There can be no doubt that these tubercules must once have supported leaflets. They are true leaf-scars, like those on the Scotch fir, and the lozenge-shaped knobs on the bark of lepidodendron. But of the form of these leaves we are still in ignorance, for no part of the plant, save the stem and roots, has yet been found. The sigillaria must have been a tree that could not long withstand maceration, for not only are its leaves gone, but, in many cases, the outer bark has partially or wholly decayed, leaving a scarcely distinguishable mass of carbonized matter.[23] When this outer rind is peeled off, the inner surface of the stem is seen to be ridged, furrowed, and tuberculed in the same way, but the markings are much less distinct than on the outside. The bark sometimes attains the thickness of an inch, and is always found as a layer of pure coal enveloping the stem where it stands erect, or lying as a flat cake without any central cylinder where the stem is prostrate. (See Fig. 14.) [Footnote 23: Another proof of the looseness of the texture of this ancient vegetable may be gathered from the almost invariable truncation of even the largest erect stems; they are snapped across at the height of a few feet from their base. The famous "Torbanehill Mineral" contains many such fragmentary stems, often of considerable thickness. Their interior consists of the same material as the surrounding bed, and displays many dissevered plants that may have been washed into the decaying trunks. For the internal structure of sigillaria see Dr Hooker's _Memoir_, and the authorities therein cited.]