Flora of New Zealand Series

Erect to spreading dioec. shrub up to 3 m. tall, us. less; branchlets pubescent when young. Lvs of juvenile plants membr., obovate, toothed or lobed; lvs of adults coriac., on petioles 2-3 mm. long; lamina obovate to elliptic to obovate-oblong, obtuse or retuse, (1)-1·5-4 cm. × 7-15 mm.; margins thickened and slightly revolute, entire to slightly sinuate, rarely with 1 to 3 blunt teeth. Fls solitary or in fascicles, 3-4 mm. diam., on straight to decurved pedicels ± 4 mm. long; bracts below middle. ♂ with ovate sepals, free almost to base; petals pale yellow, revolute at tips; scales narrow-cuneate; membrane with lanceolate, fimbriate processes. ♀ smaller, petals broader. Fr. 3-5 mm. diam., subglobose to ovoid; style 2-fid; seeds (1)-2, plano-convex.

DIST.: N., S. Rocky places, especially coastally; Kapiti Id, both shores of Cook Strait. Kirk loc. cit. gives several localities and the type is .chosen from the specimens in W labelled "T. Kirk, 1885, limestone rocks between Takaka and Riwaka". Later the number 399 has been assigned to this sheet.

Inland plants from Mount Owen, Broken River Basin and the Ashburton Mountains are us. placed under this sp. but need further study.

POLYMORPHY

The genus has been very inadequately studied and a clear separation from Melicytus is still wanting. It is evident that heteroblasty, habitat influence, and hybridism all play a part and cultural and genetic work will be essential for a satisfactory taxonomic grouping. Mueller (Veg. Chath. Is 1864, 9) considered that "In all probability only one sp. of this genus exists, for which the collective name H. Banksii might well be retained". Kirk (Stud. Fl. 1899, 45) said of his H. obovata : "The rather slender branches and the strict habit distinguish this species from all others". Cheeseman (Man. N.Z. Fl. 1906, 48) remarked: "The New Zealand species are exceedingly difficult of discrimination. They vary greatly in the leaves and vegetative characters generally; and the flowers and fruit, so far as they are known, are very similar in all." He repeated the remark in the 1925 edition, having very little further information at his disposal.

Cheeseman (loc. cit. 1906, 50) considered H. obovata "A well-marked plant at once recognised by the usually slender habit, strict branches, and entire obovate leaves". But (Man. N.Z. Fl. 1925, 575) he later said: "Apparently a very variable plant, judging from a personal inspection of it in several of the Nelson and Wellington localities," On the whole he preferred " to regard the differences as being due to the diversified character of the stations which the plant inhabits".

Cockayne (T.N.Z.I. 50, 1918, 176-179) discussed H. crassifolia and H. obovata, but did not decide whether the diversity was due to "polymorphy" or "epharmony". He describes the flowers in some detail. Allan (Genetica 9, 1927, 145) discussed an assemblage of plants at Titahi Bay, Wellington. He considered that both spp. were present and both were modified by the various habitat conditions, but that there was good floral evidence that hybrids were also present. Oliver (Rec. Dom. Mus. Wellington 1, 1944, 211) after examining Australian specimens of H. dentate R. Br. and H. angustifolia R. Br., decided that these spp. did not occur in New Zealand. He therefore raised Kirk's H. dentata var. alpina to specific rank, and included under this name both Kirk's var. and the forms hitherto treated as var. angustifolia (R. Br.) Benth. of H. dentate R. Br. Oliver states: "I have no hesitation in stating that the New Zealand forms fall outside the range found in Australia and Tasmania . . . There is a great difference in appearance between the extremes of the New Zealand forms; but there can be found every intermediate stage. Consequently I do not propose to separate specifically the slender forest form from the rigid scrub form. The species apparently is quite plastic, moulding itself to suit different stations." I have not been able, however, to find any certain characters separating our forest margin plant from Brown's no. 5480, the type of his angustifolia, nor from specimens collected by R. Gunn, 1844, in Tasmania. The question of plasticity can be solved only by experiment.

Sexual expression in the N.Z. spp. has not been fully studied. For the most part they are functionally dioec. but there appears to be a tendency to monoecism. I have not seen definitely hermaphrodite fls.