Hornea Lignieri, a plant of simi lar dimensions and habit, has tuberous swellings at the base of the shoots. The differentiation of vascular tissue, which forms a cylindrical strand in the stem, only extends for a short distance into the swollen base, whose un der surface bore absorptive hairs. The sporangia differ from those of Rhynia in having a sterile col umn of tissue extending from the base up into the midst of the spor angium. This column has been compared with the columella found in the capsules of the Liver worts and Mosses. In both Rhy nia and Hornea the sporangia look something like swollen branch-tips and in Hornea branched sporangia are known. Asteroxylon Mackei (fig. 3, A) a more complex plant, had smooth horizontal stems with small root like branches; true roots are absent just as in Rhynia and Hornea; the aerial shoots are erect and densely clothed with small scale-like leaves. The stele contains a fluted column of xylem which appears in section like a multi-rayed star. Small strands of tracheids con nected with the stele pass out towards each leaf but stop short before entering the base of the leaf which had no vascular tissue. In Thursophyton a probably closely allied plant from the Middle Devonian the leaves are small and spine-like and vary consider ably in size. These facts suggest that in plants of the small-leaved type such as the lycopods the leaves may be the homologues of spines or in other words the small leaf of the lycopod may have evolved from a small spine or emergence of the shoot by a gradual process of vascularisation, an early stage in the process being represented by the condition found in Asteroxylon where the vascularisation has not yet extended from the shoot into the leaf itself. The sporangia of Asteroxylon were borne, in all probability, on small smooth branches. Another species of Asteroxylon has been found in Germany which has leafless upper branches, com paring therefore with the supposed fertile branches of the Scot tish plant. Nothing is known about the sexual generation of these Devonian plants. A consideration of the Psilophytales as a group suggests that they may represent, in the Asteroxylon-type with its small leaves, the f ore-runners of the Lycopods. The branching shoot of Rhynia with its terminal sporangia suggests that the large megaphyllous fern frond with sporangia on its edges may have originated from such a branching shoot. The column of tissue in the sporangium of Hornea suggests comparison with the columella of the moss or liverwort. The living group, the Psilotales, which compare with Rhynia in being rootless and sometimes almost leafless but which differ considerably in the position of the sporangia on the plant, may also be related. Finally, in the absence of differentiation into stem and leaves, the plant body of Rhynia with its simple thalloid form calls at tention to the suggestion put forward by more than one writer that land-plants are descended from algal ancestors. It has been remarked that the sporangia in the vascular plants correspond morphologically to the branches of the algal thallus which bear the reproductive organs. (e.g., the stichidia in the red sea-weeds). This is interesting in view of the resemblance between sporangia and branch tips already referred to in Hornea and to the structure of the fertile branches in Sporocarpon the Upper Devonian thal lophyte in which the tips bear isolated tetrads of spores. There was clearly in Devonian times a group of plants in existence in which there was a combination of thallophyte and pteridophyte characters and, as Kidston and Lang point out, "the facts are . . . consistent with the Rhyniaceae finding their place near the beginning of a current of change from an Alga-like type of plant to the type of the simpler Vascular Cryptogams." Articulatales.—This large group of the Pteridophyta includes the Proto-articulatineae, Sphenophyllineae, and the Equisetineae. The first is represented in the Middle De vonian of Germany by Calamophyton primaevnm (fig. 4) a plant recently de scribed by Krause' and Weyland. The stems are jointed and have small bifid leaves at tached at the nodes. The sporangia are borne in pairs on modified leaves which compare closely therefore with the spor angiophores of the Equisetineae. Hyenia of the same age has small deeply cleft leaves and in vegetative features re semble the Sphenophyllineae. The fertile leaves are forked and two or three sporan gia are borne on the ends of each division.
Sphenophyllineae.— Sphenophyllum, a genus common in the Carboniferous, ranges from the Upper Devonian to the Triassic. The plant grew in the form of a slender shrub which may have supported its branches by scrambling over other plants. The stems are jointed with whorls of small wedge-shaped leaves, which are often deeply dissected (P1. I., fig. 8). The leaves of one whorl lie immediately above those of the whorl below and are usu ally in multiples of three. This arrangement is related to the structure of the stele which is triangular in section and has a central column of xylem ; at the nodes two vascular strands pass out from each corner to supply the leaves. The stems and roots have a considerable amount of secondary vascular tissue. While
not much variation is shown in the vegetative parts of the known species of Sphenophyllum there is considerable variety in their cones. In Sphenophyllum Dawsoni the cones are several inches long and at least half an inch thick. They are built up of whorls of leaf-like bracts and those of each whorl are coherent laterally for half their length from the base so that each whorl forms a cup (fig. 5). The whorls of bracts overlap so that the cone is covered on the outside by their overlapping tips. The sporangia are twice as numerous as the bracts and are attached by nar row stalks (sporangiophores) to the upper surface of the sheath of bracts close to the axis of the cone. The sporangiophores are not all of the same length so that the sporangia had the appearance of being arranged in more than one whorl in each cup. In Bowmanites a cone belonging to some closely related plant, the sporangiophores had a bract-like expansion at the end and each bore two sporangia. This arrangement suggests that bracts and sporangiophores may be equivalent structures and that the cone of Sphenophyllum may be regarded as being built up of a series of bi-lobed leaves, the ventral lobes being fertile while the dorsal assumed the form of subtending bracts. In S. fertile this theory receives support as dorsal and ventral lobes reveal their equiva lence by both being fertile. In other species there are no distinct terminal cones; nodes with fertile leaves alternate with nodes hearing sterile leaves. This hypothetical bi-lobed leaf may also be distinguished in Cheirostrobus, a very complicated cone of Lower Carboniferous age. Here the leaf is divided into three dorsal sterile and three ventral fertile segments. Each fertile segment bears four sporangia and in construction is exceedingly like the sporangiophore of the Equisetineae and it is possible that Cheirostrobus is a representative of the stock from which the Sphenophyllineae and Equisetineae have both sprung.
Equisetineae.—The fossil Equisetineae were at their zenith in Carboniferous times; in place of the small, herbaceous Equisetum (horse-tail) of to-day there were gigantic trees some with a girth of three metres and a height of 6o metres forming a conspicuous feature in some aspects of the vegetation. Calamites the best known genus had the same habit as the horse-tail with jointed stem and small leaves but differed in its much greater size. The internal structure of the stems, roots and leaves, is in many respects like that of Equisetum only there is secondary thickening in the stems and roots. The leaves, usually small, linear or lance shaped, are often united near the base to form a sheath round the stem. Like Equisetum the stems are hollow with a large central cavity. Casts of these cavities in sandstone are frequently found showing longitudinal grooves representing the courses of the primary woody strands which projected into the pith cavity. There is a constriction on the cast at each node and the grooves of one internode alternate with those on the internode above and below. Two main types of calamite-foliage are known : Astero phyllites and Annularia; in the former the leaves are long and awl-shaped and quite separate. In Annularia the leaves are usually more paddle-shaped terminating in a small sharp point. The single vein which passes out into this point had, in some species, a swelling near the tip of the leaf suggesting the presence of a water gland. In another type of Annularia the margins of the leaf were inrolled and bore a fringe of hairs which probably sheltered the stomatal surface of the leaf. The cones exhibit a considerable range in form, indicating that several families and genera may be other species are heterosporous some sporangia having large num bers of small spores while others in the same cone had a smaller number of large spores. In Palaeostachya (fig. 6, B) the spo rangiophores unlike those of Calamostachys are attached in the axile of the bracts which subtend them. In Archaeocalamites, a characteristically Lower Carboniferous plant, the leaves are long and may be twice forked while the vascular strands of the inter included in Calamites. In Calamostachys the cones were terminal an the branches and the axis of the cone bore whorls of peltate sporangiophores similar to those of Equisetum but alternating with whorls of sterile bracts (fig. 6, C). The sporangiophores are in superposed whorls while the bracts of one whorl alternate with those above and below. Each sporangiophore bears four sporangia in which the spores are found sometimes arranged in tetrads. In some species the snores in the cone are all of the same size but nodes pass straight through and do not alternate at the nodes. The cones (fig. 6, A) are very like those of Equisetum but are long and constricted at intervals where forked leaves were attached so that they resemble a branch in which the internodes are covered with sporangiophores. Although the presence of hetero spory and secondary thickening are probably stages in the direction of a gymnospermous type the Equisetineae must be considered as essentially Pteridophyta.