HN 
| 
iit 
Wi 
a1) 
} 
of Mr. Roscoe on the Scitaminece of Dr. Roxburgh, we are 
induced to believe that he has intended the present plant 
by Canna patens: if so, he has had in view quite another 
species from the Canna indica (3) patens of the first edition 
of the Hortus Kewensis, which was generally presumed 
to have been the type of his species. In that the fourth 
inner segment or nectary is described as revolute, of a yel- 
low colour, besprinkled with small red lines; and is besides — 
different, both in form and position, from that of the present, 
as may be seen by the original specimen in the Banksian _ 
Herbarium. Giganteais of much later introduction into our 
collections. “< 
Stems several, 3-4 feet high or more. Radical sheaths 
and sheathing petioles, skinned over by a white araneous 
deciduous film. Larger leaves sometimes two feet or more 
in length, elliptically or ovately lanceolate, cuspidate. Ra- 
ceme loosely spiked, flowers generally in pairs: peduncles 
longer than the germen, with a brown dry bracte to each. 
Calyx persistent, half an inch or more in depth, faintly co- 
loured, powdered, 3-parted, segments further separated at 
one side, one rather larger than the rest. Corolla tubular, 
cleft, double above the tube, 3 inches long or more, scarlet 
without spots: common tube short, narrowly turbinate or 
obconic, filled with a honied lymph: outer limb an inch 
and half long, glossy, tripartite, erectly and regularly con- 
vergent, segments linearly lanceolate, streaked, inyolutely 
concave, front fissure deepest: interior limb more deeply 
coloured and opaque, 4-parted, 2 inches and a half deep, 
segments cohering imbricately for the length of the outer 
limb, except in front, where the fourth is disunited on one 
side from the next, over which it laps down to the common 
tube, all converging into an upper semicircular reflect- _ 
ent lip; 3 similar, nearly equal, ligulate, spatulate, lanceo- 
late, somewhat pointed ; the fourth and lowermost slanting 
away towards the next on its right, of a somewhat different _ 
form, deepest coloured, irregularly retuse, not placed oppo- 
site to the filament, nor adnate to the base of that, nor 
parted deeper on the right side from the rest, than they 
from each other, as is most usual in the genus. Filament 
revolute at the top: unther placed far below the apex of the 
filament. 
Flowers in February. A stove plant. When, whence, or . 
by whom introduced, we have not yet learned. The draw- 
Fe 
