OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
205 
bundle, Invested by at least two distinct zones, the outermost of wliich bore leaves before 
any traces of a medulla could be discovered. Thus we discover at the outset a 
difference between the history of the medulla of a Lepidodendron and that of an Elm 
or an Elder tree. Tracing yet further the development of the Lepidodendroid type 
we find that in it the medulla first appears as one or two individual cells formed in 
the centre of a bundle of tracheids or vessels ; once existing, however produced, these 
cells multiply rapidly by the ordinary meristemic process of fission. So far as my 
specimens throw light upon the latter process, it exhibits some peculiarities. The 
meristemic internal subdivision of these cells was not going on continuously, but 
interruptedly; at certain periods the whole of the fully-developed cells of the medulla 
simultaneously underwent such a division. Fig. 2(3a represents a cluster of cells from 
the medulla of a branch of Lepidodendron ILarcourtii undergoing this meristemic 
multiplication. Some of these matured cells are subdividing into four or five of the 
younger generation. At this stage the latter are all thin-walled, small in size, and 
irregular in form ; but all these conditions gradually become changed. The walls 
become more strongly defined ; the area of each cell enlarges from two to two-and-a- 
half times its original size ; and its unsymmetrical form develops into that of the regular 
pentagon or hexagon seen in the primary, or mother, cells of fig. 26. 
But the eflPect of these changes is not limited to that produced upon the medulla. 
They reach the vascular bundle within which the increa.se in the number of the medullary 
cells is taking place. The first result of the internal tension occasioned by these 
cellular expansions is to develop the solid mass of vessels into an annular ring, h, of 
increasing diameter. Fig. 8 becomes progressively converted into what is seen in 
figs. 9, 10, 11, and 12. But this vascular ring, h, not only increases in diameter, but 
the vessels composing it increase in number, and change their relative individual 
positions as they do so. At later periods this process of meristemic division and 
subsequent expansion of all the new cells appears to have been repeated from time to 
time, until the medulla and its surrounding medullary vascular ring attain to their 
ultimate magnitudes—a condition which was probably coincident with the first 
appearance of the more'external exogenous zone. I was at one time inclined to think 
that some of the young medullary cells assumed a procambial form, and were converted 
into new vessels ; and even now I am not sure that this is not so in some instances. 
But it appears to me now that, in such examples as figs. 7-14, the new vessels must be 
produced on the cortical side of the medullary vascular cylinder— i.e., centrifugally 
rather than centripetally. However this may be, the enlargement of this cylinder is 
evidently effected mainly, if not wholly, through the internal tension occasioned by 
the subsequent multiplications and expansions of the medullary cells—a condition 
that has no existence amongst the exogenously-grown trees now living. 
But a partially parallel state of things does exist amongst some living plants. 
In his learned work on the ‘ Comparative Anatomy of the Vegetative Organs of the 
Phanerogams and Ferns,’ the late Dr. A. de Bary observes: “In many Ferns the 
