176 ANNALS OF THE ROYAL BOTANIC GARDEN, CALCUTTA. [.M. rumphii 
excavate at the base, otherwise exactly the same as in variety album, a little 
smaller. 
HaBrrAT.—AÀmahai in Ceram. Native name “Sagu hitam.” 
Pirate 107.—Figs. 2—6.—Metroxylon Rumphii var. ceramense  Bece.—Fig. 2 
subvar. album (from  Amahai)— Fig. 3 subvar. album (from West Ceram)—Fig. 4 
subvar. nigrum ;—Fig. 5 subvar. platyphyllum. Fig. 6 subvar. rubrum. (All specimens 
from Buitenzorg in Herb. Beccari.) | 
2e. MerRroxyLon RuMPHII var. MICRACANTHUM Bece. 
M. micracanthum Mart. Hist. Nat. Palm. ili, 215 (2nd edit.); Miq. Fl. Ind.; 
Bat. in, 146. 
Sagus micracantha Bl. Rumphia, 4, 153. 
Sagus duri rottang; Ramph. Herb. Amb. 7, 75 sub No. 1. 
Descrirtion.—The name of Sagus micracantha was assigned by Blume to the 
Sagus duri rottang of the ^ Herbarium Amboinense” described by. Rumph as 
. having very small fruits and leaves armed with shorter spines than those of M. 
Rumphi, and similar to those of a Calamus; I think, however, that a rigorous 
identification of this variety, with the description given by Rumph, to be a rather 
difficult task. Nevertheless it seems to me very probable that a variety named 
“ Sagu Tuni" growing in West. Ceram, of which I have received specimens from 
Buitenzorg, may correspond to Sagus miracantha Bl. Blume. Martius and Miquel 
have considered (doubtfully however) Rumph’s Sagus duri rottang as corresponding 
to Labillarditre’s Sagus genuina, figured in Turpin’s “ Dictionnaire des Sciences 
Naturelles (Botanique), but although Labillardiére’s species is certainly a very near 
ally of Sagus duri rotiang I prefer to consider it as a distinct variety. 
Of the Metroxylon which I believe to correspond to the Sagus duri rottang 
I have seen only the end of a leaf, and a few fruits. The leaf has the dorsum 
of the rhachis spinescent to the extreme end, and the leaflets are distinctly ciliate- 
spinulous on the margins and on the mid-costa. The fruits are very small, some 
of them (having been evidently pressed one against the other) are oblong-obovoid, 
but one regularly evolute is globular, not depressed above, terminated by a slender 
pungent beak, is 23 mm. in diameter, has the base roundish, the scales very 
shiny, straw-coloured, and very narrowly discoloured on the margins. The pericarp 
is thin above, but gradually becomes thicker and more spongy at the sides, and 
is 6—7 mm. thick at the base. 
' Haprrat—Rumph writes that the Sagu duri rottang, to which he adds the 
Amboinese name: ‘Lapia Luli-uwe," grows abundantly at Humohela (in Ceram), 
but that it ise almost unknown in Amboina. The specimen I have considered as 
corresponding to it. comes from about the place mentioned by Rumph (West 
Ceram) and was labelled by the Buitenzorg collector “Sagu Tuni,’ a name, 
however, with which Rumph distinguishes the “Sagu” that has been generally 
considered as the typical M. Rumphii. 
Prate 107.—Fig. 12.—Metroxylon  Rumphii var. micracanthum — Becc.—F ruits 
from West Ceram, named *“ Sagu Tuni". by the Buitenzorg collector, and considered 
as corresponding to Rumph's Sagu duri rottang (Herb. Beccari). 
