Celmisia gracilenta Hook.f.
Slender tufted herb with pseudo-stem ± 5 cm. long. Lamina subcoriac., flexible, rather flaccid, ascending, (5)-10-15-(20) cm. × 2-4 mm. (not including recurved portion), linear, tapering regularly from base to acute, often apiculate apex. Upper surface with us. long-persistent silvery pellicle; lower densely clad, as is evident midrib, in appressed almost satiny white tomentum. Margins entire or nearly so, strongly recurved, almost or quite to midrib. Sheath thin, pale, glab. or slightly hairy; midrib and principal veins us. evident. Scape slender, 25-40 cm. long, us. clad in appressed almost satiny white tomentum. Bracts narrow-linear, us. rather distant, with appressed soft white tomentum, lowermost lamina c. 3 cm. long, remainder regularly diminishing towards capitulum. Capitula ± 12-20 mm. diam. Phyll. up to c. 13 mm. long, linear-subulate, pale, glab. or ciliate, somewhat indurated below, darkened, especially marginally, in upper floccose part. Ray-florets up to c. 25 mm. long, tube very slender, limb gradually widened to 4-toothed apex, ascending. Disk-florets narrow-funnelform, up to 1 cm. long, teeth triangular. Achenes compressed-cylindric, ribbed, glab., c. 5 mm. long. Pappus-hairs very slender, white, up to c. 10 mm. long, hardly barbellate.
DIST.: N., S. Lowland to lower montane grassland, herbfield, bogs, throughout, from c. lat. 37° southwards.
It appears desirable to give a brief historical account.
Hooker defines his C. gracilenta thus: "Aster gracilentus, Banks and Sol. MSS. : foliis scapisque ut in C. asteliaefolia [Hook. f. Fl. Antarct. 1, 1844, 35], achaeniis glaberrimis, styli ramis longe productis gradatim acuminatis, papillis filiformibus elongatis." The localities given are: Great Barrier Id, "Lofty mountains of Waikato Lake", Mount Egmont. He remarks: "A plant so very near, even in the variable form of its leaves, to C. longifolia, that it was not till I had examined the styles that I could detect any difference; the conical papillose portion of these being much produced, gradually acuminated, and three times the length of the lower part of the arms, with the papillae almost filiform."
In Fl. N.Z. 1, 1853, 122 Hooker describes as vars: " Var. α (; foliis anguste linearibus acuminatis marginibus ad costam revolutis superne araneosis v. pube appressa argenteo-sericeis nitidis. Aster gracilenta, Banks et Sol. Mss. et Ic. Celmisia, Fl. Antarct. p. 35, in nota. Var. ß; foliis latioribus anguste lineari-lanceolatis marginibus revolutis, scapis validis. Var. γ; foliis ut in var. α, involucri foliolis brevioribus lineari-oblongis obtusis. Var. δ; foliis lineari-elongatis flaccidis submembranaceis superne glaberrimis viridibus subtus niveis marginibus tenuiter recurvis. C. graminifolia Fl. Antarct. p. 35 in nota." Localities given are: Var. ß. Auckland, Sinclair. Mount Egmont, close to the perpetual snow, Dieffenbach. Var. γ. Canterbury, Lyall. Var. δ. Bay of Islands, Sinclair, etc. Canterbury, Lyall.
In the Handbook 1864, 134 Hooker places his C. gracilenta as a synonym of C. longifolia Cass. and remarks: "Varying excessively in stature, robustness, length and breadth of leaves, one or many (8-12) slender or robust scapes, and large or small, cottony or glabrate heads . . . Some states almost pass into C. Munroi."
Kirk (Stud. Fl. 1899, 289) described his var. alpina of C. longifolia for small plants "in alpine bogs". He remarks: "The following forms depend largely upon situation and may be easily recognised, although from their not being permanent they are not eligible for even varietal rank." Under this statement he includes: gracilenta, asteliaefolia, major, graminifolia.
Cheeseman (Man. N.Z. Fl. 1925) accepts as spp.: gracilenta, major, graminifolia, alpina. Under gracilenta he remarks of specimens of C. longifolia from Mount Kosciusko: "there can be no question as to the distinctness of the Australian plant, which differs more from the New Zealand forms than most of our species do from one another." I have cursorily examined numerous specimens from Tasmania, Victoria, and New South Wales. There appear to be a number of spp., some superficially at least closely resembling N.Z. forms of gracilenta, adamsii, major and graminifolia. I found no such clear separation as Cheeseman maintained.
Recent segregations from the complex have been C. insignis Martin and var. dura Simpson et Thomson, C. spedenii Simpson, C. neo-zelandica Gandoger.
Colenso (T.N.Z.I. 21, 1889, 88) described his C. setacea from plants collected by Hill "On open ground, high slopes of Mount Tongariro . . . 1887." The description includes: "A small slender species, slightly cottony; apparently growing singly. Leaves few (6-8), 2 in.-3 in. long, 1/2 line wide, sub-setaceous, greenish-grey, flaccid, drooping, margins revolute, tips acute . . . glabrous and reddish on the outside, very loosely cottony within, midrib on under surface glabrous . . . Scape single, very slender, 7 in. to 8 in. long; bracts setaceous, 3-4 lines long, very acute . . . Heads small, loose, spreading, 1/2 in.-3/4 in. diameter . . . Ray-florets 14, distant, sublanceolate, 51/2 lines long, 4-veined . . . Achene linear, 11/2 lines long, angled somewhat 4-sided, glabrous." I have not found specimens. Cheeseman (Man. N.Z. Fl. 1925, 953), without comment, places the name as a synonym of C. graminifolia, but it appears rather to belong here.
C. neo-zelandica Gand. in Bull. Soc. bot. Fr. 17, 1918, 39 is thus described: "Praecipue ad C. longifolia DC. accedit sed ab la valde discrepens indumento pauco, gracilitate partium, foliis lineari-filiformibus, involucre pauce tomentilli phyllis subacutis apice atro-purpureis, ligulis duplo minoribus. Hab. Nova Zelandia prope Dunedin (Thomson!). Species elegans caulis tenuis sesquipedalis, folia ut in caeteres, radicalio vix 2 mm. lata junciformia, involucrum quam in typo saltem duplo minus 1 cm. diam. latum basi attenuatum."
Similar very slender forms are not uncommon throughout the range of the sp., but until a type is decided upon the status of Gandoger's specimens (not seen by me) can hardly be determined.