Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 56, 1926
This text is also available in PDF
(999 KB) Opens in new window
– 21 –

Notes on New Zealand Floristic Botany, including Descriptions of New Species, &c. (No. 4).

[Read before the Wellington Philosophical Society, 27th August, 1924; received by Editor, 31st December, 1924; issued separately, 6th March, 1926.]

44. * Acaena microphylla × Sanguisorbae.

A few seeds were sown in 1923 by L. C. of a supposed hybrid as above collected by him in the Dart Valley (Fiord Botanical District). The seed germinated well, and the seedlings betray their hybrid origin in the colour of the leaves (brown, pale green, darker green, slate-colour, &c.) and the size and shape of the leaflets. So far none have flowered. It is possible that one of the parents may be Acaena inermis, while it is not possible to say what particular variety of A. Sanguisorbae was the other parent.

45. Acaena Sanguisorbae Vahl. var. sericei-nitens Bitter.

This is the common upland Acaena of the South Otago Botanical District, and it extends into the adjacent part of the Fiord Botanical District. Seed collected by L.C. in the neighbourhood of Lake Wakatipu in 1923 has produced a number of seedlings most closely resembling one another, showing that the variety is a true microspecies (Jordanon).

46. Astelia nervosa Banks et Sol. ex Hook. f.

This aggregate species includes several distinct varieties. We here segregate the broad-leaved lowland swamp-plant with stout male inflorescence from the common narrow-leaved forest-plant with slender male inflorescence. The former has been given specific rank as A. grandis by Kirk (Trans. N.Z. Inst., vol. 4, p. 245, 1872), but it seems better to retain it in the aggregate. We thus have—


Astelia nervosa Banks et Sol. ex Hook. f. var. grandis (Hook. f. ex T. Kirk) Ckn. et Allan comb. nov.


Astelia nervosa Banks et Sol. ex Hook. f. var. sylvestris Ckn. et Allan var. nov.

Varietas distincta foliis plerumque ± 1.6 m. longis, interdum valde longioribus, ± 5 cm. latis, anguste linearo-lanceolatis attenuatis; paniculis masculis gracilibus ± 40 cm. longis.

47. Carmichaelia australis R. Br. var. egmontiana Ckn. et Allan var. nov.

Ramulis angustis applanatis ± 3 mm. latis; racemis circiter 5-floris.

North Island: Egmont-Wanganui Botanical District—Common as a member of the subalpine scrub on Mount Egmont.

[Footnote] * The numbers follow on consecutively inl this series of papers.

[Footnote] † The abbreviation we are using for “L. Cockayne.”

– 22 –

This variety is at once distinguished by its narrow branches and branchlets from the more widely-spread forms with comparatively broad branchlets. An examination of the type specimens of Kirk's var. strictissima showed that our plant is quite distinct in its much narrower branchlets, its spreading few-flowered racemes. The flowers of var. egmontiana are small, the standard being dark purple flanked by purple lines, and shading to a white margin. A seedling plant of var. egmontiana examined had the petioles and leaves on both surfaces sparsely clothed with short hairs. The leaves were pinnately 3–5-foliate, with leaflets ± 8 mm. long, broadly obovate, sessile.

48. Cassinia Vauvilliersii (Homb. et Jacq.) Hook. f. var. serpentina Ckn. et Allan var. nov.

Folia minora, laminis circa 4 mm. longis infra pallide brunneis haud. fulvis, costis minus carinatis; inflorescentia valde parviora ± 1.5 cm. diam., capitulis ± 20, circa 5 mm. longis.

South Island: Sounds-Nelson Botanical District—On Mineral Belt of the Dun Mountain: L. C.

Distinguished from C. albida (T. Kirk) Ckn. by the colour of the tomentum and the much smaller leaves. It might quite well be thought an epharmonic form of C. Vauvilliersii, due to the magnesian soil, but, if so, it is apparently “fixed,” as it has kept unchanged to any extent for five years in cultivation.

49. Coprosma Petriei Cheesem.

There are two varieties * of this species, distinguished by the colour of the drupes.


Coprosma Petriei Cheesem. var. atropurpurea Ckn. et Allan var. nov.

This is the type, described by Cheeseman (Trans. N.Z. Inst., vol. 18, p. 316, 1886). He states that the drupes are “bluish,” but perhaps it is more accurate to describe them as translucent, faintly stained very pale blue.


Coprosma Petriei Cheesem. var. astropurpurea Ckn. et Allan var. nov.

A typo drupis atropurpureis differt.

In this variety the drupes are much the colour of port-wine. The two varieties often grow side by side, but usually in pure patches of considerable size. They do not seem to hybridize. We have noted both varieties in various parts of the North-eastern and the Eastern Botanical Districts. Evidently each variety comes true from seed.

50. Epilobium melanocaulon Hook.

Two well-marked varieties constantly occur side by side on stony river-beds throughout the range of the species. Possibly the two varieties cross, but we have no notes on the matter.


Epilobium melanocaulon Hook. var. viride Ckn. et Allan var. nov. Caulibus foliisque griseo-viridibus.


Epilobium melanocaulon Hook. var. typica Ckn. et Allan var. nov. Caulibus intense atropurpureis paene nigris foliis atrorubris.

This variety is by far the more common.

[Footnote] * Since the descriptions were drawn up L. C. has seen a variety near Cass with white flowers.

– 23 –

51. Epilobium pernitens Ckn. et Allan sp. nov.

Herba parva, prostrata, affinis E. peduncularis A. Cunn. sed foliis floribusque valde distincta. Caules procumbentes radicantes saepe conferti. Folia maxime nitentia, viridia, interdum purpurascentes, convexo-concava, satis crassa, rotunda ± 6 mm. diam. vel subrotunda, marginibus paulo recurvatis. Petioli brevissimi sed distincti. Flores non multi, ut apparet, maiusculi, usque ad 16 mm. diam., axillares. Pedunculi ± 5 cm. longi, graciles, arrecti, rose0-brunnei. Calycis-lobi oblongo-lanceolati, acuti, ± 4 mm. longi, pallide virides vel rubentes. Petala alba, late obovata, ± 8 mm. longa, apice valde emarginata, ± 10 atris lineis ornata. Capsuli ± 3.1 cm. longi, glabri, pedunculis in fructu ± 5.3 cm. longis.

North Island: Ruahine-Cook Botanical District—Ruahine Mountains, Tararua Mountains, in subalpine herb-field on margins of bare places: H. H. A. South Island: North-western Botanical District—Paparoa Mountains: L. C.

This beautiful little species is distinguished at once from its nearest relatives—E. nummularifolium, E. pedunculare, and E. nerterioides—by its extremely glossy, convex, very small, rotund or subrotund leaves and its comparatively large flowers. It is to be recommended for the alpine garden, as its small patches of glistening foliage and large white flowers remind one somewhat of Linnaea borealis. It sows itself abundantly, but is not likely to become a garden-weed.

Picture icon

Fig. 1.–Leaves of Gaya : A, G. ribifolia Mount Fyffe; B, G. Lyallii; C, G.ribifolia, Mount Torlesse; D,G. Allanii, juvenile; E, F, G. Allanii, adult. × ½.

52. Gaya Allanii Ckn. sp. nov.

Arbuscula circa 4.5 m. alta; ramuli iuventute striati, plerumque stellatopubescentes, demum glabrati, cortice griseo-brunneo obtecti. Folia juvenilia ± 1.5 cm. longa, ± 8 mm. lata, petiolis ± 5 mm. longis, ovata, crenato-serrata, acuta, sparsis stellatis pilis induta. Folia

– 24 –

matura ± 5 cm. longa, ± 2 cm. lata, ovato-lanceolata, basi cuneata, submembranacea, grosse et incise bi-serrata vel crenato-serrata, pallide viridia, subtus pallidiora, acuta, supra et infra sparsis pilis stellatis albis brevibus plus minusve deciduis induta, nervis subconspicuis. Petioli graciles ± 1.5 cm. longi, stellatis pilis vestiti, canaliculati. Flores 3–5 fasciculati, brevium ramulorum apicibus versus vel axillares. Pedunculi graciles, ± 1.5 cm. longi, stellato-pubescentes. Calyx late campanulatus, 5-lobatus, dense stellato-pubescens, lobis acute triangulariis, ± 4 mm. longis. Petala ± 1.5 cm. longa, ± 7 mm. lata, oblique obovata, obtusa, retusa, unguiculata, infra satis dense stellato-pubescentia supra glabrata. Columna staminea brevis, basi distensa; filamenta gracilia ± 6 mm. longa, glabra; antherae ± 30. Ovarium ovoideum, ± 4 mm. longum, ± 2 mm. diam., plerumque 7-, interdum 6 vel 8-loculatum, loculis compressis, stellatis pilis vestitis.

South Island: Eastern Botanical District—Peel Forest, on margin of taxad forest: H. H. A. Flowers in January.

This very distinct species differs markedly in its smaller, differently shaped leaves, and rather smaller flowers with fewer carpels, from either of the other two species of the genus. Superficially it bears a much closer resemblance to Hoheria sexstylosa. It will some day become a famous garden-plant and take its place alongside the justly famed mountain-ribbonwood, G. Lyallii. It will be easily raised from seed, and probably from cuttings. Fig. 1 shows outlines of juvenile (D) and adult (E, F) leaves of Gaya Allanii, and of leaves from flowering specimens of G. Lyallii (B, Kelly's Hill, Westland) and G. ribifolia (A, Mount Fyffe, Seaward Kaikouras; C, Mount Torlesse, Canterbury).

53. Hebe as a Genus to replace a certain Section of Veronica.

F. W. Pennell (Rhodora, vol. 23, pp. 1–22 and 29–41, 1921; reprinted as Contrib. N.Y. Bot. Garden, No. 230, 1921) has restored Commerson's genus Hebe (Juss., Gen. Pl., 105, 1789). Hebe, as defined by Pennell, includes all those shrubs and trees hitherto referred to Veronica which occur in New Zealand, subantarctic South America, the Falkland Islands, Tasmania, and south-east Australia. The distinctions between Hebe and Veronica are as follows: the capsule dehisces septicidally (not loculicidally). the thick (not thin) septum splitting and each carpel opening distally by a median slit through the septal wall; the leaves are always opposite and after falling leave more or less conspicuous scars; the flowers are usually in axillary racemes, but these may be greatly reduced, spicate or corymbose; and all are evergreen shrubs, or in a few cases trees. Pennell remarks, “The austral distribution, with its suggestion of genetic remoteness, emphasizes Hebe's claim to recognition as a genus.” It is, indeed, the restriction of the group to those portions of the temperate Southern Hemisphere already noted, together with the evergreen, shrubby habit of its members, which so markedly separates Hebe from Veronica proper, that has induced us to support Pennell's action. Hebe thus becomes of great phytogeographical significance. Pennell also points out that “Hebe has an exceedingly baffling tendency to form local races, a habit at contrast with that of the other ‘Veronicas.”’

Now, L. C., at any rate, has long been of opinion that the section Hebe of Veronica should be raised to generic rank. Therefore, as soon as Pennell's paper came into his hands, he sent it to the late Mr. T. F. Cheeseman, and wrote on the matter to Professor Dr. L. Diels and

– 25 –

Dr. C. Skottsberg, whose opinion on any subject concerning subantarctic botany must receive the most careful consideration. Cheeseman replied that Pennell's action was probably justified, but that, as in the new edition of his Flora the genus Veronica was completed, he did not intend to make any alteration, and that in the absence of any other systematist of note having dealt with the question he had in the first edition followed Wettstein's treatment in Die Pflanzenfamilien. That a botanist so conservative as Cheeseman should consider the change to Hebe as probably correct speaks strongly in favour of the alteration.

Diels considered the change might quite well be made, and pointed out how certain genera had been removed from the wide conception of Veronica, which were closer to that genus than was Hebe. Skottsberg, though in general considering that there is in certain quarters a tendency to greatly overdo the splitting-up of genera, thought that the case of Hebe was in another category owing to its strictly limited distribution. In New Zealand, in addition to ourselves, Mr. W. R. B. Oliver has decided Hebe must stand, and has adopted the genus in a recently written paper.

The two objections to the change are that it will be very slowly adopted in horticulture and with more or less inconvenience, and that over one hundred species will have to be transferred. Fortunately, this will very rarely lead to changes in specific names. The following changes have already been published or are in course of publication: H. blanda (Cheesem.) Pennell, H. buxifolia (Benth.) Ckn., H. elliptica (Forst. f.) Pennell, H. salicifolia (Forst. f.) Pennell, H. subàlpina Ckn. H. vernicosa (Hook. f.) Ckn.

54. Hebe Allanii Ckn. sp. nov.

Frutex parvus, circa 30 cm. altus, ramulis patulis, profunde cicatricibus fohorum delapsorum notatis; internodia ± 8 mm. longa, dense albis pilis pubescentia. Folia sessilia, patentia vel retroflexa, paene subcordata, ± 1.8 cm. longa, ± 8 mm. lata, oblonga, subacuta, dense supra et infra pubescentes, marginibus ciliatis interdum rubro-tinctis, coriacea, glauca, paulo incurvata; costa inconspicua. Folia subfloralia ± 1.1-cm. lata, ovata vel obovato-oblonga. Spicae ± 25 florae, breves, paucae, axillares apicem ramulorum versus; pedunculi usque ad 2.5 cm. longi, dense mollibus patulis pilis albis induti. Flores sessiles, conferti. Bracteolae ± 5 mm. longae leviter carinatae, ovatae, subacutae vel obtusae, pubescentes, calyce paululum breviorae. Calyx pubescens ± 5 mm. longus; lobi profunde incisi, ovati, obtusi vel subacuti, pallide virides, marginibus scariosis. Corolla alba; tubus ± 6 mm. longus, calycis lobos excedens, extra sparsis pilis ornatus infra glaber; lobi patuli, ± 6 mm. longi, basim versus pilis sparsis infra ornati, ovati vel ovato-oblongi, obtusi. Stamina exserta, antheris atropurpureis. Ovarium pubescens. Capsula adhuc non visa.

South Island: Eastern Botanical District—On rocky outcrops in tall tussock-grassland at about 360 m., Mount Peel: H. H. A.

This well-marked species comes nearest to Hebe amplexicaulis, but is separated at once by its shorter branches, its leaves closely covered with hairs on both surfaces and with ciliate margins, its smaller, denser inflorescences, and smaller flowers, which are hairy in all parts except the upper portions of the corolla-lobes. It may easily be mistaken for H. Gibbsii on account of the presence of hairs on the leaves, calyx, &c., but in the latter species the leaves are ciliated only, and the calyx-lobes

– 26 –

are much narrower and sharply acute. Although Hebe is so richly represented in gardens, a species so distinct and striking as H. Allanii is certainly a great acquisition. It is easily raised from cuttings and from seed, and will probably grow in poor soil in almost any situation.

55. Hebe amplexicaulis (J. B. Armstg.) Ckn. et Allan comb. nov.

This species was first described by Armstrong (N.Z. Country Jour., vol. 3, p. 56, 1879) from a plant collected in the upper Rangitata by J. F. Armstrong. This (given in the description as “decumbent”) is the well-known almost prostrate garden-plant, which is to be taken as the type of the species, and which is here given the varietal name vera. The species was again described by Armstrong in 1881 (Trans. N.Z. Inst., vol. 13, p. 352), a plant collected in the upper Waiau, Nelson, being included in the conception of the species. The words “or suberect” are added to the habit description. Cheeseman's description (Manual N.Z. Flora, p. 525, 1906) differs from the original in stating the corolla-tube to be “about the length of the calyx,” and he remarks, “Mr. Armstrong describes the corolla-tube as long, but it barely equals the calyx in all the specimens I have seen, including an authentic one from him.”

Our observations in the field and in the garden show that H. amplexicaulis is an aggregate allied to the group that includes H. albicans (Petrie), but differing in the more obtuse semi-amplexical leaves, and the spicate not racemose inflorescences, with larger obtuse bracteoles. We find the corolla-tube of H. amplexicaulis distinctly longer than the calyx, but this is not shown till the flower has fully developed, so that young inflorescences may have the flowers as described by Cheeseman. Although the floral characters are very similar in all, we consider the habit differences sufficient to mark off the following varieties:—


Hebe amplexicaulis (J. B. Armstg.) Ckn. et Allan var. vera Ckn. et Allan var. nov.

This is the type, and needs no further description.


Hebe amplexicaulis (J. B. Armstg.) Ckn. et Allan var. suberecta Ckn. et Allan var. nov.

Caules non multi, pauciramosi; rami primo decumbentes deinde ascendentes, patuli, nudi; ramuli ultimi suberecti deorsum nudi, sursum foliis confertis obtecti; spicae breves floribus confertis.

South Island: Eastern Botanical District—A rock-plant from about 340 m. to 1,200 m., Mount Peel, abundant: H. H. A.

Distinguished from var. vera by its open straggling habit, with long naked branches ending in suberect branchlets with crowded leaves. In cultivation it becomes of denser, less rambling habit.


Hebe amplexicaulis (J. B. Armstg.) Ckn. et Allan var. erecta Ckn. et Allan var. nov.

Caules numerosi, erecti, ramosi; ramuli ultimi stricti, deorsum nudi, sursum foliis confertis obtecti; spicae longiores floribus numerosis.

South Island: Eastern Botanical District—On rocky outcrops at about 350 m. altitude, upper Rangitata River: H. H. A.

Only a few plants of this very distinct erect bushy variety were seen. The spikes are longer with less crowded flowers than in var. suberecta, and are more abundantly produced, so that the shrub presents a handsome appearance when in flower.

– 27 –

56. Hebe buxifolia (Benth.) Ckn. var. pauciramosa Ckn. et Allan var. nov.

Rami ultimi pauci, arrecti, foliis brevibus patentibus dense obtecti.

South Island: North-western, Western, Fiord, and South Otago Botanical Districts—In wet ground, mainly subalpine: L. C.

Hebe buxifolia is an aggregate species embracing many forms, but all distinguished by the smooth, glossy, green, keeled leaves, with short but distinct petioles (seedling and reversion leaves are deeply toothed or subpinnate), the short spikes of flowers more or less aggregated into heads, and the large bracteoles much resembling ordinary leaves. There are many intergrading forms extremely difficult to classify, but a knowledge of the species throughout its range shows that there are several well-marked varieties, which almost certainly hybridize, so producing the remarkable polymorphy. The varieties so far described are—(a) var. prostrata, which more or less closely hugs the ground; (b) var. odora, much-branched, eventually forming a globose bush; and (c) var. pauciramosa, here discussed. This last, possibly itself an aggregate, is distinguished by its erect final branches (it may be more or less prostrate at the base), straight, erect stems, leafy only in their upper portions, which branch but little and form a shrub of open habit. We have under observation other apparently distinct varieties, but they require further investigation.

57. Hebe Treadwellii Ckn. et Allan sp. nov.

Frutex humulis, ramulis elongatis, depressis, dispersis, cortice pallide brunneo obtectis, inferne cicatricibus foliorum delapsorum notatis. Folia arte imbricata, ± 2.5 cm. longa, ± 1.2 cm. lata, obovato-oblonga, subacuta, fere sessilia, supra valde concava, infra convexa, glabra, crassa, coriacea, satis nitentia, pallide viridia, marginibus pallidioribus; costa depressa. Petioli brevissimi applanati. Racemi duplices in foliorum superiorum axillis positi, aliquanto foliis obscurati, ± 1.2 cm. longi, pauci-flori; pedunculi, rhaches et bracteae (marginem versus) minute pubescentes. Bracteolae subulatae, calyce breviores. Flores albi, parvi, superiori sessiles, inferiori breviter pedicellati. Calyx 4-segmentatus; segmenta ovata, acuta, tubo corollae breviora, pallide viridia, marginibus ciliatis. Corollae tubus brevissimus, latus, calycem aequans: limbus 4-lobatus, haud patulus, lobis obtusis. Ovarium et stigma glabrum. Capsula non visa.

South Island: Western Botanical District—On stony ground at about 1,200 m. on Mount Ollivier (Sealy Range), and near the Mueller Glacier at a similar altitude: L. C.

This species, probably a common plant within its range, comes into the group which contains Hebe pinguifolia and its allies, but it is distinguished by its green leaves, its fewer few-flowered racemes almost hidden by the leaves, its subulate bracteoles, its smaller flowers with acute calyx-segments, and its glabrous ovaries and stigmas.

58. Metrosideros perforata (Forst.) A. Rich. = M. scandens Sol. ex Gaertn.

The specific name given by the Forsters (Characteres Generum, p. 72, 1776) was the first valid one, and Richard's adoption of it (Essai d'une Flore de la Nouvelle-Zélande, p. 334, 1832) must be followed.

– 28 –

59. Myosotis Colensoi (T. Kirk) Ckn. et Allan comb. nov. = M. decora T. Kirk ex Cheesem.

Exarrhena saxosa Hook. f. was described by Hooker (Fl. Nov.-Zel., i, 202) from a plant collected by Colenso in the North Island at Titiokura. In the Handbook (p. 196) he united with it a plant from the Dun Mountain, Nelson, collected by Monro and Travers. Kirk (Trans. N.Z. Inst., vol. 27, p. 351, 1895) described a plant from limestone rocks at Broken River, Canterbury, collected by Enys and himself, as E. Colensoi, and united with it the Titiokura plant, saying, “A remarkable plant which has hitherto been confused with the Nelson E. saxosa Hook. f.” But, as E. saxosa is based upon the Titiokura plant and not the Nelson one, Cheeseman appears to have adopted a suggestion of Kirk to rename the Broken Hill plant as Myosotis decora (Manual, p. 462, 1906), the Nelson plant receiving the name M. Monroi (ibid., p. 469). But, since the Broken Hill plant provided the material for Kirk's original description of his E. Colensoi, that specific name should be adopted.

60. Nothopanax arboreum × Pseudopanax crassifolium var. unifoliolatum.

Mr. E. Phillips Turner recently discovered on Kapiti Island, growing amongst Leptospermum ericoides, what he thought to be the above hybrid. In this conclusion we are strongly inclined to concur. The plants, three in number, were growing side by side close to a young plant of P. crassifolium var. unifoliolatum. From the largest plant, some 40 cm. tall, Mr. Turner collected four leaves. Fig. 2 should, give a sufficient idea of the leaf-form and size. Except B, the leaves differ greatly from those of any stage of the polymorphic P. crassifolium. The excessive increase in breadth is what might be expected from a cross with N. arboreum. The texture is similar to that of N. arboreum, and the venation with the secondary veins diverging at angle of about 50° is that of N. arboreum, the angle in the case of P. crassifolium being about 20°. The midribs are prominent above and below, as in juvenile P. crassifolium, and the marginal teething is nearer that of P. crassifolium than of N. arboreum. The base of the lamina is intermediate in form.

Picture icon

Fig. 2.—Leaves of Nothopanax arboreum × Pseudopanax crassifolium var. unifoliolatum. × ½.

– 29 –
Picture icon

Fig. 3.—Leaves of Nothopanax simplex, N. anomalum, and N. parvum. × ½.

61. Nothopanax parvum (T. Kirk) Ckn.

L. Cockayne * recently expressed the opinion that Nothopanax parvum = N. anomalum × simplex. This view is supported by a number of specimens of what was apparently N. parvum studied by him in the Spey Valley, west of Lake Manapouri. Unfortunately, too few specimens were collected, but they were present in great numbers and made up a most polymorphic group. These juvenile plants differ greatly from those of the supposed parents, but exhibit features which certainly suggest a mixture of parental characters. N. simplex has juvenile leaves ovate-serrate or 3—5-foliolate with the leaflets deeply cut (fig. 3, A—a small

[Footnote] * See L. Cockayne, Hybridism in the New Zealand Flora, The New Phytol., vol. 22, No. 3, p. 126, 1923.

– 30 –

example). These are succeeded by 3-foliolate leaves with the leaflets lanceolate-serrate (fig. 3, B, C). The adult has simple lanceolate-serrate leaves (fig. 3, D). It is almost certain that there are distinct races differing in their juvenile stages. N. anomalum has small 3-foliolate juvenile leaves with leaflets elliptic-ovate or rounded-ovate with crenate margins (fig. 3, E), succeeded by simple adult leaves oblong-orbicular, crenate-toothed (fig. 3, G, H). The leaves of the Spey Valley juveniles show intermediate characters, as can be seen from the drawings of three specimens (fig. 3, K1, K2, K3; L1, L2, L3; M1, M2, M3, M4). Two juveniles from Stewart Island are figured (fig. 3, N1, N2, N3; O1, O2, O3) which also show an intermingling of parental characters. Adult flowering specimens are figured from Inchbonnie, Westland (fig. 3, P1, P2, P3), and Stewart Island (fig. 3, R1, R2, R3; S1, S2, S3; T1, T2, T3). An examination of the drawings, traced from nature-prints, will be sufficient to show the intricate way in which the various parental characters are intermingled, without the necessity of drawing up an elaborate table. That flowering should occur on forms so different as P, R, S, T is important evidence as to their hybrid origin. This is borne out by the intermediate character of the inflorescences, bridging the gap between the few-flowered simple umbels of N. anomalum and the comparatively large compound umbels of N. simplex.

62. Ourisia Macphersonii Ckn. et Allan sp. nov.

Herba perennis, rhizomatis longis ± 8 mm. diam. Folia radicalia subcoriacea. Laminae ± 4.5 cm. longae, ± 3.5 cm. latae, supra pilis mollibus sparsis vestitae, infra glabratae, ovatae vel ovato-oblongae, apice obtusae, basi subcordata vel cuneatae, marginibus crenatis vel crenato-serratis, ciliatis; subtus purpuratae, venis viridibus distincte reticulatae. Petioli ± 3.5 cm. longi, striati, sparsis albis pilis induti, vaginantes, bases versus latescentes. Inflorescentia maiuscula; rhachis ± 30 cm. longus, basi ± 1 cm. diam., erectus, nervatus, pilis brevibus vestitus, interdum purpurascens, plerumque 5-verticillatus, verticillis sursum deminuentibus in omnibus partibus. Bracteae verticillatae, plerumque 5 per verticillum; inferiorae breviter petiolatae, late lanceolatae vel lanceolato-oblongae, obtusae, crenato-serratae vel crenatae; superiorae oblongae, obtusae, sessiles. Flores 7—8 per verticillum. Pedicelli ± 4 cm. longi, patuli, paulo crassi, albis pilis vestiti. Calycis-lobi usque ad basim 5-partiti, extra pilosi; lobis oblongis vel obovatis, apice purpuratis, obtusis vel subacutis, marginibus ciliatis, venis purpuratis distinctis. Corollae ± 2.5 cm. diam., albae, in fauces luteae glandulosis pilis ornatae; tubi ± 6 mm. longi, 2 inferiori lobi subrotundi vel late oblongi, patentes; 3 superiori lobi obovato-oblongi, suberecti, 3—4 venis distinctis, apice rotundati vel prope truncati. Stamina 4, supra apicem corallae tubi paulo exserta; antherae cremeae, reniformes. Styli longi, graciles, papillati, stigmatis 2-lobatis. Capsula paulo compressa, ± 1 cm. longa, ± 5 mm. lata, mucronata, calyce persistente induta.

South Island: Fiord Botanical District—Shady banks of River Spey at about 210 m., and in open boggy ground towards Wilmot Pass at 670 m. altitude: L. C.

This species has affinities with Ourisia macrophylla Hook. and O. Colensoi Hook. f., but differs from the former in the rather thicker, smaller leaves, the absence of cauline leaves, the lower bracts lanceolate petiolate (not linear-oblong), the yellow-throated flowers, the oblong or obovate (not

– 31 –

lanceolate or linear) calyx-lobes, and the rather larger hardly turgid capsules. From O. Colensoi it is at once distinguished by the stouter habit, bracts in whorls of ± 5 (not paired), more numerous larger flowers on longer petioles, in whorls of 7–8, and the calyx-lobes not linear.

The plant blooms for about two months, commencing under cultivation near Wellington in the middle of October, and continuing until the second week of December. It appears to be of easy cultivation, and in the southern rainy parts of New Zealand should grow with the greatest vigour. The tiers of large white flowers with yellow throats, one above another after the manner of Primula japonica, make it a most effective plant for the alpine garden. It will not tolerate drought.

Picture icon

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Fig. 4.—Leaves of Plagianthus betulinus (× ½), and of P. divaricatus and P. cymosus (× 1/1).

63. Plagianthus cymosus T. Kirk.

This so-called species invariably grows along the banks of tidal rivers not far from the sea where P. betulinus and P. divaricatus grow near one another. Mr. W. Martin, B.Sc., has recently collected it in Chatham Island. Cheeseman. (Illust. N.Z. Flora, vol. 1, t. 21, 1914) says, “If its characters are carefully compared with those of P. betulinus and P. divaricatus it will be recognized that it stands nearly half-way between the two species. The suspicion of a hybrid origin at

– 32 –

once arises, and it must be confessed that such an assumption would go far towards explaining the peculiar rarity of the plant in some of its localities. On the other hand, fruiting specimens sent to me by Mr. H. J. Matthews from the Pelorus Valley have well-developed and well-ripened seeds with a fully developed embryo, and show no signs of the impaired fertility frequently seen in hybrids.” L. Cockayne, in the New Phytol. paper already cited, listed it as a probable hybrid, the case of which required further investigation. In view of the number of hybrids now known to produce fertile seed, Cheeseman's objection does not seem to have any weight.

We have examined specimens from several localities—Pelorus Valley, Kaitaia, Chatham Island—and the leaf-characters strongly suggest a hybrid origin. P. divaricatus (fig. 4, B1, B2, B3, B4, B5) has small linear obtuse leaves, quite entire, while P. betulinus has large ovate-lanceolate crenate-serrate leaves, acute, with more or less cuneate bases (fig. 4, A). Representative leaves taken from single shoots from each locality well show the mingling of parental characters, in their intermediate size and inconstant shape and toothing—Kaitaia specimen (fig. 4, C1, C2, C3, C4, C5); Pelorus River specimens (fig. 4, D1, D2, D3, and D4, D5, D6); Chatham Island specimen (fig. 4, E1, E2, E3, E4, E5). These should be compared with the figures given in Illustrations of the N.Z. Flora from specimens collected in the Pelorus Valley, which show still another somewhat different form. Leaves taken from among the inflorescences of herbarium specimens of P. betulinus are also shown (fig. 4, E, G, H). These show a reduction in size but retain the characters of P. betulinus, and it is to be remembered that in this species the leaves are not fully developed at flowering-time. P. divaricatus has flowers solitary or in few-flowered fascicles, while P. betulinus has numerous flowers in large panicles. Our Kaitaia specimen has small male panicles just as might be anticipated on the hybrid theory, while, according to Mr. Martin, in the Chatham Island plant the fruit-clusters never exceed four. The number of anthers in our Kaitaia specimen is about 15–20 per flower. There can hardly be any doubt that P. cymosus is a hybrid with the parentage suggested. *

64. Pterostylis confertifolia Allan sp. nov.

Herba terrestris glabra, ± 7 cm. alta; erecto caule e tubere pisiforme ± 7 mm. diam. Folia conferta caulem amplectantia, inferiora scariosa squamiforma, obtusa, superiora plerumque 3, erecto-patentes, pallide viridia, paulo crassa, in siccitate tenuiora; laminae 3–5 cm. longae, 1.5–2 cm. latae, elliptico-ovatae, obtusae, venis per vitam obscuris, marginibus paulo incurvatis, basim versus in vaginas latas albas angustatae. Flores solitarii, virides, venis rubris, vix folia excedentes, aliquanto in fructu increscentes, ± 2 cm longi, ovario obovoideo ± 1.5 cm. longo excepto. Galeae erectae usque ad ± 12 mm., deinde arcuatae, apicibus acutis. Petala lateralia in inferioribus dimidiis linearo-oblonga, in superioribus falcata, acuta. Sepala lateralia linearia, usque ad ± 8 mm. connata, deinde in lobos 2, ± 1.2 cm. longos, erectos, acuminatos producta.

[Footnote] * The Waimakariri station mentioned by Kirk (Student's Flora, p. 71, 1899) appears to be based on a mistake. There are no specimens thence in his herbarium. The original specimen from Dunedin was (according to information supplied by the Hon. G. M. Thomson, taken from his original notes of many years ago) gathered from a plant culti vated in a garden.

– 33 –

Labellum subcrassum, latericium, linearo-oblongum, marginibus minute serrulatis; in canaliculatum, minute retusum, paulo exsertum apicem angustatum. Unguis curvatus, apice 2 longis et aliquis brevibus satis latis filamentis ornatus. Columna gracilis, galeam partem erectam aequans; auriculae decurrentes, superioribus lobis late triangularis, apice breviter acuminatis; inferioribus lobis late oblongis, obtusis, apice mollibus pilis vestitis.

North Island: Wellington Botanical District—On margins of subalpine scrub, and in lower subalpine herb-field, about 1,200 m. altitude, Ruahine Mountains, near Apiti: H. H. A.

Though coming into the group containing P. Banksii, P. australis, P. graminea, that have cauline leaves only, this species is very distinct. So crowded are the broadly elliptic-ovate, obtuse leaves that superficially the species more closely resembles the P. foliata group. The plant grows in small colonies, and when growing through moss cushions the stems may be much elongated.

65. Ranunculus maculatus Ckn. et Allan sp. nov.

Herba parva, caespitulosa. Folia numerosa, radicalia, maiusculis brunneis maculis notata; laminae ± 9 mm. longae, ± 10 mm. latae rotundato-cuneatae, supra paulo pilosae, nervis tribus distinctis; 3-lobatae usque ad media, mediis lobis parvioribus; lobi lati, obtusi, inaequaliter 2–3 crenato-dentati; petioli ± 16 mm. longi sine vaginis, ± 1 mm. diam., erecto-patentes, pilis longis sparsis appressis induti; vaginae ± 14 mm. longae, ± 3 mm. latae, marginibus paene hyalinis. Scapi ± 5 cm. longi, satis crass, ± 2.5 mm. diam., longis pilis induti. Sepala 5, fugacea, 3 mm. longa, 2 mm. lata, late ovata, obtusa, infra sparsis pilis obtecta, marginibus scariosis. Petala plerumque 5, ovata vet obovato-spathulata, 5 mm. longa, 2 mm. lata, flava, supra nitentia, basim versus alba. Nectarium breve, subulatum. Stamina non multa, filamentis ± 1.5 mm. longis; antherae ellipticae, 0.5 mm. longae. Carpella ± 20, glabra, aliquanto turgida. Stylus brevis, stigmate paulum recurvato.

South Island: South Otago Botanical District—In wet ground at 1,200 mm. on Rock and Pillar Range: L. C.

This seems to be a distinct species with no near relative.

66. Ranunculus rivularis Banks et Sol. ex Forst. f. var. glareosus Ckn. et Allan var. nov.

Ab varietatibus aliis speciei habitu robustiore, foliis 3-lobatis, lobulis inaequaliter 2–3 lobulatis, petalis obovato-oblongis haud linearo-oblongis, distinguitur.

South Island: Eastern Botanical District—(1) Ninety-mile Beach, L. C.; (2) mouth of Ashburton River, H. H. A. On shingly or sandy foreshore.

This variety is very closely allied to the other varieties of the species, but is apparently confined to coastal gravel or sand, a station very different from those usual for R. rivularis. It is of robust habit, with far-spreading rhizomes emitting tufts of dark-green leaves on long petioles, the blades being 3-lobed almost to the base, with the lobes again irregularly lobulate. In cultivation the plant becomes rampant, and in one case grew over and suppressed a common damp-ground form of the species.