is occupied by the cells of the medulla, which are arranged 
in vertical columns with a considerable approach to regularity, when undisturbed by 
pressure or mineralization. These columns have a mean diameter of ‘005 to ‘0075. 
Generally the cells are nearly cubical, allowance being made for the frequent obliquity 
of the transverse septa, one of which sometimes inclines upwards and the other down- 
wards at the two extremities of the same cell. Fig. 15 represents a small portion 
from a transverse section of the same specimen, illustrating the relations of the ends of 
these columns of cells to the intersected vessels of the medullary cylinder. It will be 
seen that the cells ( b ) can only be distinguished by their colour and their thinner walls 
from the vessels (<?). The colour is due to the circumstance that one or both of the 
transverse cell-walls of each cell appear in the plane of the section, their carbona- 
ceous substance giving a brown colour to the section where they exist. On the other 
hand, the vessels being long tubes filled with translucent carbonate of lime, transverse 
sections of them exhibit no such colour. The walls of the vessels also are more sharply 
defined and thicker, owing to the deposit of lignine forming the transverse bars in their 
interior ; but in every other respect of size and shape the two exhibit no material differ- 
ences. It is difficult to believe that the very peculiar arrangement of the cells in vertical 
