When we turn to the Cordaiteæ we come to a group of plants which bears distinct relationship to the preceding, but which has a number of individual characters. It is a group of which we should know nothing were it not for the fossils preserved in the Palæozoic rocks; yet, notwithstanding the fact that it flourished so long ago, it is a family of which we know much. At the time of the Coal Measures and the succeeding Permo-carboniferous period, it was of great importance, and, indeed, in some of the French deposits it would seem as though whole layers of coal were composed entirely of its leaves. Among the fossil remains of this family there are impressions, casts, and true petrifactions, so that we know both its external appearance and the internal anatomy of nearly every part of several species of the genus. For a long time the various fossil remains of the plant were not recognized as belonging to each other and together forming the records of one and the same plant—the broad, long leaves with their parallel veins were looked on as Monocotyledons (see fig. 61); the pith casts (see fig. 63) were thought to be peculiar constricted stems, and were called _Sternbergia_; while the wood, which was known from its microscopic structure, was called _Araucarioxylon_—but the careful work of many masters of fossil botany, whose laborious studies we cannot describe in detail here, brought all these fragments together and proved them to belong to _Cordaites_. [Illustration: Fig. 61.—Leaf of _Cordaites_, _l_, attached by its broad base to a Stem, _s_] We now know that _Cordaites_ were large trees, with strong upright shafts of wood, to whose branches large simple leaves were attached. The leaves were much bigger than those of any living Gymnosperm, even than those of the Kauri Pine (a member of the Araucariaceæ), and seem in some species to have exceeded 3 ft. in length. The trees branched only at the top of the main shaft, and with their huge sword-like leaves must have differed greatly in appearance from any plant now living. The leaves had many parallel veins, as can be seen in fig. 61, and were attached by a broad base directly to the main stem; thus coming closer to the Araucarias than the other groups of Gymnosperms in their leaf characters. [Illustration: Fig. 62A.—Microscopic Section of Part of a Leaf of _Cordaites_ V, Vascular bundle; W, wood of bundle; _sh_, its sheath; S^1, large sclerenchyma mass alternating with bundles; S^2 and S^3, sclerenchyma caps of bundle; P, soft tissue of leaf.]