Art. XXI.—Contributions to a Fuller Knowledge of the Flora of New Zealand: No. 3.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 42, 1909, Page 200

Art. XXI.—Contributions to a Fuller Knowledge of the Flora of New Zealand: No. 3. By T. F. Cheeseman, F.L.S., F.Z.S., Curator of the Auckland Museum. [Read before the Auckland Institute, 22nd November, 1909.] As in the two previous instalments of notes bearing this title (see Trans. N.Z. Inst., vol. xxxix, p. 439; vol. xl, p. 271), I have to acknowledge my indebtedness to many friends and correspondents in various parts of the Dominion for much of the matter now placed on record. I particularly desire to mention my obligations to the late Mr. H. J. Matthews, whose position as head of the Forestry Department gave him many opportunities of obtaining information respecting the New Zealand flora, and who for years past has liberally supplied me with specimens, notes, and photographs. His decease will be greatly regretted by New Zealand botanists, by whom he was held in high esteem, not only for his many services to botanical science, but also for his amiable and genial disposition, the simplicity of his manners, and his truly lovable character. I have also to express my thanks to the following gentlemen: to Mr. F. G. Gibbs, of Nelson, who has forwarded many interesting notes and specimens, mostly dealing with the botany of the north-western portion of the Nelson District; to Mr. B. C. Aston, of Wellington, for many specimens collected through almost the whole length of the Dominion; to Mr. R. H. Matthews, of Kaitaia; to Messrs. H. Carse, P. H. Allen, A. Allison, Rev. F. R. Spencer, F. N. R. Downard, J. H. Harvey, G. Williams, W. Townson, R. Mair, E. Linton, E. Clarke, C. Dawes, and others. I. Ranunculaceæ. Ranunculus Buchanani, Hook. f. Abundant on the slopes of Mount Balloon, McKinnon's Pass, southwest Otago; altitude, 4,000 ft.; H. J. Matthews! F. G. Gibbs! Ranunculus Matthewsii, Cheesem. Southern slopes of Mount Balloon, a few hundred feet above the track over McKinnon's Pass; H. J. Matthews! Ranunculus sericophyllus, Hook. f. McKinnon's Pass, south-west Otago; 4,000-5,000 ft.; H. J. Matthews! F. G. Gibbs! IV. Violaceæ. Hymenanthera novæ-zealandiæ, Hemsl. Mr. W. B. Hemsley (“Kew Bulletin,” 1908, p. 95) has pointed out that the New Zealand plant referred by Kirk and myself to the Norfolk Island H. latifolia is in reality distinct from that species. He states that the Norfolk Island plant has relatively more slender branches, thin entire leaves, smaller flowers, and the staminal appendages are only fringed at the tips instead of being minutely toothed all round the margin. Mr. Hemsley has also satisfied himself that Cunningham's Scœvola novœ-zealandiœ, collected at Matauri Bay, opposite the Cavallos Islands, and which Hooker referred

to his H. crassifolia, is really the same plant. That being so, Cunningham's specific name should be restored. Mr. Hemsley also suggests that the Chatham Islands' H. chathamica is not separated by characters of sufficient importance, and should be included in the circumscription of the species. As I have remarked in the Manual, it principally differs in the longer and narrower and much more sharply toothed leaves. V. Pittosporaceæ. Pittosporum ellipticum, T. Kirk. Mountain range south of Te Aroha; F. N. R. Downard! Differs from the typical form of the species in the narrower leaves and cream-coloured flowers, and may be deserving of varietal rank. XI. Tiliaceæ. Entelea arborescens, R. Br. Plentiful on Fisherman's Island, a few miles from Motueka, Nelson; F. G. Gibbs! XIV. Rutaceæ. Melicope ternata, Forst. Bush at Wakapuaka, near Nelson; F. G. Gibbs! XV. Meliaceæ. Dysoxylum spectabile, Hook. f. Fairly plentiful on Pepin Island, a few miles north of Nelson; F. G. Gibbs! XVII. Stackhousiaceæ. Stackhousia minima, Hook. f. Spooner's Range, near Belgrove, Nelson; F. G. Gibbs! XXIII. Rosaceæ. Geum uniflorum, Buch. McKinnon's Pass, south-west Otago; H. J. Matthews! F. G. Gibbs! These are the first specimens I have seen from Otago. Mr. Petrie informs me that it has also been collected on the Takitimo Ranges, Southland. Acæna glabra, Buch. Black-birch Creek, Hooker Valley, Mount Cook District; T. F. C. McKinnon's Pass, south-west Otago; F. G. Gibbs! XXIV. Saxifragaceæ. Quintinia acutifolia, T. Kirk. Mount Owen, Buller Valley; F. G. Gibbs! XXVII. Haloragaceæ. Dr. Schindler, in his recent monograph of the family (“Das Pflanzenreich,” heft 23), makes a consideráble number of changes in the nomenclature of the New Zealand species of Haloragis. As some of these appear

to be of doubtful validity, I propose to make a few remarks upon them in this place. Haloragis alata, Jacq. Schindler adopts the specific name of erecta, which was published by Murray, under the genus Cercodia, in 1780, whereas Jacquin's name of H. alata did not appear until the following year. If the rules adopted by the Vienna Congress are to be accepted, the earliest specific name must be maintained, no matter in what genus it originated, however much we may regret the necessity of changing an old-established name like that of H. alata. Haloragis cartilaginea, Cheesem. In the Manual I reduced this to the position of a variety of H. alata, a view I am still inclined to hold. Dr. Schindler treats it as distinct. Haloragis tetragyna, Labill. Schindler considers that the Australian and New Zealand plants combined under this name by Hooker and Bentham are in reality distinct. Through the kindness of Dr. Maiden, who has forwarded me an excellent series of the Australian forms, I have been enabled to make a detailed comparison, and find that the New Zealand plant can usually be distinguished by its smaller average size, shorter and less pointed leaves, with fewer serratures, and larger and narrower fruit. As the name tetragyna must be retained for the Australian plant, to which it was first applied, Dr. Schindler proposes to adopt Buchanan's name of H. aggregata for our species. But this is inadmissible, for the following reasons: (1.) Because there is an earlier specific name of incana—Cercodia incana, A. Cunn., Prodr., n. 528 (1839); Haloragis incana, Walp., Rep. ii, 99 (1843)—which consequently takes precedence. (2.) Because H. aggregata is referable to H. depressa, and not to the series of forms included by Hooker under H. tetragyna. I possess one of Buchanan's type specimens, and there is another in Mr. Kirk's herbarium, both being clearly the same as H. depressa. Further, the plate of H. aggregata given by Buchanan (Trans. N.Z. Inst., iv, pl. 13) undoubtedly represents H. depressa, the fruit being shown to be tetragonous, with smooth interspaces between the ribs, exactly as in H. depressa. In H. tetragyna the interspaces are muricate or rugose. I hold, therefore, that the name to be used is H. incana, Walp. In this place attention may be drawn to the remarkably distinct plant called by Hooker var. diffusa. This differs from the typical H. incana in the slender, much branched, procumbent or prostrate stems, and in the smaller, broader, and more obtuse leaves, with fewer serratures. These characters are constant throughout its entire range, which extends through almost the whole length of the Dominion, whereas typical H. incana has never been found to the south of Whangarei. I therefore think that it should receive the rank of a species, a view which has also been adopted by Dr. Cockayne in his report on the vegetation of Stewart Island (p. 57). Dr. Cockayne proposes the name of H. diffusa, but that is preoccupied by an Australian plant—H. diffusa, Diels in Engl. Bot. Jahr., xxxv (1904), 447. As the plant was excellently described and figured by Solander in his manuscript “Primitiæ Floræ Novæ Zealandiæ” under the name of Cercodia procumbens, I would suggest that it should bear the name of Haloragis procumbens.

Haloragis depressa, Walp. Here, too, we are indebted to Dr. Schindler for pointing out that the Australian and New Zealand plants associated by Hooker and Bentham under the name of H. depressa must be separated. The New Zealand plant has rather large oblong-obovoid tetragonous fruit, with very large persistent styles, decurved between the calyx-lobes, whereas the Australian plant has small oblong barely tetragonous fruit, and the persistent styles are much smaller and erect. Schindler applies the name of depressa to the Australian species, using Kirk's name of uniflora for the New Zealand plant. Evidently he did not fully trace the history of the name depressa, or he must have noticed that it was first applied by Allan Cunningham to the New Zealand plant (Goniocarpus depressus, A. Cunn., Precur., n. 531). Consequently our plant must retain the name of depressa. Schindler divides the species into the two varieties, genuina (which is equivalent to H. uniflora, T. Kirk) and bibracteolata. But Cunningham's type certainly falls under the second variety, to which also Buchanan's H. aggregata must be referred. The two varieties will therefore be more correctly styled var. a, aggregata, and var. b, uniflora, the latter name being equivalent to Schindler's var. genuina. Kirk's name of serphyllifolia must be abandoned on account of the fact that the New Zealand plant differs from the Australian var. serpyllifolia. Haloragis spicata, Petrie. This is placed by Schindler as a variety of H. depressa. In the Manual I have hinted that this is its correct position. The following arrangement is therefore the one which I now follow of the New Zealand species of Haloragis:— 1. H. erecta, Schindl. (H. alata, Jacq.). 2. H. cartilaginea, Cheesem. (H. alata, var. cartilaginea, Cheesem.). 3. H. incana, Walp. (H. tetragyna, Hook. f., non Labill, var. a; H. aggregata, Schindl., non Buch.). 4. H. procumbens, Cheesm. (H. tetragyna, var. diffusa, Hook. f.; H. aggregata, var. diffusa, Schindl.; H. diffusa, Cockayne, non Diels). 5. H. depressa, Walp. Var. a, aggregata, T. Kirk (H. aggregata, Buch., non Schindl.; H. uniflora, var. bibracteolata, Schindl.). Var. b, uniflora, Cheesm. (H. uniflora, T. Kirk; H. uniflora, var. serpyllifolia, T. Kirk, but not H. depressa, var. serpyllifolia, Benth.; H. uniflora, var. qenuina, Schindl.). Var. c, spicata, Schindl. (H. spicata, Petrie). 6. H. micrantha, R. Br. XXVIII. Myrtaceæ. Myrtus Ralphii, Hook. f. Tauroa (Reef Point), near Ahipara, Mangonui County; R. H. Matthews! This is a marked extension of the range of this species, the northern limit of which had been believed to be at Whangarei. XXIX. Onagraceæ. Epilobium pycnostachyum, Haussk. Shingle-slopes on Ben Nevis, Nelson, and on Waingaro Peak, north-west Nelson; F. G. Gibbs! These are the most northern localities yet recorded.

XXXIII. Umbelliferæ. Azorella. Dr. Karl Domin, in a communication entitled “Ueber eine neue australantarktische Umbelliferen Gattung,” published in Engler's “Botanische Jahrbucher” (vol. xl, p. 573), constitutes a separate genus of the whole of those species of Azorella included in Hooker's section Schizeilema, pointing out that the section is not only separated by a marked difference in habit, but by other characters of importance. There is much to be said in favour of Domin's views, which will probably receive the acceptance of New Zealand botanists. As characterized by him, the genus Schizeilema will include the whole of the species referred to Azorella in my “Manual of the New Zealand Flora,” with the exception of A. Selago, which is a typical Azorella. It will also include two species from the extreme south of South America, and the Australian Azorella Muelleri, Benth. (Schizeilema fragosum, Domin). The following synopsis of the New Zealand species, abbreviated from Domin's memoir, will show his views regarding them. Leaves 3–5-partite or -foliate. Leaflets distinctly petiolulate Divisions all sessile. 1. S. trifoliolatum. Segments entire or nearly so Segments more or less cut, lobed or dentate. 2. S. nitens. Fruit longer than its pedicel. Fruit shorter than its pedicel. Involucral leaves obtuse. 3. S. hydrocotyloides Stout; leaves coriaceous 4. S. Roughii. Slender; leaves membranous 5. S. pallidum. Involucral leaves acuminate 6. S. Colensoi. Leaves undivided. Minute; leaves scarcely 2 mm. long Large or small, but leaves always larger than above. 7. S. exiguum. Fruit longer than its pedicel 8. S. reniforme. Fruit shorter than its pedicel 9. S. Haastii. Smaller; leaves 5–10 mm.; fruit not much shorter than its pedicel Ditto, subsp. cyanopetalum. Larger; leaves –30 mm.; fruit 4 times shorter than its pedicel Ditto, subsp. Hookerianum. From the above it will be noticed that Domin adds an additional species (S. Colensoi) to those already known to inhabit the Dominion. It is based upon Hooker's Azorella trifoliolata, var. b, an obscure plant gathered by Colenso on crags at Titiokura, between Napier and Taupo, but which has not been observed since its first discovery more than sixty years ago. I have seen no specimens, and cannot express any opinion of my own as to the validity of the species; but as Dr. Domin has inspected the type in the Kew Herbarium, we may accept it on his authority. It should also be mentioned that Dr. Domin treats S. Haastii as a “collective species,” dividing it into the two subspecies—cyanopetalum, which he distinguishes by its smaller size, leaf-blades 5–10 mm. diameter, fruit not much shorter than its pedicel, and blue petals; and Hookerianum, which is much larger, the lamina of the leaf being –30 mm. diameter, and the fruit often only a quarter the length of the pedicel. Now, there is no doubt that forms of A. Haastii can be found agreeing with these characters (except the blue petals, which I have never observed, and doubt the occurrence of in a fresh state), but whether these represent permanent varieties, separated by constant characters, appears to me to be highly doubtful.

Azorella nitens, Petrie. Grassy places in the Cobb Valley, north-west Nelson; F. G. Gibbs! Not previously recorded to the north of the Clarence Valley. Aciphylla Hectori, Buch. McKinnon's Pass, near the base of Mount Balloon, forming hemispherical masses sometimes 2 ft. in diameter; F. G. Gibbs! I regret that Mr. Gibbs's specimens of this curious and little-known species do not show either good flowers or fruit. XXXIV. Araliaceæ. Pseudopanax ferox, T. Kirk. Tauroa (Reef Point), near Ahipara, Mangonui County; R. H. Matthews! A slight northward extension of the range of this fine plant, which, although it stretches throughout almost the whole length of the North and South Islands, is everywhere exceedingly local. XXXV. Cornaceæ. Corokia Cotoneaster, Raoul. Tasman Valley, Mount Cook district, ascending to 3,500 ft. altitude; T. F. C. Also in many other localities in the mountains of Nelson and Canterbury. XXXVII. Rubiaceæ. Coprosma rugosa, Cheesem. Arthur Valley and McKinnon's Pass, south-west Otago; F. G. Gibbs! Nertera setulosa, Hook. f. Vicinity of Marton; W. Townson! XXXVIII. Compositæ. Lagenophora pinnatifida, Hook. f. Tauroa (Reef Point), near Ahipara, Mangonui County; H. Carse! This is an altogether unexpected extension of the range of this fine species, which had not been previously collected to the north of the Kaipara River. Mount Starveall, Nelson, plentiful; Flora Valley and Cobb Valley, north-west Nelson; F. G. Gibbs! Celmisia Walkeri, T. Kirk. Abundant on both sides of McKinnon's Pass; H. J. Matthews! Celmisia rupestris, Cheesem. Very plentiful on most of the high peaks flanking the Cobb Valley, north-west Nelson, especially on Waingaro Peak and the Diamond Lake Range; F. G. Gibbs! Celmisia ramulosa, Hook. f. McKinnon's Pass, south-west Otago; H. J. Matthews! F. G. Gibbs!

Celmisia lateralis, Buch. Not uncommon on the mountains of north-west Nelson. Cobb Valley and adjacent mountains; Mount Cobb; Diamond Lake Range; mountains near the Heaphy River; F. G. Gibbs! Celmisia Gibbsii, Cheesem. Abundant on Waingaro Peak, north-west Nelson; F. G. Gibbs! Celmisia petiolata, Hook. f. Descends to sea-level in the sounds of the south-west coast of Otago; H. J. Matthews. The most abundant species in the Mount Cook district; T. F. C. Celmisia Petriei, Cheesem. Abundant on McKinnon's Pass and on the neighbouring mountains; H. J. Matthews! F. G. Gibbs! Celmisia linearis, Armstr. McKinnon's Pass; H. J. Matthews! Celmisia Hectori, Hook. f. An abundant plant on all the mountains in the Mount Cook district, forming extensive carpets in moist sheltered hollows; altitude, 4,500-6,000 ft.; T. F. C. McKinnon's Pass; H. J. Matthews! F. G. Gibbs! Celmisia Macmahoni, T. Kirk. On the precipitous face of Mount Rintoul, Nelson; F. G. Gibbs! The most western locality yet recorded. Haastia Sinclairii, Hook. f. Shingle-slopes on high peaks in the Mount Cook district; plentiful. Mount Ollivier, 5,500-6,000 ft.; T. F. C. Mount Kinsey, 5,000-7,000 ft.; P. Graham. Nun's Veil; E. Linton! Raoulia glabra, Hook. f. Mount Egmont; altitude, 4,500 ft.; W. Townson! The most northern locality yet recorded. Raoulia subulata, Hook. f. Mount Ollivier and other peaks in the Mount Cook district; altitude, 5,000-6,000 ft.; T. F. C.! Raoulia Buchanani, T. Kirk. McKinnon's Pass, south-west Otago; abundant; H. J. Matthews! F. G. Gibbs! Helichrysum Leontopodium, Hook. f. Grows most luxuriantly on Ben Nevis, eastern Nelson; F. G. Gibbs! Senecio saxifragoides, Hook. f. Kaikoura Peninsula, near the sea; H. J. Matthews! Senecio Kirkii, Hook. f. It is perhaps worth recording that this is an abundant plant on the lava-fields of the island of Rangitoto, often growing without protection

among the bare basaltic rocks. It is usually an inhabitant of the forest, frequently growing as an epiphyte in the forks of the rata and other trees, associated with such plants as Astelia, Freycinetia, Griselinia, Pittosporum cornifolium, &c. Senecio Monroi, Hook. f. Limestone rocks by the Mason River, North Canterbury; H. J. Matthews! Senecio revolutus, T. Kirk. McKinnon's Pass, south-west Otago; F. G. Gibbs! Senecio Adamsii, Cheesem. Abundant on Waingaro Peak, north-west Nelson; F. G. Gibbs! XLIII. Epacridaceæ. Leucopogon Richei, R. Br. Near the North Cape; E. Clarke! A most interesting discovery, of which I have given full particulars in another part of this volume. Archeria Traversii, Hook. f. McKinnon's Pass, south-west Otago; F. G. Gibbs! Dracophyllum Kirkii, Berggr. Cobb Valley and adjacent mountains, north-west Nelson; F. G. Gibbs! The most northern locality yet recorded. Dracophyllum uniflorum, Hook. f. Mount Holdsworth, Tararua Range; altitude, 4,000-4,500 ft.; W. Townson! XLIX. Loganiaceæ. Geniostema ligustrifolium, A. Cunn. Pepin Island, north of Nelson; F. G. Gibbs. This is the first record of the occurrence of this species in the Nelson Provincial District. L. Gentianaceæ. Gentiana Spenceri, T. Kirk. McKinnon's Pass, south-west Otago; F. G. Gibbs! LII. Convolvulaceæ. Ipomæa palmata, Forst. Abundant on the cliffs at Whangamumu in 1853; R. Mair. Mr. Hansen has sent me fine flowering specimens from the Tiritiri Island locality mentioned in Trans. N.Z. Inst., xl, 280. LIV. Scrophulariaceæ. Veronica elliptica, Forst. Titahi Bay, near Cape Terawhiti; B. C. Aston! This is the first recorded locality from the North Island.

Veronica Cockayniana, Cheesem. McKinnon's Pass, south-west Otago; H. J. Matthews! F. G. Gibbs! Veronica salicornioides, Hook. f. Abundant on the Cobb Valley, north-west Nelson, where it commonly attains a height of 6 ft.; F. G. Gibbs! Captain Dorrien Smith! Veronica Petriei, T. Kirk. Mountains above Lake Harris, Otago; H. J. Matthews! Veronica Cheesemanii, Benth. Occurs on all the higher mountains of eastern Nelson, as Mount Richmond, Mount Starveall, Ben Nevis, Gordon's Nob, &c.; F. G. Gibbs! Ourisia prorepens, Petrie. McKinnon's Pass, south-west Otago; H. J. Matthews! LV. Lentibulariaceæ. Utricularia protrusa, Hook. f. Waiharakeke Stream, near Morrinsville; P. H. Allen! Numerous specimens in full flower, being the first obtained in that condition since the original discovery of the species by Mr. Colenso in 1842. LXIV. Chenopodiaceæ. Rhagodia nutans, R. Br. Abundant on Mackay's Bluff, near Nelson; F. G. Gibbs! LXVI. Piperaceæ. Peperomia reflexa, A. Dietr. I am indebted to Mr. Gerald T. Williams, of Mokoiwi, near Tuparoa, East Cape district, for flowering specimens of this plant. He informs me that it is quite abundant in the valley of the Tuparoa River and its tributaries, extending to within four or five miles of the base of Mount Hikurangi. These valleys are warm and sheltered, and many coast plants like Corynocarpus, Myoporum, Entelea, &c., grow there with considerable luxuriance. Peperomia reflexa is found either on prostrate logs or on the trunks of living trees, sometimes as much as 30 ft. from the ground. According to Mr. Williams, it can often be seen clinging to the stems of Metrosideros florida and other woody climbers. There seems to be no reason why it should be confined to the East Cape district, but so far it has not been found in any other portion of the Dominion. LXX. Proeaceæ. Knightia excelsa, R. Br. I am indebted to Mr. A. Allison for specimens of a curious sport, in which the leaves are dichotomously forked near the tips, somewhat after the manner of certain “crested” ferns. They were gathered at Orakeikorako by the Upper Waikato River.

Branched Nikau-Palm.—Cheeseman.

Branched Nikau-Palm.—Cheeseman.

LXXV. Euphorbiaceæ. Poranthera microphylla, Brong. Cobb Valley, north-west Nelson; F. G. Gibbs! These are the first specimens I have seen from this portion of the Nelson Provincial District. LXXVI. Urticaceæ. Urtica ferox, Forst. Summit of the ranges between the Thames Valley and Katikati, a few miles south of Te Aroha; F. N. R. Downard! The most northern locality yet recorded. Only a few specimens were seen. LXXIX. Orchidaceæ. Pterostylis. An interesting paper on the fertilisation of the Australian species of this genus, written by Mr. Oswald H. Sargent, is printed in the Annals of Botany for April, 1909. Pterostylis foliata, Hook. f. I am indebted to Mr. Guthrie Smith for fresh specimens of this species, collected at Tutira Lake, Hawke's Bay. They show that it varies greatly in size, the specimens ranging from 4 in. to 18 in. in height. The leaves are rather flesh when fresh, and the reticulated veins are by no means obvious, except in dried specimens. A character that has not been previously mentioned is that the ovary and upper part of the peduncle are glandular-pubescent. LXXXIV. Palmaceæ. Rhopalostylis sapida, Wendl. & Drude. In the “Transactions of the New Zealand Institute” for 1906 (vol. xxxix, p. 447) I have briefly described and figured two curious branched specimens of the nikau-palm. Through the kindness of Mr. C. Dawes, of Kohukohu, Hokianga, I have received particulars of two more, photographs of which I reproduce herewith. One of the specimens has four branches springing almost from the same point; the other has three. Both trees are still growing in a patch of forest a few miles from Kohukohu. They are about 15 ft. in height, the diameter of the trunk not exceeding 9 in. (Plates XXVI and XXVII.) XCI. Cyperaceæ. Schœnus Carsei, Cheesem. Not uncommon in swamps by the Waitoa River, Thames Valley; P. H. Allen! Carex and Uncinia. A work of great importance to all followers of systematic botany has lately appeared in the shape of Kukenthal's “Monograph of the Cyperaceœ-Caricoideœ,” being heft 38 of Engler's “Pflanzenreich.” The memoir forms a bulky volume of 824 pages, is profusely illustrated with drawings and analytical figures, and contains full descriptions of all the species accepted by the author. Dr. Kukenthal has long been known as a most capable and reliable botanist, and his acquaintance with the subject is

unrivalled. Great expectations were therefore aroused as to the value and importance of his work, and it may at once be said that these have been fully realised. For the first time, botanists have been supplied with a systematic account of the Caricoideœ arranged with due regard to the affinities of the species, containing full and precise descriptions on modern lines, and with ample details of geographical distribution and other necessary factors. It appears to me that there are three reasons why New Zealand botanists are indebted to Dr. Kukenthal: First, for showing how far our species of Caricoideœ are allied to those found in other parts of the world, and what their relative positions are; second, in disposing of several dubious points of nomenclature that could only be authoritatively settled by reference to the botanical collections and libraries of the Northern Hemisphere; third, in giving the opinions of a botanist of wide training and experience, fully acquainted with the whole order, on the delimitation of the species. With respect to the first two of these, both of which are for the most part beyond the scope of inquiries that can be made in the Dominion, New Zealand botanists will gladly accept Dr. Kukenthal's conclusions. In regard to the third, which brings to the front the perennial difficulty of what constitutes a species—or, in other words, involves matters of opinion rather than matters of fact—the case is somewhat different; and I may be pardoned for presenting a few lines of explanation respecting some points on which I cannot quite, agree with the learned author. One of the chief merits of Kukenthal's memoir is that it is prepared on eminently safe and conservative lines. In this respect it is a pleasing contrast with some recent systematic work published in both Europe and America. But occasionally this conservatism appears to have been stretched too far. Taking the genus Carex, for instance, Kukethal admits 669 species. But as far back as 1885 Bentham estimated the number of valid species at considerably over 500, and since then several hundreds have been described. Mr. C. B. Clarke has alone published over 150, and Dr. Kukenthal more than 80; while other authors are responsible for a considerable number. Several recent estimates have placed the number of valid species at from 800 to 1000. With regard to the New Zealand species, Dr. Kukenthal enumerates 43. This is 10 less than the number adopted in my “Manual of the New Zealand Flora.” The difference has been caused by—(1) uniting C. appressa, C. virgata, and C. secta in one species under the name C. appressa; (2) by combining C. Forsteri, C. Cockayniana, and C. semi - Forsteri in the single species C. Forsteri; by merging C. Buchanani with C. lucida, C. Solandri with C. dissita, and C. rubicunda with C. Petriei; (4) by altogether omitting C. Sinclairii, C. plesiostachys, and C. ventosa. These are somewhat sweeping changes, involving a reduction of species that will hardly meet with the approval of New Zealand botanists as a whole. In the following notes I propose to discuss in some detail the chief alterations proposed. But before doing so I should like to say that no student of the New Zealand species of Uncinia and Carex can afford to be without a copy of Dr. Kukenthal's memoir. It is not only accurate and reliable, but contains much new matter, and many evidences of careful and painstaking research. Uncinia Sinclairii, Boott. Dr. Kukenthal considers this to be identical with the Fuegian U. macrolepis, Decne. et Hombr. in D'Urville Voy. au Pôle Sud, 3, t. b, f. A, and if so Decaisne's name will take precedence. I have not seen South American

specimens, but Hombron's plate, quoted above, so closely resembles the New Zealand species as to leave little doubt of the accuracy of the identification. The late Mr. C. B. Clarke referred U. macrolepis to U. Lechleriana, but that is a species with much longer spikes. Uncinia fusco-vaginata, Kukenth. In the Manual I referred this to U. purpurata as var. fusco-vaginata, and I still think that it must be considered as a very near ally of that plant, although possibly entitled to specific rank. Uncinia pedicellata, Kukenthal, n. sp. This is based upon a plant collected by Dr. Cockayne on Ruapuke Island, in Foveaux Strait. I have seen no specimens, but according to Kukenthal's description it is closely allied to U. australis, differing mainly in the reddish foliage, narrower spikes, obtuse glumes, and obsoletely nerved utricles, which are conspicuously longer than the glumes. In U. australis the utricles seldom exceed the glumes, and are usually distinctly nerved. (Since writing the above, I find from Dr. Cockayne's memoir on the botany of Stewart Island that he regards it as common in lowland forests on that island.) Uncinia australis, Pers. Kukenthal adopts the name of Uncinia uncinata for this; the oldest specific name being Carex uncinatus, Linn. f., published in 1781. If priority of publication is alone to be considered, the change of name is unavoidable; but it may be pointed out that the new combination comes very near to being an infraction of article 55, section 2, of the Vienna rules of botanical nomenclature, under which it is provided that specific names shall be rejected when they merely repeat the generic name. Uncinia riparia, R. Br. The circumscription adopted by Kukenthal of this variable species agrees in the main with that given in the Manual, with the exception that he also includes Boott's U. Hookeri, from the Auckland and Campbell Islands, distinguishing it as var. Hookeri. When preparing the account of the genus given in the Manual I had no good material of U. Hookeri, and in default of such followed Hooker (Handb. N.Z. Fl. 310) in placing it under Raoul's rupestris. Since then I have received an excellent series of specimens collected by Mr. B. C. Aston, and can fully confirm Kukenthal's identification. Uncinia nervosa, Boott. Dr. Kukenthal considers that the New Zealand plant referred by me to this species is identical with Raoul's U. rupestris. I regret that I am unable to agree with this opinion, which is not shared by any other worker in the genus. I take it that U. nervosa naturally falls into the neighbourhood of U. compacta, whereas the true rupestris of Raoul, judging from his plate and description, is closely allied to U. riparia. This view was also held by Mr. C. B. Clarke, with whom I had some correspondence on the subject. In all probability Raoul's rupestris should be referred, as a variety, to U. riparia, and be placed in close association with var. Hookeri. Carex teretiuscula, Good. Kukenthal adopts the older name of C. diandra, Schrank, published in 1782, whereas C. teretiuscula did not appear until 1794.

Carex appressa, R. Br. Kukenthal unites with this species both C. virgata and C. secta. There is much to be said in favour of this, for it cannot be denied that intermediate forms exist between the three plants, and also between C. appressa and the northern C. paniculata. In my revision of the New Zealand Carices, published in 1884, I therefore followed Baron Mueller and Mr. Bentham in placing all three under C. paniculata. But, as explained in the Manual, further consideration has led me to alter this opinion, and to keep the species separate. After all, the differences between the three plants are quite as well marked as those between several species of Carex admitted by most authors, and for the sake of consistency it appears best to uphold their distinctness. Carex, stellulata, Good. In the Manual I followed Bentham and other well-known authors in adopting the name of C. echinata, Murr., for this species. But the late Mr. C. B. Clarke definitely proved that Murray's plant was identical with one of the forms of C. muricata, and Dr. Kukenthal has arrived at the same conclusion. Under these circumstances, botanists are pretty well agreed in taking up the next-oldest name of C. stellulata, Good. Carex flava, Linn. Kukenthal places the New Zealand form of this species under C. Oederi, Retz, as var. cataractœ. C. Oederi mainly differs from C. flava in the smaller size and shorter straight beak to the utricle. It is little more than a matter of taste as to whether the two species should be kept up or combined. Carex Buchanani, Berggr. Kukenthal reduces this to C. lucida, as var. Buchanani, a view with which I am unable to concur. C. Buchanani differs markedly in the much more strict and erect habit, in the semi-terete and much more coriaceous leaves, in the pale-coloured glumes, and in the plano-convex utricles, which are sharply serrate above, and have a much longer and more deeply bidentate beak. Not only is the aspect of the plant most distinct, so that it can be recognised at a glance, but its characters appear to me to be fairly constant. Dr. Kukenthal remarks that it is connected with C. lucida by numerous passage forms; but I have not myself met with specimens that could be so described. I feel sure that all New Zealand botanists familiar with the plant will repudiate any proposal to associate it with C. lucida. Carex rubicunda, Petrie. Kukenthal refers this to C. Petriei, as var. rubicunda. The late Mr. C. B. Clarke suggested that it should be placed with Berggren's C. cirrhosa. I have alluded to its affinity with C. Petriei in the Manual, and suspect that its reduction to that species will be maintained. Carex Dallii, T. Kirk. Probably a mere form of C. Petriei, Cheesem. Carex plesiostachys, C. B. Clarke. Not taken up by Kukenthal. I am now inclined to consider it a form of C. comans, Berggr.

Carex Solandri, Boott. Reduced to C. dissita by Kukenthal, as var. Solandri. No doubt there are transition states between the two plants—in the river valleys of north-west Nelson, for instance, where C. Solandri is particularly abundant, such may be commonly observed. But what I take to be the typical form is a much larger and more slender plant, with longer culms (sometimes over 5 ft. in length) that are often prostrate in fruit. The male spikelets (which are seldom more than one in C. dissita) are often as many as four or five; and the lower female spikelets are usually compound, with long nodding peduncles. These characters recede so much from the ordinary state of C. dissita that I cannot refuse to regard the two plants as distinct. Carex dissita, Sol., var. ochrosaocus, Cheesem. Kukenthal unites this with the Norfolk Island C. Neesiana, placing it as var. Neesiana of C. dissita. A Norfolk Island specimen of C. Neesiana kindly forwarded to me many years ago by Baron Mueller does not seem to support this view; but the material is too scanty to form the basis of an opinion. Carex ventosa, C. B. Clarke. This is simply quoted by Kukenthal as a species unknown to him. Carex Cockayniana, Kukenth. Dr. Kukenthal now refers this to C. Forsteri. It is very much a matter of personal taste as to whether the two plants should be combined or kept apart. My own opinion is that the claims of C. Cockayniana to be regarded as distinct are quite as good as those of several species universally accepted. Carex semi-Forsteri, C. B. Clarke. This also is reduced to C. Forsteri by Kukenthal, but I feel sure that his material has been too scanty. As I have pointed out in the Manual, it has much of the habit of C. Forsteri, but the terminal spikelet is invariably largely female at the top, whereas it is wholly male in C. Forsteri. I reproduce here-with Mr. Clarke's original diagnosis, communicated to me in 1901. “C. semi-Forsteri, n. sp.: Spicis 5–9, terminali pro magna parte fœminea, basi mascula, usque ad 4–6 cm. longis; glumis fœmineis (arista inclusa) cum utriculi subæquilongis; utriculis anguste ellipsoidei, trigonis, viridibus aut fusco-viridibus, 3 ½-4 mm. longis, utrinque angustatis, glabris, 16-nervatis; rostro cum utriculi ½-¾ parte fere aequilongo dentibus 2 linearibus patulis. New Zealand, Stephenson, n. 44, partim; Hooker, n. 4189; Colenso, n. 1622. Kermadec Islands. Shakespear, n. 1662.” XCII. Graminaceæ. Microlæna polynoda, Hook. f. Tauroa (Reef Point), near Ahipara, Mangonui County; R. H. Matthews! Not previously recorded to the north of Whangarei Heads, where I collected it many years ago. XCIII. Filices. Lomaria nigra, Col. Maungataniwha Ranges, Mangonui County; H. Carse. Not previously recorded from the north of Whangarei.