Art. XLVII.—A Revision of the New Zealand Species of Carex.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 16, 1883, Page 414

Art. XLVII.—A Revision of the New Zealand Species of Carex. By T. F. Cheeseman, F.L.S., Curator of the Auckland Museum. [Read before the Auckland Institute, 12th November, 1883.] The genus Carex, the largest among Cyperaceæ, and indeed of the Glumiferæ generally, is well represented in New Zealand, as in most temperate and mountainous countries. In New Zealand, too, the species possess that tendency to variability which has made their elucidation so difficult and perplexing to the student of northern floras, and which, through the diverse views entertained by authors as to the amount of difference required for specific distinction, has led to the needless multiplication of names, and much confusion as to the limits and range of the species. Here, fortunately, the creation of species out of mere varieties has not been so regularly and consistently carried out; although, as New Zealand naturalists are no more likely than their European brethren to agree on the quæstio vexata of what constitutes a species, a very similar result may be anticipated. Notwithstanding this, the study of the genus in New Zealand has been beset with serious difficulties, few of which have yet been removed. In some cases the species have been described from insufficient material, the descriptions being consequently incomplete and inexact; in others the specimens were from one or two localities only, and, although numerous, did not always represent the predominant form of the species. Mistakes in the identifications have thus occurred, apart from the difficulty that is commonly experienced in recognizing, from descriptions alone, species of so variable a genus as Carex. But it is hardly necessary to pursue this part of the subject further—all qualified persons are aware that considerable confusion exists, and that the species are in urgent need of revision. The want of confidence in their own conclusions, openly expressed by those who have specially attended to the genus, may be taken as a fair proof of this statement. Under these circumstances, I have ventured to prepare the following re-arrangement of our species. Of its imperfections and deficiences, no one can be more sensible than myself,—many of them are inseparable from the pursuit of systematic botany in a country where there are no large public herbaria and few scientific libraries; and, as to the rest, I trust that they are not more numerous than what might be expected to attach to the first attempt made in the colony to revise the species of a difficult genus like Carex. Such as it is, I can state that it is based upon the actual study in the field of most of the species; in addition to the conscientious examination of many hundreds of dried specimens from all parts of the islands, and a comparison of these with allied plants from other parts of the world.

For the original discovery of Carex in New Zealand, we are indebted to the labours of Banks and Solander, during Cook's first voyage. Eleven species are included in their collections, now preserved in the British Museum. Most of these were named and fully described in manuscript by Dr. Solander, but were never actually published. Several of the same species were gathered by Forster, in Cook's second voyage; by D'Urville and Lesson, in the two French exploring expeditions that visited New Zealand in 1822 and 1827; and during the several visits of Allan and Richard Cunningham, from 1826 to 1838; but no fresh forms appear to have been discovered. From time to time descriptions of these were published by European botanists, who were not always cognizant of each other's work, so that in some cases the same plant received several names. About 1836 Mr. Colenso commenced his well-known botanical explorations. His collections were forwarded periodically to Kew, and seem to have included specimens of all Banks' and Solander's species, as well as eight additional ones, mostly from the previously unexplored central districts of the North Island. Some of his discoveries were collected, a little later, by Dr. Sinclair, M. Raoul, and others, with one or two fresh ones. In 1853 Sir Joseph Hooker published the first volume of his “Flora Novæ Zealandiæ.” This contains an excellent synopsis, from the pen of the late Dr. Boott, of the 22 species then known to inhabit New Zealand. Twelve or thirteen of these were published for the first time, some four or five under the manuscript names previously applied by Dr. Solander. During the ten years following the publication of the “Flora” little attention was paid to the Carices, so that Sir Joseph Hooker records only three additional species in the “Handbook of the New Zealand Flora,” issued in 1864. His arrangement is mainly that of Dr. Boott, but few changes being made. C. secta was merged with C. virgata; C. solandri was considered to be identical with the Norfolk Island C. neesiana; and C. fascicularis and C. forsteri were united. This is the last attempt to give a collected arrangement of the whole of the New Zealand species. But although no systematic sketch of the genus has appeared since the publication of the “Handbook,” numerous additions have been made to the list of species. Dr. Berggren, who specially attended to the genus during his travels in New Zealand, has described and figured five new ones in the “Minneskrift” of the University of Lund for 1878. His descriptions have been lately reproduced in the “New Zealand Journal of Science,” but it is a matter for regret that the excellent plates are not generally accessible in the colony. The Otago species have been most carefully and assiduously collected by Mr. Petrie, who has described the new ones, eight in number, in various communications to the New Zealand Institute. His C. goyeni is

clearly referable to Boott's C. raoulii, a species which has been much misunderstood. The remaining seven appear to be perfectly distinct. Lastly, in volumes xiv. to xvi. of the Transactions, I have given notices of seven additions. Further examination has led me to doubt the distinctness of one of them, C. cinnamomea, but I think that the rest will be generally admitted. To those who have not previously studied the genus, the following résumé of its general characters will be of some service. Of course it will be understood that the remarks made refer to the New Zealand species only. Size, etc.—In size the species vary from little more than an inch or two in height, as in small varieties of C. acicularis, C. inversa, C. cirrhosa, and C. uncifolia, to four or even five feet, as in large forms of C. paniculata, C. ternaria, and C. trifida. In habit most of the species form compact, close-growing tufts, very similar to those of many other Cyperaceous plants, but usually more grassy. A few have a looser and more open habit of growth. Rhizome.—All the species are furnished with a rhizome or creeping stem, clothed with rudimentary scale-like leaves, and on its lower side giving origin to the roots, while from its upper surface it throws up the culms or leafy branches. In C. pumila it is long and wide-creeping, and in consequence the culms are often far distant from one another. Some tendency to this is also shown by C. inversa, and in a lesser degree by C. colensoi; but in the other species, as a rule, the rhizome is short, though often stout and repeatedly branched. In the varieties of C. paniculata, the rhizomes are often matted together, growing under and over one another, and forming, with the roots and the lower portions of the old culms, huge tussocks resembling the stems of tree-ferns, and sometimes three to four feet high, and one to two feet in diameter. Similar but much smaller tussocks are sometimes produced by C. subdola and other species. Culms.—These ascend vertically upwards from the rhizome, and except in the above-mentioned case of C. pumila are usually placed in close juxtaposition, forming tufts more or less dense. They are unbranched in all the New Zealand species. In section they are commonly trigonous or triquetrous; but in a few species are round, as in C. acicularis, or oval, as in C. viridis. They are usually grooved or striated, and the edges are very frequently scabrid with small harsh projections, but may be quite smooth and even. Leaves.—At the very base of the culms we frequently find a number of small scale-like leaves, very similar to those clothing the rhizome, and generally of a dark colour. These pass abruptly or gradually into the foliage leaves, of which there are two main types. First, that in which the leaf is

plano-convex. This division includes C. acicularis, (which has a nearly terete leaf) C. viridis, C. buchanani, C. pulchella, C. petriei, C. comans, C. cheesemanii, and C. littoralis. Second, that in which the opposite surfaces are about parallel to one another. In this class the leaf may be quite plane, as in C. pyrenaica, C. leporina, C. raoulii, etc.; involute, as in C. colensoi, C. echinata, etc.; or keeled, as in C. lucida, C. dipsacea, etc. The leaves are usually grooved or striate; sometimes with a strong vein on either side of the midrib,—e.g., C. ternaria. C. forsteri, etc.; and the margins and veins are commonly furnished with minute sharp projections, which all point towards the base of the leaf, so that the skin is frequently cut or scratched if it be drawn sharply through the fingers. In breadth the leaves vary from nearly or quite filiform to one-half or three-quarters of an inch. Inflorescence.—The flowers, which are invariably unisexual, are arranged in spikelets of various forms and size, according to the species. The spikelets may be either solitary, as in C. pyrenaica and C. acicularis; or few in number, varying from 3 to 6, 8, or 10, as in the great majority of the species; or very numerous and arranged in a compound panicle, as in the varieties of C. paniculata, and to a lesser degree, in C. kaloides. The individual spikelets may be either few or many flowered; erect, spreading, or pendulous; sessile or stalked; close together or more or less distant; and such characters are often useful in the discrimination of the species. As regards the character of the flowers composing them, the spikelets may be androgynous, with the male and female flowers intermixed in the same spikelet, or unisexual, in which case they are placed in different spikelets. This is an important distinction, and has been made use of to divide the genus into two main groups. It is not, however, altogether absolute, as in the androgynous series spikelets composed wholly of female or male flowers can often be found, and one of the New Zealand species, C. viridis, is even occasionally diœcious. Similarly, in the unisexual group the female spikelets generally have a few male flowers at the base or apex. The male spikelets are far more rarely mixed, but C. raoulii invariably has them partly female, and occasionally this is the case in C. dipsacea and C. forsteri. Flowers.—Each spikelet is composed of an indefinite number of floral bracts or glumes, in the axils of which the flowers are situated. The glumes vary in shape, according to the species, from rounded to lanceolate or linear: they usually have a stout midrib, frequently produced into an awn of variable length; and thin and membranous, often lacerate margins. Those of the male spikelets generally differ slightly in shape from the female; and those from the lower part of a spikelet seldom exactly match those from the upper portion.

The male flowers are of very simple structure, consisting of three stamens, without any signs of hypogynous bristles, scales, or other rudiments of a perianth. They are very uniform in all the species, and call for few remarks here. The female flowers are composed of a single-celled, one-ovuled ovary, crowned by a short style terminated by two or three long and slender stigmatic branches. The ovary and lower portion of the style are enclosed in a peculiar bottle or flask-shaped organ called the perigynium, swollen in its lower part, but gradually contracted towards the top into a narrow oblique or bidentate mouth, closely surrounding the style. It varies exceedingly in shape and other characteristics, but is tolerably constant in each species, and has thus been largely used for specific circumscription. Inside the perigynium, and between it and the ovary, there often exists a minute bristle-like body called the seta, or rachilla, and usually considered to be a barren pedicel. In the allied genus Uncinia this bristle is invariably present, is much longer, hooked at the tip, and produced beyond the mouth of the perigynium. The perigynium is a structure unknown except in Carex and its immediate allies, and much discussion has arisen as to its nature and probable origin. Three principal views have been advocated. First, that it represents a perianth. Second, that it is to be regarded as of staminal origin. Third, that it is composed of one or perhaps two modified bracts. The first hypothesis was for long widely accepted, but recent researches have brought to light an almost overwhelming amount of evidence in favour of the last—or that the perigynium is morphologically to be regarded as an altered bract, and the rachilla a rudimentary axis. For a full exposition of the evidence in favour of this, reference should be made to two papers by Dr. McNab and Professor Thistleton Dyer, printed in the Journal of the Linnean Society (Botany, vol. 14, pp. 152–154). It follows from the above view of the structure of the female flowers that each perigynium must be looked upon as constituting a one-flowered spikelet. In the inflorescence of C. lucida, for instance, the terminal spike of male flowers would be regarded as a single many-flowered spikelet; the lower spikes of female flowers as each consisting of numerous one-flowered spikelets. Similarly, an androgynous spikelet, like that of C. pyrenaica, must be regarded as being composed of a single several-flowered male spikelet, and numerous one-flowered female spikelets. Fruit.—This is a minute trigonous or lenticular achenium or nut, enclosed in the persistent and hardened perigynium. Its characters are very uniform throughout the species, and are seldom of value for systematic purposes.

Geographical Distribution.—Under the remarks on each species I have briefly indicated the geographical range, so far as it is known to me; but it will be well to give a summary here. Of the 40 species admitted, 25, or almost exactly three-fifths, are peculiar to the country. Of the 15 that are found elsewhere, 11 are recorded from Australia and Tasmania. Nine of the species are found in Europe, all of which, curiously enough, also inhabit North and West Asia, and North America. Seven are natives of Southern or Eastern Asia; six of temperate North and South Africa; while four or five have been recorded from extra-tropical South America. The annexed table will show the distribution of each species at a glance. It is remarkable that most of the species with androgynous spikelets are found out of New Zealand, while in the section with unisexual spikelets nearly all are endemic. The distribution of the species within the colony is given as fully as possible, although there is much room for improvement. On the West Coast of the South Island they have hardly been collected at all, and almost nothing is known of those inhabiting the elevated central districts of the North Island. I have, therefore, not attempted to draw up any general conclusions on this point, feeling sure that such would require considerable modification within a short time. Before passing to the strictly systematic portion of this paper, I have to express my obligations to many kind friends and correspondents for assistance in preparing it. My thanks are specially due to Mr. D. Petrie of Dunedin, who has liberally supplied me with full suites of specimens of the Otago species, accompanied by many valuable notes. I am also indebted for specimens and information to Mr. G. M. Thomson, Mr. H. Tryon, Mr. T. Kirk, Mr. J. Buchanan, and others. My most grateful thanks are due to Sir J. D. Hooker for his unwearied kindness in comparing my specimens with those in the Kew herbarium. Sir F. Müeller has liberally forwarded a complete series of the Australian species, which has been of much service to me. For specimens of the European forms I am indebted to Mr. A. W. Bennett of Croydon, Dr. J. Müller of Geneva, and Dr. Levier of Florence, while I have to thank Mr. Canby and Mr. Martindale for many of the American species. In conclusion, it is my wish that this paper may be looked upon simply as an attempt to revise and re-arrange the New Zealand species of Carex, based upon the additional information that has been obtained since the publication of Sir J. D. Hooker's “Handbook.” The preparation of a more detailed monograph of the species would have been premature, until the country has been more completely explored, and the species more generally and exhaustively studied.

Geographical Distribution of the New Zealand Species of Carex. — Endemic. Australia and Tasmania Europe. North and West Asia. South and East Asia. North and South Africa, Canary Islands, etc. North America. South America. C. pyrenaica X X X X C. acicularis X C. inversa X C. colensoi X C. parkeri X C. leporina X X X X X? C. echinata X X X X X C. muricata X X X X X C. teretiuscula X X X X X C. paniculata X X X X X C. viridis X C. kaloides X C. vulgaris X X X X X X C. subdola X C. ternaria X C. buchanani X C. lucida X C. dipsacea X C. raoulii X C. devia X C. testacea X C. wakatipu X C. cirrhosa X C. cryptocarpa X C. uncifolia X C. pulchella X C. petriei X C. comans X C. cheesemanii X C. littoralis X C. dissita X C. neesiana X C. longiculmis X C. trifida X C. breviculmis X X C. pumila X X X C. flava X X X X X X C. vaccilans X C. forsteri X C. pseudo-cyperus X X X X X X X Synopsis of the Species. Section I.—Spikelet solitary, simple, terminal. (Sp. 1–2.) Leaves grassy, flat, flexuous; spikelet oblong; bract short or 0 1. C. pyrenaica Leaves terete, strict, grooved; spikelet ovoid; bract usually long 2. C. acicularis. Section II.—Spikelets several or many, androgynous or rarely diœcious, sessile, forming a compact or more or less interrupted sometimes paniculate-compound or decompound inflorescence, Stigmas 2. (Sp. 3–12.)

A. Spikelets all androgynous; female flowers above, male below. Leaves usually flat; spikelets approximate, green or pale perigynia ovate, beaked, serrate 3. C. inversa. Leaves involute, wiry; spikelets approximate, brown; perigynia broadly ovate, not beaked, serrate 4. C. colensoi. Leaves flat, grassy; spikelets approximate, purplish-brown; perigynia ovate, compressed, entire 5. C. parkeri. Leaves flat; spikelets approximate, pale brown; perigynia narrow-ovoid, margined, beak long, serrulate 6. C. leporina. Leaves narrow, keeled; spikelets spreading, close together or distant; perigynia squarrose, ovate-lanceolate, corky at the base, nerved, beak long 7. C. echinata. B. Spikelets all androgynous; male flowers above, female below. Tufted, slender; spikelets in a linear or oblong head ½ to 1½ inch long; perigynia pale, spreading, hardly nerved 8. C. muricata. Hardly tufted, slender; spikelets in an ovoid head ½ to 1 inch long; perigynia dark brown, polished, nerved below 9. C. teretiuscula. Tufted, often very tall and stout; spikelets many, in a narrow or loose panicle 6 to 18 inches long; perigynia ovoid, turgid, nerved or nearly smooth 10. C. paniculata. C. Spikelets some androgynous, others wholly pistillate or, wholly staminate, or inflorescence altogether diœcious. Perigynia narrow lanceolate, very long beaked. Diœcious, or with few females in the male inflorescence, and few males in the female; leaves semi-terete, strict; spikelets in a linear head 1 inch long 11. C. viridis. Most of the spikelets wholly pistillate, the staminate flowers few, and irregularly placed; leaves keeled; spikelets in an open panicle 4 to 9 inches long 12. C. kaloides. Section III.—Spikelets separate, usually stalked, unisexual; the male spikelets constantly uppermost, rarely mixed with female flowers; the lower spikelets all female or with a few male flowers at the base or apex. (Sp. 13–40.) A. Stigmas 2. Female spikelets often wholly unisexual, the male flowers, when present, usually at the top of the spikelet, rarely below. From 6 to 18 inches high; spikelets 3 to 6, short, erect; glumes obtuse 13. C. vulgaris. Tall, grassy, slender, 1 to 3 feet high; spikelets 4 to 8, long; glumes mucronate 14. C. subdola. Tall, stout or slender, 1 to 4 feet high; leaves often ½ inch broad; spikelets numerous, 8 to 25, long stalked and pendulous; glumes awned 15. C. ternaria. B. Stigmas 2. Female spikelets often wholly unisexual; the male flowers, when present, usually at the base of the spikelets, rarely at the top.

* Spikelets usually more or less distant, or the upper ones alone approximate (often all approximate in 19. C. raoulii). Leaves semi-terete, usually red; glumes pale, almost white; perigynia plano-convex, smooth, margins serrate 16. C. buchanani. Leaves narrow, flat, keeled; spikelets slender, loose-flowered; glumes acute, cuspidate or awned; perigynia turgid, smooth, polished, margins entire 17. C. lucida. Leaves narrow, flat, keeled; spikelets stout, close-flowered; terminal one sometimes partly female; glumes rounded, obtuse; perigynia divaricating, unequally biconvex, smooth, with serrate margins 18. C. dipsacea. Leaves broad, flat, not keeled; terminal spikelet almost invariably partly female; glumes bifid, awned; perigynia strongly nerved, with serrate margins 19. C. raoulii. Leaves short, flat, keeled, coriaceous; male spikelet stout; perigynia dark coloured, almost black, biconvex, nerved and wrinkled, margins entire 20. C. devia. ** Spikelets usually approximate (sometimes distant in 21. C. testacea) Culms slender, fiiliform, 1 to 2 feet, often much elongated in fruit; perigynia plano-convex 21. C. testacea. Culms rather stout, 4 to 12 inches high, shorter than the leaves; perigynia unequally biconvex 22. C. wakatipu. Culms very short, stout, 1 to 4 inches; spikelets concealed by the leaves; perigynia plano-convex, slightly nerved. Leaves often spirally twisted at the apex when dry 23. C. cirrhosa. C. Stigmas 3. Female spikelets either wholly unisexual, or with a few male flowers at the base or apex of the spikelet. * Perigynia quite glabrous, hardly spreading when ripe, beak short, stout. † Small, 2 to 6 inches high. Culms extremely short, so that the spikelets are hidden among the leaves. Short, stout, densely tufted; leaves flat, grooved; glumes broad-ovate or rounded; perigynia broadly ovate, turgid, unequally biconvex, serrate above 24. C. cryptocarpa. Small, slender; culms tufted but spreading; leaves narrow, concave; glumes narrow-ovate; perigynia elliptic-oblong, trigonous or fusiform 25. C. uncifolia. †† Slender, 6–18 inches high; leaves narrow, 1/15–1/40 inch, usually plano-convex; male spikelets solitary. Culms slender, flaccid; leaves longer than the culms, filiform, keeled; spikelets all sheathed, lowest one basal; perigynia trigonous 26. C. pulchella. Stouter, leaves with broad sheathing bases and fine often curled and twisted points; spikelets all sheathed and stalked; glumes pale; perigynia dark, smooth and turgid 27. C. petrici. Culms filiform, shorter than the leaves; spikelets short, sessile or the lowest shortly pedunculate, perigynia lanceolate, plano-convex, margins serrate 28. C. comans.

Culms filiform, usually longer than the leaves; spikelets slender, lower on long filiform peduncles; glumes and perigynia as in C. comans, but broader 29. C. cheesemanii. Pale-gréen; culms slender; spikelets short and broad, often close together; perigynia ovoid, turgid, biconvex or nearly trigonous, smooth, not serrate 30. C. littoralis. ††† Tall, stout or slender, from 1 ½ to 3 feet or more. Leaves broad, ⅛ to ½ inch, flat or keeled, grassy. Male spikelets usually more than one (except in C. dissita). Stout or slender, 1 to 2 feet high; spikelets 4 to 8, long or short, ½ to 2 inches long by 1/6 inch broad, male spikelets usually solitary 31. C. dissita. Tall and slender, 2 to 3 feet high; culms often elongating in fruit; spikelets 5 to 10, terminal 2 to 4 male, female slender, usually on long filiform peduncles, lower compound 32. C. neesiana. Stout, 2 to 3 feet high; spikelets 5 to 6, two terminal male, female very stout and broad, over 1/3 inch diameter; glumes and perigynia broadly ovate 33. C. longiculmis. Very large and stout, culms often 4 feet high and leaves ½ inch broad. Spikelets 6 to 12, 2 to 4 inches long, very stout, upper 2 to 4 male; glumes narrow; perigynia obovate-oblong, stipitate 34. C. trifida. ** Perigynia pubescent, not spreading when ripe, beak short. Small, 2 to 9 inches high, culms much shorter than the leaves 35. C. breviculmis. *** Perigynia glabrous, spreading after flowering, large, smooth and turgid. Rhizome long, stout, creeping in sand, etc.; leaves glaucous, longer than the culms 36. C. pumila. *** Perigynia glabrous, strongly costate-nerved, narrowed into a long and slender sharply 2-toothed beak, spreading when ripe (except in C. vaccilans). Pale yellowish-green, tufted, 4 to 9 inches high; leaves flat; spikelets 3 to 6, crowded; perigynia ovoid, suddenly narrowed into the beak, strongly nerved 37. C. flava. Culms 1 to 2 feet high; leaves flat, harsh; spikelets 5 to 9, distant, 1 to 2 ½ inches long, very slender, hardly 1/6 inch in diameter; perigynia fusiform, hardly spreading when ripe 38. C. vaccilans. Culms 1 ½ to 4 feet; leaves flat, harsh; spikelets 5 to 9, stout, distant, terminal one often mixed with female flowers; perigynia small, shortly fusiform, spreading 39. C. forsteri. Culms 2 to 4 feet; leaves flat, soft, green and grassy; spikelets 3 to 5, pedunculate, nodding, close together, or the lower alone distant; perigynia large, very narrow, fusiform or triquetrous 40. C. pseudo-cyperus.

Section I. Spikelet solitary, simple, terminal. 1. C. pyrenaica, Wahl. in Act. Holm., 139; Boott, Ill. Car., iv., 148, t. 475, 476; Hook. fil. Fl. Nov. Zeal., i., 280; Handbook N.Z. Fl., 312. North Island.—Tops of the Ruahine Mountains, Colenso (fide “Handbook.”) South Island.—Summit of Mount Arthur, Nelson, alt. 5,000–6,000 feet; Raglan Range and mountains flanking the Wairau Valley, 5,000–6,000 feet; mountains above Arthur's Pass, Canterbury, alt. 3,500–6,000 feet; Mount Dobson, near Lake Tekapo, 5,000–6,000 feet, T.F.C. Mount Aspiring, Otago, alt. 5,000 feet, D. Petrie! Usually from 3 to 9 inches high, but taller specimens are sometimes seen. The leaves are described by Hooker as “longer than the culms,” but this is only the case in the flowering stage, the culms elongating considerably as the fruit ripens. All my New Zealand specimens have the style nearly constantly 2-branched, in this respect differing from numerous European and American specimens that I have examined, and in which the style is nearly uniformly 3-branched. They thus approach the Australian C. cephalotes, F.v.M., which, judging from examples kindly forwarded by Sir Ferdinand Mueller, can only be separated from our plant by the broader and flatter, hardly stipitate perigynia, and might well be regarded as a variety only. C. pyrenaica has an extensive range out of New Zealand. It is found on the high lands of Northern and Central Europe, Northern Asia, and in America along the line of the Rocky Mountains from Alaska to Utah and Colorado. 2. C. acicularis, Boott, in Hook. fil. Fl. Nov. Zeal., i., 280, t. 63; Ill. Car., iv., 157, t. 508, f. 2; Hook. fil. Handbk. N.Z. Flora, 312; Benth. Fl. Austral., vii., 437. C. archeri, Boott, in Hook. fil. Fl. Tasm., ii., 98, t. 150; Ill. Car., iv., 156, t. 508, f. 3. C. pyrenaica, F. Muell. Fragm., viii., 251, not of Wahl. North Island.—Tops of the Ruahine Mountains, Colenso (Handbook). South Island.—Not uncommon in the mountains of Nelson and Canterbury, alt. 2000–5000 feet. Otago, mountains above Lake Harris, T. Kirk. Also in Tasmania and Victoria. A well-marked species, easily distinguished from the preceding by the strict and nearly terete leaves, shorter spikelets, and erect subulate bract. It varies greatly in size, and in the number of flowers in the spikelets, but its other characters appear to be fairly constant. Section II. Spikelets several or many, androgynous or rarely diœcious, sessile, forming a compact or more or less interrupted sometimes paniculate or decompound inflorescence. Stigmas 2.

3. C. inversa, R. Br. Prodr., 242; Boott, Ill. Car., iv., 151, t. 486–488: Hook. fil., Fl. Nov. Zeal., i., 281; Fl. Tasm., ii., 99; Handbk. N.Z. Flora, 312; Benth. Fl. Austral., vii., 438; F. Muell. Fragm., vii., 252. North Island.—Near Auckland, T. Kirk, T.F.C.; Thames Valley, Upper Waikato and Waipa, T.F.C.; Hawke's Bay and Ruahine Range, Colenso (Handbook); Waimarama, Hawke's Bay, H. Tryon! Wellington, rare, T. Kirk. South Island.—Wangapeka Valley, Buller Valley, and other localities in Nelson Province, T.F.C.; Canterbury, Armstrong (Trans. N.Z. Inst. xii., 344); Otago, Lake Wanaka, Outram, and Strath Taieri, D. Petrie! Deep Stream, G. M. Thomson! Dart Valley, T. Kirk. Altitudinal range from sea-level to over 3,000 feet. Var. radicata.—Smaller, 2–4 inches high. Culms short or almost wanting, so that the spikelets are nearly radical, and much overtopped by the leaves; leaves wiry, involute, filiform; perigynia prominently nerved, narrowed into a long tapering serrate beak. Lakes Tekapo and Pukaki, Canterbury, altitude 2,500 feet, T.F.C. The species, which is also common in many parts of Australia and Tasmania, can be easily distinguished from the following by its grassy habit, pale-coloured spikelets, and beaked perigynia. It varies greatly in height—some of my Nelson specimens are over eighteen inches, and Mr. Kirk informs me that he has gathered equally luxuriant ones in the Wellington District. On the other hand, specimens from the Upper Waikato are much depauperated. The variety to which I have applied the name radicata is a very distinct looking plant, but as at present it is only known from one district I hesitate to describe it as a new species. It differs markedly from the type in the very short culms, involute filiform leaves, and longer beak to the perigynia; but the Australian variety major has perigynia very similar in shape, and the length of the culms is a variable character in most of the species. Possibly some of its characters may be due to the dry and arid climate of the district in which it was collected. 4. C. colensoi, Boott, in Hook. fil. Flora Nov. Zeal., i. 281, t. 63, B.; Handbook N.Z. Flora, 312. North Island.—Probably not uncommon on the dry grassy plains in the interior. Upper Thames and Patetere, T.F.C.; Patangata, Hawke's Bay, W. Colenso (Handbook); Ruataniwha Plains and Taupo, H. Tryon. South Island.—Common throughout in hilly and grassy districts, ascending to 4,500 feet. Sinclair and Haast (Handbook); Travers (Handbook); Buchanan! Armstrong! D. Petrie! G. M. Thomson! T.F.C. A distinct and well-marked species, apparently confined to New Zealand.

5. C. parkeri, Petrie, Trans. N.Z. Inst., xiii., 332. South Island.—Otago, on a hill near Mount Aspiring, alt. 5,000 feet, D. Petrie! G. M. Thomson! I exceedingly regret that I only possess two immature specimens of this curious little plant, and am therefore unable to add anything to Mr. Petrie's description. It is clearly distinct from all the New Zealand species, but appears to be in some measure allied to the northern C. lagopina, Wahl., so far as the state of my specimens admits of a comparison being made. 6. C. leporina, L. Species Plant.; Cheeseman, Trans. N.Z. Inst., xiv., 301; Kirk, l.c., 384. C. ovalis, Good., Trans. Linn. Socy., ii., 148. North Island.—Ohariu Valley, near Wellington, T. Kirk! South Island.—Lower Motueka, Ngatimoti, Rosedale, Graham River, and other localities in the Nelson Provincial District: T.F.C. New Zealand specimens present no points of difference from the English plant, which is also found in Northern and Central Europe, Greenland, Siberia and Western Asia, and in America along the line of the Rocky Mountains. A reported locality in the Falkland Islands needs confirmation (Flora Antarctica, 2, p. 362). It may be easily recognized by its flat leaves, slender culms, compact pale brown heads of spikelets, and by the narrow wing which surrounds the perigynia. 7. C. echinata, Murray, Prodr. des stirp. Gott., 76; F. Muell. Fragm., viii., 252; Benth. Fl. Austral., vii., 439. C. stellulata, Good., Trans. Linn. Socy., ii., 144; Hook. fil. Flora Nov. Zeal., i., 281; Handbk. N.Z. Flora, 312. North Island.—Probably not uncommon in the elevated districts of the central and southern portions of the island. Swamps in the Upper Thames Valley (a large stout form), T.F.C.; Bogs near Lake Taupo, Colenso (Handbook). South Island.—Common in marshy places in the mountain districts of Nelson and Canterbury, alt. 1,000–4,000 feet. Stewart Island, from sea-level, D. Petrie! G. M. Thomson! T. Kirk. I have seen no Otago specimens. Easily separated from all its New Zealand allies by the squarrose perigynia. Very variable in size, robustness, number and position of the spikelets, etc. The ordinary form closely resembles the common state in Europe; but I have a large coarse variety from the Thames Valley with stems nearly two feet high, and numerous distant spikelets. A tall slender variety with distant spikelets has also been gathered in Stewart Island by Messrs. Petrie and Thomson. C. echinata is also found in Northern and Central Europe, North Africa, North and West Asia to the Himalaya Mountains, and in North America.

8. C. muricata, L. Species Plant.; Cheeseman, Trans. N.Z. Inst., xvi., p. 411. South Island.—Mount Owen, Nelson; alt. 4,000 feet, T.F.C. Elsewhere I have stated that the identification of the New Zealand plant with the northern C. muricata, L., must not be taken as finally settled until mature fruiting specimens are obtained. So far as habit, foliage and inflorescence are concerned, there is little apparent difference; but the young perigynia hardly match those of a similar age in English and Swiss specimens. C. muricata is common in many parts of Europe, and is also found in North Africa, North and Central Asia to the Himalaya Mountains, and in North America. It has not been previously recorded from the southern hemisphere. 9. C. teretiuscula, Good., Linn. Trans., ii., 150; Hook. fil. Fl. Nov. Zeal., i., 281; Handbk. N.Z. Flora, 313. North Island.—Tangoia, Hawke's Bay, Colenso (Handbook); Ruataniwha Plains, H. Tryon! South Island.—Not uncommon in marshy places in the mountain districts. Motueka Valley, Wairau Valley, Upper Waimakariri, Lake Tekapo, etc., T.F.C.; Lake Ohau, Haast (Handbook); Lake Wakatipu, Valley of the Dart, T. Kirk; Strath Taieri, Port Molyneux, D. Petrie! Altitudinal range from sea-level to over 3,000 feet. By many authors this species is united with C. paniculata, some northern forms of which approach it very closely. C. teretiuscula, however, has a very different habit, and never forms the dense tussocks of C. paniculata. It is usually much smaller, the stems more slender and wiry, the panicles (or spikes) much shorter and broader, and the perigynia are also slightly different. But there is little danger of its being confounded with the New Zealand varieties of C. paniculata, all of which are widely divergent. Outside New Zealand, C. teretiuscula is found in North and Central Europe, the Himalaya Mountains, and in North America. 10. C. paniculata, L. Species Plant., 1,383; Kunth, Enum. ii” 359; F. Muell. Veg. Chath. Isl., 59; Benth. Fl. Austral., vii., 440. Var. appressa.—Stouter, stems more acutely angled; leaves wider (often nearly ½ inch); panicles more erect and rigid, stouter and broader, with more densely packed spikelets; perigynia plano-convex, margins incurved, faces very prominently nerved. C. appressa, Br. Prodr., 242; Hook. fil. Fl. Antarct., i., 90; Flora Tasm., ii., 99; Handbook N.Z. Flora, 313; Boott, Ill. Car., i., 46, t. 119, 120. Var. virgata.—Slender, panicle longer and narrower, often interrupted below, perigynia smaller, prominently nerved. C. virgata, Sol.; Boott, l. c., t. 121, 122; Hook. fil. Fl. Nov. Zeal., i., 282; Handbook N.Z. Flora, 313.

Var. secta.—Larger, rootstocks often matted together and forming miniature trunks 2–4 feet high. Panicle larger, more slender, much more compound, often with lax drooping branches; perigynia smaller and broader, faintly nerved or quite smooth and polished. C. secta, Boott in Hook. fil. Fl. Nov. Zeal., i., 283; Ill. Car., t. 123, 124. North and South Islands.—The varieties virgata and secta abundant throughout, from the North Cape to Stewart Island, and from sea-level to an altitude of over 3,000 feet. Var. appressa.—South Island.—Near Dunedin and Otago Peninsula, G. M. Thomson! D. Petrie! Milford Sound, Dr. Hector, G. M. Thomson. Auckland and Campbell Islands, Sir J. D. Hooker. (Handbook.) I have followed Mr. Bentham and Sir F. Mueller in uniting the three plants mentioned above with the northern C. paniculata, L.; a course long ago indicated by the sagacious Robert Brown. Small specimens of C. appressa (and particularly some Tasmanian ones), are almost identical with European examples; but usually the plant differs in its larger size, stouter habit, and much longer panicles, which in C. paniculata are seldom more than 4 or 5 inches long, but in C. appressa are sometimes over 18 inches, and proportionally stout. C. virgata recedes from appressa in the narrower foliage, still longer but much more slender panicles, and smaller perigynia; but transition states are occasionally seen. C. secta has drooping panicles often 2 feet long, with lax spreading branches, and in its extreme form is widely different from any northern variety of C. paniculata. Reduced or mountain forms, however, gradually approach C. virgata, the panicles becoming shorter, and the side branches less developed. After the study and comparison of many specimens, both New Zealand and foreign, it appears to me that the differences between the three varieties and between them and the typical C. paniculata are mainly those of habit, size, and luxuriance, and that there are no structural deviations of sufficient importance available for specific distinction. C. paniculata is widely distributed in New Zealand, and perhaps contributes more to the general physiognomy of the vegetation than any other species of the genus. Everyone who has had occasion to pass through swampy districts is familiar with the huge tussocks formed by the matted rootstocks of the variety secta, sometimes from four to five feet high, with a diameter of nearly two feet. Outside New Zealand, Australia and Tasmania the species ranges through Europe and Western Asia. A variety is found in California, but the plant seems to be unknown in Eastern America. 11. C. viridis, Petrie, Trans. N.Z. Inst., xiii., 332. Pale whitish-green. Stems slender, wiry, tufted, terete below, compressed or plano-convex above, grooved, perfectly smooth, 6–24 inches high. Leaves shorter than the

stems, narrow, 1/25 inch broad, erect, strict and wiry, concave in front, convex behind, or plano-convex, grooved, margins smooth or slightly scabrid. Inflorescence nearly diœcious or altogether so; spikelets pale, almost white, few flowered, sessile, collected into a linear terminal spike ½–1 ½ inch long; those of the male plant with occasional female flowers or altogether unisexual; those of the female sometimes with a staminate flower at the top of the spikelets. Bracts short. Glumes lanceolate, thin and membranous, pale, acuminate or awned. Perigynia long and narrow lanceolate, planoconvex, nerved, tapering into a very long bidentate serrate beak. Stigmas 2. South Island.—Upper Waitaki, Mackenzie Plains, Lakes Tekapo and Pukaki, alt. 2,000–4,000 feet, T.F.C.; Rough Ridge, alt. 3,000 feet; Nevis Valley, alt. 2,000 feet; and other localities in Otago, D. Petrie! Upper Shotover, P. Goyen! A most distinct species, whose only near ally appears to be the following. I have given a full description of the plant, as Mr. Petrie does not allude to the nearly diœcious inflorescence and some other curious points. 12. C. kaloides, Petrie, Trans. N.Z. Inst., xiii., 332. South Island.—Apparently plentiful in mountain districts throughout. Common in river valleys in the interior of Nelson, T.F.C.; slopes of Mount Torlesse, Broken River, Upper Waimakariri, Mackenzie Plains, Lakes Tekapo and Pukaki, and other localities in Canterbury, T.F.C.; interior of Otago, plentiful, D. Petrie! Altitudinal range 800–3,500 feet. Allied to C. viridis, but easily separated by its larger size, less strict habit, broader, flatter, and more grassy leaves, and larger often loosely branched panicles. The Nelson specimens as a rule have much larger, more compound, and more numerously flowered panicles than those from Canterbury and Otago, but I find no other difference. I have never seen specimens wholly unisexual, as in C. viridis, but the male flowers vary greatly in number; in some cases the panicles being almost entirely pistillate, while in others a considerable portion of the upper spikelets are male. C. kaloides and C. viridis are sharply marked off by their narrow lanceolate and long-beaked plano-convex perigynia; none of the other New Zealand species even approaching them in this respect. Both are peculiar to the colony. Section III.—Spikelets separate, usually stalked, unisexual; the male spikelets constantly uppermost, rarely mixed with female flowers; the lower spikelets all female, or frequently with a few male flowers at the base or apex. 13. C. vulgaris, Fries, var. gaudichaudiana, Boott, Ill. Car., iv., 169, t. 567; Benth. Fl. Austr., vii., 442; F. Muell. Fragm., viii., 257. C. gaudichaudiana, Kunth, Enum., ii., 417; Hook. fil. Fl. Tasm., ii., 99, t. 151, A.; Handbk, N.Z. Flora, 313.

North Island.—Near Auckland; Valley of the Thames; etc.: T.F.C. Probably not uncommon in the high interior country. South Island.—Common in moist places in mountain districts, rarer in the lowlands. Nelson—Mount Arthur plateau, ascending to 4,500 feet; Red Hills; Raglan Range; Wairau Gorge, etc.: T.F.C. Canterbury—Upper Waimakariri; Arthur's Pass; Burke's Pass; Mackenzie Plains; Lakes Tekapo and Pukaki; Tasman River, etc.: T.F.C. Canterbury Plains: T. Kirk. Otago—abundant in the interior: J. Buchanan! D. Petrie! G. M. Thomson! This and the two following species form a small group possessing rather broad lenticular or flattened nerved perigynia, with remarkably short beaks. C. vulgaris is easily separated from the other two by its much smaller size, few, short and nearly sessile spikelets, and by the usually obtuse glumes. Our plant is generally kept as a distinct variety, although some specimens can hardly be distinguished from northern forms. In size it varies exceedingly—from one or two inches to nearly two feet,—and the breadth of the leaves, number and size of the spikelets, shape of the glumes, colour and shape of the perigynia are all subject to considerable variation. The perigynia are very frequently attacked by the fungus Ustilago urceolorum, and ultimately converted into a dusty mass of spores. In some districts it is difficult to find specimens free from the Ustilago. C. vulgaris has a wide geographical range. It is abundant in the colder regions of Europe, Asia, and America, extending northwards as far as Greenland and Behring Straits. In the southern hemisphere, outside New Zealand, it is abundant in Australia and Tasmania, and is also found in Chili. 14. C. subdola, Boott, Trans. Linn. Socy., 20, 142; Hook. fil. Fl. Nov. Zeal., i., 282; Handbook N.Z. Flora, 314. North Island.—Not uncommon in marshy places, from Mongonui southwards. South Island.—Nelson: Acheron Valley, Travers (Handbook); Upper Takaka and Mount Arthur plateau, 2,500–4,000 feet, T.F.C. Canterbury: various places in the Southern Alps, T. Kirk, T.F.C. Stewart Island: T. Kirk. Altitudinal range from sea-level to nearly 4,000 feet. A tall, leafy, grassy species, very common in swamps in the northern portion of the colony. It is allied to the European and Australian C. acuta, L.; but the oblong emarginate glumes, usually furnished with a stout awn from the centre of the emargination, are very different to the narrow, gradually tapering glumes of C. acuta. As in C. acuta, it often happens that in some of the lower flowers of the female spikelets the rachilla is produced beyond the perigynium, sometimes bearing one or two flowers with imperfect or fully developed perigynia.

This is an additional proof of the axial nature of the rachilla, and of the correctness of the theory that the perigynium is composed of a modified bract. 15. C. ternaria, Forst. Prodr., no. 549; Hook. fil. Fl. Nov. Zeal., i., 282; Handbk. N.Z. Flora, 314. C. geminata, Schkuhr, Car., 75. C. polystachya, A. Rich. Fl. Nouv. Zél., iii., t. 21. Var. α.—Tall and stout; leaves broad, often over ½ inch; spikelets numerous, stout, pendulous. Var. β, gracilis.—Tall, slender; leaves ⅛–1/5 inch; spikelets usually numerous, long, often over 4 inches, very slender, sometimes hardly ⅛ inch in diameter. Approaches C. subdola. Var. γ, pallida.—Stout, leaves often rigid and coriaceous; spikelets fewer, short, pale, on long filiform peduncles; perigynia broader and more turgid, hardly nerved, sometimes with serrate margins. North and South Islands.—Varieties α and β abundant throughout, from the North Cape to Stewart Island and the Auckland Isles, and from sea-level to 4,000 feet. Var. γ.—Marshy places in the mountains of the South Island, apparently plentiful, D. Petrie! T.F.C. This is one of the most variable of the New Zealand Carices, as also one of the most common and widely diffused. I have endeavoured to sort its numerous forms into three main varieties, as characterized above; but it must be borne in mind that there are numerous intermediates. Var. α is usually of general occurrence in lowland districts, by the margins of swamps, along the banks of streams and in open gullies. It is often over 4 feet in height, with broad harsh and cutting leaves ½ inch or more across, and numerous (12–24) dark brown stout pendulous spikelets, that are often 3 or 4 inches long by 1/3 inch broad. Var. gracilis is usually found in woods. Small forms appear to pass into C. subdola; but ordinarily it can be distinguished by the longer awns to the glumes, and more numerous ternate or quinate pendulous spikelets. The extreme form of var. pallida has a very distinct appearance, and is probably the same as the supposed new species briefly noticed in the “Handbook” as having been collected by Haast near Lake Ohau. But in many districts transition forms between it and var. α can be collected. C. ternaria appears to be confined to New Zealand. 16. C. buchanani, Berggren, Journ. of Botany, 1880, p. 104. C. tenax, Berggren in Physiograph. Saltskaps Minneskrift Lund, 1878, t. 3, f. 1–7 (a name already occupied). South Island.—Abundant in hilly and mountain districts throughout, rare in the lowlands. Berggren, Armstrong! Buchanan! Petrie! G. M. Thomson! T.F.C. It also probably occurs in the mountainous districts in the centre of the North Island, but I have seen no specimens from thence.

A well-marked species, excellently described and figured by Dr. Berggren. Among the species with two styles it is easily recognized by its usually reddish colour, densely tufted habit, strict semi-terete leaves, pale-coloured glumes and elliptic plano-convex serrate perigynia. Dr. Berggren considers that it is allied to C. raoulii, but I can find but little affinity with that plant, which differs altogether in habit, foliage, glumes and perigynia. Mr. Petrie informs me that C. buchanani is rapidly increasing in the central districts of Otago, spreading along the sides of water-races, and in river valleys that have had the dense native vegetation cleared away. In Canterbury and Nelson it is usually found on the shingle-beds of the large rivers; but it also ascends the mountains. 17. C. lucida, Boott, in Hook. fil. Fl. Nov. Zeal., i., 283; Ill. Car., t. 173; Hook. fil. Handbook N.Z. Flora, 314. North and South Islands.—Abundant throughout, from the North Cape to Stewart Island. Altitudinal range from sea-level to 3,500 feet. This is a well-known and widely-spread species, and in its ordinary state cannot be confounded with any other. The narrow keeled leaves; distant long and narrow somewhat loose-flowered spikelets; usually entire glumes; and the biconvex turgid smooth and polished perigynia are good distinguishing characters. The culms often elongate considerably after flowering, becoming prostrate, and sometimes attaining a length of four or five feet. This form commonly has much paler glumes and perigynia than the type, but these are unimportant differences. The style-branches appear to me to be invariably two only; unless a doubtful plant with three styles, from the Wairau Valley, Nelson, of which I possess immature specimens only, be referable to this species. 18. C. dipsacea, Berggren, Physiograph. Saltskaps Minneskrift Lund, 1878, t. 7, f. 8–14. North Island.—Tokano and Omatangi (near Lake Taupo), Berggren, l. c.; Mount Egmont Ranges, alt. 3,000 feet, H. Tryon! Upper Waikato and Patetere, T.F.C.; abundant in the swamps of the Middle and Lower Waikato nearly to the mouth of the river, T.F.C. South Island.—Nelson—Lower Motueka, Graham River, Wangapeka Valley, Wairau Valley, T.F.C.; Canterbury—Southern Alps, Berggren, l.c.; Upper Waimakariri, Lake Tekapo, T.F.C.; Otago—Manuherikia Valley, Strath Taieri, Waikouaiti, Upper Waipori, Catlin's River, D. Petrie! Altitudinal range from sea-level to 3,500 feet. A larger species than the preceding, of a deeper green colour. The spikelets are much shorter and stouter, the flowers much more closely packed, the glumes broader, rounded and obtuse at the top; and the

perigynia are sharply serrate. I find that the male spikelet is not unfrequently partly female, a point not alluded to by Dr. Berggren, whose description and plate are otherwise very accurate. 19. C. raoulii, Boott, in Hook. fil. Fl. Nov. Zeal., i., 283. Yellowish or dark green. Culms rather stout, laxly tufted, 9 inches to 2 feet in height, triquetrous, scabrid. Leaves usually longer than the culms, 1/6–¼ inch broad, flat and grassy, grooved, coriaceous, margins scabrid. Bracts long and leafy. Spikelets 4–7, pale, stout, erect, approximate and sessile, or the lower ones distant and shortly pedunculate, ½–1 inch long, ¼ inch broad; all female but usually with a few male flowers below, the uppermost generally longer with more male flowers below. Glumes pale brown, broadly ovate, thin and membranous, 3-nerved, deeply bifid, awn long, stout and hispid. Perigynia broader and longer than the glumes, elliptical, narrowed to the base and upwards into a rather long and stout 2-toothed beak, strongly nerved, unequally biconvex; margins serrate above. Stigmas 2. Ill. Car., iii., 109, t. 333; Hook. fil. Handbk. N.Z. Flora, 314. C. goyeni, Petrie, Trans. N.Z. Inst., xiv., 363. South Island.—Not uncommon in mountain districts. Nelson—Graham River, Wangapeka Valley, Mount Owen, etc.: T.F.C. Canterbury—Akaroa, Raoul; Southern Alps, Sinclair and Haast; Broken River basin, Upper Waimakariri; Lakes Tekapo and Pukaki; T.F.C. Otago—head of Lake Wakatipu, D. Petrie!; near Lake Wanaka, J. Buchanan! Altitudinal range from 250–3,000 feet. Although C. raoulii is one of the most distinct species of the genus in New Zealand, there has been considerable misconception respecting it, and by many it has been confounded with C. testacea. I therefore give revised descriptions of both species. Its main characters lie in the loose open habit, comparatively broad flat leaves, in the terminal spikelet being always partly female, and in the elliptical strongly nerved and serrate perigynia, which are unequally biconvex, and narrowed both upwards and downwards. In C. testacea the perigynia are much broader, plano-convex and almost hemispherical, and in addition to the very different habit the terminal spikelet is never mixed with female flowers. 20. C. devia, Cheeseman, Trans. N.Z. Inst., XV., 301. South Island.—Mountain districts in Nelson, not uncommon above 2,500 feet: T.F.C. Allied to C. testacea, but readily distinguished by its different habit, stouter culms, rigid and coriaceous grooved leaves, very stout clavate male spikelets, and more turgid biconvex perigynia, with a shorter broader beak and entire margins.

21. C. testacea, Sol. mss.—Stems tufted, slender, leafy, nearly smooth, from 4 to 18 inches high, in some forms elongating in fruit and becoming prostrate, occasionally reaching 4–5 feet in length. Leaves longer or shorter than the stems, flat, involute or keeled, 1/15–⅛ inch broad, striate; margins more or less scabrid. Spikelets 3–5, pale brown, close together or rarely distant, sessile, or the lower shortly peduncled; the upper one male, slender, ¾–2 inches long; lower female only, or sometimes with a few male flowers below, rarely above, short and broad, erect, ½–1 ¼ inch long, ¼–1/3 inch broad. Bracts leafy, very long, overtopping the spikelets. Glumes broadly ovate, thin and membranous, bifid at the apex, with a long or short awn, pale brownish usually dotted with chestnut, median portion more or less conspicuously 3-nerved. Perigynia rather smaller than the glumes, very broadly ovate, plano-convex, nerved, polished and shining, purplish at the apex, paler below, or wholly pale brown; beak very short and broad, with two widely-divergent teeth; margins entire or serrate above. Stigmas 2.—Boott in Hook. fil. Fl. Nov. Zeal., i., 282; Hook. fil. Handbk. N.Z. Flora, 314. North and South Islands.—Abundant throughout, from the North Cape to Stewart Island. Altitudinal range from sea-level to 3,500 feet. An extremely variable plant, especially in the length of the culms, which in some forms hardly elongate after flowering, in others reach a length of from 3 to even 5 feet, lying prostrate on the ground. The description in the “Flora of New Zealand” is much more correct than the later one given in the “Handbook,” where it is stated that the perigynia are not serrate above, which they very frequently are; and also that it can be distinguished from C. raoulii by the glumes not being 2-fid—whereas they are almost invariably bifid in both species. It seems probable that more species than one have been included in the description given in the “Handbook.” Some of the forms which I include under C. testacea should perhaps have been briefly characterized as varieties. For instance the lowland form, often common near the coast, on sand-hills, etc., differs in several respects from the mountain state. But I find so many intermediates that for the present I have thought it best to give a sufficiently wide specific description and to postpone actually characterizing trivial varieties until more is known of their range and limits. 22. C. wakatipu, Petrie, Trans. N.Z. Institute, xiv., 363. South Island.—Canterbury—mountains above the Broken River, alt. 3,500 feet; mountains above Arthur's Pass, 4,000 feet; Lake Tekapo, alt. 3,000–4,500 feet, T.F.C. Otago—Ben Lomond, near Queenstown, alt. 3,000–5,000 feet; Lake Wanaka, alt. 3,500 feet, D. Petrie!

This is principally distinguished from C. testacea by its smaller size; shorter and stouter culms; and more strongly nerved biconvex perigynia. It is usually from 6 to 12 inches high, but alpine specimens are much depauperated, only 2 or 3 inches high, bearing two or three short and stout spikelets sunk amongst the leaves. 23. C. cirrhosa, Berggren, Physiograph. Saltskaps Minneskrift Lund, 1878, 29, t. 7, f. 27–34. South Island.—Waimakariri River, Berggren. Lake Lyndon, alt. 2,500 feet, J. D. Enys! T. Kirk! T.F.C. A curious little plant, first discovered by Dr. Berggren, and excellently described and figured by him. Its only near ally is the following species:— 24. C. cryptocarpa, Cheeseman, Trans. N.Z. Inst., xvi., p. 412. South Island.—Vicinity of Lake Tekapo, Canterbury, alt. 2,500 feet. T.F.C. This is near to C. cirrhosa, but differs in its larger size and stouter habit; broader glumes; larger more turgid perigynia, with shorter beaks and serrate margins; and in having 3 stigmas. The nut is trigonous, while it is lenticular in all the fruiting specimens of C. cirrhosa that I have examined. Old tussocks of this species often have a most peculiar appearance. The centre dies out, leaving a hollow ring which grows on vigorously. In dry weather the leaves all curve inwards towards the centre of the ring, giving the tussock the appearance of a cushion with a low centre. 25. C. uncifolia, Cheeseman, Trans. N.Z. Institute, xvi., p. 412. South Island.—Mountains flanking the Wairau Valley, Nelson, alt. 3,000–4,000 feet, T.F.C. Differing much in habit from the two preceding species, never forming close compact tufts, but having more of the habit of small and fine leaved specimens of C. breviculmis. Besides this, the culms and leaves are much more slender, spikelets smaller, and the perigynia of a very different shape, being trigonous or almost fusiform, or, to take a familiar example, very near to those of C. lucida in shape. As I only have it from one locality, some little allowance must be made for the description. Small specimens are often hardly more than an inch or two in height, but large ones attain 5 or 6 inches. 26. C. pulchella, Berggren, Physiograph. Saltskaps Minneskrift Lund, 1878, 29, t. 7, f. 20–26. South Island.—Mountains near the Bealey River, Canterbury, Berggren. I am not in a position to add anything to Dr. Berggren's description of this species, for although I am inclined to refer to it some specimens of a plant gathered in the Wairau Valley, Nelson, they are much too immature to warrant a positive conclusion. Dr. Berggren remarks that it is

distinguished from C. comans by the flaccid culms; sheathed spikelets, the lowest of which is nearly basal; and by the turgid perigynia, which exceed the glumes in length. From C. testacea it is at once removed by the number of the style-branches, narrower and smoother perigynia, and by the remote lower spikelet. 27. C. petriei, Cheeseman, Trans. N.Z. Inst., xvi., p. 413. South Island.—Mountains above the Broken River, and near Arthur's Pass, alt. 3,000–4,500 feet; ravines at the foot of Mount Dobson, and between Lake Tekapo and Lake Pukaki, alt. 3,000 feet, T.F.C.; near Naseby; Dunstan Mountains, alt. 3,000 feet; Lake Wanaka, alt. 3,500 feet, D. Petrie! I take this to be a perfectly distinct and well-marked form, easily distinguished from any of the species with 3 styles by its usually reddish colour; by the broad sheathing bases to the leaves and their fine points, which are curled and twisted when dry; by the small and narrow spikelets, all of which are sheathed and stalked, and the lower on filiform peduncles; by the remarkably pale coloured glumes; and by the narrow ovoid or elliptic-oblong turgid perigynia, which are quite smooth and even, and of a dark brownish- or purplish-black colour. It seems to be conveniently placed near C. pulchella although by no means closely allied to that plant. From C. testacea and the allied species it is at once removed by the number of styles and the shape of the perigynia. 28. C. comans, Berggren, Physiograph. Saltskaps Minneskrift Lund, 1878, t. 7, f. 15–19. North Island.—Sandy shores near Hokianga, Berggren; near Kaihu, Northern Wairoa, T.F.C.; Paparoa, (Kaipara), T. Kirk. With this species I am imperfectly acquainted, for although I believe that I have gathered it in the Northern Wairoa district, my specimens are few in number and very immature. Judging mainly from Dr. Berggren's plate and description, it is separated from the allied species with three stigmas by the slender habit, filiform plano-convex leaves, which are longer than the culms, remote short-stalked or sessile spikelets, and by the narrow plano-convex perigynia, with the margins sharply and coarsely toothed towards the apex. Near Lake Tekapo, Canterbury, I have collected specimens of a plant closely allied to the above, but differing in the stouter habit, much larger and broader glumes with longer awns, and rather broader perigynia. I had placed it as a variety of C. comans, but Mr. J. G. Baker, of Kew, who has done me the favour of comparing it with an authentic specimen of C. comans received from Dr. Berggren, considers it to be distinct. I hesitate, however, to describe it as new until better acquainted with Berggren's plant.

29. C. cheesemanii, Petrie, Trans. N.Z. Inst., xv., 358. South Island.—Apparently common in hilly and mountainous districts. Wairau Valley; Buller Valley; Wangapeka Valley and other localities in the interior of Nelson, T.F.C. Upper Waimakariri and near Mount Dobson, Canterbury, T.F.C. Upper Waipori; Strath Taieri; Maniototo Plain; Nevis Valley and other places in Otago: D. Petrie! G. M. Thomson! This is probably nearest to the preceding species, and although differing in several characters might well be regarded as a variety only. The culms are longer and more slender, often considerably exceeding the leaves; the leaves are flatter and rather narrower, although variable in this respect; the lower spikelets are usually on long filiform peduncles; and the glumes and perigynia are rather broader. Like several other New Zealand species it is often a red-brown colour. Large and coarse specimens have some slight outward resemblance to small forms of C. lucida, but in reality differ altogether in leaves, glumes, and perigynia. 30. C. littoralis, Petrie, Trans. N.Z. Inst., xv., 358. North Island.—Probably not uncommon in brackish-water marshes. Whangarei Harbour; Manukau Harbour; marshes by the Thames and Piako Rivers; Tauranga Harbour, T.F.C. South Island.—Nelson Harbour and marshy places by the shores of Blind Bay, T.F.C.; Port Chalmers; Paterson's Inlet (Stewart Island), D. Petrie! G. M. Thomson! This species appears to be confined to brackish-water swamps. It has been confounded with C. testacea, with which, however, it has little affinity. It is nearer to C. comans, but is easily recognized by its larger size and stouter habit, more closely placed spikelets, and particularly by the more turgid biconvex or trigonous perigynia, not (or very obscurely) serrate above. 31. C. dissita, Solander, mss.; Boott in Hook. fil. Flora Nov. Zeal., i., 284; Ill. Car., i., t. 176; Hook. fil. Handbk. N.Z. Flora, 316. North and South Islands.—Abundant in most districts, and ranging from sea-level to 3,500 feet. alt. Var. β, lambertiana.—Much stouter, leaves broader, sometimes over ¼ inch. Spikelets much longer and stouter, sometimes 2 ½ inches long, erect, short stalked. Glumes more deeply bifid. C. lambertiana, Boott in Hook. fil. Fl. Nov. Zeal., i., 284; Handbk. N.Z. Flora, 317. North and South Islands.—Apparently not uncommon throughout. I have not myself seen specimens from Otago, but I am informed by Mr. Kirk that it occurs both there and in Stewart Island. Var. γ.—Culms usually overtopped by the leaves; spikelets paler, short-stalked, lowest often compound; glumes with longer awns. North Island.—Common in the Auckland Provincial District.

C. dissita is undoubtedly one of the most variable species we possess. Commencing with the small form found abundantly in mountain swamps, and which is often hardly more than 6 or 9 inches high, with narrow' leaves and three or four short and small spikelets, every step can be traced into the tall grassy variety so generally distributed in both lowland and upland districts, and which is often more than two feet high, with 6–8 loosely placed, often pendulous, spikelets. This again varies into the stouter and harsher form hitherto kept distinct under the name of C. lambertiana, but which is certainly only entitled to rank as a variety—all its differences being those of size and luxuriance, and none of them being even tolerably constant. Another curious state, which is found in many localities in the Auckland Provincial District, has many of the characters of C. lambertiana, but differs in the softer foliage, shorter and much paler spikelets, the lowest of which are almost invariably compound, and much shorter culms. In the compound spikelets it approaches C. neesiana, but differs altogether in habit and the short stalk to the much larger spikelets. C. neesiana is certainly a close ally of C. dissita, but appears to be sufficiently distinct in the different habit, numerous male spikelets and compound lower female ones on long pendulous stalks. The distinction of entire glumes, given by Hooker in the “Handbook,” breaks down when a large suite of specimens is examined, the glumes varying in both species from quite entire to deeply bifid, and the length of the awn is equally inconstant. 32. C. neesiana, Endl. Prodr. Fl. Ins. Norfolk, 24; Boott, Ill. Car., iv., 136, t. 436; Hook. fil. Handbk. N.Z. Flora, 316. C. solandri, Boott in Hook. fil. Fl. Nov. Zeal., i., 284; Ill. Car., i., t. 176. North Island.—Not uncommon in wooded districts. South Island.—Queen Charlotte Sound, Banks and Solander (Handbook); Wangapeka, Upper Wairau and other localities in Nelson, T.F.C.; Banks Peninsula, Armstrong!; Akaroa, T. Kirk; near Dunedin and on Stewart Island, D. Petrie! This is principally distinguished from C. dissita by the taller, more slender habit; longer culms, often becoming elongated in fruit and prostrate; more numerous male spikelets (the male spikelets are seldom more than one in C. dissita); and by the compound long peduncled lower female spikelets. The perigynia and glumes are much alike in both species. C. neesiana is found in Norfolk Island, but does not seem to extend into Australia or Tasmania. 33. C. longiculmis, Petrie, Trans. N.Z. Inst., xiv., 363. South Island.—Paterson Inlet, Stewart Island, D. Petrie! G. M. Thomson! T. Kirk!

A very distinct species. Easily separated from C. dissita var. lambertiana, by its different habit; shorter, much stouter, pale brown spikelets, and longer beak to the broader and somewhat stipitate perigynia. From C. trifida it is removed by its smaller size and more slender habit, less numerous and smaller spikelets, broader glumes, and very different perigynia. The short broad blunt spikelets remind one of those of some species of the Vesicariæ section, but the perigynia are unlike. 34. C. trifida, Cav. Ic., 41, t. 465; Boott in Hook. fil. Fl. Antarct., i., 89; Fl. Nov. Zeal., i., 284; Handbk. N.Z. Flora, 316. C. incrassata, Solander, mss. C. aristata, D'Urv. South Island.—Totaranui (Queen Charlotte Sound) Banks and Solander (Handbook); Akaroa, Raoul; Dusky Sound, Lyall; near Dunedin, D. Petrie! G. M. Thomson! Auckland and Campbell Islands, Sir J. D. Hooker (Flora Antarctica). Also common in temperate South America, from Chili and Fuegia to the Falkland Islands. Its large size, stout habit, and numerous massive spikelets readily distinguish it from any other species found in New Zealand. 35. C. breviculmis, R. Br. Prodr., 242; Boott in Hook. fil. Fl. Nov. Zeal., i., 283, t. 63; Fl. Tasm., ii., 101; Handbk. N.Z. Flora, 316; Boott, Ill. Car., iv., 181; F. Muell. Fragm., viii., 255; Benth. Fl. Austral., vii., 445. North and South Islands.—Abundant in dry places from the North Cape to Stewart Island. Altitudinal range from sea-level to 3,000 feet. Easily distinguished by its pubescent perigynia. It has a wide range outside New Zealand, being found in Lord Howe Island, South-eastern Australia and Tasmania, the Himalaya Mountains, and in China and Japan. 36. C. pumila, Thunb. Flora Japon., 38; Benth. Fl. Hongkong, 482; Fl. Austral., vii., 445; Boott, Ill. Car., iv., 217; Hook. fil. Handbook N.Z. Flora, 315. C. littorea, Lab. Fl. Nov. Holl., ii., 69, t. 219; R. Br. Prodr., 243; Hook. fil. Fl. Nov. Zeal., i., 284; Fl. Tasm., ii., 200. North and South Islands.—Abundant on sandy shores and on sand-dunes, from the North Cape to Stewart Island. A well-marked plant. The long running rhizomes sending up distant stems; the glaucous keeled leaves; and large, smooth and turgid perigynia, exceeding in size those of any other New Zealand species, are conspicuous characters. Common in Eastern Australia, in Tasmania, and in some parts of extra-tropical South America, and also found on the coasts of China and Japan. 37. C. flava, L. Species Plant.; Kunth Enum., ii., 446; Benth. Fl. Austral., vii., 444. C. cataractæ, R. Br. Prodr., 242; Hook. fil. Fl. Tasm., ii., 101, t. 151; Handbook N.Z. Flora, 315; Boott, Ill. Car., iv., 204.

South Island.—Amuri, T. Kirk! Southern Alps, Sinclair and Haast (Handbook); Canterbury Plains, T. Kirk; Lakes Tekapo and Alexandrina, and by the Tasman River, T.F.C.; Waitaki River, Haast; Dunedin, D. Petrie! Lake District of Otago, J. Buchanan! Stewart Island, T. Kirk. Altitudinal range from sea-level to over 3,000 feet. I have followed Mr. Bentham in uniting the Tasmanian and New Zealand C. cataractæ with the northern C. flava; the differences between the two forms being hardly of specific value. New Zealand specimens never seem to attain the size of European, and as a rule the spikelets are much more closely compacted, and the perigynia smaller. In all my specimens the beak of the perigynium is shorter than in the typical C. flava, in this respect approaching the variety æderi, often kept as a distinct species. Specimens from Swanport, Tasmania, kindly forwarded to me by Baron Müeller, have indeed perigynia almost indistinguishable from those of C. æderi, as is stated in the “Flora Australiensis;” but I have not seen any specimens collected in New Zealand that exactly match them. C. flava has a wide distribution; being found through the greater portion of Northern and Central Europe, in temperate North America, Madeira, Western and Central Asia to the Himalaya Mountains, and in Tasmania. 38. C. vaccilans, Sol., mss.; Boott in Hook. fil. Fl. Nov. Zeal., i., 285; Handbk. N.Z. Flora, 317. North Island.—Not uncommon on declivities in dry woods, especially near the sea. I have seen no specimens from the South Island. This is a handsome and in my opinion very distinct species. Sir J. D. Hooker remarks (Handbook, p. 311) that it should perhaps be united with C. dissita; but the long and slender spikelets, narrow entire glumes, and much longer and narrower fusiform beaked perigynia do not appear to show any close alliance to that plant. In habit and in the shape of the perigynia there is considerable resemblance to the next species, which has induced me to place the two plants close together. C. vaccilans is, however, much smaller in all its parts, and the perigynia do not spread when mature. 39. C. forsteri, Wahl. in Act. Holm., 1803, 154; Boott in Hook. fil. Fl. Nov. Zeal., i., 285; Hook. fil. Handbk. N.Z. Flora, 315 (in part only). C. debilis, Forst. Prodr., 92, non Michx. C. recurva, Schkuhr Car., f. 84. C. punctulata, A. Rich. Fl. Nouvelle Zél., t. 22. C. cinnamomea, Cheeseman, Trans. N.Z. Inst., xiv., 301, not of Olney. North Island.—Whangarei Heads; Thames Goldfields; Raglan; near Gisborne, T.F.C.; Wellington, T. Kirk. South Island.—Lower Motueka, Graham River, Wangapeka Valley, and other localities in Nelson, T.F.C. Akaroa, T. Kirk. Altitudinal range from sea-level to 3,000 feet.

In the opinion of many botanists, C. forsteri ought not to be separated, save as a variety, from the plant named C. fascicularis by Solander. The two were kept as distinct by the late Dr. Boott in his account of the genus given in the “Flora of New Zealand,” but were united by Sir J. D. Hooker in the “Handbook.” Sir Fred. von Müeller, in the eighth volume of his “Fragmenta,” adopts Hooker's views as to the identity of C. forsteri and C. fascicularis, and reduces both to the widely distributed C. pseudo-cyperus, L.; and in this he is apparently followed by Mr. Bentham (“Flora Australiensis, 7, p. 449). With the reduction of C. fascicularis to C. pseudo-cyperus I perfectly agree; but at present I am not prepared to adopt similar views with respect to C. forsteri. After an examination of numerous specimens from both Islands, and a comparison with New Zealand, Australian, and European examples of C. pseudo-cyperus, it appears to me that C. forsteri differs in several prominent characters. Its habit is by no means the same, being nearer that of C. vaccilans, but stouter; the leaves are harsher and more rigid, differing much from the soft and grassy foliage of C. pseudo-cyperus; the spikelets are much more distantly placed, the males being often two or three in number; and the perigynia are smaller and broader, with much shorter beaks. The plant described by me as C. cinnamomea (a name, by the way, preoccupied by a North American species), I am now convinced must be reduced to C. forsteri. It is usually of a more slender habit, with longer awned glumes, bright red-brown spikelets, and still smaller perigynia with shorter, obscurely-toothed beaks. 40. C. pseudo-cyperus, L. Species Plant., 1387; Kunth Enum., ii., 501; F. Muell. Fragm., viii., 249; Benth. Flora Austral., vii., 448; Boott, Ill. Car., iv., t. 451, 452. C. fascicularis, Sol. ex Boott in Hook. fil. Fl. Nov. Zeal., i., 283; Boott, Ill. Car., i., t. 189, 140. North and South Islands.—Abundant throughout, in marshes or swampy woods. Altitudinal range from sea-level to 3,000 feet. This is the C. forsteri, β minor and γ fascicularis of the Handbook. As stated in the remarks to the preceding species, I quite agree with Mueller and Bentham in uniting it with the northern C. pseudo-cyperus. The species, in some of its forms, is found in Europe, Western and Southern Asia, North and South Africa, North and South America, and in Australia. Addendum. In addition to the above species, Mr. Kirk* Trans. N.Z. Inst., vol. x., appendix, p. 42. has added C. chlorantha, Br., a common Australian and Tasmanian plant closely allied to C. paniculata. His specimens were gathered, some years ago, in a locality within

the precincts of the City of Auckland. In this station it has since become extinct; and as the plant has not been observed elsewhere in New Zealand, I am of opinion that it should be regarded as a chance introduction only. Mr. Colenso* Trans. N.Z. Inst., vol. xv.,p.335. has also described a C. spinirostris as new. His description, however, is exceedingly vague; and the only specimen I have seen is in such poor condition that I have been unable to form an opinion as to the validity of the species.