But the most interesting point about the _sigillariae_ is the root. This was for a long time regarded as an entirely distinct individual, and the older geologists explained it in their writings as a species of succulent aquatic plant, giving it the name of _stigmaria_. They realized the fact that it was almost universally found in those beds which occur immediately beneath the coal seams, but for a long time it did not strike them that it might possibly be the root of a tree. In an old edition of Lyell's "Elements of Geology," utterly unlike existing editions in quality, quantity, or comprehensiveness, after describing it as an extinct species of water-plant, the author hazarded the conjecture that it might ultimately be found to have a connection with some other well-known plant or tree. It was noticed that above the coal, in the roof, stigmariae were absent, and that the stems of trees which occurred there, had become flattened by the weight of the overlying strata. The stigmariae on the other hand, abounded in the _underclay_, as it is called, and were not in any way compressed but retained what appeared to be their natural shape and position. Hence to explain their appearance, it was thought that they were water-plants, ramifying the mud in every direction, and finally becoming overwhelmed and covered by the mud itself. On botanical grounds, Brongniart and Lyell conjectured that they formed the roots of other trees, and this became the more apparent as it came to be acknowledged that the underclays were really ancient soils. All doubt was, however, finally dispelled by the discovery by Mr Binney, of a sigillaria and a stigmaria in actual connection with each other, in the Lancashire coal-field. Stigmariae have since been found in the Cape Breton coal-field, attached to Lepidodendra, about which we have already spoken, and a similar discovery has since been made in the British coal-fields. This, therefore, would seem to shew the affinity of the sigillaria to the lepidodendron, and through it to the living lycopods, or club-mosses. Some few species of stigmarian roots had been discovered, and various specific names had been given to them before their actual nature was made out. What for some time were thought to be long cylindrical leaves, have now been found to be simply rootlets, and in specimens where these have been removed, the surface of the stigmaria has been noticed to be covered with large numbers of protuberant tubercles, which have formed the bases of the rootlets. There appears to have also been some special kind of arrangement in their growth, since, unlike the roots of most living plants, the tubercles to which these rootlets were attached, were arranged spirally around the main root. Each of these tubercles was pitted in the centre, and into these the almost pointed ends of the rootlets fitted, as by a ball and socket joint. [Illustration: FIG. 17--_Section of stigmaria_.]