OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
293 
in my cabinet, each of which has a diameter of fully an inch and a half, I found each 
of the numerous root-bundles consisting of an average of about forty vessels. These 
facts show that as the diameter of the exogenously developed root increases with 
age, this growth is accompanied by a corresponding increase both in the dimensions 
of each rootlet and in the number and magnitude of the vessels constituting its vascular 
bundle. 
Fig. 17 represents the inner cylinder, c, with the vascular bundle, a, of a rootlet of 
the same size as fig. 16, but enlarged 270 diameters. The vascular bundle, a, consists 
of but three small eccentrically placed vessels. The space b, occupied in the living 
Selaginellct by the large, delicate cells regarded by M. Van Tieghem as liber-cells, 
(fig. 13, b), is here empty. We rarely find a specimen in which this is not the case ; 
even in transverse sections of recent Lycopodiaceous roots these cells frequently become 
ruptured, and their almost constant absence from the fossil ones is a proof that in the 
latter these structures were similarly delicate. I find them preserved in a few 
specimens—as in fig. 15—where they prove to be of much smaller size relatively to 
the vessels than in the living plants. 
Fig. 15 represents a yet larger rootlet, enlarged 37 diameters, in which all the 
tissues are preserved. The vascular bundle, a, has increased to six vessels of larger 
dimensions, and, as already stated, the ring of inner cortex, c, encloses the representa¬ 
tives, b, of the.“ cellules liberiennes ” of M. Van Tieghem. The outermost cells, c , of 
the inner cortex exhibit a somewhat defined transition from those of the middle bark, 
suggesting an homology with the “membrane protectrice ” of the French botanist. 
The cells of the middle bark, e, are here well preserved and consist of parenchymatous 
tissue, very much more delicate and thin-walled than those of either the inner, c, or 
the outer cortex, d. 
Fig. 14 is a section of a rootlet of the most usual size and aspect. It is enlarged 
21 diameters, its actual breadth being about '2. The vascular bundles, a, and their 
surrounding cylinders of inner bark c of two similar rootlets are further enlarged in 
figs. 18 and 19. We now find that the vessels of fig. 18 have increased to 18, and 
in fig. 19 to 22 in number. But even this is not the final increase, since I have 
two rootlets in my cabinet each of which has a diameter of fully '4. The central 
portion of one of these is represented in fig. 20, in which the vessels composing the 
bundle have increased to 38 in number, the last formed ones, a\ being much smaller in 
size than those which immediately preceded them. The first formed ones, a, have, in 
this instance, become accidentally detached from the inner bark cylinder, which is a 
rather unusual occurrence. 
The illustrations I have given will, I think, demonstrate the fact that we have in each 
of these rootlets a single vascular bundle that commences its development eccentrically 
in relation to the centre of the rootlet to which it belongs, and that, proceeding from 
this one pole, it is developed cen tripet ally, the extent of that development, including 
the number of its constituent vessels, bearing a general relation to the age of the 
