BRYOPHYTA

Hepaticae (Liverworts).

The range of form and structure of both generations in the liverworts is so great that no one form can be taken as a satisfactory type. It will, however, be of use to preface the more general description by a brief account of a particular example, and we may take for this purpose a very common and easily recognized thalloid liverwort belonging to the Jungermanniales.

Pellia epiphylla (fig. 2) can be found at any season growing in large patches on the damp soil of woods, banks, &c. The broad flat thallus is green and may be a couple of inches long. It is sparingly branched, the branching being apparently dichotomous; the growing point is situated in a depression at the anterior end of each branch. The wing-like lateral portions of the thallus gradually thin out from the midrib; from the projecting lower surface of this numerous rhizoids spring. These are elongated superficial cells, and serve to fix the thallus to the soil and obtain water and salts from it. No leaf-like appendages are borne on the thallus, but short glandular hairs occur behind the apex. The plant is composed throughout of very similar living cells, the more superficial ones containing numerous chlorophyll grains, while starch is stored in the internal cells of the midrib. The cells contain a number of oil-bodies the function of which is imperfectly understood. The growth of the thallus proceeds by the regular segmentation of a single apical cell. The sexual organs are borne on the upper surface, and both antheridia and archegonia occur on the same branch (fig. 3, A). The antheridia (an) are scattered over the middle region of the thallus, and each is surrounded by a tubular upgrowth from the surface. The archegonia (ar) are developed in a group behind the apex, and the latter continues to grow for a time after their formation, so that they come to be seated in a depression of the upper surface. They are further protected by the growth of the hinder margin of the depression to form a scale-like involucre (in). Fertilization takes place about June, and the sporogonium is fully developed by the winter. The embryo developed from the fertilized ovum consists at first of a number of tiers of cells. Its terminal tier gives rise to the capsule, the first divisions in the four cells of the tier marking off the wall of the capsule from the cells destined to produce the spores. In fig. 4, C, which represents a longitudinal section of a young embryo of Pellia, these archesporial cells are shaded. The tiers below give rise to the seta and foot. The mature sporogonium (fig. 3, B) consists of the foot embedded in the tissue of the thallus, the seta, which remains short until just before the shedding of the spores, and the spherical capsule. It remains for long enclosed within the calyptra formed by the further development of the archegonial wall and surmounted by the neck of the archegonium. The calyptra is ultimately burst through, and in early spring the seta elongates rapidly, raising the dark-coloured capsule (fig. 2). In the young condition the wall of the capsule, which consists of two layers of cells, encloses a mass of similar cells developed from the archesporium. Some of these become spore-mother-cells and give rise by cell division to four spores, while others remain undivided and become the elaters. The latter are elongated spindle-shaped cells with thick brown spiral bands on the inside of their thin walls. They radiate out from a small plug of sterile cells projecting into the base of the capsule, and some are attached to this, while others lie free among the spores. The latter are large, and at first are unicellular; but in Pellia, which in this respect is exceptional, they commence their further development within the capsule, and thus consist of several cells when shed. [v.04 p.0702]The cells of the capsule wall have incomplete, brown, thickened rings on their walls, and the capsule opens by splitting into four valves, which bend away from one another, allowing the loose spores to be readily dispersed by the wind, assisted by the hygroscopic movements of the elaters. On falling upon damp soil the spores germinate, growing into a thallus, which gradually attains its full size and bears sexual organs.

While the general course of the life-history of all liverworts resembles that of Pellia, the three great groups into which they are divided differ from one another in the characters of both generations. Each group exhibits a series leading from more simple to more highly organized forms, and the differentiation has proceeded on distinct and to some extent divergent lines in the three groups. The Marchantiales are a series of thalloid forms, in which the structure of the thallus is specialized to enable them to live in more exposed situations. The lowest members of the series (Riccia) possess the simplest sporogonia known, consisting of a wall of one layer of cells enclosing the spores. In the higher forms a sterile foot and seta is present, and sterile cells or elaters occur with the spores. The lower members of the Jungermanniales are also thalloid, but the thallus never has the complicated structure characteristic of the Marchantiales, and progress is in the direction of the differentiation of the plant into stem and leaf. Indications of how this may have come about are afforded by the lower group of the Anacrogynous Jungermanniaceae, and throughout the Acrogynous Jungermanniacae the plant has well-marked stem and leaves. The sporogonium even in the simplest forms has a sterile foot, but in this series also the origin of elaters from sterile cells can be traced. The Anthocerotales are a small and very distinct group, in which the gametophyte is a thallus, while the sporogonium possesses a sterile columella and is capable of long-continued growth and spore production. The mode of development of the sporogonium presents important differences in the three series that may be briefly referred to here. In fig. 4 young sporogonia of a number of liverworts are shown in longitudinal section, and the archesporial cells from which the spores and elaters will arise are shaded. In Riccia (fig. 4, A) the whole mass of cells derived from the ovum forms a spherical capsule, the only sterile tissue being the single layer of peripheral cells forming the wall. In other Marchantiales (fig. 4, B) the lower half of the embryo separated by the first transverse wall (1, I) forms the sterile foot and seta, while in the upper half (ka) the peripheral layer forms the wall of the capsule, enclosing the archesporial cells from which spores and elaters arise. In the Jungermanniales (fig. 4, C, E, F) the embryo is formed of a number of tiers of cells, and the archesporium is defined by the first divisions parallel to the surface in the cells of one or more of the upper tiers; a number of tiers go to form the seta and foot, while the lowest segment (a) usually forms a small appendage of the latter. In the Anthocerotales (fig. 4, D) the lowest tiers form the foot, and the terminal tier the capsule. The first periclinal divisions in the cells of the terminal tier separate a central group of cells which form the sterile columella (col). The archesporium arises by the next divisions in the outer layer of cells, and thus extends over the summit of the columella. In none of the liverworts does the sporogonium develop by means of an apical cell, as is the rule in mosses.

Leaving details of form and structure to be considered under the several groups, some general features of the Hepaticae may be looked at here in relation to the conditions under which the plants live. The organization of the gametophyte stands in the closest relation to the factors of light and moisture in the environment. With hardly an exception the liverworts are dorsiventral, and usually one side is turned to the substratum and the other exposed to the light. In thalloid forms a thinner marginal expansion or a definite wing increasing the surface exposed to the light can be distinguished from a thicker midrib serving for storage and conduction. The leaves and stem of the foliose forms effect the same division of labour in another way. The relation of the plant to its water supply varies within the group. In the Marchantiales the chief supply is obtained from the soil by the rhizoids, and its loss in transpiration is regulated and controlled. In most liverworts, on the other hand, water is absorbed directly by the whole general surface, and the rhizoids are of subordinate importance. Many forms only succeed in a constantly humid atmosphere, while others sustain drying for a period, though their powers of assimilation and growth are suspended in the dry state. The cell-walls are capable of imbibing water rapidly, and their thickness stands in relation to this rather than to the prevention of loss of water from the plant. The large surface presented by the leafy forms facilitates the retention and absorption of water. The importance of prolonging the moistened condition as long as possible is further shown by special adaptations to retain water either between the appressed lobes of the leaves or in special pitcher-like sacs. In thalloid forms fimbriate or lobed margins or outgrowths from the surface lead to the same result. Sometimes adaptations to protect the plant during seasons of drought, such as the rolling up of the thallus in many xerophytic Marchantiales, can be recognized, but more often a prolonged dry season is survived in some resting state. The formation of subterranean tubers, which persist when the rest of the plant is killed by drought, is an interesting adaptation to this end, and is found in all three groups (e.g. in species of Riccia, Fossombronia and Anthoceros). No examples of total saprophytism or of parasitism are known, but two interesting cases of a symbiosis with other organisms which is probably a mutually beneficial one, though the nature of the physiological relation between the organisms is not clearly established, may be mentioned. Fungal hyphae occur in the rhizoids and in the cells of the lower region of the thallus of many liverworts, as in the endotrophic mycorhiza of higher plants. Colonies of Nostoc are constantly found in the Anthocerotaceae and in Blasia. In the latter they are protected by special concave scales, while in the Anthocerotaceae they occupy some of the mucilage slits between the cells of the lower surface of the thallus.

Other adaptations concern the protection of the sexual organs and sporogonia, and the retention of water in the neighbourhood of the archegonia to enable the spermatozoid to reach the ovum. In thalloid forms the sexual organs are often sunk in depressions, while in the foliose forms protection is afforded by the surrounding leaves. In addition special involucres around the archegonia have arisen independently in several series. The characters of the sporogonium have as their object the nutrition and effective distribution of the spores, and only exceptionally, as in the Anthocerotaceae, are concerned with independent assimilation. In most forms the capsule is raised above the general surface at the time of opening, usually by the rapid growth of the seta, but in the Marchantiaceae by the sporogonia being raised on a special archegoniophore. The elaters serve as lines of conduction of plastic material to the developing spores, and later usually assist in their dispersal. The spores, with few exceptions, are unicellular when shed, and may develop at once or after a resting period. In their germination a short filament of a few cells is usually developed, and the apical cell of the plant is established in the terminal cell. In other cases a small plate or mass of cells is formed. With one or two exceptions, however, this preliminary [v.04 p.0703]phase, which may be compared with the protonema of mosses, is of short duration.

The power of vegetative propagation is widely spread. When artificially divided small fragments of the gametophyte are found to be capable of growing into new individuals. Apart from the separation of branches by the decay of older portions, special gemmae are found in many species. In Aneura the contents of superficial cells, after becoming surrounded by a new wall and dividing, escape as bi-cellular gemmae. Usually the gemmae arise by the outgrowth of superficial cells, and become free by breaking away from their stalk. When separated they may be single cells or consist of two or numerous cells. In Blasia and Marchantia the gemmae are formed within tubular or cup-shaped receptacles, out of which they are forced by the swelling of mucilage secreted by special hairs.

Marchantiales.—The plants of this group are most abundant in warm sunny localities, and grow for the most part on soil or rocks often in exposed situations. Nine genera are represented in Britain. Targionia is found on exposed rocks, but the other forms are less strikingly xerophytic; Marchantia polymorpha and Lunularia spread largely by the gemmae formed in the special gemma-cups on the thallus, and occur commonly in greenhouses. The large thallus of Conocephalus covers stones by the waterside, while Dumortiera is a hygrophyte confined to damp and shady situations. Among the Ricciaceae, most of which grow on soil, Ricciocarpus and Riccia natans occur floating on still water. The dorsiventral thallus is constructed on the same plan throughout the group, and shows a lower region composed of cells containing little chlorophyll and an upper stratum specialized for assimilation and transpiration. The lower region usually forms a more or less clearly marked midrib, and consists of parenchymatous cells, some of which may contain oil-bodies or be differentiated as mucilage cells or sclerenchyma fibres. Behind the apex, which has a number of initial cells, a series of amphigastria or ventral scales is formed. These consist of a single layer of cells, and their terminal appendages often fold over the apex and protect it. Usually they stand in two rows, but sometimes accessory rows occur, and in Riccia only a single median row is present. The thallus bears two sorts of rhizoids, wider ones with smooth walls which grow directly down into the soil, and longer, narrower ones, with peg-like thickenings of the wall projecting into the cell-cavity. The peg-rhizoids, which are peculiar to the group, converge under shelter of the amphigastria to the midrib, beneath which they form a wick-like strand. Through this water is conducted by capillarity as well as in the cell cavities. The upper stratum of the thallus is constructed to regulate the giving off of the water thus absorbed. It consists of a series of air-chambers (fig. 6, B) formed by certain lines of the superficial cells growing up from the surface, and as the thallus increases in area continuing to divide so as to roof in the chamber. The layer forming the roof is called the "epidermis," and the small opening left leading into the chamber is bounded by a special ring of cells and forms the "stoma" or air-pore. In most species of Riccia the air-chambers are only narrow passages, but in the other Marchantiales they are more extended. In the simplest cases the sides and base of the chambers perform the work of assimilation (e.g. Corsinia). Usually the surface is extended by the development of partitions in the chambers (Reboulia), or by the growth from the floor of the chamber of short filaments of chlorophyllous cells (Targionia. Marchantia, fig. 6). The stomata may be simply surrounded by one or more series of narrower cells, or, as in the thallus of Marchantia and on the archegoniophores of other forms, may become barrel-shaped structures by the division of the ring of cells bounding the pore. In some cases the lowermost circle of cells can be approximated so as to close the pore. In Dumortiera the air-chambers are absent, their formation being only indicated at the apex.

The sexual organs are always situated on the morphologically upper surface of the thallus. In Riccia they are scattered singly and protected by the air-chamber layer. The scattered position of the antheridia is also found in some of the higher forms, but usually they are grouped on special antheridiophores which in Marchantia are stalked, disk-shaped branch-systems (fig. 5). The individual antheridia are sunk in depressions from which the spermatozoids are in some cases forcibly ejected. The archegonial groups in Corsinia are sunk in a depression of the upper surface, while in Targionia they are displaced to the lower side of the anterior end of a branch. In all the other forms they are borne on special archegoniophores which have the form of a disk-shaped head borne on a stalk. The archegoniophore may be an upgrowth from the dorsal surface of the thallus (e.g. Plagiochasma), or the apex of the branch may take part in its formation. When the disk, around which archegonia are developed at intervals, is simply raised on a stalk-like continuation of the branch, a single groove protecting a strand of peg-rhizoids is found on the ventral face of the stalk (Reboulia). In the highest forms (e.g. Marchantia) the archegoniophore corresponds to the repeatedly branched continuation of the thallus, and the archegonia arise in relation to the growing points which are displaced to the lower surface of the disk. In this case two grooves are found in the stalk. The archegonia are protected by being sunk in depressions of the disk or by a special two-lipped involucre. In Marchantia and Fimbriaria an additional investment termed in descriptive works the perianth, grows up around each fertilized archegonium (fig. 1, 3, d). The simple sporogonium found in the Ricciaceae (fig. 4, A) has been described above; as the spores develop, the wall of the spherical capsule is absorbed and the spores lie free in the calyptra, by the decay of which they are set free. In Corsinia the capsule has a well-developed foot, but the sterile cells found among the spore-mother-cells do not become elaters, but remain thin-walled and simply contribute to the nutrition of the spores. In all other forms elaters with spirally thickened walls are found. The seta is short, the capsule being usually raised upon the archegoniophore. Dehiscence takes place either by the upper portion of the capsule splitting into short teeth or falling away as a whole or in fragments as a sort of operculum. The spores on germination form a short germ-tube, in the terminal cell of which the apical cell is established, but the direction of growth of the young thallus is usually not in the same straight line as the germ-tube. The Marchantiales are divided into a number of groups which represent distinct lines of advance from forms like the Ricciaceae, but the details of their classification cannot be entered upon here. The general nature of the progression exhibited by the group as a whole will, however, be evident from the above account.

Jungermanniales.—This large series of liverworts, which presents great variety in the organization of the sexual generation, is divided into two main groups according to whether the formation of archegonia terminates the growth of the branch or does not utilize the apex. The latter condition is characteristic of the more primitive group of the Anacrogynous Jungermanniaceae, in which the branch continues its growth after the formation of archegonia so that they (and later the sporogonia) stand on the dorsal surface of the thallus or leafy plant. In the Acrogynous Jungermanniaceae the plant is throughout foliose, and the archegonia occupy the ends of the main shoot or of its branches. The antheridia are usually globular and long-stalked. The capsule opens by splitting into four halves.

Jungermanniaceae Anacrogynae.—The great range of form in the sexual plant is well illustrated by the nine genera of this group [v.04 p.0704]which occur in Britain. One thalloid form has already been described in Pellia (fig. 2). Sphaerocarpus, which occurs rarely in stubble fields, is in many respects one of the simplest of the liverworts. The small thallus bears the antheridia and archegonia, each of which is surrounded by a tubular involucre, on the upper surface of distinct individuals. The sporogonium has a small foot, but the sterile cells among the spores do not develop into elaters. The same is true of the capsule of Riella. The plants of this genus, none of the species of which are British, grow in shallow water rooted in the mud, and are unlike all other liverworts in appearance. The usually erect thallus has a broad wing-like outgrowth from the dorsal surface and two rows of rather large scales below. No provision for the opening of the capsule exists in either of these genera. In Aneura the form of the plant may be complicated by a division of labour between root-like, stem-like and assimilating branches of the thallus. The sexual organs are borne on short lateral branches, while in the related genus Metzgeria, which occurs on rocks and tree trunks, the small sexual branches spring from the lower surface of the midrib of the narrow thallus. In these two genera the elaters are attached to a sterile group of cells projecting into the upper end of the capsule, and on dehiscence remain connected with the tips of the valves. Pallavicinia and some related genera have a definite midrib and broad wings formed of one layer of cells, and are of interest owing to the presence of a special water-conducting strand in the midrib. This consists of elongated lignified cells with pitted walls. Blasia pusilla, which occurs commonly by ditches and streams, affords a transition to the foliose types. Its thallus (fig. 7) has thin marginal lobes of limited growth, which are comparable to the more definite leaves of other anacrogynous forms. The ventral surface bears flat scales in addition to the concave scales which, as mentioned above, are inhabited by Nostoc. This interesting liverwort produces two kinds of gemmae, and in the localities in which it grows is largely reproduced by their means. In Fossombronia, of which there are a number of British species, the plant consists of a flattened stem creeping on muddy soil and bearing two rows of large obliquely-placed leaves. The sexual organs are borne on the upper surface of the midrib, and the sporogonium is surrounded by a bell-shaped involucre which grows up after fertilization. Treubia, which grows on rotting wood in the mountain forests of Java, is similarly differentiated into stem and leaf, and is the largest liverwort known, reaching a length of thirty centimetres. Lastly Haplomitrium, a rare British genus, forms with the exotic Calobryum, an isolated group which is most naturally placed among the anacrogynous forms although the archegonia are in terminal groups. The erect branches bear three rows of leaves, and spring from a creeping axis from which root-like branches destitute of rhizoids extend into the substratum.

Jungermanniaceae Acrogynae.—The plant consists of leafy shoots, the origin of which can be understood in the light of the foliose forms described above. The great majority of existing liverworts belong to this group, the general plan of construction of which is throughout very similar. In Britain thirty-nine genera with numerous species are found. With few exceptions the stem grows by means of a pyramidal apical cell cutting off three rows of segments. Each segment gives rise to a leaf, but usually the leaves of the ventral row (amphigastria) are smaller and differently shaped from those of the two lateral rows; in a number of genera they are wanting altogether. Sometimes the leaves retain their transverse insertion on the stem, and the two lobes of which they consist are developed equally. More often they come to be obliquely inserted, the anterior edge of each leaf lying under or over the edge of the leaf in front. The two lobes are often unequally developed. In Scapania the upper lobe is the smaller, while in Radula, Poretta and the Lejeuneae this is the case with the lower lobe. The folding of one lobe against another assists in the retention of water. Pitcher-like structures have arisen in different ways in a number of genera, and are especially common in epiphytic forms (Frullania, Lepidolaena, Pleurozia). In some forms the leaves are finely divided, and along with the hair-like paraphyllia form a loose weft around the stem (Trichocolea). The rhizoids spring from the lower surface of the stem, and sometimes from the bases of the leaves. The branches arise below and by the side of the leaves.

The sexual organs may occur on the same or on distinct individuals. The antheridia are protected by leaves which are often modified in shape. The archegonia are borne at the apex of the main stem or of a lateral branch. A single archegonium may arise from the apical cell (Lejeunea); more commonly a number of others are formed from the surrounding segments. The leaves below the archegonial group are frequently modified in size and shape, but the chief protection is afforded by a tubular perianth, which corresponds to a coherent whorl of leaves and grows up independently of fertilization. The perianth serves also to enclose and protect the sporogonium during its development. In a number of forms belonging to different groups the end of the stem on which the sporogonium is borne grows downwards so as to form a hollow tubular sac enclosing the sporogonium; in other cases this marsupial sac is formed by the base of the sporogonium boring into the thickened end of the stem. The sac usually penetrates into the soil and bears rhizoids on its outer surface. Kantia, Calypogeia and Saccogyna are British forms, which have their sporogonia protected in this way. The sporogonium is very similar throughout the group (figs. 8, 9). At maturity the seta elongates rapidly, and the wall of the capsule splits more or less completely into four valves, allowing the elaters and spores to escape. In the Jubuloideae, which in other respects form a well-marked group, the seta is short and the elaters extend from the upper part of the capsule to the base; at dehiscence they remain fixed to the valves into which the capsule splits. The germinating spore usually forms a short filament, but in other cases a flat plate of cells growing by a two-sided apical cell is first formed (Radula, Lejeunea). In one or two tropical forms the pro-embryonic stage is prolonged, and leafy shoots only arise in connexion with the sexual organs. In Protocephalozia, which grows on bare earth in South America, this pro-embryo is filamentous, while in Lejeunea Metzgeriopsis, which grows on the leaves of living plants, it is a flat branched thallus closely applied to the substratum. Other cases of the plant being, with the exception of the sexual branches, apparently thalloid, are on the other hand to be explained as due to the reduction of the leaves and flattening of the stem of a shoot (Pteropsiella, Zoopsis).

The Acrogynous Jungermanniaceae fall into a number of natural groups, which cannot, however, be followed out here. They occur in very various situations, on the ground, on rocks and stones, on tree trunks, and, in the damp tropics, on leaves. Usually they form larger or smaller tufts of a green colour, but some forms have a reddish tint.

Anthocerotales.—This small and very natural group includes the three genera Anthoceros, Dendroceros and Notothylas, and stands in [v.04 p.0705]many respects in an isolated position among the Bryophyta. Three species of Anthoceros occur in Britain, growing on the damp soil of fields, ditches, &c. The dark green thallus has an ill-defined midrib, and is composed of parenchymatous cells. In each assimilating cell there is usually a single large chloroplast. The apical region, which has a single initial cell, is protected by mucilage secreted by the mucilage slits, which are small pit-like depressions between superficial cells of the lower surface. Mucilage is also often formed in intercellular spaces within the thallus. Colonies of Nostoc are constantly found living in some of the mucilage slits which then become enlarged. The sexual organs are scattered over the upper surface. The stalked globular antheridia are exceptional in being formed endogenously, and are situated in groups in special intercellular spaces. The superficial layer of cells bounding the cavity does not break down until the antheridia are nearly mature. Occasionally antheridia develop on the surface of shaded portions of the thallus. The necks of the archegonia hardly project above the general surface of the thallus. In structure and development they agree with other Hepaticae, though differences of detail exist. The young sporogonium is protected by a thick calyptra derived from the tissue of the thallus around the archegonium. The sporogonium consists of a large bulbous foot, the superficial cells of which grow out into processes, and a long capsule, which continues to grow for months by the activity of a zone of cells between it and the foot, and may attain the length of an inch and a half. The wall of the capsule is several layers of cells thick, and since the epidermis contains functional stomata and the underlying cells possess chlorophyll it is capable of assimilation. In the centre of the capsule is a strand of narrow elongated cells forming the columella, and between this and the wall spores mixed with elaters are formed from the dome-shaped archesporium, the origin of which has already been described (fig. 4, D). The capsule opens by splitting into two valves from the apex downwards, and the mature spores escape while others are developing in succession below. In Dendroceros, which grows as an epiphyte in the tropics, the thallus has a well-defined midrib and broad wings composed of a single layer of cells. The capsule is similar to that of Anthoceros, but has no stomata, and the elaters have spirally thickened walls. Some species of Anthoceros agree with it in these respects. Notothylas resembles Anthoceros in its thallus, but the sporogonium is much smaller. In some species, although the columella and archesporium arise in the usual way, both give rise to mingled spores and elaters, and no sterile columella is developed.

Musci (Mosses).

Though the number of species of mosses is far greater than of liverworts, the group offers much less diversity of form. The sexual generation is always a leafy plant, which is not developed directly from the spore but is borne on a well-marked and usually filamentous protonema. The general course of the life-history and the main features of form and structure will be best understood by a brief account of a particular example.

Funaria hygrometrica is a moss of very common occurrence even in towns on the soil of paths, at the foot of walls and in similar places. The small plants grow closely crowded in tufts, and consist of short leafy shoots attached to the soil by numerous fine rhizoids. The latter, in contrast to the rhizoids of liverworts, are composed of rows of elongated cells and are branched. The leaves are simple, and except for the midrib are only one layer of cells thick. The structure of the stem though simple is more complicated than in any liverwort. The superficial cells are thick-walled, and there is a central strand of narrow cells forming a water-conducting tissue. The small strand of elongated cells in the midrib of the leaf runs down into the stem, but is not usually connected with the central strand. The sexual organs are developed in groups at the apices, the antheridial group usually terminating the main axis while the archegonia are borne on a lateral branch. The brown tint of the hair-like paraphyses mixed with antheridia (fig. 15) makes the male branch conspicuous, while the archegonia have to be carefully looked for enclosed by the surrounding leaves (fig. 16, B). The sporogonium developed from the fertilized ovum grows by means of a two-sided apical cell (fig. 16 A), and is at first of uniform thickness. After a time the upper region increases in diameter and forms the capsule, while the lower portion forms the long seta and the foot which is embedded in the end of the stem. With the growth of the sporogonium the archegonial wall, which for a time kept pace with it, is broken through, the larger upper part terminated by the neck being carried up on the capsule as the calyptra, while the basal portion remains as a tubular sheath round the lower end of the seta (cf. figs. 16, C, and fig. 11, A, B). The seta widens out at the base of the capsule into a region known as the apophysis. The peripheral cells of the seta are thick-walled, and it has a central strand of elongated conducting cells. In the epidermis of the apophysis functional stomata, similar to those of the higher plants, are present and, since cells containing chlorophyll are present below the superficial layers of the apophysis and capsule, the sporogonium is capable of independent assimilation. The construction of the capsule will be best understood from the median longitudinal section (fig. 11, C). The central region extending between the apophysis and the operculum is composed of sterile tissue and forms the columella (c). Immediately around this is the layer of cells from which the spores will be developed (s), and the layers of cells on either side of this form the walls of the spore-sac, which will contain the spores. Between the wall of the capsule, which is composed of several layers of cells, and the spore-sac is a wide intercellular space (h) bridged across by trabeculae consisting of rows of chlorophyll-containing cells. At the junction of the operculum (d) with the rest of the capsule is a circle of cells forming the annulus (a), by help of which the operculum is detached at maturity as a small lid. Its removal does not, however, leave the mouth of the capsule wide open, for around the margin are two circles of pointed teeth forming the peristome. These are the thickened cell-walls of a definite layer of cells (p), and appear [v.04 p.0706]as separate teeth owing to the breaking down of the unthickened cell-walls. The numerous spores which have been developed in the spore sac can thus only escape from the pendulous capsule through narrow slits between the teeth, and these are closed in damp air. The unicellular spores when supplied with moisture germinate (fig. 12) and give rise to the sexual generation. A filamentous protonema is first developed, some of the branches of which are exposed to the light and contain abundant chlorophyll, while others penetrate the substratum as brown or colourless rhizoids. The moss-plants arise from single projecting cells, and numerous plants may spring from the protonema developed from a single spore.

The majority of the mosses belong to the same great group as Funaria, the Bryales. The other two subdivisions of the Musci are each represented by a single genus. In the Andreaeales the columella does not extend to the upper end of the capsule, and the latter opens by a number of lateral slits. The Sphagnales also have a dome-shaped spore-sac continued over the columella, and, though their capsule opens by an operculum, they differ widely from other mosses in the development of the sporogonium as well as in the characters of the sexual generation. The three groups are described separately below, but some more general features of the mosses may be considered here.

On the whole mosses grow in drier situations than the liverworts, and the arrangements they present for the conduction of water in the plant are also more complete and suggest in some cases comparisons with the higher plants. In spite of this, however, they are in great part dependent on the absorption of water through the general surface of the shoot, and the power of rapid imbibition possessed by their cell-walls, the crowded position of the small leaves on the stem, and special adaptations for the retention of water on the surface, have the same significance as in the foliose liverworts. The different appearance of exposed mosses in dry weather and after a shower illustrates this relation to the water supply. The protonema is always a well-marked stage in the life-history. Not only does a moss-plant never arise directly from the spore, but in all cases of vegetative reproduction, apart from the separation of branches by decay of older regions of the plant, a protonema is found. Usually the protonema is filamentous and ceases to be evident after the plants have developed. But in some small mosses (e.g. Ephemerum) it plays the chief part in assimilation and lives on from year to year. In Sphagnum, Andreaea and some genera of the Bryales the protonema or some of its branches have the form of flat plates or masses of cells. The formation of the moss-plant on the protonema is always from a single cell and is similar in all mosses. The first three walls in this cell intersect one another, and define the three-sided pyramidal apical cell by means of which the shoot continues to grow. In Fissidens and a few other mosses the apical cell is two-sided. The leaves formed by the successive segments gradually attain their normal size and structure. Each segment of the initial cell gives rise to a leaf and a portion of the stem; the branches arise from the lower portion of a segment and stand immediately below a leaf. The leaves may form three vertical rows, but usually their arrangement, owing to the direction of the segment walls at the apex, becomes more complicated. Their growth proceeds by means of a two-sided apical cell, and the midrib does not become more than one cell thick until later. In addition to the leaves the stem often bears hair-like structures of different kinds, some of which correspond to modified branches of protonema. The branched filamentous rhizoids which spring from the lower region of the stem also correspond to protonemal branches. The structure of both stem and leaf reaches a high grade of organization in some mosses. Not only are thick-walled sclerenchymatous cells developed to give rigidity to the periphery of the stem and the midrib of the leaf, but in many cases a special water-conducting tissue, consisting of elongated cells, the end walls of which are thin and oblique, forms a definite central strand in the stem. In the forms in which it is most highly developed (Polytrichaceae) this tissue, which is comparable with the xylem of higher plants, is surrounded by a zone of tissue physiologically comparable to phloem, and in the rhizome may be limited by an endodermis. The conducting strands in the leaves show the same tissues as in the central strand of the stem, and in the Polytrichaceae and some other mosses are in continuity with it. The independent origin of this conducting system is of great interest for comparison with the vascular system of the sporophyte of the higher plants.

The sexual organs, with the exception of the antheridia of Sphagnum, are borne at the apices of the main shoot or of branches. Their general similarity to the mature antheridia and archegonia of liverworts and the main difference in their development have been referred to. The antheridia open by means of a cap cell or groups of cells with mucilaginous contents. The details of construction of the sporogonium are referred to below. In all cases (except Archidium) a columella is present, and all the cells derived from the archesporium produce spores, no elaters being formed. In a few cases the germination of the spore commences within the capsule. The development of the sporogonium proceeds in all cases (except in Sphagnum) by means of an apical cell cutting off two rows of segments. The first periclinal division in the region forming the capsule separates an inner group of cells (the endothecium) form the peripheral layer (amphithecium). In Sphagnum, as in Anthoceros, the archesporium is derived from the amphithecium; in all other mosses it is the outermost layer of the endothecium.

Vegetative propagation is widely spread in the mosses, and, as mentioned above, a protonema is always formed in the development of the new plant. The social growth of the plants characteristic of many mosses is a result of the formation of numerous plants on the original protonema and on developments from the rhizoids. Besides this, gemmae may be formed on the protonema, on the leaves or at the apex, and some mosses have specialized shoots for their better protection or distribution. Thus in Georgia the stalked, multicellular gemmae are borne at the ends of shoots surrounded by a rosette of larger leaves, and in Aulacomnium androgynum they are raised on an elongated leafless region of the shoot. In other cases detached leaves or shoots may give rise to new plants, and when a moss is artificially divided almost any fragment may serve for reproduction.

Even in those rare cases in which the sexual generation can be developed without the intervention of spore production from the tissues of the sporogonium, a protonema is formed from cut pieces of the seta or in some cases from intact sporogonia still attached to the plant. This phenomenon of apospory was first discovered in mosses, but is now also known in a number of ferns (see Pteridophyta).

Sphagnales.—The single genus Sphagnum occupies a very distinct and isolated position among mosses. The numerous species, which are familiar as the bog-mosses, are so similar that minute structural characters have to be relied on in their identification. The plants occur in large patches of a pale green or reddish colour on moors, and, when filling up small lakes or pools, may attain a length of some feet. Their growth has played a large part in the formation of peat. The species are distributed in temperate and arctic climates, but in the tropics only occur at high levels. The protonema forms a flat, lobed, thalloid structure attached to the soil by rhizoids, and the plants arise from marginal cells. The main shoot bears numerous branches which appear to stand in whorls; some of them bend down and become applied to the surface of the main axis. The structure of the stem and leaves is peculiar. The former shows on cross-section a thin-walled central tissue surrounded by a zone of thick-walled cells. Outside this come one to five layers of large clear cells, which when mature are dead and empty; their walls are strengthened with a spiral thickening and perforated with round pores. They serve to absorb and conduct water by capillarity. The leaves have no midrib and similar empty cells occur regularly among the narrow chlorophyll-containing cells, which thus appear as a green network. The antheridia are globular and have long stalks. They stand by the side of leaves of special club-shaped branches. The archegonial groups occupy the apices of short branches (fig. 13, A.). The mature sporogonium consists of a wide foot separated by a constriction from the globular capsule (B). There is no distinct seta, but the capsule is raised on a leafless outgrowth of the end of the branch called a pseudopodium (C, qs). The capsule, the wall of which bears rudimentary stomata, has a small operculum but no peristome. There is a short, wide columella, over which the dome-shaped spore-sac extends, and no air-space is present between the spore-sac and the wall. In the embryo a number of tiers of cells are first formed. The lower tiers [v.04 p.0707]form the foot, while in the upper part the first divisions mark off the columella, around which the archesporium, derived from the amphithecium, extends. The sporogonium when nearly mature bursts the calyptra irregularly. The capsule opens explosively in dry weather, the operculum and spores being thrown to a distance. The spore on germination forms a short filament which soon broadens out into the thalloid protonema. Some twelve species of Sphagnum are found in Britain.

Andreaeales.—The species of the single genus Andreaea (fig. 14) are small, dark-coloured mosses growing for the most part in tufts on bare rocks in alpine and arctic regions. Four species occur on alpine rocks in Britain. The spore on germination gives rise to a small mass of cells from which one or more short filaments grow. The filament soon broadens into a ribbon-shaped thallus, several cells thick, which is closely applied to the rock. Erect branches may arise from the protonema, and gemmae may be developed on it. The stem of the plant, which arises in the usual way, has no conducting strand and the leaves may or may not have midribs. The leaf grows by a dome-shaped instead of by the usual two-sided initial cell. The antheridia are long-stalked. The upper portion of the archegonial wall is carried up as a calyptra on the sporogonium, which, as in Sphagnum, has no seta and is raised on a pseudopodium. The development of the sporogonium proceeds as in the Bryales, but the dome-shaped archesporium extends over the summit of the columella and an air-space is wanting. The capsule does not open by an operculum but by four or six longitudinal slits, which do not reach either the base or apex. In one exotic species the splits occur only at the upper part of the capsule, and the terminal cap breaks away. This isolated example thus appears to approach the Bryales in its mode of dehiscence.

Bryales.—In contrast to the preceding two this group includes a very large number of genera and species. Thus even in Britain between five and six hundred species belonging to more than one hundred genera are found. They occur in the most varied situations, on soil, on rocks and trees, and, in a few instances (Fontinalis), in water. Although exhibiting a wide range in size and in the structural complexity of both generations, they all conform to a general type, so that Funaria, described above, will serve as a fair example of the group. The protonema is usually filamentous, and in some of the simplest forms is long-lived, while the small plants borne on it serve mainly to protect the sexual organs and sporogonia. This is the case in Ephemerum, which grows on the damp soil of clayey fields, and the plants are even more simply constructed in Buxbaumia, which occurs on soil rich in humus and is possibly partially saprophytic. In this moss the filamentous protonema is capable of assimilation, but the leaves of the small plants are destitute of chlorophyll, so that they are dependent on the protonema. The male plant has no definite stem, and consists of a single concave leaf protecting the antheridium. The female plant is rather more highly organized, consisting of a short stem bearing a few leaves around the group of archegonia. The sporogonium is of large size and highly organized, though it presents peculiar features in the peristome. Buxbaumia has been regarded by Goebel as representing a stage which other mosses have passed, and has been described by him as the simplest type of moss. In Ephemerum also we may probably regard the relation of the small plants to the protonema as a primitive one. On the other hand, in the case of Ephemeropsis, which grows on the leaves of living plants in Java, the high organization of the sporogonium makes it probable that the persistent protonema is an adaptation to the peculiar conditions of life. A highly developed protonema provided with leaf-like assimilating organs is found in Georgia, Diphyscium and Oedipodium, all of which show peculiarities in the sporogonium as well. The cells of the protonema of Schistostega, which lives in the shade of caves, are so constructed as to concentrate the feeble available light on the chloroplasts.

We may perhaps regard the persistent protonema bearing small leafy plants as a primitive condition, and look upon those larger plants which remain unbranched and bear the sexual organs at the apex (e.g. Schistostega) as representing the next stage. From this condition different lines of specialization in the form and structure of the plant can be recognized. A large number of mosses stand at about the same grade as Funaria, in that the plants are small, sparingly branched, usually radial, and do not show a very highly differentiated internal structure. In others the form of the plant becomes more complex by copious branching and the differentiation of shoots of different orders. In these cases the shoot system is often more or less dorsiventral, and the sexual organs are borne on short lateral branches (e.g. Thuidium tamariscinum). The Polytrichaceae, on the other hand, show a specialization in structure rather than in form. The high organization of their conducting system has been referred to above, but though many species are able to exist in relatively dry situations, the plants are still dependent on the absorption of water by the general surface. The parallel lamellae of assimilating cells which grow from the upper surface of the leaf in these and some other mosses probably serve to retain water in the neighbourhood of the assimilating cells and so prolong their activity. As common adaptive features in the leaves the occurrence of papillae or outgrowths of the cell-walls to retain water, and the white hairlike leaf tips, which assist in protecting the young parts at the apex of many xerophytic mosses, may be mentioned. The leaves of Leucobryum, which occurs in pale green tufts in shaded woods, show a parallel adaptation to that found in Sphagnum. They are several cells thick, and the small assimilating cells lie between two layers of empty water-storage cells, the walls of which are perforated by pores.

With the possible exception of Archidium, the sporogonium is throughout the Bryales constructed on one plan. Archidium is a small moss occurring occasionally on the soil of wet fields. The protonema is not persistent, and the plants are well developed, resembling those of Pleuridium. The sporogonium has a small foot and practically no seta, and differs in the development and structure of its capsule from all other mosses. The spores are derived from the endothecium, but no distinction of a sterile columella and an archesporium is established in this, a variable number of its cells becoming spore-mother-cells while the rest serve to nourish the spores. The layer of cells immediately around the endothecium becomes the spore-sac, and an air-space forms between this and the wall of the capsule. The very large, thin-walled spores escape on the decay of the capsule, which ruptures the archegonial wall irregularly. On account of the absence of a columella Archidium is sometimes placed in a distinct group, but since its peculiarities have possibly arisen by reduction it seems at present best retained among the Bryales. In all other Bryales there is a definite columella extending from the base to the apex of the capsule, the archesporium is derived from the outermost layer of cells of the endothecium, and an air space is formed between the spore-sac and the wall. In the Polytrichaceae another air space separates the spore-sac from the columella. There is great variety in the length of the seta, which is sometimes practically absent. The apophysis, which may be a more or less distinct region, usually bears stomata and is the main organ of assimilation. In the Splachnaceae it is expanded for this purpose, while in Oedipodium it constitutes most of the long pale stalk which supports the capsule. A distinct operculum is usually detached by the help of the annulus, and its removal may leave the mouth of the capsule widely open. More usually there is a peristome, consisting of one or two series of teeth, which serves to narrow the opening and in various ways to ensure the gradual shedding of the spores in dry weather. In most mosses the teeth are portions of thickened cell-walls but in the Polytrichaceae they are formed of a number of sclerenchymatous cells. In Polytrichum a membranous epiphragm stretches across the wide mouth of the capsule between the tips of the short peristome teeth, and closes the opening except for the interspaces of the peristome.

In a number of forms, which were formerly grouped together, the capsule does not open to liberate the spores. These cleistocarpous forms are now recognized as related to various natural groups, in which the majority of the species possess an operculum. In such forms as Phascum the columella persists, and the only peculiarity is in the absence of arrangements for dehiscence. In Ephemerum [v.04 p.0708](and the closely related Nanomitrium which has a small operculum) the columella becomes absorbed during the development of the spores. Stomata are present on the wall of the small capsule. Such facts as these suggest that in many cases the cleistocarpous condition is the result of reduction rather than primitive, and that possibly the same holds for Archidium.

The former subdivision of the Bryales into Musci Cleistocarpi and Musci Stegocarpi according to the absence or presence of an operculum is thus clearly artificial. The same holds even more obviously for the grouping of the stegocarpous forms into those in which the archegonial group terminates a main axis (acrocarpi) and those in which it is borne on a more or less developed lateral branch (pleurocarpi). Modern classifications of the Bryales depend mainly on the construction of the peristome.

It remains to be considered to what extent the several natural groups of plants classed together in the Bryophyta can be placed in a phylogenetic relation to one another. Practically no help is afforded by palaeobotany, and only the comparison of existing forms can be depended on. The indications of probable lines of evolution are clearest in the Hepaticae. The Marchantiales form an obviously natural evolutionary group, and the same is probably true of the Jungermanniales, although in neither case can the partial lines of progression within the main groups be said to be quite clear. Such a form as Sphaerocarpus, which has features in common with the lower Marchantiales, enables us to form an idea of the divergence of the two groups from a common ancestry. The Anthocerotales, on the other hand, stand in an isolated position, and recent researches have served to emphasize this rather than to confirm the relationship with the Jungermanniales suggested by Leitgeb. The indications of a serial progression are not so clear in the mosses, but the majority of the forms may be regarded as forming a great phylogenetic group in the evolution of which the elaboration of the moss-plant has proceeded until the protonema appears as a mere preliminary stage to the formation of the plants. Parallel with the evolution of the gametophyte in form and structure, a progression can be traced in the sporogonium, although the simplest sporogonia available for study may owe much of their simplicity to reduction. The Andreaeales may perhaps be looked on as a divergent primitive branch of the same stock. On the other hand, the Sphagnales show such considerable and important differences from the rest of the mosses, that like the Anthocerotales among the liverworts, they may be regarded as a group, the relationship of which to the main stem is at least problematical. Between the Hepaticae, Anthocerotales, Sphagnales and Musci, there are no connecting forms known, and it must be left as an open question whether the Bryophyta are a monophyletic or polyphyletic group.

The question of the relationship of the Bryophyta on the one hand to the Thallophyta and on the other to the Pteridophyta lies even more in the region of speculation, on slender grounds without much hope of decisive evidence. In a general sense we may regard the Bryophyta as derived from an algal ancestry, without being able to suggest the nature of the ancestral forms or the geological period at which they arose. Recent researches on those Algae such as Coleochaete which appeared to afford a close comparison in their alternation of generations with Riccia, have shown that the body resulting from the segmentation of the fertilized ovum is not so strictly comparable in the two cases as had been supposed. The series of increasingly complex sporogonia among Bryophytes appears to be most naturally explained on an hypothesis of progressive sterilization of sporogenous tissue, such as has been advanced by Bower. On the other hand there are not wanting indications of reduction in the Bryophyte sporogonium which make an alternative view of its origin at least possible. With regard to the relationship of the Bryophyta and Pteridophyta the article on the latter group should be consulted. It will be sufficient to say in conclusion that while the alternating generations in the two groups are strictly comparable, no evidence of actual relationship is yet forthcoming.

For further information consult: Campbell, Mosses and Ferns (London, 1906); Engler and Prantl, Die naturlichen Pflanzenfamilien, Teil i. Abt. 3 (Leipzig, 1893-1907); Goebel, Organography of Plants (Oxford, 1905). Full references to the literature of the subject will be found in these works. For the identification of the British species of liverworts and mosses the following recent works will be of use: Pearson, The Hepaticae of the British Isles (London, 1902); Dixon and Jameson, The Student's Handbook of British Mosses (London, 1896); Braithwaite, British Moss Flora (London, 1887-1905).

(W. H. L.)