\\\E-UN!VER% % I I 3 ^JOJIIVD-W^ %(OJIW3-JO^ HOKAUFO^ 9 I/ i I I ^IDS-ANCtUr^ if ~ $ y "%< -OF-CAllFO/5 !i * E-UNIVERS/A ^ g I ^OF-CAllFOty* IU=\I y 0Aavaain^ ^Aavaan-^ ^UN!VER% ^los-Awcncr^ 3 g/-^ -^ j^.iuwuan.!^ v^i 8 ^ HIBRARWfe I \ a a II 53 wU & % 1 1 - I _ BBW-SaV^ vflHAlNMVVv ^-UBRARY^. BHJNlVERtfA vvlOSAVCElfj> < ^ X-V_A^ ^ - - Q fytmwtf*' I 3 A\tf-UNIVERJ/A ^ | OF-CALIF(%, . . & ^. ^^ i t 5 ^ ^AHvaam^ ^. ^ " 0AEUKIVER% v^-lOS-ANCEl^ ' il^- 4 I i i ?3 O S? HWr*> ^\9 DUTCH N.W. NEW GUINEA. A CONTRIBUTION TO THE PHYTOGEOGRAPHY AND FLORA OF THE ARFAK MOUNTAINS, &c. BY L. S. GIBBS, F.L.S., F.R.M.S. With Four Plates and Sixteen Text-figures. (Read before the Newcastle Meeting of the British Association for the Advancement of Science, September 1916.) LONDON: TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. JULY 1017. Price 12s. 6 3U CONTENTS. PAGE INTRODUCTION 1 1. HISTORICAL ... ... ... ... ... ... ... ... 1 2. SUMMARY OF PREVIOUS WORK 2 3. SCOPE AND CONDITIONS OP PRESENT WORK ... ... ... ... 5 GENERAL CHARACTERS OF THE COUNTRY 7 1. TOPOGRAPIIY ... ... ... ... ... ... ... ... 7 a. The Immediate Shore-Line ... ... ... ... ... ... 8 1. " Korang " or Coral-Limestone Zone ... ... ... ... 9 c. Inhabited Zone of Foot-hills and Lower Ranges ... ... ... 11 d. Crests of Main Range and Lake Basins ... ... ... ... 11 2. METEOROLOGY 13 3. PlIYTOGEOGRAPHY 14 ITINERARY AND DESCRIPTIVE ACCOUNT OF VEGETATION ... 15 a. BEACH FORMATION : Immediate Shore-Line ... ... 15 1. INUNDATION FOREST BELT : " Korang " or Coral- Limestone Zone ... 16 c. SECONDARY ASSOCIATIONS : Inhabited Zone of Foot-hills and Lower Ranges 18 d. Low MOUNTAIN FOREST FORMATION ABOVE 7000' (to which syste- matic collection was limited) : Crests of Main Range and Lake Basins ... ... ... ... ... 21 PLANT ASSOCIATIONS OF THE LOW MOUNTAIN FOREST FORMATION- 27 A. FOREST ASSOCIATIONS ... ... ... ... ... 27 1 a. Mossless Forest. 1 ft. Aruucaria Forest. 2 a. Intermediate Mossy Forest. 2 /3. Dwarfed Intermediate Mossy Forest. 2 y. Shrubberies. 3. Mossy Forest. PAGE a. S.W. RIDGE .' 27 la. Mossless Forest, 7000-7500' (trees 40-50') 27 2 a, ft. Intermediate Mossy Forest, 7500-8500' (trees 20-40')... 28 3. Mossy Forest, 8000' (trees 10-20') 29 1. LAKE BASIN, 7000' 29 2 a. Intermediate Mossy Forest (trees 40-50') 29 1/3. Araucaria Forest ... ... ... ... ... ... 31 3. Mossy Forest 32 2 j3. Dwarfed Intermediate Mossy Forest (trees 10') 33 c. SLOPES OF KOEBRE MOUNTAIN ... ... ... ... ... 33 2 a. Intermediate Mossy Forest, 7000-7500' (trees 30-40') ... 33 2 /3. Dwarfed Intermediate Mossy Forest, 7500-8500' (trees 10-30') 33 2 y. Shrubberies bordering Crest of Mountain, 8500-9000' ... 34 B. OPEN "OPPORTUNITY" ASSOCIATIONS ... ... ... ... ... 34 a. S.W. EIDGE, 7000-8500' ' 34 Papuan Rest- and Camping-places ... ... ... ... 34 Small Landslips 35 b. MARSH BY $ LAKE, 7000' 35 Sand Pans with running water ... ... ... ... ... 35 Denser Sedge Growth in Boggy Areas ... ... 36 Shallow Standing Water 37 Grass-jungle on Edge of Forest ... ... ... ... ... 37 c. OPEN SUMMIT OF KOEBRE MOUNTAIN, 9030' ... ... ... 37 Cladonia Association ... ... ... ... ... ... 37 PHYTOGEOGRAPHICAL CONCLUSIONS 39 SOME PLANT ASSOCIATIONS OF THE N.W. COAST 49 SYSTEMATIC RESULTS 50 PRINCIPAL BIBLIOGRAPHY 58 SYSTEMATIC ACCOUNT OF THE PLANTS COLLECTED 59 1. ARFAK MOUNTAINS ... ... ... ... ... 59 2. DOREI BAT, ISLANDS TO HUMBOLDT BAY ... ... 185 EXPLANATION OF THE PLATES 223 \_The altitudes given are approximate^ A CONTRIBUTION TO THE PHYTOGEOGRAPHY AND FLORA OF DUTCH N.W. NEW GUINEA. INTRODUCTION. 1. HISTORICAL. THE history of the navigation and exploration of Now Guinea from the earliest times has been most exhaustively summarised by Dr. Wichmann in the two volumes of history included in " Nova Guinea." The following account is, therefore, strictly limited to records of botanical collections and observations in N.W. New Guinea alone. From historical times the N.W. coast of Dutch New Guinea paid tribute to the Sultans of Tidor, and was subsequently included in the Residency of Termite, which now administers the whole of the Sultan's dominions. Termite, the capital, has always been the centre of the N. New Guinea trade in Bird*-of-Paradise, the " Passaros del Sol " of the old Portuguese navigators, who widely distributed these highly prized objects both east and west, Oriental potentates and Moorish sultans with their courtiers vying in the possession of such treasured symbols of royal power and magnificence, which adorned the headgear on ceremonious occasions. Subsequent to this the golden returns from the spice trade in the Moluccas, a monopoly most stringently held by the Dutch East India Co., attracted the attention of other European nations. In the search both for wild varieties of the precious trees and suitable areas for possible exploitation, N. New Guinea offered the most promising field, the Dutch, from their base in the Moluccas, again holding most of the trumps. All navigation or detailed survey of the N.W. coasts of the country, with investigation into the numbers and condition of the inhabitants, dates from this period about the 18th century. In j-ecent years, thanks to the dictates of fashion and enhanced value, the volume of the trade in Birds-of-Paradise has again enormously increased, regulated, however, in the Dutch Possessions by most adequate measures, stringently enforced, for the protection of these beautiful creatures. All intercourse in N.W. New Guinea during the open season is dependent on this trade. Thanks to the enterprise of Tidorese, Malay, Arab, and Chinese traders, in whose hands it is chiefly concentrated, the coast Papuans have been brought more or less into contact with the outside world. This has resulted in a certain amount of intermarriage, as a ready means of con- solidating and extending trade relations in the chief centres of distribution, and also in spreading a knowledge of Malay, even in remoter communities. In the season the chief trading-stations swarm with the miscellaneous agents engaged in this lucrative business, whilst the Papuans are occupied in hunting in the interior on their own account, or for the various Ternate traders or agents, so that it is next to impossible to obtain quarters or procure native hunters or carriers. It is advisable to take this fact into consideration in connection with biological work in this part of the country. 2. SUMMARY OF PREVIOUS WORK. Geelvink Bay was first discovered and mapped out in 1705 (21, i. 138-152) by Jacob Weyland, who commanded the ships 'Geelvink' and ' Nova Guinea/ In the course of his work he touched at Dorei Bay, situated to the N.W. of Geelvink Bay. In 1775 Forrest (1, 95-114) in the 'Tartar,' a 10-ton ship, in which he had sailed from Balambangan to obtain nutmeg and clove trees for the purpose of introducing them into that island, l spent part of January and February at Dorei, enjoying fine weather all the time, and his excellent straightforward account is quite one of the best relating to this place. The 'Tartar' lay off Wousi, where Forrest describes the whaleback Papuan houses spreading over the water, as they are still to be seen at the present day, and also the Arfak Mountains rising below Dorei. He even mentions Oranswari and Wariap as two inhabited places along the coast to the south, and notes the correct time to reach them by fast sailing in a favourable wind. He states that the people had many prnus, and were said 1 Off N. Borneo. to deal honestly with the Chinese who traded with them, conditions which still prevail. In July 182-1, Duperrey (21, i. 316) in the 'Coquille' spent over a fortnight at Dorei, and the first collection of plants from there was made by Lesson, the surgeon accompanying the expedition (2, ii. 534). Dumont d'Urville (3, iv. 578-612) in the Voyage of the ' Astrolabe ' gives a verv good general account of Dorei Bay, where he stayed in August 1827, after the discovery of Humboldt Bay, which was named, but not touched ;it. He mentions taking in water " au limpide ruisseau de Wirsi," and refers to the very rainy weather. The local Papuans are described as living in terror of the Arfakis, some mountain tribes established above Wousi on what is now the site of Manokoeari, who were always treacherously attacking the coast people, one such attack occurring duriug the stay of the ' Astrolabe/ Lesson, who accompanied him, made another collection of plants. Wallace in 1858 (5, ii. 298-326) spent three months and a half, from March to July, at Dorei Bay collecting zoological specimens, and was there when the S.S. 'Etna' (8, 78), fitted out by the Dutch Netherlands Indian Government for ethnological and natural history investigations, put into the bay in June of that year. Some of the 'Etna's' company attempted to ascend the Arfak (8, 73) from the N. side, but turned back after reaching 1500'. In September 1872, D'Albertis (9, 67-71), having spent three months at Andai with Beccari making zoological collections, succeeded in reaching Hatam, an Alfuero village, situated at 5000', three days' journey from Andai, where he spent a month, enduring great privations. This intrepid explorer was the first European to penetrate into these mountains and reveal their ornithological treasures, for, of the magnificent collection of Birds-of- Paradise and other types obtained by him, nearly all proved new to science. In June 1875, Hatam was again visited, this time by Beccari (11, 35-38), who spent a month in this inhospitable place, much hindered by rains and swollen torrents. He attained 2040 in. in exploring the surrounding heights, and brought back, in addition to much valuable zoological material, a large collection of plants unfortunately not yet fully worked out. Rosenberg in 1869 and 1870 (10) made several journeys to Dorei and the adjacent islands of the N.W. coast, of which he gives an interesting description. Accompanied by Malay hunters he spent three months at Andai, staying with the Missionary VVoelders, with the intention of attaining Hatam (10, 99). Several abortive attempts were made, and he claims this achievement for his hunters, a claim since disproved (21, ii. 142). A vocabulary of the Noemfoer (coast) and the Alfuero (mountain) languages, with some excellent original drawings, are included in his book. 32 Teysmann, the well-known Dutch botanist, visited Dorei in 1871 (14, 61-95), and also touched at Andai and Wariap in coasting up Geelvink Bay, making large collections of plants. In 1891 one of Veitch's orchid collectors, David Burke (21, ii. 539), a gardener, made his way to Small Hatam, a remarkable feat, referred to by E. St. Vraz, who in 1896 (15, 232-235) spent several weeks in Great Hatiim, some distance beyond Small Hatam, which he states was D'Albertis' and Beccari's objective. Though much hampered by rain and the usual difficulties with carriers and hunters, the standing barrier to all work in the mountains of this country, St. Vraz's account is detailed and interesting. In 1898 Manokoeari (21, ii. 677), the old Alfuero "campong" 1 on Dorei Bay, was raised to the status of a Government Station, to which Mr. van Oosterzee was appointed Assistant Resident, an appointment held until 1913, when he was obliged to retire through ill-health. In April 1904 (17, 998-1021) van Qosterzee visited the Angi lakes, the first European to explore the more southern portion of the Arf'ak. On arriving at Ternate, on the way to Manokoeari, I heard with great regret of the serious illness of this very able administrator, who during his 16 years' residence at Manokoeari had made himself familiar with the Papuans and their language, his sympathetic rule giving him unusual influence with the chiefs, or "Koranos" as they are called. Manokoeari also bears witness to his initiative in the beautifully situated Residency and " Pasangrahan," 2 in avenues of Casuarinas shading the level white roads, in the efficient water supply, substantial pier with well-arranged godowns, and the broad street of Chinese " tokos " 3 . In 1908 Pratt, the well-known collector, with his two sons, spent some time at the Angi lakes. They were accompanied in the first instance by van Oosterzee, who, however, returned immediately to the const. In 1912 (25, 77-78) Gjellerup, attached ;is Officer of Health to the Exploration Detachment 1911-12, proceeded to the lakes, accompanied by a mining engineer named Hubrecht. They spent April and May in that region, where the former made important botanical collections, now in course of publication in " Nova Guinea/' Unfortunately, many specimens were lost owing to the desertion of his carriers. I had the pleasure of meeting Dr. Gjellerup in Java soon after his return from New Guinea, and acknowledge with keen appreciation his generous kindness in giving me the fullest benefit of his experiences, not only in relation to valuable details in organization and the areas to which he had devoted most attention, but also for most welcome data on Papuan 1 Campong, a native settlement or village. a Rest-house. 3 Shops. idiosyncrasies, with special reference to the mountain tribes and the best method of dealing with them. On my return from Wariap in 19 L3 I was fortunate in finding Mr. Pratt and his sons at Manokoeari, where they had arrived to organize another expedition to the Angi lakes. I gladly welcome this opportunity of thanking them for permission to reproduce the successful photographs subsequently taken in that region, including several purely botanical subjects selected especially at my request. My own results there had not proved satisfactory. 3. SCOPE AND CONDITIONS OF PRESENT WORK. The chief objective of the present work was a direct outcome of results obtained and observations made on Mt. Kinabaln in British North Borneo in 1910. The great interest in the higher regions of that, mountain lies in the extra- Malayan facies of the vegetation and elements in the systematic composition of the flora. The predominance there of species in Orchidacese, Myrtacea3, Ericaceae, chiefly Rhodoilemlrons and Vacciniums, is such a striking feature, that one felt it must be derived from some more continental type of vegetation than the limited occurrence on an isolated mountain could account for. Celebes and the Moluccas did not suggest sufficient scope for the proposed key to the solution ; but the mountains of New Guinea with their huge area, approximate rainfall, and great altitude offered a perfect basis of comparison ; while Dr. Beccari's description in "Malesia" of his work and observations on the plants found by him on the Arfak mountains showed a very close approach to the phytogeographical facies of the Kinahnlu " Massiv." This impression was further confirmed by the account of Mr. Pratt's first expedition to the more southern region of the range, and a few plants collected by him in an open marsh by the larger of the two Angi lakes, kindly shown to me at Kevv by Dr. Stapf, finally led me to decide on that locality for comparative work. Through the broad-minded interest and influence of my friend the late Sir Kenelm Digby, permission for the expedition was accorded, through the kind offices of the Foreign Office, by the Government, of the Netherlands. I must express my deep gratitude to His Excellency Mr. A. W. F. Idenburg, then Governor-General of the Netherlands-India, for the extreme courtesy of his reception and the interest he evinced in the scientific objects of the projected work, the success of which was assured, thanks to his generous assistance and detailed instructions. My thanks are also due to Mr. 0. Lulofs and Mr. J. C. Kielstra for much kind attention during my stay at'Buitenzorg. At Ternate the Resident, Mr. Gh. L. J. Palmer van den Broek, was again most kind in forwarding my plans, and from Mr. L. J. J. M. (now Captain) Tabbers, Acting Assistant-Resident at Manokoeari, I met with every sissistance which military training in grasp of the situation, judgment of men, and appreciation of fact and detail could do to ensure success. To Captain J. .F. E. ten Kloosten, who most kindly procured me Dr. Gjellerup's sketch-map of the Angi lakes from the military carto- graphical office at the Wousi Bivouac, with permission to reproduce the same in this work ; to Mr. J. W. Langeler, of the Dutch Navy, attached to the Mamberamo expedition ; and to Mr. F. J. F. van Hasselt, Chief Missionary at Manokoeari, I am further indebted for much help and thoughtful consideration. Any difficulty in reaching Wariap, about 60 miles south of Manokoeari, the best point of departure for the mountains, was obviated through the kind offices of Captain J. V. L. Opperman, Commander of the Mamberamo expedition 1913-1914. He gave me, with the escort kindly provided by the Dutch Government, a passage in the Government boat the ' Valk,' which was conveying his expedition to its destination. The ' Valk ' landed us at Waren, 10 miles south of Wariap. As there had been recent fighting between the coast and hill people, Mr. Tabbers judged it advisable to send an unusually strong escort, com- prising five Amboinese armed policemen, " Pradjoerit," and ten convicts, " Orang rante," attached to them, to help in the carriage of provisions should the Papuans desert, according to their usual custom. It was no doubt partly owing to these two circumstances that all my carriers remained with me on this occasion ; I noticed great reluctance to stray far from the protection of the guns. The sergeant in charge, always called " Serzan," a Timorese, who had served for 20 years through the Achin campaigns, was a. most capable and tireless little man, a thorough soldier, not only very clever in handling his own men, who did him credit, but also in dealing with the Papuans, not such an easy matter. The " Orang rante," all enormous men, who made a great impression on the Papuans, turned out very handy and willing. One of them acted as my cook, both in the mountains and until my departure from Manokoeari, while several became quite expert in carrying out botanical processes. It is a great tribute to the humanity of the Dutch rule that men like these Malays, to whom confinement is death, should be sent to expiate their offences, chiefly crimes due to jealousy and gambling, in outlying stations, where they have congenial work under easy and almost independent conditions. Escape is out of the question, as the Papuans shoot at sight any stragglers in the forest so it is not even necessary to guard them, while the best behaved are allowed to earn money by gardening or acting as " boys " to the civil and military officers. GENERAL CHARACTERS OF THE COUNTRY. 1. TOPOGRAPHY. The north-west coast of New Guinea is bleak, mountainous, and sparsely inhabited. The mountains, from 5000-7000', rise abruptly from the sea, there being little sloping foreground and few small bays to shelter FIG. 1. N.W. NEW GUINEA. the schooners and Papuan praus, which only trade during the favourable inonsun. There is no good anchorage tor large boats between Sorong, an island on the extreme north-west, the first point of call, and Dorei Bay, a fine harbour, consisting of two bays, sheltered by the islands of Mansinam and the much smaller Meoswar. A low mangrove spit, on which a Papuan " cainpong " is situated, separates the two bays, the further one being much smaller, very shallow, and probably rapidly silting up. Manokoeari lies in the centre, on the lower slopes of a very recent u korang '' or coral limestone range about 500' in height, while to the east the coast runs out into a low peninsula, where most of the Papuan plantations are situated, the soil there being much better than that of the sterile " korang " of the range. To the south rise the Arfak Mountains, 9000' high, which run in parallel ranges down the western shore of Greelvink Bay, at about forty to fifty miles inland, except at one point Tanaroeboe, one day's journey along the beach from Andai, where spurs from the mountains abut steeply on to the sea. How far south the range extends has not been determined, nor whether it is continuous with the mountain chain of the north-west coast, as some writers have inferred. The main buttress of the Arfak rises above Andai, a Papuan " campong " at the mouth of the Andai River, about four to five hours by native prau across the bay from Manokoeari. From Andai there is a tract to Amberbaki on the north-west coast, which passes through Hatam, a four days' journey according to Rosenberg (10, 79). The two small Angi lakes, to the vicinity of which this collection was limited, lie at an altitude of 7000' and 9000' respectively, on the southern portion of the range. The lakes are accessible from several points on the coast, of which the immediate shore-line only is sparsely inhabited, the few inhabitants being restricted to the beach, hence their name, " Orang pantai " or " beach people/' A huge intervening low-lying belt of coral limestone or " korang " extends from the beach to the loot-hills of the Arfak. This tract of country, intersected by the alluvial terraces and large inundation-areas of the rivers, which pour down from the mountains in the rainy season, is devoid of inhabitants and suggests very recent elevation. The main topographical features of this part of the W. coast of Geelvink Bay may therefore be roughly divided into four zones, which also corre- spond to the general plant-formations : (a) The Immediate Shore-Line ; (b) "Korang" or Coral-Limestone Zone; (c) Inhabited Zone of Foot-hills and lower Ranges; (d) Crests of main Range and Lake Basins from 7000'. a. THE IMMEDIATE SHORE-LINE. The immediate shore-line is sparsely inhabited, though many fine rivers and torrents sweep down from the mountains. With the exception of some mangrove formation, between Dorei Bay and Andai, and Oranswari and Wariap, the beach sweeps southward, consisting of big stones, shingle, or 9 loose sand. Large blocks of coral, as it' just broken off from the reef, lie at the water's edge, where the surf beats ceaselessly and the dip of the beach is general!)' too steep for native praus to land without running the risk of being smashed to pieces. It is the presence of landing-places, with good water, whicli limits the stages in the tedious journey along the beach from Andai to Sjari. The rivers in flowing into the sea either break up into many shifting mouths or are barred by banks of sand or shingle into semi-lagoons, with only a small exit to the sea. Most of the trees fringing the beach are prostrate or semi-prostrate, as if torn up by the force of the waves. Stagnant lagoons, impenetrable bog, and shallow standing water occur in parts, while the undergrowth is covered and the ground strewn with seaweed, evidence of the retreating swirl of great waves. The natives told me that when the north wind blows the sea washes all over this belt of country, fish being often found stranded on bushes far inland. The few small native ; 'campongs" are placed just above the beach, whore sand-banks have accumulated, on which some Casuarinas mark the permanence, as at Wariap and Waren. I returned from Wariap by the beach in December 1913, the first time this journey had been made by a European, to be followed by Mr. Pratt in April 1914, at the height of the north monsun. He described the whole region as then more or less under water, the rivers, pouring down from the mountains in floods, being beaten back over the land by the huge surf raised by the north wind, which bars the exit of their waters to the sea. This interesting observation accounts for the shifting river mouths, and also explains why the native habitations are generally so far from water and so few in number. 1. " KORANG " Ott CORAL-LIMESTONE ZONE. Behind the beach the low-lying belt of " korang," covered with forest, stretches uniformly from the coast to the foot-hills of the Arfak, a sterile porous formation showing so little depth of soil that it gives the impression of walking over a reef. Rosenberg (10, 80) in 1870 refers to the recent appearance of this coral- limestone area, which he concluded must be still rising, and he quotes the older inhabitants of Andai as saying that they remembered low scrub where the forest now stands. Van Gelder (20, 94) considers that a gradual rising of the whole of the north coast of New Guinea is taking place, or a lowering of the sea-level, which amounts to the same thing. He found evidence of this fact at 10 Manokoeari, in the presence along the beach, at slight elevation above the sea, of an undoubtedly very recent " Schelp Gonglomeraat " (shell-limestone) that must have been formed in the surf zone. The rivers cut their way through alluvial deposits of mud or sand, or form great open spaces their present inundation-areas, all sand and stones in the dry season, but which, in the wet, are lakes of standing water. MAP OF ANOI LAKES (reduced from Dr, K. G. Gjellerup's sketch-map). Dr. Gjellerup's route. At Wanessi, the head of such an area on the Momi River, a day's journey from Wariap, the remains of an old " campong " can still be seen, where some of the hill people tried to establish themselves, but were forced to retire to Wariap on account of the floods. This belt is uninhabited. 11 c. INHABITED ZONE OF FOOT-HILLS AND LOWER RANGES. Once on the foot-hills the character of the country changes : the ground is broken by rocky outcrops and boulders, and the streams are now mountain torrents, plunging over great boulders which form their beds. Crossing the Soedomi River, a rocky stream which falls into the Momi, one mounts steeply up the flanks of the spurs to the crests of the ridges, which are cleared. Native houses are first seen at about 1000', and from thence upwards, on the crests and slopes of the ridges, where the soil is deep and easy to work, there is evidence of extensive cultivation, past and present. All the region of the subsidiary spurs and lower ranges seems to be inhabited. From commanding views it is easy to trace the line of the ridge, by which the approach to the lakes is made from Sjari, by the smoke rising from the houses, which are always perched in a conspicuous position, probably for the double purpose of observation and defence. Only in the case of one or two head-hunting tribes, who seem to inspire the whole neighbourhood with terror, were the houses situated on the lower slopes. On the third, or Sjari ridge, from the Soedomi River, one looks down on to the Momi again, which rises in these mountains at right angles to its coast course. From below Soekoem the course of the Momi River (fig. 2) is erroneously indicated ; it should be that suggested for the Runsiki River (fig- 2). Van Oosterzee (17, 999), who went up from Sjari, 20 miles along the coast from Wariap and 10 miles south o Warcn, mentions crossing the rivers Waren and Waidiri, both flowing into the Momi, but the Waren enters the sea at Waren, halfway between Sjari and Wariap. The Soedomi was the only river we crossed, coming up from Wariap, till the Momi was again met with as a rocky torrent. Crossing to the west bank of the Momi we climbed another long spur in a westerly direction, till the latter joined the crest of the main range at 7000', called the S.W. ridge in this work, where the inhabited zone suddenly ceases, a fact probably determined by the steeper gradients and narrower crests, and the exposed and poor gravelly soil. d. CRESTS OF MAIN RANGE AND LAKE BASINS. The route to the lakes lies along the narrow crest of this ridge, marked as " Bonjas Gebebergte " on fig. 2, which runs S. t S.W. from 5000-8500' in height, bounded on the west by a valley, said by the Papuans to be that of the river issuing from the largest of the two lakes (Tilaan of fig. 2), and on the east by short spurs which run down to the Momi valley. The Angi lakes are situated at about 7000' and 8000' respectively, separated on the west by the long Koebre Mountain, 9000' high, and 12 bounded to the east and the south by the S.W. ridge. The largest lake is called the Wannasin Bean or " Woman " Lake by the Papuans, in contra- distinction to the higher and smaller one, the Warmasin Snoon or "Man" Lake. These names are so well recognised that the Malays with me invariably referred to each lake as "Prempuan" or "Anak," the Malay translation. According to Gjellerup (25, 77), the greatest length of the lowest or " Woman " lake is 9 km. south-west to north-east, and the greatest breadth is 4 km. on the south side. This lake is so deep that van Oosterzee (17, 1010) found that, on the south side, at a few hundred in. from the edge, he could no longer touch the bottom after 90 m. Grjellerup (25, 77) states that eels are the only living fish in it. On the north side the con- verging hills gradually slope into a marsh, the exit of the Tilaan River, which, according to both van Oosterzee (17, 1018) and Gjellerup (25, 77) joins the Ransiki River, which is stated by the former to fall into the sea at Wariap, obviously a mistake for the Momi River. The western shore of this lake is inhabited by a small Alfuero tribe, their communal houses being in four groups on the extreme edge of the water, in separate bays, the " kebuns " stretching up the slopes behind. Further on, the sides of Koebre rise steeply from the water, while to the south a low ridge, about a couple of hundred feet high, connects with the S.W. ridge, which forms the boundary to the east. The smaller or " Man " lake is reckoned by Gjellerup (25, 77) as roughly about 4 km. from the " Woman " lake, and he gives its north and west sides as inhabited. Van Oosterzee (17, 1013), who spent most of his time on the upper lake, estimated it to be 7 km. long, and 2 broad in the southern portion and he counted some thirty houses on the western shore with about 1000 souls ; the houses seemed to me more numerous to the south- west, and I hardly think there can be so many inhabitants at the present time. He mentions the morass at the north end of the lake, also rafts on its surface, the cultivation of potatoes on the shores, and that eels are the only fish found in its waters. St. Vraz (15, 234), while at Great Hatam, heard from the natives that at four days' journey south there was a large lake called " Tschemti," with many fish and crocodiles, on whose banks Manikianos lived. Inquiring of some of the Papuans with me, who had been up at the lakes before, if it were possible to reach the north coast and Manokoeari from the upper lake, they answered " Yes," and that it took twenty days, but, of course, in their estimate of time they would lose count beyond a certain number. Van Oosterzee met some Papuans (17, 1010) by the lower lake, who told him they had come from the north, and Gjellerup (25, 77) also mentions a way to the north coast, facts which agree with St. Vraz's information. As I heard it was possible to return to Wariap in two days by keeping to the S.W. ridge, I decided at once on such a welcome alternative to the 13 & fc I CO CN t- co CO 2 CO C^l CO CD S CM (N CO ^f O s co cs i-H C5 3 0. i five days spent in coining up. Our return route therefore followed the S.W. ridge, leaving the spur by which we had ap- proached it from the Momi to the right. Continuing due north we gradually de- scended to about 5000', then bearing east along a lateral spur which finally dipped very steeply to the rocky bed of a tribu- tary of the Moini. After crossing the latter, we emerged on to the north bank of the Momi River, whore we bivouacked for that night, returning to Wariap the next day. The chief difficulty of this somewhat strenuous route is the want of water, there being none between the vicinity of the lakes and the tributary of the Momi. According to fig. 2 this route would follow the " Bonyas Gebebergte," which join the spurs of the Arfak indicated on the sketch-map to the north of the sup- positional position of the Ransiki River. The altitudes in the sketch-map, judging by the character of the vege- tation, seem somewhat underestimated. 2. METEOROLOGY. The following facts referring to the north-west coast are taken from Braak's (24, 210) paper on the climate of New Guinea, which summarizes all the infor- mation at present available. On the north coast the east or dry monsun prevails from May and June to September and October, though even then much rain may fall. In Mano- koeari the east monsun is the driest time (24, 219). It is probable that in the interior the difference between the cha- racter of the E. and W. monsun is much less pronounced. In N. New Guinea it is at present not possible to give actual figures, though 14 it may be accepted that the temperature throughout the whole year is very constant, and that the mean variation is probably between 26 and 27 C. (24,223). Braak's table (24, 221) quotes the average rainfall for several years for Sorong, Manokoeari, Djende on Roon Island, and Windesi to the south of Geelvink Bay. In both the latter stations the rainfall is much higher than on the more exposed N. const at Sorong and Manokoeari. In the historical summary the prevailing weather has been quoted whenever recorded, and it will be seen to vary considerably. At Mano- koeari, January and June were considered the wettest months. During a stay there in November 1913 the weather was very fine; jbut on my return in December the rainy season had set in and there was rain every day, generally in the afternoon, the temperature being markedly cooler. In the intervening period, spent at Waren and Wariap, on the coast of Geelvink Bay, in the Arfak, and on returning- to Manokoeari along the coast, only three half-days of rain were experienced, and these occurred on the way up and while at the lakes, where the rain was also accompanied by strong \A'ind. Otherwise the fine still weather was a constant source of amazement to the Papuans, the rainy season being well overdue, and they attributed this lucky chance to the fact that sometimes at the ch;mge of the monsun a short halcyon period of fine weather sets in. At Waren and Wariap a very strong N.W. breeze invariably sprang up between 3 and 4 P.M., causing a sudden fall of temperature. Unfortunately my only available calendar was lost at Waren. Con- sequently, no readings were taken either there or at Wariap, and only one or two in the mountains, which were as follows : On the Momi River, 3500'. 75 F. 6 P.M. On the crest of the S.W. ridge, 8500'. 68 F. 4 P.M. On the "Woman" lake, 7000'. 60 F. 7 A.M. ; and on the following day at the same hour, 55 F. Up at the lakes it was always cool in the early morning, the sun being very hot in the middle of the day, but cooling down in the afternoon and at night, though warmer thau at the same altitude on Kinabalu. Both Pratt and Gjellerup reported very wet weather during their several stays at the 3. PHYTOGEOGRAPHY. The general plant-formations, as is shown below, agree in broad outline with the topographical zones already described. They are : (a) Beach Formation : Immediate Shore-Line. (6) Inundation Forest Belt : " Koran g " or Coral-Limestone Zone, 15 (c) Secondary Associations : Inhabited Zone of Foot-hills or Lower Ranges. (d) Low Mountain Forest above 7000' : Crests of Main Range and Lake Basins, 7000-9000'. All systematic collection was limited to the last formation. ITINERARY AND DESCRIPTIVE ACCOUNT OF VEGETATION. (a) BEACH FORMATION. Permanent sand-spits, Waren and Wariap. The beach at Waren forms a long sweep on ench side as far the eye can reach. A plantation belonging to a Japanese, who with his son had permission to accompany me to the lakes, was situated at the mouth of the Waren River, which, dammed up by a sand-bank, formed a green lagoon, with only a very narrow outlet to the sea. A clnmp of Casuarina equisetifolia proved a certain stability, but the river-exit, with the dip of the beach, must be always in a state of flux with each N. monsun season. The great accumulation of s;md to which Waren and also Wariap owe their security from the inroads of the surf, must be due to the amount brought down and deposited by the rivers at their mouths. Where the beach broadens out in the immediate vicinity of Waren, a Pes-caprcB association with Tacca pinnatifida is formed, to be succeeded by typical beach-shrubs, like Tliespesia populnea, Canavalia obtusifolia, Sccevola Koenigii, Vitex trifolia, Clerodendrun inerme, Premna nitida, and a Gmelina, probably villosa, which must successively bank up the sand against the wash of the surf, as the Japanese had cleared behind them and planted coco-nuts on the pure sand, with cotton and pineapples asundercrops. This is the only spot along the coast besides Wariap where such a risk could be taken. Where the beach was lower and narrower, the surf w:ithed through beach-jungle or under Casuarinas to the overhanging fringe of forest trees. Sub-emerged Beach. It is several hours from Waren to Wariap along the coast to the north, by what may be styled a sub-emerged beach. A little beyond Waren the sand decreases in volume, strand plants disappear, and the beach narrows considerably. Huge trees of Barringtonia speciosa lie prostrate to semi- prostrate over the sea. Young plants of Pandamis &\>.,Draca>na angustifolia, an immense Crinum, probably C. macrantheruin, with giant stools of Asplenium Nidus, no doubt displaced from the branches of trees as they fell, crowd the ground, all dusted over and growing plentifully amongst much water-washed debris and plant detritus, both terrestrial and marine, 16 which, with myriad prostrate Barringtonia-seedVmgs, attached by the one anchor-root, all pointing seawards, bear witness to the force of great waves retreating from their rush inland. In many places this undergrowth was so thick, or the prostrate trunks so numerous, that it was easiest to walk through the surf, outside the branches of the fringing trees. The rivers, where there is no sand to bar their exit, form small mangrove swamps at their mouths, which have to be waded through. In one of these small swamps a tree covered with a vine of Mucuna Kratkei,a\so known from the N.E. and S.W., whose numerous huge racemes formed a dome of brilliant red flowers, was a magnificent sight. Wariap. Here, where the people all remembered Mr. van Oosterzee and the Pratts, I was welcomed as an old friend the "korano," a very fine man physically and quite a personality, and the " guru " (teaclier), to whom I had a letter from Mr. van Hasselt, having already paid their respects at Waren. It was arranged that the "korano," Manao, should act as guide to my party to the Likes, and the Wariap people of themselves offered to accompany me as carriers, promising to remain as long as I stayed there a promise sealed on " Pinang " and " Zabacco," as they call the latter, and faithfully kept. " Pinang " replaces betel-nut on the coast of N. New Guinea, being obtained from the wild Areca macrocalyv Zipp. (12, i. 18) and eaten with lime and the fruit of Piper Siriboa (14, 69). Wariap, situated on a sand-spit through which the Momi has cut its broad way on one side, forming a good harbour for praus, while on the other Casuarinas are massed, is quite a large and busy " campong," where much prau-building and making of Pandanus mats (nokes) and sago-holders is carried on. The long whale-backed houses are built above the beach, on a level spit of very fine sand, which, overgrown with grass and Pes-Caprce, is broken by shallow green lagoons shadowed by a jungle upgrowth of Thespesia populnea, Abrus precatorius, Ccesalpinia Nuga, Wedelia tiflora> etc. (V) INUNDATION FOREST BELT. Just behind the beach formation this forest forms a huge unbroken green wall, in which the pyramidal branching of Terminalia Catappa is easily distinguished from the outside, whilst most of the trees are covered with the heavy green curtains, falling straight from the crowns, of Zanonia macro- carpa, a Cucurbitaceous liane. In this forest Ficus, Macaranya, and Artocarpus sp., the latter with enormous leaves often about 1 m. long and m. or more broad, mostly predominate their trunks screened with immense fronds of climbing ferns, spreading radially all the way up, or Epipremnopsis Hugeliana, fiaphidophora Peepla, other Philodendron spp., 17 Piper Forstenii, Potlios, and various large-leaved root-climbers. Piles of these huge leaves accumulate under the trees, to be dispersed by the floods of the rainy season. A giant Korthalsia, its interlaced stems scrambling up and down the trees or spreading in tangled mass over the ground, was the only " rotan " seen. Asplenium Nidus was abundant ; but epiphytic orchids were lew and far between, and the absence of flowers or fruits was most striking. In this forest there is little undergrowth. Sodden leaves mostly fill up the interstices of the porous korang. Where light shoots through, some thin grass or patches of the creeping Geoplnla renifornris and Ilemigraphis re plans, or single specimens of the small semi-herbaceous shrub Amarocarpus Wichmannii, with horizontal dorsi-ventral branches, appear ; but the most conspicuous objects are the huge Z&ntaua capsules, the size of large pumpkins, in all stages of decay, scattered over the " korang," which is so porous that surface-water soon drains through; but where there is standing water, groups of Sago Palms occur, often forming swamps covering large areas. A track from Waren to the Arfak runs for a couple of days through Sago swamps. Native Plantations. These are dotted through this forest where little islands of soil accumulate, as at Waren, or alluvial deposits have been formed by the rivers, as at Wariap. Here the well-stocked " kebuns " 1 surprised me by their extent, many kinds of bananas and plantains, coco-nuts, papaya, cassava, " kladi," 2 " ubi," 3 "labu," 4 egg-fruit, and various " sayur" 5 etc. being grown, with clumps of bamboo, probably planted. On passing through this " korang " zone we followed a new route, keeping to the south bank of the Momi, so obviating the necessity of crossing the river, which is rather deep at its mouth. This track passes through the Wariap plantations and the sterile "korang" forest beyond, cutting off a great angle of the river. Then it skirts the shifting banks of the river as the latter cuts through alluvial deposits of mud and sand, or crosses wide sand-banks, the splash of crocodiles heralding our approach. These sand-banks afforded a fine view over the Momi, showing Camarinas and the symmetrical Terminalia, the latter veiled in all-obliterating Zanonia, backed by the distant mountains. Thence the track alternates between the dome-like " korang " forest and the thicker undergrowth of alluvial flats, or, when nearer the river, over oozy slime which spreads over the stems and leaves of a small Licuala palm and the giant KorlJuilsia, the chief under- growth in such areas. 1 Gardens. a Colocasia antiquorum Schott. ' Sweet potatoes. 4 Gourds. ' Vegetables. C 18 Momi Inundation Area. From the forest one emerges into the blazing sunshine, on an open plain, which it takes several hours to cross, all stones, gravel, sand, and " lalang," dotted with small trees of Casuarina equisetifolia. This open space forms an inundation-area of the river, and is under water during the rains. Amongst the stones Geodorum pictum, a pink-flowered orchid, was characteristic, and Peristylus goodyeroides was found among the " lalang " which covered the sandy areas. The most remarkable feature was abundant clumps of a new Pteris, P. bamliisoides, with erect rhachises resembling stems rising from an underground rhizome, each rhachis about 2 m. in height, clothed \vith segments from the base to the apex, originally bilateral, becoming spiral later through the twisting of the cortex. The species was also seen near Wariap, in passing through the " kebuns." Professor Bower, to whom I submitted this interesting new fern, suggests that " the general habit might possibly compare with that of Pteris grandiflora, in so far that both are probably creeping rhizomatous." Pandanus Trees. The second day, while still in the " korang " forest, we passed through a striking group of old Pandanus trees, about 30 in. high, each rising out of the forest on numberless grotesque stilt roots for about a quarter of the height, succeeded by a straight stem with much-branched top ; the old leaves hung in limp masses from every possible resting-place and strewed the wet ground underneath. These weird trees gave a pregnant impression of the scenic possibilities of this genus under primitive conditions. Even the Malays were impressed ; but the Papuans spoke of similar isolated groups scattered through this forest. These groups possibly represent the first vegetative covering of the " korang/' displaced later by more rapidly growing dicotyledonous forest trees, as only those plants which could respond by equal vertical growth and so maintain the same level as their competitors would have a chance of survival under such enveloping conditions. No fruit was seen on the trees or underneath them. (c) SECONDARY ASSOCIATIONS. Once on the foot-hills the character of the forest changes, the sodden effect of the "korang'* belt works out, and it loses the mud-washed look and steaminess suggestive of constant inundation. The ground, rocky and broken, is strewn with Sapindaceous and Anacardiaceous fruits, red Pometia, and other brightly-coloured kinds, and undergrowth in the shape of plants and shrubs appears, Musscenda frondosa being general. The foliage of both trees and lianes is less monstrous and more varied in form, while a graceful epiphytic flora puts in an appearance. From time to time small cleared 19 spaces were reached, evidently known rest camps, where we halted ten minutes to rest the carriers. After a tedious climb up the flanks of the spurs, clothed in high forest, one emerges on to cleared narrow ridges, covered with long grass, showing the first signs of cultivation, where a beautiful view opens out, on the one side to the blue island-dotted waters of Geelvink Bay over the foot-hills and the flat inundation-belt just passed through, which spreads out like a green table, and on the other side across the Momi valley to the central mountain range with its many outlying spurs. After more climbing, bamboo thickets evidently planted, with the magic plant of Malaya, Justida Gandarussa, never known to set seed, further confirm the impression of man's vicinity. Fine forest succeeded these abandoned gardens, from which we emerged on to old " kebuns " on the broad crest of the lower Serao range. Here were the Serao houses, surrounded by present cultivation, where we were very well received by the korano and his charming wife, the prettiest Papuan woman seen, though many are nice-looking. The Serao people great friends of Manao's cordially invited me to sleep in their house ; but as all the Papuans with the Japanese streamed in, to say nothing of the original inhabitants, I decided to camp outside in a newly made " kebun," with the " Pradjoerit" and " Orang ranteV' The korano's house was very large, with split-bamboo flooring and a few small partitions, while against each side a narrow strip was thickly sanded over for fires. Opposite the entrance a second door opened on to a balcony, commanding a lovely view over the dip of the ridge to the immediate Momi valley and the further spurs of the Arfak. A house inhabited by a head- hunting tribe was pointed out on the slopes below. Native Plantations. In the " kebuns " the luxuriance of the crops and method in cultivation is surprising. Sweet potatoes of very fine quality, gourds, plantains etc., and papayas, the latter replaced by maize and tobacco as the altitude increases, with some of the finest sugar-cane I have seen, are planted, the standing crops in diagonals alternately, with sweet potatoes and gourds, chiefly Lagenaria hispida, as undercrops. Some of these plantations were situated on the steepest slopes, where, toiling up in the pitiless sun, one sinks to one's kneos in fine deep soil. Fortunately there are always many logs lying in succession as they were felled, which facilitate the ascent. The plantations or gardens are surrounded by a strong double stockade against wild pig, with notched poles slanting both ways at certain points for ingress and egress. One or two communal houses are generally built in the middle of the plantation, each on a maze of criss-crossed poles, about 20 from the ground, with a veranda back and front, approached by a notched C2 20 pole from the front only. Where the crest is narrow the houses are built at the edge, so that, entering on the level in front, the elevation at the hack accords with the slope of the hill. Strict etiquette demands that your name and business be shouted out by the most sonorous voiced Papuan at the point of ingress into the plantation, before intrusion on their domain, for the information of the owners. You are then received by the inhabitants, both men and women, all standing or sitting unarmed on the balcony, and after friendly greeting and distribution of tobacco one passes on in peace. This custom is described in the account of the ' Etna' expedition for the northern part of the Arfak (8, 74), and by van Oosterzee (17, 1002 & 1004) on the occasion of his expedition to the Angi lakes, in the Sjari region. From the third ridge on leaving the Soedomi River, we looked down on to the Momi again, and descended to it over secondary slopes overgrown with Rnbus roscefolius, the fruit dirty-red in colour, hard, and like a small rasp- berry in shape, just as insipid but very different in appearance to the large scarlet, strawberry-like fruit of the same species in the uplands of British N. Borneo. A boggy bamboo-thicket lined the bed of the river, which we crossed, to camp for the third night on the other side in an old "kebun." Whole families of the hill people came down to visit us, even with babies in arms, each party, after wandering round gloating over the various sights of the camp, building its own shelter, to which they retired to cook their meal and spend the night. I distributed tobacco to the men, women, and bigger' children, and rice to the babies, of which the very tiniest demanded its quota. Some of these people came on with us to the lakes, as others had done from Serao a source of considerable relief to the coast carriers, many of whom were getting tired from the steady climbing. These mountain people are splendid carriers, but it is next to impossible to get them to go down to the coast. The next day, proceeding through secondary forest up a lateral spur, old plantations opened out at 5000'. At about 6000' a couple of bushes of the copper-coloured Rhododendron Icetum, red when older, one of the glories of the Arfak, heralded the approach of the mountain collecting-grounds, which I alone intended to work. The korano of Wariap and his grandson Waspiri pointed out, on the southern flanks of the spur near the valley below, two houses of the " Orang Jatoe," or bad people, noted head-hunters according to my informants, Waspiri adding that the victims were decapitated at the house observed from the Serao ridge, the resulting trophies being brought up here. With glasses the people could be seen standing on a rise near by, while in front of the house, in a cleared space, twelve men were sitting, in two rows of six each, singing some barbaric chant, accompanying their song of defiance with reiterative movements of arms and legs, in Polynesian fashion. Certainly, unlike all the other mountain people, they did not attempt to 21 approach our party, which was, of course, much too large to invite any form of attack. Higher up, a large solitary house inhabited by very friendly people marked the limit of the inhabited zone. Immediately above this the path enters true " rimbu," l then strikes the track along part of the main range, which, running approximately N. to S.W., is clothed in low mountain forest, the plant-formation common to the crest of the main range and the lake basins. (d) Low MOUNTAIN FOREST FORMATION ABOVE 7000'. From 7000' low mountain forest prevails up to the lakes, and systematic collection was limited to this formation. Advancing due south, with increase of altitude the crest of theS.W. ridge soon becomes narrower and the trees smaller, showing a gradual transition from mossless forest to an intermediate mossy forest with a fine variety of mosses and hepatics ; while on higher points true mossy forest prevailed, but very limited in incidence and species. Coniferous trees, Phyllocladus, Podocarpus, Uacrydium, and Libocedrus, were conspicuous on this ridge, the facies of the vegetation being strongly reminiscent of that of the ridges of Kinabalu at the same altitude, but suggesting a wider and more continental origin. Small open rest- and camping-places from time to time gave vantage points of observation, where the ground was always bright with clumps of a brilliant orange Dendrobium and the pretty mauve Burmannia disticha : otherwise, white and yellow predominated amongst the shrubs and trees of the mountain forest. Other open spaces were caused by small landslips of the loose granite- gravel soil, on which no rock was seen exposed. In one place quite a large part of the crest had slipped away, leaving a steep wall about 100' high, difficult to scramble up, as the gravel gave beneath the feet. Probably the higher points of the ridge, now overgrown with roots of trees, are due to this agency. These open spaces afforded beautiful views over the Sjari ridge with its tiny trails of smoke, to the blue waters of the Bay in the distance on the east, and limited on the west by the densely-wooded slopes of the Tilaan valley. That night, the fourth after leaving Wariap, we camped on the highest part of this ridge, on an open space above a mountain-torrent. I was awakened in the night by a wild stamping of feet, accompanied by a stentorian chant of powerful voices in unison, taken up by each Papuan party in turn. The " Papuas," as they are always called, rigidly kept to their separate clans, each putting up its own shelter. This somewhat alarming incident proved to be a primitive method of keeping warm. 1 Virgin forest. fiven on the march when resting for meals the Papuans invariably sorted themselves into their own family groups. Notwithstanding the heterogeneous nature of the party over forty carriers, men, women, and children, the men in most cases accompanied by their wives and other children, with the hill people joined on, I never heard any complaint from the Serzan nor a discordant note or quarrel. No doubt the presence of my staunch friends Manao and Waspiri, both very fine men, and that of old Basi, the korano of another "campong," a fund of quaint humour and good temper, who had been up several times before, had a great deal lo do with this result. There were, however, many outside elements over which their control \>as not acknowledged; but the absence of bad temper and quarrelsomeness is always a distinguishing feature of primitive people who have been spared contact with our so-called civilization. The next morning, after some hours' progress along the crest, we emerged from the small forest into a lower scrub, to look down on to the brilliant blue waters of a lovely lake, surrounded by slopes and ridges wooded to the water's edge as they ran into the lake, intersected by the white beaches of many bays of varying size and outline. This lake lies in a trough between the ridge we were on and Koebre Mountain, of which the bare summit with a few scattered trees limits the sky-line west, as it rises straight from the water's edge, except for a little cultivated land where the lew Alfuero houses were dotted along the shores. One or two moving spots on the water showed that these people were observing us from the frail rafts made of three palm-stems tied together, which are their only means of transport (PL 1. fig. 2). On the eastern side at one's feet the prevailing forest runs down to the water's edge where the banks are steep (PL 2. fig. 4), but where the slopes are gradual it is replaced by an open marsh bordering this side of the lake, on which some long spinneys and an isolated forest-patch intrude (PL 2. fig. 3). Intersected about the centre by a few forest-clad lateral spurs from the S.W. ridge, the marsh sweeps round to the north, where a break in the hills marks the exit of the Tilaan River. This northern portion was not touched by me in the course of this work. Angi Lakes. Turning abruptly to the right, we descended steeply over roots and trunks of trees through a sheltered mossy forest, then splashing over logs and bog on to the marsh. Here it was open and easy walking where water streams over the coarse quartz sand, bright with Riedelicu, Dendrobiums, and Rhododendrons, but impassible where boggy and covered with fern and sedge. The marsh is not a natural association, but it is kept open by the Alfueros, who were busy burning it off during our stay ; a practice which no doubt 23 facilitates drainage, as the small rivulets, issuing from the slopes, which stream over the marsh to drain into the lake, would form a water-logged area under forest conditions, impeding access to and from the lake. This is no doubt what the Papuans imply when they explain the burning by saying that it keeps the ground u panas" or dry. But more important still is the fact that the houses on the opposite side command the whole of this cleared area, so that the arrival of strangers can be controlled and warning received in case of hostile attack. The practice of burning areas for observation or otherwise must always have been prevalent in the Arfak, as Forrest, in 1705 (1, 108), whose stay at Dorei Bay coincided with a very dry January, saw from there " many great fires on the mountains of Arfak." Passing over the central portion of the marsh, we made our way to some rising ground, where the camp was pitched in front of an isolated forest patch facing the lake, from which it was separated by a muddy tract, where Juncus lampocarpus predominated. This tract soon became a morass from the constant Papuan traffic to and from the lake, where a long white beach formed the landing-stage of the native rafts. The day after our arrival the Alfueros streamed over to visit us, accom- panied as usual by wives, children, and babies. Wearing no clothes, many were plastered over with some black pigment, possibly for greater warmth. They brought to trade "ubi," corn-cobs, tobacco, and splendid potatoes. The latter, grown on the upper lake, were a most welcome delicacy after unlimited sweet potatoes, and as much appreciated by the Malays and Papuans as by myself. My people traded everything that could be scraped together for them and for the tobacco, which was said to be of very good quality. The " Pradjoerit" and " Orang rante" exchanged their salt rations and matches, relying with touching faith on my supplies, even parting with the buttons on their uniforms. The Papuans traded their rice and sago rations, to return to the coast exhausted as a result of an " ubi " diet. St. Vrsiz (15, 235) mentions potatoes as doing well at Hatam, where they had been introduced for twenty years through the Missionary Woelders from Andai. The tobacco was carried at the top of long poles, rolled into large pointed ellipses, which looked like clubs. In 1857 (8, 75), when the northern part of the Arfak was visited by members of the * Etna ' Expedition, they were told that tobacco was not grown on the north side but on the east, and that it was distributed from Hatam to Amberbaki, Dorei, and the south-west coast of Geelvink Bay. That it should be easiest to bring this appreciated article three days' journey down from Hatam to Andai, and then by " prau " along the coast, in prefer- ence to the two to five days' journey from the Angi lakes, proves how little intercourse there is between the Hill and Coast tribes, and also what a natural boundary the inundation "korang" belt forms. 24 The korano of Koebre was quite a superior man, a blood-brother of Manao's, who brought him up to me, when he presented me with splendid potatoes and corn-cobs, and I gave him knives and a "kain" J in return. It was interesting to note how the character of the surrounding forest, even in such a small area as this lake-basin, varied with the exposure. The eastern slopes were characterized by mossy forest, while to the south-e;ist Araucaria Beccarii predominate.!, gregarious and in groups, to near the water's edge (PL 1. fig. 1). To the north and north-east the forest was not so homogeneous, older Libocedrus arfakensis and Podocarpus papuanus, with the graceful palm Kentia Gibbsiana, standing out above the mass level, both on the slopes and the ridge. On the western slopes of Koebre it was much drier in type. The most fertile part was the isolated patch of intermediate mossy forest behind our camp, which reminded me of Fiji in its beautiful moss-flora and wealth of creeping orchids. The possibilities of this patch, though continually worked through, seemed inexhaustible. Accompanied by two of the " Pradjoerit," Manao and his friend the Alfuero korano, with the latter's two delightful boys, most keen to help in collecting and looking for plants, I spent a day on Koebre. We crossed the lake on two of the rafts tied together, following Dr. Gjellerup's advice. It was a very tedious journey, taking about an hour and a half; while coining back in the dark, with stormy gusts of wind and rain, we spent about two hours in crossing. The two rafts, attached by a rotan-tie at each end, worked against each other as the waves splashed up between. The summit of Koebre is a bare, open, lichen-covered plateau, of which the wind-swept character is revealed in the shrnbs, either prostrate and spreading on the ground or of clipped, erect, and compact habit. A few single trees which have survived the fires to which the open character of this summit is due, dot the surface, while in gullies and depressions small trees are crowded into shrubberies surrounded by a ring of burnt wood. It was amazing to see solitary grotesque Myrmedomas, over a metre high, also recorded by van Ooster/ee (17, 1008) plants of such size, to say nothing of the terrestrial habit, being quite unknown to me (PL 3. fig. G). The same may be said of an extraordinary Hydnophytutn^ just like a collection of pipes standing upright on the ground, each pipe representing a hollow stem, about one dm. across, bearing flowering branches round the rim. A couple of small isolated trees of Dacrydium novo-gumeense bore an abundance of small red cones. From the summit there is a splendid view over the smaller, or " Man " (<$ ), lake, beautiful in outline, with much cultivation round its shores, of which the upper slopes are much more densely wooded and the lower more thickly inhabited than the "Woman" (?), especially towards the south, 1 Cloth. 25 where the excellent potatoes are grown. To the north the surrounding hills slope on both sides to a marshy area, which marks the exit of a river, as on the lower lake. To the south-west the houses of another head-hunting tribe were pointed out, who seem to dominate this part of the country, as those before men- tioned cause a reign of terror at lower altitudes. These people possibly represent the tribe called " Hiraj," about which St. Vraz (15, 234) was told at Hatam thnt they lived beyond the lake Tschemti, and were spoken of as cannibals ; but I never heard any of these " Orang jatoe " referred to as anything but head-hunters by my informants, Manao and Waspiri. Our stay at the <$ lake, much as I should have liked to prolong it, was limited to six days. The camp, never very dry at the best of times, became sodden and under water from so much trampling, and many of the Wariap people, who had faithfully kept their promise to stay with me, were suffering from bad colds and rheumatism, such a sudden change of tempera- ture necessarily telling on people accustomed to tropical heat, when warm clothing cannot be provided. The Malays of my escort were also becoming depressed, as they always do when out of their accustomed environment, so the order to break camp was hailed as a happy deliverance by everybody but myself. The beauty of these surroundings, their extraordinary variety, afforded such a wealth of material for work and reflection, which, with crisp air, cool temperature, and splendid weather, made one long to spend more time in this lovely spot. Our return was along the S.W. ridge again, leaving on the right the spur by which we had ascended. Beyond this point a fine high forest developed as the altitude decreased, the huge climbing ferns still wreathing the trunks of the much finer trees, the undergrowth showing less herbaceous variety with more sub-staging of shrubs and young trees. A small clump of Corsia arfakensis, a new species, grew on some dead wood, and at about 5000' a group of two or three magnificent Agatliis Dammara, with towering white steins, too large to climb, and very small crowns, occurred. Great lumps of white dammar stood out on the trunks, which the " Pradjoerit," to whom these trees were familiar in Amboina, immediately wanted to fire, so like the wasteful Fijians with their beautiful " Dakua " trees, which I sternly forbade. This practice, I was glad to see, seemed unknown to the Papuans. Many young trees showing the fastigiate youth form were seen, and one of the Papuans found me a young seedling, about 2 m. high, from which I took the foliage, but had to content myself, so far as fruiting material went, with some old scales found beneath the old trees ; they were sufficient, however, to determine the species. Continuing down a lateral spur, running in an easterly direction, we 26 passed through a somewhat dense undergrowth of a small Licuala sp., where the pretty white Medinilla arfakensis, almost a small tree, a very handsome Bulbophyllum covering a prostrate log, and the climbing Diclirotricliium brevipes, another of Beccari's Hatam records, were also growing. From an open space we caught a glimpse of the buttress of the ridge we had descended from, with the glorious Agathis trees towering far above the rest of the forest. Farther on, at another unusually large cleared resting-space. vaulted over by trees, we came upon some horizontal sticks, resting on a couple of forked supports stuck in the ground, on which a number of little forked prongs were arranged in groups. Waspiri exclaimed when he saw this peculiar arrangement, explaining that it was the practice of the Coast people to have these places, which show what parties have recently been in the mountains, the arrangement of tho little prongs indicating how many and whether women or children, if the parties had returned, or if any of their members had been killed. On this occasion it was m:ide out that a man and woman, missing from Wariap, had been killed in the hills. Finally, always working east, we passed through the bamboo clumps, which herald cultivation, on to a large plantation, with a solitary house in the centre. Here an old man showed us the way down to the Momi River, an impossibly steep track, up which, considering the way it was worn, the people of the house must fetch all their water every day. From this plantation a view up to the Serao ridge showed the " kebun," in which we had camped on the second night, and also the house of the head-hunters on the lower slopes. Camping by the river that night we reached Wariap next day through the " korang " forest, joining our old tract later, just before the Pandanus group. At Wariap my first inquiry was for the ' Valk/ as the Commander had most kindly promised to call in there on his return from the Mamberamo River, on the chance of our being there. As there was no news I decided not to wait in tliat sand-fly stricken spot, but, giving the men two days' rest, to return by the beach to Manokoeari, which everybody at Manokoeari, even Mr. van Oosterzee, had told me was quite impossible. The iSerzan, however, after searching inquiries, found that this route was feasible and well-known to the Wariap people, taking four days. Two " praus" from Wariap carried kit and provisions, landing at night at the native camping-places, to which it is always wisest to keep, though somewhat long, as they are determined by good water and landing-stages. I had hoped to be in time for the December boat to Java, but on crossing in u praus " from Andai to Manokoeari, once past the mouth of the estuary, we could see the smoke of the steamer across the bay. Watching intently which direction she took, the Serzan exclaimed, " Poelang ! " (Home). This 27 entailed a month's wait at Manokoeari for the next boat ; I put in the time working about Dorei Bay, as Dr. Grjellerup told me it had not been collected over. As the next boat called at Humboldt Bay, the limit of the Dutch possessions, which is only visited every other month, I was able to take that trip as well, collecting at each stopping-place, with very good results. The coast-collections proved very interesting, but phytogeographically so distinct from the Arfak plants, no two species proving common to both regions, that they have been separately enumerated. PLANT ASSOCIATIONS OF LOW MOUNTAIN FOREST FORMATION. A. FOREST ASSOCIATIONS. [Endemic species are marked , and those of wider distribution *.] a. S.W. RIDGE. 1 a. Mossless Forest. On the main range, or S.W. ridge, at 7000', a mossless forest asso- ciation prevails, of slender straight trees about 13-] 6 m. high, with a very open undergrowth of chiefly herbaceous plants. Undergrowth. Alpinia domatifera, 1-2 m., always in appreciable colonies of one height, the flowers varying from white to red with red fruit, and A. arfakensis var. subsessilis, with pink flowers and white fruit, were dominant, more or less covering the open ground. Liancs. * Gleichenia linearis, spread over supports up to 7 m., while the trunks of the trees were wreathed in the climbing ferns *Nephrolepis acuminata and P^lybotrya arfakensis, from base to branches, the long fronds standing out radially from the stems. Freycinetia Gibbsece, with very hand- some red sheaths, hung bunched from the trees or spread in thick masses underneath, and F. flaviceps, with yellow fruit, was more slender in habit. Trees. A group of Quercus Lauterbachii, the ground underneath strewn with magnificent acorns of all sizes, some of those collected having proved the largest known, represented a family recorded by D'Albertis (9, 69), Beccari (12, i. 177), and St. Vraz (15, 33) from Hatam. *Podocarpus Rumphii, recorded by Beccari, but not seen in fruit, was abundant ; likewise Phyllodadus kypophyttvSj the Kinabalu species, and Podocarpus papuanus, recorded as P. imbricatus (which it very much resembles, the seedling form being indistinguishable) by Beccari from Hatam, and since found by Kloss on Mt. Carstensz. Advancing due south, as the crest of the ridge narrows, a gradual transition to an intermediate mossy forest of smaller trees with branched stems and denser crowns, the trunks and buses covered with small hepatics and mosses, takes place as the altitude increases. 2 a, /3. Intermediate Mossy Forest. Undergrowth. Sphagnum novo-guineense, with *Rhacopilum spectabile and the magnificent and abundant Dawsonia gigantea, ull in fruit, with the creeping *Lycopodiutn cernuum, formed part of the prevailing moss-carpet, from which rose the orchids Platantliera elliptica, vars. longicalcarata and elatior, green in colour and varying in size, gregarious and general in this association ; Cryptostylis arfakensis, with red lip and green perianth, and Bulbophyllum muricatum, with large red-spotted yellow flowers, were found in single examples. Clumps of Alpinia domatifera (dwarfed), *Gahnia psittacorum, previously known only from Australia and Tasmania, * Histiopteris incisa, *Dipteris conjugata, gregarious as usual where more open, with the small shrubs Diplycosia Lilianea?, about "25 m. high, with rigid branches and striking white-tipped red flowers, Vaccinium pilosifiorum with pretty, very hairy rose-pink corollas, also seen as an epiphyte, and V. leptospermoides, with red-pink flowers, were often grouped together at the base and between the dwarfed trees. Tiny tufts of Gentiana Vanderwateri, with large white flowers, showed up in damp places and, where drier and the small trees opened out, the minute Lobelia arfakensis spread its large patches on the ground, dotted with white flowers, while innumerable seedlings of all the conifers previously mentioned, with Dacrydium and Libocedrus, formed the most general and conspicuous part of the undergrowth, including the slender tree-ferns Alsopldla arfakensis, with stems 1 dm. through and about 1 m. high, the fronds 1 m. long, and Cyathea arfakensis not much larger. Epiphytes. The small ferns collected on the moss-grown trunks of the trees were * Trichomanes palmattfidum, Hymenopliyllum cincinnatum, *Lindsaya hymenopliylloides, Poly podium remigerum, *P. stenopliyllum, *P. clavifer, and the minute yellow orchids, Octarrhena cylindrica, with Dendrobium glauco-viride (magenta), Phreatia spatlmlata (white), and Piper arfakianum. Lianes. * Gleiclienia linearis and *G. volubile abounded, with Freycinetia GibbsecK, much less luxuriant, F. ftaviceps, and Calamus arjakianus ; a slender bamboo, identical in habit and appearance with the Kinabalu plant, but not seen in flower, and Lyonsia albifiora were also pretty general. Trees. Dacrydium novo-guineense with Libocedrus arfakensis, Podo- carpus papuanus, *P. Rumphii, and * Phyllodadus hypopliyllus wore dominant, as the profusion of seedlings testified ; but a great variety of other small trees were associated, especially towards the southern portion of the ridge, where it broadens out again and is consequently more sheltered Urimys arfakensis, with a dense round crown, bore its small, white, later pink flowers on pendent pedicels ; Spirceanthemum bullatum, *13ceckea frutescens ; Back- housia arfakensis, flat-topped like a Leptospermum, with very small coriaceous leaves, and smothered in golden-orange flowers. A Psyclwtria sp., resembling P. sarmentosa, with white panicles of flowers, Idenburyia arfakensis repre- senting a new natural order, Myrtus flavida var. glabrescens, a glabrous form of the Kinabalu specie?, Jambosa arfakensis, with small thick leaves and white flowers, Palmervandenbroefcia papuana, an interesting new genus in Araliacese, and Timonius filipes, were in full flower. 3. Mossy Forest. The occurrence of this sterile type of mossy forest, as distinct from the intermediate form, was limited to the highest points of the ridge, where the prostrate and erect trunks of the small stunted trees with the ground between were swathed in long moss, which, as usual in this type of associa- tion, stands out straight from its supports, rigidly turgid and generally in vegetative condition. *Schizcea malaccana, *Gahnia psittacorum, 1 m. high, Halorrhagis suf- fruticosa, Nepenthes maxima v. l>revi folia of the exposed peaks of the serpentine ridges of Kinabalu. * Lycopodium cernuum, * Gleichenia vulcanica, * Dip/ten* conjugata, grew densely, also * Spathoglottis aurea and *Dianella ccerulea. At another more recent slip, Gleichenia vulcanica with *Lycopodium cernuum were associated with *Gahnia psittacorum, 1'50 in., * Burmannia disticha, the woody herbaceous llalorrhagis suffruticosa, the single shoots with verticillate spinous leaves terminated by a panicle of red flowers, Oldenlandia nutans, of similar habit, but quite herbaceous, with white flowers ; Myrtus arfakensis, as a small spreading shrub, and the prostrate M. koebrensis, with reddish foliage and stems, both with yellow flowers, spread over the easier gradients. b. MARSH BY ? LAKE. Sand Pans with running water. In certain areas of the marsh, more towards the centre, where coarse quartz-sand formed a solid surface, intersected by shallow streams or bathed in films of streaming water, many small herbaceous plants grew spaced between the larger clumps of more showy species. Herbaceous Plants. The little *Schizcea malaccana witli the cosmopolitan * Lycopodium carolinianum, creeping tightly to the ground, and *L. cernuum, P2 36 the minute *Bulbostylis capillaris var. trifida, Centrolepis novo-guineensis in tufts, the first record of the genus in New Guinea, *Xyris pauciftora, Eriocaulon leucogenes, very variable in size, * llurmannia disticha, up to 1 m., one or two examples of the little white Spiranthes papuana ?, *Polygonum alatum, the small yellow *Utricularia bifida, and the mauve *U.racemosa from a few cm. to 2 dm. in height, both new records for New Guinea, dotted the surface ; while abundant patches of the yellow and red RiedeHa montana var. puberula, also var. arfakensis, cream and dark red, both about 50 m. high, with the beautiful terrestrial orchids, Dendrobium fruticicola, the foliage '50 m., above which the brilliant orange flowers, red when older, rose to '30 m., D. latifrons, also yellow, and the splendid D. rhomboglossum, 1 m. high with large convex flowers, magenta-pink outside, white inside, borne the whole length of the rhachis, made glorious patches of colour, blended with small slender plants of Halorrhagis su/ruticosa, Oldenlandia nutans, and Coleus Gibbsece, white to mauve, all of similar habit and dominant on the marsh in all situations ; young plants of Nepenthes maxima var. nana, the mature trailing over any support, with *Lycopodium divaricatum abounded. Shrubs. As the above plants eased oft' and a ranker growth inhibited close burning, shrubs occurred in small groups, chiefly *Bceckea frutescens, Rhododendron Devriesianum, of which the magnificent white flowers, turning- pink later and fragrant at night, are about 1 dm. across, the whole inflor- escence being 10 dm. in diameter, R. Icetum, the little profuse flowering R. Vonroemeri, and the bright red R. Gibbsece, all further emphasize the wonderful colour-scheme of the more open portions of this interesting marsh association. Denser Sedge Growth in Boggy Areas. Boggy areas with deep ditches, standing water, or larger streams bore an impenetrable sedge-growth which included *Dryopteris Beddomei, *Lyco- podium divaricatum, * Cladium falcatum and *C. germanicum, *Sdrpus setaeeus, Trisetum latifolium, Ischcvmum aristatum var. arfakensis, *Juncus lampocarpus, Phaius Tankervillece var. papuanus, the flowers brown, white inside, Halorrhagis suffruticosa, Oldenlandia nutans, Coleus Gibbsece, and * Emilia sonchifolia, all drawn up to the level of the surrounding sedge. In other boggy areas sloping down to the lake, generally under water, *Juncus lampocarpus predominated, easing off with * ' Carex Gaudichaudiana to the sand-pan association. Edging the lake, limiting the white coarse quartz-sand beach, a higher and drier sandy bank, well bound together by the roots of the sedges, and broken in parts where small streams ran into the lake, carried *Marchantia polymorpha, * Lycopodium divaricatum and *L. complanatum, Cladium arfak- ense and *C. germanicum, * Carex Gaudichaudiana, *Gahnia psittacorum, * Rhynchospora aurea and *R. glauca, Trisetwn latifolium, * Dianella 37 ccerulea, Riedelia orchioides, '75 in. high with red perianth-segments and red-tipped lip, replaced, where the sandy ground was more exposed, by the creeping *Isachne miliacea, * Centella asiatica, *Gna sorbifolia, *Gleichenia vulcanica, and Oleandra cuspidata were common all round, the scrambling *Gleichenia Icevigata abundant in the open to under the trees, *G. glauca up to 6 m. \\ith *G. linearis, *Cladium falcatum, * Gahnia psittacorum up to 2 m., Halorrhagis sujfruticosa, Uldenlandia nutans, and Coleus Gibbsece, all drawn up to the level of their environment, with the shrubs Medinilla Forbesii, * Mclast oma malabathricum var. adpressum, the lanky single shoots of young plants of Poikilogyne arfakensis, Rhododendron Devriesianum, It. Ict'tum, and the pink R. undulaticalyx. The gregarious Sccevola Lauter- hachU, with long semi-scandent shoots, shining leaves, and for the genus large yellow flowers, known hitherto from N.E. New Guinea, dominated in the tangled mass of fern and sedge. c. OPEN SUMMIT OF KOEBRK MOUNTAIN, 9000'. Cladonia Association. The open summit of Koebre forms a flat plateau with a hard surface of disintegrated quartz-granite on which quartz-gravel lies so thickly that it gives quite a white effect. Where the rock is exposed it snows as large a proportion of quartz-veins as of granite. Herbaceous plants. On this open plateau, for which systematic burning is again responsible, a remarkable association of Cladonia spp. prevailed, spreading over the whole area as a uniform grey carpet, about 3 cm. high, composed of * Cladonia verticillata, *C. didyma, and *C. cocci/era, displaced 38 in certain parts by stretches of short *Pteridium aquiliniini var. lanuginosum or *Gleiclienia vulcanica, and in damper places dense mats of C&ilrolepis novo-guineensis. Associated with the Cladonla were clumps of Riedelia montana var. arfakeusis, and colonies of * Burmannia disticha and * Gahnia psittacorum, the latter dwarfed to '50 in. high. Patersonia novo-guineensis in full flower, the gleaming white or pale mauve corollas well exposed, was dotted in tufts all along the summit, a most interesting first record of a genus known from Kinabalu, Mt. Halcon in the Philippines, and Australia. A small example of the gi'eenPlatanthera elliptica was a remarkable find, the species alone being previously known from S.W. New Guine;i, while the two new varieties, so abundant in the moss-grown forest of the S.W. ridge and the small high mossy forest-patch by the lower l:ike, were not seen on the slopes of Koebre. Patches of the familiar D. rhomboglossum and D. fruticicola, both as conspicuous as on the marsh, with the magenta D. infractum, represented the terrestrial orchids in flower, with Glomera Gibbsece epiphytic on the upper branches of a wind-swept shrub. Didiscus koebrensis with many radiating stolons, near a species already collected on Mt. Scratch ley in the S.E. supposed to bo a variety of I), sanicitlce/olius of Kinabalu and Mt. Halcon. Oldenlandia nutans var. alpinum was dwarfed to '50 m. ]\]yrmedoi>ia arfakiana (PI. 3- fig. 6) with its slate-blue flowers the size of a shilling, showing all round the formless bristly steins, of which the fleshy consistency no doubt resists the fires, grew in solitary examples, one of which, about a metre in height, sent out a metre- long shoot from the apex at right angles to the main stem. Shrubs. The exposed and wind-swept character of this summit plateau was shown in the shrubs, which either spread prostrate over the surface or were clipped back into small compact shapes, Hibbertia novo-guineensis, a plant with large flowers and spreading habit and very near to H. scandens of E. Australia, is the first species in this genus, hitherto known from A-Ustralia and New Caledonia, to be described for Malaya or New Guinea. The ever-present *J3ceckea frutescens took on a prostrate form, Acronyclna arfakensis, with insignificant white flowers and reddish fruit, was either pros- trate or erect, while Myrtus prostrata, with reddish stems and small coriaceous leaves, had developed quite dorsiventral shoots, resting on the ground. Small erect shrubs were *Bensloivia umbellata with yellow foliage, Diplycosia Liltanete, and Vaccinium villosiftorum, while the dark green Tel- mmt/iodia rotundifolia, the typical Styirtielia Gjellerupii, with white flowers and pink berries, and another Styplielia sp., of which the material was not sufficiently complete for identification, Rhododendron angiense, Sericolea novo-guineensis with Psycliotria vaccinioides, were of clipped compact habit. A couple of small trees of Dacrydium novo-guineense dwarfed to about 4 111., with stiff, Araucaroid, ascending branches, bore many tiny ripe cones, 39 glinting red through the dark green foliage, a fortunate fact, deciding the genus of a very prevalent species of conifer, of which it was certain that the abundant seedlings in various growth-forms could only represent a new Dacrydium sp., but the search for fruiting specimens at lower altitudes proved unsuccessful. In reference to this exposed lichen-covered plateau a quotation from Lorenz (22, viii. (1909) 178) in "Nova Guinea" suggests the presence of a similar association on the Charles Louis Mountains, while Mr. Stroeve of the Dutch Navy, whom I had the pleasure of meeting at Manokoeari after my return from the Arfak, described what seemed a very similar formation as occurring on the mountains to the south of Geelvink Bay at about the same altitude, where not only the practice of burning obtains but potatoes are also grown. I heard later with great regret of the untimely fate of this promising young officer, who, after most successful exploration and surveying work on the Ruffaier River, was treacherously shot by arrow by Papuans on the Waroza River, near the coast to the east of Geelvink Bay (26, 782). PHYTOGEOGRAPHICAL CONCLUSIONS. These may be summarized as follows : (1) New Guinea, the centre of distribution for many so-called Polynesian, Australian, and to a lesser extent Malayan types, of which the Papuan species are not only older in type, but also show extraordinarily pronounced specific differentiation. (2) Wide distribution in New Guinea of endemic Mountain types. (3) Low Mountain Forest formation approximates to the Ridge forma- tion of Mt. Kinabalu and the Philippines. (4) Mountain Forest Flora of the Arfak Mountains possibly represents the oldest Papuan type. (5) Pteridophytic and Bryophytic Floras are more Polynesian than Malayan in incidence and luxuriance. (6) Open " Opportunity " Plant Associations offer further proof of the autochthonous character of the Papuan Flora. 1. NEW GUINEA THE CENTRE OF DISTRIBUTION FOR MANY SO-CALLED MALAYAN, POLYNESIAN, AND AUSTRALIAN TYPES. Evidence of New Guinea as the centre of distribution for many plant- types, so far considered Polynesian or Australian, has been strikingly confirmed on the present occasion. Emphasized by the well-known botanists who have worked out the results, such evidence has been a marked feature of recent Dutch and German exploration, and was also shown in Kloss's Mt. Oartensz collections 1 . 1 Ridley, II. N., " Report on the Botany of the Wollaston Expedition to Dutch New Guinea," Trans. Linn. Soc., ser. 2, Bot. ix. (1916) 1-269, pis. i.-vi. 40 Geological support for this theory is quoted by Wallace (7, 443) in ' Island Life,' who points out that the " 1000 fathom line, which indicates the land area which would be produced if the sea-bottom were elevated 6000', extends in a broad mass westwards, then sending out two great arms, one reaching to beyond Lord Howe's Island, while the other stretches over Norfolk Island to the great barrier reef, thus forming a connection between Tropical Australia and New Guinea." Sieberg 1 explains the same configuration in greater detail as follows : " Neuguinea und der Bismarckarchipel gehoren nach E. Suess dem inner- sten unter den Bogen junger Faltengebirge an. die vom Osten her sich gegen die alte starre Festlandstafel Australien anlegen und die in Neu- seeland zusatnmentreffen. Dieser innerste australishe Bogen setzt sich aus zwei annahernd parallelen Teilstiicken zu?ammen, die beide in die Nord- nordwesthalbinsel Neuseelands miiriden. Die Nordwestlich streichenden Ketten holier Faltengebirge, die das innere Neuguineas der Lange nach durchziehen (Bismarck und Kratkegebirge, Viktor-Emanuel, (Charles-Louis Gebirge) reichen untermeerisch noch weithin und ragen in ihren hochsten Spitzen als die Inselgruppe der Louisiaden und Neukaledonien ernpor ; der ostliohe parallelbogen setzt sich zusammen .aus Neumecklenburg, den Salomonen und Neuen Hebriden." Of the preponderating influence of the New Guinea Flora, when suffi- ciently well-known, we have an instance in the Orchidacese, a family which in the N.E. has received the expert attention of Schlechter (23), the well-known authority and collector. Dr. J. J. Smith, of Buitenzorg, an equally keen expert in this family, has not only described nearly all the orchids of the N.W. and S.W. so far known, but has also consistently enjoined on collectors in the Dutch possessions to devote particular atten- tion to these plants, so that from both these sources we can form some idea of the general incidence and distribution of tin's one order throughout the country. Such is the wealth of material in a single instance in this little-explored island that Schlechter, whose unrivalled experience in N.E. New Guinea has spread over a number of years, summarizes his results in orchids alone as 116 genera with 1450 species, of which 1102 are new (23, i. 14, xiii). He considers that nowhere in the world is the Orchid flora so rich, quoting 2600 species as being known from there. Orchids known from Australia and common to New Guinea he (23, i. 14, xix) looks upon as branches of the Malayan-Papuan flora in Australia, the Orchid flora of tropical Australia having developed under strong Malayan-Papuan influence, but it has had little influence on the Papuan flora. This remark, to those who have 1 Sieberg, A., " Die Ei'dbebentatigkeit in Deutsch-Neuguinea (Kaiser- Wilhelms-land und Bismarckarchipel)," Peterm. Mitth. Ivi. pt. 2 (1910) 118. 41 worked over the two areas, succinctly summarizes the whole relation of both floras. This Malay-Papuan influence is the determining factor extending to the South Sea Islands, while the Moluccas show Malay with Philippine ground- types and a strong Papuan influence (23, i. 14, xx), and Schlechter describes New Guinea as the centre of distribution of an endemic Orchid flora (" Aus- gangscentrum einer eigener Orchideenflora ") (23, i. 14, xx) as evidenced by Corysantltes, supposed to be Australian, but now truly Papuan, 13 species being known from N.E. New Guinea alone. The presence of this genus in Australia, New Caledonia, Samoa, Java, Philippines, and the Himalayas suggests radiation from a Papuan centre of development (23, 14, xxiii). This overwhelming Papuan influence is amply demonstrated in many other families, which not only show an actual numerical predominance in species, but also a greater range of specific differentiation than is known from elsewhere : for example, Libocedrus (4), Drimys (19), Myrtus (6) ; of Pandanus and Freycinetia, I have never seen such a wide range in form ; while in Rhododendron, Vaccinium, Styplielia, the numerical predominance along with the great morphological range of form is unsurpassed in any other region, except perhaps, in the case of the first, Central China. Further interesting proof of the soundness of this point of view is afforded by comparison of the Australian and New Guinea representatives of the same genera, the Papuan forms showing decidedly the oldest types. In the present collection perhaps Trimenia is the most striking case in point. T. weinmannicefolia Seem., described in 1852 from Fiji, a dioecious plant, remained the type of a supposed monotypic Polynesian genus, closely allied to Piptocalyx Moorei, also dicecious, ranking as a monotypic Australian one, till Ridley described T. papuana from Mt. Carstensz in 1916. T. arfak- ensis is included in the present paper. Two species of Trimenia are now known from New Guinea, both hermaphrodite, while two new species in a closely allied new genus, Iden- burgia, show a syncarpous bilocular ovary, proving not only that the dioecious habit of the two isolated outliers of this order is probably derived, but also that the systematic position of Trimenia and Piptocalyx in the apocarpous Monimiacese is untenable, necessitating the new order Trimeniacese. An equally convincing example is that of Pullea, a genus established by Schlechter in Saxifragacese to include two plants with inferior ovary from N.E. and N.W. New Guinea respectively, to which P. papuana is now added, with a N. Queensland species, P. Stutzeri = CaUicoma Stutzeri F. Muell., first distributed as Stutzeria by him, but afterwards included in Callicoma. Pullea, therefore, now includes three distinct Papuan species and one N. Australian, but, had the latter plant first been accorded proper generic position, the former would have been cited as evidence of a wave of Australian immigration into New Guinea. 42 Of the two Papuan species in BackJiousia, another of the so-called Australian endemic genera, one was described by Ridley from Mt. Carstensz, to which a second, one of the commonest trees in the Arfak, is now added. Both have normal calyx-lobes, while the Australian species show a secondary petaloid development of the lobes, evidently derived in character, though previously considered a generic distinction. The same case may be quoted for Didiscus, where the older types, according to the Monographer, 1 showing a normal calyx, occur in Malaya and New Guinea, with one species in N. Australia, whereas those with aborted calyx-lobes, by far the most numerous, are known from the rest of that continent, with one species from some Polynesian Islands. Further evidence is also forthcoming in interesting new records of genera not previously known from New Guinea, namely, Hibbertict, hitherto considered Australian and New Caledonian ; Centrolejns, abundant at 7000' and 9000', connects New Guinea with the open summit of Kinabalu in N. Borneo at 13,000', Mt. Halcon in the Philippines at 7000', and S. China on the one hand, and Australia and New Zealand on the other ; while the genus Patersonia, supposed on inadequate knowledge to be endemic Australian, shows the same distribution, with the exception of S. China, and now in- cludes three very distinct Mahiyan mountain types. In the case of species, Gahnia psittacorum, abundant everywhere in the Arfak from 7000', is the first record for Malaya of a plant widely distributed in E. Australia to Tasmania. 2. WIDE DISTRIBUTION IN NEW GUINEA OF ENDEMIC MOUNTAIN TYPES. In considering questions bearing on the phytogeography of New Guinea, as a whole, it is important to take into consideration the fact that it is a country of 786,000 km. in area (23, i. 14, i), most of which is mountainous in character and undisturbed in condition. A huge region of vast ranges, in which all intercourse between the relatively few and scattered inhabitants is not only restricted by natural barriers, but also by the many different languages, no two tribes having a common speech, even when living in apparent proximity on the same range of mountains. Both these facts have effectually debarred outside penetration into the mysterious back-country. Though our knowledge of the plant-covering of these mountains is extremely limited, such phytogeographical exploration so far accomplished points, as would be expected under such virgin conditions, to the homogeneity and stability of the flora as a whole. This fact is evinced in the pronounced endemism and wide distribution throughout the whole country, further 1 Domin, K., " Monographie der Gattung Didiscus (DC.)." Sitz. Kon. bohm. Ges. d. Wise. ii. Cl. (1908) 21. 43 convincing proofs of which are afforded in the present collection. The following instances may he quoted. All the mountains of New Guinea show a preponderating number of species belonging to Rhododendron and Vaccinium in Ericaceae and Styphelia in Epacridacese, though but few identical species have been so far recorded ; but that this point is only a question of further investigation is proved by the distribution of the very distinct R. Vonroemeri, which, most abundant in the Arf'ak, is recorded for the ( Cyclops Mountains in the north, and from the Hellwig Mountains and Mt. Carstensz in the south-west ( = 72. calceo- larioides Wernham). Medinilla Forbesii, collected in the south-east and the south-west, is now established for the north-west as well, while Timonius filipes and Podocarpus papuanus, first collected on Mt. Carstensz, are common on the Arfak at higher altitudes. Quercus Lauterbachii 1 and / SccBvola Lauterbachii,^ generally collected in the north-east, were not pre- viously known from Dutch N.W. New Guinea. In small herbaceous plants two new species of Didiscus link up the Arfak with the Owen Stanley range in the south-east, while Eriocaulon leucogenes, Trisetum latifolium, and Gentiana Vandencateri are common to the former and Mt. Carstensz. Spirantlies papuana is now known from the north-east and north-west, Platanthera elliptica from north and south-west. New species in genera first recorded from New Guinea through Kloss's Carstensz collections, such as Trimenia and Backhousia, are now proved to be common to the Arfak as well ; also the genera Pullea and Sericolea, the latter represented by six species in the north-east, one in the south-east, three on Mt. Carstensz, and now by two in the Arfak; while Libocedrus, known from the north-east, south-east, and south-west, and Dacrydium from the north-east, south-west, and south-east, are abundantly represented by one very distinct species each in the Arfak. 3. THE LOW MOUNTAIN FOREST FORMATION APPROXIMATES TO THE RIDGE FORMATION OF MT. KIN ABA LU AND THE PHILIPPINES. The low forest formation approximates very closely to that of the serpentine ridges of Kinabalu from 7000', but with a larger proportion of what I would have previously described as southern hemisphere types, but prefer now to refer to as Papuan. Identical plants so far recorded are Phyllocludus liypophyllus, Myrtus jiavida var. ylabrescens, a glabrous variety of the Kinabalu plant, while closely allied species occur in Dawsonia, Podocarpus, Ducrydium, Centrolepis, Patersonia, Didiscus, Qentiana, and others. 1 Schumann, K., and Lauterbacli, K., ' Flora der deulachen Schutegebiete in der Siidsee,' Leipzig, 1901. Naclitrage, 1905. 44 The prominence of orchids, Myrtacese, Rhododendrons, Styphelias, and Vacciniums, in relation to other plants, is equally emphasized, but members of the Araliacese, an old type of plant so prominent in the Arfak of New Guinea, are, so far as we know, absent on the N. Bornean mountain. From descriptions of the Philippine mountain-ridge vegetation it would appear that the Papuan facies again predominates, thus approximating it to that of the Arfak and Mt. Kinabalu, the absence of araliaceous types being less marked, though of the typical P;ipuan genera Anomopanax is represented by one species, and Kissodendron and Mackinlaya are absent. Phyllocladu* hypophyllus is common to all, with allied Dacrydium and Podocarpus spp. Glochidion Merriltii is also common to the Arfak, but a thinning out in the number of Styphelia and Vaeciniuni spp. is apparent, while the Rhododendrons show a great reduction in variety of type. 4. THE MOUNTAIN FOREST FLORA OF THE ARFAK MOUNTAINS POSSIBLY REPRESENTS OLDEST PAPUAN TYPE. It is in the Arfak alone, of all the Papuan mountain ranges of similar or greater altitude, that two different localities have been worked over on three separate occasions. This fact gives a wider field of comparison in relation to this range. On comparing the general type of vegetation and the generic and specific distinctness of the plants so far collected on this range with those known from similar localities, one feels inclined to look upon these granite mountains as carrying an older part of the Papuan flora. Where so little is known, however, of the mountain flora of the country as a whole, and collections remain limited to vertical sequence, it is impossible to base such an impression on any actual fact, and the hypothesis is only worth advancing as a possible consideration for future workers. The contrast is very striking between the recent " korang " belt, from which this range on the east rises almost without transition as abruptly as the mountains on the north from the sea, and the well-defined limits and deep soil of the foot-hills and lower ranges. The extreme homogeneity of the mountain forest zone, the extraordinarily small incidence of outside or immigration plants, combined with the relatively broad crests of the ridges and the tremendously weathered condition of the main range, are all facts which point to general stability in condition, spreading over a considerable epoch of time. The climate of the Arfak would also seem to be more favourable than on other ranges, and the nature and structure of the plants are not so limited by the edaphic and atmospheric factors which play such a large part in the restriction of floral conditions on Kinabalu. For instance, all the Arfak plants show coriaceous to very small leaves, and the hairy covering or tomentum, common to many of the Kinabalu 45 plants, is quite absent a fact possibly to be attributed to the warm well- drained soil of disintegrated granite and the even conditions of illumination with less direct exposure to wind. The " Nebelwald " of the German botanists, described as predominating on the mountains of the north-east, with trees swathed in lichen, is absent, also the " Krlippelwald" association, while the turgid vegetative form of mossy forest is limited in incidence. But, of course, so far as the two former associations are concerned, the lower altitude is a possible and limiting factor. 6. PTERIDOPHYTIC AND BRYOPHYTIC FLORAS ARE MORE POLYNESIAN THAN MALAYAN IN INCIDENCE AND LUXURIANCE. The Pteridophytes and Bryophytes of the Arfak proved exceedingly varied and luxuriant in incidence, more Polynesian than Malayan in facies, but endemic Papuan in type. In mosses Spiridens, with a limited distribu- tion in Polynesia, and since recorded from the north-east and north-west of New Guinea and Borneo, is probably a Papuan generic type. 6. OPEN " OPPORTUNITY " PLANT ASSOCIATION. In a mountainous forest country like New Guinea, sparsely inhabited, with little or no intercommunication between the different tribes, and no migratory herds of grazing animals, there is everything to conserve and nothing to modify natural conditions. Lying in what may be called the centre of the monsun region, the whole country is subjected to a more or less regular rhythm of alternating air currents. Beccari (12, i. 216) has emphasized that the north-east monsun, blowing regularly from November to April, must affect the general distribu- tion of plants in the regions which come under the immediate influence of these prevailing winds. This opinion has been already advanced by myself, 1 and later 2 from observations on the granite core of the exposed summit of Kinabalu, before I had seen Beccari's convincing remarks in relation to his own observations throughout Malaya. Beccari aptly remarks that seeds are lighter than grains of sand from volcanoes, and are adapted to remain longer in the air. Authenticated instances of the possibilities of long-distance transport of grains of sand and volcanic ash by the agency of wind are quoted by him (12, i. 216-220). Warming 3 advances the same theory, for which extraordinary proof is forthcoming in recent work of the Geological Survey of India. La Touche 4 1 L. S. Gibbs, " A Contribution to the Montane Flora of Fiji, with Ecological Notes," Journ. Linn. Soc., Bot. xxxix. (1909) 137. 2 . " A Contribution to the Flora and Plant Formations of Mt. Kinabalu and the Highlands of Brit. N. Borneo," I. c. xlii. (1914) 47. 3 T. Warming, ' History of the Flora of the Faeroes,' Botany of the Faeroes II. Copen- hagen, 1903. 4 T. H. D. La Touche, Mem. Geol. Surv. Ind. xxxv. (1902) 42. 46 discovered small undamaged fora minif era in the desert sand of Barmar and Bikanir, which must have reached the heart of the desert by wind trans- portation over a distance of 500 miles from the coast of Cutch. This interesting observation gave the first clue to the remarkable work organized by T. H. Holland, 1 to explain the origin of several intermittent saline lakes in the Rajputana desert, in which the quantity of salt stored is in excess of the amount that could be accumulated by normal freshwater rivers acting within any reasonable geological period under present physio- graphic conditions. The Rann of (.hitch dries up in the hot dry season, to be covered with a thin incrustation of salt. This salt is transported by strong winds from the south-west, which blow regularly from April to June, to be followed by the rainy season, when the salt, deposited on the surface of the desert, is washed in solution into convenient hollows, forming small lakes. It was found, as a result of this investigation, which should be widely known amongst botanists, that during four months of the hot season of 1908 the amount of wind-borne salt passing a front of 300 km. broad and 100 m. high must have been something of the order of 1HO,000 tons. Recently our airmen 2 operating in Mesopotamia have found the " dust chokes the engines and the sand above blows as high as 4000'." Further interesting proof of the constant direction of air-currents is afforded in the first report on upper air research in Australia, 3 when difficulties were encountered owing to the fact that Melbourne is on the south coast of Australia, and the prevailing winter winds, as well as the upper currents in advance of cyclonic disturbances, are from a northerly direction, and thus carry the balloons out to sea. Material carried by wind is deposited on reaching contrary currents, when should the seeds carried be precipitated on to a suitable habitat germination takes place (12, i. 220). Cross-currents would be most likely met with on the summits of high mountains, where conditions in the tropics would alone be favourable to ombrophobous plants requiring temperate conditions. We have now sufficient evidence to prove that the area of high open country on the immense chains of mountains in New Guinea must carry an enormous number of species of this type of plant. The summits of Mt. Kiuabalu in Borneo, Mt. Halcon and others in the Philippines, and Bonthain Peak in Celebes, would offer the only suitable habitat for such plants between New Guinea and the Himalayas in the west monsun region of distribution. In those cases where not only the suitable area but also the fauna is much restricted, though identical conditions of temperature and rainfall prevail, 1 T. H. Holland : successive Annual Reports of the Geological Survey of India published in Recordt G.S.I, during 1904-09; "Discussion on the Physiography of Arid ' Lands," Report Brit. Asaoc. Adv. Sci. for 1914, 363 (1915). a Candler, Edmund, ' Daily Mail,' Dec. 19, 1916. Commonwealth Bureau of Meteorol., Bull. 13 (Melbourne, 1915). 47 possibilities in the evolution of new species are practically inhibited, and we get the same or very closely sillied representatives. But when cross-currents from the Antarctic (24, 222) cause precipitation in Australia, the opportunity habitat is that of a large and open arid plateau with little or no elevation, low rainfall, and a different and unlimited insect fauna. Where constant conditions in illumination, temperature, and limited rainfall prevail, there is practically nothing to limit the multiplication of species able to survive the widely prevailing nrid conditions of this continent, of which the rainfall is less than 10" over 100,000,000 sq. miles. 1 A multitude of plants, so similar in appearance that it is difficult on casual acquaintance to separate the different species, are evolved, especially as annuals or ephemerals of the eremaea. To quote the genus Didiscus again, which well exemplifies this theory : according to Domin 2 the Calycina section represents the oldest form, limited to four species, of which two are Papuan, one extending to N. Borneo and the Philippines ; the others occur in Celebes and N.E. Australia respectively. Pseudo-calycina, considered atavistic, is represented by one Papuan species and two in N.E. Australia. The rest, or Eudidiscus, are all mostly annuals, younger types, represented largely in W. Australia, with one in the Polynesian Islands, where restricted conditions wouid .limit the evolution of the many closely allied species, so marked a feature of Australian xerophytic types. Open plant associations at the Angi lakes belong to the " Opportunity " category, as under normal conditions these areas would be in forest, being only kept open by the means of artificial burning at regular intervals. It is extremely interesting in this respect that it should be possible to compare two distinct associations of this type, developed under absolutely opposed conditions, viz. : 1. The open marsh at 7000', sheltered in position, showing varying conditions of soil and constant conditions of moisture. 2. The Cladonia association of Koebre at 9000', where constant con- ditions of exposure and drainage prevail with a hard surface soil, sterilised by the constant burning. No more certain proof is afforded of the derived character of a flora than that offered in a forest country by open spaces due to artificial circum- stances, providing conditions, rainfall, and temperature are favourable. This fact is well exemplified in New Zealand and the Pacific Islands, where both rainfall and original plant covering once approached present Papuan conditions, but now, under the stress of competition with ombrophobous immigrant plants, not a single native species will be found in such areas. 1 Griffith Taylor, " Discussion on the 1 hysiography of Arid Lands," Report Brit. Assoc. Adv. Sci. for 1914, 303 (1916). 2 Domin, K., " Monographic der Gattung Didiscus (DC.)." Sitz. Ktin. bo'hni. Ges. d. Wiss. ii. Cl. (1908) 13, 20-23. 48 So much is this the case that the history and country of origin of immigrant man, to whose agency the presence of this " Opportunity flora " is indirectly due, can be pretty accurately determined by the nature of the invading plants. It is the absence of aliens in the " Opportunity " associations of the Arfak which form the engrossingly interesting feature in the phytogeography of this region. 1. Open Marsh. Endemism is the dominant note in the plant-covering of this marsh. Most of the dominant plants found there are now described for the first time, or were first collected by Beccari and Gjellerup. One or two have been previously described from New Guinea, as Erwcaulon leucoyenes and Trisetum latifolium ; others, to quote new Papuan records alone, are incidental wind immigrants from the Himalayas, as Xyris pauciflora, Polygonum strigosum, and Viola distans. The latter is unknown in Malaya with the exception of the Philippines, while the others reach N.E. Australia, and therefore their incidence in New Guinea was to be taken for granted. In Utricularia racemosa and 7. bifida, the former shows the Himalayan range and the latter is limited to Malaya, while Gahnia psittacorum, abundant on both these open areas and also at home in the forest, reaches E. Australia and Tasmania. Most of the other Cyperacese are cosmopolitan tempei-ate types, of incidental wind distribution, like the cryptogams, of which, in relation to the freshwater Alga?, Professor West writes " that all the species observed are ubiquitous, few of the tropical ascending to 7000', the one exception being Closterium Bacillum, known only from Burma." The lichens, most mosses and ferns, including the Lycopodiums of these areas, are also cosmo- politan, while the only Selaginella collected is endemic. The one plant to suggest man's agency was J)esmodium Scalpe, an unexpected representative of a genus that, may almost be described as alien to the Papuan mountain flora, but, growing on the site of van Oosterzee's and the Pratts' camps, it may be considered the one relic of alien intrusion. On Kinabalu this plant has so far only been found at Lobang, on the invariable camping-site. 2. Cladonia A ssociation of Koelre. Nonb of the ombrophobous herbaceous plants with the exception of the Riedelias, a feature of the open spaces of the S.W. ridge and of this summit plateau, were collected in the surrounding forest. They were all plants requiring constant illumination and low temperature for their development, of which the germination of the seeds would be inhibited under shade conditions. Some of these plants, such as the Dendroliums and Centrolepis, are common to the open spaces of the S.W. ridge and to the marsh ; of the 49 others, Didiscus has been recorded from Mt. Scrutchley in the south-east, and the Platanthera from the south-west. The presence of the other plants must be due to wind-incidence, and they would be derived from natural exposed areas above the tree-level on the mountains of greater altitude to the east and south of the country. The cryptogams, again, as is the case on the marsh, are all cosmopolitan. This remarkable ridge association of Koebre combines some of the most peculiar elements of what have been considered the Malayan, Polynesian, and Australian floras. The plants found there show roughly what the systematic enumeration of the species collected proves in detail, that the flora of the mountains of New Guinea, almost unknown outside the last ten years, must now be considered the axle of a wheel of distribution, of which the spokes alone have so far been familiar to us. This is in agreement with all recent work at similar or greater altitudes. Had that axle, even now barely investigated, been worked out first, we would, as a matter of course, speak of the dominance of Papuan elements in neighbouring floras as the German and Dutch botanists have already rightly suggested. SOME PLANT ASSOCIATIONS OF THE N.W. COAST. Dorei Bay. The chief plant association of Dorei Bay is that of the " korang " forest clothing the low coral-limestone range which rises immediately behind Mano- koeari to the height of about 500', in a gradual slope from the sea-shore. This forest is still in its pristine condition, as all the surface-water drains through the sterile and porous subsoil, to a certain level line, about 200' above the beach, which marks the issue of the small streams representing the drainage of the ridge. This line also limits possible cultivation, as below it the " korang " is covered with sufficient depth of soil, due mostly to the erosive action of thse streams, to allow of necessary but not luxuriant cultivation. The old "pisang" 1 plantations of the Alfueros, now run to seed, with secondary jungle upgrowth, abut on to the natural forest at this level, on which both the reservoirs collecting for the water-supply of Manokoeari are sit.u.-ited at different points. The peculiarities of this " korang " forest were noted by Forrest in 1750 (1, 111), who wrote "there being no underwood it is easy travelling under the lofty trees "; and Dumont d'Urville in 1827 (3, iv. 581) estimated the trees in the forest as 80-200' high, writing of a " sol degage, arbrisseaux clairsemes, 1 Banana. 50 fongeres de petite taille, et fort peu de plantes herbacees," and further on (3, iv. 602), " Tous les environs du Havre proprement ditsont occupes par des tbrets a 1'etat de nature, situees sur un sol entierement madreporique, qui s'eleve en pente tres douce " ; and, finally, Wallace, in 1852 (6, 173), describes " the Dorey promontory is a raised coral reef, and, geologically speaking, a very recent one. The beach is a mass of dead and broken coral, not yet ground into sand, quite impracticable for walking, and from this beach up into the jungle, and even on to the hill, to the height of 200' or 300', there is scarcely a perceptible change in the coral rock, and the masses of coral and shells that everywhere strew the surface. In some of the gulleys, however, I found traces of a core of stratified rock." I did not work over any of the gullies, as once off the "korang" range the conditions are all secondary, every inch of ground having been under present or past cultivation. With regard to the beach, the coral mentioned by Wallace had possibly been washed up by a heavy N.W. monsun, as our own beaches are often covered with shingle during the winter gales, to be dispersed again later. At the period of my stay it was certainly not an apparent factor. On the "korang" the soil is so thin that the coral is always visible, mostly covered with dead leaves. The most interesting portion is along the flat-topped summit where the surface is more even and advantageous to plants, and in parts small soak-areas hold shallow standing water. The immediate flanks proved barren of results, being very dry with great over- hanging outcrops of pure " korang "-like cliffs, too porous to offer much hold for plants. On what may perhaps be referred to as the drainage-line of the streams issuing from the range, quite a different type of undergrowth prevailed, almost luxuriant in character, comprising chiefly ferns, Zingiberaceous and Araceous plants. Trees. Most conspicuous were fine isolated examples of that magnificent palm *Pigafetta pilaris. Too beautiful to cut down, I only took some old fruit and the measurements of the immense leaves, those shed, with the old flowering rhachises, remaining piled around each tree, which in con- sequence form isolated spots in the forest. Dr. Beccari, however, with his personal knowledge of this locality and expert interest, had no difficulty in determining this splendid species. *Ficus myriocarpa, *F. celebica, *F. botryo- carpa, and F. Irachiata, the two latter with green receptacles which all contained water, were very general, with the large-leaved c Macaranga riparia, *Mallotus tilicvfolia, Aglaia Gibbsece, with large branching white racemes, * Euonymus javanicus, and *Albizzi.a moluccana. Climbing plants. All the trunks of the trees were covered with root- climbing epiphytes, as in the "korang" forest of the inundation-zone of 51 Geelvink Bay Pothos sp. not seen in flower or fruit, *Piper Forstenii with liuge leaves, also sterile. The scandent fern Thysanosoria dimorpho- phylla (PI. 4. fig. 7) with fertile fronds at the apex of the shoots, the type of a new genus, was very common with *Lygodium digitatum, Freycinetia lanceolata, *Drac,tvna angustifolia, *Flagellaria indica in huge examj)les running up the tallest trees in the forest, and *Zanonia macrocarpa enveloping most of them in its heavy curtains. Epiphytes. As in the littoral korang forest these were not numerous and not a single orchid uas collected. Possibly the thick swathing growth of climbing epiphytes and stem-clasping lianes may, to a certain extent, account for their absence, also perhaps the porous nature of the subsoil, which reduces the constant evaporation so essential to the support of the large epiphytic flora common to the primary high forest of the foot-hills. The moss *Pelekium trachypodum on dead wood, the fern *Antrophyum reticu- latum, with the white-flowered Mymecodia puloinata, were collected. Undergrowth. The hepatic *I)iimortiera velutina was found in patches, and the terns * Stenosemia aurita always in colonies, while *Asplenium laser- pitiifolium, *Aspidium Lenzeanum, with *Dryopteris truncata and *I)iplazium proliferum, the two latter like small tree-ferns in habit, were dotted about. Small colonies of the creeping * Hemiyraphis reptans, *Geophila reniformis, with the orchids, always grouped, Microstylis Gibbsete and the larger orange- green M. xanthocheila, \\ ith Liparis maboroensis var. bistriata. Of larger plants, *Centotheca lappacea and *Schleria margaritifera, the small white Draccena novo-guineensis with very screwed leaves, were abundant; Pellionia Vanhasseltii massed on a prostrate trunk and on the ground beneath, the stinging Laportea armata about 1-2 in. high, gregarious on a small soak- area, and the shrubby Arnarocarpus Wichmannii, about 1 m. high with dorsiventral branches, were scattered over the surface, which showed no understaging of shrubs and little even of young trees. In the denser undergrowth at the base of the forest, on the drainage-line, abounded *Aspidium pachyphyllum with fertile and sterile fronds, Aglaionema novo-guineensis about 1 in. high, the leaves crowded towards the top and the flowers with green spathes and white spadix and red fruit, and Cyrtosperma macrot urn ; Alocasia acuta, a peculiar plant with large fleshy leaves on long petioles, crowded at the apex of a stem about 2 m. high, with the flowers bunched in the axils of the leaves, was common, with the white *Peristrophe ialappcefolia, not previously recorded outside Java. Schismatoglottis dorensis spread in large colonies in more open places where the white translucent Clavaria Gibbsece and * Dictyophora phalloidea, always yellow, were plentiful, the presence of the latter being invariably revealed by the smell, with *G easier Jim hiatus und the smaller G. mirabilis var. tricliifer. 52 Clearings at Edge of Forest and Secondary Jungle. This form o tropical upgrowth was not so rampant as is usually the case, on account of the poor " korang " subsoil, but it proved more than usually interesting in character in showing such a large proportion of endemic Papuan and Moluccan species. In clearings Gigantochloa novo-guineensis, apparently cultivated, was in flower, the fine Pandanus Tabbersianus bearing fruits almost 1 m. long, *Mallotus tilicefolia, *MellocTiia arborea, *Kleinhofia hospita, *Tamarindus indica, *Rubus moluccanus, * Callicarpa erioclona, white, Premna nitida, also white-flowered with black fruit, and the ubiquitous * Wedelia biflora, with the scandent * Allopliyllus Cobbe, *Flagellaria indica, *Rhyssopteris timorensis, the bright yellow Sccevola novo-guineensis, *Merremia nymphceifolia, with the epiphytic -Loranthus Versteegii, of which the flowering, vertically hanging shoots bear dense red racemes, standing out at right angles to the stems for half their length, were all general. There has been some question lately about the fertilization of LorantJius sp. by butterflies. I can only say that, on vigorously pulling one of the long shoots in the first excitement at seeing such a peculiar Loranthus form, I was covered by showers of very large red ants. The denseness and position of the racemes would certainly favour fertilization by these insects. Just below the drainage-level of the forest-clad range, a damp area of some extent, shaded by a secondary unidentified upgrowth, proved a very good collecting-ground for fungi. In fact, almost the whole total of the species collected were found on this area, the " korang " forest itself being too dry in character to encourage this form of growth. Cultivation. In the grounds of the Residency, bread-fruit, Avocado pears, pumiloes, custard-apples, mangoes, Canarium nuts, jambu, limes and lemons, pine- apples, with small water-melons and hill-rice, have been successfully grown. Roses were always in flower in the garden, of which the chief interest was a well-grown young Araucaria Beccarii about 5 m. high, brought down from the Angi lakes by Mr. van Oosterzee, who had laid out and planted these gardens. The old convict gardener who looked after them lo.st no oppor- tunity of denouncing the " korang," which certainly outcrops in most inopportune places, and the shallow soil. Surrounding the " Pasangrahan " and the quarters of the " Pradjoerit," all the usual " sayur " were grown by the " Orang rante " in the well-kept gardens, and also by the Chinese, who all have their own plots of cultivated ground. 53 Mangrove Association at Langg&n. On the spit of land between the two bays a Papuan " campong " was built among the mangroves, where *^Egiceras floridum flourished as a round shrub about 1 m. high, covered with the white flowers and quaint fruit. On a tree overhanging the sea-water, * Lycopodium phlegmariodes with Dendro- bium pseudo-ealceolum in flower and various Dischidia spp. abounded. *Pandanus polycephalus with small red fruit bunched at the apex of the peduncle, common through the Moluccas, with *Excacaria Agallocha and the climbers * Tristellateia australasica, a mass of yellow flowers, *Derris uligi- nosa and Sarcolobus retusus, marked the land-edge of the mangrove-spit ; while Freycinetia Beccarii, Erythrospermu>n candid^,, the yellow-flowered *JDurandea parvifolia, a Gardenia sp., with *Pollia sorzogonensis as under- growth, were found where the ground was more consolidated, and young colonies of *Pigafetta pilaris were quite abundant towards the shores of the second bay. Wousi and Genbela. Wousi, in earlier times the watering-place for all the boats calling at Dorei Bay, where the Papuan " campong," with a fringe of houses built over the sea, still stands, is a tiny valley cut out of the " korang " range by the action of the stream, the range from this point gradually sloping to the level ground. At the time of my stay Wousi was also the site of the Military Bivouac, and the base for the work of the Exploration detachments so suc- cessfully organized from 1907 by the military authorities at Amboina. The survey having been completed by the successful results of Captain Opper- man's Expedition (27, 542-3) the bivouac has now been closed. In the military cartographical office hung a huge map of Dutch New Guinea, originally blank, filled in by degrees with tracings of the work of each Exploration detachment in turn. When I was shown this map only one blank space remained, viz. the source of the Mamberamo River. With the successful results of Captain Opperman and Mr. Langeler that blank space has also disappeared, and with it this admirable era of organized exploration is closed, the whole of the Dutch possessions in North New Guinea being now mapped out. At Wousi, under the shade of the overhanging trees, the ideal and classical anchorage was reserved for naval and military needs, and all the ground underneath the beautiful century-old trees behind the beach was cleared and grass sown, forming a fine green sward, where the quarters of the military and naval officers were built. Farther up the valley were the open barracks for the native troops, carefully arranged with intersecting white paths, beyond which again excellent gardens had been made up the bu.iki of the stream, where all the vegetables for the needs of the forces were 54 grown, a herd of the beautiful little Bali cattle being also kept for regular killing ; on these occasions the requirements of the official residents of Manokoeari were also taken into account. Genbela. Beyond Wousi secondary forest with Malay houses at intervals bounded the beach to Genbela, where fine sands run out to the cape that limits the bay to the east. In the. forest the fungus *Favolus scaber, the lianes * Entada scandens and Mucuna Krdtkei were collected. *Peristroplie jcdappcefolia, *Hemi(/r aphis reptans and H. dorensis, with *Geopldla reniformis, formed constant undergrowth. On the edge of the sandy beach *Pandanus dubius grew in clumps, the young plants unbranched with thin leaves about 3 m. long; the old trees about 8 m. high, with many branched crowns and shorter leaves of much stiffer consistency, bore large round heads of glaucous mericarps. Behind this beach, on a level stretch of country with intermittent sago-swamps, the native plantations were mostly situated. Island of Roon. We stopped a day at Djende, the chief place on the island, of which the gneiss J formation carries a mainland type of vegetation. Djende lies in a deep sheltered bay, the native houses being all built over the water, like a miniature Brunei. It is surrounded by hills about 500' high, which, from the little seen, seemed to carry a very interesting association of plants. Along the road bordering the bay, Anthoceros bullato-spongiosus associated with the minute mosses *Garckea phascoidea and *Wilsoniella pellucida and young plants of * Lycopodium cernuum grew in the shade, while *Sclileria margaritifera and 0tanthera novo-gmneensis were found in the open. On the forest slopes Cyailiea runensis grew as undergrowth, where the orchid Vrydagzynea elongata was growing sporadically with Centotlieca lappacea ; * Trichomanes bipunctatum was collected as an epiphyte, also * Piper Forstenii in flower, with hanging yellow ? spikes, about 4 dm. long. The forest was in a very dry condition, rather a surprising fact, taking the heavy rainfall into account (p. 13), also considering it was the rainy season ; this fact again proves the fallacy of the all-embracing term rain- forest applied in general to all and sundry tropical forest formations. The actual rainfall is by no means the dominant factor, as it is the limiting characters, no matter how small their incidence, which have to be taken into account ; that is to say exposure, soil, and drainage, while should drier A. Wichmann's Uerichte in Jiull. nos. 43, 44 & 40 v/d Maatscli. ter bev. v.in het Natuurk. Onderzoek d. Nederl. Kolon. (N. Guinea Exped. 1903, Bull. nos. 3, 4 & 6). 80. Leiden. 55 conditions obtain for only one week, that week will limit the conditions of the forest formation. In N. New Guinea as in Brit. N. Borneo I saw no forest that answered to the description of rain-forest. 1 Humboldt Bay. This bay forms a most beautiful inlet, broken in outline, surrounded by mountains and protected by a small island, on which a very picturesque Papuan pile-village with an imposing " spirit house " is situated. As the Tidorese never penetrated so far (8, 87) the natives here are still unspoilt in primitive condition. The Government Station is at the head of the bay, just behind the beach. On a marshy spot near the " campong " * Tlioracostachyum hypolytroides, a sedge with white upper leaves, bracts, and inflorescence, grew abundantly. On the banks of the river Gigantocldoa novo-guineensis was in flower, as at Manokoeari, also *I)raccena angustifolia. We entered the " rimbu/' the Controleur having most kindly found me two Malays who knew the forest, on the edge of which a handsome Casuarina with a spreading crown, most distinct in habit, was abundant. Ascending the ridge, clothed in fine high forest on deep brown soil, we found as undergrowth * Trichomanes javanicum var. rhomboideum, *Diplaziuin maximum, with *Selaginella plumosa, creeping, the palms Licuala montana with red fruit and *Arenga microcarpa, the aroid Q Holochlamys Seccarii, like a Caladium in habit, and the under- trees *Pipturus argenteus and Clerodendron Lindawianum var. glabrescens, the latter with handsome white flowers and black fruit, also the climbing *Polypodium normale and Calamus humboldtianus. Over the crest of the ridge, where great mounds of dead leaves testified to Megapode activity in building their nests, we descended by a stream into quite a different type of vegetation, one of those sudden changes in comparatively small areas, which so constantly characterize the mixed tropical forest. Along the course of the stream, in which I picked up pure alabaster, much fancied by the Papuans for nose-ornaments, * Trichomanes humile was found on rocks with * Vittaria elongata, the hanging fronds to 2 m. in length. The fungi *Polyporus arcularius and *Hirneola polytriclia grew on dead wood. As undergrowth the handsome Pteris torricelliana with fronds 1*50 m long, and //i/j><>lt;j>is grandifrons on a rhachis about 3 cm. through, bearing magnificent single deltoid fronds about 5 m. long, the petiole about 3 m., and lamina 2 in. long, rising at intervals from an underground rhizome, \vas 1 L. S. Gibbs, " A Contribution to the Flora and Plant Formations of Mt. Kinabalu and the Highlands of Brit. N. Borneo," Journ. Linn. Soc., Bot. xlii. (1914) 8. 56 quite distinct from anything in ferns so far familiar. *Angiopteris evecta was abundant, and Alsophila straminea well represented. Begonia humboldtiana with mottled leaves and pink flowers grew all along the banks in large clumps, but it was impossible to find any normal fruit, the capsules being all swollen through the action of some insect. Ficus conocephalifolia, most aptly named, with enormous leaves and bearing red receptacles, was a common tree. Bosnik, on the Island of Wiak. On the return to Manokoeari, half a day spent on Bosnik, on ne of the Schouten Islands, was interesting, it being the first time the steamer had called there, this new Government station having been built as an alternative to Mosmer as a point of call. The island is coral-limestone, a high ridge rising behind Bosnik, and low-lying flats in front of the latter are rapidly being cleared for the " campong " plantations. Under the strand trees bordering this area * Cycas ctrcinalit, the ferns * Dryopteris stenobasis, *Diplazium polypodioides, with *Fleuria ruderalis, Ophiorrhiza insularis, Piper bipunctatum, and P. bosnicanum, the latter epiphytic on a tree, were found. As this was quite a new station, the strand trees were not yet stripped of their epiphytic treasures, and one splendid example, a veritable garden of various plants, yielded *Lycopodium phlegmaroides, *Psilotum flaccidum, the orchids Hippeopliyllum alboviride, Dendrobium inconspicuum and D. potomo- philum, Eria rigida var. papuana, and Sarcanthns bicornis, all with incon- spicuous flowers, with the handsome Medinilla rhodorhachis with pink flowers. All the streams drain through the " korang " into the sea at the beach level, so that it is possible to drink fresh water from the salt. At all the other stopping-places the time was too short to do more than collect a few plants, or, the islands being entirely under cultivation, possibilities in collection were limited to the beach. SYSTEMATIC RESULTS. Over 330 plants were collected in the Arfak, of which 100 have proved new to science, with one new natural order and five very distinct new genera. Of. the new species perhaps the most interesting are a Dacrydium, the first species in fruit to be described from New Guinea, a Libocedrus, a genus new for Dutch N.W. New Guinea, and a Kentia. A new species each in Trimenia, Telminthodia, and Backhousia, establish Papuan preponderance in those genera, while a new species in Patersonia, Centrolepis, and Hibbertia represent new generic records for New Guinea. I/- ,-2/A ^e'*<.* . He. Y* V - 57 The few widely distributed plants included comprise about 30 new records for New Guinea, of which the most interesting are Xyris pauci/ora, Bulbostylis capillaris var. trifida, Spathoglottis aurea, Polygonum strigosum, Viola distans, and Hydrocotyle rotundifolia. Several of Beccari's Hatam plants have been re-collected, viz. Riedelia orchioides, Palmeria arfakiana, in a J example, of which the ? is so far de- scribed, Styphelia trochocarpoides,and Dicrotrichium brevipes. Of Gjellerup's plants, as would be expected in working over the same ground, the dupli- cate examples are more numerous, especially in the orchids, but even in that order Smith has described 20 new species with four new varieties, and in Ericaceae two Rhododendrons, four Vacciniums, and one Diplycosia prove new, with one variety in Styphelia in Epacridacese. Collections made subsequently round Manokoeari (Dorei Bay) on some of the islands along the coast, and at Humboldt Bay, have been separately enumerated, no two species proving common to both the mountain and coast flora. About 150 plants are comprised in this list, which includes one new genus and several new species in ferns, with 27 new species in other genera, and interesting new records, of which the larger portion is more Malayan in type. Wide distribution of endemic plants is again a very marked feature. In working out these collections I must express my thanks to the expert botanists who have so kindly determined those orders in which they were interested. To Dr. J. J. Smith of Buitenzorg I am under especial obligation, not only for undertaking the Orchidacese, Ericacese, and Epacridacese, but also for many valuable details concerning work already done in the Arfak, and in arranging for the services of one of his trained native collectors who accompanied me to New Guinea. I am indebted to Dr. Odoardo Beccari for working out the Palmse ; Dr. Valeton for the Ziugiberaceae and Rubiacea? ; and M. Casimir de Candolle and Professor J. Macfarlane for the Piperacere, Meliaceajj and the Nepenthacese. In London my thanks are due to Drs. Stapf and Rendle and the staffs of Kew and the British Museum for much kind help ; and in particular to Professor G. S. West for the deter- mination of the Freshwater Algoe ; Miss A. Lorraiu Smith for the Lichens ; Dr. A. B. Rendle for Pandanacese, Cyperacese, and Gramineje, and a new genus in Urticacese ; Dr. 0. Stapf for Utriculariaceae and valued criticism ; Messrs. J. Ramsbottom, A. Gepp, L. Hutchinson, E. G. Baker, H. N. Ridley, and J. R. Drummond for Fungi, Bryophytes and Pteridophytes, Euphor- biacese, Melastomacese, Moracea3, and Solanacese ; and to Mr. S. Moore for the Labiata?, Acanthacese, and Composites, and his great kindness in revising the proofs. The plants may be consulted at the British Museum, Kew, Leiden, and Buitenzorg, and in my own collection (on loan to the British Museum). r 58 PRINCIPAL BIBLIOGRAPHY FOR DUTCH N.W. NEW GUINEA. 1. FORREST, T. A Voyage to New Guinea and the Moluccas from Balambangan. London, 1779. (2nd ed. 1780.) 2. LESSON, P. Voyage autourdu Monde sur la Corvette La Coquille. Paris, 1838. 3. DUMONT D'URVILLE, J. S. C. Voyage de V Astrolabe, 1826-29. iv. Paris, 1832. 4. DE BRUIJN KOPS, G. F. Bijdrage tot de kennis der Noord- en Oostkusten van Nieuw-Guinea. Natk. Tijdschr. Niederl. Indie, Deel i. (1850) 163-222. 5. WALLACE, A. R. Malay Archipelago. London, 1869. 6. . Notes on a Voyage to New Guinea. Journ. R. Geogr. Soc. xxx. (1860) 172-177. 7. . Island Life. London, 1880. 8. GOES, H. D. A. VAN DER. Nieuw Guinea, ethnographisch et natuurkundig onder- zocht in 1858. Bijdr. t. de Taal-, Land- en Volkenkunde, (2) v. (1862) 72. 9. D'ALBERTIS, L. M. Una mesa fra i Papuani del Monte Arfak. Boll. Soc. Geog. Ital. x. 2 (1873) 67-71. 10. ROSENBERG, H. VON. Reistochten naar de Geelvinkbai. 's Gravenhage, 1875. 11. BECCARI, O. Esplorazione dei Monte Arfak. Boll. Soc. Geog. Ital. i. (1876) 35-38. 12. . Malesia : i., ii., iii. 1877-1890. 13. SCHEFFER, R. H. Plantes de la Nouvelle Guinee. Ann. Jard. Buit. i. (1876) 1-60. 14. TETSMANN, M. J. E. Voyage a la Nouvelle Guinee. Loc. cit. 61-95. 15. ST. VRAZ, E. Reise nach Neu-Guinea. Petermann's Mitth. xliv. (1898) 232-35. 16. WARBURG, O. Monsunia, Beitrage zur Kenntniss der Vegetation des Slid- und Ostasiatischen Monsungebietes. i. Leipzig, 1900. 17. OOSTERZEE, L. A. VAN. Tijdschr. Kon. Ned. Aardr. Gen. (2) xxi. (1904) 998-1021. 18. VALETON, Tir. Plant* Papuanae. Bull. Soc. Dep. Agric. Indes Neerl. x. (1907) 1-70. 19. HARTMANN, A. Repertorium op de Literatuur betreffende de Nederlandsche Kolonien. 1866-1893. 1894-1905. 1906-10. Published by Murtinius Nijhoff. 20. GELDER, J. K. VAN.^ Verslag omtrennt eene geologische verkenning van de Mambei-amo-rivier op Nieuw-Guinea. Jaarb. Mijnw. Ned. Oost-Ind. Bat. xxxix. (1910) 87-112. 21. WICHMANN, C. E. A. Nova Guinea. History, i.-ii. (1909-12). 22. NOVA GUINEA, Botany, viii. (1907-13), (1913-15) ; xii. (1913-15). 23. SCHLECHTER, R. Die Orchidaceen von Deutsch-Neu-Guinea. (1911-1914.) 24. BRAAK, C. Het Klimaat van Nieuw-Guinea. Natk. Tijdschr. Nederl. Indie, Deel Ixxiii. (1914) 179-266, with Map. 25. SMITH, J. J. De Exploratie van Nieuw-Guinea. in Tijdschr. Kon. Ned. Aardr. Gen. (2) xxx. (1913) 77-8. 26. . Ibid. xxxi. (1914) 782. 27- . Ibid, xxxii. (1915) 542-3. [References to papers not reftrring to Dutch N. If. Jfeu- Guinea are girtn in the text.'] PLATE I. FIG. 1. FIG. 2. PLATE II. FIG. 3. Fio. 4. PLATE III. FIG. 6. PLATE IV. P. Highley, photo. THYSANOSORIA DIMORPHOPHYLLA, Gepp. 59 SYSTEMATIC ACCOUNT OF THE PLANTS COLLECTED AT THE ANGI LAKES, 7000-9000', IN THE ARFAK MOUNTAINS, IN DUTCH N.W. NEW GUINEA, IN DECEMBER 1913. [An asterisk denotes new records for New Guinea. Collectors' names and localities are given for Dutch New Guinea only.~^ THALLOPHYTA. FRESHWATER ALG^l. (G. S. WEST.) The geographical distribution has not been included, since all the species observed are ubiquitous except Closterium Bacillum Joshua, known only Erom Burma. The reason for this general ubiquity is the altitude, few of the tropical types ascending to 7000'. MYXOPHYCE^:. CHROOCOCCUS MINOR Nag. Gatt. einzell. Alg. 1849, 47, t. 1 A, f. 4 ; Rabenh. Flor. Europ. Alg. ii. 30. No. 5971. CHROOCOCCUS TURGIDUS Nag. Gatt. einzell. Alg. 1849, 46 ; Rabenh. Flor. Europ. Alg. ii. 32. No. 5971. MERISMOPEDIA GLAUCA (Ehrenb.) Nag. Gatt. einzell. Alg. 1849, 55, t. 1 D, f. 1. No. 5971. OSCILLATORIA AMPHIBIA Ag. in Flora, x. 1827, 632. No. 5971. OSCILLATORIA ANGUSTISSIMA W. & G. S. West in Journ. Bot. 1897, 300. No. 5723. OSCILLATORIA LIMOSA Ag. Dispos. Algar. Suec. 1812, 35. No. 5723. OSCILLATORIA TENUIS Ag. Algarum Decades, ii. 1813, 25. No. 5723. OSCILLATORIA TEREBRIFORMIS Ag. in Flora, x. 1827, 634. No. 5971. LYNGBYA VERSICOLOR (Wartm.) Gom. in Ann. Sc-i. Nat. 7 e serie, Bot. xvi. 1892, 147, t. 4, f. 4-5. No. 5971. STIGONEMA OCELLATUM Thuret, Essai de classification des Nostochinees, Ann. Sci. Nat. 6 e serie, Bot. i. 1875, 380. No. 5971. SCYTONEMA MIRABILE (Dillw.) Thuret. \_ = S. figuratum Ag. Syst. Algar. 1824, 38.] No. 5971. BACILLARIE^E. FRAGILARIA PARASITICA (W. Sin.) Grun. Van Heurck, Synops. Diat. Belg. 1885, t. 45, . 30. Odontidium parasiticum W. Sin. No. 5978. Epiphytic on Surirella robusta var. splendida. EUNOTIA BICAPITATA Grun. Van Heurck, 1. c. 1885, t. 35, . 11. No. 5971. EUNOTIA LUNARIS (Ehrenb.) Grun. in Van Heurck, 1. c. 1885, 144, t. 35, f. 3, 4, ot 6. No. 5971. EUNOTIA ROBOSTA Ralfs. Van Heurck, 1. c. 1885, 144, t. 33, f. 11-13. No. 5971. Both forms originally described by Ehrenberg as E, octodon and E. polyodon were plentiful. NAVICDLA APPENDICULATA Kiitz. Bacill. 1844, 93, t. 3, f. 28. Nos. 5723 and 5971. NAVICULA BICAPITATA Lagerstedt, Spitsberg. Diat. 1873, 23, t. 1, f. 5. No. 5971. NAVICULA EXILIS Kiitz. Bacill. 1844. Van Heurck, 1. c. 101, t. 12, f. 11, 12. No. 5723. NAVICULA MAJOR Kiitz. Bacill. 1844, 97, t. 4, f. 19. No. 5971. NAVICDLA PAR VA (Greg.), jy.gibba var. brevistriata Van Heurck, 1. c. 78, t. 6, f. 5. No. 5971. NAVICULA VIRIDIS Ktttz. Bacill. 1844, 97, t. 4, f. 18. No. 5971. COCCONEMA LiEVE (Nag.) G. S. West in Journ. Linn. Soc., Bot. xxxviii. 1907, 159. Cymbella laivis Nag. in Kiitz. Spec. Alg. 1849, 59. No. 5971. NITZSCHIA COMMUNIS Rabenh. Alg. no. 949 ; Flora Europ. Alg. i. 1864, 159. Van Heurck, 1. c. 184, t. 69, f. 32. No. 5723. 61 NITZSCHIA CONSTRICTA (Kutz.) Pritch. Infus. 1861, 780. N. dulia W. Sm. Brit. Diatom, i. 1853, 41, t. 13, f. 112. No. 5978. NITZSCHIA PALEA (Kutz.) W. Sm. Brit. Diatom, ii. 1856, 89. Synedra Palea Kiitz. Bacill. 1844, t. 3, L 27 ; t. 4, . 2. No. 5978. NITZSCHIA FRUSTULUM (Kiitz.) Grun., var. PERMINUTA (Grun.). Van Heurck, 1. c. t. 69, f. 4. No. 5723. HANTZSCHIA AMPHIOXYS (Ehrenb.) Grun. in Kongl. ST. Vet.-Akad. Handl. xvii. no. 2, 1880, 103. No. 5723. SURIRELLA ROBUSTA Ehrenb. in Ber. Akad. Berlin, 1840, 215. S. nobilis W. Sm. Brit. Diatom, i. 1853, 32, t. 7, f. 63. Var. SPLENDIDA (Ehrenb.) Van Heurck, 1. c. 185, t. 72, f. 4. No. 5978. All the specimens were covered with the small epiphyte Fragilaria parasitica (W. Sm.) Grun. CHLOROPHYCE.E. SCENEDESMUS ACUTiFORMis Schroder in Forschungsb. Biol. Stat. Plon, v. 1897, 17, t. 2, . 4. No. 5971. ULOTHRIX SUBTILIS Kutz. Phyc. Germ. 1845, 197 ; Tab. Phyc. ii. 1852, t. 85, f. 1. No. 5971. MIOROSPORA PACHYDERMA (Wille) Lagerh. Conferva pachyderma Wille in Ofvers. a K. Vet.-Akad. Forh. 1881, no. 8, 20, t. 1, f. 28-35. No. 5971. SPIROGYRA sp. (sterile). No. 5978. This species was undeterminable, but may possibly be one of the forms of Sp. decimina (Mull.) Kiitz. NETRIUM OBLONGUM (De Bary) Liitkem. in Cohn's Beitrage zur Biol. der Pflanz. viii. 1902,407." Var. CYLINDRICUM W. & G. S. West in Journ. Bot. 1903, 8 (sep.), t. 446, f. 10. No. 5971. 62 PENITJM sp. No. 5971. Several specimens were observed of a small Penium, but they were not in good condition for identification. It is possible that it is a form of P. pliy- matosporam Nordst. CLOSTERIUM BACILLUM Joshua in Journ. Linn. Soc., Bot. xxi. 1885, 652, f. 4-6. Long. 224 p ; lat. 34 /t. No. 5971. Known previously only from Burma. MICRASTKRIAS DECEMDENTATA (Nag-.) Archer in Pritch. Infus. 1861, 726. Long. 68 ft ; lat. 68 fi ; lat. isthm. 15 /A ; crass. 24 /*. No. 5971. COSMARIUM L^VE Rabenh. Flor. Europ. Alg. iii. 1868, 161 ; G. S. West in Journ. Linn. Soc., Bot. xxvii. 1899, 386, t. 10, f. 1-6. No. 5971. COSMARIUM PSEUDOPYRAMIDATUM Lund in Acta R. Soc. Scient. Upstila, ser. 3, viii. 1871, 41, t. 2, f. 18. No. 5971. COSMARIUM PUNCTULATUM Breb. W. & G. S. West, Brit. Desm. iii. 1908, 206, t. 84, f . 13, 14 ; t. 102, f. 22. No. 5971. HYALOTHEOA DISSILIENS (Sin.) Breb. in Ralfs' Brit. Desm. 1848, 51, t. 1, f. 1. No. 5971. LICHENES. (A. LORRAIN SMITH.) *CLADONIA COCCIFERA (L.) Willd. Fl. Berol. 1787, 361. Arfak Mts., Koebre ridge, growing thickly on burnt open summit, 9000'. Dec. 5598. Distrib. Cosmopolitan. *CLADONIA DIDYMA (F^e) Wain., var. MUSCIGENA Wain., Monogr. Clad. i. 1887, 141. Arfak Mts., Koebre ridge, 9000', growing thickly on burnt open summit Dec. 5720. Distrib. New Caledonia; West Indies; Central America j Brazil; Chile j Peru. 63 *CLADONIA VERTICILLATA (Floerk.) Schaer. Lich. Helv. Spic. 1831, 31. Arfak Mts., Koebre ridge, 9000', carpeting on burnt open summit. Dec. 5739. Distrib. Cosmopolitan. STICTA VARIABILIS Achar. Lichenogr. 445. Schum. & Laut. N. 30. Arfak Mts., Angi lakes, epiphytic in forest by ? lake, 7000'. Dec. 5905. Distrib. N.E. New Guinea. Polynesia to E. Africa. FUNGI. (J. RAMSBOTTOM.) PYRENOMYCETES. H YPOCREACE^E. PODOCREA COHNU-DAM^: (Pat.) Lind. in Engl. & Prantl, Naturfam. i. 365 (1897). Lower foot-hills, Arfak Mts., 500', terrestrial in forest. Jan. 5703. This species was described from China by Patouillard, who placed it in the genus Hypocrea. The stromata of typical species of this genus are crustaceous, cushion-shaped or hemispherical, and those s| ecies which have an elongated or vertical stroma are better separated off as a distinct genus. Saccardo placed them in the subgenus Podocrea, which Lindau afterwards (/. c.) raised to generic rank. Atkinson (Bot. Gaz. xl. 401 (1905)) showed that the generic name Podophyllum had meanwhile been proposed by Karsten (Hedw. xxxi. 294 (1892)) and therefore had priority. As Podophyllum was a name used by Linnaeus (1735 and Sp. PI. i. 723 (1753)) for a genus of Berberidaceae it cannot be duplicated amongst the fungi, and Podocrea must stand. XYLABIACE^E. XYLARIA POLYMORPHA (Pers.) Grev. Flor. Edin. 355 (1824). On dead trunk in forest. Lower foot-hills by Momi River, Arfak Mts. 500'. Dec. 6151. Distrib. World-wide. XYLARIA DOMINGENSIS (Berk.) Sacc. Syll. i. 315 (1882). On dead wood in forest, foot-hills by Momi River, Arfak Mts., 500'. 6136. Distrib. West Indies, etc. 04 BASIDIOMYOETES. AGARIC AC E^:. OMPHALIA ARFAKENSIS Ramsbottom, sp. nov. Tota alba. Pileo carnuloso, ex umbilicato infundibuliforme, margine primum inflexo, crenato, 1-2 cm. lato ; stipite fistuloso, 1 cm. longo, 2 mm. crasso ; lamellis decurrentibus, valde distantibus, postice latissimis (hinc triangulis), interdum dichotomis, venoso-connexis ; sporis suballantoideis, 3-guttulatis, 6-7 p X 3-4 p ; basidiis c. 25 p. X 5 p. Caespitosa ad lignum putridum. Bab. In forest, foot-hills by Momi River, ArfakMts., 500'. Dec. 6148. XEROTUS CINNAMOMEUS Ramsbottom, sp. nov. Mesopus ; cinnamomeus, coriaceus ; pileo convexo, centra papillate, c. 2 cm. diam., ad centrum plicato sulcato, margine acuto, integro ; stipite subfistuloso, deorsum cylindrico, sursum compresso sulcato, basi strigoso ; lamellis parcis (c. 10), valde distantibus, non decurrentibus. latis, intermixtis brevioribus, interdum venae- formibus ; sporis ellipsoideis, subgranulosis, 6-7 p X 4-5 p ; basidiis c. 50 p x 8 p. Ad lignum putridum. The radiating depressions on the upper surface of the pileus mark the position of the gills below. A trsmsverse section of the stipes shows a pseudo- sclerenchymatous tissue, which greatly simulates the sclerenchyma seen in Gramineae and certain other glumiferous Monocotyledons. Hab. In forest, lower foot-hills by Momi River, Arfak Mts., 500'. Dec. 6150. P O L Y P O R A C E JR. HKXAGOKA APIARIA (Pers.) Fr. Ii.picr. Syst, Myc. 497 (1838). Foot-liills by Momi River, Arfsik Mts., plentiful, 400'. Dec. 6149. Distrib New Guinea. India, Ceylon, Java, Philippines, etc. Fungi Imperfecti. LEPTOTHYRELLA SERICOLE/E Ramsbottom, sp. nov. Pycnidiis ampbigenis, sparsis, dimidiat-o scutellatis, radiato-cellulosis, atris c. 500 p diam., zona purpurascenti cinctis, sporis fusoideo-ellipsoideis vel clavulatis, rectis vel curvulis, hyalinis, multiguttulatis, continuis, dein medio uniseptatis. non constrictis, 25-35 p x 4-5 p. In foliis Sericoleae. On Sericolea arfakensls Gibbs (p. 147). 65 BRYOPHYTA. (A. GEPP.) The number of species brought home is but small, and obviously fails to represent adequately the richness of the moss-flora in the districts visited. Attention was exclusively directed to the collection of fruiting plants. HEPATICJE. *RICCAHDIA MAXIMA Schiffn. He[>at. Flor. von Buitenzorg, i. 57 (1900). Arfak Mts., Angi lakes, forest patch by ? lake, on dead wood in forest, 7000'. Dec. 5680. Distrib. Java ; Sumatra. *MARCHANTIA POLYMORPHA Linn. Spec. Plant, ed. i. 1603 (1753). Arfak Mts., Angi lakes, on open gravel-bank by ? lake, 7000'. Dec. 5906. Distrib. Cosmopolitan. MUSCI. SPHAGNUM JUNGHUHNIANUM Doz. et jMolk. Bryologia Javanic-a, i. 27, tab. 18 (1855). Arfak Mts., Angi lakes, in open marsh by ? lake, 7000'. Dec. 5967. Distrib. New Guinea. Malay Islands ; Formosa ; Japan ; India. SPHAGNUM NOVO-GUINEENSE Fleisch. et Warnst. in Engler, Pflanzenreich, Heft 51, Sphagnales, p. 520 (1911); Nova Guinea, xii. (1914) 127, t. xxxiv. B. Arfak Mts., ridge running up to Angi lakes, terrestrial in moss-grown forest, 8000'. Dec. 6006. Distrib. New Guinea (D.S.W., Goliath-Gebirge, de Kock}. *FUNARIA CALVESCENS Schwaegr. Spec. Muscorum, Suppl. i. sect. 2, 77, tab. 65 (1816). Arfak Mts., Angi lakes, on ground amongst bracken, "here burnt, on bank of ? lake, 7000'. Dec. 5938. Distrib. Warmer regions of the world. *RHODOBRYUM GIGANTEUM Paris, Index Bryolog. 1116 (1898). Arfak Mts., Angi lakes, abundant on ground in isolated t'orrst patch by ? lake, 7000'. Dec. 5892. " Very handsome moss." Distrib. Malay Islands ; India ; Bourbon ; Hawaii. RHIZOGONIUM SPINIFORME Bruch in Flora, 1846, 134. Arfak Mts., Angi lakes, on living tree in isolated forest patch by ? lake, 7000'. Dec. 5979. Also, on dead wood in same forest patch, 7000'. Dec. 5894. Distrib. New Guinea (N.E.). Throughout the tropics. DAWSONIA GIGANTEA C. Mull, ex Geheeb in Bibliotheca Botanica, Heft 44. 13 (1898). Arfak Mts., abundant in moss-grown forest of ridge running up to Angi Likes, in mossy forest sloping down to ? lake and in the open by same, 7000-8500'. Dec. 5523. Angi lakes on open banks by edge of ? lake and in forest, 7000'. Dec. 5935. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari ; D.S.W., Went- Gebirge, von Roemer ; Hubrecht-Gebirge, van Noulmys ; Siriwo-Fluss, Janowsky) . DAWSONIA BECCARII Brotherus et Geheeb in Bibliotheca Botanica, Het't 44, 13 (1898). Arfak Mts., ridge running up to Angi lakes, steep open slopes of gravel, 8000'. Dec. <$ & ? together. 5521. Ridge running up to Angi lakes in open gravelly spaces, 8500'. Dec. 6005. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari). RHACOPILUM SPECTABILE Reinw. et Hornsch. in Nov. Act. Acad. Cses. Leop. xiv. 721, tab. 40 (1828). Arfak Mts., ridge running up to Angi lakes, in forest, 8000'. Dec. 6121, 6122. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari ; D.S.W., Went- Gebirge, von Roemer ; N.E.). Malay Islands to Fiji and New Caledonia. SPIRIDENS REINWARDTI Nees ab Es. in Nov. Act Acad. Cses. Leop. xi. 143, tab. 17 (1823). Arfak Mts., Angi lakes, in isolated forest patch by $ lake, 7000'. Dec. 5591. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari ; S.E., Armit, MacGregor, Miclwlitz ; N.E.). Malay Islands. ENDOTRICHELLA ARF^KIANA C. Mull, ex Geheeb in Bibliotheca Botanica, Heft 44, 16, tab. 14 (1898). Arfak Mts., Angi lakes, with Rliizogonium spiniforme on living tree in isolated forest patch by ? lake, 7000'. Dec. 5979 pro parte. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari}. TAXITHELIUM SUBSTIGMOSUM Broth, in Engl. & Prantl, Nat. Pflanz. i. Abt. 3, 1092 (1908). Arfak Mts., Angi lakes, small forest by ? lake, on dead wood, 7000'. Dec. 5965. Distrib. New Guinea (D.N.W., MacCluer Bay, Naumann ; N.E.). 6? ECTROPOTHECIUM ARFAKENSE Broth, et Geheeb in Bibliotheca Botanica, Heft 44, 24 (1898). Arfak Mts., Angi lakes, on dead wood, in isolated forest patch by ? lake, 7000'. Dec. 5904. Distrib. New Guinea (D.N.W., Mt. Arfak, Beccari). HYPNODKNDRON DIVEKSIFOLIUM Broth, et Geheeb in Ofvers. Finsk. Vet. Soc. Forh. xl. 191 (1898). Arfak Mts., Angi lakes, forest patch by ? lake, carpeting in forest, 7000'. Dec. 5667. Distrib. New Guinea (S.E., Mt. Dayman, Artnii). PTERIDOPHYTA. (A. GEPP.) FILICALES. The references to the descriptions of the following ferns are to be found in Christensen's ' Index Filicum,' 1905-13. *TRICHOMANES DIGITATUM Sw. Arfak Mts., S.W. ridge running up to Angi hikes, 9000'. epiphytic in moss-grown forest. Dec. 6000. Distrib. Malay Islands to the Mascarenes and New South Wales. *TRICHOMANES PALMATIFIDUM K. Mull. Arfak Mts., S.W. ridge, Angi lakes, 8000', epiphytic in moss-grown forest, Dec. 5520: Distrib. Java. TRICHOMANES PALLIDUM Bl. Arfak Mts., Koebre Mt,, 8000-9000', epiphytic in forest. Dec. 5728. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Hellwig- Gebirge, von Roemer ; Mt. Carstensz, Kloss ; S.E.). Trop. Asia. Polynesia. TRICHOMANES APHLEBIOIDES Christ. Arfak Mts., Angi lakes, 7000', epiphytic in forest by ? lake, 5947 ; forest patch by ? lake, epiphytic, 6137. Dec. Distrib. New Guinea (D.N.W., R. Begowri, Gjellerup ; D.S.W., Noord Rivier, Versteeg). TRICHOMANES MEIFOLIUM Bory. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 5678. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Mt. Carstensz, Kloss}. Malay Islands to Polynesia and Reunion. 68 HYMENOPHYLLDM AUSTRALE Willd. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 5893. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari- N.E.). India. Malay Islands. Australasia. *HYMENOPHYLLUM PANICULIFLORUM Presl. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by $ lake. Dec. 6141. Distrib. Malay Islands and Japan. *HYMENOPHYLLUM SALAKENSE Racib. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 6138. Distrih. Java. *HYMENOPHYLLUM KUHZII Prantl. A rials. Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 5679. Distrib. Java. HYMKNOPHYLLUM (LEPTOCIONIUM) CEKNUUM Gepp, sp. nov. Rhizoma longe repens pilosum; stipites remoti erecti pilosi obsolete alati, ad 10 cm. longi, 1 mm. crassi. Frons 20-22'5 cm. longa. 5-7'5 cm. lata, lanceolata, rhachi omnino sed anguste alata pilosa ; pinnis alternis 20-jugatis contiguis ovato- lanceolatis pinnatifidis, costa alata pilosa ; pinnulis 1-2-dichotome lobatis ; segmentis ultimis planis serrulatis 2 inin. long., 0'5 nun. lat., costula anguste et dentate cristata. tSori in lobis brevibus pinnarum superiorum terminales, 1-8. Indusium vix ad medium divisum, valvis rotundatis integris, basi obovatum, longitudinaliter cristatum. Receptaculum exserturn. Hab. Arfak Mts., Angi lakes, 7000', terrestrial in spinneys by ? lake. Dec. 5964. The tall narrowed frond is in general habit somewhat like IF. Zol- lingerianum as figured by Van den Bosch (Hymen. Javan. t. 50), and bears about 20 pinna) on each side. In the dried specimens the frond-apex is cernuons and the pinnae complicato-decurved. The narrowly winged stipes and rhachis, tlie flat serrulate ultimate segments, the narrowly dentato-cristate costules, and the hairiness of stipes, rhachis, costfe, and costulesare characters to be noted. HYMENOPHYLLUM (LEPTOCIONIUM) CINCINNATUM Gepp, sp. nov. Rhizoma longe repens pilosum ; stipites remoti erecti pilosi obsolete alati, 3-5 cm. longi, 0'5 mm. crassi. Frons circ. 5 cm. longa, 2 cm. lata, lanceolata. rhachi alata pilosa ; pinnis alternis 10-12-jugatis contiguis (circ. 1 cm. longis, 0'5 cm. latis) pinnatifidis ; pinnulis 1-2-dichotome lobatis ; segmentis ultimis planis paucidentatis, 2-3 mm. longis, 0'5 cm. latis. Sori solitarii in lobis brevibus pin- narum terminates. JiH/itsimn ad medium divisum, valvis obtusis truncatisve integris, basi obovatum, parce longitudinaliter cristatum. JReceptaculum exsertum. Hab. Arfak Mts., S.W. ridge, Angi lakes, 8500', epiphytic in moss-grown forest. Dec. 5989. This species has sori much as in H. cernuum (No. 5964), but is not one- quarter the size, and is not cristate on the costules of the segments. It approaches //. holochilum Van den Bosch (Hymen. Javan. t. 34), hut has a stouter, more hairy rhizome and stipes (the latter winged throughout its length), and much narrower ultimate segments. The dried plants are very convolute. OYATHEA ARFAKENSIS Gepp, sp. nov. Stipes circ. 18 cm. longus muricatus inferne fuscus supeme griseo-purpurascens ; rhachis purpurea baud nitens supra ferrugineo-pubescens glande juxta cujusque pinnae basin instructa. Frons lineari-lanceolata, circa 80 cm. longa, 13 cm. lata, bipinnata ; pinnae alternae 35-jugatae stipitatae lanceolatae, 8 cm. long*, 2 cm. lata?. obtusae, pinnnlis 15-20-jugatis, stipitatis 1 cm. longis, 0'4 cm. latis, imbricatis oblongis obtusis, inferioribus paucipinnatis (segmentis rotundatis), superioribus lobatis vel crenatis vel integris (versus apicem) ; rhachibus pinnarum supra pubescentibus, infra squamulis albidis sparse instructis. Sori costales 4-5-jugati ; venuke pinnularum 5-6 parum conspicuae simplices f urcatuave. Textura coriacea ; frons supra griseo- purpurea, infra pallide brunnea. Indusium persistens hemisphericum. Hab. Arl'ak Mts., S.W. ridge, Angi lakes, 8000', undergrowth in moss- grown forest. Dec. 6008. " Tree-fern." . Tlie pinnae are not contiguous, but are attached to the rhachis at intervals of about 3 cm. In the dried specimen the shortly-stalked opposite rows of pinnulse are appressed to one another and directed upwards (apically), displaying the sori and concealing the upper surface. CYATHEA FUSCA Baker. Arfak Mts., Angi lakes, 7000', undergrowth in forest by ? lake. Dec. 5932. "Tree-fern, 3 in. in height (pinnae)." Distrlb. New Guinea (S.E.). ALSOPHILA ANGIENSIS Gepp, sp. nov. Stipes (?). Frons tripinnatifida ; rhachis purpureo-fusca sparse aculeata impolita sparse et breviter atro-hispida ; pinna brevi-stipitata, 37 cm. longa, 19 cm. lata, oblongo-lanceolata ; rhachis pinnse supra atro-tomentosa purpureo-fusca, infra sulcata furfuraceo-squamulata ; pinnula? suboppositse 30-jugatai contiguae imbricataeve sessiles horizontales lineari-lanceolata} acuminatse 1'5 cm. lataa, fere ad costam pinnatisecta? ; costa pinnulae supra fulvo-tomentosa, infra sulcata furfuraceo-squamu- lata ; segmenta 2G-jugata lineari-oblonga falcatula integra 3 mm. lata, supra f usco-purpurea glabra, infra glauca ; costula segmenti supra glabrum, infra pauci- pilosa et squamulis flavidis timbriatis versus basin vestita ; venulae inconspicuae tenues 14-jugat, 1-2-furcata?, pauci-pilosa?. Sori (?). Textura coriacea. 70 [Jab. Arfak Mts., Angi lakes, 7000', spinney by ? hike, undergrowth in forest. Dec. 5968. " Tree-fern, 3 m. in height, brown paleze on young stein and fronds. Brown and with thorns on old." This plant is represented by two pinnae which are entirely sterile. It is tentatively referred to Alsophila. In some respects it recalls A. glauca, but differs in having the pinnules crowded together and imbricated, and the rhachises dull and unpolished. The pinnules are inserted on the pinna- rhachis at intervals of about T25 cm. and overlap ; the middle pinnules are horizontal, tlie lower are deflexed. ALSOPHILA ARFAKENSIS Gepp, sp. nov. Stipes (?). Frons tripinnatifida, pinnis remotis alternis stipitatis. Rhachis (versus apicem) pallida, supra breviter tomentosa, infra glabra ruguloso-aspera ; rhachis pinnulae similis est. Pinnae infimae 28 cm. longse, 9-10 cm. latse. lineari- lanceolat* breviter acuminatae ad apicem pinnatifidae. Pinnulae rt 12-jugatae remotse alternae stipitatse, usque ad 5 cm. longse, 1 cm. latse, lineari-lanceolatai obtuse acuminatse, ad | pinuatifidse ; costa pinnulae supra tomentosula, sed versus apicem glabra, infra rugulosa squamulis paucis lauceolatis rubris ornata ; segmenta 12- jugata (apice excluso), 3'5 mm. lata, oblonga obtusissima crenata margine recurvata ; costula segmenti supra glabra, infra squamulata ; venulse 5-jugatae furcata? simplicesve. Sori 4-5-jugati prope costain dispositi, segmenti latitudinem haud obtegentes. Textura coriacea. Hal. Arfak Mts., S.\V. ridge, Angi lakes, 7000-8500', common under- growth in moss-grown forest. Dec. 5990. Also common about ? lake, in forest. " Slender tree-fern, stem 1 dm. in diameter and 1 m. in height. Fronds 1 m. long." The material consists of the two .lowest pinnoe and the top 37 cm. of a frond. A small dark gland is present at the base of the stalks of the pinnules. *DRYOPTERIS (LASTHEA) BEDHOMEI 0. Kuntze. Arfak Mts., Angi lakes, 7000', common in open marsh by ? lake. Dec. 5939. " Creeping rhizome." Distrib. Philippine Islands ; Ceylon ; South India ; South China. DKYOPTEHIS (LASTRKA) VILLOSIPES Gepp, sp. nov. Rhizoma erectum squamulis lineari-lanceolatis ferrugineis dentatis vel subintegris vestitum. Stipites 12'5 cm. longi, c. 1 mm. crassi, atro-pui-purei, inferne squamulis angustis capillaribus c. 1 mm. longis ferrugineis vestiti, superne breviter et sparse pubescentes et squamulis paucis lineari-lanceolatis ornati (ut etiam rhachis). Frons lanceolata 10 cm. longa, 3 cm. lata; pinnae subcontiguae alteruae plerumque brevi-stipitatae, versus frondis apicem sessiles, elliptico-lauceolatse 30-jugatae, horizontales, inferne 1'3-1'7 cm. longae, 0'4-0 g 5 cm. latae, fere ad if pinnatiticlae, ad apices crenatae vel subintegrae obtusss; segmenta 0-8-jugata approximata sub- falcata obtusa, segmento infimo superiori oblongo quain reliquis majori interdum 71 libero ; venulse segment! paucae pinnate dispositae inconspicuae. Sort singuli plerumque ad media segmenta, prope pinnae costam dispositi, indusio persistente. Textura coriacea ; lamina superne fusca, inferne griseo-viridis. Hal. Arfak Mts., Koebre Mt., 7000-8000', epiphytic in forest. Dec. 5627. The lower pinnules are slightly deflexed, and the lowest pair are a trifle shorter than the pair above them. The plant differs in every respect from I), viscosa. POLYBOTRYA ARFAKENSIS Gepp, Sp. nOV. Stipes paleis linearibus brunneis deciduis 0'5 cm. longis vestitus et earum cicatricibus muricatus, 15 cm. longus, atro-purpureus. Frons circa 1 m. longa, 40 cm. lata, bipinnata obovato-lanceolata ; rhachis atro-purpurea paleacea muricata (velut stipes) ad apicem haud evoluta, guperne breviter ferrugineo-tomentosa, inferne parum furfuracea, pinnas circa 9 alternas remotas dimorphas (inferiores 6 fertiles, superiores 3 steriles) gerens; pinnae brevi-stipitatae ad apicem abrupte (haud evolutse), steriles circa 30 cm. longse, 13 cm. latae, lanceolatae, fertiles circa 25 cm. longse, 7'5 cm. latae ; rhachis pinnae superne ferrugineo-tomentosa ; pinnulse alternae circa 12-jugatae stipitatae ; pinnulae steriles 6'5 cm. longae, 1'2 cm. latae, oblongo- lanceolatae acuminatae apice serrato, usque ad \ pinnatifidae, lobis (fere 20-jugatis) rotundatis crenulatis, venulis in lobis pinnatae dispositis c. 3-jugatis; pinnulae fertiles 3'5 cm. longae, 4-8 mm. latse, lineares usque ad costam pinnatisectae, lobis parvis oblongis sessilibus propter soros copiosos omnino obtectis. Textura subcoriacea. Lamina superne purpureo-brunnea, inferne pallidior glabra. Hal. Arfak Mts., S.W. ridge, Angi lakes, 7000', climbing in forest, clasping trunk to top. Dec. 5984. " Fronds 1 in. long, sterile like barren portion of fertile frond ; fertile frond distinct." This fern is remarkable for its dimorphous fronds and pinnae and for the arrested growth of its frond-apices. The material consists of one frond, the six lower pinnse of which are fertile, and the three upper are sterile. The apex of the main rhachis has failed to attain its full development, as also have the apices of the secondary rhachises (pinnse both fertile and sterile). Hence, the proper apices are absent. Whether the apical growth has been temporarily interrupted or permanently nrrested, and whether the arrest of growth is due to injury or is of normal occurrence in the life-history of the plant is uncertain. The lower pinnse are shorter than the others. The sterile pinnules are inserted at intervals of about 2 cm., the fertile at about 1*5 cm. The fertile pinnules recall those of Osmunda javanica in shape, but not in arrangement. Dimorphism of frond and pinnse occurs in the tropical American P. osmundacea. P. arfakensis climbs by means of its rhizome and puts out its fronds at right angles to the axis of the tree up which it climbs, 72 DIPTEKIS CONJUGATA Reinw. ArFak Mts., Angi lakes, 7000', growing on bank by edge of ? lake, in open. Dec. 5913. " Also abundant on open steep gravel slopes on S.W. ridge, running up to lakes from 8000-8500'." Distrib. New Guinea (D.N.W., Arfak Mts., Andai, Beccari ; D.S.W., Noord Rivier, Versteeg : submontane region, von Roemer : Mt. Carstensz, Kloss ; S.E.). Asia. Polynesia. OLEANDRA CUSPIDATA Baker. Arfak Mts., Angi lakes, 7000', common on edge of forest patch by ? lake, growing in clumps. Dec. 5559. " Up to 2 in. in height, leaves in interrupted whorls up the stem Also seen on S.W. ridge." Distrib. New Guinea (D.N.W., Arfak, Bcccari ; D.S.W., Noord Rivier, Versteeg ; Mt. Carstensx, Kloss}. NEPHROLBPIS ACUMINATA Kuhn. Arfak Mts., S.W. ridge, Angi lakes, 7000', climbing in forest. Dec. 6123. "Fronds with apical fertile portion or with entire fertile fronds." Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss). Malay Islands. Perak. HUMATA PUSILLA Carr. Arfak Mts., Angi lakes, 7000', epiphytic in forest, spinneys by ? lake. Dec. 5960. Distrib. New Guinea (N.E.). Melanesia. HUMATA NEOGUINENSIS C. Chr. Arfak Mts., Angi lakes, 7000', epiphytic in forest or creeping on edge. Dec. 5588. " Sterile and fertile frond." Distrib. New Guinea (D.S.W., Hellwig-Gebirge, von Roemer). HUMATA ALPINA Moore. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 5674. Distrib. New Guinea (D.S.W., Low country and Hellwig-Gebirge, von Roemer; N.E.). Malay Islands. Polynesia. *DAVALLIA DISSECTA J. Sm. Arfak Mts., Angi lakes, 7000', epiphytic in forest patch by ? lake. Dec. 5592. Distrib. Java ; Sumatra. DAVALLIA (PROSAPTIA) SCHLECHTERI C. Chr. Arfak. Mts., Koebre Mt., 8000-9000', epiphytic in forest slopes. Dec. 5634, 5640, 5625. Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss ; N.E.). 73 DAVALLIA (PROSAPTIA) CONTIGUA Spr. Art'ak Mts., Angi lakes, 7000', epiphytic in isolated Forest patch by ? lake. Dec. 5895. Distrib. New Guinea (D.S.W., Noon! Rivier, Virsteey : Mt. Carstensz, Klo948. Distrib. New Guinea (D.S.W., Hellwig-Gebirge, von Roemer; S.E.). Malay Islands. Polynesia. R 76 DRYOSTACHYUM SPLENDENS J. Smith. Arfak Mts., Angi lakes, terrestrial on edge of spinneys by ? lake, 7000'. Dec. 5970. Distrib. New Guinea (D.S.W., coastal lowlands, von Roemer}. Malay Islands. CHEIROPLEURIA BICUSPIS Presl. Arfak Mts., lower part of S.W. ridge, terrestrial in high forest, 5000- 6000'. Dec. 6134. Fertile frond only. Distrib. New Guinea (S.E.). Java ; Philippine Islands ; Formosa ; Liu-kiu Islands. GLEICHENIA VULCANIOA Bl. Arfak Mts., Koebre ridge, terrestrial on burnt open summit, common, 9000'. Dec. 5611. Angi lakes, abundant at edge of forest and spinneys by ? lake, 7000'. 5727. S.W. ridge, terrestrial on open steep gravel slopes, 8000-9000'. 5996. Distrib. New Guinea (S.B.). Malay Islands. GLEICHENIA GLAUCA Hook. Arfak Mts.. Angi lakes, abundant on edge of forest by ? lake, 7000'. Dec. 5668. " One frond climbing up to 7 m." Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss). Asia. Australia. Polynesia. *GLEICHENIA L^VIGATA Hook. Arfak Mts., Angi lakes, abundant on edge of forest patch by ? lake, 7000'. Dec. 5577. "3 in. in height, almost arboreal in habit. Leaves glaucous underneath." Distrib. Malay Islands. GLEICHENIA LINEARIS Clarke. Arfak Mts., Angi lakes, running up trees at edge of forest patch by ? lake, 7000'. Dec. 5575. " Massed/' S.W. ridge, common in forest, where more open, 7000-8000'. 5991. "Scrambling up to 7 m. ; half frond/' Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Noord Rivier, Verstteg ; S.E.). Tropics and subtropics. *ScHiz^A MALACCANA Baker. Arfak Mts., Angi lakes, common on open drier and gravelly parts of marsh by ? lake, 7000'. Dec. 5945. S.W. ridge, 850^ in moss in forest undergrowth. 6011. Distrib. Malay Islands. Burma. 77 LYCOPODIALES. LYCOPODIUM SERRATUM Tlmnb. Arfak Mts., Angi lakes, forest patch by ? lake, terrestrial in humus, 7000'. Dec. 5724. Distrib. New Guinea (N.E.). Asia. Polynesia. Mexico. Bourbon. LYCOPODIUM SQUARROSUM Forst. Arfak Mts., Angi lakes, spinneys by ? lake, 7000', epiphytic in torest. Dec. 5725. Distrib. New Guinea (N.E. ; S.E.). Asia. Polynesia. Mascarene Islands. LYCOPODIUM PINIFOLIUM Bl. Arfak Mts., Angi lakes, epiphytic under Araucaria forest by ? lake, 7000'. Dec. 5936. Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss ; N.E.). Java ; Borneo. LYCOPODIUM CERNUUM L. Arfak Mts., Angi lakes, common on open marsh by ? lake, 7000'. Dec. 5658, 5925. S.W. ridge, creeping in forest and on open steep slopes, 8000- 9000'. 5995. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Mt. Carstensz, Kloss ; N.E. ; S.E.). Tropics and some subtropics. LYCOPODIUM CASUARINOIDES Spring. Arfak Mts., Angi lakes, scrambling on edge of forest by ? lake, 7000'. Dec. 5944. Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss). Trop. Asia. LYCOPODIUM CLAVATUM L., var. WALLICHIANUM Spring. Arfak Mts., Angi lakes, open gravel bank by ? lake, 7000'. Dec. 5911. Open marsh by ? lake. 5556. Distrib. The species is recorded for New Guinea (D.N.W.*, Arfak Mts., Beccari ; S.E.). The species is cosmopolitan in temperate regions. The variety occurs in Java and India. LYCOPODIUM COMPLANATUM L. Arfak Mts.. Angi lakes, by open marsh and bank of ? lake, 7000'. Dec. 5912. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari). Asia. Polynesia. Africa, America; but mostly in the north temperate zones. H2 78 LYCOPODIUM VOLUBILE Forst. Arfak Mts., Angi lakes, scrambling up to 10 in. in forest spinneys by ? lake, 7000'. Dec. 5954. S.W. ridge, scrambling- in moss-grown forest, 7000-8500'. 5982. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; S.E.). Malay Islands. Australasia. Polynesia. PsiLOTUM FLACCIDUM Wall. Arfak Mts., Koebre Mt., epiphytic in forest, 7500'. Dec. 5633. Distril. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Noord Rivier, Versteeg ; Mt. Carstensz, Kloss ; N.E.). Tropics. SELAGINELLA ANGUSTIRAMEA F. Muell. & Baker. Arfak Mts., Angi lakes, terrestrial on edge of forest by ? lake, 7000'. Dec. 5726. " Growing in patches." Distrib. New Guinea (N.E. ; S.E.). SPERMATOPHYTA. CONIFERS. TA x ACE^E. DACRYDIUM NOVO-GUINEENSE Gibbs, sp. nov. Arbor parva, dioica, in statu juvenili erecta ; ramulis foliis longioribus 5-seriatim imbricatis pncditis, postea ramulis gracilioribus folia breviora gignentibus. Folia in statu plantae juvenili laxiuscula, squarrosa, acicularia, incurvato-pungentia, sectione triangulare ; folia in statu adultiori parva, dense imbricata, 5-seriata squami- formia, triangularia, apice pungentia, facie plana, dorso carinata ; folia ramoruin fertilium arete applicata, rhombica, acuta, dorso carinata, sectione sub-tetragona. Strobili masculi ignoti. Strobili feminei ad apicem ramulorum brevissimorum axillarium erecti, parvi ; bractese 24, lineares, apice apiculata?, incurvat^, carinata?, margine membranaceae, basi incrassata?, demum carnosse, rubrse, bractea fertilis unica, terminalis. Squama ovuligera fere usque ad basin libera, sub anthesi teres, cucullata. Ovulum in statu pollinifero liberura, squama ovuligera adhuc immatura circumdatum. Seminn erecta vel obliqua, bracteam summum sterilem superantia, ovoideo-angulata, viridia, nitida, basi ad | vel i squama ovuligera incrassato- cupuliforme cincta. Hob. Arfak Mts., crest of ridges and forest by ? lake, 7000-9000'. Seedling, ? lake, 7000'. Dec. 5508. Koebre" ridge, open summit, 9000'. ? . Fr. with mature foliage. 5648. Foliage of seedling branches 1 cm. across, the leaves 4 mm. by '5 mm. ; tbat of more adult branches 1*5 mm. across, the leaves 1*5 by *4 mm. ; foliage of fruiting branches 2 mm. across, the adpressed lozenge-shaped leaves 1*7 mm. by 1 mm. Strobilus 7 mm. by 3 mm., imbricating, with bracts 79 FIG. 3. Dacrydium noro-fluitieense Gibbs. A. Youth foliage ; B. Mature $, nat. size; C. Ovule in pollination-stage, ovuliferous scale appearing; D. a. Lateral, b. ventral view; E, F. Stages in development of ovule; H, I, J, K. Stages in development of strobilus; L. Seed with ovuliferous scale still attached; M. Strobilus, showing swollen bases of sterile bracts forming strobilus (seed shed) ; M.c. Ovuliferous scale. 80 gradually increasing in length till at the apex they are 3 mm. by - 5 mm., of which the apical bract only is fertile ; the mature strobilus, with swollen bract-bases, is 7 mm. by 4 mm., not including the seed. Ovuliferous scale is 2 mm. in height and 2 mm. broad, persisting in the axil of the fertile bract after the seed is shed. Seed 5 by 2 mm. All measurements of the fruiting-stages and drawings are from material in formalin. Pollination drops were seen on the young ovules. The mature fruiting-form of this species was only seen in one or two isolated examples on the open summit of Koebre* ridge, small trees +10 m. high, with short trunks and rounded crowns of rigid ascending branches, bearing numerous small red and fleshy cones. In younger stages this plant was very abundant on the crest of the S.W. ridge and in the forest round the ? lake. The foliage of the seedlings is plumose in character to '50 in. in height, when the small adpressed scaly leaves begin to appear. In the peculiar habit and the strap-shaped laminae of the bracts composing the strobilus, this species is quite distinct from known members of the genus. PODOCARPUS PAPUANUS Ridl. in Trans. Linn. Soc. ser. 2, Bot. ix. 158, quoad spec. Klossianum de Gamp III apportatum, non aliorum. Arfak Mts , Angi lakes, common on the surrounding ridges and in spinneys by the ? lake, 7000-9000'. Fl. ? ., Fr. (yg.). Dec. 5540. Diittrib. New Guinea (D.N.W., Arfak Mts., Hatam, Beccari ; D.S.W., Mt. Carstensz, Kloss). A fine tree, 30 m. in height, very like P. imbricatus Bl. in habit and in the dimorphic foliage, both seedling and youth form of foliage being identical tn both species in fact, like Beccari, I took the young plant for that species in the field, but the foliage of the mature tree is more spreading and distinct, the scales of the $ cones differ in shape, while the ? cones are larger and very glaucous in appearance. The fusion of the lamina of the fertile bract with the ovuliferous scale is also less complete than is the case in P. imbri- catus and the position of the seed is more oblique. Dr. Beccari most kindly sent me some of his Arfak material, published as P. imbricatus^ which, on comparison, proved identical with the above. The description of this plant being limited to the mature foliage and one <$ cone, I append what is necessarily wanting in the original diagnosis : Arbor alta ; truncus teres, erectus ; rami fere penduli, copiose ramulosi ; ramuli flexuosi, graciles. Folia dimorpha, juvenilia 1 cm. longa, 2 mm. lata, biseriatim expansa, plana, anguste linearia, mucronato-pungentia, decurrentia, apicem et basin versus decrescentia. Strobili feminei 1*2 cin. longi, '5 mm. lati, ad apicem ramu- lorum brevium erecti. Bractece 2 ve!3, 2 mm., glaucescentes, laminae teretiusculae, apice obtusae, 2 mm. longse, '8 mm. latse (inter se connatse), basi incrassata, verru- culosa 4-5 mm. longa, 1 vel 2 superioribus fertilia, lamina bracteis fertilioribus cum FIG. 4. Podocarjnts pupuanus llidl. A, B. Youth foliage ; C. Mature foliage, sterile ; D. Branch, bearing 2 $ strobili, one allowing two fertile bracts (pollination-stage) ; E. $ strobilus (fertilization-stage) : all uat. size. G, H, I. Same stages : J. Strobilus with oblique ovule ; K. Mature foliage : all X 4. 82" squama ovuligera tota longitudine connata. Squama ovuligera cum ovulo connata, viridis, nitida. Ovula juvenilia in statu archegoniale, oblique erecta, bracteas superantia, 7 mm. longa, 5 mm. lata. All the measurements given are from formalin material. On a mature tree in fruit small branches of the young foliage occurred up the stem. The ? strobili were in the pollination to the archegonial stage, and may be com- pared with similar stages in P. imbricalus (Gibbs, in Ann. Bot. xxvi. (1912) pi. xlix. figs. 1-6). The terminology given on p. 518, /. c., is that followed on the present occasion. There seems little doubt that Giulianetti's specimen, included by Mr. Ridley in his description of P. papuanus, represents P. imbricatns Bl., as Dr. Stapf had already nnmed it on the Herbarium sheets at Kew ; the two ? cones on the specimen prove the correctness of this determination. These cones were not seen by Reorders (Nova Guinea, viii. (1911) 615). PODOGARPUS RUMPHII Bl. Rumphia, iii. (1847) 214 ; Becc. Malesia, i. 179. Arfak Mts., S.W. ridge, in forest from 7000-9000'. Veg. Dec. 5985. Distrib. New Guinea (D.N. W., Arfak, Beccari ; D.S.W., Lobo, Zippelius ; N.E.). Moluccas, Amboina, Celebes. A tree, + 16 m. high, very common on the crest of the ridge, but not seen in fruit ; therefore the determination must remain uncertain. Leaves over 2 dm. long. PHYLLOCLADUS HYPOPHYLLUS Hook. f. Ic. PI. 889 ; F. Muell. in Trans. Roy. Soc. Viet. i. (1888) 32. Arfak Mts., common on ridges. Koebre ridge, open summit, 9000'. ? (very yg.). Dec. 5657. Seedling, 5657 a. S.W. ridge, foliage glaucous. ? (yg.). 5992. Distrib. New Guinea (S.E.). N. Borneo, Philippines. Very plentiful on ridges and in the forest, showing glaucous and non- glaucous foliage as on Kinabalu. Only the remains of some mature cones were collected. I fail to distinguish any difference between the above species and P. protractus Pilg. It is a very variable plant like other Phyllocladus spp., differing according to the age of the plant and whether the fertile branches occur on the old or the young wood. The series of variations obtained by me on Kinabalu are duplicated in the Arfak material, and appear also marked in the large amount of material from the Philippines available at Kew for comparison. P I N A C E M. AGATHIS DAMMAKA (Lum) ; A. Rich. Conif. Ixxxiii. t. 19 ; Rumph. Herb. Amboina, ii. 174, t. 57 ; Becc. Malesia, i. 180 ; Warb. Monsunia, i. 182 ; K. Laut. in Engl. Bot. Jahrb. 1. (1913) 48. Arfak Mts., S.W. ridge, 5000', in high forest. ? . Dec. 6127. Veg. (yg. plant). 5747. Distrib. New Guinea (D.N.W., Waigiou, La Billardiere ; Ramoi, Jobi Island, Ansus, Beccari}. Auiboina, (Celebes, Borneo, Java, Philippines, Malay Peninsula. A beautiful tree, about 40 in. high, with straight white shaft branching at the very top into a small, not very spreading crown with yellowish- green foliage. The distinct habit of these trees, of which three to four were growing near together, the crowns rising above the surrounding forest, was especially noted, to be confirmed by the excellent description given by Rumphius above, quoted by Parlatore (DC. Prod. xvi. 2, 374). Great lumps of white resin stood out on the straight white trunks. The trees were all too big to climb, so it was only possible to collect the old scales underneath, still in sound condition and in some cases attached to the axes of the cones. Many young trees, + 20 in. high, showed the fastigiate youth form characteristic of the genus (16, t. ix. f. 1). The leaves in the above collection were from a young plant 2 m. high, with one whorl of single branches + 2'75 in. long. As the leaves vary in size and texture with the age of the plant, it is impossible to base a species on vegetative characters alone, and A. Labillardieri Warb. is no doubt synonymous with the above, as already suggested by Lauterbach, I. c. ARAUCAKIA BECCARII Warb. Monsunia, i. 187. A. Cunningliamii Becc. in Malesia, i. 180 (non Ait.). A. Cunning kamii Ait., var. papuana Laut. in Engl. Bot. Jahrb. 1. (1918) 51. Arfak Mts., gregarious in parts in forest by ? lake. Seedlings. Dec. 5934. Young plant (1 m.). 5748. Old foliage and cones (pro-embryo stage). 5749. Distrib. New Guinea (D.N.W., Arfak Mts., Hatam, Beccari). These handsome trees were quite abundant on the eastern bank of the lake. They were about 25 in. in height and mostly in fruit. The large erect cones are borne on the horizontal uppermost branches of the trees. Many of the mature cones seen on the largest trees must be double the size given by Beccari, but the indurated bases of old leaves densely investing the trunks rendered climbing out of the question, and, having no axes with us, they could not be felled. A smaller tree, felled by " parang, v yielded two ? cones, one of which approximated to Beccari's measurements, while the other was smaller. Both the cones were in pro-embryo stage, with the seed- coat already quite indurated. The ovuliferous scale, not shown in A. Cun- ninghamii, is very noticeable in this species, as Beccari has described, and in this character it approximates to J. JIunsteinii K. Schum. (Fl. Kais. Wilhelmsland, 12 (1889)). The cone-scales a re more elongate than in A. Cun- nuiijkamii, with narrower base, more swollen apophysis, and a more pungent apex. The leaves of the fertile branch are more spreading, 1 cm. long and 84 3 mm. broad at the dilated base, with pungent apex more or less reflexed. In habit these trees differ from A. Cunningliamii, for, though fastigiate- pyramidal in growth, the branching is not so defined or symmetrical, showing little trace of the candelabra-like habit so familiar in the latter. There is also a difference in the size and shape of the leaves, in their stomatal markings, and in the much larger size of the ? cones. I must thank Dr. Beccari for material of his species, which he most kindly sent me for purposes of comparison ; also Dr. Stapf for carefully considering the points of difference between the two plants. (Pis. 1, 3, figs. 1, 5.) In both the available cones an apparent orifice (ori., JB) is visible on the swollen pulvinus of most of the bracts, behind the apex of the ovuliferous scale, possibly due to rapture of tissue. FIG. 5. Araucaria Beccarii Warb. A. Sporophyll; o.s., ovuliferous scale, w., wing. B. Lateral view ; ori., apparent orifice. C. Dorsal view. LlBOCEDRUS ARFAKENSIS Gibbs, Sp. nov. Arbor alta, in diversis ramis monoica ; rami teretes, cortice fusco obducti, ramuli oppositi, distichi. Folia decussatim opposita, in statu juvenili omnia conformia, linearia, in statu adulta adpresse quadrifariatim imbricata, diffonnia, marginalia navicularia, subacuta, 'coriacea, maxima ex parte adnata, complicata, apice solum libera, facialia plana, squamiformia, triangularia, carinata, acuminata. Strobili masculi in ramulis lateralibus solitarii, terminales, cylindracei. Antheroe GO -seriatse, spiraliter dispositse, stipite breve, connectivi appendicula squamiformi, late ovata, leviter peltata, chartacea, loculi 3-6, globosi, deorsum 2-valves. Strobili feminei in ramulis brevibus erecti ; bractese 4, elongato-ovatae, demum lignosse, appendicem magnum late ovatum obtusum antice proferentes. Nucula elliptica, subacuta, alata, altera subobsoleta, altera elongato-ovata. Hob. Arfak Mts., on ridges and in the forest by ? lake, 7000-8000'. 85 Seedling, ? lake, 7000'. Dec. 555(1 Youth form, S.W. ridge, 8000'. 5500. Koebre ridge, 8000'. formibus, subteretibus ; spicis parvis densifloris. Scandent, and of moderate size. Sheathed stem apparently 15-20 mm. in diam. Leaf-sheaths (in the small portion seen by me) closely armed with short small prickles. Ocrea extraordinarily large and long, as much as 50 cm. in length, very gradually long-acuminate, slightly inflated, enfolding the younger part of the stem, covered with thin fuscous-furfuraceous scurf, chartaceous, exsuccous, rigid, later splitting longitudinally on the outer side and not dissolving into fibres, fugaciously furfuraceous outside and orna- mented all over, and rather closely, with oblique, interrupted, slightly raised, spinulit'erous lines or ridges. Leaf-slieatli flagella slender and very long, armed, but not very regularly, with ternate or quinate claws, intermingled with smaller scattered prickles. Leaves non-cirriferous ; in one specimen about 1 m. long on the whole ; the petiole alone 35 cm. long, almost biconvex, slightly prickly, and with obtuse margins ; rhachis smooth, bifaced, with an acute salient angle above, and armed beneath, at first, with quinate and higher up with ternate claws. Leaflets very conspicuously approximate into very few distant groups (3 in the specimen at hand) with long vacant spaces of rhachis interposed ; in one specimen the leaflets are 7 on each side of the rhachis, of which 8 (4 on each side) form a basal group, and 7 (3 on each side with a deeply bilobed apical one) form the terminal group ; between these two main groups is another intermediate 12 94 formed of 2 opposite leaflets onlv ; the leaflets are lanceolato-ensiform, or oblanceolate, and taper almost equally to both ends, have the base rather acute, and are shortly and suddenly acuminate in a slightly bristly-spinulous tip ; are papyraceous, rather firm, glossy on both surfaces, but especially above, very slightly paler beneath, with an acute mid-costa and a few slender side-nerves, all quite naked on both surfaces, though at times the mid-costa is slightly spinulous above near the apex ; margins remotely and minutely ciliate-spinulous, more closely near the apex ; transverse veinlets very crowded, fine and sharp ; intermediate leaflets 40-50 cm. long, 3'5-4 cm. broad, the lowest smallest, those of the terminal group shorter, but not narrower. Male spadix somewhat shorter than the leaves (70 cm. long in one specimen), erect, strict, having an elongate pedicellar part bearing only one dense and narrow cupressiform panicle, about 20 cm. long (perhaps at times the spadix is longer, and with more than one panicle) ; the primary spathe is very elongate, and at first enfolds the spadix ; it is tubular, flattened with acute edges, closely sheathed in its lower part, and is produced above into an open, thinly membranous, lacerate, lanceolate-acuminate limb, it is sprinkled all over outside with minute tuberculiform prickles ; the panicle is composed of several gradually diminishing, very approximate, very densely flowered, short spikelet-bearing branchlets, inserted at an angle of 45, subtended by secondary thinly membranous, dry, lanceolate-acuminate, more or less lacerated, secondary spathes, only a little shorter than their respective branchlets ; the branchlets have the appearance of being small, simple, cylindrical spikes, 5-6 cm. long in the basal part of the panicle, gradually a little shorter above ; in fact, however, they are composed of several, very short, very closely drawn together, alternate, distichous spikelets, the lowest of which are 8-10 mm. long, with 7-8 very approximate flowers on each side ; the upper spikelets gradually diminish in length and number of flowers ; spathels bracteiform, membranous, concave with a triangular acute apex, surpassing the involucre; this is deeply cupular, or subcampanulate, truncate, not distinctly two-keeled, and bidentate on the side next to the axis. Male flowers very closely packed together, oblong, 6 mm. long; the calyx deeply 3-lobed; the corolla twice as long as the calyx, narrowing a little above to a bluntish apex. Female spadix and fruit unknown. Hab. Humboldt Bay, ridge behind the village, 700', scrambling in high forest. (?. Jan. 6267. A very distinct species, but with very marked affinities with C. macro- chamis Becc. It belongs to Group V. of my monograph, characterized mainly by non-cirriferous leaves, and by the leaf-sheaths furnished with a long clawed flagellum ; in that group it falls into the division which contains C.macrocldamys and other Papuan species, all having very large and elongate chartaceous ocrese. Even among these species C. humboldtiantti 95 is easily distinguishable by its relatively short leaves, having few, lanceolate, grouped leaflets, glossy and smooth on both surfaces ; by the sheaths prolonged into an excessively long, slightly inflate ocrea, which is orna- mented with oblique spinuliferous slightly raised ridges ; by the strict long-pedicelled spadix with a very densely flowered panicle ; and by the primary spathe having an elongate, thinly-membranous, lacerate, externally prickly limb. CALAMUS ARFAKIANUS Becc., sp nov. Gracilis, scandens, caudice 5-10 mm. diarn., vaginis flagelliferis, spinis gracilibus, interdum brevissimis, parce armatis, in ore et in ocrea brevi barbatis ; foliis non cirriferis, brevibus, petiolo brevissimo, segmentis paucis per greges paucos inter se remotos approximatis, lineari-lanceolatis, subulato-acuminatis, tenuiter papyraceis, unicostulatis, nervis secundariis utrinque 1-2 tenuibus, costa media et nervis omnibus levibus; segmentis intermediis circiter 20 cm. longis, 15-20 mm. latis, duobus terminalibus basi breviter connatis vel fere liberis; spadicibus valde elongatis et flagelliformibus ; spathis elongatis arete vaginantibus ; spadicis 3 ramis prinuiriis paucis, remotis, spicas numerosas detiexas f erentibus ; spicis majoribus 15-18 mm. longis, floribus utrinque 13-15 fere contiguis ; spatbellis concavis, bracteiformis ; spadicis $ ramis primariis paucis, remotis ; spicis majoribus 4-5 cm. longis, floribus utrinque 8-10 ; fructibus late ovato-ellipticis, abrupte mucrouatis, 15 mm. diam. ; scininLs albumin e homogeneo. IScandent and very slender. Sheathed stem 5-6 or at most 8-10 mm. in diam. Leaf-sheaths flagelliferous, not or only slightly gibbous above, at times almost, smooth, and only slightly (and very minutely) tubercled- spinulous in their upper part; but in some specimens rather densely covered with scattered or slightly confluent bristle-like spines, which become closer, considerably longer, and more hair-like, near and on the margins of the mouths of the same leaf-sheaths, and of the ocrea, which because of them looks bearded. Leaf-sheath flagella very slender and long, and armed irregularly with very minute scattered claws. Leaves non-cirriferous, short, 40-60 cm. long; petiole very short, 2-3 cm. long at most; rhachis fuga- ciously rusty-furfuraceous like the petiole, armed beneath with unequal, rather long-tipped, scattered or ternate claws. Leaflets few, 15-19 in all, very irregularly set, usually approximate in three groups separated by long vacant spaces of rhachis; they are narrowly lanceolate or linear-lanceolate, gradually subulately acuminate, thinly papyraceous, almost shining and about the same colour on both surfaces ; have the mid-costa very sharp above, and one, or at times two slender secondary nerves on each side of it ; mid-costa and nerves quite naked and smooth on both surfaces ; transverse veinlets rather sharp and not very crowded ; the margins have only a few distant, almost inconspicuous, appressed spinules, closer and more distinct in the apical part ; the leaflets of the interme iiute and lower groups are about 96 20 cm. long and 15-20 mm. broad ; those of the terminal group somewhat smaller ; the two terminal are the smallest, free, or more or less connate, at the base. Male and female spadices similar, flagelliform, very long and slender, with a very few distant partial inflorescences ; primary spathes very narrowly tubular and elongate, very closely sheathing, more or less armed with scattered small claws ; the lowest slightly compressed, the others cylindraceous. Male partial inflorescences (primary branches) zigzag sinuous, spreading, inserted outside their respective spathes with a distinct axillary callus ; the lowest 18-20 cm. long, with 7-8 rather distant spikelets on each side; the upper ones gradually shorter, and with fewer spikelets; secondary spathes 8-10 mm. long, smooth or at times very slightly spinulous, tubular -infundibuliform, or with a narrow flattened base, and slightly enlarged above, very closely sheathing, obliquely truncate and glabrous at the mouth, where acute at one side ; spikelets flattened, 6 mm. broad, comb- like with very approximate or contiguous bifarious flowers ; spikelets strongly deflexed, inserted just at the mouths of their respective spathes ; the lower spikelets 15-18 mm. long, bearing 13-15 flowers on each side; the others gradually shorter; spathels bracteiform, broad, concave, acute, strongly striately-veined; involucre cupular, obsoletely two-keeled and two-toothed on the side next to the axis. Male /lowers very close together, ovate, sub- acute, 2 - 5 mm. long; the calyx deeply 3-lobed, strongly striately-veined ; the corolla twice as long as the calyx. Female spadix has shorter partial inflorescences and fewer but larger spikelets than the male spadix ; the largest (lowest) spikelets are 4-5 cm. long, and carry 8-10 flowers on each side; spathels broadly infundibuliform, truncate; involucrophorum obliquely campanulate ; involucre deeply cupular, truncate, as long as the involucro- phorum ; areola of the neuter flower rather conspicuous, subcupular. Female flowers ovate, inserted at an angle of about 45. Fruiting perianth almost explanate, not pedicelliform. Fruit broadly ovoid-elliptical, 17-18 mm. long (not taking into account the beak), 15 mm. broad, equally rounded at both ends, suddenly topped by a beak 2'5 mm. long ; scales in 21 longitudinal series, rather deeply grooved along the centre, straw-yellowish with a very narrow dark brown marginal line; tip bluntish. Seed irregularly globose, with homogeneous albumen. Hob. Arfak Mts., Angi lakes, common all through the forest round ? lake at 7000-8000'. "A small rotang, excellent for tying." 5977 (? type- specimens) with the sheathed stem 8-10 mm. in diam. and rather densely prickly leaf-sheaths. The male plant (6144, Dec., Mt. Arfak, 5000' on the S.W. ridge, scrambling in high forest) : very slender; sheathed stem 8-10 mm. in diam. ; leaf-sheaths nearly smooth. It is evidently related to the Australian C. Muelleri, from which it mainly differs in the leaflets not being distinctly 3-costulate, and in the female spadix 97 having the flowers inserted at an angle of about 45 (not horizontal), and in its larger fruit. C. arfakianus is characterized by its slender stem, by the cirriferous leaf- sheaths, and short non-cirriferous leaves, with a few narrowly lanceolate leaflets approximated into a few (three) distant groups; the two terminal leaflets free, or but slightly united at their bases ; by the very slender and long filiform spadices, not very dissimilar in the two sexes, and by the ovoid- elliptical, suddenly-beaked fruit, having the seed with equable albumen. CALAMUS ARFAKIANUS var. IMBERBIS Becc. Vaginis inermibus, in ore ininiine barbatis ; ocrea brevi truncata ; segmentis interdum obsolete 3-costulatis. Differs from the type in having the leaf-sheaths unarmed and strongly striate longitudinally, and not bearded at their mouths ; in the ocrea being short, truncate, and marcescent, and also without bristles ; the leaflets in this variety have the side nerves more distinct, and at times appear almost 3-costulate. Hob. Arfak Mts., S.W. ridge, leading to Angi lakes ; scrambling in forest at 7000'. ? . Dec. 6143. CALAMUS PRATTIANUS Becc., sp. nov. Gracilis, scandens, caudice circ. 1 cm. diam. ; vaginis flagelliferis, spinis graci- libus subsetiformibus densiuscule obsitis, in ore dense longiuscule hispido-barbatis ; foliis non cirriferis, brevibus ; petiolo brevissimo ; segmentis paucis, inaequidistantibus, saepe utrinque geminatis (non per greges remotos approxiinatis), anguste lanceolatis, subulato-acuminatis, chartaceis, fimiis, unicostatis, utrinque nudis vel superne in costa media et in nervis secundariis duobus inconspicue remoteque spinulosis ; segmentis intermediis 15-18 cm. longis, 20-25 mm. latis, superioribus sensim minoribus, duobus terminalibus basi liberis vel breviter unitis ; spadice c? gracillimo, elongato ; inflorescentiis partialibus paucis, brevibus, remotis, spicas paucas erecto- patulas ferentibus ; spicis majoribus 3 cm. longis, flores 13-15 utrinque ferentibus, remotiusculis, spathellis breviter lateque infundibularibus, truncatis. Caetera desunt. Scandent and very slender. Sheathed stem about 1 cm. in diam. Leaf- sheaths flagelliferous, not or but slightly gibbous above, rather densely covered with scattered or slightly confluent bristle-like spines, becoming closer and considerably longer and more hair-like near the mouths of the same leaf-shaths and their ocrese, which latter are very short and on account of these bristles look densely bearded. Leaf-sheath flagella very slender, long, and armed irregularly with very minute scattered claws. Leaves non- cirriferous, short (the few seen by me are 35 cm. long); petiole very short, or almost obsolete ; rhachis fugaciously rusty-t'urfuraceous, armed with a line of solitary, rather long-tipped claws. Leaflets few, about 16 in all ; very irregularly set, usually approximate in pairs on each side of the rhachis, with irregular vacant spaces of rhachis interposed ; they are narrowly 98 lanceolate, gradually tubulately acuminate, papyraceous, rather rigid, almost shiny and of about the same colour on both surfaces, have the mid-costa very sharp above, and one or two slender secondary nerves on each side of it ; underneath they are quite smooth, but above have occasionally a few very minute spinules on the mid-costa, and on a secondary nerve on each side of it ; transverse veinlets much interrupted and not very sharp ; the margins almost smooth, or very inconspicuously spinulous, more distinctly so at the apex; the intermediate leaflets are 15-18 cm. long and 20-25 mm. wide, the lowermost and the upper ones gradually smaller, the two terminal being the smallest, free, or more or less connate at the base. Male spadix flagelliform, very long and slender, with very few and distant partial inflorescences ; primary spathes very narrowly tubular and elongate, very closely sheathing, more or less armed with scattered small claws, the lowest compressed with rather acute margins ; the others cylindraceous, somewhat produced and lacerated at apex; partial inflorescences ascendent, rather short, the lowest and largest about 10 cm. long with 5-6 spikelets on each side ; 'secondary spathes very narrowly tubular-infundibuliform, or with a flattened base, and slightly enlarged above, obliquely truncate at their mouths, and there slightly produced at one side into a triangular point, which embraces the bases of their respective spikelets ; spikelets spreading, the largest (lowest) about 3 cm. long, with 13-15 flowers on each side; the upper are shorter, and have fewer flowers ; spathels shallowly and broadly infundibuliform, truncate, strongly striately- veined ; involucre cupular, obsoletely two-keeled and slightly two-toothed on the side next to the axis. Male flowers ovoid, inserted at an angle of about 45, not in contact with each other, but separated by the blades of their respective spathels. Female spadix and fruit unknown. Hob. Arfak Mts., near the Monswoon Snoon (cT lake) at about 8000'. Collected by Mr. A. E. Pratt. It is closely related to C. arfikianus, from which it differs in the leaves having unequidistant, but not distinctly grouped leaflets, on that account resembling C. Muelleri more than C. arfakianus. From C. Muelleri it differs also in having larger spikelets, and having ascendant non-contiguous flowers. C. Prattianus is characterized by its slender stem with cirriferous leaf- sheaths and short non-cirriferous leaves ; by the ocreai and the mouths of the leaf-sheaths being furnished with long bristles ; by the lanceolate unequidistant (but not distinctly grouped) and almost smooth leaflets ; by the very slender and long-flagelliform male spadix having only a few partial inflorescences furnished with but few spreading spikelets ; these have briefly infundibular spathels and bear the flowers inserted at an angle of about 45, a little apart one from the other. 99 CENTKOLEPIDACEiE. CENTROLEPIS NOVO-GUINEENSIS Gibbs, sp. nov. Planta perennis, densissiine csespitosa, multifoliata. Folia basi vaginata, vagina hyalina, margine longe pilosa, lamina lineare, canaliculata, leviterscabriuscula, obtusa. Pedunculus gracilis, scabriusculus ; gliunis 2, inaBqualibus. calyptratis, margine hyalinis, unitioris. Flares bractea unica hyalina stipati. Stamen 1. Ovarium 2-loculare, loculis 2, superpositis. Styli 2. Hah. Arfak Mts., Angi lakes. Abundant in marsh by ? lake where open on sand, 7000'. FJ., Fr. Dec. 5566. Koebre ridge, abundant on open burnt summit, where damp, 9000'. Fl., Fr. 5646. Plants 2'4 cm. high, with branched stems, all densely matted with white hairs at the base. Leaves 8 mm. long, with vagina 3 mm. long and 1 mm. broad, the lamina 4 mm. by 5 mm. Peduncle 1*5-1'8 cm. in length, much longer exported in 5646, each bearing one spikelet, 4 mm. long. Glumes 2, the largest 3'5 mm. long, and the smaller 3 mm. Hyaline bract 4 mm. long, o>ate acute. Stamen with versatile anther 1 mm. in length; filament 3'5 mm. long. Ovary 1*5 mm. long. Styles 3 mm. Very near C. pliilippinensis Merr., but is a smaller plant with shorter, radially arranged, spreading leaves, hairy at the base, with, as far as seen, one flower in each glume. This interesting record affords further striking proof of the wider distribution of a so-called Australian genus. One species alone was previously known from Asia, viz. C. cambodiana Hance, till Merrill in 1906 found C. pliilippinensis on Mt. Halcon at 2400 in. In 1910 1 C. kinabaluensis was found by me in N. Borneo, on Mt. Kinabalu at 13,000', which now, wich the above, gives 4 well-defined Asiatic species, indeed 5, if an undescribcd plant in the Kew Herbarium should belong to this genus ; and no doubt further exploration of the magnificent mountain ranges of New Guinea with their unlimited possibilities will yield many more. 1 CKNTBOLEPIS KINABALUENSIS Gibbs, = C". philijipinemis W. B. Turrell, ex Gilbs in Journ. Linn. Soc., Bot. xlii. (1914) 172 (non Merr.). Planta perennis, 2-3 cm. alta, caespitosa. Folia disticha, imbricata, + 1'2 cm. longa, vagina 6 mm. longa, 2 mm. lata, hyalina, glabra, lamina 8 mm. longa, 5 mm. lata, obtusa, sctosa. Pedunculiis 1'2 cm. longus; glumis 2, inaequalibus, inferne latis concavisque, superne angustatis, apice obtusis, dorso carinatis, 4*5 mm. long. Flares 6-8, in quaque gloma 34, flos qnisque bractea hyalina, 3'5 mm. longa. Filamentum 3 mm. longum, anthera 1'2 mm. longa. Ovarium 1'5 mm. longum, stylis 2 mm. Hab. Kinabalu, granite core, summit, cracks in granite, forming mats. Fl.,Fr. Feb. 4207. A note of Turrell'* on the herbarium sheet at Kew states, " The number of flowers in a ppikelet varies in this plant. I have found 4, 6 or 8 flowers in one spikelet. The Philippine type is described as having 4 flowers and all the flowers of Merrill 6160 that I have dissected have this number in each spikelet." The plant also differs in the glabrous distichous leaves and the shape of the glumes, which have bluut apices. 100 X YllID A C E M. *XYRIS PAUCIFLORA Wildm. Phytogr. i. 2, t. i. f. 1. Arfak Mts., Angi lakes, common in open marsh by ? lake. FL, Fr. Dec. 5928. Distrib. India (" Foot-hills of the Himalayas, in marshes, from Nepal eastwards, N. Bengal to Burma," Hook. f. Fl. Br. Ind. vi. 365), Ceylon, Malay Peninsula, N. Borneo, Celebes, Philippines, and China. N.E. Australia. ERIOCAULACE^E. (A. B. RENDLE.) ERIOCAULON LEUCOGENES 1 Ridl. in Trans. Linn. Soc. ser. 2, Bot. ix. (1916) 240. Arfak Mts., Angi lakes, common on marsh by $ lake, where open and sandy, 7000'. Fl. Dec. 5567. Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss). " Flower heads mauve." The plants show a great range in size from 6 cm. to 20 cm., the heads varying in diameter from '5 to 1 cm. JUNCACE^E. (A. B. RENDLE.) JUNO us LAMPOCARPUS Ebrh. Calam. n. 126 ; Ridl. in Trans. Linn. Soc. ser. 2, Bot. ix. (1916) 231. Arfak Mts., Angi lakes, open marsh by ? lake, 7000'. Fl. Dec. 5927. Distrib. Temperate Europe and Asia, N. and E. Africa, Eastern N. America, S.E. Australia and New Zealand. LILIACE^E. DIANELLA CGERULEA Sims, Bot. Mag. (1801) t. 505 ; Schum. & Laut. 219 ; Nova Guinea, viii. (1914) 996 ; Ridl. in Trans. Linn. iSoc. ser. 2, Bot. ix. (1916) 230. Arfak Mts., in high forest below, and in open spaces on the S.W. ridg", running up to Angi lakes, also in open marsh by ? lake, 7000-8000'. Fl. Dec. 5519. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup; D.S.W., G.-lieb, Brandenliorst; Mt. Carstensz, Kloss; N.E.). Philippines, New Caledonia, and Fiji ; N.E. Australia to Tasmania. Acaulescent, and with light blue flowers. LUZURIAGA ASPERICAULIS Hall. f. in Nova Guinea, viii. (1914) 991, t. clxxxi. Arfak Mts., twining in mossy forest on S.W. ridge running up to Angi 1 A species of Eriocanlon was also collected by Miss Gibbs on Mt. Kinabalu at 12,000' (no. 4209) by Kadamaiau torrent, on the granite core near the summit of the mountain ; it was mixed with Centrolepis kinabaluensis Gibbs (no. 4209) (see p. 99). .It is a csespitose plant, forming small cushions 2 - 5 cm. high, with glabrous leaves 2-2'5 cm. long, jhl mm. wide in the middle, linear-tapering from a broad membranous base. The specimens are all Bterile. No Eriocaulon has hitherto been recorded from the grauitd core of the mountain. 101 lakes, 8000-0000'. Fl. Dec. 5536. Common in forest slopes by ? lake, 7000'. Fl., Fr. 5744. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). A common and very pretty twiner with white flowers. The fruit, which was not available in Dr. Gjellerup's specimens, is a roundish berry, 8 mm. loi'g by 6 mm. broad, 1-2-seeded, with persistent style and remains of perianth- tube at the base. IRIDACE.E. PATERSONIA NOVO-GUINEENSIS Gibbs, sp. nov. Planta perennis ; caulis brevissimus. Folia circiter 5-8, scapum sequantia vel superantia, disticha, linearia, rigida, acutissima, in marginibus dense rufo-pilosa, tenuiter inultistriata. Spathce subsequales, lineari-oblongae, acutissirnse, carinatie (carina ad f rufo-pilosa vel glabra) striatae, scarioso-marginatse. Bracteoe spathis similes nisi magis membranacese necnon angustiores brevioresque, ad apicem piloso- marginatae vel glabrse. Flores in spatha 2-3, sessiles; tubus gracillimus, vix exsertus; lobi exteriores obovati, patentes, in carina dorsali ad apicem dense barbati, interiores nulli. Filamenta in tubum integrum connata. Stylus superne, leviter incrassatus ; lobi stigmatici lineari-oblongi, papillosi. Hob. Arfak Mts., Koebre ridge, plentiful on open burnt summit, 9000'. Fl., Fr. Dec. 5600. A small rigid plant, with gleaming white or pale mauve flowers, + 1-3-5 dm. in height. Leaves from 1-2-3-5 dm. by 5-6 mm., characterized by a line of matted branched hairs, containing brown colouring-matter, up the dorsal keel, and from the vagina to the apex on the adaxial margin of the leaf. Sheathing scales at the base of the leaves from 3'5 cm. long, lanceolate acute, brown, the old bases persisting round the stem. Scape 7'3 cm. long, more or less exserted from the sheathing leaves, which it exceeds or equals, or it may be shorter. Spathes 3'8 cm. by 5 mm. Perigonium tube 2 cm. long, lobes 8 mm. long, including apical tuft of hairs 1 mm. long. Anthers 3 mm. Stigma-lobes 2 mm. This species is near P. Lowii Stapf. It differs in the somewhat broader leaves with marginal line of brown hairs and the generally white corollas with an apical tuft of hairs to each lobe. The transverse section of the leaf shows many more fibro-vascular strands and is more attenuated at the margin than in the above species. There is also no trace of an inner perianth, as Stapf also found for the Kinabalu plant. The marginal line of matted branched hairs is common to some Australian species, viz. P. sericea R. Br. and P. pygmcca Lindl., but in every case apparently the hairs are hyaline and much finer in texture. This genus, for many years supposed to be endemic in Australia, now includes three Malayan mountain types, viz. P. borneensis Stapf and P. Lowii from Mt. Kinabalu in N. Borneo, the latter also common to Mt, Halcon in the Philippines, and the above. 102 ZlNGIBEEACE^. (TH. VALETON.) ALPINIA ARFAKENSIS K. Sch. in Engl. Bot. Jahrb. xxvii. (1899) 296. Var. SUBSESSJLIS Val., var. nov. Folia subsessilia minora, glaberrima, siccando valde convoluta coriacea. Ligula 20 mm. longa vel longior, valde macerata. Anther cm. longum, 112 clinandrio parvo, concavo cum costula longitudinal! in rostellum decurrente, auriculis obtusis. Anfhera cucullata, longe ultra thecas producta, ambitu subquadrangula vel melius 7-angula, apice abrupte subulato-acuminata, basi truncata, c. 0'14 cm. longa, connective in costulam longitudinalem incrassato, thecis suborbicularibus. Pollinia 4, geminata, lateraliter compressa, triangula, extus convexa. Ovarium pedicellatum 6-costatum, c. O9 cm. longum. Gapsula oblongo-obovoidea, basi acuta, c. 1-1 cm. longa, pedicello c. I'l cm. longo. Hob. Arfak Mts., Angi lakes, forest patch by $ lake, edge of forest in humus, terrestrial, 7000'. Fl., Fr. Dec. 5G89. The first species of the section Rhachidilulbon recorded from New Guinea. It much resembles L. brgvistylis J. J. S. (L. montana Lindl. var., brevistylis J. J. S.). Flowers green with a brown labellum. LIPARIS ( PLATYCHILUS ?) GIBBSIJR J. J. S., sp. nov. Rliizoma repens, radicans, teres, vaginis magnis ad c. 3'25 cm. longis tectuin. Pseudobulbi c. 2'5 cm. distantes, cum rhizomate angulum anguste acutum facientes, sicco tenues, teretes, c. 2'5 cm. longi, 1-folii, basi nonnullis vaginis magnis acute acuminatis ad c. 4 cm. longis cincti. Folium inarticulatum, erectum, lanceolatum, acuminatum. acutum, basi sensini in petiolum angustatum, nervis 5 majoribus subtus prominentibus, sicco membranaceum, c. 9'75-10 - 5 cm. longum, 2'l-2'25 cm. latum ; petiolus canaliculatus, nervosus, c. 2'5-4 cm. longus. Inflorescentia in pseudobulbo mature terminalis, folio brevior, satis multiflora, pedunculo c. 4'5 cm. longo, vaginula angusta acuta c. 2'75 cm. longa ad basin, rhachide fractiflexa, c. 6 cm. longa. Bractcce adpressse, lanceolato-trian guise, acutissimae, concavse, 1-nervise, ad c. 0'55 cm. longse. Flores c. 16, quaquaversi, patentes, sepalis patentissimis, convexis. Sepalum dorsale oblongum, superne angustatum, obtusum, 3-nervium, costa media dorso apicem versus prominente, c. 0*65 cm. longum, 0'25 cm. latum. Sepala lateralia divergentia, oblique oblonga, f'alcatula, obtusa, vix apiculata, 3-nervia, costa media dorso pro- minente, c. - 67 cm. longa, 0'3 cm. lata. Petala reflexa, linearia, obtusa, basi leviter oblique dilatata, leviter convexa, 1-nervia, c. 0'57 cm. longa, medio O'OG cm. lata. Labellum porrectum, cum gynostemio angulum acutum faciens, basi gyno- stemio adnatum, in utraque parte canaliculse latse longitudinalis convexum, subtus convexum, e basi contracta quadrangula sensim dilatatum, 2-lobum, obcordatum, lobis remotis obtusis grosse crenatis cum lobulo brevi obtuso in sinu, basi contracta excepta ciliolatum, 3-nervium, nervis exterioribus ramosis, callo majusculo antice dentes 2 erectos parallelos falcato-triangulos acutiusculos utrosque basi antice in costulam brevem exeuntes gerente in basi, explanatum c. 0'67 cm. longum, antice 0'45 cm. latum, parte contracta c. 0'2 cm. longa, 0'275 cm. lata. Qynostemium superne valde hamato-incurvum, dorso convexum, basi valde dilatatum, medio labello adnatum, ovario multo latius, c. O32 cm. longum, basi 0'17 cm. latum, clinandrio concavo, intus ad basin rostelli dente acuto instructo, auriculis subquadrangulis Anthera cucullata, transverse quadrangulo-ovalis, apice truncata, basi emarginat? bidentata. Stigma, obverse rotundato-triangulum, margine elevatum. Ovariuii 6-sulcatum, cum pedicello c. 0'4 cm. longum, curvatum. Hob. Arfak Mts., Angi lakes, forest patch by ? lake, epiphyte in forest 7000'. Dec. 5Gi*trib. New Guinea (D.N.W., Arfak Mts., Gjellervp). 116 CERATOSTYLIS ( EUCERATOSTYLIS) ANGTENSIS J. J. S., sp. nov. Caules approximati, teretes, c. 12'5-15 cm. longi, macerati 0'17 cm. diam., basi vaginis tubulosis superne accrescentibus sese amplectentibus sicco membranaceis laxe reticulato-venosis brunneis ad c. 2*25 cm. longis tecti. Folium erectum, subulato- teres, acutiusculum, antice leviter longitudinaliter costatum cum sulco tenuissimo, c. 2'7 cm. longum, maceratum c. 0*125 cm. diam. ; vagina tubulosa, antice rumpens, c. 0'3 cm. longa. Inflorescentia fascicularis, pluriflora, squamata, pedunculis partialibus tenuibus, 1-floris, c. 0'3 cm. longis, basi vaginulam latam obtusam membranaceam c. 0'26 cm. longam gerentibus. Bractea ample cucullata, obtusa, ovario multo brevior, membranacea, c. 0'14 cm. longa. Flores parvi, bene 0'4 cm. longi, sepalis dorso parce pubescentibus. Sepalum dorsale oblongum, obtusum, 3-nervium, c. 0'25 cm. longum, bene O'l cm. latum. Sepala lateralia lacinia oblique oblonga concava pedem gynostemii bene superante decurrentia, mentum reversum ovario parallelum oblongum apice subinflatum obtusum bene O'l cm. longum formantia, oblique oblonga, apice angustata, obtusa, 3-nervia, c. 0'24 cm. tota fere 0'4 cm. longa, 0'125 cm. lata. Petala oblique oblonga, acutiuscula, falcatula, 1-nervia, c. 0'2 cm. longa, fere 0'07 cm. lata. Lalellum cum pedi gynostemii angulum fere rectum faciens, gynostemio parallelum, ima basi ultra apicem pedis gynostemii obsolete productum, supra basin leviter recurvum, con- cavum, gynostemium bene superans, basi unguiculato-angustatum, -| parte superiore contractum carnoso-incrassatum et utrinque convexum, obtusum, parte mediana marginibus incurvum et ciliolatum, 3-nervium, lineis 2 elevatis puberulis parallelis intus usque ad medium, explanation c. 0'35 cm. longum, medio 0'14 cm. latum, apice incrassato bene O'l cm. longo, fere O'l cm. lato. Gynostemium ultra medium bifidum, c. 0'14 cm. longum, brachiis parallelis, oblongis, obtusis, extus convexis, intus concavis. AntTiera cucullata, transverse suboblongo-elliptica, apice truncata, c. 0'0(5 cm. lata. Pes gynostemii reversus, cum ovario angulum acutum faciens, apice a sepalis lateralibus libero obtusangule incurvus, c. O'l cm. longus. Ovarium pubescens, c. 0'375 cm. longum. Hob. Arfak Mts., Augi lakes, isolated forest patch by ? lake, epiphytic, 7000'. Dec. 5718. A very inconspicuous species. The colour of the flowers is white. DENDKOBIUM ( CADETIA) SUBRADIATUM J. J. S. in Fedde Rep. xii. (1913) 27 ; in Nova Guinea, xii. (1915) 273, t. c. 171. Arfak Mts., Angi lakes, terrestrial in forest by ? lake, 7000'. Dec. 5542 and 5908. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). DENDROBIUM ( LATOURIA) RHOMBOGLOSSUM J. J. S. in Bull. Jard. Boh Buit. 2 e ser. n. ii. (1911) 9 j in Nova Guinea, xii. (1913) 44, t. xiii. 34. Arfak Mts., Angi lakes, marsh by ? lake, terrestial, abundant, 7000'. Dec. 5511. Koebre ridge, between c? an d ? lake, burnt open summit plateau, 9000'. Dec. 5009. Distrib. New Guinea (D.N.W., Goliath Mts., De Kock). 117 " Plant up to 1 ra., perianth segments magenta outside and white inside ; abundant." DENDROBIUM ( LATOURIA) CURVIMENTUM J. J. S., sp. nov. Caules (1 adest) tenuiter clavati, c. 6-nodes (absque internodiis summis foliiferis brevissimis), c. 19'5 cm. longi, apice 3-folii, internodio maximo incrassato infra folia c. 6-3 cm. longo. Folia erecto-patentia, ovato-elliptico-lanceolata, apicem versus angustata, acuta, .basi breviter contracta, sicco rigide coriacea, c. 5-67 cm. longa, 1-2-1'G cm. lata. Infloresceiitice e nodis summis inter folia, folia superantes, c. lo-florae, pedunculo tenui, c. 5'2 cm. longo, vaginulis c. 4 tubulosis acutis donato, rhachide sicco angulata, c. 4 cm. longa. Bractece late triangulae, acute, concava?, c. 0'3 cm. longae. Flores parvi, incurvi, carnosuli, glabri. Sepalum dorsale cuin ovario anguluin obtusum fornians, oblongo-triangulurn, dimidio superiore angustutnin, obtusum, concavum, 5-nervium, c. 0'54 cm. longurn, bene 0'3 cm. latum. Sepala latcralia ad pedem gynostemii decurrentia, mentum ovario parallelum valde incurvum breve crassum obtusum basi subconstrictum bene 0'3 cm. longum formantia, late oblique triangula, subobtusa, basi antice in laciniarn brevem triangulam dilatata, margine superiore valde rotundata fere quadrangula, concava, 5-nervia, fere 0'5 cm. longa, basi 0'5 cm. lata. Petala oblique anguste oblonga, falcatula, obtusa, basi leviter contracta, apice erosula, concava, 1-nervia, c. 0'46 cm. longa, 0'14 cm. lata. Lalelluiii pedi gynostemii insertum, unguiculatum, 3-lobum, concavum, carnosum, subexplanatum c. 0'4o cm. longum, 0'6 cm., bene explanatum 0'65 cm. latum, ungue pedi gynostemii parallelo valde recurvo, canaliculato, fascia longitudinal! bicostata ad basin laminae dilatata et in dentes 3 conicos exeunte instructo, dentibus lateralibus membra na angusta labello adnatis, callo rotundato verruculoso ante dentem inter- medium, carinula transversa utrinque ad basin lobi intennedii, lamina transversa, c. 0'3 cm. longa ; lobi laterales inexplanati porrecti, paralleli, gynostemium et lobum intermedium superantes, explanati lobum intermedium aequar.tes vel vix superantes, late oblique ovati, rotundati, concavi ; lobus intermedius transversus, bipartitus, concavus, subtus convexus et crasse carinatus, carina infra sinum in apiculum crassum conicum exeunte, lobulis oblique triangulis, margine antico rotundatis, obtusis, explanatis divergentibus, explanatus c. 0'14 cm. longus, 0'325 cm. latus. Gynostemium a dorso compressum, in utraque stigmatis parte obtusangule dilatatum, absque anthera c. 0-175 cm. longum, clinandrio concavo, margine denticulate, fila- mento incurvulo, oblongo, rotundato, auriculis paulo brevioribus, obtusis, leviter denticulatis. Anthera cucullata, transverse ovali-reniformis, basi rotunda to-biloba, apice breviter rotundato-producta, c. 0-16 cm. lata. Pollinia 4, in corpuscula 2 oblique obovata supra convexa subtus plana unita, lateraliter compressa, interioni qua m exteriora angustiora, c. 0'07 cm. longa. Stigma obverse triangulum, basi rotundatum, apice truncatum. Pes gynostemii cum ovario angulum acutum faciens, reversus, inedio fere rectangule incurvus, oblongus, obtusus, valde excavato-concavus, iiifxplunatus c. 0'3o cm. longus. Orariutn valde incurvum, 6-sulcatum, c. 0'33 cm. longum ; pedii-i'llus I'l cm. longus. Hub. Arfak Mts., Koebre ridge, forest, 8000-9000'. Fl. (green). Dec. 5647. 118 This Dendrolium belongs to the small-flowered species of the section Latouria. The shape of the labellum calls to mind that of I), dendrocolloides J. J. S. and D. latifrons J. J. S., but the habit is very different. DENDROBIUM ( TRACHYRHIZUM) LATIFRONS J. J. S., sp. nov. Rhizoma breve, vaginis brevibus tubulosis sese amplectentibus omnino obtectum, radicibus dense verrucosis. Caules simplices, elongati, supenie flexuosi, c. 47 cm. longi, internodiis c. 3'7-l'3 cm. longis. Folia patentia, oblonga, oblique bidentata, dentibus acutiusculis, costa media supra sulcata subtus prominente, marginibus in sicco recurvis, sicco tenuiter coriacea, ad c. 5'5 cm. longa, 1*8 cm. lata, superiora et inferiora decrescentia ; vaginae tubulosse, apice truncate, internodiis breviores. InJlorescenticB in caulibus foliatis axillares, vaginam dorso ad basin perforantes, arcuatae, laxe 3-5-florse, pedunculo cauli adpresso, c. 3-4 cm. longo, nonnullis vaginulis tubulosis ad basin, rhachide c. l'5-2 cm. longa. JBractets adpressse, triangulse, concavse, c. 0%3 cm. longae. Flores quaquaversi. Sepalum dorsale late triangulum, marginibus curvatum et base utrinque angulatum, obtusum, concavum, 7-nervium, c. O8 cm. longum et latum. Sepala lateralia cum pede gynostemii inentum obtusum formantia, antice omnino libera, late oblique triangula, obtusa, breviter crasse obtusa apiculata, margine postico basi angulata, 7-nervia, costa media dorso obtuse prominente, c. O95 cm. longa, basi 1'15 cm. lata. Petala oblique subovalia, basi lata, breviter obtusiuscule triangulo-acuminata, concava, minutissime erosulo-ciliolata, basi 3-nervia, c. 0'77 cm. longa, O53 cm. lata. Labellum ungue brevi lato 5-nervio pedi gynostemii insertum, 3-lobum, curvatum, subtus supra basin concavum cum nervis 5 obtuse prominentibus, intus fascia latissima longitudinali carnoso-incrassata subquinquecostata cum costis 3 mediants alte elevatis basi in appendicem reversam rotundato-trapeziformem obsolete retusam concavam carnosam unguem sequantem, c. 0'2 cm. longam, O35 cm. latam, producta in basi lobi intermedii in lamellam transversam porrectam excavationem anticam obtegentem verrucosam antice dentes c. 5 porrectis iiTegulariter conicis gerentem utrinque nonnullas costulas transversas emittentem terminante, intus papillosum, explariatum c. 1'13 cm., usque ad apicem loborum lateralium O93 cm. longum, ad lobos laterales c. 1'43 cm. latum, ungue c. O2 cm. longo ; lobi laterales erecti, patentes, oblique ovati, rotundati, basi lati, concavi, margine postico manifesto incurvi ; lobus intermedius latissimus, concavus, 2-lobus, medio plica supra concava subtus convexa et in sinu dente brevi triangulo donatus, lobis transverse oblique rotundato-quadrangulis, apice truncatis, irregulariter crenatis, angulum exteriorem versus repandulis, expknatus c. 0'43 cm. longus, 1'3 cm. latus. Gynostemium (foecundatum) bene 0'3 cm. longum, apice porrecto utrinque dente donato, auriculis triangulis. Pes gynostemii cum ovario angulum obtusum fere rectum formans, late oblongus, intus pubescens, basi dorso convexus, apice altius excavatus et dorso roturidato-gibbosus, c. 0'7 cm. longus. Ovarium (foecundatum) obconicum, 6- sulcatum, c. 0'4 cm. longum ; pedicellus c. 1*3 cm- longus. Hob. Arfak Mts., Angi lakes, terrestrial in marsh by ? lake, 7000'. Dec. 5558. This species is nearly allied to D. appendiculoides J. J. S., from which it differs, so far as the single specimens allow comparison, by its more robust 119 stems, less flowered racemes with the peduncle pressed against the stem, broader petals, the lip with different-shaped keels, shorter, blunter, not falcate lateral lobes, a shorter and much broader mid-lobe and a broader basal appendix, whereas the column-foot is much less excavated. Moreover, the plant was collected in a much higher altitude above the sea. The species recalls I), prosteriglossum Schltr. var. oltusilolum Schltr. to mind ; it is possible that the two plants will prove to be identical. The flowers are described as yellow. DENDROBIUM ( OXYGLOSSUM) TRIFOLIUM J. J. S., sp. nov. Caules approximati, c. 4-5 cm. longi, inferne vaginis mox fatlscentibus, apice c. 3-foliatL Folia erecto-patentia, lanceolata, oblique obtusa, breviter apiculata, omnino sed praesertim apice minutissime denticulata, costa media subtus prominente, sicco rigide coriacea, c. 2'6-4 cm. longa, 0'67-0'83 cm. lata ; vaginae inferne tubu- losse, superne canaliculato-conduplicatae. Injlorescentia pseudoterminalis, brevissima, c. 3-flora. Uractete rachidem amplectentes, late triangula3, acute acuminatae, ad c. 07 cm. longse. Flores erecti, majusculi, c. 3'1 cm. longi, sepalis sicco dorso parco nigro-punctatis. Sepalum dorsale ellipticum, obtusum, apice minutissime denticu- latum, 5-nervium, c. 1'25 cm. longum, 0'65 cm. latum. Sepala lateralia kcink elongata partim anguste lineari ad pedem gynostemii decurrentia, inentum rectum ovario adpressum c. 1'7 cm. longum, superne marginibus auticis longitudine c. 0'6 cm. connatis anguste calcariforme apice retusum formantia, parte anteriore oblique trkngulum, marginibus curvatis, subacuta, acute apiculata, apice minutissime denticulate, 7-nervia, costa media dorso prominente, c. 1-2 cm. longa, basi 2 cm. margin e antico 2*9 cm. metientia. Petala spathulato-rhombea, vix acuminata, acuta, superne papilloso-ciliolate, 3-nervia, c. 1'23 cm. longa, 0'57 cm. late. Labellum pedi gynostemii et gynostemio pamllelum et adpressum, pedi gynostemii longitudine c. 0'95 cm. adnatum, gynostemium superans, anguste lineare, apicem versus cliktetum, apice 3-lobuin, concavum, infra apicem subventricosum, ecallosum, inferne 3-, superne 7-nervium, expknatum c. 2'6 cm. longum, ad lobos kterales fere 0'4 cm., basi partis liberse 0'175 cm. ktum; lobi kterales breves, rotundati, papillosi ; lobus intermedius revolutus, triangulus, acutus, papilloso-cilioktus, c. 0-25 cm. longus, basi O16 cm. ktus. Gynostemium a dorso compressum, in utraque stigmatis parte obtusangulum, absque anthera c. 0'2 cm. longum, clinanclrio in utraque fikmenti parte kcinukto, filamento subukto, incurvo. Anthera cucullata, antice visa quinquangukris, apicem versus angustete, truncate et parte superiore adpresse puberuk, basi 2-loba, c. 0'2 cm. late. Pollinia 4, in corpuscuk 2 oblique obovate supra convexa subtus concava unite, oblique oblonga, 2 interiora quam cxtci'iora paulo angustiora. Stigma suborbicukre. Pes gynostemii lineans, canalicu- ktus c. 1'7 cm. longus. Ovarium pedicelktum ckvatum, suj^erne 3-alatum et 3-costatum, ak superiore quam cetera? longiore et in dentcm libenHD trkngulum producte, sparsa nigro-punctetum, c. 3 cm. longum. llab. Arfak Mts., Angi lakes, on bank by ? lake. 7000', terrestrial. Dec. 5907. This species, the flowers of which agree in colour with I), pe 120 Svhltr., differs from the latter in the larger flowers, a blunt dorsal sepal, differently shaped petals and lip, lacinulate clinandrium, not 5-winged ovary. The flowers are described as dull yellow, with a red tip to the labellum. DENDROBIUM AGATHOD^MONIS J. J. S. in Bull. Dep. Agr. Ind. Neerl. n. xxxix. (1910) 7 ; etc. Arfak Mts., ridge running up to Angi lakes, terrestrial in open gravel spaces, 8000'. Dec. 5530 and 5596. Disifib. New Guinea (D.N.W., Cyclops Mts., Gjdlerup ; D.S.W., Aga- thodiimonsberg, von Roemer, Hell wig Mts., van Noukuys). Perhaps it would be better to keep the typical I), agatliodcemonis apart as a species and not to unite with it the different forms I successively added. However, these forms are, in the material at hand, almost all represented by single specimens, and therefore it is very difficult to make a decision. The flowers of the Arfak specimens are described as magenta. DENDROBIUM ( CALYPTROCHILUS) PAPUANUM J. J. S., sp. nov. Caules approximati, elongati, superne flexuosi, sicco inferne teretes superne alto sulcati, c. 46 cm. longi, internodiis ad c. 2'6 cm. longis. Folia basi semitbrta, ovato- lanceolata, longe acuminata, acutissima, longe mucronata, acumine denticulata, costa media subtus prominente, sicco rigidiuscula, c. 6-7'5 cm. longa, 1-2-1-9 cm. lata ; vaginae tubulosse, internodia paulum superantes, minute verrucosse, novelise sicco fusee punctatse. Inflorescentiae ad nodos caulium defoliatorum, abbreviate, c. 4-flora3, pedunculo nonnullis vaginulis brevibus tubulosis tecto. Bractece valde approximatse, patentes, ovato-trian guise, acute acuminataa, concava^, sicco c. 0'85 cm. longae. Flores parvi, macerati c. 1*4 cm. longi. Sepalum dorsale subovale, apice brevissime productum, obtusum, valde concavum, 3-nervium, c. 0'65 cm. longum, 0'4 cm. latum. Sepala lateralia lacinia oblique triangula ad pedem gynostemii decurrentia, marginibus anticis longitudine c. O65 cm. connata, men turn magnum conicum ovario parallelum suberectum anguste obtusum c. 0'9 cm. longum formantia, parte libera oblique ovato-triangula, pra^sertim margine postico valde curvata, obtusa, apiculata, concava, c. 4-nervia, c. 0'625 cm. longa, margine antico c. 1'35 cm., basi 1 cm. metientia. Petala oblique subelliptico-oblonga, subulato-apiculata, concava, superne erosula, 3-nervia, c. 0'65 cm. longa, 0'27 cm. lata. Labellum gynostemium sequans, basi gynostemii longitudine c. 0'4 cm. adnatum, concavum, apice cucullatum, ecallosum, c. 5-nervium, nervis exterioribus ramosis, cucullo plicato margine apicali truncate dense et breviter subulato-lacinulato, explanatum cuneatum, late obverse rhombeum, fere I'l cm. longum, 0'84 cm. latum, ab apice cuculli usque ad marginem apicalem c. 0'37 cm. metiens. Gynostemium recurvuluin, a dorso compressum, apicem versvis dilatatura, in utraque stigmatis parte obtusangule dilatatum, c. 0'2G cm. longum, filarnento curvulo, lineari, obtuso, dorso convexo, auriculis magnis, quadrangulis, obtusissimis, dimidio superiore apice leviter productis et denticulatis, dimidio inferiore rotundatis. Antliera cucullata, arnbitu transverse subquadi-angula, leviter 6-angulata, basi biloba, apice producta, truncata et papillosa, c. 0'22-j cm. lata. Pollinia 4, oblonga, in corpuscula 2 oblique subovalia supra convexa subtus concava conglutinata. PCS gynostemii linearis, canaliculatus, 121 c. 0'9 cm. longus. Ovarium pedicellatum clavatum, minutissime furfuraceo- puncticulatum, c. l - 25 cm. longum. Hal. Arfak Mts., Angi lakes, epiphytic in forest patch by ? lake, 7000'. Dec. 5712. Amongst the small-flowered purple or pink coloured species of the section Cali/ptrochilus this one differs from D. roseum Schltr. in its lonf acuminate leaves, smaller flowers, and much broader lip without callus. The flowers are pink. DKNDROBIUM ( CALYPTROCHILUS) INFRACTUM J. J. S. in Fedde Rep. xii. (1913) 118 ; in Nova Guinea, xii. 340, t. cxxiii. 225. Arfak Mts., Koebre' ridge between <$ and ? lake, terrestrial on burnt open summit plateau, 9000'. Fl. Dec. 5655. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). DENDROBIUM ( CALYPTROCHILUS) RIPARIUM J. J. S. in Fedde Rep. xii. (1913) 117 ; in Nova Guinea, xii. 343, t. cxxiv. 227. Arfak Mts., Angi lakes, forest patch by ? lake, epiphytic in forest, 7000'. Fl. Dec. 5896. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). DENDROBIUM ( CALYPTROCHILUS) FRUTICICOLA J. J. S. in Fedde Rep. xii. (1913) 116 ; in Nova Guinea, xii. t. cxxv. 229. Arfak Mts., Angi lakes, terrestrial in open marsh by ? lake, growing ia patches, abundant, 7000'. Fl. Dec. 5510. Forest patch by ? lake, 7000'. Fl. (orange). Dec. 5553. Koebre ridge between and ? lake, terres- trial on open burnt summit plateau, 9000'. Dec. 5608. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). The flowers show the same slight differences as in the original specimens. DENDROBIUM ( CALYPTROCHILUS) PARVIFOLIUM J. J. S., sp. nov. Planta parva (?). Caulis ramosus, valde flexuosus, sicco longitudinaliter alte suleatus, pars adest 13'5 cm. longus, internodiis c. O'G-O'S cm. longis. Folia parva, patentissiina, oblongo-ovata, apicem versus angustata, breviter acuta, apiculata, prsesertim superne erosula, costa media sicco supra sulcata subtus cum nervis sequen- tibus prominente, rigida, sicco ad c. 1*1 cm. longa, fere 0.4 cm. lata; vagina? tubulosse, verruculosae, internodia aequantes. Ltflorescentice (1 adest) in flexubus caulium foliatorum, brevissimse, 1-florae, pedunculo abbreviate, nonnnllis vaginulis tubulosis verruculosis obtecto. Bractea cucullata, apiculata, c. 0-85 cm. longa. Flos c. 1-8 cm. longus. Sepalum dorsale ovatum. apiceru versus angustatum, anguste obtusum, breviter obtuse apiculatum, basi latum, concavum, 5-nervium, c. U'7o cm. longum, 0'4 cm. latum. Sepala lateralia lacinia falcatulo-trian^nla a(tlum dorsale horizontal, oblongum, obtusum, breviter obtuse conico-apiculatum, concavum, superne vix erosulum, 3-nervium, c. 0'37 cm. longum, 0'15 cm. latum. Sepala lateralia porrecta, cum pede gynostemii mentum subrectanguluin cum ovario angulum obtusum faciens obtusum postice c. 0'13 cm. longum fonnantia, margiue proximo ultra medium conglutinata, oblique triangula, falcata, margine extoriorc rotundata, obtusa, breviter obtuse conico-apiculata, concava, 3-nn-via, tola c. ();}.") cm. longa, 0'17 cm. lata. Pffnln gyoostemio :i(l])rcssa, parallela, oblique oblonga, suj>erne denticuluta, longe suhulalo-aruininata, l-ncrvia, c. 0'3 cm. longa, O'l cm. latu, L 124 acumine O'l cm. longo. Labellum membrana tenui 3-nervia pecli gynostemii insertum, mobile, erectum, sigmoideum, e basi contracta dentibus brevibus reversis falcato-triangulis et intus callo erecto quadrate ornata dilatatum, fascia mediana longittidinali convexo-incrassatum et subtus convexum, lateribus supra convexum subtus concavum, 3-nervium, \ parte superiore recurva et utrinque plica parva sed distincta supra convexa a parte inferiore separata, itaque subtrilobum, medio utrinque minute ciliolatum, inexplanatum c. 0'15 cm. longum, explanatum spathulato- obovatum, c. 0'17 cm. longum, O075 cm. latum, lobo intermedio rotundato, semi- orbiculari, convexo, subtus concavo, carnosulo, eciliato, c. O05 cm. longo, 0'075 cm. lato. Gynostemium cum ovario angulum obtusum faciens, curvatum, totum c. 0-12 cm. longum, clinandrio concavo, filamento elongate, subulate, auriculis filamentum bene superantibus, subulatis, falcato-incurvis, margine inferiore lobulo subobsoleto obtuso munitis. Anthera cucullata, semiglobosa, apice producta incurva truncata, connectivo gibboso-incrassato papilloso, c. 0"05 cm. alta. Pollinia in cor- pusculum semiglobosum unita. Rostellum triangulum, obtusum. Stigma magnum, fere tetam faciem inferiore.m gynostemii occupans, profunde excavatum, obovato- triangulum. Pes gynostemii cum ovario angulum obtusum faciens, subrectus, basi dorso convexus, apice vix recurvulus, oblongus, obtusus, canaliculatus, crassiusculus. Ovarium 6-sulcatum, c. 0'075 cm. longum. Arfak Mts., Angi lakes, forest patch by ? lake, epiphyte in forest, 7000'. Fl. Dec. 5662. Differs from B. myrtillus Schltr., probably its nearest ally, by the mentum, the broader slightly three-lobed lip biauricled at the base and only ciliolate in the middle beneath the semiorbicular mid-lobe, and the papillate connective. The flowers are described as yellow ; in the dried plant the petals are distinctly red. BULBOPHYLLUM ( MONOSEPALUM) MURICATUM J. J. S. in Bull. Dep. Agr. Ind. Need. n. xlv. (1911) 9; in Nova Guinea, viii. (1911) t. C, B ; xii. 374. Monosepalum muricatum Schltr. Orch. D. Neu-Guinea, (1912) 682. Arfak Mts., ridge running up to Angi lakes, terrestrial in mossy forest, 8000'. Fl. Dec. 5997. Distrib. N.E. New Guinea. Flowers yellow with red spots. BULBOPHYLLUM ( NEMATORHIZIS) OVALITEPALUM J. J. S., sp. nov. Eliizoma elongatum, repens, filiforme, longinode, ad nodos plerumque 1-subpseudo- bulbis pauces radices emittens, pars adest 25 cm. longa, internodiis c. O5-1 cm. longis, vaginis tubulosis, quarn internodia multo brevioribus. Pseudotulbi c. 1 '5-3-2 cm. distantes, erecti, oblique ovoidei, sicco valde rugosi, c. 0'5-0'65 cm. longi, 1-folii. Folium erectum, breviter petiolatum, oblongum ad ligulatum, obtusum, apice leviter insequaliter vel subsequaliter breviter obtuse bilobulatum, basi acutum, costa media subtus prominente, c. rS-3'1 cm. longa, sicco 0-45-O55 cm. lata ; petiolus cana- liculatus, c. G'l-0-2 cm. longus. Inftorescentice ad basin pseudobulborum et ad 125 nodos rhizomatis solitariae, pedunculo 1-floro filiformi, c. 1'3 cm. longo, mferne c. 2-vaginulis tubulosis superne leviter arnpliatis obtusis c. 0'125 cm. longis donato. Bractea brevis, cupuliformis, obtusa, c. 0'08 cm. longa. Flos parvus. Sepalum dorsale ovatum, obtusum, concavum, 3-nervium, c. 0'34 cm. longum, 0*225 cm. latum. Sepala kteralia oblique ovalia, obtusissima, breviter obtuse apiculata, 3-nervia, c. 0'35 cm. longa, 0'22 cm. lata. Petala vix obovato-ovalia, obtusissima, 3-nervia. Labellum mobile, curvum, carnosum, ambitu subquadrangulo-ovale, basi truncatum, apice rotunda turn, dimidio inferiore excavato-concavum cum marginibus erectis, antice convexum, glabrum, c. 0'2 cm. longum, 0'12 cm. latum. Gynostemium breve, apice obtuso, auriculis brevibus. Anthera cucullata, explanata reniformis, obtusa, c. 0-075 cm. lata. Pollinia 4, lateraliter compressa, a latere visa obovata, exteriora extus convexa, interiora manifeste tenuiora. Stigma majusculum, trans- versuna. Pes gynostemii cum ovario angulum obtusum faciens, quadrangulus, apicem versus paulo angustatus, truncatus, c. 0'075 cm. longus et fere sequilatus. Ovarium G-sulcatum, parce punctatum, c. 0'25 cm. longum ; pedicellus tenuis, supra basin articulatus, c. 0'45 cm. longus. Hab. Arfak Mts., Angi lakes, forest patch by ? lake, epiphytic in forest, 7000'. Dec. 5683. This seems to be the first known species of the section NematorJuzis from the Arfak Mts. The column was not in a very good state, hence the description of it wants completion. The flowers are green. BlILBOPHYLLUM ( PELTOPUS) OCTARRHENIPETALUM J. J. S. in Fedde Rep. xii. (1913) 400 ; in Nova Guinea, xii. 400, t. cxlriii. 277. Angi lakes, forest patch by ? lake, epiphytic, 7000'. Fl. (white). Dec. 5504. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). BULBOPHYLLUM ( DiALEiPANTHE) FRiSTis J. J. S. in Fedde Rep. xii. (1913) 399 ; in Nova Guinea, xii. 419, t. clviii. 296. Arfak Mts., Angi lakes, slopes by " campong," ? lake, terrestrial in forest, 7500'. Dec. 5638. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). Only an inflorescence. BULBOPHYLLUM ( HYALOSEMA) TRICANALIFERDM J. J. S. in Fetfde Rep. xii. (1913) 398 ; in Nova Guinea, xii. 426, t. clx. 302. Arfak Mts., Angi lakes, epiphytic in forest patch by ? lake, 7000'. Dec. 5548. hi ft nb. New Guinea (D.N.W., Arfak Mts., Angi lakes, Gjellerup). Flowers yellow with brown spots. L2 126 PHREATIA ( BULBOPHREATTA) SPATHULATA J. J. S., sp. nov. Pusilla. RMzoma elongatum, repens, ramosum, radicans, vaginis tubulosis tectum. Pseuclobulbi parvi, remoti, omnino vaginis obtecti, 2-folii. Folia erecta, insequimagna, lanceolato-spathulata, obtusa, retusa, apiculata, sicco ad c. 2'3 cm. longa, 0'35 cm. lata. Inflorescentia folia bene superans, laxe c. 13-flora, pedunculo tenui, c. 4'5 cm. longo, vaginulis c. 6-tubulosis apiculatis inferne magis approximatis ad c. 0-35 cm. longis donate, rhachide tenui, c. 3 cm. longa. Bractece ovato-triangula?, acuminatse, .acutissima?, ad c. 0'225 cm. longse. Flores parvi. Sepalum dorsale ovatum, obtusum, concavum, 1-nervium, c. 0'2 cm. longum, O14 cm. latum. Sepala lateralia divergentia, late oblique ovata, acuta, oblique concava, 1-nervia, c. 0-2 cm. tota 0'25 cm. longa, 0'2 cm. lata. Petala oblique ovata, subacuta, concava, 1-nervia, c. 0-16 cm. longa, Ol cm. kta. Labellum leviter sigmoideum, unguiculatum, con- cavum, explanaturn c. 0'14 cm. longum, ungue quadrangulo, basi utrinque in lobulum parvum rotundatum dilatato, probabiliter 2-glanduloso, c. 0'07 cm. longo, basi O06 cm. lato, abrupte in laniinam dilatato, lamina semilunato-reniformi, apice latissime rotundata, medio leviter retusa cum lobulo distincto obtuso in sinu, lobis lateralibus obtusa, incrassationibus 2 convexis intus ad basin, intus papillosa, c. 0-06 cm. longa, O175 cm. lata. Gynostemium breve, cum ovario angulum obtusum faciens, dorso convexum, papillosum, absque anthera c. O075 cm. longum, clinandrio alte excavato. Bostellum latum. Stigma transversum, margine elevatum. Pes gynostemii reversus, c. Ol cm. longus, basi ovario adpressus et intus convexus, parte superiore incurva, quam partem inferiorem angustiore, truncata. Ovaritim cum pedicello trigono torto clavatum, c. 0'35 cm. longum. Hob. Arfak Mts., ridge running up to Angi lakes, epiphytic in forest, 8000'. Dec. 6004. A small plant remarkable for the two convex thickenings at the base of the blade of the lip. Flowers white. PHREATIA ( RHIZOPHYLLUM) DENSISSIMA J. J. S. in Fedde Rep. xii. (1913) 26 ; in Nova Guinea, xii. 438, t. clviii. 314. Arfak Mts., Angi lakes, epiphytic in forest patch by ? lake, 7000'. Fl. (green). Dec. 5550. Distrib. New Guinea (D.N.W., Arfak Mts., Gjellerup). OCTARRHENA CYLINDRICA J. J. S., Sp. DOV. Caules approximati, compressi, simplices, basi valde radicantes, c. 5-5 cm. longi, cum vaginis c. 0'375-0'4 cm. lati, internodiis c. O5-O7 cm. longis. Folia equitantia, erecto-patentia, articulata, lateraliter compressa, linearia, interdum falcatula, acuta, basi valde obliqua, dorso c. 2-3'5 cm. longa, medio 0'275-0'35 cm. lata; vaginse sese amplectentes, tubulosse, lateraliter cornpressse, antice alte fissa3, apice valde obliqua?, internodia superantes. Inflorescentia axillares, erectae, dense multiflora?, cylindricse, pedunculo tenui, c. 1-75-2 cm. longo, inferne nonnullis vaginulis tubulosis longe acuminates, superne numerosis vaginulis bracteiformibus donate, rachide angukta, c. 2'4-2'S cm. longa. Bractece e basi lata longe lineari-acuminatse, concave, iiTe- gulariter marginatae, c. 0'14 cm. longae. Flores vagi, minimi c. 0'17 cm. lono-i. 127 Sepalum dorsale erectum, triangulum, obtusiusculura, convexum, c. 0'06 cm. longura et latuin. Sepala lateralia oblique ovato-triangula, apice interdum plus ininusve contracta, anguste obtusa, c. 0-075 cm. longa, 0'07-0'075 cm. lata. Petala diver- gentia, anguste oblique triangula, subfalcatula, acuta, convexa, 1-nervia, c. 0'07 cm. longa. Labellum simplex, decurvuin cum ovario angulum obtusum faciens, medio fere obtusangule incurvurn, subsigmoideum, manifeste concavum, 1-nervium, basi medio tantum affixum, callis 2 longitudinalibus parallelis approximatis oblongis in f partibus inferioribus labelli, explanatum ovato-oblongum, anguste obtusum, c. Ol cm. longum, fere 0'05 cm. latum. Gynostemium cum anthera 0. O'OG cm. longum, clinandrio cum dorso gynostemii angulum rectum faciente, concavo. Anthera cucullata, ovato-triangula, basi leviter emarginata, apice anguste truncata, c. 0'04 cm. longa. Pollinia 8, clavato-pyriformia. Bostellum productum. Stigma trans- versum. Ovarium obconicum, cum pedicello angulum obtusum faciens, c. O'OG cm. longum ; pedicellus crassus, tortus, c. 0'06 cm. longus. Hob. Arfak Mts., ridge running up to Angi lakes, epiphytic in forest, 8000'. Dec. 5993. From the other species this one is readily distinguished by its moderately dense and broad leaves and dense spikes. The flowers are said to be yellow. Var. MAJOR J. J. S., var. nov. Gaules elongati, inferne defoliati cum radicibus adpressis, ad c. 21 cm. longi. Folia 2-4-3 cm. longa, 0-37-0-475 cm. lata. Pedunculus c. 2*25-25 cm. longus ; rhachis 4-5 cm. longa. Flores majores. Sepalum dorsale c. O'l cm. longum, 0-075 cm. latum, lateralia 0'12 cm. longa, 0'8 cm. lata. Hob. Arfak Mts., Angi lakes, mossy slopes by ? lake, 7000-8000'. Fl. Dec. 5536. This differs from the type-specimen only by its longer stems and larger, differently coloured flowers. The flowers are white. In the dried specimens the flowers of the type as well as those of the variety are greenish with a large white anther. DICOTYLEDONE.E. P I P B R A C E M. (0. DB CANDOLLE.) PIPER ARFAKIANUM C. DC., sp. nov. Kamulis glabris, junioribus in nodis hirtellis; foliis modice petiolatis glabris, lirnbo ovato-acmninato basi ima sequilatera acuto apice longe et acute acuminate, 5-nervio, petiolo basi ima vaginante ; pedunculo glabro quam petiolus breviore, spica quam limbus pluries breviore cylindrica et apice obtusa, rhachi pilosa, bractese glabrae, pelta transverse elliptica centre late et brevissime pedicelkta, ovario inferne rhachi at baud profunde iinmerso superue libero ovato et glabro, stigmatibus rotundato-ovatis. 128 Hob. Arfak Mts., S.W. ridge running up to Angi lakes, in forest, 8000'. Fl. Dec. 5525. Dioicum, epiphytum. Ramuli spiciferi 1 mm. crassi, collenchyma libriforme in fascicules discretes dispositum, fasciculi intramedullare, 1-seriati, canalis lysigenus unicus centralis. Limbi in sicco firmi fuscescentes et pellucido-punctulati, superi usque ad 5 cm. longi et 2'5 cm. lati. Petioli usque ad 5 mm., pedunculi usque ad 3 mm. longi. Spica matura 7 mm. longa, 4 mm. crassa, in sicco nigra. PlPEK PILOSULINODUM C. DC., sp. nOV. Ramulis tantum in nodis pilosulis, primum Isevibus postea lineatim lenticellatis ; foliis parvis modice petiolatis, limbo ovato-acuminato basi obtuso apice obtusiuscule et sat longe acuminate supra glabro subtus baud dense piloso, 5-nervio, petiolo piloso basi ima vaginante ; pedunculo glabro quam petiolus breviore, spica florente quam limbus paullo breviore, rhachi hirsuta, bractese glabrse, pelta rotunda, staminibus 2, antheris tetragonis 4-valvatis. Hab. Arfak Mts.. slopes of Koebre ridge between Angi lakes, 8000'. FL Dec. 5624. Dioicum, epiphytum. Ramuli spiciferi 1 mm. crassi, collenchyma libriforme in fascicules discretes dispositum, fasciculi intramedullares 1-seriati, canalis lysi- genus unicus centralis, in ramulis 2 mm. crassis lineatum lenticellatis cellulae sclerosse circum collenchyma creberrimse. Limbi in sicco subcoriacei fuscescentes et creberrime pellucido-punctulati usque ad 4'2 cm. longi et 2 cm. lati. Petioli circiter 5 mm., pedunculi 3 mm. longi. Spicse florentes 3 cm. longse, 1 mm. crassse, bracteae pelta 0'75 mm. diam. F AGACE^I. QUERCUS LAUTERBACHII Seemen in Engl. Bot. Jabrb. xxiii. Beitr. Ivii. 54 ; Si:hum. & Laut. 264, t. iv. figs. A-E. Arfak Mts., S.W. ridge, 6000'. Fr. Dec. 6124. Distrib. New Guinea (D.S.W., Arfak Mts., 6000', A. E. Pratt ; (Herb. Brit.)N.E.). A tall tree with slender stem. Pratt's collection consists of a single acorn, very striking in size, 6'5 cm. long, with cupule 6 cm. by 5 cm. and nut 4'5 cm. long and about the same in breadth, enclosed in the cupule for I'D cm. from the base. I came across the same group of trees on the exact spot kindly described to me by Mr. Pratt ; the ground was strewn with acorns of all sizes, of which the one originally collected would represent the largest. The leaves are larger than in Seemen's measurements, with the veins pilose on the under surface, but they were collected from young plants. URTICACEJE. PIPTURUS PAPUANUS Gibbs, sp. nov. Arbuscula vel frutex : ramulis internodiis brevibus, cortice rugoso-lenticelloso ; innovationes tomentosae. Folia parva, petiolata, lanceolata, sensim angustata, acuminata, basi obtusa, integerrima, leviter revoluta, rigide membranacea, supra 129 hirsuta, pilis albidis adpressis dense obtecta, demum asperrima, subtus subtiliter brunneo-velutina, trinervia, costa media prominente nervis 2 lateralibus, arcuato- anastomosantibus, reticulo inter venas conspicuo. Glomeruli feminei axillares, sessiles, pisifonnes, densiflori. Perigonium tomentosum, fructiferurn albo-carnosum. Hab. Arfak Mts., Angi lakes, edge of forest by ? lake, 7000'. Fl. ? , Fr. Dec. 5955. Leaves +6 cm. by 1*5 cm., green, but drying a dusty brown, densely covered on the upper surface with punctiform cystoliths. Petioles 1*3 cm., slender, densely puberulous. Stipules 4 mm., lanceolate, acute, tomentose on the outside. Glomerules 4 mm. across, in fruit, forming a white swollen receptacle. Flower ? 3 mm. long ; style filiform, hirsute, 2'5 mm. Achene + 1 mm. long, covered with short stiff white hairs. This plant is distinct in the small narrow leaves with scabrid upper surface and the very well-marked network of minor veins, forming minute interstices, covered with dense brown tomentum on the under surface. In the shape of the leaves it is nearest to P.mindartensis Elm., but differs in the small size, more rigid texture and entire margins, the prominent venation with fewer lateral veins, and the brown coloration when dry. GIBBSIA Rendle, gen. nov. (A. B. RENDLE.) Flores inonoici, unisexuales : HCasculi alabastro depresso-globosi, mucronati ; perianthio fere ad basin 5-partito, segmentis valvatis, ovatis. Stamina 5, antberis subrotundis, dorsifixis, in alabastro inflexis. Ovarii rudimentum lanuginosum. Feminei perianthio brevi, late cupuliforme, persistente et basin fructus cupula carnosula adnata tegente. Ovarium ovoideum valde obliquum ; stigma apicale, sessile, discoideum, papillosum, margine setulis fimbriato; ovulum a basi erectum, micropyle elongata, superne ampliata et fimbriata. Fructus subdrupaceus, parvus, valde obliquus, exocarpium tenue, carnosulum, et super cupulam demum separabile, endocarpium crustaceuin. Semen conforme, testa membranacea ; albumen carnosum ; cotyledones parvae, ellipticae ; radicula superior. Frutices, foliis alternis, petiolatis, crenato-serratis, 3-nerviis, subtus canescentibus. Stipiilae membranaceae, in unam intrapetiolarem alte bifidam connatae, deciduie. Inflorescentice axillares, solitariae vel binae, foliis multo breviores, dichotomse. Flores in cymulis parvulis androgynis apice ramulorum ultiinorum breviter pedi- cellati. Bracteee minutae, ovatse, scariosae. Recalls the genus Delregeasia in its leaf-characters, but in characters of flower and fruit is nearest the Malayan and Pacific Island genus Leucosyke, which, however, has a penicillate stigma and the fruits crowded on a fleshy receptacle ; there is no trace of a fleshy receptacle in Gibbsia. The fruit is very characteristic ; the upper portion of the thinly succulent exocarp separates like a cap, and the endocarp enclosing the seed is then readily separable from the cup formed by the union of the somewhat fleshy perianth with the lower part of the exocarp. 130 GIBBSIA INSIGHTS Rendle, sp. nov. Fruteoc ramulis foliiferis colore cinerascente et pilis brevibus appressis hispidis. Folia breviter petiolata, lanceolata, acuminata, margine crenato-dentato recurvato, in facie superiore eleganter reticulato-impressa, hispidula, subtus inter nervos rubros conspicuos dense albo-tomentosa ; petiolus ut in ramulo hispidus. Stipules infra medium bifidse, triangulari-acutse, in margine superiore ciliolatse, uninerviaB. In- florescentice geminatse, ssepius bis vel ter dicbotoma?, pedunculo ramisque tenuibus, foliis 4-2-plo breviores. Cymulce ssepe floribus evolutis 7. Antherae connective umbonato ; filamenta perianthium aequantia. Exocarpium laete-brunneum. Leafy branchlets 2-2'5 mm. thick. Leaves 5-8 cm. long, 1'2-1*7 cm. wide ; petiole 3-7 mm. long. The leaves have a striking appearance ; the three main nerves are impressed on the upper face, which is beautifully embossed with small chequer-like areas representing the transverse and small connecting veins ; the red main nerves on the lower face stand out conspicuously on the intensely white tomentum covering the rest of the surface. Stipules 3'5 mm. long, united in the lower third, each with a strong median nerve. Inflorescences generally in pairs in the leaf-axils, to 2'5 cm. long ; the small dense cymes are crowded at the ends of the short ultimate branches, the number of flowers in each varies a typical one is represented in fig. C, of which a diagrammatic analysis is given in fig. D. Bracts about *5 mm. long. Flowers jointed on very short pedicels, which are less than 1 mm. long. Bud of male flower 1'5 mm. in diameter, perianth- segments 1'6-1'7 mm. long. Ovary slightly compressed, ovoid, with a narrow rounded keel running along the back and bearing a row of short forwardly-pointing setse, at the base of each of which is a black dot ; similar but slightly larger setse surround the stigmatic disc, covering it when young but ultimately becoming reflexed (figs. I, J). Fruit about 1 mm. long ; crowned with the withered stigmatic disc ; exocarp bright chestnut-brown. Hob. Arfak Mts., Angi lakes, edge of forest by ? lake, 7000'. Fl., Fr. Dec. 5961. A second species of this genus was collected by Mr. Kloss on Dr. Wollaston's Dutch New Guinea Expedition. The material is scanty, but there is no doubt that it represents a plant congeneric with the Arfak specimen and a distinct species. A description is appended : GIBBSIA CARSTENSZBNSIS Rendle, sp. nov. Frutex ramulis foliiferis veluti in G. insignis hispidis. Folia longius petiolata, elliptica, acuminata, margine crenato-dentato, recurvato, in facie superiore reti- culato-impressa, hispidula, subtus nervis exceptis albido-tomentosa ; petiolo tenui, hispido. Stipulce supra medium bifida3, acutse. Inflorescentia et fruclus velut in G. insignis. Leaves 4'5-7 cm. long, 2-2'5 cm. wide ; petiole l-2'5 cm. long; bracts 75 mm. long. 131 Hab. Dutch New Guinea :i Mt. Carstensz, 5500-6700'. Collected by 0. B. Kloss. Fr. Jan. 1913. Distinguished from G. insignis by its relatively much broader elliptical leaves and larger petioles ; the upper face is less conspicuously chequered. The bracts are also somewhat larger. FIG. 8. Gibbsia insignis, Rendle. A. Branch, bearing leaves and inflorescences ; B-L. Description in text. A. B. Rendle, anal. ; P. Highley, del. 132 Description of Fig. 8 (B-L) (p. 131). B. Cymules clustered at the ends of branchlets ; X 4. C. A typical cymule with its subtending bract, B ; the terminal female flower was incompletely developed, a, b, lateral bracteoles, each subtending- a female flower the lateral bracteoles of which (a', b' in fig. D) subtend each a male flower the four male flowers indicated; the stalk of each of these male flowers bears a pair of minute bracteoles, each of which subtends an undeveloped male flower with a lateral bracteole. The apex of the inflorescence and the bracteoles a and b are pushed forward away from the axis, x 15. D. Diagram of same. E. Male flower, with dehiscent anthers ; X 4. F. Stamen ; X 15. G. A stamen taken from a bud, showing the umbo-like dorsal connective ; X 15. H. Female flower ; x 15. I. Young stigma, covered by the ring of inflexed setae. J. Mature stigma. I and J, x 40. K. Fruit, showing detached upper part of exocarp (a), endocarp containing seed (b), and cupule (c) ; X 8. L. Ovule : X 20. SANTALACE^S. EXOCAEPUS sp. ? Arfak Mts., Koebre ridge, common in forest undergrowth from 8000- 8500'. Veg. 5617. A leafless parasite, with dark green, rigid, flattened, much branched shoot?, about 1 m. high. The cladodes are 5 mm. broad, showing swollen round scars alternately up the nodes, on which the flowers are evidently borne in fascicles. I am indebted to Mr. S. Moore for this determination, but, as he points out, in the absence of flowers or fruit the genus must remain uncertain. HENSLOWIA UMBELLATA Bl. Mus. Bot. Lugd.-Bat. i. (1850) 243 = //. Rein- wardtiana Bl. cf. Koords, Excurr-ionsfl. Java, ii. 168 ; Schum. & Laut. 300; Nova Guinea, viii. (1910) 287 ; Ridl. in Trans. Linn. Soc. ser. 2, Bot. ix. (1916) 146. Arfak Mts., Koebre ridge, open summit, 9000'. FL, ) 100). It differs in the smaller leaves with much shorter tomentum underneath, the few-flowered racemes, and the truncate to 3-lobed petals. SERICOLEA ARFAKENSIS Gibbs, sp. nov. Fruticulus epiphyticus, sparsim ramosus. Kami teretiusculi, innovationes fulvo- sericei. Folia opposita, ovata, acuminata, basi rotundata, minute dentata, incurvata, superne glabra, subtus fulvo-sericea, multinervia, reticulationibus conspicuis ; petiolo brevi, sericeo. Raeemi abbreviati, pedunculati; floribus parvis, oppositis vel sub- umbellatis, pedicellis gracilibus fultis. Bractece knceolatae, acuminatse. Sepala 4, lanceolata, acuta, extus sericea, carinata. Petala 4. cuneato-obovata, apice truncata et brevissime trilobata, basi ciliata. Stamina 1/-2, puberula. Discus 5-lobatus. Ovarium conicum, glabrum, biloculare ; stylus subulatus ; stigma truncatum ; ovula in loculo utroque 2, pendula, anatropa. Hob. Arfak Mts., Angi lakes, ridge above ? lake, 8000'. Fl. Dec. 6009. Epiphyte with hanging branches + *50 m. long, in moss-grown small forest. Leaves 3'5 cm. including hair-like acumen + 5 cm. long by 1'9 cm. broad. Petiole 2 mm. long. Racemes +3 cm. long, including peduncle + 1'4 cm. and pedicels 1*1 cm., both slightly sericeous ; flowers 4 mm., pink. Sepals 2 - 5 mm. long. Petals 4 mm. Stamens 1'5 mm. long ; anthers shorter than the filaments, with apical dehiscence. Ovary and style +3 mm. long. This delicate little plant is, so far, distinct in Sencolea in the more remote pairs of leaves and the 4-merous pink flowers. The leaves were infested with Leptotliyrella sericolece Ramsbottom (p. 04). Pyrsonota Ridl. and Mischopleura Wern\mm = Sericolea. In the generic diagnosis of the latter 10 stamens are given, but in one bud of S. ovalifolia examined 15 stamens were seen. DlLLENIACEJS. HlBBERTIA (SUBSESSILES) NOVO-GUINEENSIS Gibbs, Sp. HOV. Frutex prostratus ; rami volubiles, teretes, junioribus pubescentibus, demum glab- rati. Folia lineari-lanceolata, acuminata, in petiolum basi dilatatum longe attenuata, Integra, chartacea, supeme glabra, inferne villosula. Flores solitarii, terminales, subsessiles, flavi. Sepala longe ovata, acuminata, coriacea, concava, subtus villosa. Petala 5, obovato-obtusa, integra. Stamina calyce breviora, antberis linearibus, filamentis filiformibus. Ovaria 5 ; styli flexuosi ; stigmata simplicia, apice concava. Hab. Arfak Mts., Koebre ridge, spreading on burnt open summit, 9000'. Fl. Dec. 5619. A twining woody plant with large yellow flowers. Largest leaves +8'5 by 1 cm., the midrib much impressed on the upper surface and prominent on the lower. Flowers +4 cm. across. Sep;ils unequal, l'j-1'8 cm. by 9 mm. Petals 1'5 by l'2-3 cm., equalling the sepals. Stamens unequal, red-brown in colour (dried), the longest 6 mm. long; anthers 2 mm. long, with apical 149 pores ; filaments 4 mm. long. Ovaries 4 mm. and styles 6 mm. long. The material is insufficient to determine the number of ovules in each ovary. A species which resembles in habit and flowers //. scandens (Willd.) Gilg, from Queensland and New South Wales, but is at once distinguished by the more linear leaves and smaller flowers, and the longer stamens with shorter anthers and longer filaments. This is the first record of the genus from New Guinea and adds another to the long list of genera so far considered limited to Australia. The genus is also common to New Caledonia, and further investigation will no doubt prove it to be widely spread in New Guinea. GUTTIFER^E. HYPERICUM MUTILUM L. Sp. PI. ed. 1, 787 ; cfr. Maximowicz in Bull. Ac. Sc. St. Petersburg, xi. (1881) 171. Arfak Mts., Angi lakes, 7000', open marsh by ? lake. Fl. (yellow), Fr. Dec. 5963. Distrib. New Guinea (D.S.W., Mt. Carstensz, Kloss; N.E.). Sumatra, Java, N. Borneo, Philippines, Central and S. China; New Zealand and Tasmania. Madagascar. N. and S. America. VIOLACE.E. *ViOLA DISTANS Wall. Cat. n. 4022. Arfak Mts., Angi lakes, open marsh by ? lake, 7000', creeping amongst grass. Fl., Fr. Dec. 5962. Distrib. Temperate Himalayas, Nilghiri Mts. to Ceylon. S. China. Flowers white, striped violet on lower lip. Drawings of the style on herbarium sheets at Kew and the British Museum, including some excellent dissections by the late C. B. Clarke, show complete identity with the above specimens, which only differ in the so far recorded colour of the flowers (violet and blue). BEGONIACE^E. SYMBEGONIA ARFAKENSIS Gibbs, sp. nov. Herba, caule erecto in sicco rufo, pilis paucis longis suberectis pubescentibus onusto. Folia sessilia, oblique lanceolata vel ovata, basi uniauriculata, apice angus- tata acuta, grosse serrato-dentata, supra glabra, subtus albo-punctata, in nervis rufo-pubescentia. Inflorescentia terminalis, a basi ramosa, bracteis multis, latis, apice obtusis, albidis. Flares d 1 sepalis 2 ; staminibus haud multis in columnain gracilem connatis, supra liberis ; antheris globosis. Fructus solitarii vel bini, pedicellati, 3-alati, pedunculo fructifero puberulo, alis apice divaricatis, protractis* acutis, sparsim serratis. Placenta bitida. Hob. Arfak Mts., Angi lakes, in forest by ? lake. Fl. h papery cortex, very glandular on the young wood. Leaves' + 1 cm. by 8 mm., with midrib slightly impressed on the upper surface and indistinct on the lower; petioles 4 mm. long, glandular. Peduncles sparsely villous when young, 2-3 mm. long. Flowers subtended by two villous, narrow- linear bracts 3 mm. long. Calyx-lobes villous when young, 2 mm. lung, narrowing from a broad base. Petals glandular, 4 mm. by 1 mm. Stamens as long as or just exceeding the petals ; anthers globose. Style thick, 152 5 mm. long. Ovary 1'5 mm. long. Fruit gland-pitted, 2'G mm. by 3 mm. Seeds large, 2 mm. long. This plant is nearest to M. rufo-punctata Panch., a New Caledonian species, but differs in tbe prostrate habit, ovate leaves, and pedunculate flowers. MYRTUS (AUSTROMYRTUS) ARFAKENSIS Gibbs, sp. nov. Frutex prostratus, valde ramosus, glaberrimus et glanduligerus. Folia opposita, petiolata, late elliptica, basi subattenuata, apice obtusa vel emarginata, margine revoluta, coriacea, supra in sicco fuscescentia, subtus pallidiora, pagina utraque glandulis rufis prominentibus conspersa. Flores cymosi, pedicellati, bibracteolati, cymis axillaribus, subpedunculatis, 3-floris vel solitariis. Calyx 4-lobatus, lobi triangulares, obtusi. Petala 4, flavida, unguiculata. Stamina oo , biseriata. Stylus filiformis. Ovarium biloculare, ovulis biseriatis. Bacca carnosa. Hob. Arfak Mts., Angi lakes, S.W. ridge, on open steep slopes, 8000'. Fl., Fr. Dec. 5994. In this plant the stems are terete, with the older wood dark brown in colour, the younger reddish and rugulose with larger glands ; the small leaves are closely arranged up the branches, the younger shoots being reddish in colour. The leaves are +7 mm. by 5 mm., but with the incurved margins appear 3 mm. across ; the midrib is thick, somewhat raised on the u-iderside but not conspicuous. Petiole thick and glandular, 1'3 mm. long. The flowers form little subsessile cymes, or are on solitary pedicels + 1-3 mm. long, with small, triangular, translucent, gland-dotted bracts at the base of the calyx. Calyx-tube 2 mm. long, with obtuse lobes 1'5 mm. long and the same in breadth at the base. Petals suborbicular, +2 mm. by 1'5 mm. Stamens 4'5 mm. long, with minute globose anthers. Style 4 mm. long. Berry black, crowned by persistent calyx-lobes, 5 mm. by 5 mm. This plant is nearest to M. prostrata, described above, and in the field was very similar in habit and appearance. It is easily distinguished by the entire glabrous and more glandular leaves, tbe longer stamens, and bilocular ovary. MYRTUS (AUSTROMYRTUS) KOEBRENSIS Gibbs, sp. nov. Ffutex vel arbor parva, ramosissima ; ramis cortice cinerascente obductis, tota planta utrmque glaberrima. .Folia parva, ovata, obtusa; basi in petiolum breveiu attenuata, margine leviter recurvata, glandulosa, coriacea. Flores 4-meri, axillares, solitarii vel bini, pedicellis foliis minoribus. Calyce tubo turbinato, basi bibracteolato. Sepala late obtusa, glandulosa. Petala flava, reflexa, basi unguiculata. Stamina biseriata. Ovarium biloculare. Ovula biseriata. Hob. Arfak Mts., in shrubberies on open summit of Koebre Mt., 9000'. Fl. Dec, 5614. A compact shrub to small tree, with very small leaves +'9 mm. by 5 mm., smooth and shining above, with the midrib indistinct on both surfaces ; petiole 1 mm. long. The flowers arise in the axils of the leaves 153 all along the branches, on pedicels 4 mm. long, with two small hracts at the base. Calyx-lobes triangular, 1 mm. long and 1 mm. broad at the base, obtuse to acute and erect in flower. Petals 2*5 by 1'5 mm., the lamina reflexed in flower. Stamens with filaments 3 mm. long j anthers minute. Style 3*5 mm. long. Ovary 2 mm. long by 1 mm. Pedicels, calyx, and ovary thickly gland-dotted. This plant approaches some forms of M. tenuifolia (Sm.), Mez, but is distinguished by the glabrous glandular habit, much smaller leaves, and the yellow 4-merous flowers. JAMBOSA (CLAVIMYRTUS) ARFAKKNSIS G-ibbs, sp. nov. Arbor parva ; ramuli teretes, saturate brunnei. Folia coriacea, opposita, petiolata, oLovata, spathulata, rigida, verisimiliter impellucida, glabra, leviter recurvata, basi attenuato-cuneata, apice obtusa vel emarginata, supra crebre punctata, subtus ]jarit, fronds 1 m., naked stipes 50 m., young plantlets make roots on parent plant." Distrib. New Guinea (D.N.W., Ramoi, Beccari-, D.S.W., Hellwig-Gebirge, ron Roemer ; N.E. ; S.E.). Malay Islands. Polynesia. Trop. Australia. HYPOLEPIS GRANDIFRONS Gepp, sp. nov. Frons maxima subdeltoidea (pinna basalis c. 72 cm., pinna media c. 92 cm. longa) tripinnata ; pinna? primariae lanceolate acuminate 22-25 cm. latae ; pinna? secundaria} alternse c. 40-jugate breviter stipitatae, intervallis c. 2'5 cm. insert*, lineari-lanceolate acuminate usque ad 12*5 cm. longe 2'5 cm. late ; pinnule alternae 20-22 -jugate breviter stipitate oblongse obtusissime pinnatifide, lobis anticis 5 et posticis 4 instructs, apice flabellatae ; lobi oblongi paucicrenati vel integri, venulis 3 anticis 1 postico percursi ; lobus anticus inferior duplo major, venulis 3-jugatis furcatis percursus. Sori in lobis solitarii, in sinu marginis antici positi, lobulo (indusio) parvo luteo-fusco involute obtecti ; in lobo antico inferiore duo sori adsunt. Sporangia circum penicillum paraphysium articulatarum disposita. Textura membranacea ; color valde viridis. Rhachis primaria (? 1 cm. crassa) strain im-a ; rliachis secundaria et tertiaria atque pinnularum coste supenie sulco pilis articulatis brevibus villoso exaratae ; costa et venule pilis paucis hinc illinc instructae. Hab. Humboldt Bay, ridge behind " campong," 500', terrestrial by stream in high forest. Jan. 6258. "Magnificent single fronds from underground creeping rhizome, 5 m. long, petiole 3 m., lamina 2 in. deltoid." The huge fronds, the stipitate pinnules, and the few sori distinguish this species from the rest. The material consists of two pinnae with fragments of the main rhachis attached. PTERIS (EUPTERIS) BAMBUSOIDES Gepp, sp. nov. Rhachis erecta crassa atro-purpurea subnitens minute pubescens. Pinnae inter- vallis circa 4 cm. insert*, erecto-patentes, alternse (?) brevi-stipitatae, ad basin imam 1-2 -dichotomy, segmentis usque ad 40 cm. longis e basi cuneata linearibus sensim attenuatis, fertilibus integris, sterilibus serrulatis, versus apicem spinuloso-serratis, margine incrassata venuliformi, lamina nitente, costa inferne plerumque rubella et ad latera saepe pubescenti, venulis plurimis (circa 33 in centimetri spatio) conspicuis simplicibus furcatisve. In Just urn angustum brunneolum membranaceum. Ildb. Arfak Mts., inundation area of Momi River, and common down to Wariap. Dec. 5732. "Each shoot up to 3 m. tapering; growing in clumps. Pinnre on young fronds plane, but on rhachis twisting in growth, finally arranged spirally." 196 The material collected consists of a longitudinal half of a fragment of rhachis, about 38 cm. long, with pinnae attached. The pinna-segments closely resemble the pinnse of P. moluccana, but differ, of course, in being dichotomously disposed at base. The remarkable fronds of this fern, rising to a height of 3 m. from the ground and growing in clumps, must form a conspicuous feature in the savannah landscape. It is surprising that so well developed a species should not have been recorded previously, unless, indeed, its distribution is extremely limited. PTERIS TORRICELLIANA Christ. Humboldt Bay, ridge behind " campong/' 400', by stream in high forest, undergrowth. Jan. 6252. " Petiole 1'50 m., frond 1 m. long." Distrib. New Guinea (N.E.). VlTTARIA ELONGATA Sw. Humboldt Bay, ridge behind "campong," 200', epiphytic by stream in high forest. Jan. 6254. " Pendent from rock, fronds 2'80 m. long." Distrib. New Guinea (D.N.W., Soron, Ramoi, Beccari ; D.S.W., coastal lowlands, Versteeg ; N.E. ; S.E.). Trop. Asia. Polynesia. N.E. Australia. ANTROPHYUM RETICULATUM Kaulf. Manokoeari, track to Ambani, 700', epiphytic in high forest. Jan. 6193. Distrib. New Guinea (D.S.W., coastal lowlands, Versteeg ; N.E. ; S.E.). Madagascar to Polynesia and tropical Australia. POLYPODIDM (PLEOPELTIS) NORMALE Don. Humboldt Bay, ridge behind " campong/' 500', common, climbing in high forest. Jan. 6271. Distrib. New Guinea (N.E.). Trop. Asia. China. Madagascar. *LYGODIUM DIGITATUM Presl. Manokoeari, track to Ambani, 500', common in forest and clearings. Jan. 6208. Distrib. Philippine Islands. Malacca. ANGIOPTERIS EVECTA Hoffm. Humboldt Bay, ridge behind " campong," 500', very common by stream in high forest. Jan. 6259. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari; D.S.W., Mt. Car- stensz, Kloss ; S.E.). Tahiti. Trop. Asia. Africa. OPHIOGLOSSUM PEDUNCTJLOSUM Desv. 'Dammar Island, in open " kebun " near seashore. Jan. 6289. Distrib. New Guinea (N.E.). Trop. Asia. Australasia. 197 LYCOPODIALES. LYCOPODIUM PHLEGMARIOIDES Gaud. Manokoeari, Langgen, epiphytic on tree hanging over sea. Dec. 6221. Distrib. New Guinea (N.E.). Malay Islands. Polynesia. LYCOPODIUM CERNUUM L. Island of Roon, young plants, spreading in the open, in shade by road round bay. Jan. 6225. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Mt. Car- stensz, Kloss ; N.E. ; S.E.). Tropics and some subtropics. PsiLOTUM FLACCIDUM Wall. Schouten Islands, Bosnik, Wiak, epiphytic on strand trees. Jan. 6273. Distrib. New Guinea (D.N.W., Arfak Mts., Beccari ; D.S.W., Noord River, Versteeg ; Mt. Carstensz, Kloss ; N.E.), Tropics. SELAGINELLA PLUMOSA Baker. Humboldt Bay, ridge behind " campong," creeping in high forest, 500'. Jan. 6265. Distrib. New Guinea (N.E.). Solomon Islands. Trop. Asia. SPERMATOPHYTA. CYCADALES. CYCADACE^E. CYCAS CIRCINALIS L. Sp. PI. ed. 1, 1188 ; F. Muell. Pap. PI. ii. 71 ; Schum. & Laut. 153 ; Nova Guinea, viii. (1910) 343. Schouten Islands, Wiak, Bosnik, under trees on seashore. Veg. Jan. 6278. Distrib. New Guinea (D.S/W., Noord River, Versteeg ; S.E. : N.E. and adjacent islands). S. Asia to Polynesia. MONOCOTYLEDONS*;. P A N D A N A C E M. (A. B. RENDLE.) FREYCINETIA OBLANCEOLATA Martelli in Webbia, iii. 176. Manokoeari, "korang" forest, 500'. Fl.,