In many species of _Lepidodendron_ the quantity of secondary wood formed round the primary stele was very great, so that (as is the case in higher plants) the primary wood became relatively insignificant compared with it. In most species of _Lepidodendron_ the primary stele is a hollow ring of wood (cf. fig. 38, p. 62) round which the secondary wood developed, as is seen in fig. 49. These two cases illustrate a peculiarity of fossil plants. Among living ones the solid and the simple ring stele are almost confined to the Pteridophytes, where secondary wood does not develop, but the palæozoic Pteridophytes, while having the simple primary types of steles, had quantities of secondary tissue, which was correlated with their large size and dominant position. [Illustration: Fig. 50.—Diagram of Steles of the English _Medullosa_, showing three irregular, solid, steles A, with secondary thickenings S, all round each. _a_, Small accessory steles] Among _polystelic_ types (see p. 63) we find interesting examples in the fossil group of the _Medulloseæ_, which are much more complex than any known at present, both owing to their primary structure and also to the peculiar fact that all the steles developed secondary tissue towards the inner as well as the outer side. One of the simpler members of this family found in the English Coal Measures is illustrated in fig. 50. Here there are three principal protosteles (and several irregular minor ones) each of which has a considerable quantity of secondary tissue all round it, so that a portion of the secondary wood is growing in towards the actual centre of the stem as a whole—a very anomalous state of affairs. In the more complex Continental type of _Medullosa_ there are _very_ large numbers of steles. In the one figured from the Continent in fig. 51 but a few are represented. There is a large outer double-ring stele, with secondary wood on both sides of it, and within these a number of small steles, all scattered through the ground tissue, and each surrounded by secondary wood. In actual specimens the number of these central steles is much greater than that indicated in the diagram. No plant exists to-day which has such an arrangement of its vascular cylinder. It almost appears as though at the early period, when the Medulloseæ flourished, steles were experimenting in various directions. Such types as are illustrated in figs. 50 and 51 are obviously wasteful (for secondary wood developing towards the centre of a stem is bound to finally meet), and complex, but apparently inefficient, which may partly account for the fact that this type of structure has not survived to the present, though simpler and equally ancient types have done so. [Illustration: Fig. 51.—Continental _Medullosa_, showing R, outer double-ring stele with secondary wood all round it; S, inner stellate steles, also surrounded in each case by secondary tissue]