The New Zealand Species of Metrosideros with a Note on Metrosideros collina (Forst) Gray.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 59, 1928, Page 419

The New Zealand Species of Metrosideros with a Note on Metrosideros collina (Forst) Gray. By W. R. B. Oliver, M.Sc, F.N.Z. Inst., Director Dominion Museum, Wellington. [Read before the Wellington Philosophical. Society, 24th June, 1928; received by Editor, 31st July, 1928; issued separately, 31st August, 1928.] Plate 67. The purpose of the present paper is merely to revise the nomenclature of the New Zealand species of Metrosideros. Discussions on affinities and distributions are therefore included only so far as they have a bearing on the name to be applied to the species. Investigations into the nomenclature of the genus revealed an amazing state of confusion, for, as will be seen by what follows, no fewer than seven of the eleven names used in Cheeseman's Manual of the New Zealand Flora will require to be changed. Unfortunately, in two cases the names are transferred to different species within the genus. It will be convenient to compare in tabular form the names herein proposed with those in Cheeseman's Flora. Cheeseman's Manual. Present Paper. M. florida M. scandens M. lucida, M. umbellata M. Parkinsonii M. Parkinsonii M. albiflora M. albifiora M. diffusa M. carminea M. hypericifolia M. diffusa M. Colensoi M. Colensoi M. scandens M. perforata M. robusta M. robusta M. robusta. var. intermedia × M. subtomentosa M. tomentosa M. excelsa M. villosa M. kermadecensis Of the above changes M. perforata has already been noticed recently by the author and by Cockayne and Allan (Trans. N.Z. Inst., vol. 56, pp. 5, 27, 1926), while M. scandens (= M. florida) was recorded by Druce in 1917. In the case of the hybrid, M. subtomentosa, this name is adopted instead of intermedia under the authority of a rule which states that the first name used in a specific sense must stand. The rule is obviously not in harmony with the law of priority. In order to verify the results of my researches, I applied to the Department of Botany of the British Museum where the types of the Forsters are preserved. Mr. A. W. Exell kindly took the matter up, fully investigated the problems I placed before him, and drew up a synonymy which not only confirmed mine but added several references not available to me. Thanks to Mr. Exell, therefore, I believe the names now put forward are in accordance with the International Rules of Botanical Nomenclature, and the authori-

ties given are correctly stated. Mr. Exell kindly examined the type specimens in the British Museum and in the Linnean herbarium. Metrosideros scandens (Forst.) Druce. Leptospermum scandens Forst. Char. Gen. 72, pl. 36, 1776. Melaleuca florida Forst. Fl. Ins. Austr. Prodr. 37, 1786. Metrosideros fulgens Sol. ex Gaertn. Fruct. Sem. Pl. 1, 172, pl. 34, f. 7, 1788. M. florida Smith, Trans. Linn. Soc. 3, 269, 1797. M. speciosa Colenso, Trans. N.Z. Inst., 22, 463, 1890. M. aurata Colenso, Trans. N.Z. Inst., 23, 385, 1891. M. scandens Druce, Rep. Bot. Exch. Cl. Br. Isles, 1916, 635, 1917. Not M. scandens Sol. ex Gaertn. Fruct. Sem., pl. 1, 172, 1788. The Forsters introduced Leptospermum scandens twelve years before Gaertner founded Metrosideros scandens for a different species. As Gaertner's specific name is not valid, being pre-occupied by Forster's earlier perforatum, Forster's name scandens becomes available for the present species. Metrosideros umbellata Cav. Melaleuca lucida Forst. Fl. Ins. Austr. Prodr. 38, 1786. Metrosideros umbellata Cav. Ic. Desc. Pl. 4, 20, pl. 337, 1795. Melaleuca umbellata Raeuseh, Nomencl. Bot. ed. 3, 143, 1797. Metrosideros lucida A. Rich. Voy. Astrol. Bot. 333, 1832. Agalmanthus umbellatus Hombr. & Jaeq. ex Decne. Voy. Astrol. et Zel. 78, 1853. Not Melaleuca lucida Linne. f. Suppe. 342, 1781. Forster's specific name lucida is not valid for this species on account of an earlier homonym of Linne f. (= M. collina.). Cavanille's name umbellata therefore comes into use. Metrosideros Parkinsonii Buch. Metrosideros Parkinsonii Buchanan, Trans. N.Z. Inst., 15, 339, 1883. Metrosideros albiflora Sol. ex Gaertn. Metrosideros albiflora Sol. ex Gaertn. Fruct. Sem. Pl. 1, 72, pl. 34, f. 11, 1788. Melaleuca albiflora Raeunch, Nomencl. Bot. ed. 3, 143, 1797. Metrosideros diffusa A. Cunn., Ann. Nat. Hist. 3, 114, 1839. Hook f. Ic. Pl. pl. 569, 1843. Not M. diffusa Smith. Trans. Linn. Soc. 3, 268, 1797. Metrosideros carminea W. R. Oliv. new name. Metrosideros diffusa Hook, f., Fl. Nov. Zel. 1, 67, 1853. Not. M. diffusa Smith, Trans. Linn. Soc. 3, 268, 1797, nor Melaleuca diffusa Forst. Fl. Ins. Austr. Prodr. 37, 1786. This species has been known under the name of M. diffusa since the publication of Hooker's Flora Novae Zelandie in 1853. It is not however the Melaleuca diffusa of Forster, better known as M. hypericifolia, nor the Metrosideros diffusa of Smith, which is a Polynesian species (M. collina), nor yet M. diffusa of Cunningham, which

Plate 34 of Gaertner's De Fructibus et Seminibus Plantarum, 1788.

is M. albiflora. The name diffusa has thus been applied to four different species of Metrosideros. Forster did not land at any point in the North Island, so that his name could not rightly be applied to the present species which is confined to the northern portion of that island. Metrosideros diffusa (Forst.) W. R. Oliv. new comb. Melaleuca diffusa Forst. Fl. Ins. Austr. Prodr. 37, 1786. Metrosideros myrtifolia Sol. ex Gaertn. Fruct. Sem. Pl. 1, 172, pl. 34, f. 12, 1788. M. hypericifolia A. Cunn. Ann. Nat. Hist. 3, 114, 1839. M. subsimilis Col. Trans. N.Z. Inst., 12, 361, 1880. Cheeseman in his Illustrations of the New Zealand Flora, pl. 48, 1914, discusses the identity of Melaleuca diffusa, of Forster and concludes that “in all probability his plant is identical with the species usually known as M. hypericifolia.” In order to verify this I sent to Mr. Excell specimens illustrating the range of variation of M. hypericifolia. Mr. Excell kindly compared the specimens with Forster's type and reported that they belonged to the same species. Metrosideros diffusa of Smith, is, as shown under M. collina not the same as Forster's Melaleuca diffusa. Metrosideros myrtifolia of Gaertner is as shown by his plate reproduced herewith the same species as Forster's Melaleuca diffusa. Metrosideros Colensoi Hook. f. Metrosideros Colensoi Hook. f. Fl. Nov. Zel. 1, 68, 1853. M. pendens Colenso, Trans. N.Z. Inst., 12, 360, 1880. Metrosideros perforata (Forst.) Rich. Leptospermum perforatum Forst. Char. Gen. 72, 1776. Melaleuca perforata Forst. Fl. Ins. Austr. Prodr. 37, 1786. Metrosideros scandens Sol. ex Gaertn. Fruct. Sem. Pl. 1, 172, pl. 34, f. 10, 1788. M. perforata A. Rich. Voy. Astrol. Bot. 334, 1832. M. buxifolia A. Cunn. Ann. Nat. Hist. 3, 111, 1839. M. vesiculata Colenso, Trans. N.Z. Inst., 16, 327, 1884. M. tenuifolia Colenso, l.c. 24, 387, 1892. Not Leptospermum scandens Forst. Char. Gen. 72, 1776. I have myself examined Forster's type of Melaleuca perforata, and found it to agree with the plant known to New Zealand botanists as Metrosideros scandens. Gaertner's Metrosideros scandens, which should not be confused with Forster's Leptospermum scandens, is shown, by Gaertner's plate reproduced herewith, to be the present species. Metrosideros robusta A. Cunn. Metrosideros robusta A. Cunn. Ann. Nat. Hist. 3, 112, 1839. M. florida Hook. f. Bot. Mag. pl. 4471, 1849 (not Forst.) M. robusta var. retusa Kirk, Students Fl. N.Z. 162, 1899. × M. subtomentosa Carse. Metrosideros robusta var. intermedia Kirk. Students Flora, 162, 1899. × M. subtomentosa Carse, Trans. N.Z. Inst., 57, 92, 1927.

This hybrid was first described by Kirk as a variety of M. robusta. Kirk's specimens were collected on Rangitoto Island where both M. robusta and M. excelsa are found. There are also in his herbarium similar specimens from Great Omaha, where this form is known, to bushmen as the “inland pohutukawa.” Plants of different appearance, but intermediate between M. robusta and M. excelsa have been collected at Titirangi, Mount Tarawera, and Taupo Lake, in every case localities where both M. robusta and M. excelsa are found, so that their hybrid origin is practically certain. Metrosideros excelsa Sol. ex Gaertn. Metrosideros excelsa Sol. ex Gaertn. Fruct. Sem. Pl. 1, 172, pl. 34, f. 8, 1788. M. tomentosa Rich. Voy. Astrol. Bot. 336, pl. 37, 1832. My attention was first drawn to the identity of M. excelsa through the examination of specimens in the Dominion Museum herbarium, collected by Banks and Solander during Cook's first voyage. The specimens were labelled Metrosideros excelsa. In order to verify the identification I applied to Professor J. Arthur Harris, of the University of Minnesota, who kindly sent me a typed copy of the Gaertner's description of M. excelsa and a photographic reproduction of his plate 34. The description runs: “Calyx tomentosus, quinquedentatus, capsulae ad medium usque adnatus. Capsula ovata, extra calycem prominens, pubescens, trilocularis.” The plate is reproduced herewith. There can be no doubt that Gaertner's plant, which was collected by Banks and Solander, is that commonly known as M. tomentosa. Mr. Exell is in agreement with me. Metrosideros kermadecensis W. R. Oliv. new name. Melrosideros polymorpha Hook, f. Handb. N.Z. Fl, 73, 1864, (not Gaud). M. villosa Kirk, Students Fl. N.Z. 163, 1899 (not Sm.). This species has hitherto been referred to the very complex group known collectively under the name Metrosideros collina (Forst.) Gray, or one of its numerous aliases, and which is found throughout the Pacific from Fiji to Tahiti and the Hawaiian Islands. Leaving aside the Kermadec species, M. collina may be divided into two groups, one confined to the Hawaiian Islands and consisting of 11 varieties, and another found in the South Pacific Islands and separable into three varieties. In the latter group I have seen no form corresponding to the Kermadec plant. Throughout its range it is for the most part either entirely glabrous or it has the inflorescence, young leaves, and to a greater or lesser extent, the under-surface of the leaves, canescent or tomentose. Moreover the leaves are typically elliptic, moderately thin and have flat margins. The Kermadec species has oblong coriaceous leaves with recurved margins and is densely tomentose below. The inflorescence and young leaves are also densely covered with white tomentum. In both M. collina and M. kermadecensis the flowers are sessile. To find a form within the complex collina at all comparable with the Kermadec species it is necessary to go to Hawaii where the variety incana approaches it. But this variety differs from M. kermadecensis in its larger leaves with longer petioles and pedicellate flowers. Moreover it is generally an erect tree with grey bark, and

with the capsule slightly exserted or immersed in the calyx-tube, from which it is nearly free. The variety incana is said to come close to the variety typica which has cordate leaves, and if these and the other named varieties are properly included in a single species then its genetic constitution is entirety different from that of the South Pacific group and still more so from the Kermadec species. M. kermadecensis is really a close ally of M. excelsa of New Zealand, differing mainly in its smaller oblong leaves and smaller flowers. Note on Metrosideros collina (Forst) Gray. As the nomenclature of this species is involved in that of some of the New Zealand species, the following remarks are worthy of record in this place. At my request Mr. Exell kindly examined for me the type of Melaleuca lucida in the Linnean herbarium. He reports that it agrees well with Leptospermum collinum of Forster. It was collected by Sparrman and given to Linné. Though it was said to have come from New Zealand it was probably collected at Tahiti. As there has been a considerable amount of confusion through the application of the name diffusa to New Zealand species of Metrosideros, I applied to Dr. Daydon Jackson for particulars of the specimens in Smith's herbarium. I received back extracts of Smith's account in Volume 3 of the Transactions of the Linnean Society, and a tracing of the specimen in Smith's herbarium with a copy of the label attached thereto. The tracing might well be M. collina. Smith in his account states that he had seen the specimen given by Sparrman to Linné, and then synonymizes Melaleuca lucida Linn, with his diffusa. But M. lucida Linn, is as above stated, Metrosideros collina. The label of Smith's specimen reads “Otateite, Nelson. Herb. Banks, 1797.” Otateiti is a form of Tahiti. “Nelson” must be a late addition as the place so called in New Zealand would not be named for more than forty years after Smith's account was published. Although Smith uses Forster's name diffusa he evidently had not seen Forster's type, for he states that “I have only seen one specimen which was given to Linneaus by Dr. Sparrman” (that is, M. lucida Linn.). From this it is clear that Smith's Metrosideros diffusa is identical with Leptospermum collinum of Forster. Finally, Hooker and Arnott identify Tahiti specimens with Smith's diffusa. The synonymy of the Hawaiian group, which I recommend it would be convenient to call M. polymorpha Gaud, as was done by Gray, is given fully by Rock in The Ohia Lehua Trees of Hawaii. The following synonymy covers the South Pacific Group, all the names except vitiensis being founded on Tahiti specimens. Leptospermum collinum Forst. Char. Gen. 72, pl. 36, 1776. Melaleuca lucida Linn, f. Suppl. Pl. 342, 1781. M. villosa Linn. f. l.c. M. aestuosa Forst. Fl. Ins. Austr. Prodr. 38, 1786. Metrosideros spectabilis Sol. ex Gaertn. Fruct. Sem. Pl. 1, 172, pl. 34, f. 9, 1788. M. villosa Smith, Trans. Linn Soc. 3, 268, 1797. M. diffusa Smith l.c. Hook. & Arn. Bot. Beech. Voy. p. 63, 1841 (not Forst.). M. villosa var. glaberrima Bertero ex Guill. Zeph. Tait. p. 57. M. collina Gray, Bot. U.S. Expl. Exp. 558, pl. 68, 1854. M. collina var. vitiensis Gray l.c. p. 559. M. tahitiensis Decne., Bot. Voy. Venus, p. 30, 1855. Nania collina O. Ktze, Rev. Gen. Pl, 1, 242, 1891.