204 REPORTS ON THE STATE OF SCIENCE. 
On September 12 I sailed for Aldabra, but again visited the island 
on my return, spending two days in photography and checking previous 
observations. 
ALDABRA. 
I arrived in Aldabra on September 18, and, owing to the excessive 
drought, had perforce to confine myself mainly to geological investiga- 
tion until the end of the period of south-east trade winds. The wet 
season commenced in November, after which month most of the plants 
were found in flower, while land animals, previously invisible, became 
fairly numerous. I had four camps—t.e., on Michel Island, at Taka- 
maka on Main Island, on Esprit Island, and on Picard Island—from 
which I examined every portion of the atoll. Owing to the dense and 
impenetrable scrub, exploration was always attended by considerable 
difficulty, as paths could only be cut at the cost of great labour; in 
addition I cleared several broad sections from the sea to the lagoon, in 
order to get a clear idea of the sequence of the rocks and vegetation and 
of the relative elevations. 
The nature of the ground and of its vegetation is such that the land 
may be divided into four somewhat irregular zones, from the lagoon 
outwards, as follows :— 
(1) Mangrove swamp—varying in size up to nearly a mile in 
maximum breadth. 
(2) Champignon—the surface much metamorphosed, coral rock, 
usually with sharply defined dark portions, which appear to consist of 
guano included during metamorphosis. It has evidently been subjected 
to heavy rain denudation, its surface being a mass of points and pits. 
The vegetation is a scrub of Pemphis acidula. 
(3) Platin—fairly smooth, composed mainly of coral fragments and 
reef débris, with a few shells, weathering into large flat slabs with soil 
accumulating in the crevices. In places are larger depressions, in which 
there are usually clumps of trees. The soil is guano, with a mixture 
of disintegrated rock. The vegetation is varied, containing numerous 
small bushes and trees, Pandanus, Ficus, Euphorbia, &c.; the fauna 
also is varied and comparatively rich. 
(4) Shore zone—largely of blown sand, with a stunted and wind- 
swept vegetation ; large clumps of Pandanus, Tournefortia, and Scaevola 
everywhere very numerous. 
In a broad sectional clearing which I made at Takamaka, the sea- 
ward reef commences with a fissured edge, succeeded by a sand flat, 
the sand being bound together by beds of grass-like Cymodocea, its 
rhizomes greatly overgrown by Lithothamnia; the buttresses between 
the fissures are themselves largely covered with sand; live coral is 
almost absent; not far from the edge are a few small boulders of dead 
coral, all much encrusted with Lithothamnia; a few species of seaweed 
are found in the pools left at low tide. The landward edge of the reef 
is formed of cliffs, 12 feet to 15 feet high, just outside which is usually 
a small depression in the reef with 2 feet or 3 feet of water. The cliffs 
are sloping, not overhanging, and are divided into buttresses ; they con- 
sist of a mass of corals cemented together with lime. The corals are all 
in the position in which they grew, and so perfect as to give the 
