174 
FLORA AND SYLVA 
I formed the combination Cortaderia \ 
argentea for the Pampas-grass. [ 
It is not my intention to describe \ 
fully the characters which divide Gy- 
nerium and Cortaderia, but I will just ' 
remark that Gynerium (of which only 
one species is known) is a very tall grass 
with the habit of the Sugar Cane and 
growing in companies at the waterside. 
It is a native of tropical America and 
a good specimen exists in the Victoria- 
Regia House at Kew. Nor will I for 
the present discuss whether Cortaderia 
should not take in certain species which 
are still (though in my opinion wrongly) 
referred either to Gyfteriiun ox Arimdo; 
in any case the five species which I 
referred to Coi^taderia five years ago 
(no new kinds have been added since) | 
would have to stand as a distinct natural 
group. They would in any case re- { 
present Cortaderia proper, and to them 
I confine myself They are all similar 
in general appearance, and although 
they might be easily distinguished if 
seen side by side in a garden, they are 
rather difficult to recognise when dried 
and trimmed to size for the herbarium, 
but on such material I have to fallback, 
as only three out of the five species are 
in cultivation. In these circumstances 
I shall need to introduce technicalities, 
and to understand them I would ask the 
reader to keep in mind what has been 
said as to the general structure of the | 
spikelets of the Pampas-grass, the re- j 
marks on this point holding good for i 
all the Cortaderia proper. The follow- i 
ing are the species at present known: — 
Cortaderia argentea. — Flowering culms 6 to 8 feet 
high without the plume (in California they are stated to 
attain as much as 20 feet), uppermost internode 2 to 2\ 
feet long. Leaf-blades up to 6 feet by 3 to 5 lines. Panicles 
I to 2 feet by 4 to 6 inches, often with the branches directed 
towards one side, looser, more nodding and silvery in the 
male, silky and feathery in the female. Spikelets 3 to 6 
flowered ; glumes very narrow with a long fine point 6 to 
8 lines long; valves very similar to the glumes, glabrous 
or very scantily hairy in the male (hence its more silvery 
lustre) and densely hairy in the female (whence its silky 
panicle), hairs 3 to 4 lines long. Staminodes very minute, 
Trg- to y\ lines long, cylindrical, without (or almost without) 
trace of an anther. 
This is the common Pampas-grass of gardens. 
It varies under cultivation in the size, colour, 
shape, and looseness of the plume, and also 
produces forms with variegated leaves. Some 
of these variations have been named by growers 
as though they were distinct species or fixed 
varieties, but as they are usually merely indi- 
vidual variations which do not come true from 
seed, it seems best to avoid this method of 
naming. Thus, Cortaderia IVesserlingiim'xght be 
known asC. argentea '■'■Wesserling" and C.Sten- 
ackeri as C. arge?itea StenacJier^'' and so on. 
The home of this species is southern extra- 
tropical Brazil, Paraguay, Uruguay, and the 
eastern parts of the Argentine, about as far as 
40° S. lat. It is mainly found in the neigh- 
bourhood of water-courses, in marshes and 
similar places, but is by no means one of the 
characteristic grasses of the vast plains gener- 
ally called " Pampas." The name Pampas- 
grass, which first appears in " Paxton's Flower 
Garden" of 1850, can therefore hardly be 
considered appropriate. I have already traced 
the history of the plant up to its introduction 
into this country, whence it soon spread to 
the continent. Vast plantations of Pampas- 
grass have sprung up in California, where it 
was grown for profit as long ago as 1874. 
The crop of 1889 from the Santa Barbara 
plantations was estimated at one million plumes. 
C. araucana.— This is only known to me from herbarium 
specimens. It is evidently a smaller plant than C. argentea 
with slender and shorter culms. Leaf-blades up to 4 feet, 
by 3 lines wide. Panicles 9 to 15 inches by 2 to 2\ inches, 
oblong, erect, and very dense, straw to light bronze-coloured 
in the dry state, very lustrous, almost alike in both sexes. 
Spikelets 4 to 7 flowered ; glumes somewhat broader than 
in C. argentea, prolonged as a very long and fine point 7 
to 8 J lines long ; valves very similar to the glumes, hairy 
in both sexes but more so in the female, hairs 2 lines long, 
middle nerves drawn out into a fine bristle 3 to 5 lines long. 
Staminodes with very minute short thick filaments, and 
minute rudimentary anthers. 
This is a native of Southern ChiH ( Valdivia, 
Chiloe) but nothing is known as to its manner 
of growth. It is a handsome plant and would 
