Lepidosperma. | CYPERACE. 791 
NortH Isntanp: 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 ; 
margins involute, scabrid. Panicle large, terminal; sometimes 
broad and efftuse, with drooping branches; sometimes narrower and 
more erect. Spikelets clustered, black or dark-brown, 1—2-flowered ; 
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 want- 
ing. 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 ofthe flowering glumes. They generally remain attached 
to the base of the nut after it has fallen away, and as the other end of the fila- 
ment 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 ex- 
tent. 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, 3-7ft. Panicle 1-3ft., nodding. Glumes 7-8; 4-5 
empty, unequal. Nut small, }in., red-brown 2: 
Smaller, 2-4ft: Panicle 14-2} ft., rigid, erect. Glumes 
6-7 ; 3-4 empty, subequal. Nut small, tin, red-brown 
Slender, 2-4ft. Panicle 13-24 ft., narrow, eiongate, 
branches distant. Glumes8; 5 empty, unequal. Fruit 
large, 4-4in., red-brown... oe 5 e 
Very tall and stout, 5-12ft. Panicle 2-5ft., nodding. 
Glumes 6-7; 4-5 empty, subequal. Nut large, 4-+in., 
black when fully ripe or ae nie Er 
1. G. setifolia. 
bo 
. G. rigida. 
3. G. pauciflora. 
4. G. xanthocarpa. 
