base, where they continued into the epidermis of the sporangiophore. In Sphenophyllum fertile both the ventral lobes of the sporophyll (corresponding to the sporangiophores in other species) and the dorsal lobes, which in other species are sterile, were developed as peltate sporangiophores. In other species of Sphenophyllum, which are known only as impressions, single sporangia, or groups of four, appear to have been inserted directly on the upper surface of the bracts. In Cheirostrobus a similar relation of sporangiophores to bracts existed, but here each bract was divided into three segments. From each segment, near its base, a stalked peltate sporangiophore arose; this bore four sporangia, which hung parallel to the stalk. That these three sterile segments, with their sporangiophores, are together comparable to one of the bracts of Sphenophyllum, with its sporangiophores, is shown by the vascular supply in each case being derived from a single leaf-trace. So far as is at present known, the Sphenophyllales were homosporous. The differences between the two genera described above are sufficiently marked to justify the division of the Sphenophyllales into the two orders Sphenophyllaceae and Cheirostrobaceae. A consideration of the characters of both shows that the Psilotales are the nearest living representatives of the Sphenophyllales, while resemblances suggesting actual relationship exist between this group and the Equisetales and Lycopodiales. It has been suggested that the Sphenophyllales may have sprung from a very old stock which existed prior to the divergence of the latter groups. So long, however, as our knowledge of these phyla is confined, as at present, to specialized forms, the nature of the relationship between them must remain to some extent hypothetical.
III. Psilotales.—The two genera Psilotum and Tmesipteris, which are provisionally isolated in this group, have usually been classed with the Lycopodiales. Recent work both on their anatomy and on the morphology and structure of their spore-producing organs has however tended to show that their peculiarities can be best understood in the light of our knowledge of the Sphenophyllales. Some authorities place them in this group and there is much to be said in support of the close relationship implied. The Psilotaceae, however, differ from the Sphenophyllales in a number of definite features, such as the arrangement of the leaves singly and not in whorls, and the mode of branching. These differences and our comparatively imperfect knowledge of the Sphenophyllaceous plants which most closely resemble the Psilotaceae appear to justify the provisional isolation of the latter as a distinct group, showing affinities with both the Sphenophyllales and Lycopodiales. In both Psilotum and Tmesipteris the functions of the root-system, which is completely absent, are performed by leafless rhizomes bearing absorbent hairs and inhabited by an endophytic fungus. Psilotum lives epiphytically or in soil rich in humus, while Tmesipteris is epiphytic (and, it has been suggested, partially parasitic) upon stems of tree ferns: the former has small scale-like leaves; those of the latter are of considerable size. The stem is monostelic, the protoxylem groups being towards the periphery of the xylem, the development of which is thus centripetal; the centre of the stele is occupied by sclerenchymatous tissue. The leaves, which bear the sporangia, are dichotomous, and do not form definite cones, but alternate in irregular zones with the foliage leaves. The sporophylls may exceptionally undergo further dichotomies and bear more numerous synangia. The sporangia of the Psilotaceae are associated in synangia, which occupy the same position relatively to the sporophyll, as the single sporangium of Lycopodium or the group of sporangia in Spenophyllum majus. The careful study of the development of the synangium of Tmesipteris, which consists of two loculi, and of Psilotum, which consists of three, has shown that their structure can be explained as originating by the septation of a single sporangium resembling that of Lycopodium. Other views of the nature of the Psilotaceous synangium are, however, possible, and indeed the existence of both simple and complicated sporangiophores in the Sphenophyllaceae leaves the question open as to whether the synangium in existing Psilotaceae is a relatively simple type of sporangiophore which has persisted unaltered or is the result of reduction from a more elaborate structure. There is some reason to believe that the prothallus of Psilotum resmbles some Lycopodium prothalli, but conclusive evidence is wanting; that of Tmesipteris is unknown.
IV. Lycopodiales.—The living representatives of this group are of small size compared with the related plants which lived in Palaeozoic times. A large proportion of the living species are tropical, though others have a wide distribution. As general characteristics of the Lycopodiales, the simple form of the leaves, which are generally of small size, and the situation of the sporangia on the upper surface of the sporophylls, which are often associated in cones, close to their insertion on the axis, may be mentioned; there are both homosporous and heterosporous forms, the prothalli exhibiting corresponding differences. A number of species of Lycopodium are epiphytic and those of Isoëtes live submerged in water. Vegetative reproduction is effected in various ways: by the separation of the branches of a creeping stem in some Lycopodia, the persistence through the winter of the apex of the shoot in L. inundatum, and by the formation of leafy bulbils on the aerial stem of L. Selago and others. A highly specialized means of vegetative reproduction is seen in the tubers of Phylloglossum and the embryos of some Lycopods. The modifications shown by the gametophyte of Lycopodium will be described below. All such special relations of the plant to its environment, which might be expected in the few forms of a large group which has persisted beyond the others, are less marked in the genus Selaginella. It would appear as if the latter was more suited to the conditions of the existing flora, and many of the specific forms within it may rather be regarded as recently evolved than as simply persistent.
Lycopodiaceae.—This order contains the two genera Phylloglossum and Lycopodium; the former has a single species, confined to Australia, Tasmania and New Zealand, while nearly one hundred species of Lycopodium are known. Erect and creeping terrestrial plants and pendulous epiphytes occur in the latter genus. The simple leaves, which are of small size and do not possess a ligule, are arranged spirally around the branched stem in the majority of the species. The roots of the erect forms often grow downwards in the cortex of the stem to reach the soil. The anatomy of Lycopodium presents considerable variety in detail, but the stem is always monostelic and the development of the xylem centripetal, the protoxylems being situated at the periphery of the stele; pericycle and endodermis surround the stele, and the wide cortex may be more or less