Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 29, 1896
This text is also available in PDF
(741 KB) Opens in new window
– 501 –

Art. XLVII.—On Carmichaelia, Corallospartium, Huttonella, and Notospartium.

[Read before the Wellington Philosophical Society, 17th February 1897.]

Carmichaelia is perhaps the most characteristic of the many remarkable endemic genera of New Zealand phanerogamic plants, as, although absent from the larger outlying islands, Stewart Island, and the Chathams, it is represented from the North Cape to Foveaux Strait by numerous species, differing widely in habit and general appearance, but in all localities recognised by the settler as native broom and by the Maori as makaka.

Although one of the very earliest New Zealand genera collected by Europeans, our acquaintance with its numerous species has been of slow growth. In 1769 Banks and Solander collected the first species at Poverty Bay, and gave it the MS. name of Genista compressa. The next species was discovered by Forster in Dusky Sound, in 1772, and described in G. Forster's “Prodromus” as Lotus arboreus. The genus Carmichaelia was established by Robert Brown in 1825, but no additions were made until 1853, when four new species, discovered by Colenso, were described in Hooker's “Flora Novæ-Zelandiæ,” Forster's plant being considered identical with the Banksian Genista compressa. In the “Handbook of the New Zealand Flora,” published in 1864, the number of species is increased to nine, two of which were discovered by Colenso, but treated as varieties in the original Flora, the other two being remarkable southern plants. In 1871 the lamented Baron Von Mueller described C. exsul, of Lord Howe's Island, the tallest member of the genus, and the only species found outside New Zealand, and in 1878 C. williamsii was added to the flora. Numerous species have been described since that date; the number enumerated in the “Students’ Flora of New Zealand,” now in the Press, is twenty-five, arranged under the genera Carmichaelia, Corollospartium, and Huttonella. Two species of Notospartium must also be enumerated.

The species of Carmichaelia are admittedly difficult of definition. The learned author of the “Flora Novæ-Zelandiæ” wrote: “The species are extremely difficult to distinguish, and require much further elucidation on the spot; the characters employed appear far too variable” (i., 50); and in the

– 502 –

supplement to that work he adds, “I have received many forms of Carmichaelia, and they certainly do not tend to clear up the difficulty of discriminating the species, but rather complicate them, several of these being intermediate between those already defined. The whole genus requires careful revision in New Zealand, and a judicious selection of ticketed specimens from the same and different individuals at many different localities, various periods of growth, different seasons of the year, &c. Their habits and variations should also be narrowly watched in a growing state” (ii., 327). Speaking for myself, I am convinced that until this has been done it will be impossible to define the species in a thoroughly satisfactory manner, on account of the great amount of variability exhibited by many of them at different stages of growth. It is therefore obvious that I do not claim immunity from error in the descriptive portion of the following arrangement.

Carmichaelia, R. Br. In Bot. Reg., xi., 912 (1825).

With considerable variety in stature, habit, foliage, and inflorescence, all the species of this genus exhibit a 2-valved pod, with the margins and placentas consolidated and thickened, the valves falling away from this framework, or more rarely opening at the base or apex, leaving the solitary or numerous seeds attached to the placentas for a longer or shorter period.

Habit, Etc.

The species may form dense leafless patches, less than an inch in height, on the surface of the ground (C. enysii), or large rounded masses 2in.—5in. high (C monroi), or bushy leafless shrubs 2ft.—6ft. high (C. australis, C. virgata), or leafy shrubs 8ft.—9ft. high, with elegant pendulous branches (C. odorata). The tallest species is said to attain the height of 14ft., but is restricted to Lord Howe's Island.* Some species have elegant whip-like shoots, while others are inelegant bushes, with numerous short irregular branches. The whole aspect of certain species is completely altered when growing in situations open to the attacks of sheep.

The branches may be flat, ¾ in. broad (C. williamsii), or almost filiform, terete or plano-convex, erect, distichous or spreading. Some species exhibit flattened branchlets at the beginning of the season, which become terete or plano-convex before the fruit reaches maturity; more slowly this characteristic is exhibited by many of the larger species, the branches of which almost invariably become terete with the development of the plant. The branchlets are usually striated

[Footnote] * C. exsul, F. Muell. The only species not found in New Zealand.

– 503 –

or grooved, and rigid, but rarely flexuous and twining amongst other shrubs; lastly, they may be pubescent or glabrous, while very slender and very robust branchlets may sometimes be developed on the same plant.


The leaves are alternate, and almost invariably of brief duration. In some species they are restricted to very young plants (C. williamsii), the mature state being always leafless; in others they are reduced to mere scales (C. monroi). The whole plant may be leafy until the maturation of the pod, when the leaves fall away in a few days (C. odorata); in species which exhibit this peculiarity the young plants are invariably leafy. The leaves are articulated at the junction of the petiole and blade, and may be 1-foliolate or pinnately 3—7-foliolate; the leaflets are sessile, and may be ovate, ovate-cuneate, or oblong-cuneate, orbicular, &c., but are always emarginate or obcordate In a few species they are glaucous beneath. In some species the leaves vary at different stages of growth—in the young state 3—5-foliolate, large, distant; but on mature specimens 1—3-foliolate, small, dense, and often fascicled (C. flagelliformis). Generally speaking, 3- or 1-foliolate leaves are the most common, but 1-, 3-, 5-, or 7-foliblate leaves may be found on the same plant. They are often pubescent in the young state.


The flowers are developed in the alternate notches which are formed on each margin of the branchlets. They are usually, arranged in racemes, which, by non-development of their internodes, are often much shortened, and frequently reduced to fascicles. The flowers are rarely solitary (C. uniflora), when the peduncles are elongated and very slender; or in 2—5-flowered racemes (C. nana), or 10—40-flowered (C. odorata). Occasionally the racemes are crowded, forming a false corymb. The rhachis and pedicels may be glabrous, puberulous, or silky; usually the pedicels carry one or more bractlets.

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

The flowers are constructed on the ordinary papilionaceous type, and vary in size from 3/16 in. to 1in. in length, and may be pendulous or erect. In colour they may be of a dull red or yellow, violet, or lighter shades of purple, or very pale purple with darker streaks and markings; very rarely they are pure white. Those of a red tinge are almost invariably produced on species of extremely dwarf habit.

The calyx may be truncate, campanulate, or almost tubular, while the sinus between the teeth may be broadly rounded, or simply acute. Sometimes the teeth are extremely

– 504 –

tremely minute. The whole calyx may be glabrous, silky, or puberulous.

The standard in most species is longer than broad, and is invariably reflexed; in a few species it is broader than long, the lobes being slightly auricled at the base. It may be entire or emarginate. The wings are usually linear and auricled, the length of the narrow claw varying in the different species; usually they are shorter than the keel, in a few species they exceed it. The keel is usually slightly longer than the style, which is enclosed by it at least before anthesis; in one or two instances the two petals of which it consists show an obvious want of cohesion at the apex. The stamens are invariably diadelphous, the filaments being slightly unequal in length, but the free portion is very short, so that the tube invariably encloses the style, except at its apex. The ovary may be nearly globose or oblong, and is rarely stipitate; the style may be straight or curved, and is usually naked, but in one or two species it is sparsely ciliated although not bearded (C. kirkii). The number of ovules is extremely variable. Normally the ovary is glabrous in all species of Carmichaelia, but occasionally several species develope few or many silky or pilose ovaries; this condition is, however, never permanent.*

The most valuable characters for differentiation of the species are afforded by the pods; in most species identification is difficult and uncertain in their absence. On the other hand, the pod alone is sufficient for positive identification in the majority of species (C. monroi, C. williamsii, C. kirkii, &c.) The pods may be erect (C. odorata), or spreading (C australis), or pendulous (C. monroi). They may be turgid (C. williamsii), rounded (C. kirkii), or much compressed (C. angustata), elongate, elliptical, or subglobose, while the beak may be straight and well developed, oblique, or curved, or almost obsolete. The valves are excessively rugose or corrugated (C. monroi), faintly reticulate, or almost smooth (C. australis). In some species the pods are 1-seeded, others exhibit from two to twelve seeds in each pod. The seeds are reniform or subreniform, rounded or compressed, mottled or perfectly black (C nana), rarely red (C. australis). In all the species the radicle exhibits a double flexure.

In most cases the seeds are liberated by the valves dehiscing along the replum and falling away; sometimes the seeds remain attached to the placentas for weeks, or even months, after the valves have fallen (C robusta). In a few cases the valves open at the base or apex only, and are not separable for their entire length.

[Footnote] * The same phenomenon is often exhibited by Festuca pratensis, L.

– 505 –

Corallospartiuim, J. B. Armstrong. In Trans. N.Z. Inst., xiii. (1880), 333.

No technical definition of this genus was given by its author, but it differs essentially from Carmichaelia in the absence of the replum, and in the ovary and pod being invariably silky or pilose. It is a singular plant, with sparingly branched cylindrical leafless stems, ½ in. ¾ in. in diameter, and 2ft.—4ft. high. The stems are yellow, with numerous deep longitudinal grooves, which are filled with short black lax tomentum. In the young state the stems are compressed or subterete, and produce a few unifoliolate leaves, which speedily fall away. The flowers are produced in dense fascicles; pedicels, bracteoles, and calyx being excessively woolly. The ovary is almost deltoid, rounded at the back, and excessively villous, and the pod is 1-seeded, with very thin valves.

Huttonella, n.g. T. Kirk.

The four species comprised in this; small genus have the general habit of Carmichaelia, but the pods are indehiscent and turgid, the breadth usually exceeding the length, beak ascending or sharply turned upwards, sometimes forming a right angle with the axis of the pod. Seeds 1—3. Branchlets terete or compressed. Leaves 1—3-foliolate, but only known in a single species. One species is said to attain the height of. 6ft.-8ft.; the others are of more humble stature, or prostrate.

The pods remain on the branches until the new flowers are produced, when they fall to the ground and decay, the valves being inseparable from the replum, which is usually imperfect.

(Described fully in “Student's Flora of New Zealand,” now in the press.)

Notospartium, Hook. f. In Hook. Kew Jour., ix. (1857), 176, t. 3; Handbk. of N.Z. Fl., 51 (1864)

This remarkable genus consists of only two species, usually leafless shrubs, with slender, flexuous or pendulous, much-compressed narrow branchlets and small racemose flowers. The flowers differ but little from those of Carmichaelia; the wings are shorter than the hatchet-shaped keel; the ovary is linear, tapering into a long curved style, which is bearded on the upper surface. The pod is linear-elongate, shortly stipitate, straight or falcate, compressed, shortly beaked, torulose, 5—10seeded, indehiscent. Unifoliolate leaves are developed on very young plants Only.

One species was described by Sir Joseph Hooker as above; the other, which at present is not thoroughly understood; was included with Carmichaelia kirkii by Mr. J. B. Armstrong in his description of “Carmichaelia gracilis,” in Trans.N.Z. Inst., xiii. (1880), 336

– 506 –


It is remarkable that, while six species of Carmichaelia are common to the North and South Islands, no species extends over the entire area of the genus.

C. australis, R. Br., is found from the North Cape to the northern part of the Wellington. District, and has a solitary outlying habitat in Queen Charlotte Sound, South Island. It ascends from sea-level to 3,000ft.

C. odorata, Col., occurs from the Ruahine Mountains to Cook Strait, with a single outlying habitat in Pelorus Sound, South Island.

C. flagelliformis, Col., is found from the East Cape to Milford Sound and Te Anau Lake, which can hardly be its extreme southern limit.

C. diffusa, Petrie, which is scarcely satisfactory as a species, has a single habitat in the Wellington District, and is also found in Canterbury and Otago, although remarkably local.

C. subulata, n.s., occurs on the East Cape, if the identification be correct, but is plentiful in Marlborough and Canterbury.

C. enysii, T. Kirk, is found on the Rangipo Plain, in the North Island, and in several localities in Canterbury and Otago.

C. nana, Hook. f., occurs from Taupo to Central Otago, and is probably more plentiful over a wider area than any other species.

The following are restricted to the North Island:–

C. williamsii, T. Kirk. Only found at the northern extremity of the East Cape; is the rarest and most local of all the species.

C. acuminata, n.s. White Rock, East Coast.

C. hookeri, n.s. Akitio River to Pencarrow Lagoon.

The following are only found in the South Island:—

C. uniflora, T. Kirk. Nelson; Canterbury; North Otago.

C. monroi, Hook. f. Marlborough to Otago; ascends to 4,500ft.

C. grandiflora, Hook f. Nelson, Canterbury, and Otago.

C. robusta, n.s. Canterbury Plains and Broken River basin.

C. petriei, n.s. Upper valley of the Clutha, Otago.

C. violacea, n.s. Coleridge Pass, Canterbury.

C. virgata, n.s. Otago(?), Southland. Attains the extreme south limit of the genus near the mouth of the Waiau River.

C. kirkii, Hook. f. Canterbury and Otago, but local.

C. angustata, n.s. Valley of the Buller, Nelson.

– 507 –


The only species is restricted to the mountains of Canterbury and Otago. Local, but more plentiful in Otago than in Canterbury.


H. juncea. From the East Cape, Rotorua, Taupo, to Canterbury and Central Otago. Possibly two species are included under this name.

H. curta. Waitaki Valley.

H. compacta Central and Northern Otago; local.

H. prona. Broken River basin, Canterbury; 2,800ft.


N. carmichaeliæ, Hook, f. Marlborough, from the Awatere to the southern boundary of the province, but, unhappily, has been extirpated in many habitats. Ascends to nearly 2,000ft. in ravines arid valleys.

N. torulosum, n.s. From the Mason River, Nelson, to Mount Peel, Canterbury, but rare and local.

From the above it will be seen that seven species of Carmichaelia are common to both islands, but in very unequal degree; four having a wide distribution in one island, but only a solitary habitat in the other; the other three are more widely distributed in each. Three species are absolutely restricted to the North Island, two of them being remarkably local. Nine species are restricted to the South Island, two of which are found from Marlborough or Nelson to Otago, the remaining seven being decidedly local. The different species range from sea-level to nearly 5,000ft.

Corallospartium crassicaule, J. B. Armstrong, is local, and alpine or subalpine in the Districts of Otago and Southland.

Huttonella juncea, T. Kirk, is diffused from the East Cape to Central Otago, but is decidedly local.

The other three species are confined to the South Island, one being restricted to the Canterbury District and two to Otago.

Notospartium, Hook, f., as already stated.

The following is the arrangement of genera and species adopted in the forthcoming “Student's Flora of New Zealand”:—

1. Corallospartium, J. B. Armstrong.

I. C. crassicaule.

2. Carmichaelia, R. Br.

I. Nana. Depressed leafless plants with fastigiate compressed branchlets, forming dense patches 1in.—4in. high. Flowers red.

– 508 –

*Branchlets linear or narrow-linear, thin.

Flowers solitary or racemose. Pods 1-seeded 1. C. enysii

Flowers solitary, on long peduncles. Pods 3-seeded 2. C. uniflora.

Flowers racemose. Pods 2—4-seeded 3. C. nana.

Branchlets robust.

Flowers racemose. Pods 6—12-seeded C. monroi.

II.Eucarmichaelia. Erect or spreading species, 1ft.-8ft. high, leafless when in flower, except 7, 13, and 15. Flowers usually streaked with purple, rarely yellowish, white, pink, or deep purple.

*Pods convex or turgid.
A. Stems stout, erect, leafless, except 7 and 13.

Flowers few, 1in. long. Pods 1in. long, straight,

erect 5. C. williamsii.

Flowers numerous, small. Pod oblong, seeds red 6. C. australis.

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

Leafy. Branchlets flat, 1/12 in.— 1/4b in. broad. Pods

erect 7. C. grandiflora.

Branchlets robust. Racemes lax. Pods elliptic-oblong; seeds 3–6 8. C. robusta.

Branchlets spreading, terete. Pods broadly oblong; seeds 1 or 2 9. C. petriei.

Branchlets deeply grooved, terete. Flowers violet.

Pod 2—4-seeded 10. C. violacea.

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

Branchlets almost filiform. Pods 1/6 in.— 1/5 in. long;

1-seeded. 11. C. diffusa.

Branchlets terete or plano-convex. Pods drooping, 1—2-seeded 12. C virgata.

Leafy. Branchlets compressed or plano-convex.

Pods subulate 13. C. subulata.

B. Stems slender, flexuous, interlacing.

Leafy, racemes short. Pod with a straight subulate beak 14.C. kirkii.

Pods compressed.
† Leafy when in flower.

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

Branchlets compressed, 1/20 in.— 3/16 in. broad, drooping. Pods in erect racemes, 2-seeded 15. C. odorata.

Branchlets compressed or terete. Pods many in spreading racemes 16. C. angustata.

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

Branchlets compressed, fastigiate, 1/20 in.— 1/10 in. broad. Seeds 2—4 17. C. flagelliformis.

‡ Leafless or nearly so when in flower.

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

Branchlets compressed, 1/12 in. broad. Pods abruptly acuminate 18. C. acuminata.

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

Branchlets compressed, 1/16 in.— 1/12 in. broad. Pod

ovate oblong, small 19. C. hookeri.

3. Huttonella, T. Kirk, n.g.

*Leafless when in flower.

Branchlets numerous, erect, terete. Racemes lax 1. H. compacta.

† Racemes compact.

Branchlets few, erect, terete. Pods broadly oblong, 2—3 seeded 2. H. curta.

Erect or prostrate. Branchlets terete or compressed. Pods 1—3 seeded 3. H. juncea.

– 509 –
Leafy when in flower.

Prostrate. Branchlets compressed. Pods 1-seeded 4. H. prona.

4. Notospartium, Hook. f.

Branchlets drooping. Pods 3–5-seeded, crowded, straight 1. N. carmichaeliæ.

Branchlets flexuous or pendulous. Pods 3–10-seeded, curved, distant 2. N. torulosum.

In conclusion, I venture to urge the advisability of students of the genus obtaining flowering and fruiting examples from the same plant before attempting to identify their specimens, and of observing the species over as wide an area as possible. It would be of the greatest benefit if many of the species could be raised from seed, so that the young plant could be observed from the seedling state, and the period of leaf duration exactly determined.