( 
174 ANNALS OF THE ROYAL BOTANIC GARDEN, CALCUTTA. [M. rumphii 
with small fruits, whieh I have considered. as a subvariety ‘of var. micracanthum 
(M. Rumphit var. mieracanthum subvar. Makanaro) and which on account of its small 
fruits, evidently cannot be Rumph’s Sagus longispina. 
Under the name of M. longispinum were forwarded to me, also from Buitenzorg. 
some fruits a little smaller than those of the typical M. Rumphit (high and broad 
35—37 mm.) flattish above and at the base, having the nucleus covered with a 
copious integument, and the chalazal cavity wide and shallow. 
Pirate 107.—Fig. 9.—Fruit and seed entire, and seed cut vertically, from the 
specimens Tecelved from Buitenzorg under the name of Metroxylon longispinum, 
mentioned above. 
2c. MerRoxyLoN RuMPHH var. SYLVESTRE Becc. 
M. sylvestre Mart. Hist. Nat. Palm ili, 215, 343; Miq. Fl. Ind. Bat. iii, 146. 
Sagus sylvestris Rumph. Herb. Amb. i, 75; Blume, Rumphia, ii. 153. 
a Lo wi vAR. B sylvestris fructu perforato Giseke, Prael. in Ord. 
Descriprion.—Rumph describes his Sagus sylvestris as having a very high trunk 
covered with spines, shorter but more densely set than in any other; most pro- 
bably, however, it is not precisely the trunk that is spinescent, but the spines are 
on the bases of the leaves that cover the trunk. The fruits are said to be smaller 
than in “Lapia Tuni” (the type of M. Rumphit), globular, depressed at the base 
and above, and having ihe seed perforatel, or in other words, with the chalazal 
Cavity so deep as to traverse the/.albumen from side to side. I have, however, 
often. observed this in M. Sagus, where frequently the embryo penetrates the entire 
mass of the albumen, and with its apex attains the base of the chalazal fovea, in 
which case the albumen looks as if it were perforated. 
HaBrrAT.—According to Rumph this variety grows frequently in Ceram, but it 
is rare in Amboina. Native names “ Lapia Ihur” and “Ihul.” 
Osservations.—I have received from the Buitenzorg Herbarium a specimen 
named “Sagu Ihor” consisting of the terminal part of a leaf, and 3 fruits, 
gathered in the West of Ceram. The rhachis of the leaf is rather strongly spines- 
cent beneath; the leaflets are long-acuminate, prolonged into a narrow tip and 
rather densely ciliate at the sides; the mid-costa is furnished with some long 
bristles on the upper surface towards the apex, and the margins are closely 
ciliate spinulous. The fruits are. globular, inconspicuously beaked, very much like 
those of M. Sagus, 35—38 mm. across and 30—35 mm. high, and have the base 
excavate ; the largest scales are 11—12 mm. broad in the exposed part, glossy, 
straw-coloured, with a darker faint intra-marginal line, and very - narrow, 
discoloured, erosely-toothed margins. The pericarp has the uniform thickness of 
2—3 mm. all round, and is not considerably more spongy at the base than 
~- elsewhere. One of the 3 fruits I dissected contained an abortive seed with 
hypertrophic integument as in M. Sagus. This specimen we may assume to represent 
really the Sagus sylvestris of Rumph, but in fact, as far as I can judge from the 
specimens at my disposal, it differs from M. Rumphi only in its smaller fruit, and 
in the leaflets being furnished with some long bristles. 
