K. squarrosa. KORTHALSIA. 153 
longitudinally along the centre, finely striate; the floral braeteoles are narrow, elon- 
gate, very slightly falcate, covered densely at apex and on the keeled backs with 
paleaceous, not woolly, hairs. Flowers relatively large, 13 mm. long, protruding 
during the anthesis by the full length of the corolla above their respective spathels ; 
the calyx is oblong, parted ‘to the middle into 3 broadly ovate, obtuse, thinly 
cartilaginous lobes; the corolla is. two and a half times as long as the calyx; 
when in bud fully developed.it is clavate and about 5 mm. broad, rounded on the 
top and parted down to about the middle into 3 oblong or spoon-shaped, thickish 
cartilaginous, very obtuse, strongly striate segments; it is fleshy in its lower 
undivided part; the stamens are inserted at the throat of the corolla, where 
iis tubular part begins, and where. also the filaments are united together to 
form a ring or collar, crowned by 6 thick and short teeth; anthers erect, 5 
mm. long, linear or linear-sagittate, apiculate, auricled at the base; the ovary | 
is ovoid and narrows above gradually into a rigid style, angular and grooved 
longitudinally, ending in 3 briefly subulate, connivent stigmas, which usually 
protrude beyond the ring formed by the filaments of the stamens at the throat 
of the corolla, and apparently reach to about midway of the anthers (the amalysis 
of the flowers in the specimens at my disposal was, however, very difficult, as all 
had been infested by insects, which had destroyed all the fleshy parts of th 
flower, leaving only the ovary untouched). Fruit unknown. 
Hasrrar.—The Philippines. In Palawan near Jwahig (Curran No. 7185-—May 
1906 in the Manila and Beccari Herbaria); and in Balabac Island (Merrill No. 
5384 in the Manila Herbarium). 
OssERvATIONS.— lhe type specimens—Curran No. 4185 consist in a few 
detached spikes and fragments of stem and leaves. In these specimens the leaf- 
lets are cuneate-rhomboidal, considerably longer than broad, and tae stems have had 
their ocrea destroyed by age; I consider, however, as conspecific with it the 
sterile specimen collected by E. D. Merrill (No. 5384) on Balabae Island. provided 
with the well conformed ocree described above, but bearing broader and more 
regularly rhomboidal leaves than Curran’s No. 4185. y 
A note of Merrill’s annexed to his specimen No. 5384 informs us that ants 
live in the ocreæ, which, however, are not inflated, and that a quantity of rubbish: 
is accumulated by the ants among the spines that cover the surfaces of the 
ocreæ— 
K. squarrosa is related to K. robusta Bl. of Sumatra, and to K. macrocarpa Bece 
of Borneo, and forms with. these two a small group, characterized by the spikes 
looking quite different from those of the great majority of Korthalsias. owing to 
their large loose imbricating spathels, and also to the floral bacteoles not being 
woolly. 
K. squarrosa differs from K. robusta in the shorter truncate ocreæ, in the smalleg 
spikes, and in the flowers having the style produced above the throat of the corolla ; 
with K. macrocarpa it agrees in the flowers, but differs, as from K. robusta, in the 
shorter truncate ocreæ:. i * 
ANN. Roy. Bor. GARD., CALCUTTA, VoL. XII. 
~ 
=» 
