Flora of New Zealand Series

Type locality: Mountains near Lake Guyon at c. 900 m., H. H. Travers, 1871. Buchanan's description is based on a small plant.

Main stems creeping and rooting, up to 5 dm. or more long; branches ascending, often crowded, c. 1 dm. long, stout, woody, glab. Lvs up to 5 cm. long; stipules up to 1 cm. long, 3-5-fid; lflts 7-11, inciso-serrate, obovate-oblong, upper pair up to 1·5 cm. long, pale brownish green, shining above, subglaucous below; terminal lflt with petiolule c. 2 mm. long. Scape glab., up to 1 dm. long, naked or with one or two pairs of bracts. Heads up to c. 2 cm. diam., green to purplish; bracts linear, ciliate. Cupules c. 5 × 4 mm., compressed, 4-winged, lateral pair conspicuous. Spines 4, seldom > 2 mm. long, without barbs. Sepals 4, broad-ovate, glab., purple-margined. Stamens 2, anthers white, filaments long. Stigma simple, fimbriate. Achene 1, obpyriform.

DIST.: S. Montane to subalpine screes and riverbeds, occ. to frequent throughout the Southern Alps.

Var. heteranthera Bitter in Bibl. bot., Stuttgart 74, 1911, 280. Kirk (Stud. Fl. 1899, 134-135) for A. glabra : "Stamens 25-40, all perfect, in the male flowers, but only 1 or 2, and usually imperfect, in the female flowers, with fimbriate stigmas. Some imperfect female flowers have 2 penicillate stigmas." Cheeseman (Man. N.Z. Fl. 1925, 511-512) says: "Flowers most commonly bisexual, but occasionally unisexual or polygamous. In the bisexual flowers the stamens are 2 only; while in the male flowers the stamens vary from 20 to 40 . . . The peculiarity was first detected by Mr. T. Kirk in some specimens collected in the Clarence Valley, and I still possess a specimen given to me by him in which it is clearly shown. But I have failed in obtaining any additional examples.

Bitter suggests the varietal epithet diandra for the plants with perfect, 2-stamened fls, and the above for the forms with ♂ 20-40-stamened. I have seen and grown plants from several localities; all had perfect fls with 2 stamens. Bitter suggests the possibility that A. glabra crosses with a sp. of Poterium. Sanguisorba minor does sts come within the range of A. glabra.