OF THE FOSSIL PLANTS OF THE COAL-MEASUKES. 
301 
which it becomes a true Diploxylon , this plant possesses a well-defined cellular pith. Its 
central axis is not composed of bundles of vessels, but of vertical piles of true cells. As 
soon as the outer ligneous cylinder makes its appearance, true medullary rays also pre- 
sent themselves. Simultaneous with the formation of these true medullary rays is that 
of my primary rays, or cellular spaces through which the vascular foliar bundles pass 
outwards through the ligneous zone, and which differ from the others in no respect, 
either of structure or of origin, save in the circumstances that they are larger and that 
the foliar bundles are lodged within them. The specimen from which Plate XLXXI. 
fig. 22 is taken clearly proves this. The vascular bundles proceeding from the interior 
of the ligneous zone to the leaves, when once formed, evidently became permanent struc- 
tures, undergoing neither increase nor diminution of number ; but as the diameter of 
the stem steadily increased, these bundles obviously became lengthened, by some process 
as yet unascertained, so as to accommodate themselves to the altered dimensions of the 
tree, especially of its bark. It follows that when the pseudo-cambium-layer commenced 
its work of producing new vessels, which were added exogenously to the exterior of the 
preexisting vascular cylinder, it was penetrated by these leaf-bundles, and the arrange- 
ment of the newly formed vessels was modified by their preexistence. On studying these 
tissues in the original of Plate XLIXX. fig. 22, where the arrangements of the new 
growths are very distinct, no essential difference can be observed between those inter- 
vascular areas filled with cells through which a vascular bundle passes, and which are 
destined to become what X have designated primary medullary rays, and those which 
ultimately assume smaller dimensions and become secondary ones. It appears to me 
that as the new, longitudinally arranged vessels of the young growth increased in size, 
the intermediate cellular tissue seen in Plate XLXXX. fig. 22 was gradually absorbed to 
make room for them. In the secondary medullary rays this absorption was carried so 
far, in consequence of the pressure occasioned by the steady growth of the vessels, that 
nearly all the cells disappeared ; whereas in the primary rays, where a vascular foliar 
bundle interposed between two adjacent enlarging vessels, the bundle resisted their 
pressure, protecting the cells immediately above and below it from its effects. Hence 
a lenticular space was left permanently occupied by unabsorbed cells ; but at the upper 
and lower angles of this space it contracts to the dimensions of the true secondary 
medullary rays. If this explanation is correct, it establishes my conclusion that these 
large spaces, seen in Plate XLXX. figs. 12 & 13, in, are but modified medullary rays, and 
that they are so modified, not for the purpose of transmitting the vascular foliar bundles, 
but as an effect of their presence, which is a very different thing. 
In my last memoir X called attention to the fact that the foliar bundles originated 
from the line of junction between the vascular medullary cylinder and the ligneous zone A 
* I have to correct an error into which I fell on this point in the text of my previous memoir. I had clearly 
ascertained that the foliar hundles sprang from small vessels occupying the plane where the outer surface of the 
vascular medullary cylinder and the inner one of the ligneous zone were in contact, and I came to the con- 
clusion that they belonged to the latter rather than to the former ; but I now see that this was a mistake. I 
2 s 2 
