Art. XLI.—Descriptions of New Native Flowering-plants, with a few Notes.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 53, 1921, Page 365

Art. XLI.—Descriptions of New Native Flowering-plants, with a few Notes. By D. Petrie, M.A., Ph.D., F.N.Z.Inst. [Read before the Auckland Institute, 15th December, 1920; received by Editor, 31st December, 1920; issued separately, 8th August, 1921.] Plates LVII, LVIII. 1. Note on Pittosporum cornifolium A. Cunn. As existing descriptions of this interesting Pittosporum appear to be imperfect, if not inaccurate, the following observations, which may help to clear up the position, seem worthy of record. In July last I was able to study, on Kawau Island, a number of these shrubs showing good flowers as well as ripe capsules. Most of them grew on clay soil just above low sea-washed rocky faces. The plants were of two kinds, which may be distinguished as male and female. The male

plants were less common than the female, and of less vigorous growth. Not a single capsule was to be found on any of the males, while on the females they were plentiful, and just ripe enough for the valves to be opening and exposing the numerous rather large pitchy seeds. The male inflorescence was terminal on the branchlets, and consisted of an umbel of 10 or usually fewer (6–8) flowers on long very slender peduncles. These flowers had every appearance of being hermaphrodite, the pistils being well developed and equalling the stamens in height, while the stigmas were more or less coated with pollen. But as no capsules occurred on any of the male plants it is clear that fertilization did not take place. The pistils are probably sterile to the pollen of their own flowers; but more extended observations are needed before this conclusion can be considered established. In the female plants the inflorescence was reduced to a single terminal flower, rather smaller than the male ones, seated on a short stouter peduncle about as long as the flowers. Here the pistil hardly differs from that present in the males. The stamens are much shorter, with anthers greatly aborted, and destitute of pollen. As already noted, these plants produced a capsule at the end of almost every branchlet. In every case the capsule was two-valved, and this is the case in the considerable number of dried fruiting specimens in my herbarium, gathered from various widely-scattered stations in the North Island. Mr. H. B. Matthews, a very careful observer, has examined many plants bearing capsules, and in one case only has he seen a three-valved one, and in it the extra valve was much smaller than the other two. Both Mr. Kirk and Mr. Cheeseman give the number of the valves as three, but there can be no doubt that two is the normal number. Though the Kawau plants all showed solitary terminal female flowers, binate flowers are sometimes met with. Mr. H. Carse has sent me one or two such flowers, and I have one fruiting specimen with binate immature capsules. The facts set forth above seem to show that the flowers of this Pittosporum are not truly polygamous. The conclusion that the plants are never truly terrestrial seems also devoid of warrant, for the great majority of the plants seen at Kawau grew in clay soil, though certainly close to a rocky shore. 2. Notospartium glabrescens sp. nov. (Plate LVII, fig. 1.) Arbor subhumilis ramosus N. Carmichaeliae Hk. f. similis; differt truncis crassioribus; rhachide pedicellis et calyce glabris; floribus paullo majoribus purpureis; leguminibus multo crassioribus, 15–25 mm. longis ± 4 mm. latis, subteretibus oblongis ± coriaceis subacutis, breviter apiculatis, haud torulosis; seminibus subreniformibus haud complanatis, 2 mm. longis 1.75 mm. latis, rubris, punctibus atris ± maculatis. A small round-headed leafless tree 15–30 ft. (4.5–9 m.) high, usually with several trunks up to 8 in. (2 dcm.) in diameter springing from the ground, in habit not unlike a weeping-willow, the lower branches and branchlets more or less pendulous, the upper erect or ascending; twigs flattened striate, often closely placed on the branchlets; stems, branches, and older branchlets terete. Inflorescence usually erect or ascending, springing from the upper nodes of the twigs; rhachis glabrous, 4–7.5 cm. long, usually many-flowered; flowers like those of N. Carmichaeliae but somewhat larger and purplish in colour; pedicels glabrous, ± 8 mm. long, generally with two minute

bracteoles near their tips; calyx ½–⅔ as long as the pedicels, broadly campanulate with short subtriangular teeth, more or less distinctly ribbed, glabrous or with faint pubescence at the teeth and more or less on the edges between; standard broadly rounded in front, marked by numerous close delicate purplish nerves diverging from the rather broad claw, and with a large purplish blotch above the base covering more than half its upper surface; wings shorter than the keel, oblong, obtuse, narrow-clawed and with a triangulo-hastate expansion at the base opposite the claw. Pods 15–25 mm. long, ± 4 mm. broad, oblong, a little flattened or semiterete, subacute, shortly apiculate, not torulose, more or less wrinkled and marked by obvious distant divaricating veins; seeds subreniform, not flattened, 2 mm. long, 1.75 mm. broad, red when mature, more or less mottled with small black spots. Hab.—Awatere Valley, Marlborough: T. Kirk! Mouth of Clarence River: G. Stevenson! Throughout the upper basin of the Clarence River and its tributary valleys: B. C. Aston! I am deeply indebted to Mr. B. C. Aston for a fine series of specimens of this plant which he collected in 1915, the fruiting pieces in April and the flowering in December. The pods in his specimens are, however, still immature. Mr. G. Stevenson also deserves my warmest thanks for very fine flowering and fruiting specimens gathered near the Clarence Bridge. These show the mature pods. I have put off describing this species for several years, as I was long uncertain whether it might not prove to be a form of one of the species already described. Thanks to the much-valued help of Dr. L. Cockayne, and Mr. H. H. Allan of Ashburton, good specimens of the pods of N. Carmichaeliae and N. torulosum have now been available for study, with the result that I am satisfied that the present plant is quite distinct from both. 3. Observations on the Genus Notospartium. (Plate LVIII.) The accurate investigation of this genus has been greatly delayed by the very incomplete material that has been available for examination by local botanists. The species appear to be nowhere plentiful, and some of them are confined to small areas more or less difficult of access. In my view the genus contains at least three well-marked species, the habitats of which seem nowhere to overlap—N. Carmichaeliae Hk. f., N. torulosum T. Kirk, and N. glabrescens, a new species described on page 366. The flowers of these species are all very much alike in size and form, and one has to fall back on the pods for characters that can be depended on in distinguishing the species. N. Carmichaeliae appears to be confined to the valleys of tributaries flowing northwards into the Wairau River, and the Awatere Valley immediately to the south of these. It is distinguished when flowering by the pubescent rhachis, pedicels, and calyx, and by the pink colour of the flowers. Its pods were apparently first collected by Mr. Teschemaker a few years ago at the Avon River, a tributary of the Waihopai, a district in which only N. Carmichaeliae is known to grow. These fruiting pieces were sent to Dr. Cockayne, from whom I received about a dozen of the pods. They are thin, narrow, more or less curved, and much flattened. I consider it certain that the pods figured in the Botanical Magazine as those of N. Carmichaeliae do not truly belong to that plant; they may be immature pods of N. glabrescens. These pods were figured from material

collected by Waitt in the northern part of North Canterbury, but it is now almost certain that the species does not extend even so far south as the Clarence Valley. N. torulosum T. Kirk is, I think, a valid species. It has the widest distribution of all the species of the genus, extending from Amuri County and Mason River to Mount Peel in the Canterbury Alps. In this species the inflorescence is quite glabrous, the petals are purple, and the pods long, narrow, strongly torulose, almost square in cross-section, and produced into a fairly long slender apical bristle. The seeds are larger than those of N. Carmichaeliae. This is doubtless the plant of which, in its flowering state, Dr. Cockayne has given a minute description in Trans. N.Z. Inst., vol. 49, p. 59, 1917. The third species, N. glabrescens, is more distinct from the two others than these are one from another. So far as is at present known, it occurs only in the Clarence basin, coming down almost to sea-level at the Clarence mouth. It reaches the dimensions of a small tree, and has much thicker trunks than its congeners. Some specimens are as much as 30 ft. high, with trunks 8 in. in diameter. The inflorescence is glabrous, the petals are purplish, the pods are not torulose and are much stouter than those of the allied species. In order to make it easier to obtain further material, and especially flowers and fruit from the very same plants, the places from which the specimens in my herbarium came are set out below:— N. Carmichaeliae.—Awatere Valley (J. Stevenson); Upper Awatere Valley (T. Kirk); Avondale, near Renwicktown (H. J. Matthews); Omaka River, near Blenheim (B. C. Aston); Avon River, tributary of Waihopai (Mr. Teschemaker). N. torulosum.—Mason River, south-east Nelson (L. Cockayne); Whaleback, Amuri County (H. J. Matthews); Mount Kautu (back of), Waipara watershed (R. M. Laing); The Point, Rakaia Gorge (A. Wall); river-terrace scrub, Mount Peel (H. H. Allan); Lynn Stream, Mount Peel (R. M. Laing). N. glabrescens n. sp.—Clarence mouth (G. Stevenson); Swale River, Clarence Valley (B. C. Aston); Nidd Valley, Clarence Valley (B. C. Aston); Dee River, Clarence Valley (B. C. Aston); Mead Valley, Clarence Valley (T. Kirk—this specimen was sent me named “N. Carmichaeliae”); Mead Gorge, Clarence Valley (B. C. Aston); Ure River, Clarence Valley (B. C. Aston). All the species of the genus flower late in December or early in January, according to the altitude of the station; the pods are not ripe till well on in the following year. Since this paper was written Mr. James Stevenson has sent me flowers and pods of N. Carmichaelia from the same plant. These pods exactly match that shown in Plate LVIII, fig. 3. 4. Coriaria thymifolia var. undulata var. nov. A typo differt foliis tenuioribus ac secundum margmes emorso-undulatis, floribus minoribus. Hab.—Both flanks of the Kaimanawa Range: B. C. Aston! Te Whaiti (Whakatane County), c. 1,500 ft. Mr. Aston writes me that this is the only form of the species that grows on the Kaimanawa Mountains. The edges of the leaves look as if a small insect had made a regular series of closely-placed bites all round.

Fig. 1.—Notospartium glabrescens in flower, Nidd Valley, Clarence River, Marlborough. Fig. 2—Notospartrum Carmichaeliae, Tynterfield, Wairau Valley. (Inserted for comparison.)

Fig. 1—Pods of Notospartium glabiescens, collected near Clarence Bridge by Mr. G. Stevenson. The actual length of the specimen here figured is 3 ½ m. (9 cm., nearly). Fig. 2—Pods of Notospartium torulosum, collected at Lynn Stream, Mount Peel, by Mr. H. H. Allan. Fig. 3—Pods of Notospartium Carmichaeliae, collected at Avon River, a tributary of the Waihopai, Marlborough, by Mr. Teschemaker. All the figures in this plate are shown on the same scale

5. Epilobium nerterioides A. Cunn. Most local botanical workers refer A. Cunningham's Epilobium nerterioides to E. nummularifolium R. Cunn. as a variety. The seeds of E. nerterioides, however, always have a smooth testa, whereas the testa is papillose in all the forms of E. nummularifolium. As in addition to this constant difference the vegetative characters are also fairly distinctive, it seems to me that E. nerterioides should be considered a valid species. Its leaves are usually quite entire, though occasionally slightly sinuate at the edges, are generally marked by irregular shallow fairly-wide semidepressions on the upper surface, and when fresh show a general green hue more or less mottled with pale yellow. Its stems are prostrate, creeping and rooting, little branched, and up to 10 cm. long, though commonly shorter. The plants form more or less matted tufts from 8 cm. to 13 cm. across. There are two prevalent forms: one, occurring on damp sandy soil, has small rather thin orbicular leaves somewhat distantly placed, fairly long peduncles that may reach 5 cm. when in ripe fruit, and rather long capsules (= 2.5–3 cm.); the other form has the leaves closer, more coriaceous, larger, and longer (sometimes broadly elliptic and slightly reflexed with shallow sinuation at the edges), and peduncles equalling the short rather stout capsules. I find no variation in the size or form of the seeds. Mr. Cheeseman's variety angustum of E. nummularifolium and Mr. Kirk's variety minimum both belong to E. nerterioides. The species is widely spread throughout the Dominion, and occurs also on the Subantarctic Islands of New Zealand. I have examined specimens from Peria and Fairburn (Mongonui County), Mercury Bay, Gordon's Knob (Nelson), Cass River (Lake Tekapo), Speargrass Flat (Vincent County), Pembroke (Lake Wanaka), Ashburton, Mount Cargill (Dunedin), and Fortrose and Bluff (Southland). 6. Note on Epilobium antipodum Petrie. For a considerable time I have been satisfied that my Epilobium antipodum is no other than E. crassum Hook. f. The latter was in cultivation for a year or two in my garden, and chance seedlings of it grew up where the seeds of the Antipodes Island plant had been sown some considerable distance away. Satisfactory foliage-bearing pieces of the island plant had not been seen, hence the regrettable failure to recognize what had happened. 7. Epilobium Matthewsii n. sp. This name is proposed for my Epilobium arcuatum, a combination that now proves to have been preoccupied when the original name was published (Trans. N.Z. Inst., vol. 45, p. 266, 1913). The plant was only coming into flower when I visited the Clinton Valley. Subsequently the late Henry J. Matthews collected a few ripe capsules, but no entire plants. Though the plant is still very imperfectly known, I consider it one of the most distinct of the native species. The unavoidable change of name affords me a welcome opportunity to commemorate the services of an enthusiastic student of the native flowering-plants of our Dominion. 8. Aciphylla Poppelwelli sp. nov. Planta A. Traillii T. Kirk subsimilis, differt inflorescentia principali recta (haud flexuosa), bracteis floralibus pernumerosis (nonnunquam 40 v.

ultra) arctissime confertis trifoliolatis, umbellis quam bracteae multo brevioribus vaginas subtumidas vix excedentibus, fructibus paullo majoribus oblongis, apicem versus ± contractis. Small, 12–22 cm. high, with scapes solitary or occasionally two from the same main root. Leaves 6–12 cm. long, ± 3.5 mm. wide, linear, pungent-pointed, subcoriaceous scarcely stiff, striate, thickened along the margins, midrib inconspicuous; petiole short, gradually dilated downwards into a broad membranous subhyaline sheathing base. Scapes rather stout for the size of the plant, ± 3.5 mm. across, the naked part (as long as the leaves) supporting a much longer spike-like main inflorescence with very numerous (sometimes 40 or more) densely-crowded floral bracts enclosing the umbels, bracts ± 3 cm. long trifoliolate with rather short slightly tumid sheaths; umbels both male and female compactly branched, much shorter than the bracts. Fruit oblong, rather large, more or less contracted at the top. (A. Trailii Kirk, Students' Flora, p. 210, pro parte; also A. Traillii Kirk, Cheeseman's Manual, pp. 211–12, pro parte.) Hab.—Mount Kyeburn (Maniototo County), 3,000 ft.; Rock and Pillar Range (Taieri County), 3,800 ft.: B. C. Aston! Arnold Wall! Garvie Mountains (Southland), 4,000 ft.: D. L. Poppelwell! Dr. L. Cockayne! The Garvie Mountain plant is taken as the type. Var. major var. nov. Elatior, scapo nudo quam in typo ter quaterve longiore, inflorescentia principali scapo nudo multo breviore, bracteis umbellisque paucioribus, bractearum vaginis brevioribus paene aeque latis ac longis. Hab.—Mount Buster (part of Mount Ida Range, Maniototo County), 3,500 ft. I have authentic specimens of A. Traillii from Mr. T. Kirk, collected on Mount Anglem, Stewart Island, and others that exactly match these, collected in the same place by Mr. W. R. B. Oliver. In the Mount Anglem plant the floral bracts are placed far apart; the inflorescence is markedly zigzag, the bracts being seated on the bends; the bracts are long, few (5–8), and nearly always simple; the midribs are very prominent, and the space on the underside of the leaves between the midrib and the thickened margins is of a dull-brown colour; the male umbels have long delicate branches that greatly exceed the elongated sheaths enclosing them; the fruits also appear to be smaller than in the present species. 9. Note on Veronica Willcoxii Petrie. Though this plant has been in cultivation for several years in the alpine garden of the late H. J. Matthews (now Dr. Hunter's) at Mornington, Dunedin, I have never succeeded in getting specimens in flower or fruit from there. For a few years two plants grew in my garden in Auckland, but they languished season by season, and died without flowering. The flowers and capsules described by me from plants growing in the University grounds at Dunedin most likely belong to this species, but it is highly desirable that flowering and fruiting pieces should be got from plants known to have come from the Lake Harris habitat, where alone the wild plant has so far been found. For the present there must remain some doubt as to whether the flowers and capsules ascribed to the species truly belong there

10. Veronica angustifolia A. Rich. var. abbreviata var. nov. Racemi foliis breviores 2–3 cm. longi, 1.2 cm. lati, obtusi, dense multiflori; folia quam in forma typica subbreviora; capsulis haud visis. Valley of the Ure River, Marlborough: B. C. Aston! Collected early April, 1915. Though collected very late in the season, Mr. Aston's specimens are in full flower. When ripe capsules can be examined this plant may be found to deserve specific rank. 11. Carex Wallii sp. nov. Planta humilis laxe caespitans, in locis humidis v. uliginosis crescens. Folia pauca filiformia flaccida plana v. leviter complicata striata apices versus delicatule scaberula, 6 cm. longa v. breviora; vaginis valde tenuibus ± striatis in ligulam latam truncatam desinentibus. Culmi folia longe excedentes suberecti filiformes ± trigoni flaccidi nudi ad 12 cm. longi. Spiculae solitariae terminales parvae subovatae ad 5 mm. longae ebracteatae, floribus superioribus masculis, inferioribus (ad 6) femineis; florum femineorum glumis membranaceis late ovatis subacuminatis 1-nerviis pallide viridibus, marginibus ± scariosis, gluma infima nonnunquam bractiformi. Utriculi glumas excedentes 2 mm. longi semiteretes v. late biconvexi ovati subpaullo alati, dorso leviter 5-nervii, a basi ± rotundati, supra gradatim in rostrum gracile vix longum integrum abeuntes. Styli rami 3. Nux ± triquetro-biconvexa. A more of less matted slender plant, growing in wet or damp spots. Leaves, few, filiform, flaccid, flat or more or less folded striate, finely scaberulous towards the tips, 6 cm. long or less; sheaths very thin, finely ribbed, and ending in a broad truncate ligule. Culms much exceeding the leaves, suberect, usually more or less curved, glabrous, filiform, more or less trigonous, flaccid, 12 cm. long or less. Spikelets solitary, terminal, small, ovoid in outline, ebracteate, 5 mm. long or rather less; the upper flowers male, the lower (6 or fewer) female; glumes of the female flowers membranous, broadly ovate, subacuminate, pale green, the edges more or less scarious; the lowermost glume occasionally produced into a bractiform elongation. Utricles longer than the glumes, ± 2 mm long semiterete or broadly biconvex, slightly winged, rather faintly 5-nerved on the back, ovate rather wide near the more or less rounded base and gradually narrowed above into a slender short entire beak; style-branches 3; nut more or less triquetrously biconvex. Hab.—Wet ground at Centre Hill, Southland: Arnold Wall! Collected February, 1920. I have not seen much material of this plant, and most of the specimens were over-mature. It is very distinct from any other native species of Carex, but its position can hardly be determined with certainty until fuller material is available for examination. The present description may then prove to need amendment in some details.