• Stewart Island: Forest trees near Freezer, Port Pegasus, 389, 11077, W. M.

• Campbell Island: Beeman Hill, 1958–59 Expedition, comm. K. W. A.

• Macquarie Island: 52W, Dr. M. E. Gillam, comm. J. H. Willis.

What appears to be specimens of M. furcata with marginal gemmae have been met with, but it is difficult to be certain about this identification. In the Notrhern Hemisphere the gemmiferous state with marginal gemmae is known as var. ulvula.

Metzgeria disciformis Evans is so intimately bound up with M. furcata, that it seems correct to consider it a variety of M. furcata on an equal footing of course with var. furcata. Even so, a large number of specimens will be unidentifiable as to variety, but they will at least have a specific name. M. disciformis Evans which is here transferred to M. furcata var. disciformis (Evans) Hodgson, resembles the typical form of M. furcata in the flat-margined thallus with single marginal hairs, the ventral alar hairs and the structure of the mid-rib. It differs from M. furcata sensu stricto in the occasional presence of dorsal gemmae and in the bluish or blue colouring which may appear after drying. The following is the original description: “Pale green more or less tinged with blue, growing in depressed mats; thallus prostrate, repeatedly dichotomous, well-developed branches about 1.2 mm wide and from 1–2 mm long between the forks, plane or slightly convex; costa bounded antically by 2 rows of cortical cells and below by 4 rows (rarely 2 or 3); wings from 15 to 20 cells broad, the cells with slightly thickened walls, but without evident trigones, averaging about 25 × 21μ, but not varying much in size in different parts of the thallus; hairs restricted to the margin and to the postical surface of both wings and costa, sometimes few but usually abundant along the costa, marginal hairs averaging about 90μ in length, straight or nearly so, occurring singly, slightly displaced to the postical surface: inflorescence not seen; gemmae arising from the antical surface of the wings, in the form of circular concave disks, one cell thick throughout, abruptly contracted at the base into a short stalk composed of 2 or 3 cells, apical cell single, antical hairs few or wanting, borne near the centre of the convex surface, marginal hairs numerous but much shorter, usually displaced to the postical surface.

On leaves. Colenso No. 1907. The specimens were communicated to the writer under the name of M. australis Steph.

Except for its small cells and peculiar colour the present species agrees pretty closely with M. furcata in its vegetative structure, the costa being built up according to the same plan. The antical gemmae will at once serve to distinguish it.”

Unfortunately the bulk of the blue specimens do not bear gemmae at all, but nevertheless it must be assumed that they belong to M. disciformis Evans, because M. furcata var. furcata does not turn blue. The point is, however, that some wholly green specimens which would ordinarily be referred to M. furcata, bear antical (dorsal) gemmae. The small cell-size mentioned by Evans is not constant. Thus it would appear that M. disciformis and M. furcata are intricately mixed and must both belong to the aggregate species M. furcata, which ensures that they all at least have a specific name.

As regards the surface gemmae, Evans writes (p. 290, 1910): “In the early development of a gemma one of the alar cells projects from the surface in the usual way, but it becomes at once the mother-cell of the gemma without undergoing one or more preliminary divisions. In this respect, M. disciformis differs markedly from most of the other species which bear antical gemmae, but agrees with most of those which bear marginal gemmae.…” There remains, then, only the position of the gemmae to be considered. Macvicar definitely states that M. fruticulosa in addition to bearing marginal gemmae has them on both surfaces near the apex. Their presence on the surface therefore in M. disciformis. Evans can have little significance as a specific character.

The type of M. disciformis was epiphyllous. Other epiphyllous but non-gemmiferous specimens are: On tawa (Belschmiedia tawa) 2518 E. A. H.; on Pseudo-wintera axillaris, Doubtful Sound, 2431, G. Simpson, comm. H. H. Allan; on rewa-rewa (Knightia excelsa) Herb. Colenso.

Localities of green and blue plants with surface gemmae:—North Island: On tree by watercourse, Mahia Peninsula (green), 2380, on tree by watercourse, Kiwi Valley, Wairoa (green) 1653, on trees, edge of bush, Morere Hot Springs (green) 1657, (green) 11074, E. A. H.; on manuka, Paeroa stream, 11073, A. L. H.; Ruahakune Bush, E. of North end of L. Taupo, 11071, bare face of shady rock near Atiamuri, 11007, K. W. A.; on trees under rocks opposite Rosie Bay, L. Waikare-moana, 9378, H. M. D., on smooth barked trees, coastal forest, Kapiti Island, 11072, A. P. D.

South Island: On bark of Fuchsia excorticata, Balgueri V. Akaroa, 9575, on F. excorticata Pine Hill, Dunedin, 11497, W. M.; on manuka in open above Ravens-bourne, 11498, K. W. A.; on lacebark with Frullania patula, Waihopai Forest Reserve, Invercargill, 9664, W. M.

Bluish non-gemmiferous specimens from the outlying islands, classed as M. furcata var. disciformis are as follows:

Auckland Islands: Port Ross on Coprosma ciliata in Metrosideros forest at the shore, 2241, on Suttonia diuaricata from the same place, 2237, G. Einar Du Rietz.

The Snares: 50, F. Newcomb (calyptras almost devoid of hairs).

Antipodes Islands: Low shrubs of Coprosma antipoda in the tussock country 2598 G. Einar Du Rietz. This specimen was erroneously named M. decipiens var. violacea (Hodgson, 1948).

Specimen No. a894 Herb. Mitten, labelled M. australis Steph. is M. furcata var. disciformis.

Beyond mention of a stout nerve, there is nothing in Colenso's description of his M. flavo-virens (1888) by which one could identify the species. Stephani places it in M. furcata.

Specimen No. 000722 Herb. Stephani, from Mt. Winterslow, Beckett, labelled M. nitida is M. furcata var. disciformis.

Metzgeria hamata. Lindb. (Text-fig. 1, 6–9)

Metzgeria linearis Lindb. Acta. Soc. Sci. F. et. 10, 494, 1875. Not M. linearis (Sw.) Aust.

• M. hamata Lindb. Acta. Soc. Sci. Fenn. 12, 25, 1877.

• M. leptoneura Spruce, Trans. Bot. Soc. Edin. 15, 555, 1885.

• M. nitida Mitt., Journ. Linn. Soc. Bot. 22, 243, 1885.

• M. australis Steph., Hedwigia 28, 267, 1889.

Plants dioicous, variable, apparently nongemmiferous in New Zealand, in loose or compactly layered cushions or patches, on earth, rock, or in grass, or even epiphyllous on filmy ferns, pale green or yellow-green to pale brown, glossy, spotted, common. Thallus to 5 cm, 0.10—1.3 mm broad, flimsy or rigid, closely to laxlv irregularly dichotomously branched, convex with margins decurved to sub-revolute, portions sometimes almost explanate, mid-rib usually conspicuous of a lighter colour, convex both dorsally and ventrally, bounded on both surfaces by 2 cortical cells; wings naked on both dorsal and ventral surfaces, margins with short or long hairs, single or in pairs when they are divaricate, but there may be long stretches without hairs. Cells variable in shape and size, maybe even in different parts of the same thallus, 40–50μ in diameter, or even to 70μ in extreme cases, usually hexagonal, but occasionally lax and shapeless, walls thin or thick, trigones often present. ♀ branches usually hairy along the margins, calyptra clavate, occasionally to 4 mm in length, hairy or only slightly so, seta wavy, vaguely cellular, in one instance increasing in width upwards to 1.7 mm at the junction of the valves; capsule valves 0.6–0.9 mm long, width to ca. 0.3 mm of ca. 14 cells, median line of outside cell layer without nodular thickenings, thickenings otherwise crowded; the cell walls are sometimes not in a straight line, when the pattern of thickenings is confused. ♂ branches usually with very few hairs.

In the northern hemisphere the marginal hairs are mostly hamate (hooked), hence the name, but this is exceptional in New Zealand.

Metzgeria nitida Mitt. common in New Zealand and recognized by some authors, was considered by Evans (1923) to be inseparable from M. hamata, and this view certainly seems to be correct. It was based on 2 specimens, one from Apollo Bay, Australia, and the other, Colenso's 279 with Homalia pulchella. Both of these are on the sheet with others, in Mitten's Herbarium, but Evans missed No. 279, owing to a badly formed figure 7, but a stem of the Homalia is present, which definitely identifies the specimen. However, Evans examined another one of Colenso's, No. 1100, similar to the Apollo Bay specimen. The thallus of each of these specimens is almost plane, the hairs fewer than usual, and the cells larger, averaging in No. 1100, 70 × 55μ; but the costa shows the 4 rows of cortical cells, 2 dorsal and 2 ventral as called for in the description, and where the marginal hairs are abundant, they do occur in pairs, and throughout the range of specimens, there are transitional sizes in the cells. It seems that excessive moisture may induce flimsy thalli with few hairs and large cells.

According to Schiffner (1898), the species listed in the Flora Novae Zelandiae 11 (1855, p. 166) and in Hooker's Handbook New Zealand Flora 11 (1867, p. 542) as M. furcata were M. hamata.

Metzgeria procera Mitt. mentioned in Flora Novae Zelandiae ii, p. 166, is not given as a New Zealand species. It is from Guadeloupe, and Lindberg (1877) made it a variety of M. hamata.

Metzgeria australis Steph. type 000718. Geneva, Rogers Creek, W. Bauerlein, from Melbourne Herbarium is certainly M. hamata. Stepham (1900) synonymized M. australis with M. nitida.

No. a894 Herb. Mitten labelled M. australis is M. disciformis Evans.

M. hamata is common throughout New Zealand and is usually recognized by the convex to sub-terete thallus with decurved, scarcely revolute wings, crowded marginal hairs mostly arising in pairs, the wings without hairs, and the mid-rib bounded both dorsally and ventrally by 2 rows of cortical cells, and bearing hairs on its ventral surface. Before a dichotomy the nerve may be thickened. Less typical forms have scarcely convex thalli, larger cells and fewer marginal and costal hairs. Occasionally the cells of the lamina may be shapeless with thick walls, and the thalli without hairs at all, such as 2353 from swampy ground amongst Cyperaceae near Atiamuri, K. W. A., 4/10/30.

Two localities outside the main islands are: Laurie Harbour, Port Ross, Auckland Islands, on Metrosideios umbellata, 2239, G. E. Du Rietz, and Tasman Stream, Three Kings Islands, M. Holdsworth, 5566.

Also found in England, Ireland, North and Scuth America, Polynesia, Southern Asia, Japan.

Metzgeria decipiens (Massal.) Schiffn. & Gottsche. (Text-fig. 1, 10–12)

• M. furcata var. decipiens Massal. Nuovo Gior. Bot. Ital. 17, 256, 1885.

• M. decipiens Schiff. &. Gottsche in Schiff. Forshungsreise Gazelle, 4, 43, 1890.

• M. glaberrima Steph Bull. Herb. Boissier. 7, 939, 1899.

• M. nuda Steph. Kungl. Svenska Vet-Akad. Handl. 46, 10, 1911.

Plants dioicous, rarely fruiting hyaline to dull green or shiny dull brown, very common, on trees and ground in layered mats, very variable, often gemmiferous (in New Zealand). Thallus flat, usually 1–2 cm long, but may be as long as 4 cm and scarcely branched ca. 0.6–0.8 mm, but may be more than 1 mm wide, mid-rib bounded both dorsally and ventrally by 2 rows of cortical cells, ventral branching often present; wings naked on both dorsal and ventral surfaces, margins plane, hairs short or long, occasionally abundant, but more often absent for long stretches or altogether, single or very rarely in pairs, mostly truly marginal, occasionally branched at the apex as in rhizoids. Cells almost always irregular in size on the same thallus, averaging 38 × 29μ (Evans), with individual cells deviating widely from this measurement, largest usually adjacent to the mid-rib. ♀ branches with few or many marginal hairs, calyptra clavate, 1.1 mm long, lower portion bare, upper sparsely hairy in some cases, densely hairy in others, hairs ca. 0.2 mm long; capsule valves not constant in length and width, to 0.8 mm long × 0.4 mm broad, but may be smaller. The local thickenings on the

outer wall of the valves are mostly on the inner cell-walls, those furthest from the margin. ♂ branches destitute of hairs. Gemmae arising from a marginal cell, broaden into a rounded confusion of small cells, appearing sessile, which ultimately elongates and develops with or without a mid-rib, occasionally persisting to form a branch. Chilian and other specimens examined by Evans were without gemmae.

A sterile specimen in the Riksmuseum, Stockholm, labelled Metzgeria glaberrima by Stephani, from W. Patagonia, T. Halle a256, and determined by Evans as M. decipiens, does not differ from New Zealand and Australian specimens of M. glaberrima in Stephani's Herbarium. Therefore, as Evans conclusively shows that M. dicipiens and M. glaberrima are one and the same thing (1923, p. 302), one can confidently treat our New Zealand plants as M. decipiens.

Speaking generally, the plane margins, the scarcity of costal and even marginal hairs, which are practically always single, the absence of ventral hairs, the 4 cortical cells bounding the mid-rib, 2 dorsal and 2 ventral, excepting before a fork, the occurrence of marginal gemmae, and the irregularity of cells in size and shape, all taken together, will serve to differentiate the species. Massalongo, the author of M. furcata var. decipiens, states that it is almost intermediate between M. furcata and M. hamata, which seems correct. It is unfortunate that, in separating a closely related species from M. decipiens, M. epiphylla (Evans, 1923), a taxon with more or less narrowed and specialised gemmiferous branches, Evans restricts the conception of M. decipiens to a species without gemmae. Our plants do at times produce gemmae, but there is nothing exceptional about this, as M. hamata, without gemmae in most countries, including New Zealand, is gemmiferous in Japan and North America.

Another character of the restricted M. decipiens is to do with the nodular thickenings on the outside wall of the capsule valves. According to Evans the median wall or line in M. decipiens, has thickenings on both sides arranged alternately, while that of M. epiphylla is bare of thickenings. In New Zealand, the median wall or line is bare as in M. epiphylla, while the arrangement in adjacent cells appears to be haphazard and conforms with neither that of M. decipiens or M. epiphylla, being sometimes on neither of the visible vertical walls but in between them, while there may be a net-work of cells with no apparent median line at all. One could not propose even a sub-species or a variety on the basis of such a slender and variable difference as the arrangement of cell-thickenings. It seems best, therefore, to treat M. decipiens as an aggregate species and not a restricted one.

It is clearly evident that M. quadriseriata Evans from Japan (1906) is a variant of M. decipiens, differing only in the more elongate marginal gemmae.

As regards thallus with arrested development, or perhaps reversionary, some specimens consist almost entirely of these. One such, 7691 from Paeroa Stream, Auckland Province, is made up of tiny unbranched thalli, less than 2 mm long and only 0.15 mm wide, with thick-walled alar cells in 3 rows. These minute plants may or may not bear gemmae, and some may even show calyptras and capsules. In others the mid-rib may be lacking. At the other extreme are long, flat thalli, wide and scarcely branched, indistinguishable from forms of M. hamata. Indeed Stephani refers M. decipiens to doubtful synonymy with M. nitida (M. hamata).

No. 9953 from Half Moon Bay (Martin) might be mentioned as one in which the thalli are margined with narrow cells, though not as narrow as in M. linearis (Swartz). Another specimen examined was 22099 Auckland Museum Herbarium, from fallen logs and bases of trees, Mt. Pitt, Norfolk Island, F. C. Allen, which is undoubtedly M. decipiens. M. glaberrima (Dusen 394) from Is. of Guaitecas, and M. glaberrima (Naumann) from Tuesday Bay, Straits of Magellan, both of which Evans has identified as M. decipiens, have also been examined.

Gemmiferous specimens of M. decipiens are from the following localities:—

North Island: On bark, Morere Hot Springs bush, 7695, on dead branch of tree near outlet, L. Waikaremoana, 9306, on bark, near Wairoa, 11504, Pohangina

Totara Reserve, abundant, small tree in bush near Manawatu Gorge, 10668, Bush below Mt. Egmont Hostel, 11232, E. A. H.

South Island: On bark, Riccarton Bush, Christchurch, abundant, 2406, E. A. H.; on rock amongst mosses, Mt. Herbert, 11385, K. W. A.; tree base, Waihopai Scenic Reserve, Invercargill, 9575, W. M.

According to Evans (1923, p. 302), it seems possible that Schiffner's M. decipiens and Stephani's M. glaberrima were in part at least, based on the same material from Straits of Magellan.

Metzgeria atrichoneura Spruce (Text-fig. 11, 13–18)

M. atrichoneura Spruce. Trans. Edin. Bot. Soc., p. 556, 1885. Steph. Spec. Hep. 1, 298, 1900.

Plants dioicous, small to medium, more or less alpine, pale, whitish to yellowish green, rarely gold, creeping loosely or in layers on bark or twigs, or in sub-erect compact tufts, or creeping on or through other bryophytes on earth, gemmiferous. Thallus convex, more or less dichotomously branched to 1.5 cm, but often shorter, ordinarily appearing ca. 0.5—0.6 m wide when not explanate, but branches often 0.3 mm or narrow-linear, almost tubular, to 0.15 mm with or without gemmae, or resembling stolons, mid-rib 0.6–0.9 mm wide, bounded by 2 dorsal and 2 ventral cortical cells, as in the type to as many as 8 dorsal and 8 ventral cortical cells and 35 small interior ones, with or without ventral costal hairs, wings naked on both surfaces, margins strongly revolute often reaching to halfway across the ventral surface, hairs single or rarely in pairs, 0.15—0.25 mm long and often intermeshing across the mid-rib, short and thick on gemmiferous thalli. Cells convex, hexagonal, ca. 40–45μ in diameter but often smaller, walls usually thin, trigones minute; excessive moisture seems to produce large shapeless cells with thick walls, and a yellow nerve at times. The convexity of the cells is noticeable along the fold of the margin, and is described by Spruce, the author, as “valde papulosae prommulae”, but it is not always so pronounced. ♀ branches obcordate-reniform with dentate to ciliate margins. Calyptra elongate-clavate with few or many apical hairs, or variously setose; capsule valves 0.85 mm, median width 0.35–0.4 mm; nodular thickenings small and crowded. ♀ branches naked, antheridia usually 3 or 4. Gemmae slender with crenulate margins, crowded along the margins of narrow modified thalli and sometimes of the ordinary thalli, along the fold of the thallus, or somewhat displaced towards the ventral side. Commencing as a rounded body, it elongates considerably to 1 mm or more, but remains narrow, ca. 6–8 cells wide, 0.1–0.14 mm, from a still narrower base of 2–4 cells. The margins become strongly revolute at an early stage, showing dorsiventrality. Occasional secondary gemmae may arise from the margins. Masses of detached gemmae may be among the stolon-like branches, which may make up most of a specimen. Gemmiferous thalli may bear either ♂ or ♀ branches, and adventive ventral branches may also be present.

The nerve varies considerably in thickness. Narrow, modified gemmiparous branches usually have a thick nerve. In fact the thin nerve with the 2 over 2 cortical cells as described by Spruce is mostly at the ends of the branches. In specimens from seepage amongst rocks, 6,000ft on Mt. Egmont, a very thick nerve preponderates throughout.

The distinctive gemmae and the whitish stolon-like branches will distinguish the species when present, otherwise the small size, the strongly revolute margins usually with copious hairs, and the crenulate margins at the fold of the wings are the signs one must look for M. hamata is larger and always has a 2 over 2 sheath to the nerve.

Localities of gemmiparous plants:—

North Island: On bark of Olearia ilicifolia Hauhangaroa Range, W. of L. Taupo, 2,400ft, 1016, on manuka, Otupae, N. W. Ruahines, 2434, Mt. Kapakapanui, 869, 883. On stems of Olearia lacunosa with Lepidolaena Berggrenii, slopes of Mt. Alpha, 3,900ft, Tararuas, 2461, Akatarawa. V., 2537, A. P. D.; on shrub plateau above Stratford Mountain House, Mt. Egmont, 3–4,000ft, 11293, E. A. H.; Oroua V. Ruahines, H. M. H.; bark in bush, Ohakune Mountain road, 11522, E. A. H.

South Island: Pelorus Bridge, Nelson, 1652, A. L. H.; Arthur's Pass, 2485, F. MacDonald.

Some specimens have lax shapeless cells with thick walls, perhaps caused by excessive moisture. Such specimens are: In edge of pool under waterfall near Atiamuri (tip not quite reaching surface) 2360, swamp, growing in water E. of Waiotapu V., Rotorua, 2358, 2361, K. W. A.; Kaimanawas ca. 5,000ft, 5977, A. P. D.; Otonga, Whakapara, W. A. Setchell, det. Pearson, Manchester University

Herbarium; a 438 Colenso, Herb. Mitt. unidentified, and another unidentified one in M. nitida folder in Mitten's Herbarium, New York.

Localities of non-gemmiparous plants, North Island: Tree trunk, Waipoua Kauri Forest, 2359, K. W. A.; Subalpine Ruapehu, 3375 in part, S. Berggren; Urewera National Park, 438, G. O. K. S., 2505, 9707, E. A. H.; Waiau R. Gorge, 2519, B. Teague; 7 specimens from Kaimanawas, including Mt. Makarako, 5,000ft, A. P. D.; 10 specimens from Ruahine Ranges, including Mokai Patea, 5,000ft, A. P. D., N. M. Elder, E. A. H.; Mt. Climie, Rimutakas, A. P. D.; Waiopehu Ridge, 2453, G. O. K. S.; common on Mt. Egmont to 6,000ft, A. P. D., R. Svihla, G. O. K. S., W. M., E. A. H., a681 Herb. Colenso; Maungapohatu, 6,400ft, 9695, A. P. D.; W. of Kime Hut, Tararuas, 7680, Botany Division (D.S.I.R.) Herb., V. D. Z.

South Island: Arthur's Pass (approaching M. hamata) 119, 113, W. M.; Little River, Akaroa, 3386 in part, S. Berggren; Misery Creek, Cass, B33, L. Visch comm. W. R. Philipson; on rock, de la Beche Hut, Tasman Glacier, 2528, G. O. K. S.; mid Godley R. valley, 4–5,000ft, D. Scott, 206, comm. K. W. A.; Stillwater Base Camp, Caswell Sound, 5181, V. D. Z.; Facile Harbour, Dusky Sound, 2352, H. H. A.

Stewart Island: Base of tree, Pegasus, 588, W. M.

Apart from New Zealand, Spruce gives no details as to collector or locality. The type is reported as missing from Spruce's collection in Manchester University Herbarium, but there seems to be no doubt whatever as to the group it represents, as indeed there is nothing else it could be that I know of, except perhaps a form of M. hamata, which is very unlikely.

Metzgeria saccata Mitt. (Text-fig. 11, 19–22)

• M. saccata Mitt, Journ. of Linn. Soc. Bot., 22, 145, 1886.

• Steph., Spec. Hep. 1, 290, 1900.

• Rod. Pps. & Proc. Roy. Soc. Tas. Bry. 11, 16, 1916.

Plants dioicous, medium, pale or yellowish green to whitish, rarely fruiting, compact or loosely sprawling on bark and twigs, usually amongst other bryophytes. Fronds to 2 cm convex to 1.2 mm wide, mid-rib bounded dorsally by 2 cells and ventrally by 3–4, apparently naked, but ventral branching may be present; wings segmented with inflated saccate lobes on the ventral surface, more or less continuous, sometimes involute, to 1 mm long, 0.5 mm broad, marginal hairs absent or few, short or long, usually between the lobes, but some were noticed at the fold-over of the lobe, also at the apex of the frond, sometimes with terminal suctorial branches. Cells 30μ fairly regularly hexagonal with thickish walls, trigones minute. ♀ branches 0.3–0.4 mm long with few hairs at the apex Calyptra ca. 2 mm, clavate, clothed but not densely with laminae, irregularly shaped, 3–6 cells deep, somewhat lobed, margined with stiff hairs. All the calyptras present are crowned with these laminae. I do not think they are in any way connected with gemmae; median line of outside capsule wall bare of thickenings, spores minutely papillose, ca. 0.05 mm. in diameter. ♀ branches with 3 antheridia, destitute of hairs.

The saccate marginal lobes at once distinguish this species from other New Zealand ones, and of course the laminiferous calyptra when this present, but M. inflata from British Guiana is suspiciously close. Stephani noticed the similarity, but stated that as the margins of M. saccata are segmented into inflated lobes, and those of M. inflata are continuously inflated, they must be separate species. However, New Zealand specimens show long stretches cf unstrictured inflated margins, to mention only two, No. 831 from N. W. Ruahines, and 2420 from Arthur's Pass. Calyptras of M. inflata would show definitely whether these two are identical as suspected, though it is a far call from New Zealand to British Guiana.

North Island: Mt. Pukeamaru, E. Cape, 3,200ft, 11441 Alison Druce; 6 specimens from manuka bark (Leptospermum scoparium) E. of Taupo; Opepe, near Taupo, K. W. A.; on manuka, W. of L. Taupo, Hauhangaroa Range, ca. 2,500ft, 7018, Wharite, South Ruahines, 2,500ft, 2417, A. P. D.; Aniwaniwa V. Waikaremoana, 9730, H. M. D.; Eastbourne, 3161, N. J. Butler.

South Island: Arthur's Pass, on beech bark, 6a, 60, W. M.; on bark, 2420, 2419, H. M. H.; L. Rotoiti, Nelson, 2451, G. O. K. S.; Middle Bush, Cass, West Canterbury, C. Burrows.

The only specimen with calyptras so far seen is from beech tree with Frullania aterrima, by path to Tawhai Falls, Tongariro National Park, 11531, E. A. H.

The type was from Tasmania. Helms, mentioned by Stephani, was probably the first to collect the species in New Zealand.

Metzgeria Colensoi Steph. (Text-fig. 11, 23–24)

M. Colensoi Steph. Spec. Hep., vi, 48, 1917.

Plants dioicous, dull fawn or pale green, variable, mostly on bark, gemmiferous. Thalli to about 1.5 cm, rarely more than twice dichotomous, flat or slightly convex, variable in width, 0.8 mm, but usually broader to 1.2 mm, the wide thalli having shorter spaces between the forks, mid-rib thin, bounded by 2 dorsal and 2 ventral rows of cortical cells; ventral branching sometimes evident, ventral hairs abundant in places, short or long and sometimes curved; wings 20–30 cells across, naked on the dorsal surface, with hairs on the ventral, numerous to scarce, straight or bowed, marginal hairs very variable, 0.2—2.8 mm, may be slightly displaced to the ventral surface, single or twinned, straight or bowed, sometimes ending in branched suctorial surfaces, thus serving as rhizoids, cells fairly uniform, hexagonal, ca. 30–35μ, near the costa ca. 35–45μ, but occasionally they are larger; walls medium to thick. ♀ branches broadly reniform, margins and surface densely hairy with curved to hamate hairs. Calyptra clavate, very setose, hairs straight, or curved to hamate; capsule valves 0.5 × 0.2 mm nodular thickenings irregular in size and shape, median line wavy with minute thickenings. ♂ branches naked, reaching from the mid-rib to the margin, cells quadrate, antheridia 3–6 (in plant examined). Gemmae marginal, round at first, becoming oblong, nerved and with a few marginal hairs.

Stephani does not mention the ventral alar hairs, but a few are present in the type. Stephani also describes the plant as bipinnate, and draws it so, but it is not a marked characteristic and was not obvious in the piece of type examined. The species is distinguished from hamata by the planer margin, small cells, and the ventral alar hairs, and probably by the gemmae when these are present.

I would venture to suggest that M. Colensoi may be M. Magellanica (Schiff. Exped. Gazelle p. 43, 1889). I would not agree that M. Magellanica is synonymous with M. nitida as determined by Stephani.

Apart from the occasional presence of gemmae, M. Colensoi appears to differ from the African M. camerunensis Steph. (1899) only in the smaller number of interior nerve cells.

Localities of gemmiferous plants: On manuka near headquarters, Waipoua Forest, H789, K. W. A., Toa Toa, Bay of Plenty, 2389, Mrs. Haskell; on tree-fern caudex, McKinnon's Bush, near Wairoa, 2311, with moss and lichen, track near North Egmont Hostel, 11178, E. A. H., with Papillaria sp., bush at Ball's Clearing, Puketitiri, H. B., 2339, E. C. West; base of beech tree by Mahuia R., Tongariro National Park, 11521, E. A. H.; on totara trees, E. side of Rimutakas, 11421, A. P. D.

Twinned hairs are much in evidence in these specimens, and have served to distinguish the Mt. Egmont one from M. decipiens, in the apparent absence of alar hairs.

Localities of non-gemmiferous plants:—

North Island: Epiphytic in tanekaha bush, Great Barrier Island, 18, R. Lloyd; Papakauri, S. Berggren, on small branches, Onetangi Beach Bush, Waiheke Island, 1391, A. L. H.; on manuka bark, Puaiti, Nr. Rotorua, 2310, on manuka, near Atiamuri, 11402, 11400, H90 in part, on manuka E. of Taupo, 2390, 11396, Mokoia, S. of Hawera, 11401, upper surface of nikau palm fronds in damp bush gully, 11501, K. W. A.; small trees, Makaretu Falls Bush, Wairoa, 9698, Manawatu Gorge, 10670, E. A. H.; Kapakapanui Mt., Ruahines, 878, 853, on kanuka tree, E. side of Rimutakas, ca. 500ft, 2819, Upper Hutt, 909, A. P. D.; on kanuka tree, N.W. of L. Wairarapa Rd., 11502, Silveistream Bush, 2747, Mt. Holdsworth, 2531, H. M. D.; on Melicytus ramiflorus, Wilton's Bush, Wellington, R. Mason; near outlet, Lake Waikaremoana, 9305, E. A. H.