Notes on New Zealand Floristic Botany. Including Descriptions of New Species, ETC.—No. 7.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 69, 1940, Page 270

Notes on New Zealand Floristic Botany. Including Descriptions of New Species, ETC.—No. 7. By H. H. Allan, Plant Research Bureau. [Read before the Wellington Philosophical Society, October 12, 1938; received by the Editor, March 7, 1939; issued separately, September, 1939.] I. Taxonomic. 1. Aciphylla leighii n.sp. Affinis A. dobsoni, sed differt statu minore, foliis vix rigidis, non pungentibus, umbellis vix capitatis. Herb, forming patches up to 2 m. diam. Rosettes in flower, c. 6 cm. tall, with spread of c. 7 cm. Tap root stout c. 6 mm. diam., remaining so for at least 6 cm., with numerous descending lateral rootlets. Leaves c. 5 cm. long, standing at c. 45° with horizontal. Sheathing bases c. 3–5 cm. long, white with thin hyaline margins, becoming yellowish-green on back towards leaflets; concave, thin below, thickening upwards and curving outwards, striate and shallowly grooved on back; c. 1 cm. wide at base, narrowing to c. 6 mm. at base of leaflets; thickened upper portion spongy, with copious milky juice from dorsal vessels; leaflets 3, diverging, triangular in projection, the two lateral facing each other, the central slightly longer. Segments c. 1.6 cm. long, c. 1 mm. thick at midrib, c. 4 mm. wide, shallowly channelled above; dark green, margins cartilaginous, about linear-oblong, convex on back, distinctly veined, cross veinlets evident, slightly constricting margins; apices yellowish, blunt to ret use, with short mucro. Peduncle central, c. 5 cm. long, striate, surmounted in male by rather dense compound umbel c. 4 cm. across; primary rays 3 to 4, with general involucre of leaf-like bracts; rays c. 4 cm. long, unequal, diverging, somewhat flattened, thickened slightly upwards, ending in umbels, or again branched and enclosed in sheaths of bracts; secondary umbels c. 2.5 cm. across, with leaflike involucral bracts, or with only central leaflet developed, or reduced to sheaths; rays 7 to 10, c. 6 mm. long, ending in 3 segments with finely acicular tips; bracteoles scalelike. Final umbellules of c. 20 flowers, the filiform bracteoles exceeding the flowers. Sepals greenish, incurved, linear, acute, enlarging after petal fall to 2 mm. long, concave, narrowed to short claw; stamens exserted at anthesis, with filaments up to 3 mm. long, disc yellowish. Female plants not seen. Mount Milne, Darran Range, Western Otago (Fiord Bot. Dist.) at c. 1980 m. D. Leigh! Type at Herbarium, Plant Research Bureau, Wellington. Mr. Leigh provides the folio-wing interesting note on his discovery of this is notable addition to the genus Aciphylla, specimens of which are now growing in the scree garden of Mr. J. Scott-Thomson: “The Darran Range occupies nearly all the west coast territory of Otago, “and is isolated from all neighbouring alpine country principally by the deeply cut valley of the Hollyford River. The

northern section is connected with the Central Darrans by a narrow, precipitous barrier of peaked ridges coming off Mount Madeline and bearing due south to Mount Underwood between Tutoko River and the Lower Hollyford. It was on these intervening peaks that I collected this distinct species of Aciphylla at an altitude of 6,500 feet, beneath the summit rocks of the peak which the late Samuel Turner called Mount Milne. Growing on stabilized scree terraces and forming circular patches sometimes six feet in diameter, this Aciphylla provided a unique appearance. The habitat is very exposed, being directly opposite the massive ice-face of Mount Tutoko, and above a branch of the Madeline Ice Plateau. There were no signs of this plant on the north-east face of Tutoko, and I am of the opinion that the snowfield on Mount Underwood would prevent its dissemination over the extensive central Darrans. The plant did not occur anywhere below the Madeline Ice Plateau.” 2. Coprosma. A remarkable “sporting” Coprosma (plate 23) has been observed on Mayor Island, Bay of Plenty, by Misses L. B. Moore and J. Wakelin. The following account has been drawn up from notes and specimens kindly sent me. The species known to occur on the island are: C. australis, C. lucida, C. macrocarpa (the North Auckland form), C. repens (baueri) and C. robusta. The “C robusta var. angustata Kirk?” listed by Allan and Dalrymple (1926, p. 36) may have been gathered from one of the sporting plants. These were noted by the discoverers on the south-east slopes in second-growth scrub of Leptospermum, with Knightia excelsa, Suttonia australis and other species, and on the scrub margins. They were about as plentiful as C. robusta. They reached 3 m. in height, and were always in a sufficiently open station to be leafy almost to the ground, at least on one side. The sport is characterized by bearing branches carrying leaves of two forms and sizes. Some branches have large leaves, somewhat of robusta size, shape and texture, but slightly more coriaceous and lustrous, and usually with a mucronate tip. Average dimensions are 8 cm. by 3.5 cm. Other branches bore leaves similar in texture, but much smaller and narrower, rather like C. lucida var. angustifolia Cheeseman. Average dimensions are 4 cm. by 1.25 cm. “Specimens like that shown in the photograph were never difficult to find, but often the two parts of the plant separate out at one of the lower forks, so that at a glance one would think that there were two distinct plants. Though the big-leaved form is plentiful, we found no case where a plant had nothing but small leaves. Both parts of the plant fruit, but the smaller-leaved part more abundantly. Fruit, though green, was heavy and conspicuous. The flowers were often ☿, with persistent perianths.” Unfortunately, no ripe fruit was obtained, so that progeny tests have not yet been carried out. 3. Coriaria kingiana Colenso. Although this name was effectively and validly published by Colenso (1844, p. 20) it has been ignored by all subsequent writers. He states of it, “an elegant procumbent plant with undulated ovate-acuminate leaves. It seldom rises above two feet in height, and is

mostly found quite prostrate, and very abundant.” Petrie (1921, p. 368) described his var. undulata of C. thymifolia as “a typo differt foliis tenuioribus ac secundum margines emorso-undulatis, floribus minoribus,” and appropriately remarked that the edges of the leaves look “as if a small insect had made a regular series of closely-placed bites all round.” The taxonomy of Coriaria is in an extremely unsatisfactory state, but there is now general agreement that the true C. thymifolia does not occur in New Zealand. Kirk's name C. lurida is often used to apply to all the New Zealand “thymifolia” forms, though he himself (1899, p. 98) specially separated out his lurida from the rest of the complex. Colenso's name has priority for this group, if it is not to be retained for the special undulate variety described. But the whole Coriaria complex needs fresh investigation. It is now known that undnlate form also occurs in the “ruscifolia” or “sarmentosa” group, in the “parvifolia” group, and is transferred by hybridism to the “arborea” group (cf. Allan, 1930, p. 446). 4. Dracophyllum rosmarinifolium (Forst. f.) R. Br. This is based on specimens collected by Forster at Dusky Sound. Hooker (1864, p. 183) included under it what is now known as D. pronum W. R. Oliver, and described as new D. uniflorum, with pungent leaves, from specimens collected in various parts of South Island. Oliver (1928, p. 689) rightly separated Epacris rosmarinifolia Forst. f. from D. pronum, but treated D. uniflorum as a synonym of D. rosmarinifolium. His action was largely based on the views of Professor A. Wall, whom he quoted as follows:“… Forster's type. There is one small twig only with no fruit or flower … The only other specimen there [British Museum] is one of Kirk's from Amuri, which is exactly the same plant as Forster's. It was plain to me at once that Forster's plant (and Kirk's) is simply what we now call D. uniflorum … Dr. Rendle agreed with me that they, specimens of D. uniflorum, perfectly match the D. rosmarinifolium of Forster.” Oliver added, “as pointed out by Hooker the characters by which this species may be recognized are the short pungent leaves and the large solitary flowers.” My examination of the material at the British Museum and at Kew lead me to a somewhat different conclusion. The leaves in the type specimen of Epacris rosmarinifolia are from 15 to 25 mm. long, slightly over 1 mm. wide, the upper third of the blade trigonous, narrowed rather abruptly to the tip, the lower part channelled; the sheaths not shouldered. Taking Travers's Wairau mountain specimens as the type of D. uniflorum, we find: leaves 30 to 40 mm. long, hardly 1 mm. wide, the upper half trigonous, narrowed gradually to the acute apex, very shallowly channelled below; sheaths abruptly shouldered. Kirk's Amuri specimens agree very well with D. uniflorum. Oliver (loc. cit., p. 690), however, recognizes that these differences exist, mentioning two forms, “apparently not depending on habitat”: (a) leaves less than 1 mm. wide, up to 35 or 40 mm. long—“Tararua Ranges, Nelson, and Canterbury”; (b) leaves up to 25 mm. long and 1.5 mm. broad—“Otago (Mount Barber).” He adds, “Forster's type agrees with this broad-leaved form.”

That D. rosmarinifolium and D. uniflorum are different is clear. The evidence is all against either being a “habitat form” of the other. Whether D. uniflorum should be accepted as a species or given the status of a variety of D. rosmarinifolium cannot be decided till a full comparison of flower and fruit characters has been made. Dr. Oliver has kindly given me specimens of the Mount Barber plant. These have rather stouter leaves than those of Forster's type, but otherwise have the same characteristics. 5. Halorrhagis colensoi Skottsberg. This species was described by Skottsberg (1922, pp. 152, 153, 156) from a collection by Colenso, present among a set of H. erecta in the Kew Herbarium. But though figured and validly described, it was not referred to by Cheeseman (1925). It was rediscovered by Mr. T. Latta on Mount Oporae, near Weber (Ruahine-Cook Bot. Dist.) and specimens were sent to Professor Skottsberg, who has supplied the following fuller diagnosis: “Suffrutex laevis glaberrimus, H. masatierranae Skottsb. simil-limus. Folia ovata usque lanceolata, acuta, basi cuneata, lamina 10–15 mm. et ultra, longa et 6–8 mm. lata, argute serrata dentibus utrinque 5–7, petiolo 3–4 mm. longo suffulta. Flores subsessiles. Achaenia ovoideo-coniea, 2.9-3.3 mm. longa (sepalis inclusis) et 1.7-2 mm. diam., tetragona, laevia, sed in parte dimidia inferiore transverse tuberculato-rugoso, exalata vel alis 4 angustis triangulatis parvis nec non humilibus infra medium positis ornata. Cetera ignota. Hab. Novae Zelandiae ins. bor. (Colenso in herb. Kew); in collibus Puketoi dictis, commun. H. H. Allan. As pointed out (l.c.) H. colensoi is very close to H. masatierrana, from which it differs mainly in the smaller leaves and fruits, which, in the latter, lack the tubercles and are perfectly smooth; but it should be noted that of H. masafuerana (l.c., p. 156) I have forms with or without tubercles. Petals, stamens and stigmas will probably show some differences, and in any case I do not find it possible to refer H. colensoi to masatierrana, or to any other species I know of. The close affinity between the Halorrhagis species of New Zealand and Juan Fernandez is of considerable interest. We have, in fact, two pairs of vicarious species: H. colensoi—H. masatierrana and H. erecta—H. asperrima, whereas no Halorrhagis has ever been found in South America.” 6. Hebe raoulii (Hook, f.) Ckn. et Allan var. maccaskillii nov. A typo differt ramis gracilioribus, elongatis; foliis minus coriaceis, interdum vix dentatis, floribus manifeste pedicellatis. Weka Pass, Canterbury, on limestone, L. W. McCaskill! White Rock, on limestone, R. M. Laing! Type in Herbarium, Plant Research Bureau, Wellington. This graceful plant is easily distinguished from the type by the slender, rambling habit, the less coriaceous leaves with few or no teeth, the floral ones suborbicular, the longer racemes, the flowers definitely stalked. It appears to be confined to limestone. An interesting feature is the presence on many flowers of a scale-like fifth sepal. In exposed places the stems lie flat on the rock, while amongst

Sophora prostrata and other shrubs they become semi-lianoid. I am indebted to Mr. McCaskill for a large suite of specimens and for field note on its habit and occurrences. 7. Hebe scott-thomsonii n.sp. H. traversii et H. brachysiphon affinis, sed sepala 1–1.5 mm. longa; corollae tubus brevis; sursum aliquatenus dilatatus, sepala vix aeqiuns; capsula minute pubescens, 5–6 mm. longa, sepalis 3–4-plo longior. South and North Otago Bot. Dists. By streams in tussock grassland Deep Stream, Lammerlaw Hills, Rock and Pillar Range, Macrae's Hill, Upper Waikouaiti River, Mount Hyde, Stoneburn. Type in Herbarium, Plant Research Bureau, Wellington. Shrub up to 1 m. tall; leaves oblong-obovate, rather abruptly acute, to narrow-oblong subacute, up to 2.5 cm. long and 0.8 cm. broad, slightly concave, drying yellowish brown; racemes slender, c. 4 to 5 cm. long, elongating in fruit; pedicels c. 2 mm. long, pubescent; bracts ovate to linear, variable in size; calyx segments ovate, obtuse or occasionally subacute, ciliolate; corolla tube very short, widening upwards, hardly attaining the apices of the calyx lobes; corolla lobes broadly obovate to subrotund, c. 2 mm. long and broad; capsules finely pubescent, 5 to 6 mm. long, c. 3 to 4 times length of calyx. The “traversii” complex has long been a difficulty to students. Summerhayes (1927, p. 395) established the type as being the plant with the corolla tube two to five times the length of the calyx. From the short-tubed forms he selected the species figured by Hooker in Bot. Mag., t. 6390, and gave it the name of H. brachysiphon. He includes a specimen gathered by Petrie at Lake Wakatipu (South Otago Bot. Dist.), the other localities cited being in the Sounds-Nelson, North-eastern and Eastern Bot. Districts. H. traversii appears to be confined to the Eastern District. The three species thus provide an interesting example of geographical distribution. Possibly all should be united as varieties of H. traversii in a wide sense, but other members of the complex need examination before a full revision can be made. The account of cytology within the complex promised by Frankel and Hair (1937, p. 685) should present important data in this connection. The following synopsis may be useful: H. traversii—calyx 1.5 mm. long; corolla tube slender, cylindric, 2–5 times length of calyx; capsule 5 mm. long, 3–4 times length of calyx. H. brachysiphon—calyx 2–3 mm. long; corolla tube rather broad, widening upwards, equalling or up to twice length of calyx; capsule 5–6 mm. long, 2–2 ½ times the length of calyx. H. scott-thomsonii—calyx 1–1.5 mm. long; corolla tube rather broad, widening somewhat upwards, hardly equalling calyx; capsule 5–6 mm. long, 3–4 times length of calyx. This species is further referred to by Simpson and Scott-Thomson (1938, p. 435).

8. Nertera balfouriana Ckn. The flowers have not hitherto been described. The following is drawn up from specimens collected on Mount Mistake, Lake Tekapo, in early January. Flowers markedly protogynous. Calyx truncate, obsolescent. Corolla 4-lobed; tube 2–25-3 mm. long, glabrous, funnelform; lobes bluntly triangular, reflexed, papillose on margin. Styles projecting beyond anthers, recurved, united in basal third. Stamens c. 5 mm. long, erect, often dehiscing before extending beyond corolla lobes and pollinating lower part of styles. Much pollen is caught, on the corolla lobes. 9. Veronica. The question here is whether Veronica in the strict sense, i.e. the subgenera Veronicella and Euveronica, is represented among indigenous plants. Veronica L. and Hebe Comm. are separated taxonomically as follows: Veronica—capsule dehiscing loculicidally, its walls and septum thin; herbs, the stems dying with the leaves. Hebe—capsule dehiscing septieidally, the thick septum splitting and each carpel opening distally by a median slit through the septal wall. Hooker (1854, p. 190) under Veronica says: “Capsula compressa, bisulcata; carpellis dorso loculicide dehiscentibus, marginibus inflexis columnae placentiferae adhaerentibus v. abea septicide solutis; v. capsula septicida cum columna placentifera bipartibilis.” In the loculicidal group he places V. catarractae and its allies. In his later treatment Hooker (1864, p. 204) makes the same groups, separating them by (1) “capsule dorsally compressed, ovoid, turgid,” and (2) “capsule laterally compressed, didymous.” He gives, however, the capsule dehiscence as septicidal, and makes no reference to loculicidal dehiscence. Cheeseman (1925, p. 776) states that the dehiscence is either septicidal or loculicidal. For his Division Hebe he gives, “Capsule turgid or dorsally compressed, the septum across the broadest diameter,” and for his Division Euveronica, “Capsule laterally compressed, the septum across the narrowest diameter.” Pennell (1921) revived Commerson's genus Hebe on the grounds: (1) dehiscence of capsule septicidal, (2) leaves opposite throughout, (3) flowers always in axillary racemes, (4) all shrubs or trees, (5) distribution austral, “with its suggestion of genetic remoteness.” He adds, “Moreover, Hebe has an exceedingly baffling tendency to form local races, a habit at contrast with that of the other Veronicas’.” The New Zealand species transferred are H. salicifolia, H. blanda and H. elliptica. Cockayne and Allan (1926) accepted Pennell's action, and made the necessary new combinations. They, however, left Cheeseman's division Euveronica in Veronica sens. strict. Frankel and Hair (1937) examined the cytology of a number of New Zealand species. For Hebe they found the chromosome numbers: 20, 21, 40, 42, 60; and for Veronica 20, 21. They point out that in Veronica of the Northern Hemisphere the numbers 4, 8, multiples of 7, 9, 17 have been recorded. In V. plebeia, almost certainly an introduced plant

in its New Zealand occurrences, the number is also 17. They conclude: “From the cytological evidence it is apparent that the genus Hebe and the New Zealand Veronicas have closely allied chromosome complements which are equally far removed from those of the Veronicas endemic in the Northern Hemisphere. The austral distribution of the New Zealand Veronicas is a further characteristic linking them closer to the genus Hebe than to the Northern Veronicas. Therefore, although from the viewpoint o1 systematic convenience the New Zealand Veronicas belong to Veronica, according to the cytological evidence they must be grouped with Hebe. Here, then, we have probably the first example in which cytological findings are clearly contradictory to the systematic evidence.” In view of this I have examined the capsule dehiscence of those species of indigenous and introduced Veronicas of which suitable material was available. The ultimate result of the dehiscence is much the same in both. In Hebe the dehiscence is as described by Pennell (loc. cit., p. 2):“the carpels part, thus splitting sagitally the septum, after which a distal median suture through the septal wall of each carpel permits the seeds to escape.” In all the examples studied by me loculicidal stage commences early, before the septicidal dehiscence is complete. In Veronica (introduced species) the carpels split loculicidally, and later separate septicidally as noted by Hooker. The septicidal stage is usually long delayed and often imperfectly developed. Often the seeds are all shed before the septieidal split becomes evident. In the New Zealand species placed by Cheeseman in Euveronica, so far as I have been able to examine them, the dehiscence is essentially as in Hebe, though occasional capsules show a slight loculicidal split before the septicidal dehiscence has developed far. The septum is narrow, but the capsules are rather thick-walled and often turgid, and often not markedly didymous. The plants are small shrubs, or at least woody towards the bases, the stems not “dying with leaves.” The group thus occupies a somewhat intermediate position, but is much closer to Hebe proper than Euveronica, If, as on most grounds seems preferable, Hebe is maintained as a separate genus, I agree with Frankel and Hair that the species of “Veronica” mentioned below should be transferred to Hebe. It is very desirable that the remaining species, especially the herbaceous V. canescens (of which all my specimens are in young flower) be examined. V. anagallis is a true Veronica, but the evidence is all against it being indigenous. I, therefore, make the following new combinations:— Hebe bidwillii (Hook.) Allan, comb. nov. Veronica bidwillii Hook., Icon. Plant., 1848, t. 814. Hebe catarractae (Forst. f.) Allan, comb. nov. Veronica catarractae Forst. f., Flor. Ins. Aust. Prodr., 1786, p. 3. Hebt hookeriana (Walp.) Allan, comb. nov. Veronica hookeriana Walp., Rep., III, 1844, p. 341. Hebe linifolia (Hook, f.) Allan, comb. nov. Veronica linifolia Hook, f., Handb. N.Z. Flora, 1864, p. 214. Hebe lyallii (Hook, f.) Allan, comb. nov. Veronica lyallii Hook, f., Fl. Nov.-Zel., 1, 1854, p. 196.

II. Phytogeographic. 1. Eleocharis gracilis R. Br. Mr. S. Blake, of the University of Queensland, who has kindly examined my specimens, finds that there is no specific difference between the E. cunninghamii Boeck. of New Zealand and the E. gracilis R. Br. of Australia. In both countries this species shows considerable habitat modifications, and the varietal names so far applied are of very doubtful status. 2. Eleocharis pusilla R. Br. E. acicularis R. Br. was admitted to the New Zealand flora by Cheeseman (1925, p. 217) on immature specimens collected by Petrie at Lake Te Anau. He remarks: “Mr. C. B. Clarke, who has examined them, states that he is satisfied that they belong to the small group consisting of E. acicularis and a very few closely allied species, and most probably to E. acicularis itself.” Mr. Blake places specimens sent by me under E. pusilla R. Br. It is represented in the herbarium of the Plant Research Bureau, Wellington, by specimens from: Lake Taupo (Volcanic Plateau Bot. Dist.) at water's edge, K. W. Allison!; Foxton (Ruahine-Cook Bot. Dist.) in damp sand hollows among dunes, V. D. Zotov!; Lake Tekapo (Eastern South Island Bot. Dist.), margins of tarns and moraine. 3. Gaultheria colensoi Hook. f. This species was rightly revived by Burtt and Hill (1935, p. 620). The stations given by them are: base of Mount Tongariro, near Lake Rotoaira, and Mount Ruapehu. Mr. N. Elder, who has devoted great attention to the Kaimanawa and adjacent ranges, found it on the Kaweka Range (at 900-1,200 m.), Black Birch Range (at 1,100 m.), and the southern Kaimanawas (at 1,370 m. and more).He finds that it usually occurs on rocky or shingly faces, flowering in December or January. 4. Hebe acutiflora (Benth.) Ckn. Mr. K. W. Sexton has forwarded excellent specimens of this, well matching the type from Kerikeri, collected near Herekino, also in the North Auckland Bot. Dist. 5. Isolepis crassiusculus Hook, f. Mr. Blake so identifies specimens, in ripe fruit, collected by me in bogs above the forest-line on Mount Hauhungatahi (Volcanic Plateau Bot. Dist.). Possibly some of the records of I. lenticularis R. Br. really referred to this species. Mr. A. L. Poole has recently collected the same species on the Kaimanawas. 6. Poa lindsayi Hook. f. Near Lake Taupo (Volcanic Plateau Bot. Dist.), K. W. Allison! Previously reported for North Island only from the Ruataniwha Plains (Ruahine-Cook Bot. Dist.). 7. Triodia pumila (T. Kirk) Hack. ex Cheesem. Near Lake Taupo, K. W. Allison! Cheeseman (1925, p. 182) gives only South Island stations.

Literature Cited. Allan, H. H., 1930. The Significance of Hybridism in the New Zealand Flora, Rep. 20th Meet. A. and N.Z. Ass. Adv. Sci., pp. 429–477. Allan, H. H.—and Dalrymple, K. W., 1926. Ferns and Flowering Plants of Mayor Island, N.Z., Trans. N.Z. Inst., vol. 56, pp. 34–36. Burtt, B. L., and Hill, A. W. The Genera Gaultheria and Pernettya in New Zealand, Tasmania and Australia, Journ. Linn. Soc. Bot., vol. 49, pp. 611–644. Cheeseman, T. F., 1925. Manual New Zealand flora, ed. 2, Wellington. Cockayne, L., and Allan, H. H., 1926. The Present Taxonomic Status of the New Zealand Species of Hebe, Trans. N.Z. Inst., vol. 57, pp. 11–47. Colenso, W., 1844. Journal of a Naturalist in Some Little-travelled Parts of New Zealand, Lond. Journ. Bot., vol. 3, pp. 1–62. Frankel, O. H., and Hair, J. B., 1937. Studies in the Cytology, Genetics, and Taxonomy of New Zealand Hebe and Veronica (Part I), N.Z. Journ. Sci. and Tech., vol. 18, pp. 669–687. Hooker, J D., 1852–55. Flora Novae-Zelandiae, Part I, London. Hooker, J D.—1864. A Handbook of the New Zealand Flora, Part I, London. Kirk, T., 1899. The Students' Flora of New Zealand, Wellington. Oliver, W. R. B., 1928. A Revision of the Genus Dracophyllum, Trans, N.Z. Inst., vol. 59, pp. 678–714. Pennell, F. W., 1921. “Veronica” in North and South America, Contrib. N.Y. Bot. Gard., no. 230. Petrie, D., 1921. Descriptions of New Native Flowering Plants, with a Few Notes, Trans. N.Z. Inst., vol. 53, pp. 365–371. Simpson, G., and Scott-Thomson, J., 1938. The Dunedin Sub-district of the South Otago Botanical District, Trans. Roy. Soc. N.Z., vol. 67, pp. 430–442. Skottsberg, C., 1922. The Phanerogams of the Juan Fernandez Islands, Nat. Hist. Juan Fern, and Easter Is., vol. 2, pp. 95–240, Uppsala. Summerhayes, V. S., 1927. The Type of Veronica traversii, Kew Bulletin Misc. Inf., 1927, no. 9, pp. 395–398.