Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 49, 1916
This text is also available in PDF
(371 KB) Opens in new window
– 47 –

Art. III. — On a New South Polynesian Palm, with Notes on the Genus Rhopalostylis Wendl. et Drude.

[Read before the Auckland Institute, 13th December, 1916; received by Editors, 30th December, 1916; issued separately, 28th June, 1917.]

In a visit paid in August, 1887, by Mr. T. F. Cheeseman to the Kermadec Islands Group, lying in latitude 30° S. in the South Polynesian Sea, thère was found growing very abundantly on Sunday Island, one of the group, a fine palm, which at the time of the visit was not bearing flowers, but was loaded with large bunches of nearly mature fruits. Specimens of this palm, supposed to belong to Rhopalostylis Baueri, or to a variety of it, were sent to Kew from the herbarium of the Auckland Museum, and transmitted to me in Florence some time ago. These specimens (wanting the flowers, and only a few detached fruits being available for study) were not sufficient for a rigorous specific identification. Quite recently, however, having received, also from Kew, portions of the flowering spadix of the same palm, its affinity to Rhopalostylis Baueri and R. sapida was rendered quite clear; but yet it proved to be a palm specifically distinct from either, on account especially of the shape and other peculiarities of its fruits. The flowers of the Kermadec palm were derived from a plant that as Mr. Cheeseman informs me, flowered recently in Auckland from seeds originally collected in those islands.

It is for me a doubly pleasant task to distinguish with the name of its discoverer this new palm from the temperate Australian regions, as I hope that through Mr. Cheeseman's exertions it may soon become a new inmate in our Mediterranean gardens, as have the two already-known Rhopalostylis; and also in memory of the fine trip made on the 12th March, 1878, under Mr. Cheeseman's guidance, to the Titirangi Ranges, near Auckland, where I was able to admire Rhopalostylis sapida, the nearest ally of the Kermadec palm, in its native home.

Rhopalostylis Cheesemanii Becc. n. sp.

A fine palm with a straight solitary stem, attaining 60 ft. (about 18 m.) in height, and producing large bunches of fruits 2 ft. (60 cm.) in diameter. The primordial leaves have the blade 12–20 cm. long, deeply bifid, or formed by two lanceolate slightly falcate leaflets, united by their bases; their petiolar part is elongate, and besprinkled with appressed rusty scales.

One leaf from a young plant is pinnate, and carries several leaflets on each side of the rhachis, upon which they are attached by means of a broad base; such leaflets are narrowly lanceolate-subfalcate, and very acuminate to a capillary tip; their lower surface is besprinkled with very minute brown scales, considerably smaller than those of the petiole and rhachis. The leaflets of the adult plant are ensiform, have the upper end gradually acuminate and very slightly falcate, and the lower end slightly narrowed and very slightly curved (sigmoid); the bases are relatively broad and the margins not reduplicate; are green above, subglaucescent beneath, and very distinctly 3-costulate; the costae are especially prominent on the lower surface, and slightly scaly-furfuraceous there, especially near the base; the side costae are not so strong as the middle one, and become feebler from the middle upwards; the secondary nerves are slender, and barely

– 48 –

distinguishable from the tertiary ones, which are rather sharp, so as to render the two surfaces (in the dry specimens) distinctly striate; transverse veinlets obsolete; the margins acute. The single leaflet (from an adult plant) seen by me was 75 cm. long, and 4 cm. broad at about its middle. The spadix is large and much branched; its flowering branchlets are thickish, angular, and deeply and closely scrobiculate along 6 vertical series; the scrobiculi are shaped like a swallow's nest, and have their lower margins truncate, but with a small acute point at their middle. The flowers have their bases immersed in the scrobiculi, are arranged in glomerules of 3 (the central a female, and the side ones male, as usual) almost to the end of the flowering branchlets. At the time of the anthesis of the male flowers the female flowers are rather well evolute, but not yet ready for impollination. The male flowers are asymmetrical, 7 mm. long, lanceolate-acuminate, and somewhat falcate at apex; the calyx is formed by 3 free sepals, thinly membranous, lanceolate-acuminate, very briefly imbricate at the base; the corolla is divided down to the base into 3 concave ovate-lanceolate acuminate segments, one-third, or even less, longer than the sepals; are cartilaginous, strongly striately veined outside; the stamens are 6, have the filaments linear with the apex subulate and inflected; the anthers are erect when in the bud and versatile later, are attached about their middle, linear-sagittate, their apices and their auricles are obtuse, have the cells disjointed in their lower third part, and dehisce laterally. The rudimentary ovary is conspicuous, subclavate, almost as long as the stamens (in the unopened flower). The female flowers, at the time of the flowering of the male flowers, are broadly conical, acute, 4 mm. broad (probably slightly larger later); sepals thinly cartilaginous, very broadly imbricate, suborbicular, briefly apiculate; the corolla slightly longer than the calyx; petals similar to the sepals, broadly imbricate, having only the short apices valvate; sepals and petals minutely ciliolate; ovary oblong, conical above, terminated by 3 trigonous, acute, at first connivent, then spreading and recurved, stigmas; ovule relatively large, attached laterally in the highest part of the cell. Staminodes 3, dentiform.

Fruit globular; in the dry condition it is 11–13 mm. in diameter, and has a terminal discoid-mammillaeform apiculus, showing in its centre the minute remnants of the stigmas; the pericarp is in its totality 1 mm. in thickness; its epicarp is very thin, and very finely l ineolate by very minute fibres (sclerostomes); the mesocarp is apparently very slightly fleshy in the fresh state, and contains 4–5 rows of flattened rigid fibres; the endocarp is very thin, pellicular-cartilaginous, tightly enveloping the seed.

Seed globose, 10.5–11 mm. in diameter, the outer coat is light-coloured and almost polished, and faintly marked by the nearly simple vascular branching of the raphe; several of the branches descend spirally and posticously from the summit, and only very few descend the sides. The hilum is conspicuous, broad and suborbicular on the vertex of the seed, it narrows gradually downward to its base, in close proximity to the embryo. Albumen very hard, homogeneous. Embryo exactly basal. Fruiting perianth very slightly concave, or almost explanate.

Rhopalostylis Cheeseman ii appears to be more closely allied to R. sapida, the seed of which has, like that of R. Cheeseman ii, a light-coloured and polished surface, than to R. Baueri, of which the seed has a brown and dull surface. R. Cheeseman ii is, however, a much taller plant than R. sapida, and in that respect approximates more closely to R. Baueri. But this new species differs from both R. sapida and R. Baueri in the globular form of its fruit.

– 49 –

Rhopalostylis Wendl. et Drude.

Wendl. et Drude in Linnaea, xxxi (1875), 180, t. 1, f. 2 (the ovary); Drude in Bot. Zeit., 1877, t. 6, f. 18–21 (anatomy of the ovary); Benth. et Hook., Gen. Pl. iii, 890. Kentia and Areca of several authors.

Arboreous unarmed palms. Stem marked with annular rings or scars left by the dead fallen leaves. Leaves pinnate, furnished with a long basal sheathing part. Leaflets narrow and elongate, acuminate, slightly falcate at apex, 3-costulate, and with several distinct secondary nerves; their margins acute (not thickened by a marginal nerve). Spadix infrafoliar, briefly stalked, enveloped by 2 complete similar papyraceous or membranous deciduous spathes; the outermost 2-winged. Flowering branches thick, closely and deeply scrobiculate. Flowers in glomerules of 3 (one male between two females) from the base up nearly to the apices of the branches. Male flowers opening somewhat before the female ones, asymmetrical; sepals subulate, entirely free, slightly imbricate at the base; petals valvate, almost entirely free, more or less obliquely lanceolate, or ovate. Stamens 6, free, their filaments filiform, inflected at apex; anthers linear or linear-sagittate, versatile, dorsifixed. Rudimentary ovary conspicuous, columnar or subclavate. Female flowers globose-ovoid; sepals br cadly imbricate; petals slightly longer than the sepals, convolutive-imbricate in their very broad basal part, and having very brief valvate apices; ovary ovoid, unicelled; ovule attached all along one side of the cell; stigmas triangular, short, at first connivent, later recurved; staminodes minute. Fruit small, globular-ovoid or elliptical, symmetrical, bearing the remains of the stigmas exactly on the apex. The whole pericarp thin; the epicarp furnished with very minute linear fibres (sclerosomes); the mesocarp very slightly fleshy and furnished with a few rows of rigid fibres; the endocarp very thin, cartilaginous. Seed globular, ovate, or ellipsoid, attached all along one side of the endocarpal cavity, and marked with a conspicuous hilum, running from the apical chalaza down to the base, in close proximity to the embryo; vascular branches of the raphe very slightly anastomosing, starting mostly from the chalaza, running down the ventral side of the seed, and only a very few starting from the sides of the raphe; albumen homogeneous. Embryo basal.

The generic name Rhopalostylis was proposed by H. Wendland and Oscar Drude for the two palms of the Southern Hemisphere known by the old names of Areca sapida and A. Baueri. The genus Rhopalostylis is now represented by three species, which form a small but well-characterized group: it is somewhat related to the true genus Kentia, as understood by me (see Webbia, iv, 1913, p. 143); but perhaps even more to the genus Actinokentia of U. Dammer (Kentiopsis divaricata A. Brongn., a palm indigenous to New Caledonia), on account of its ovule attached all along one side of its cell; of its symmetrical fruit, having the remains of the stigmata exactly on the apex; and of its seed being marked with a conspicuous hilum, running from apex to base, and with homogeneous albumen and basilar embryo.

Rhopalostylis sapida from New Zealand and R. Baueri from Norfolk Island are two palms very well characterized and distinct, but have a some-what uncertain synonymy, and have been frequently confounded, or considered as representing one species only. Martius himself has apparently, in his description of Areca sapida, cumulated the characteristics of the palm of New Zealand with those of the Norfolk Island palm; but the fine

– 50 –

plates 151 and 152 of his Historia naturalis Palmarum, reproduced from Bauer's drawings, represent only Areca (Rhopalostylis) Baueri.

The name of Areca sapida Solander appears for the first time, I believe, in the work of Georg Forster, De Plantis esculentis insularum Oceani australis Commentatio botanica, 1786, p. 66, n. 35; but apparently Solander has never given a description of that palm, and Forster evidently considers the New Zealand palm the same as that growing in Norfolk Island, as he writes of A. sapida, “Reperitur spontanea in Nova Zelandia usque ad aestuarium Charlottae reginae, et frequens in Norfolciae insula deserta.”

H. Wendland, in the “Enumeration of all Known Palms,” published in the work of Oswald de Kerchove de Denterghem, Les Palmiers, considers as Rhopalostylis sapida Wendl. et Drude only that which goes by the horticultural name of Kentia sapida, figured in the Botanical Magazine in plate 5139, under the name of Areca sapida. To Rhopalostylis Baueri he refers A Baueri Hook, of the Botanical Magazine (plate 5735), Areca sapida Sol., Kentia sapida Mart., A. Banksiia A. Cunn., and Seaforthia robusta Hort.

I have not now the leisure to clear up the exact synonymy and to adduce the entire relative literature of these two palms. It is sufficient for me at present to establish the fact that these two palms are quite distinct, and that (1) Rhopalostylis sapida is now generally considered to be the species growing in New Zealand; (2) the palm indigenous to Norfolk Island is R. Baueri. Of the first I collected specimens myself in its native land; of the second I have examined fruiting specimens forwarded to me many years ago by my lamented friend Baron Ferdinand von Mueller, and collected in Norfolk Island by Mr. Isaac Robinson.

The three species of Rhopalostylis are characterized as follows:—


Rhopalostylis sapida Wendl. et Drude in Kerch, Les Palmiers, p. 255. Areca sapida (Sol.) Hook. f., Fl. Nov. Zel. i, 262, t. 59, 60; Bot. Mag. t. 5139.

A middle-sized palm, 5–6 m. high. Spadix 30–40 cm. long or less. Fruit ovoid or ovoid-elliptical, 12–14 mm. long, 7–8 mm. through. Seed ovoid, with light-coloured polished surface and marked with a broadly linear hilum. Fruiting perianth cupular-campanulate, 6–7 mm. high.

Hab.—New Zealand, from 35° to 38° S latitude.


Rhopalostylis Baueri Wendl et Drude in Linnaea, xxxix (1875), 180 et 234, t. 1, f. 2. Areca Baueri Hook. f. in Fl. Nov. Zel. i, 262, in obs.; Bot. Mag. t. 5735. A. sapida Mart, Hist. Nat. Palm. iii (partly, as to description), t. 151–52.

A tall palm. Spadix 60–90 cm. long. Fruit ovoid-elliptical, 15–17 mm. long; 12 mm. through. Seed with dull-brown surface; hilum linear but slightly narrowing towards the base. Fruiting perianth cupular-campanulate, 6–7 m. high.

Hab.—Norfolk Island.


Rhopalostylis Cheesemanii Becc. n. sp.

A tall palm, about 18 m high. Spadix in fruit about 60 cm. in diameter. Fruit globose, 11–13 mm. in diameter. Seed globose, with light-coloured polished surface; hilum broad and suborbicular above, narrowing considerably towards the base. Fruiting perianth almost explanate.

Hab.—Kermadec Islands