Sporophyte with roots, stems and whorled leaves. Protostelic (solid or medullated). Some with secondary thickening. Sporangia thick-walled, homosporous (or heterosporous), usually borne in a reflexed position on sporangiophores arranged in whorls. Antherozoids multiflagellate.

List 7


Until 1957 the earliest known representatives of the Sphenopsida were Hyenia and Calamophyton, both of which are of Middle Devonian age, but in that year Ananiev discovered the remains of a most interesting plant in Lower Devonian rocks of western Siberia. This he named Protohyenia (Fig. 15A). Although lacking some of the features which are characteristic of the Sphenopsida, yet, as the generic name suggests, it might well represent an early ancestor of the group. From a creeping axis, erect branches arose at intervals, bearing either sterile or fertile appendages in rather indefinite whorls. The sterile appendages forked several times and, although having the appearance of tiny lateral branches, they probably functioned as leaves. The fertile appendages were very similar, but terminated in sporangia. These were unlike those of almost all other members of the Sphenopsida, in that they were not reflexed.

Hyenia elegans (Fig. 15E) had a similar growth habit, as we now realize from the work of Leclercq, for it had a stout horizontal rhizome bearing roots and erect aerial stems up to 30 cm high, some sterile and others fertile. The sterile axes bore whorls of forking appendages, alternating at successive nodes and, as in the case of Protohyenia, it is difficult to decide whether they should be regarded as leaves or as stems performing the functions of leaves. The fertile axes bore whorls of sporangiophores (Fig. 15F), which were similar to the ‘leaves’, except that two segments were reflexed and usually terminated in two sporangia each.

Protohyenia janovii

Other species of Hyenia are known, in which the aerial axes were branched and which, therefore, resembled quite closely the other Middle Devonian genus Calamophyton. The illustration of C. primaevum (Fig. 15B) is taken from an early description which emphasizes the articulate nature of the aerial axes, a feature which used to be regarded as essential in defining the genus. However, other species, e.g. C. bicephalum, are not so clearly articulated and it has been suggested that the two genera merge into one another. Certainly the sterile and fertile appendages of C. bicephalum (Figs. 15C and 15D) were very similar to those of Hyenia, the chief difference being that the fertile appendages of the former forked more profusely and bore twelve sporangia, instead of three or four, as in the latter. The lateral appendages of C. primaevum are said to have been much simpler, forking only once, and the sporangiophores are said to have borne only two sporangia; the way in which they were restricted to special fertile branches may represent early stages in the evolution of the strobilus, which is so characteristic of later sphenopsids.

Little is known of the internal anatomy of the Hyeniales, but in Calamophyton there are indications of a triangle of pith surrounded by tracheids with reticulate or scalariform thickenings; it is also suggested that there may have been some degree of secondary thickening.


This group first appeared in the Upper Devonian and persisted until the Lower Triassic, remains of stems as well as of leaves being referred to the genus Sphenophyllum. Many species are known, all of which are characterized by the whorled arrangement of the leaves, usually in multiples of three at each node (Fig. 15K). The stems were usually very delicate, in spite of secondary thickening, for they seldom exceeded 1 cm in diameter. Presumably, therefore, they were unable to support their own weight and must have been prostrate on the ground, or must have depended on other plants for support. In general appearance, they probably looked rather like a Galium (‘Bedstraw’). The anatomy of the stem was most peculiar in its resemblance to that of a root, for in the centre was a triangular region of solid primary wood, with the protoxylems at the three corners in an exarch position. In the Lower Carboniferous species, S. insigne, the protoxylem tended to break down to form a ‘carinal’ canal, but in the Upper Carboniferous species this rarely happened. Outside the primary wood, a vascular cambium gave rise to secondary wood, first between the protoxylems, and then later extending all round. However, the wood opposite the protoxylems was composed of smaller cells than on the intermediate radii, resulting in a pattern which is quite characteristic and which is recognizable at a glance in transverse sections (Fig. 15L). The primary wood consisted entirely of tracheids (i.e. without any admixture of parenchyma) and they bore multiseriate bordered pits on their lateral walls. The tracheids of the secondary wood also bore multiseriate pits, but they were restricted to the radial walls. Between the tracheids, there were wood rays. These were continuous in S. insigne, but were interrupted in S. plurifoliatum where they were represented only by a group of parenchyma cells in the angles between adjacent tracheids. Large stems had a considerable thickness of cork on the outside, formed from a deep-seated phellogen.

The leaves of Sphenophyllum showed a wide range of structure, some being deeply cleft, while others were entire and deltoid (Figs. 15M-O); yet all received a single vascular bundle, which dichotomized very regularly within the lamina. Some species were markedly heterophyllous, as illustrated in Fig. 15K, and in these the deeply cleft leaves were usually near the base, while the entire ones were higher up on lateral branches, an arrangement that suggests that the former might represent juvenile foliage.

A number of cones, referred to the genera Sphenophyllostachys or Bowmanites, have been found attached to the parent plant; others, found detached, are placed in these genera on the basis of their general similarity. A number of other genera of cones are also referred to the SphenophyHales, but on less secure grounds. Some of them represent the most complex cones in the whole plant kingdom. One of the earliest to appear in the fossil record is Eviostachya, described by Leclercq, from the Upper Devonian of Belgium. Less than 6 cm long and less than i cm in diameter, each cone had at its base a whorl of six bracts. Above this were whorls of sporangiophores, six in each whorl. Each sporangiophore was itself highly complicated (Fig. 15G) and branched in a very characteristic way (Fig. 15H), bearing a total of twenty-seven sporangia in a reflexed position. Sporangiophores in successive whorls stood vertically above each other, as is characteristic of the Sphenophyllales, but there were no bracts between them.

Cheirostrobus, from the Lower Carboniferous of Scotland, was a large cone, 3-5 cm across, and had thirty-six sporangiphores in each whorl, subtended by the same number of bracts, each with bifurcated tips (Fig. 15I). The arrangement of the vascular supply to these appendages is interesting in that a common ‘trunk-bundle’ supplied three sporangiophores and the three bracts subtending them. This has led some morphologists to suggest a more complicated interpretation of the cone structure than is really necessary, based on the supposition that each trunk-bundle represented the vascular supply to one compound organ made up of three fertile leaflets and three sterile leaflets.

Sphenophyllostachys fertilis (= Sphenophyllum fertile, = Bowmanites fertilis) from the Upper Carboniferous (Fig. 15J) was also a complex cone. Up to 6 cm long and 2-5 cm in diameter, it was made up of whorls of superimposed sporangiophores, six in a whorl, each subtended by a pair of sterile appendages (possibly homologous with one bifid bract). Each sporangiophore terminated in a ‘mop’ of branches, about sixteen in number, each bearing two reflexed sporangia. Only detached cones have, so far, been found, but they are presumed to have belonged to some member of the Sphenophyllales, because of the triarch or hexarch arrangement of the primary wood in the axis.

Sphenophyllostachys (= Bowmanites) Dawsoni, on the other hand, is known to have been borne on stems like those of Sphenophyllum plurijoliatum. The cone was up to 12 cm long and 1 cm in diameter and bore whorls of bracts, fused into a cup near the base, but with free distal portions. In the axils of these bracts, and fused with them to a certain extent (Fig. 15P), were branched sporangiophores. In one form (forma a) each sporangiophore had three branches arranged in a very characteristic way (Fig. 15Q), each terminating in a single reflexed sporangium. In another form (forma y), there were six branches.

Sphenophyllostachys (= Bowmanites) Roemeri was similar in its organization to S. Dawsoni, forma a, except that each branch of the sporangiophores bore two reflexed sporangia (Fig. 15R).

In recent years, a number of relatively simple cones have been described, which are nevertheless believed to belong to the Sphenophyllales. Thus, in Bowmanites bifurcatus, each sporangiophore forked only once, while in Litostrobus iowensis the sporangia were borne singly on a short unbranched stalk. In the latter species, the sporangia were not reflexed but, despite this very simple organization, an affinity with Bowmanites is presumed, because of the number of bracts and the number of sporangia in a whorl (twelve and six respectively). The discovery of these simple cones has led to the suggestion that, within the Sphenophyllales, evolution has involved progressive simplification.

While the vast majority of the Sphenophyllales were homosporous, at least one, Bowmanites delectus, was heterosporous with megaspores about ten times the size of the microspores.


This group reached the peak of its development in the Upper Carboniferous, when a large number of arborescent species was co-dominant with the Lepidodendrales in coal-measure swamp forests; yet by the end of the Permian the group had become extinct. The first representatives to appear, in the Upper Devonian, were the Asterocalamitaceae, a group which differed from the later Calamitaceae in a number of interesting details. Asterocalamites (= Archaeocalamites) (Fig. 16A) had woody stems up to 16 cm in diameter, strongly grooved on the outside, with the grooves continuing through successive nodes (i.e. not alternating). The leaves, up to 10 cm long, were in whorls at the nodes and forked many times dichotomously, in a manner strongly reminiscent of Calamophyton leaves. At intervals along the more slender branches, there were fertile regions, in which there were superimposed whorls of peltate sporangiophores, each bearing four reflexed sporangia (Fig. 16B). Sometimes the fertile regions were interrupted by a whorl of leaves, but these were apparently normal leaves and could not be regarded as bracts. The absence of any regular association between bracts and sporangiophores makes an interesting comparison with the cones of the later Calamitaceae.

Protocalamites was one of the earliest representatives of the Calamitaceae, being present in the Lower Carboniferous rocks of Pettycur, Scotland. Its stems were ridged, with the ridges alternating in successive internodes, like those of most members of the family, but they differed in one important respect. Examination of a transverse section of a petrified stem (Fig. 16C) reveals a marked development of centripetal wood, as well as centrifugal (i.e. the primary wood was mesarch). As in Calamites and in Equisetum, the protoxylem tended to break down to form a carinal canal. Secondary wood was laid down to the outside of the metaxylem, but the primary wood-rays were so wide that it gives the appearance of having been formed in separate strands, although in fact it was formed from a continuous vascular cambium.


Protocalamostachys is the name given to a peculiar cone described by Walton from Lower Carboniferous rocks in the Island of Arran. Two small pieces of the cone had dropped into the hollow stump of a Lepidophloios before it became petrified. Unlike the cones of other members of the Calamitales, its sporangiophores branched twice (Fig. 16D), instead of being peltate. In this respect, it showed some affinities with the Sphenophyllales and also with the Hyeniales. However, Walton compares it most closely with Pothocites, a cone associated with leaves of the Asterocalamites type. Furthermore, within the axis of the cone there is centripetal primary wood as in the stem of Protocatamites.

The height to which Calamites grew is difficult to determine, because of the fragmentary nature of the remains, but it is almost certain that some specimens must have attained a height of 30 m with hollow trunks whose internal diameter was up to 30 cm. Strictly speaking, the generic name Calamites should be applied only to pith casts of stems and branches, while petrified wood should be described under the form genus Arthropitys, but common usage has extended the application of the name Calamites to include all methods of preservation. The pith casts exhibit ridges and grooves, corresponding in number to the protoxylem strands, running up the inside of the secondary wood and alternating at successive nodes. In this respect they differ from Mesocalamites, in which there was some variability from node to node, the ridges sometimes alternating and sometimes continuing straight across the nodes. A number of subgenera are recognized which differed in their mode of branching and, hence, in their general form. The subgenus Eucalamites branched at every node. Fig. 16E is of Eucalamites carinatus, in which there were only two branches at each node, but other species branched more profusely. By contrast, the subgenus Stylo calamites branched only near the top of the erect organpipe-like trunk.

Transverse sections of Calamites (Arthropitys) show very little primary wood indeed, for secondary thickening provided most of the wood (Fig. 16H). The protoxylem was represented by carinal canals and the small amount of metaxylem present was entirely centrifugal. The wood rays varied, according to species, dividing the secondary wood into segments in some species, or losing their identity in a continuous cylinder of wood in others. In all cases the wood contained small wood-rays in addition, but otherwise was composed entirely of tracheids with scalariform pitting or with circular bordered pits on the radial walls.

The leaves of Calamites were unbranched, with a single midvein, and occurred in whorls of four to sixty. In most species, they were free to the base, but in a few they showed some degree of fusion into a sheath. They are placed in one or other of two form genera, according to their overall shape, Annularia being spathulate or deltoid (Fig. 16F), while Asterophyllites were linear (Fig. 16G). The latter were peculiar in being heavily cutinized, with the stomata restricted to the adaxial surface, suggesting that the branches bearing them may have been pendulous. It is probable, therefore, that the cones were pendulous, too.

The cones of Calamites were borne in a variety of ways, in some species singly at the nodes, in others in terminal groups or infructescences or on specialized branches. Many species are known, but most of them are placed in one or other of the two genera Calamostachys and Palaeostachya. As originally defined, these two genera were clearly distinct, but, in the light of many newly discovered species, Andrews has questioned whether the distinction is now justified. In both genera there were whorls of peltate sporangiophores bearing four reflexed sporangia (Figs. 161 and 16J), alternating with whorls of bracts fused into a disc near their point of attachment. Whereas the sporangiophores were in vertical rows, the bracts in successive whorls alternated with one another. While the number of bracts in a whorl bore a definite relationship to the number of sporangiophores, the actual numbers varied from species to species and, sometimes, from individual to individual. Calamostachys Binneyana, a cone about 3-5 cm long and 7-5 mm wide, had six sporangiophores in each whorl and twelve bracts. C. magnaecrucis was more complicated, in having alternating vascular bundles in successive internodes within the cone and in having sporangiophores and bracts so numbered that, if ‘n’ were the number of vascular bundles, then the number of sporangiophores was 2n in each whorl and the number of bracts 3n; the number ‘n’ could be either seven or eight. Most species were homosporous, but some were definitely heterosporous. Thus, in C. casheana the megaspores were three or four times the size of the microspores, while in C. americana they were about twice the size.

Whereas the sporangiophores of Calamostachys stood out at right angles to the cone axis, those of Palaeostachya stood out at an angle of about 45 °, and in some species they appear to have been in the axil of the bract whorl below. P. vera had eight to ten sporangiophores in each whorl and twice as many bracts (according to an early description by Williamson and Scott, but according to Hickling, the same number). Perhaps the most interesting feature shown by this species is the course taken by the vascular bundle supplying the sporangiophore. As illustrated in Fig. 16I, it travelled up in the cortex of the cone axis to a point about midway between the bracts, and then turned downwards, before entering the stalk of the sporangiophore. To those morphologists who regard vascular systems as highly conservative, this implies that Palaeostachya must have evolved from some ancestral form in which the sporangiophore stood midway between the bract whorls, as in Calamostachys, and that during the ‘phyletic slide’ the vascular supply had lagged behind. Palaeostachya Andrewsii showed the same feature, but in P. decacnema the sporangiophore bundle took a direct course. One concludes, therefore, that this last species is more advanced than the others in this respect. In P. Andrewsii, the numbers of sporangiophores and bracts in a whorl were twelve and twenty-four respectively, while in P. decacnema they were usually ten and twenty.

The above brief review of the Calamitales brings out some interesting evolutionary trends, which are paralleled very closely in the Lepidodendrales. Thus, the production of increasing amounts of secondary wood was accompanied, in both groups, by a reduction of the primary wood, of which the centripetal metaxylem was the first to go, being replaced either by pith or by a central hollow. At the same time, there was a trend in the fertile regions from a ‘Selago condition’ to a compact cone, in which the sporangia were protected by overlapping sporophylls in one group and by bracts in the other. Then, having reached their zenith together, both groups became extinct at about the same time.


The only representatives of the Sphenopsida that are alive today belong to the single genus Equisetum and, of this, only some twenty-five species are known. Eleven of them occur in the British Isles, where they are known as ‘horse-tails’. The genus is distributed throughout the world with the exception of Australia and New Zealand, from which countries it is completely absent. All the species are herbaceous perennials, but there is an interesting range of growth habits, for some are evergreen, while others die back to the ground each year. Early statements that a limited amount of secondary thickening occurs are now discredited, for there is no evidence that a cambium is present in any species. Most species are, therefore, very limited in size; the largest species, E. giganteum, has stems up to 13 m long, but since they are only 2 cm thick the plant depends on the surrounding vegetation for its support. The largest British species, E. telmateia, sometimes attains a height of 2 m and is free-standing in sheltered localities, but most species are much smaller than this and are between 10 and 60 cm tall.

In all species there is a horizontal rhizome from which arise aerial stems that branch profusely in some species (e.g. Equisetum telmateia, E. arvense) or remain quite unbranched in others (e.g. E. hiemale). The leaves, in all species, are very small and are fused into a sheath, except for their extreme tips which form teeth round the margin of the sheath. They are usually without chlorophyll, photosynthesis being carried out entirely by the green stems. In the past, there have been discussions as to whether the small leaves of Equisetum represent a primitive or a derived condition, but, in the light of the fossil record, it is now clear that they have been reduced from larger dichotomous structures (i.e. that they are derived). The stems are ridged, each ridge corresponding to a leaf in the node above, and the ridges in successive internodes alternate with one another (as, of course, do the leaves in successive leaf-sheaths). There are, however, some departures from this regular alternation, as the number of leaves in a whorl diminishes from the base to the apex of the stem.

The sporangia are borne in a cone, which in some species (Equisetum arvense) terminates a special aerial axis that lacks chlorophyll, is unbranched and appears before the photosynthetic axes. In other species the fertile shoot is green and may subsequently give rise to vegetative branches lower down (e.g. E. limosum and E. pratense), after the cone has withered. In yet other species, most of the lateral branches may terminate in a cone (e.g. the Mexican species E. myriochaetum). This last arrangement is commonly regarded as the primitive condition, on the basis that it involves the least specialization, but it must be realized that real evidence for this view is lacking.

The internal anatomy of the stem of Equisetum presents an interesting association of xeromorphic and hydromorphic characters, together with a vascular system which is without parallel in the plant kingdom today, and whose correct morphological interpretation has long been the subject of controversy. The ridges in the stem are composed of sclerenchymatous cells, whose thick walls are so heavily sihcified as to blunt the edge of the razor when cutting sections. Stomata are restricted to the ‘valleys’ between the ridges and are deeply sunken into pits whose openings may be partly covered by a flange of cuticle. The walls separating the guard cells from their accessory cells bear peculiar comblike thickenings which are known elsewhere only in the leaves of Calamites. Beneath each of the valleys is a ‘vallecular canal’ and the central region of the internodes of aerial stems consists of a large space (but, in subterranean stems, the centre may be occupied by pith). At the nodes, there is a transverse diaphragm. Such an arrangement of air channels, together with a very reduced vascular tissue, are features normally found in water plants and contrast strikingly with the heavy cuticle, sunken stomata and reduced leaves.

The internodal vascular bundles lie beneath the ridges of the stem and are quite characteristic (Fig. 16Q). As in Calamites, the protoxylem is endarch and is replaced by a carinal canal (4), in which may be seen lignified rings which are all that remain after the dissolution of annular tracheids. To the outside of each carinal canal, and on the same radius, lies an area of phloem, flanked on either side by a lateral xylem area. This lateral xylem may contain further protoxylem tracheids with annular thickenings, but otherwise consists of metaxylem elements which may be tracheids with helical thickening, or with pits, or may even be true vessels. Two types of vessel element occur, one with simple perforation plates and the other with reticulate, but it must be emphasized that they are restricted to the internodes and that they seldom occur more than three in a row. They do not, therefore, form conducting channels of great length as do the vessels of flowering plants. In some species (e.g. Equisetum litorale) each internodal bundle is surrounded by its own separate endodermis, in others (e.g. E. palustre) there is a single endodermis running round the stem outside all the bundles, while in yet other species (e.g. E. sylvaticum, Fig. 16Q) there are two endodermes, one outside and the other inside all the bundles.

At the nodes (Fig. 16O) the vascular bundles (1) are connected by a continuous cylinder of xylem, from which the leaf traces (2) and branch traces (3) have their origin. Neither vallecular canals nor carinal canals are present in this region and there have been disagreements as to whether there is any protoxylem here either, but the most recent investigations confirm its presence as a constant feature. This disposes of the view, held by some, that the internodal bundles represent leaf traces extending down through the internode to the node below. An alternative view used to be held—that the vascular network represents a kind of dictyostele, in which the spaces between the internodal bundles represent leaf-gaps. However, this is unlikely, in view of the arrangement known to have existed in the earliest relatives of the genus, such as Asterocalamites, where there was no alternation at the nodes. Furthermore, this view overlooks the peculiar way in which the internodes of Equisetum are formed from an intercalary meristem. If analogies are to be sought with other pteridophytes, then it is not the internode, but the node, which should be compared. The vascular structure of the node can best be looked upon as a medullated protostele. The internodal spaces then appear as perforations, albeit of a peculiar (intercalary) origin.

Growth at the stem apex takes place as a result of the activity of a single tetrahedral apical cell, daughter cells being cut off in turn from each of its three cutting faces. Despite the spiral sequence of such daughter cells, subsequent growth results in a whorled arrangement and three daughter cells together give rise to all the tissues which make up a node and an internode. It is interesting that, in the first-formed stem of the young sporeling, there are three leaves in each whorl, but, nevertheless, it is stated that their initiation is in no way determined by the position of the cutting faces of the apical cell. Each leaf primordium grows from a single tetrahedral apical cell, and in the angle between the leaf sheath and the axis, but on radii between the leaves, lateral bud primordia arise, also with a single apical cell. The lateral bud primordia subsequently become buried by a fusion of the base of the leaf sheath with the axis, with the result that, when it grows, it has to burst through the leaf sheath, so giving the appearance of an endogenous origin. However, not all branch primordia do grow, for in species such as Equisetum hiemale, although present, they are inhibited from growing beyond the primordial stage, unless the main stem apex should be destroyed or damaged. Each branch primordium, besides bearing leaf primordia, also bears a root primordium which in aerial axes is also inhibited from growing further. In underground axes, however, they are not inhibited in this way. It is interesting to note that the roots which are apparently borne on a horizontal rhizome are, in fact, borne by the axillary buds hidden with its leaf sheaths, and not directly upon it.

The root grows from an apical cell with four cutting faces, the outermost of which gives the root cap. It may be triarch, tetrarch or diarch in its vascular structure and there is usually just a single central metaxylem element. The stele is surrounded by a pericycle, whose cells correspond exactly in number and radial position with those of the endodermis, since they are formed by a periclinal division in a ring of common mother cells. This has led to the statement that the root has a double endodermis, but this is incorrect, since the cells of the inner ring are without Casparian strips, and must be regarded as pericycle.

The cone (Fig. 16L) invariably terminates an axis, whether it be the main axis or a lateral one, and bears whorls of sporangiophores, without any bracts or other leaf-like appendages interposed, although there is a flange of tissue at the base of the cone called the ‘collar’. Each sporangiophore is a stalked peltate structure, bearing five to ten sporangia which, although having their origin on the outer surface, become carried round during growth into a reflexed position on the underside of the peltate head (Fig. 16M). Within the cone axis, the vascular system forms a very irregular anastomosing system, without discernible nodes and internodes, from which the sporangiophore traces depart without any regular association with the gaps. The sporangiophore trace branches within the peltate head and each branch terminates near a sporangium.

The sporangium has its origin in a single epidermal cell, which divides periclinally into an inner and an outer cell. The inner cell gives rise to sporogenous tissue. The outer cell gives rise to further blocks of sporogenous tissue and also to the wall of the sporangium. Adjacent cells may also add to the sporogenous tissue. The sporangium may therefore be described as eusporangiate in the widest sense and at maturity the sporangium is several cells thick. The innermost wall cells break down to form a tapetum, as also do some of the spore mother cells, and the ripe sporangium is two cells thick, of which the outer layer shows a characteristic spiral thickening.

Each spore, as it matures, has deposited round it four spathulate bands, which are free from the spore wall except at a common point of attachment (Fig. 16N). These are hygroscopic, coiling and uncoiling with changes in humidity, and are referred to as ‘elaters’, although what function they perform during dehiscence of the sporangium is not clear. McClean and Ivimey-Cook have shown that a distribution curve of the size of spores in Equisetum arvense is a bi-modal one, suggesting that a slight degree of heterospory exists, the large spores being some 25 per cent larger than the small ones. Furthermore, the smaller spores give rise to small male prothalli, whereas the large spores produce hermaphrodite ones. Whether this represents the early stage of the evolution of heterospory, or the last stages of a reversion to homospory, cannot be determined but, in any case, little outbreeding advantage is likely to accrue, because the elaters become so entangled as the spores are being released that they are usually distributed in groups. Reports of completely dioecious prothalli have been published, but these are probably based on observations made at a single moment in time. The prothalli of Equisetum are long lived, and extended observations would probably show that any one prothallus has archegonia alone for a time and then antheridia alone, as has indeed been demonstrated in some species. Differences in nutrition can also influence the behaviour of the prothalli for, under favourable conditions, only male prothalli result. Further work on this fascinating subject is clearly necessary.

The prothallus consists of a flat cushion of tissue, varying in size from 1 mm across to 3 cm in some tropical species. From the underside are produced abundant rhizoids, and from the upper side numerous irregular upright plates, or lobes, which are dark green and photo synthetic. Archegonia are formed in the tissue of the cushion between the aerial plates and have projecting necks of two or three tiers of cells in four rows. There may be a single neck canal cell or there may be two boot-shaped cells, lying side by side, as illustrated in Fig. 16R. There is also a ventral canal cell. The antheridia are sunken in the tissue of the basal cushion, but may also occur on the aerial lobes. They are massive and give rise to large numbers of antherozoids, which are spirally coiled and multiflagellate (Fig. 16P).

The first division of the zygote is in a plane more or less at right angles to the axis of the archegonium. No suspensor is formed and the embryo is exoscopic. Fig. 16S shows the spatial relationships of the stem apex (x), the first leaves (l), the root (r) and the foot (f), as described as long ago as 1878, but it is now becoming clear that the various parts of the embryo are not so constant in position and origin as was formerly thought.

There can be little doubt that the Equisetales are related to the Calamitales, but it is most unlikely that they represent their direct descendants. Remains of herbaceous plants resembling Equisetum are placed in the genus Equisetites, They are traceable right back through the Mesozoic to the Palaeozoic, where several species have been described from the Upper Carboniferous. The situation is thus closely comparable with Selaginella, whose herbaceous ancestors were living alongside the related arborescent Lepidodendrales in Carboniferous times.