The gradual recognition of this group from among the numerous plant fragments of Palæozoic age is one of the most interesting of the accumulative discoveries of fossil botany. Ever since fossil remains attracted the attention of enquiring minds many “ferns” have been recognized among the rich impressions of the Coal Measures. Most of them, however, were not connected with any structural material, and were given many different names of specific value. So numerous were these fern “species” that it was supposed that in the Coal Measure period the ferns must have been the dominant class, and it is often spoken of even yet as the “Age of Ferns”. From the rocks of the same age, preserved with their microscopical structure perfect, were stems which were called _Lyginodendron_. In the coal balls associated with these stems (which were the commonest of the stems so preserved) were also roots, petioles, and leaflets, but they were isolated, like the most of the fragments in a coal ball, and to each was given its name, with no thought of the various fragments having any connection with each other. Gradually, however, various fragments from the coal balls had been recognized as belonging together; one specimen of a petiole attached to a stem sufficed to prove that all the scattered petioles of the same type belonged also to that kind of stem, and when leaves were found attached to an isolated fragment of the petiole, the chain of proof was complete that the leaves belonged to the stem, and so on. By a series of lengthy and painstaking investigations all the parts of the plant now called _Lyginodendron_ have been brought together, and the impressions of its leaves have been connected with it, these being of the fernlike type so long called _Sphenopteris_, illustrated in fig. 77. [Illustration: Fig. 77.—_Sphenopteris_ Leaf Impression, the fernlike foliage of _Lyginodendron_] [Illustration: Fig. 78A.—Diagram of the Transverse Section of Stem of the _Lyginodendron_ _p_, Pith; P, primary wood groups; W, secondary wood; _l.t_, leaf trace; _s_, sclerized bands in the cortex; S, longitudinal view of wood elements to show the rows of bordered pits.] The anatomy of the main stem is very suggestive of that of a Cycad. The zones of secondary wood are loosely built, the quantity of soft tissue between the radiating bands of wood, and the size of the pith being large, while from the main axis double strands of wood run out to the leaf base. The primary bundles, however, are not like those of a Cycad stem, but have groups of _centripetal_ wood within the protoxylem, and thus resemble the primary bundles of _Poroxylon_ (see p. 97), which are more primitive in this respect than those of the Cycads.