North Island: Auckland— Clay hills between Mongonui and Kaitaia, H. Carse! August–September.
I am indebted to Mr. C. B. Clarke for identifying this with the Australian L. filiforme. So far, it has only been gathered in New Zealand by Mr. Carse, but it will probably prove to be not uncommon north of Auckland. In Australia it has been recorded from Victoria and Tasmania.
11. GAHNIA, Forst.
Tufted perennial herbs, usually of large size. Stems tall and stout, leafy throughout their length. Leaves usually long, very coarse and harsh, narrowed into long subulate or filiform points; margms involute, scabrid. Panicle large, terminal; sometimes broad and effuse, with drooping branches; sometimes narrower and more erect. Spikelets clustered, black or dark-brown, 1–2-flovered; the upper flower hermaphrodite and fruit-bearing; the lower flower sterile or male. Glumes many, imbricated all round; the outer 2–5 or more empty, keeled, often mucronate; flowering glumes minute at first, but enlarging in fruit. Hypogynous bristles wanting. Stamens usually 4 in the hermaphrodite flower, often 4–6 in the male flower; filaments greatly elongated after flowering, and often holding the nut. Nut hard and bony, ellipsoid or ovoid or obovoid, obscurely trigonous or terete, red or reddish-brown or black.
Species about 30, most of them natives of Australia and New Zealand, but extending through the Pacific islands to the Sandwich Islands and Malay Archipelago. Of the 8 found in New Zealand, one occurs in Lord Howe Island and another in the Sandwich Islands, the remaining 6 are endemic. The genus is remarkable for the extraordinary extent to which the filaments lengthen after flowering. In G. procera they are often quite 2 in. long, or from 8 to 10 times the length of the flowering glumes. They generally remain attached to the base of the nut after it has fallen away, and as the other end of the filament is usually entangled with the glumes or with the filaments of other flowers the nut remains swinging by the filaments quite free from the spikelet. Mr. Colenso (Trans. N.Z. Inst. xviii. 281) has suggested that some of these filaments are in reality hypogynous scales, giving as a reason for this belief that in his G. scaberula and G. exigua he has noticed within the same flower stamens with the filaments still very short, and filaments already lengthened to the full extent. He failed to notice that the lower male flower expands long before the hermaphrodite flower placed just above it, so that its filaments have lost their anthers and lengthened long before the expansion of the upper flower takes place. The two flowers are placed so close together that it is quite easy to take the two for one.
| Tall, 3–7 ft. Panicle 1–3 ft., nodding. Glumes 7–8; 4–5 empty, unequal. Nut small, ⅙ in., red-brown | 1. G. setifolia. |
| Smaller, 2–4 ft. Panicle 1½–2½ ft., rigid, erect. Glumes 6–7; 3–4 empty, subequal. Nut small, ⅙ in, red-brown | 2. G. rigida. |
| Slender, 2–4 ft. Panicle 1½–2½ ft., narrow, elongate, branches distant. Glumes 8; 5 empty, unequal. Fruit large, ⅕–¼ in., red-brown | 3. G. pauciflora. |
| Very tall and stout, 5–12 ft. Panicle 2–5 ft., nodding. Glumes 6–7; 4–5 empty, subequal. Nut large, ⅕–¼ in., black when fully ripe | 4. G. xanthocarpa. |
