orbicular in shape, are either close together or somewhat spaced out, and those near the top of the stem are larger than the lower leaves. The firm, stout perigynium is 7 mm long. C. gibbsiae is the smallest of the species. The stems are from 5 to 14 mm tall, of uniform width throughout, and slender as compared with any of the preceding species. The leaves are close together, deltoid or rhombate in shape, with an acute apex, and are translucent green in colour when fresh, changing to a pale yellow or yellowish-green when dried. The fleshy perigynium is up to 8 mm long. Dehiscence of the capsule in this, as in the two preceding species, is lengthwise into four valves which remain united at the apex and the base. Frequent branching of the stem is mentioned by Stephani but is not always evident. When it does occur it is no more frequent than that shown in Goebel's figure of C. blumei (Goebel, 1891).

C. gibbsiae resembles some forms of C. rotundifolium fairly closely but differs in the smaller size of the plant, the more translucent leaves of deltoid or rhombate shape, and the fact that the stem of the archegoniophore does not widen above.

Distribution of Calobryum gibbsiae in New Zealand.

Calobryum gibbsiae Steph. was first collected by Gibbs in November, 1907, at Wairongomai Mine, Te Aroha, on clay banks in forest at 1,000ft (Gibbs, 1911). The description was drawn up by Stephani (Stephani, 1924) from female plants which Gibbs had forwarded to him. The locality as given by Stephani, namely Wairongomai Main at 2,000ft, is probably less accurate than that recorded by Gibbs.

The present writer made visits to Wairongomai in October, 1955, and in August, 1956, in an unsuccessful attempt to collect plants from the type locality. In the early years of the twentieth century at a time of extensive gold-mining activity at Te Aroha Mountain, there was constructed on the forested hillsides a series of tram tracks, contoured at different altitudes and connected by steep cableways up to 300 m in height. At the same time wooden flumes several km in length were built to carry water. These activities would involve the formation of numerous newly-cut clay banks, on some of which no doubt the Calobryum established itself. By now, however, the flumes, although still traceable, are broken and decayed, the tram tracks and cableways are recognizable only by the rusted rails, and the banks are overgrown by ferns and shrubs. It seems unlikely that Calobryum still grows in the region.

K. W. Allison collected Calobryum in 1934 under manuka on the steep bank of a washout between Taupo and Rangitaiki, and in December, 1937, at Ngahunga crossing, Rangitaiki River, above Murupara, on a steep, damp bank in semi-shade. However, these areas have since been planted out in conifer forest, and in October, 1955, no trace of Calobyrum could be found.

In January, 1939, G. O. K. Sainsbury collected Calobryum on a damp bank, Whakapapanui Stream, Mt. Ruapehu, and in January, 1946, K. W. Allison found plants growing in a sheltered ledge of steeply-sloping ground in open country on top of Maungatua Range, south of Dunedin, at an altitude of about 800 m. This last record is noteworthy in that the Calobryum occurred above the forest level and about 630 km south of previously recorded stations. In 1954 G. P. Harris found Calobryum growing in the Dunedin area at a much lower altitude.

It would appear that Calobryum occurs throughout New Zealand wherever conditions are suitable, but, since it is a pioneer on steep, damp clay banks which are liable to slipping, it would always be local in occurrence. It is rapidly disappearing as a result of man's activities in stabilizing slips and in opening roadside banks to sunlight, wind, and dust.

Source of Material.

The plants used for the present paper were collected in Tongariro National Park at an altitude of 1,075 m on almost perpendicular banks of andesitic clay. This particular day has the texture of a sand rather than a clay, but is retentive of water and does not dry out readily. Bacteria are very abundant in the uppermost zone. The surface of the banks is liable to break away at times, but is protected from snow and heavy rain by the projecting, overhanging upper edge and is shaded by the surrounding Nothofagus cliffortioides (Hook. f.) Oerst. forest. The Calobryum either occurs as a monospecific stand or in more deeply shaded places grows intermixed throughout the mat of other liverworts and mosses which form a low-growing cover to the clay surface. Schistochila splachnophylla (Tayl.) Steph. plays a prominent part in the composition of this mat and, especially when the branches are forced into an erect position by crowding, it bears a superficial resemblance to Calobryum. However, Schistochila may be distinguished even in the vegetative state by the prostrate habit of its leafy stems, its two-ranked, imbricate leaves and tufts of brown rhizoids. Other plants present in the mat are a species of Lepidozia, the moss Ditrichum flexifolium (Hook.) Hampe, and sometimes the filmy fern Hymenophyllum multifidum (Forst.) Swartz.

Description of the Plant

In many respects the habit is unlike that of a typical liverwort (Figs. 1, 2). The main stem is at first prostrate, growing either interwoven in a felt of other liverworts or in the brown leaf-litter derived from these, or sometimes under the soil. This part of the stem is white in colour, up to 0·6 mm in diameter, and on the outside has a colourless mucilaginous sheath as thick again as the stem itself. Neither leaves nor rhizoids are present. When the subterranean stem is 1 to 2 cm long the tip turns erect and gives rise to a translucent, pale green, leafy stern up to 14 mm in height. Lateral shoots grow out from the base of the erect stem. Some of these remain as short white shoots 1 to 2 mm long, growing either horizontally or downwards and anchoring the plant to the substratum. But one or more grow out strongly and these after a period of horizontal growth turn erect as new leafy shoots.

The appearance of the plant differs according to the number and position of the new leafy shoots. As a rule if the plant is growing amongst other liverworts only one strong lateral shoot develops from the base of each erect shoot and eventually itself turns erect. But when the underground stem is actually beneath the soil on banks liable to drying out several strong lateral shoots develop resulting in an interwoven tangle of branches.

The lateral shoots may grow out from higher up the erect stem, even just beneath the tip if the surrounding liverworts are growing rapidly and tending to overtop the Calobryum; such lateral shoots tend to develop immediately into erect leafy shoots which crowd one another. Even detached pieces of the plant readily give rise to lateral shoots.

The leaves on the erect stem are transversely inserted and spirally arranged in three vertical rows. One row is centrally placed on the morphologically lower side of the stem and corresponds in position to the row of underleaves of the Jungermannineae; the other two rows are lateral rows lying to either side of the midline on the morphologically upper side of the stem. There is no difference in size of leaf as between the rows, but the leaves on the basal portion of the stem, when overtopped and shaded by the growth of surrounding liverworts, are smaller than the exposed upper leaves. The larger, vegetative leaves are 1 mm long and 1 mm wide at the broadest part. In shape they are deltoid or rhombate, acute, and entire. In moist air at intervals round the margin up to 6 club-shaped, colourless cells project singly from the surface, but in dry air these soon wither off.

There are distinct antheridial and archegonial plants which grow either intertwined or in separate clumps. (Figs. 1, 2.) The antheridial plant is the more slender of the two, with erect stems no more than 6 mm high and horizontal stems 0·3 to 0·4 mm in diameter. Also the leaves tend to be more widely spaced than on female plants. The male shoots continue to grow for some time, giving rise to successive series of leaves and antheridia. The appearance differs with the stage of development. In September (spring), when new growth begins, the young antheridia are present. They lie in a mucilage-filled cavity at the tips of the erect branches, completely overarched and surrounded by the three youngest leaves. By the end of October these leaves, rhombate in shape with a truncate base and slightly toothed margin, now fully expanded to a maximum breadth of 1·5 mm, lie in a horizontal plane around the flattened disc-like apex of the stem. The disc itself is covered with 30 or more club-shaped antheridia of a cinnamon colour interspersed with mucilage hairs. At this stage the male plants are most easily recognized, especially after a dry period when many unopened antheridia are present. Already the disc shows a green, knob-like centre where the 3 leaves of the new series are beginning to enlarge. Several successive series of leaves and antheridia may be formed and at a later stage old, empty antheridia of a cinnamon colour may still be seen persisting amongst leaves farther down the stem. However, with continued upward growth, the male shoots become more and more exposed to desiccation and during dry periods in summer the tops die and turn brown. From February until early September at National Park there is no evidence on the soil surface of their existence. However, the subterranean part of the plant persists and in the next spring lateral shoots grow up above the soil surface.

The archegoniophores, like the antheridiophores, have continued growth for some time provided fertilization does not occur. In this case 3 or 4 series of leaves are formed before the next group of archegonia. Young archegonia were forming in September. The tip of the stem forms a flat or dome-shaped disc on which up to 20 archegonia develop. At first these lie in a mucilage-filled cavity enclosed and overarched by the surrounding leaves, but later the leaves enlarge and flatten. These leaves are transversely inserted in 3 rows on the stem but are much larger than the vegetative leaves. They attain a maximum width of 4 mm, are rhombate to almost orbicular in outline, with the coarsely and irregularly dentate margins slightly recurved. If fertilization fails to take place, the archegoniophores remain in position for several months, the disc covered with old bronze-coloured archegonia and surrounded by the large dentate leaves. The whole head is up to 7 mm in diameter at the top. Eventually the archegoniophores either turn brown and die during dry periods or undergo renewed terminal growth. Lateral shoots from the base of the erect stem continue the horizontal growth and form new erect shoots.

Ripe sporogonia are found in November at the tips of some of the archegoniophores (Fig. 5). The base of the sporogonium is surrounded by a colourless, fleshy pengynium 5 to 8 mm long. The slender, colourless seta stands erect to a height of 22 mm beyond the top of the perigynium and carries at its tip the dark brown capsule 3 to 4 mm long and 0.5 mm wide. Usually only one sporogonium develops on an archegoniophore, but in one instance two had developed, each enclosed in its own pengynium.

Anatomy of the Subterranean Stem

The subterranean stem in transverse section is either circular in outline or flattened from above into a broad oval (Fig. 4) and, as noted above, is ensheathed in mucilage. Mucilage hairs may project from the exterior. Where the surface cells lie close to the substratum, the outer walls are convex and arch outwards in a papillate shape. The surface cells and those for a depth of 6 to 7 cells below are densely stocked with oil droplets, the whole zone appearing opaque in a

transverse section of fresh material. In contrast the colourless central core, 5 to 6 cells in diameter, consisting of narrower cells, contains only a few, very small oil droplets. The cells in this zone are more elongated, being 100 to 175μ ong and 15 to 20μ broad (Fig. 3) and possibly function in water storage as suggested by Goebel (Goebel, 1891).

At the tip of the stem there is an apical cell with three cutting faces lying at the base of a shallow depression filled with slime.

Anatomy of the Erect Shoot

The erect stem shows a central colourless zone which may be up to 16 cells in diameter in stems bearing sporogorria, and an outer zone 5 to 8 cells wide, containing small oil droplets and pale green plastids. Male shoots are more slender, with a central zone only 4 cells in diameter and an outer zone 5 to 7 cells wide.

The surface cells are approximately isodiametric, have a very thin cuticle, and contain more plastids than deeper cells; some of them give rise to mucilage hairs consisting of a club-shaped terminal cell and a short stalk cell. The deeper cells gradually become more elongated until in the central zone the cells are 200 to 300μ long and 30μ broad.

At the tip of the stem is an apical cell with 3 cutting faces, all three faces being equal in size except immediately after cell division (Figs. 6, 7). From each segment cut off by the apical cell a leaf arises. These leaves are at first uniseriate, but as they enlarge the basal portion becomes multiseriate, usually 4 cells in thickness (Fig. 8). However, vegetative leaves on male shoots are only 2 cells thick at the base and small leaves at the base of the shoot are uniseriate throughout.

A lateral branch is initiated by the formation of an apical cell on the side of the stem at a point midway between two leaves of a vertical row (Fig. 8). Mucilage hairs arising from adjacent cells curve over the apical cell and protect it with a mucilage sheath. Some of the apical cells after cutting off 3 or 4 segments remain dormant for lengthy periods, whereas others immediately produce lateral branches. The number of apical cells produced, their time of appearance, and their height on the erect stem varies from plant to plant.

The Anthendium

The antheridia occur interspersed with mucilage hairs at the tip of the antheridiophore (Fig. 9). The mucilage hairs have a stalk 3 cells in length, and a large, club-shaped terminal cell. Young antheridia develop close to the apical cell and also amongst the older antheridia.

The antheridium arises from a superficial cell which projects from the surface and then divides transversely into an inner cell embedded in the receptacle tissue and an outer protruding cell. The outer cell divides transversely into the primary antheridial cell and the primary stalk cell. The latter divides by two intersecting vertical walls followed by transverse walls to produce the antheridial stalk.

In the primary antheridial cell the first wall is an obliquely vertical one, dividing the antheridial cell into 2 cells of unequal size (Fig. 10a). In the larger cell 2 further obliquely vertical walls are formed, which meet the first wall above. At this stage there is a densely-staining central cell, shaped like a three-sided pyramid, enclosed by 3 jacket cells (Figs. 12a, 10b). The first-formed jacket cell divides vertically, and when the resulting cells enlarge the central cell expands in this direction and forms a four-sided pyramid (Fig. 10c). Later the other jacket cells divide also. Transverse divisions follow, producing the uniseriate jacket of the fully developed antheridium. The central pyramidal cell divides transversely (Fig. 12b) and then by repeated divisions in various planes gives rise to the spermatogenous cells (Figs. 11, 12c, 13). Either all the fertile cells divide simultaneously or else those belonging to the top half of the anthendium divide prior to those of the

bottom half. The last division in the formation of the spermatogenous cells is not necessarily a diagonal one, such as frequently occurs in the Hepaticae, but lies in any plane.

The mature antheridium is oval in shape and stands away from the surface of the receptacle on a stalk composed of 4 rows of cells 3 to 5 tiers in height (Fig. 13). When water is present it opens at the top, the terminal cells separate from one another and bend outwards as the mass of sperm cells is slowly extruded.

The Archegonium

The archegonium in its early stages of development cannot be distinguished from a young antheridium. It arises from a superficial cell alongside the apical cell and in position replaces a leaf rudiment (Fig. 15). The superficial cell projects from the surface and divides transversely into an inner cell embedded in the receptacle tissue and an outer protruding cell. The latter, instead of dividing by vertical walls as is usual in the Hepaticae, divides transversely in the manner recorded for Pellia epiphylla (Hutchinson, 1915), so forming a stalk initial and a terminal cell. At this stage the resemblance to a young antheridium is very striking. The stalk initial soon divides by vertical and by transverse walls giving rise to the base of the venter, and to the short stalk of the archegonium standing above the surface of the receptacle. Meanwhile in the terminal cell 3 inclined walls are laid down successively in such a way as to cut out a pyramidal central cell from 3 jacket cells. (Figs. 16, 17a, 17b.) The central cell gives rise to the axial row. (Figs. 18, 19, 22.) One of the jacket cells divides vertically so producing 4 jacket cells (Fig. 17c). For a long time thereafter the jacket cells in the neck region divide only transversely and give rise to a four-rowed neck (Fig. 21a), but at a late stage of development vertical divisions occur in the terminal region and the neck here becomes from 6 to 10-rowed (Fig. 21b). In the venter region the jacket cells divide both vertically and transversely, and just before the archegonium is mature they divide once periclinally (Fig. 22).

In the mature archegonium there is a short stalk projecting above the surface of the receptacle, a biseriate venter only, slightly broader than the base of the neck, and a spirally twisted, often curved, neck up to 570μ in length (Fig. 22). The axial row comprises the egg, the ventral canal cell and as many as 40 neck canal cells. At a late stage of development vertical or obliquely vertical divisions take place amongst the uppermost neck canal cells. (Figs. 20, 22.)

While the archegonia are being initiated there is little extension growth of the stem, and the upper part broadens into a convex, disc-shaped receptacle on which archegonia intermixed with mucilage hairs are closely aggregated (Fig. 14). After about 20 archegonia have been formed, the apical cell ceases division, and, if fertilization fails to occur, it remains in position at the centre of the receptacle. Several months later it may function again and give rise to new leaves and additional archegonia, but growth of a new erect shoot eventually brings its activity to an end.

Development of the Sporogonium and its Protective Sheath

Insufficient young embryos were obtained to enable the development to be followed in detail, although it could be determined that, after the zygote has divided by a transverse wall, a vertical wall is laid down both in the epibasal and in the hypobasal cell (Figs. 23–25). Evidently the early stages are passed through very rapidly until the sporogenous tissue is clearly delimited. Subsequent development is dependent on external conditions, possibly day-length and temperature, as at any one time all sporogonia are found to be at the same stage.

While the embryo is developing, the cells of the archegonial venter divide a few times forming a calyptra which becomes 4 cells wide in the basal portion

(Fig. 26). Meanwhile the tissue of the receptacle surrounding the base of the archegonium divides actively forming a massive perigynium roofed over above by the outstretched calyptra. Scattered over the outer surface of the perigynium, especially in the basal portion but sometimes high up, are old archegonia, and above are reduced leaves.

The young sporogonium shows a foot, a seta, and a capsule region (Fig. 27). The fertile cells in the capsule are differentiated early. They become diamond-shaped. Some elongate greatly to form elaters, and others give rise to files of 4 spore mother cells. The spore mother cells separate from one another. The nucleus of

Text-fig. 4.—Fig. 27—Vertical section of a young sporogonium showing foot, seta and capsule regions, X 200. Fig. 28—Vertical section of part of a nearly mature capsule, showing the cap, X 75. Fig. 29—Part of the jacket of a ripe capsule in transverse section, showing the nature of the thickenings, X 250. Fig. 30—Spore mother cell, X 80. Fig. 31—Division of the spore mother cell, X 80. Fig. 32—Ripe spore, X 390. Fig. 33a-f—Stages in germination of the spore, X 390.