Such very young stems are rare, for the development of secondary wood began early, and it soon greatly exceeded the primary wood in amount. Fig. 114 shows a photograph of a stem in which the secondary wood is well developed. The primary triangle of wood is still to be seen in the centre, and corresponds to that in fig. 113, while closely fitting to it are the bays of the first-formed secondary wood, which makes the wood mass roughly circular. Outside this the secondary wood forms a regular cylinder round the axis, which shows no sign of annual rings. The cells of the wood are large and approximately square in shape, while at the angles formed at the junction of every four cells is a group of small, thin-walled parenchyma, see fig. 115. There are no medullary rays going out radially through the wood, such as are found in all other zones of secondary wood, and in this arrangement of soft tissue the plants are unique. [Illustration: Fig. 114.—_Sphenophyllum_, Transverse Section with Secondary Wood W. At _c_ the cork formation is to be seen. (Microphoto.)] Beyond the wood was a zone of soft tissue and phloem, which is not often preserved, while outside that was the cork, which added to the cortical tissues as the stem grew (see fig. 114, _c_). [Illustration: Fig. 115.—Group of Wood Cells _w_, showing their shape and the small soft-walled cells at the angles between them _p_] Petrified material of leaves and roots is rare, and both are chiefly known through the work of the French palæobotanist Renault. The leaves are chiefly remarkable for the bands of sclerized strengthening tissue, and generally had the structure of aerial, not submerged leaves. The roots were simple in structure, and, as in _Calamites_, had secondary tissue like that in the stems. In the case of the fructifications it is the English material which has yielded the most illuminating specimens. The cones were long and slender, externally covered by the closely packed tips of the scales, which overlapped deeply. Between the whorls of scales lay the sporangia, attached to their upper sides by slender stalks. A diagram will best explain how they were arranged (see fig. 116). Two sporangia were attached to each bract, but their stalks were of different lengths, so that one sporangium lay near the axis and one lay outside it toward the tip of the bract. [Illustration: Fig. 116.—Diagram of Arrangement of Scales and Sporangia in Cones of _Sphenophyllum_ A, Axis; _br_, bract; S, sporangium, with stalk _st_.]