OF THE FOSSIL PLANTS OF THE COAL-MEASTTEES. 
227 
lary sheath, whilst in these fossil Cryptogams they are given off from the more external 
woody cylinder*; but this difference may be explained by the fact that in the former 
plants the spirals of the medullary sheath are altogether different from the non-spiral 
ones in the woody zones, whilst in the latter the two classes of vessels have the same 
structure, and differ only in size. Hence in the Cryptogams the one set may be sub- 
stituted for the other, which could not be done in the ordinary Exogens. 
My supposition respecting the relations subsisting between the inner vascular ring of 
Lepidodendron and the medullary sheath of Exogens receives fresh support from the 
structure of Stigmaria. In the latter plant, now well known to be a Sigillaroid root, 
we find no inner or medullary cylinder of vessels. The ligneous zone receives a won- 
derful development ; it is furnished with an abundance of medullary rays, and gives 
off numerous vascular bundles which are supplied to the epidermal rootlets that here 
occupy the place of leaves. We have here a parallel state of things to that seen in the 
roots of Exogens, in which in like manner the medullary sheath is wanting. This 
curious coincidence has not escaped the observant eye of M. Brongniart, who calls 
attention to it in his memoir on Sigillaria elegansf. 
My specimens throw no direct light upon the structure of the vascular and medullary 
axis of the true Sigillarite as distinguished from the Favularian type ; but the cortical 
portions of all the plants, including the true Sigillarife, exhibit what is practically an 
identity of structure. In all we have a remarkably thick spongy bark, reminding us in 
many of its features of that found in the living Cycads. This consisted either of paren- 
chyma, prosenchyma, or of both combined, enclosed externally in a bast-layer of elon- 
gated prosenchymatous tubes, which in turn is invested by a layer of cellular paren- 
chyma supporting the bases of leaves, the latter invariably consisting of the same form 
of parenchyma as the epiderm. M. Brongniart’s specimen of Sigillaria ( Favularia ) 
elegans exhibits a central axis, the structure of which is nearly identical with that of my 
Plate XXVIII. figs. 33, 34. This, in its turn, only differs from the more ordinary forms 
of Diploxylon , in the crenulated outline which separates the ligneous zone from the 
cylinder of medullary vessels, giving to the exterior of the latter a fluted aspect like that 
of a Calamite, but without the transverse nodal constrictions of the latter genus. The 
Diploxylons again, as I have already shown, shade off into the ordinary forms of Lepido- 
dendron , and are undoubtedly Lepidodendroid plants which have lost the central por- 
tions of their medullary axes. Remove the cellular tissues from the centre of the plant 
which I have represented in figs. 8 & 9, and we have, at once, the closest resemblance to 
Witham’s Anabathra and Corda’s Diploxylon , as well as to those now under consider- 
ation. That Witham’s plant is identical, in type, with mine, is further indicated by his 
tab. 8. fig. 12, where he exhibits one of the large compound medullary rays shown in my 
Plate XXVII. fig. 23. The cellular tissues have not been preserved in the medullary rays 
of Brongniart’s Sigillaria elegans-, but tab. 4. fig. 2 of his memoir shows that his plant 
* The reverse proves to be the case, hence this objection disappears. See note on page 237. 
t Loc. cit. p, 433. 
