54 
Flora of Narcondam and Barren Island. 
a thicket of Mussuenda macrophylla —the accident of its situation has 
converted the species into a straggling shrub and imparted to it a very 
distinct facies. On the lava itself nothing grows, though further inland 
and to the south of the stream it is in several places partially covered by 
beds of Aganosma marginata, which, rooted in the adjacent soil, and 
having no trees on which to climb, prefers sprawling over the bare black 
lava to spreading along the ground among the grass. This grass, 
Ischcemum muticum, almost completely occupies the plain between the 
lava flow and the inner wall of the outer cone, which is thus a great 
meadow in which, however, there are some patches of scrub jungle, the 
chief constituents being Dodoncea viscosa, Flueggia microcarpa, Gelonium 
hifarium, Phyllanthus reticulatus, Trema amboinensis, Dalbergia tama- 
rindifoUa, and stunted examples of Gallicarpa arborea. 
The inner cone is merely a “ cinder-heap,” with hardly any vegeta¬ 
tion ; a few very stunted examples of Trema amboinensis on its southern 
face, about 650 feet up, and small shrivelled tussocks of Fimbristylis 
ferruginea scattered unevenly over all the sides except the western, being 
the only plants present. The interior of the crater has more vegeta¬ 
tion than the whole outside of the cone ; near the crevices in the inner 
wall, and especially on the south side where the soil is moistened by the 
condensation of escaping steam, occur NepJirolepis tuberosa (also obtained 
elsewhere in the island), Gheilanthes tenuifolia (very small and stunted 
specimens). Lycopodium cernuum (all over the stones in the western, 
more shallow depression of the crater), Psilotum triquetrum (also found 
in Java, on the crater of Gunong Boddas Preanger, by H. O. Forbes), 
Pholidota imbricata, Vandellia Crustacea and Oldenla7idia corymbosa ; in 
the sand at the bottom of the deeper eastern craterine depression occur 
luxuriant patches of Fimbristylis ferruginea. 
An attempt was made to land at Anchorage Bay; owing, however, 
to the heavy surf that rolls in this was found to be impossible. The 
beach in this bay is sandy ; behind it could be seen the usual sea-fence 
of Pandanus, a species seen nowhere else on the island. Just within the 
Pandanus fence rise 13 coco-nut trees tall enough to be seen and counted. 
Judging from the analagous beaches in the Coco Group and hTarcon- 
dam it may be anticipated that there are many seedlings besides. To 
verify this surmise an attempt was made later on to cross the outer cone 
from the amphitheatre and work down to this beach. The attempt did 
not succeed ; the sea was reached at a point too far to the east and the 
attempt was not considered worth repeating.* Rowing round the island 
* Those who have been engaged in similar work will understand how difficult 
it is under such circumstances to strike the jiroper ridge or ravine. The results 
of the journey, which it took a day to accomplish, were not sufficiently remunerative 
2G8 
