Stewart Island Herekopere Island and headlands near the South Cape, Kirk!The Snakes Kirk! Chapman! Auckland and Campbell Islands: Abundant, Hooker, Buchanan! Kirk! Chapman! Antipodes Island; Kirk! Macquaeie Island: Fraser, Professor Scott, A. Hamilton!

Easily distinguished from all other New Zealand species by the great size, very broad flat leaves, and large dense panicle. It is closely allied to the Kerguelen Island P. Cookii,Hook, f., and to the well-known tussock grass of the Falkland Islands and Fuegia, P. flabellata,Hook. f. (Dactylis cœspitosa,Forst.). The flowers seem to be partly if not altogether unisexual, most of the specimens that I have seen being females with the anthers much reduced in size.

P. novæ-zealandiæ, Hack. in Trans. N.Z. Inst. xxxv.(1903) 381.—Perennial, tufted, innovation-shoots intravaginal or mixed. Culms erect, 3–18 in. high, slender, glabrous, 3-noded, upper node about the middle of the culm. Leaves usually much shorter than the culms, 2–10 in. long, 1/12–¼ in. broad, linear, suddenly acuminate at the tip, flat or those of the innovation-shoots complicate, erect, quite glabrous, finely striate; sheaths lax, compressed; ligules ovate, acuminate, often dentate. Panicle broadly ovate to ovate-oblong or linear-oblong, dense, nodding or more rarely erect, 1–4 in. long; rhachis smooth, terete; branches binate or ternate, once or twice divided, smooth, capillary. Spikelets pale-green or whitish-green, much compressed, elliptic-oblong, ¼–⅓ in. long, 5–6-flowered. Two outer glumes slightly unequal, acuminate, glabrous; lower subulate-lanceolate, 1-nerved; upper longer and broader, about half as long as the whole spikelet or rather more, lanceolate, 3-nerved. Flowering glumes lanceolate, acuminate, often incurved at the tip, usually 5-nerved, but the intermediate nerve on each side faint and sometimes obsolete, callus with a tuft of crisped woolly hairs more than half as long as the glume, remainder of the glume glabrous, smooth. Palea ⅓ shorter than the glume, linear-oblong, bidentate, pubescent on the keels.—P. foliosa var. b, Hook. f. Handb. N.Z. Fl. 338; Buch. N.Z. Grasses, t. 43. Festuca foliosa, Hook. f. Fl. Nov. Zel. i. 308 (not of Fl. Antarct.).

Var. subvestlta, Hack. l.c.—Flowering glumes rather longer, clothed with crisp hairs in the lower ⅓, exterior lateral nerves more prominent. Spikelets often tinged with violet.

North Island: Mount Hikurangi, Petrie!Mount Egmont, Buchanan. T. F. C-Tararua Mountains, Buchanan! Towson! South Island, Stewart Island: Abundant in damp alpine and subalpine localities. 2500–6500 ft.

Separated without any difficulty from P. foliosa, with which it was placed by Hooker, by the different habit, much smaller size, shorter narrower and smoother leaves, smaller and proportionately broader panicle, and less prominently nerved flowering glumes, which are less silky at the base, and almost smooth above.

South Island: Otago—Abundant on the cliffs of the east and southern coasts, from Port Chalmers southwards, Lyall, Kirk! Petrie! H. J. Matthews! Stewart Island, The Snares, Auckland and Campbell Islands, Antipodes Island: Plentiful on rocks near the sea.

A very distinct species, with a good deal of the habit and appearance of small states of Festuca littoralis, which no doubt induced Sir J. D. Hooker to place it in the same genus. But it has the keeled flowering glumes and punctiform hilum of Poa; and, as Professor Hackel has pointed out to me, must be transferred to that genus. As there is already a Poa scoparia(Kunth, Rev. Gram. ii. 535) a new name is required. Hooker describes the flowering glume as "basi longe villoso-barbata," but it is frequently quite free from hairs.

Auckland and Campbell Islands: Abundant near the sea, Sir J. D. Hooker!

Of this species I have only seen a single panicle from one of Hooker's Campbell Island specimens, and in default of further information, the above description is based upon that given in the "Flora Antarctica." Hooker remarks that "this is a very abundant grass in both groups of islands, and of a most singular habit of growth. The culms are invariably prostrate and quite simple for a foot or so, when they suddenly ascend and divide into many short leafy branches, each bearing a panicle of flowers. It forms a copious, soft, green herbage, especially on the banks near the sea, always throwing its long culms over the edges of the cliffs, which are thus fringed with a delicate festoon of green."

Kermadec Islands: Abundant on Sunday and Macaulay Islands, chiefly near the sea, T. F. C. Miss Shakespear!

Distinguished by the branching habit, distichously spreading leaves, short contracted panicle, and narrow acuminate flowering glumes, which are sharply scabrid on the keel, and scaberulous on the surfaces and veins.

Var. condensata,Cheesem.—Culms 4–18 in. high, often overtopped by the leaves. Panicle shorter and much more compact, dense-flowered. Spikelets rather smaller, 2–4-flowered. Glumes smoother, hardly scaberulous.

Var. gracilis,Cheesem.—Culms slender, 4–18 in. high. Leaves narrow, often involute. Panicle lax; branches few, 2–3 from each node or solitary; spikelets fewer, towards the tips of the branchlets, smaller, 2–5-flowered. Glumes smoother. This appears to connect the type with P. seticulmis.

North and South Islands: The typical form throughout the whole of the North Island, ranging from sea-level to 3500 ft., apparently rare and local in the South Island, but recorded from Marlborough and Nelson, and extending along the West Coast to the south of Westport. Var. condensata not uncommon as far as Canterbury; var. gracilis to Foveaux Strait.

What I consider to be the typical state of this variable plant includes the two varieties elataan d foliosa of the Handbook, and can be distinguished by the tall stout culms often branching at the base, broad and flat subdistichous smooth leaves, large usually lax panicle, and numerous rather large spikelets, with subacute flowering glumes prominently nerved and usually more or less finely scaberulous. But it runs on all sides into numerous varieties exceedingly difficult to define, if, indeed, they are capable of exact circumscription.

North and South Islands, Stewart Island Not uncommon in dry places throughout. Sea-level to 2500 ft.

What may be taken as the typical form of this species is abundant on sandy soil near the sea in the northern part of the North Island, and from its very slender filiform culms and leaves and lax few-flowered panicle presents a very distinct appearance. But, as Mr. Petrie remarks, there is a widely spread inland state that cannot be separated from it by any characters of importance, but which gradually varies into small and slender states of P. anceps, the var. gracilis of that plant forming a direct connection between the two species.

South Island: Wet places in mountain districts, from the Wairau Valley, Nelson, to the south of Otago. Stewaet Island Kirk!Sea-level to 5000 ft.

I am greatly puzzled with this species. Forms very closely resembling Berggren's plate and description are not uncommon in subalpine localities in the South Island, but they appear to pass insensibly into a larger lowland state, with a more developed panicle and larger spikelets. This in its turn approaches so near to P. seticulmis that it is difficult to draw a strict line of demarcation between the two plants.

South Island: Nelson—Raglan Range, T. F. C.Canterbury—Wet places near the sources of the Broken River, Petrie! T. F. C.;Craigieburn Mountains, Cockayne! 3000–5000 ft.

This seems to be a distinct species, recognised without much difficulty by the long decumbent bases of the culms, very lax few-flowered panicle, and large spikelets clustered at the tips of the branchlets. Depauperated states approach P. pusilla, but are easily distinguished by the larger spikelets and more distinctly nerved scaberulous flowering glumes.

South Island: Nelson—Lake Tennyson, T. F. C.; near Westport, Townson!

Professor Hackel remarks of this species that it is allied to P. anceps, but differs markedly in the stoloniferous rhizome, the rhizome of P. anceps being invariably tufted and without stolons. The spikelets are also broader, the two outer glumes longer in proportion and smoother, and the flowering glumes much more hairy at the base and smoother above. P. dipsacea differs in the more flaccid habit, larger spikelets, and in the flowering glumes being scabrous above.

Chatham Islands: Abundant in Sphagnum swamps, Cox and Cockayne!

Closely allied to P. anceps, but sufficiently distinct in the creeping rhizome, more coriaceous erect leaves, ligule composed of short stiff hairs, and short dense panicle with few branches and rather large spikelets.

Var. leioclada, Hack. MSS.—Panicle-branches smooth. Spikelets larger, ⅓ in. long or more.

Var. australis, Benth. Fl. Austral. vii. 653.—Leaves rough and scabrous. Panicle very lax and spreading. Perhaps naturalised.

North and South Islands, Stewart Island: The typical state abundant from the Upper Thames and Waikato southwards. Var. leioclada:Mount Egmont, Petrie! near Westport, Townson!Var. australis:Marua, near Whangarei, H.Hawkins! near Auckland, T. F. C. Sea-level to 4000 ft." "Tussock Grass."

Also in Australia and Tasmania. The most abundant grass through wide districts in the South Island, also plentiful in the elevated central portions of the North Island. Unfortunately, it is not relished by stock, and is seldom eaten, save in the absence of better food.

Var. intermedia, Cheesem.—Taller, with more of the tussocky habit of P. cæspitosa, 9–18 in. high. Ligules as in the typical state. Panicle larger, 2–3 in. long. Spikelets more numerous, larger, ⅓ in. long, but flowering glumes as in the type.—P. intermedia, Buch. N.Z. Grasses, t. 48A.

North Island: Mountainous localities and dry elevated plains of the interior, from Moehau (Cape Colville) southwards, but rare and local to the north of Lake Taupo. South Island, Stewart Island: Plentiful throughout. Usually from 1000 to 5000 ft., but descends almost to sea-level in the south of Otago, and ascends to over 7000 ft. on Mount Egmont.

A very remarkable species. I have reunited Mr. Buchanan's P. intermedia with it, there being no differences save those of size and habit, in which respect the two forms pass into one another by insensible gradations. Professor Hackel also takes the same view. Small states of P. cæspitosa can always be distinguished by the almost obsolete ligules and by the long crisped hairs on the callus of the flowering glume.

P. Colensoi is one of the most important of the indigenous pasture-grasses, It is eaten by all kinds of stock, and is a specially valuable sheep-grass in mountain districts.

South Island: Nelson—Mount Arthur, A. Mackay!Canterbury—Limestone rocks in the Broken River Basin, Enys! Kirk! T. F. C. 2000–4000 ft.

Very closely allied to P. Colensoi, and chiefly distinguished by the peculiar habit, short rigid acicular leaves, and densely silky flowering glumes.

South Island: Otago—Summit of Mount Pisa, altitude 6000 ft., Petrie!A very remarkable little species, quite unlike any other.

Var. Mackayi,Hack. MSS.—Taller and stouter; leaves often ¼ in. broad. Spikelets larger, ⅓ in. long. Flowering glumes often with a tuft of long crisped hairs on the callus. Anthers 1/15 in. long.—P. Mackayi, Buch. N.Z. Grasses, t. 50A.

Var. Collinsii,Hack. MSS.—Slender, pale-green, laxly tufted, 1–2 ft. high; nodes of the culm usually naked. Panicle 3–6 in. long, very lax; branches long, slender. Spikelets as in the type, but rather larger.—P. Collinsii, Kirk ex Petrie in Trans. N.Z. Inst. xxviii. (1896) 589 (name only).

North Island: Mount Egmont, Petrie! T. F. C.;Tararua Range, H. H. Travers! T. P. Arnold! South Island: Abundant in subalpine localities throughout. Var. Collinsii:Mount Fyffe (Marlborough), Kirk!Hooker Valley, T.F.C. 2000–5000 ft.

A variable plant. Buchanan's P. Mackayi looks different at first sight, from its larger spikelets and broader leaves, but is connected with the type by intermediate forms. On the Mount Arthur Plateau, Nelson, the two varieties can be seen to merge into one another. P. Collinsii is a taller and more slender plant, with a larger and laxer panicle, but the structure of the spikelets is the same as in the type. P. Kirkii is a valuable grass for all kinds of stock in cool elevated localities, and is well worth cultivation.

North Island: Hawke's Bay—Ruataniwha Plains, H. Tryon! South Island: Not uncommon from the south of Nelson to Foveaux Strait. Sea-level to 5000 ft.

A pretty and distinct species, easily recognised by its small size and slender delicate habit, lax panicle, small silvery-brown spikelets, and faintly nerved silky flowering glumes. Hooker describes the flowering glumes as glabrous and nerveless, but I do not find them so.

Auckland Islands: F. R. Chapman!

I have seen very few specimens of this species, and the above description will probably require modification when a larger series is obtained. It appears to be nearest to P. exigua, but the panicle is much more lax, the spikelets larger and more turgid, and the outer glumes are much shorter.

South Island: Otago—Lake district, Hector and Buchanan;Mount Pisa, Hector Mountains, Mount Cardrona, Petrie!Humboldt Mountains, Cockayne! 3500–6000 it.

South Island: Canterbury—Broken River Basin, Mackenzie Plains, T. F. C.Otago—Dry plains in the interior, Kurow, Maniototo, Bendigo, Mount Pisa, Petrie!Lake Wanaka, Kirk! 1200–3000 ft.

South Island: Probably not uncommon on dry shingle slopes in alpine localities. Nelson—Mount Percival, T. F. C.;Mount Captain, Kirk,Canterbury—Mount Torlesse, Berggren, Petrie! mountains above the Broken River, T. F. C.;Mount Dobson and Mount Darwin, Haast.Otago—Mount St. Bathan's, Mount Ida, Mount Kyeburn, Petrie! 3500–6000 ft.

A very peculiar and distinct little species, quite unlike any other.

Var. Matthewsii,Hack. MSS.—Taller, 10–20 in. high. Panicle larger, 4–8 in. long. Spikelets rather larger, 4–6-flowered. Flowering glumes closer, usually 5-nerved, but the intermediate nerves on each side often very faint.—P. Matthewsii, Petrie in Trans. N.Z. Inst. xxxiv. (1902) 392. P. breviculmis, Hook. f. Handb. N.Z. Fl. 337 (in part).

North and South Islands, Stewart Island, Chatham Islands: Not uncommon in shaded places throughout. Sea-level to 4000 ft.

As a species, P. imbecilla is well characterized by the slender flaccid habit, small spikelets with minute outer glumes, and obtuse glabrous flowering glumes, which are usually 3-nerved in the typical form, but generally 5-nerved in var. Matthewsii.It is said to occur in Australia.

Auckland Islands: Kirk! Chapman! Campbell Island: Sir J. D. Hooker!

Of this species I have only seen a fragment of one of Hooker's Campbell Island specimens, and two or three collected on the Auckland Islands by Kirk and Chapman. All these differ from P. imbecilla in the very unequal and much more minute outer glumes, and in the prominently nerved and acute flowering glumes. How far these characters are constant can only be ascertained from the inspection of a larger series of specimens. The New Zealand examples referred to P. breviglumis in the Handbook are probably referable to P. imbecilla var. Matthewsii.