OF THE FOSSIL PLiNTS OF THE COAL-MEASURES. 
291 
where, instead of being detached from the medullary vascular cylinder, the boundary 
between the two was only indicated in the transverse section by two slight lateral 
constructions in the otherwise unbroken circle of vessels. Other sections of the 
same specimen in Mr. Aitken’s cabinet exhibit similar conditions. We have here 
a clear proof that, contrary to the conclusions of M. Renault, age does bring about 
very important changes in the form and arrangement of the tissues in the branches of 
these plants. Other examples of stems in a very much younger state, for which 
I am indebted to Mr. Aitken and Mr. Cash, exhibit corresponding phenomena. 
My most recent researches have brought to light another very remarkable series of 
facts indicating the Lycopodiaceous character of the Lepidodendroid and Sigillarian 
stems. Whatever else may be doubtful, there is no doubt that Stigmciria jicoides, 
with its peculiar rootlets, is the root alike of Lepidodendron and of Sigillaria. I had 
long noticed the circumstance that the vascular bundle in each of these rootlets was 
rarely in the centre of the small innermost circle of cortical cells by which it is usually 
surrounded ; but such structures are so often pushed out of their normal position in 
these fossil plants that I regarded this eccentric position of the rootlet bundle as 
accidental. The perusal of a very important paper by M. Ph. Van Tieghem,* first 
called my attention to a peculiarity in the structure of the true roots of the Lycopods, 
especially of the Selaginellece , and led me to re-examine the rootlets of Stigmaria 
with the results to be described. M. Van Tieghem has shown that when a rootlet 
belonging to any class of plants is first formed its primitive vascular elements commence 
by developing a small vascular bundle at each of two or more points on the periphery 
of a central cellular circle, from which points each bundle grows by the centripetal 
additions to the number of the vessels composing it, until the several bundles approach 
each other centripetally, and thus constitute a symmetrical group of vessels. These 
primary rootlet bundles are never less than two in number, with the exception of the 
ultimate rootlets of the Lycopodiacece and the Opliioglossece. Each terminal rootlet in 
these classes of plants contains but one triangular vascular bundle. This bundle first 
appears as one or two very minute vessels attached to one side of the small cellular 
cylinder within which it is located. New and larger vessels are subsequently added 
centripetally to these primary ones, but the bundle always remains eccentric in 
reference to the cellular sheath within which it is enclosed. The remainder of the 
cavity within the sheath is occupied by some large thin-walled cells, which M. Van 
Tieghem regards as the equivalents of the liber of other plants. Fig. 13 represents 
the central portion of a rootlet of Selaginella Martensii, in which a indicates the first 
formed vessels of the vascular bundle, and a the later, centripetally developed and 
larger ones. The large cells, l>, are those which M. Van Tieghem identifies with the 
liber, whilst c indicates the next inneimost zone of the cortex. 
Figs. 14 to 20 exhibit sections of rootlets of Stigmaria Jicoides in various stages of 
* “ Recherches sur la symmetric de structure des plantes vasculaires.” Par M. Ph. Yan Tieghem. 
Annales des Sciences Natnrelles.’ Cinquieme serie—Botanique, tome 13, 1870-71. 
2 Q 
MDCCC’LXXX I. 
