Art. XXIV.—On the Flowering Plants of Stewart Island.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 17, 1884, Unnumbered Page

Art. XXIV.—On the Flowering Plants of Stewart Island. By T. Kirk, F.L.S. [Read before the Southland Institute, 9th December, 1884.] Until within the last five or six years botanists have been almost entirely ignorant of the flora of Stewart Island, our knowledge being restricted to about a dozen flowering plants and a similar number of mosses and Hepaticæ collected by Dr. Lyall in 1848–49, recorded in the Handbook of the New Zealand Flora. Of late years, however, this reproach has been removed by the labours of several collectors whose work may be briefly mentioned. Mr. Charles Traill has done a large amount of good work in the investigation of the flora and fauna of the island: I am specially indebted to him for dried specimens of about 200 species of flowering plants and ferns, accompanied in many cases by valuable notes on their habits and distribution, as well as the native names in use on Stewart Island. Mr. and Mrs. A. W. Traill have most kindly formed for me a copious collection of the plants of Ruapuke Island, numbering about 140 species, several of which have not been observed on Stewart Island proper. Mr. W. Pearson, Commissioner of Crown Lands, has laid me under obligations for numerous dried plants collected in out-of-the-way places during several visits to the island. In 1870 Professor Black, of the Otago University, visited the island, on the part of the Otago Provincial Government, for the purpose of investigating its natural productions, but no plants of special interest were obtained. Mr. G. M. Thomson has, I believe, paid several visits to Stewart Island, during which he collected the new species of Brachycome, to which his name is attached, and Myrsine chathamica, previously known on the Chatham Islands only; he also collected between thirty and forty species of ferns, with other plants of interest. In 1880 he was accompanied by Mr. D. Petrie, who read a valuable paper on the flora of Stewart Island before the Otago Institute, and gave a catalogue of 200 flowering plants observed by him, this being the first published account of the plants of the island.**Trans. N.Z. Inst., vol. xiii., p. 323. During this expedition Actinotus bellidioides and Liparophyllum gunnii were added to the New Zealand flora, a matter of great interest, as previously they were only known as indigenous to Tasmania, where they are extremely rare. In

December, 1881, Mr. W. S. Hamilton and Mr. Goyen made the ascent of Rakiaua, during which they collected an Aciphylla, the first observed on the island, Raoulia goyeni and Hymenophyllum rufescens, the last being extremely rare and local. I had the pleasure of examing the flora in January, 1882, when I landed on Herekopere Island; and again in January, 1884. Although my stay on the last occasion was but short, I was able to make the ascent of the Ruggedy Range, on the west coast, and of Mount Anglem, the highest peak on the island, but was unable to visit the extreme north or the extreme south. During my excursions a few plants new to science were discovered, and much important information collected with regard to the vertical and horizontal distribution of many interesting species. Stewart Island is about 42 miles in length from north to south-west, and about 26 miles in its greatest breadth from east to west. Its area is estimated at 640 square miles. It is composed chiefly of slates and granitic rocks, the latter often in a highly decomposed condition, and in some localities intersected by dykes. In a few instances the slates are more or less metamorphosed by the overlying or interjected volcanic rocks occasionally passing into a kind of novaculite, which is utilised for whetstones by the settlers. On the eastern side the coast is deeply indented by Paterson's Inlet, a fine arm of the sea, which extends half way across the island and includes several secure harbours—Glory harbour, Abraham's Bosom, Glory Cove, etc. At the head of the inlet a swampy valley, but slightly elevated above high-water mark, runs across the island to a point between Ruggedy on the west coast, and the Mount Anglem range, where it is abruptly closed by low hills. About six miles from the head of the inlet a break in the hills on the southern side of the valley leads into Mason's Bay, where a large extent of flat land of good quality is sheltered from the sea by hills of blown sand, now more or less covered with low forests, although in some places advancing inland. From Mason's Bay a considerable extent of undulating table-land extends southward to Port Pegasus at an elevation of from 1,200 to 1,700 feet. To the north the country is much broken and rugged, culminating in Mount Anglem at an elevation of 3,200 feet. Between seven and eight miles from the head of Paterson's Inlet the main valley rises somewhat rapidly, and a grand system of river terraces comes almost suddenly into view; the terraces are two in number and are carried along each side of the valley; at a point where the valley is two miles wide the lower terrace is about 45 feet high and the upper about 40. Near the base of Ruggedy they form two branches, one running along the south-western flank, the other along the northern; both branches appear to

terminate abruptly at the base of the low hills which shut off the valley from the sea; but I was not able to visit the termination of either. These terraces are chiefly composed of loose sand and contain water-worn fragments of slate-rocks; their faces are often covered with a dense growth of scrubby manuka. No traces of fossils were to be seen on the surface. In its lower part the valley is traversed from east to west by low ranges of sandhills from three to six feet high which, in many cases, run parallel with each other for two or three miles and are covered with a scanty vegetation. With the exception of the valleys already mentioned, and the elevated table-land south of Mason's Bay, there are no tracts of level land. The country is more or less broken and covered with forest. Much of the soil is of a peaty character, and not well adapted for general cultivation. Where the peat is mixed with sand, as at Mason's Bay, introduced grasses can be grown with but little trouble. Mr. Walker has utilized this part of the island for a sheep run, and, after having had to face many difficulties, is realizing a fair measure of success, which will increase with the extent of land laid down in grass. At Halfmoon Bay, Horseshoe Bay, and Port William, the chief places of European settlement, the soil is of a fertile character, and introduced grasses are easily grown. At the Neck, which forms the southern head of Paterson's Inlet, a large acreage of excellent land is cultivated by the Maori and half-caste population. Two or three families reside as far south as Bravo Island in Paterson's Inlet, the extreme southern point of permanent residence in the colony. Port Adventure and places further south have long ago been abandoned by the Maoris, although I am assured that peach trees and other cultivated plants still mark the sites of their gardens. No observations have been recorded with regard to temperature, amount of rainfall, prevailing winds, etc., so that no precise data exist by which to compare the climate of Stewart Island with that of other parts of the colony. The abundance of luxuriant tree-ferns, the luxuriance of the varied ligneous and herbaceous vegetation, afford conclusive evidence of a mild, equable, and moist climate. In all probability the actual rainfall does not exceed that of Wellington or Auckland; but the atmosphere must be more continuously moist. It will be necessary to return to this subject before the close of the paper; but a remarkable result of the great amount of moisture in the atmosphere may be recorded here: in many instances the duramen of old trees is converted into peat, while the alburnum is still discharging its functions. Still more striking is the fact of the dense lower leaves on the stems of Raoulia goyeni becoming changed into peat, while the upper leaves are performing their usual functions.

So far as at present known the flora of the island comprises about 380 species of Phænogams and nearly 70 species of Ferns and allied plants. The lower Acotyledons have not been sufficiently investigated to allow of their number being estimated with any approximation to correctness. It is not my intention to present a complete list of the plants observed by me, but simply to offer a brief account of the most striking characteristics of the flowering plants, with special remarks on the endemic species. The traveller who visits Stewart Island in December or January will have his attention arrested by the blaze of crimson presented by the rata (Metrosideros lucida), often flecked with the pale racemes of the kamahi (Weinmannia racemosa), the deep green leaves of the puheretaiko and the tupari (Senecio rotundifolius and Olearia colensoi), and the grass-like leaves of the inaka (Dracophyllum longifolium), all of which form a fringe of greater or less breadth at the water's edge, and (except the puheretaiko) often expand into large masses ascending to the crests of the hills. Of these however the kamahi and rata alone attain the dignity of timber trees, and are excelled in dimensions by the rimu and the totara alone. The rimu (Dacrydium cupressinum) is abundant at the lower levels and attains large dimensions, the specimens comparing favourably with those from the west coast of the South Island, and yielding timber of greater durability. Podocarpus totara comes next in order of frequency, but is much less common than the rimu, although individual specimens attain a larger size. Dacrydium colensoi and D. intermedium are abundant in certain localities and yield the “white pine” of the saw mills: the true white pine (Podocarpus dacrydioides) does not appear to occur on the island. Podocarpus ferruginea and Griselinia littoralis, the miro and the kapuka, are plentiful, but the matai (Podocarpus spicata) is extremely rare. Amongst smaller trees and shrubs the most prominent are the horoeka (Panax crassifolia), and its congeners Panax simplex and P. edgerleyi, both exhibiting such excessive luxuriance as frequently to require an examination of the fruit before their identification can be determined; the rautawhiri (Pittosporum tenuifolium, var. colensoi), the kotukutuku (Fuchsia excorticata), the mako (Aristotelia racemosa), the mapau (Myrsine australis), the pokako (Elæocarpus hookerianus), also Carpodetus serratus, Myrsine divaricata, both of which are abundant. The remarkable “weeping tree,” Myrsine montana, with its crowded recurved branches and small foliage, affords a picturesque effect of a peculiarly attractive character altogether unique amongst New Zealand plants. On the slopes of the hills Leptospermum scoparium, the manuka, forms a small tree, but in the lower lands it is usually scrubby. It should be observed that the trees and shrubs here mentioned form the greater portion of the

forest, and with the exception of Dacrydium colensoi, are of general distribution in the colony. Olearia nitida, O. avicenniæfolia, and Panax colensoi are common by the sides of streams, or on the margin of forests. Veronica salicifolia occurs by the sides of water-courses, but is far from common. Schefflera digitata is abundant in damp gullies. The undergrowth is often extremely dense, and consists in many places chiefly of Coprosma fætidissima and C. lucida. C. colensoi is common, C. tenuicaulis and C. rotundifolia are local. Metrosideros hypericifolia is abundant, and is the only scandent species on the island. Myrtus pedunculata is plentiful, and the supplejack (Rhipogonum scandens) occurs in moist places, but scarcely ascends above sea-level. Amongst the shrubs the soil is often carpeted with a compact growth of the charming liliaceous plant, Callixene parviflora, with its elegant drooping flowers, mixed with numerous ferns, orchids, and mosses. The orchids form a marked feature in some parts of the forest. Corysanthes oblonga, C. rivularis, and others produce their attractive flowers literally by thousands; in no other locality have I seen these interesting plants in such vast profusion. Gastrodia cunninghamii is rare, having been observed only on the small island of Ulva. Caladenia bifolia is frequent, one of its forms making a close approach to C. lyallii. Chiloglottis cornuta occurs on Ulva, the glands on the labellum vary considerably in their shape and arrangement. In the majority of cases there are five depressed coloured glands arranged in a symmetrical manner, in a few specimens they were reduced to three, and in a solitary plant numerous stalked glands were arranged in a double row down the middle of the labellum exactly as in the Tasmanian C. gunnii, which is probably a state of the New Zealand plant. The dwarf variety of Pterostylis banksii, with abbreviated sepals, is common in open places in the forest. In addition to the terrestrial forms, the epiphytic forms are well represented, with the exception of Sarcochilus, which appears to be rare, and Bolbophyllum, which has not been observed on the island. Arborescent ferns are abundant alike on the outskirts of the forest and in its deepest recesses, often occurring in large masses especially on sheltered slopes, where they frequently rise above the level of the surrounding shrubs and produce an effect rarely seen elsewhere in the colony. The most common species are the poka (Dicksonia squarrosa) and the katote (Hemitelia smithii); the stem and fronds of the latter are sometimes used as food for cows. The mamaku (Cyathea medullaris) is local, and has not been observed south of Halfmoon Bay. Lomaria discolor frequently developes an erect caudex 2′–3′ high. Of the numerous filmy ferns I will only mention Hymenophyllum bivalve, one of the commonest ferns on the island, sometimes

covering moist banks with its rigid drooping fronds half hidden amongst Dicranum menziesii, at others exhibiting exceptional luxuriance and grace on the trunks of trees. The most striking feature of the fern flora (next to the abundance of tree-ferns) is afforded by the numerous varieties of Asplenium bulbiferum, A. obtusatum, A. falcatum, and A. flaccidum; in many cases these varieties are extremely beautiful and attractive, but merge into each other by such minute gradations that it is often a matter of difficulty to determine the species to which certain varieties should be referred. The occurrence of Asplenium falcatum, A. obtusatum, etc., in masses covering many square yards, to the exclusion of all other forms, is not often seen in other localities. The trees are more or less clothed with mosses and foliaceous lichens, while terrestrial Hepaticæ are unusually abundant, and on the whole form the most striking feature in this section of the flora. The genus Gottschea is especially well developed. Specimens of G. appendiculata were collected from 7″–9″ in length: a species new to science was remarkable on account of its dichotomous branching and habit forming loose rounded patches of a yellowish-green tint. The vegetation of the open peaty valleys between Paterson's Inlet and Mason's Bay is of a very different character. On the margins of streams a dense growth of shrubs usually exhibits great luxuriance, and consists chiefly of Olearia avicenniæfolia, O. nitida, Leptospermum scoparium, Veronica salicifolia, V. buxifolia: the last named exhibiting much greater luxuriance than in its usual mountain habitat, while the former is of smaller size than in the north. Under the shelter of the larger shrubs Gaultheria perplexa is abundant, and is frequently associated with Gleichenia cunninghamii. On the open land Gleichenia dicarpa is abundant, varying from stunted specimens 2″ high with a single whorl of branches, to luxuriant stems 18″ high with from 6 to 9 whorls. Lycopodium ramulosum covers acres of ground, when growing in the open its short stout branches present a stunted appearance. In places where fire had passed over the valley, the blanched dichotomous stems with their persistent leaves presented so close a resemblance to the European Selaginella helvetica, that one was repeatedly impelled to examine the plant afresh. In shade the plant is extremely luxuriant, but invariably prostrate in habit. Hundreds of acres were covered with the wiry stems of Hypolæna lateriflora, almost to the exclusion of other plants. Carpha alpina was abundant as well as Oreobolus stricta; but O. pumilio occurred only in small quantity, and in a single locality on the west coast. Drosera arcturi was extremely rare; D. binata and D. spathulata were plentiful. Centrolepis monogyna formed rounded velvety-looking masses at the

margins of bogholes, etc., the leaves being more hairy at the base than in alpine specimens. Actinotus bellidioides occurred plentifully in the fruiting state. Liparophyllum gunnii is common in all the swamps, its white roots sometimes a foot or more in length. Potamogeton oblongus, Myriophyllum pedunculatum, and Hydrocotyle muscosa are not unfrequent. Schizæa australis is plentiful in swamps, forming compact tufts 1″–3″ high. Dichondra brevifolia is common on moist peat, its large white flowers being conspicuous at a considerable distance. In wet places Carex echinata is frequently intermixed with the small bladder-wort, Utricularia monanthos, while Cladium glomeratum, Carex ternaria and other species fringe the margins of small streams and bogholes. In drier places several common mountain plants may be found—Pentachondra pumila, Cyathodes empetrifolia, Hierochloe alpina, Dacrydium laxi-folium varying from one inch in height to three or four feet, and in exposed places a stunted depressed form of D. intermedium. These montane forms are associated with the northern Lindsaya linearis, which is usually in poor condition, resembling specimens from the clay hills about Auckland, but in a few favourable spots luxuriant specimens were obtained. On the sandy ridges described in a previous paragraph, Drapetes lyallii was abundant, with occasional patches of Actinotus; large tussocks of Danthonia raoulii served to protect many smaller plants. A curious pigmy form of Viola filicaulis was not uncommon. Uncinia rubra and Danthonia semi-annularis formed large patches. Haloragis micrantha, H. uniflora, Libertia ixioides, Prasophyllum colensoi, Thelymitra uniflora, and an undescribed species of the genus, were frequent on many of the ridges. A few littoral plants deserve special mention. The suffruticose trailing stems of Tetragonia trigyna, often many feet in length, cover the rocks in sheltered places at the margin of the sea. A striking variety of Gentiana saxosa, distinguished by its much-branched, prostrate, or semi-erect habit, and deeply-cut calyx, with the segments subulate and recurved, is common on maritime rocks. A large Myosotis with white flowers, doubtless the plant described by Mr. J. B. Armstrong as M. capitata, var. albiflora, is abundant on rocks, often in situations exposed to the spray of the sea. Brachycome thomsonii is common on maritime banks and cliffs. Convolvulus soldanella is restricted to a solitary habitat in Sydney Cove. Festuca littoralis occurs in several localities, but is by no means abundant. Lomaria dura is found all round the coast, but does not extend inland. Atriplex billardieri is not uncommon, and Festuca scoparia is specially characteristic of sheltered bays. The typical form of Poa foliosa, a noble species, is found on Hereko-pere Island, and on certain headlands frequented by mutton birds south of Port Pegasus.

A brief account of the plants observed during the ascent of Mount Anglem, the highest peak of the island, will serve as a fitting introduction to the alpine portion of the flora. Landing at Sentry Point, our track led through a swampy forest, with large specimens of red pine and other trees, interlaced with supplejack. Tree-ferns were numerous, but restricted to two species, Hemitelia smithii and Dicksonia squarrosa, their trunks often shrouded with a luxuriant growth of filmy ferns. The ascent at first was very gentle, the supplejack (Rhipogonum scandens) did not ascend above 200 feet, and the tree-ferns disappeared below 400 feet. The timber gradually became of smaller dimensions, and at about 1,000 feet on the ridges the forest was to a large extent replaced by Leptospermum scrub, with solitary plants of Drosera stenopetala and Ehrharta thomsonii in moist places. In the gullies large timber occurred up to 1,700 feet or higher. Still ascending, Leptospermum scoparium gradually assumed an arboreal habit, but none of the trees exceeded 25 feet in height, although the trunks were from 1′–2′ in diameter. Progress was very easy up to 1,500 feet, when a tangled belt of Olearia colensoi, Dacrydium colensoi, and other shrubs, interspersed with inclined ratas (Metrosideros lucida) formed an almost impenetrable barrier, although rarely exceeding 15′–20′ in height. The branches were so closely interlaced that it was impossible to break through them, while they were so tough and elastic that our tomahawks were useless. Under these circumstances our progress was extremely slow, and in no way favoured by the continuous heavy rain. The vertical range of this belt did not exceed 800 or 900 feet, but it required four hours of excessively fatiguing work to force our way through; on the ridge of the spur the scrub was only breast high and offered less difficulty. From this point the ascent of the last slope was comparatively easy, except in a few hollows where the scrub attained 8 or 10 feet in height and was excessively dense. In many places it was varied by large patches of open peaty land studded with alpine plants—Celmisia discolor, Forstera sedifolia, Donatia novæ-zealandiæ, Phyllachne clavigera, Dracophyllum rosmarinifolium, and a remarkable prostrate form of D. scoparium, Carpha alpina, etc., etc.; owing to mingled hail, rain, and sleet at this elevation but little could be made out beyond the point on which we stood for the moment. At about 2,700 feet Dracophyllum menziesii, one of the most remarkable species of the genus, was collected; it attains the height of from 1′ to 3′ with the habit of a miniature D. latifolium, the native branches terminating in an almost globose head of recurved leaves with several racemes of waxy white flowers springing from beneath: the flowers are the largest in the genus. In swampy places near the crest of the slope

a small rush was obtained closely allied to Juncus antarcticus of Campbell Island, if not identical with that species, also an undescribed species of Uncinia and a Carex new to science. Before reaching the crest we encountered a dense snowstorm, which later on was varied by fierce blasts of sleet and hail, so that hands and face were stung almost past endurance, and the use of note-book and pencil became impossible for the rest of the day. The scene on reaching the crest was quite unexpected; right in front towered the highest peak, half-obscured by the driving snow, steep and precipitous, but between the crest on which we stood and the beetling cliffs was a crateriform hollow apparently 300 feet in depth, the bottom of which was occupied by a lake. The ridge from which we looked down to the lake sloped gradually to the water's edge, causing the hollow to present the appearance of a cup-shaped crater. I was unable to detect the outlet of the lake, but one of the party caught a view of it as the falling snow was momentarily swept on one side by a fiercer blast than usual. On its inner face the slope was sparingly clothed with scrub apparently of the same character as that amongst which we had been struggling' but time pressed too hardly to allow of its examination. On the crest itself stunted Olearia and Dacrydium were the commonest plants, but in some places a strange sight was presented: Dracophyllum muscoides formed a compact dark-green sward, thickly gemmed with white flowers, occasionally rising into pillar-like hummocks two feet high, and so extremely dense that it was a matter of great difficulty to thrust a knife into the woody mass; at a somewhat lower elevation short straggling branches were given off at the margins of patches, forming a kind of loose fringe. The line of ascent lay along the eastern side of the hollow, the mountain rising into a high peak not greatly inferior to the highest or northern peak. The ascent to the eastern peak was impeded by numerous patches of dense scrub of no great height but of extremely rigid habit; on the open spots several interesting plants were collected, amongst which may be mentioned a new Aciphylla of flaccid densely-tufted habit, which I have described as A. traillii, occasional specimens of Ranunculus lyallii, a prostrate form of Coprosma colensoi (identified in the absence of flowers), patches of C. pumila, a silvery Celmisia, destitute of flowers but apparently allied to C. sessiliflora, Schænus pauciflorus, Danthonia florescens, and two undescribed species of the same genus. In rock crevices Hymenophyllum villosum was plentiful, and the curious fern usually referred to Polypodium australe, v. alpinum; but the most interesting plant was unquestionably Ourisia sessili-flora, its large handsome flowers were white as the snow by which it was partially hidden.

The principal plants observed on the highest peak were Ourisia cæspitosa, which was luxuriant and plentiful. A Raoulia, or Haastia, destitute of flowers but new to science has been provisionally described as Raoulia goyeni. These were associated with the prostrate form of Coprosma colensoi and other plants already mentioned. The rapid approach of evening prevented any detailed examination of the vegetation of the highest and western peaks as it was desirable to commence the return to camp during daylight, but we lingered sufficiently long to frustrate our wish, and had the doubtful pleasure of camping in the open. It is obvious that under the unfavourable atmospheric conditions which prevailed during two-thirds of the ascent, and which continued until nightfall, the above must necessarily be an incomplete account of the alpine portion of the flora. Not only was investigation restricted to a comparatively small area, but many inconspicuous plants were hidden by the snow and hail, and others doubtless were overlooked, owing to the weakened condition of the observing faculties caused by our benumbed and saturated condition. A few of the endemic and rarer plants found on the island deserve special mention. Aciphylla traillii, T. Kirk. A small flaccid species found on Rakiaua and Mount Anglem. Ligusticum intermedium, Hook. f., var. oblongifolia. A handsome plant, to some extent combining the characters of L. lyallii and L. intermedium, but distinguished from both by the narrow, oblong, erect leaves. Mason's Bay and inland base of Ruggedy. Actinotus bellidioides, Benth., var. novæ-zealandiæ. At sea-level, head of Paterson's Inlet, and on Mr. Petrie's authority at Port Pegasus. On the peaks of Ruggedy, and at nearly 3,000 feet on Mount Anglem, but not observed at intermediate levels. Found also on the west coast of the Nelson district altitude 2,000 feet. Specimens collected on the dry sandy ridges between Paterson's Inlet and Ruggedy cannot be distinguished from the Tasmanian plant, but marsh specimens are characterized by numerous short barren stems. Aralia lyallii, T. Kirk (Stilbocarpa polaris, Dcn. and Pl. in part). I have elsewhere given my reasons for removing this plant from Stilbocarpa, and need only state here that it is altogether a littoral plant, being restricted to sea-cliffs or to small outlying islets, etc. One of the most striking plants in the flora. Olearia angustifolia, Hook. f. Tete-a-weka of the natives. This is one of the grandest flowering plants, and was first observed by Dr. Lyall, who, however, was not fortunate enough to obtain flowers. It appears

to be confined to exposed portions of the coast on both sides of the island, and to the adjacent islets, but has not been observed north of Paterson's Inlet. It varies in size from a small shrub six feet high to a tree of twenty feet, with a stout trunk and compact dome-shaped head. The leaves are of a deep glossy green, excessively coriaceous, lanceolate-acuminate with curious indurated obtuse teeth, white with closely appressed tomentum beneath. Head solitary, 1 ½″–2″ in diameter, with snow-white rays and rich purple disc, carried on stout foliaceous peduncles crowded at the tips of the branches. It is not easy to conceive of a grander floral display than is afforded by a fine specimen of this plant when viewed from above. The regular outline of the head, the glossy green of the leaves, which, when stirred by the wind, show the white tomentum beneath; the snowy rays and dark purple discs of the myriad flower-heads form a never-tiring source of attraction, while a grateful aromatic perfume is constantly exhaled. Olearia traillii, T. Kirk. A noble species of sparing occurrence on Stewart Island, but found also on Puyseygur Point. Brachycome pinnata, Hook. f. A small but elegant species discovered by Dr. Lyall at Port William, but not observed elsewhere until its recent discovery on the Canterbury Plains. Brachycome thomsonii, T. Kirk. The largest New Zealand species. Common near the sea on Stewart Island, and on Ruapuke, etc. Cotula traillii, n. s. A handsome littoral species, plentiful near the Neck. Allied to C. squalida, Hook. f. Senecio muelleri, T. Kirk. A fine species, allied to S. huntii, F. Muell., of Pitt Island. Only known on Herekopere Island, South Cape Island, and the Snares. ? Raoulia goyeni, T. Kirk. A small species only known at present from Mount Anglem and Rakiaua. Flowers not seen. Dracophyllum pearsonii, n. s. I am indebted to Mr. Pearson for a muchbranched specimen of a Dracophyllum, collected either on Codfish Island or in Chew-tobacco Bay. The plant is evidently erect, the leaves about 1″ long are close-set and densely imbricating, appressed to the branches, which, with the leaves, are about ¾″–1″ diameter. Flowers not seen. The habit differs so widely from that of any other species known to me, that I venture to describe it provisionally in the absence of flowers, and have great pleasure in attaching the name of its discoverer, to whom I am greatly indebted for numerous specimens of Stewart Island plants.

Gentiana saxosa, Forst., var. recurvata. A littoral form with prostrate or suberect stems excessively branched, flowers produced in profusion. Calyx divided fully two-thirds of its length, segments thickened, subulate, recurved at the tips. Found also on the outlying islets and on the Bluff Hill. Liparophyllum gunnii, Hook. f. Plentiful in swampy places on the low ground between Paterson's Inlet and Mason's Bay. Observed by Messrs. Petrie and Thompson at Port Pegasus. Myosotis antarctica, Hook. f., sub-sp. traillii. Mason's Bay. Myosotis capitata, Hook. f., sub-sp. albida. A rather coarse littoral plant which might almost claim specific honours. The radical leaves are much longer than in the type and on longer petioles; the cauline leaves are narrower and more distant, continued to the base of the inflorescence. The flowers are smaller, densely crowded, corolla white. Also on the outlying islets and on Campbell Island, but I am assured that it does not occur on the Auckland Islands. Dacrydium intermedium, T. Kirk, var. gracilis. Usually a smaller plant than the type, from which it differs in its slender branchlets and monœcious flowers. Plants 3–6 feet high are often laden with flowers. Two specimens of a small epiphytic orchid were obtained on the descent from Mount Anglem. It seems probable that they will form the type of a new genus closely allied to Burnettia and Chiloglottis. Juncus antarcticus, Hook f. (?) A plant doubtfully referred to this species for the present was collected on Mount Anglem. Scirpus (Isolepis) muscosus, n. s. A minute species, less than one inch in height, forming moss-like patches at the head of Paterson's Inlet. Also on the Bluff Hill. Scirpus (Isolepis) ebenocarpus, n. s. A tufted species resembling a luxuriant state of S. cartilaginea, but easily distinguished by the shining jet-black nuts. The Neck. Ehrharta thomsonii, D. Petrie. A small tufted species not uncommon in moist situations on the hills. Danthonia crassiuscula, n. s. A sparsely-tufted rigid glabrous plant 1′–1 ½′ high, with a lax broadly-ovate panicle, allied to D. ovata, J. Buch. Mount Anglem. Danthonia flaccida, n. s. A small slender species allied to D. buchanani Hook. f. and to D. pumilio. Mount Anglem, 2,800 feet. Poa walkeri, n. s. A remarkable plant, densely tufted, strict, with the panicle excessively contracted. Local. East coast, Stewart Island. I have great pleasure in attaching Mr. Walker's name to this distinct species, as a slight recognition of his ready assistance during my stay on the island.

Poa foliosa, Hook. f. The typical form; a noble species, producing an immense yield of herbage. Observed only on Herekopere Island; but a native informed me that it was to be found on several headlands near the South Cape. Plentiful on the Auckland and Campbell Islands. The total number of Phænogams observed on the island is about 380, one half being species generally distributed throughout the colony. From the small amount of variation exhibited in the plants of different districts, except under changed conditions, I see no reason to anticipate any large additions to this number, not more than from 12 to 15 per cent. at most, and in all probability less than 10 per cent.; but amongst these additions several species new to science may fairly be anticipated, as well as others at present known only from the Auckland and Campbell Islands. The absence of many species of general distribution is most remarkable, and in most cases not easy to be accounted for. Not a single leguminous plant has yet been observed, although Sophora and Carmichælia are found on the opposite side of the strait. Dodonæa, Melicope, Pennantia, Hoheria, are altogether wanting; still more singular is the apparent absence of Oxalis, Pelargonium, Daucus, Galium, Microseris, Scleranthus, and other genera. The spear-grasses of the South Island are represented by a single endemic species only, Aciphylla traillii. The white pine, Podocarpus dacrydioides, does not appear to occur on the island, the common name being applied by the bushmen to Dacrydium intermedium. Podocarpus spicata is extremely rare, and P. nivalis has not been observed. More remarkable still is the total absence of the beeches, although Fagus cliffortioides, F. menziesii, and F. fusca descend to the sea-level on the opposite mainland. The absence of the toatoa (Phyllocladus alpinus) is equally noteworthy. A striking peculiarity of the flora of Stewart Island is the occurrence at sea-level of numerous species usually restricted to alpine or sub-alpine situations. Amongst these may be enumerated:— Claytonia australasica, Drosera arcturi, Hydrocotyle muscosa, Liparophyllum gunnii, Actinotus novæ-zealandiæ, Veronica buxifolia, Olearia colensoi, Drapetes lyallii, Celmisia petiolata, Dacrydium colensoi, Senecio elæagnifolius, " laxifolium, Oreostylidium subulatum, Caladenia bifolia, Cyathodes empetrifolia, Centrolepis monogyna, Pentachondra pumila, Oreobolus stricta, Dracophyllum longifolium, " pumilio, Gentiana saxosa, Uncinia rubra, Drosera stenopetala, Hierochloe alpina.

The above list does not include several species which, although generally found on the mountains, are occasionally seen in littoral situations. Amongst these may be mentioned Hymenanthera crassifolia, Colobanthus billardieri, Pimelea lyallii, Danthonia raoulii, Haloragis uniflora, Senecio lagopus, Gentiana montana, Utricularia monanthos, Herpolirion novæ-zealandiæ, all of which are found on Stewart Island,** Olearia colensoi and Senecio elæagnifolius descend to the sea-level in the West Coast Sounds. but there is no other locality in the colony where so large an assemblage of alpine plants can be found growing at the sea-level. It has been hastily assumed that this remarkable feature affords direct evidence of a severe climate. Mr. Petrie, in his Notes on Stewart Island, after recording the occurrence of Donatia novæ-zealandiæ at sea-level in Paterson's Inlet, writes, “It is extremely remarkable that a plant which does not descend below 3,000 feet in the latitude of Dunedin should flourish at sea-level in that of Paterson's Inlet, and the fact bears emphatic testimony to the severity of the climate of Stewart Island.”†† Trans. N.Z. Inst., vol. xiii., p. 325. Again, referring to Senecio lyallii at intermediate levels in Port Pegasus, he writes, “The occurrence of this alpine plant at so low a level in Stewart Island, as well as its dwarfed proportions, give additional proof of the severity of the climate in this part of the colony.”‡‡ l.c., p. 328. It may be mentioned incidentally that I was unable to find Donatia at sea-level, and did not observe it below 2,700 feet. Senecio lyallii is usually dwarfed when growing on undrained peat soil; its favourite habitat is the side of a mountain stream, where it exhibits its greatest luxuriance, while specimens growing within a few yards on undrained soil are dwarfed and stunted. If this theory were correct it would be necessary to account for a still greater difficulty—the occurrence of a varied and extremely luxuriant arboreal vegetation, including arborescent ferns, under the conditions incidental to a severe climate. Nowhere in alpine districts are the alpine plants enumerated on the preceding page associated with a varied and exuberant forest growth, still less with tree-ferns as is the case on both sides of the low moory ground at the head of Paterson's Inlet. All the facts of the case are opposed to such a conclusion: the climate is remarkable for its mildness, and so far from being severe that it is much more favourable to a luxuriant plant-growth than that of many parts of the South Island, for instance the Canterbury Plains. It is true that the atmosphere is almost constantly moist, but snow does not fall on the lowlands, and frosts are not felt. Not only are many plants of mild climates cultivated in the gardens of the settlers, but the Clianthus, karaka, nikau, and other plants from the northern part of the colony, grow luxuriantly in the

open air. The peach is never touched by frost. Eucalyptus globulus grows luxuriantly—but it is needless to multiply instances. Similar testimony is afforded by the robust health of the settlers, their freedom from affections of the respiratory organs, and the rarity of deaths from sickness, as well as by the vigour of the children. To what cause then must the occurrence of numerous alpine plants at low levels be attributed? Unquestionably to the remarkable equability of the moist mild climate characteristic of the low lands. For it is especially worthy of remark that, while the majority of the alpine plants found at sea-level on Stewart Island occur also on the higher slopes of Mount Anglem, they are rarely to be found at intermediate heights below 2,000 feet. It is to this equability of climate that the occurrence of a few characteristic northern plants in the extreme southern part of the colony must be attributed. Lindsæa linearis, a fern common in the Auckland District, rare and local elsewhere, and altogether absent from the eastern side of the South Island** It occurs in somewhat small quantity on the western face of the Bluff Hill. is abundant and often luxuriant, growing side by side with Pentachondra pumila, Cyathodes empetrifolia, Centrolepis monogyna, etc. Eleocharis sphacelata, not unfrequent north of the Waikato, has a single habitat in the Taupo District: but in the South Island has only been observed at Okarito and Bluff Island. It occurs freely in deep bogholes at the head of the northern creek running into Paterson's Inlet. Dacrydium kirkii, only known elsewhere to the north of Auckland, occurs on Ruggedy, and, as I was informed, attains a large size in localities on the western side of the island, the trunks being 3 feet in diameter. It should, however, be mentioned that the identification is made in the absence of flowers or fruit. Campylopus kirkii, previously known only from the Great Barrier Island, is abundant in the low valleys between Paterson's Inlet and the west coast. Microlæna stipoides, a grass decidedly impatient of frost, is found at the Neck and other places. I do not propose to include a list of the Phænogams in the present paper, as one has been already given by Mr. Petrie, and although the number recorded by him has been nearly doubled, so much of the island remains unexplored that a complete list could not be presented; but I purpose giving a detailed account of the ferns and allied plants in a future paper. The equable character of the climate, especially in sheltered inlets of the sea, is so conducive to health, that in a few years it will doubtless attract many residents from amongst the wealthy merchants of Dunedin and Invercargill. The shores of Paterson's Inlet, Port William, Horseshoe Bay, and similar localities, will be dotted with villas, relieved by the

crimson glory of the rata and the pale racemes of the akamai. But speaking generally the quality of the soil is not of the first class, and with the abundance of land still waiting inhabitants in both the North and South Islands, it would be most wasteful policy to destroy the fine timber of the lower levels simply for the sake of forcing settlement. Except in the localities indicated above, a wise policy would retain the island as a timber reserve, for not only is timber abundant, but it is of more durable quality than much of that grown on the west coast of the South Island. Past attempts to force settlement on Stewart Island have resulted in the expenditure of large sums of money without any return, as may be seen in the Government barracks now decaying at Port William, happily in this case without requiring the continuous annual outlay rendered necessary by premature settlement in other localities.