those of the Coal period." Returning from this digression to the forests of the Coal period, we may first notice that which is the most conspicuous and abundant tree in the swampy levels--the Sigillaria or seal-tree, so called from the stamp-like marks left by the fall of its leaves--a plant which has caused much discussion as to its affinities. Some regard it as a gymnosperm, others as a cryptogam. Most probably we have under this name trees allied in part to both groups, and which, when better known, may bridge over the interval between them. These trees present tall pillar-like trunks, often ribbed vertically with raised bands, and marked with rows of scars left by the fallen leaves. They are sometimes branchless, or divide at top into a few thick limbs, covered with long rigid grass-like foliage. On their branches they bear long slender spikes of fruit, and we may conjecture that quantities of nut-like seeds scattered over the ground around their trunks are their produce. If we approach one of these trees closely, more especially a young specimen not yet furrowed by age, we are amazed to observe the accurate regularity and curious forms of the leaf-scars, and the regular ribbing, so very different from that of our ordinary forest trees. If we cut into its stem, we are still further astonished at its singular structure. Externally it has a firm and hard rind. Within this is a great thickness of soft cellular inner bark, traversed by large bundles of tough fibres. In the centre is a core or axis of woody matter very slender in proportion to the thickness of the trunk, and still further reduced in strength by a large cellular pith. Thus a great stem four or five feet in diameter is little else then a mass of cellular tissue, altogether unfit to form a mast or beam, but excellently adapted, when flattened and carbonised, to blaze upon our winter hearth as a flake of coal. The roots of these trees were perhaps more singular then their stems; spreading widely in the soft soil by regular bifurcation, they ran out in long snake-like cords, studded all over with thick cylindrical rootlets, which spread from them in every direction. They resembled in form, and probably in function, those cable-like root-stocks of the pond-lilies which run through the slime of lakes, but the structure of the rootlets was precisely that of those of some modern Cycads. It was long before these singular roots were known to belong to a tree. They were supposed to be the branches of some creeping aquatic plant, and botanists objected to the idea of their being roots; but at length their connection with Sigillaria was observed simultaneously by Mr. Binney, in Lancashire, and by Mr. Richard Brown, in Cape Breton, and it has been confirmed by many subsequently observed facts. This connection, when once established, further explained the reason of the almost universal occurrence of Stigmaria, as these roots were called,