138 PERCY SLADEN TRUST EXPEDITION. 
latter represent the perpendicular-sided fissures commonly found in other reefs. They 
are more free of weed than other parts of the reef, and are regular in proportion to the 
breadth of the reef, representing no doubt the channels for the rise and fall of the tide. 
The exposed masses are mainly covered and formed by Lithothamnia of fine branching 
facies intergrown with Halimeda, both classes of organisms being richly covered in their 
lower parts with thin brilliant-green Tunicates. Corals are scarce, small colonies of 
fine-branching Madrepora, Pocillopora, and Millepora alone being common. 
Passing inwards from the edge of the reef, such as it is, a reef-flat and a boulder-zone 
should be crossed before reaching the shore-flat. Actually one passes on to a flat, on 
which the Cymodocea becomes more and more dominant as the shore is approached. 
Here, extending from about 35 to 60 yards back from the breakers, are found many 
loose masses of dead coral, mainly Jadrepora and Porites, more or less imbedded in sand, 
with colonies of Heliopora formed of thick densely-packed laminz. Some of these 
may be exposed at low tide; and this area is all that represents the boulder-zone. 
Behind this the shore-flat is as densely covered with Cymodocea as a sown meadow, 
the soil beneath consisting largely of foraminiferal sand mixed with rubble. While 
generally it has less than a foot of water at low tide, a series of pools are commonly 
found at the base of the beach. They average 3 to 6 feet deep, with sandy bottoms, and 
in most places are only 20 to 30 yards broad; but at the north end of the island, where 
there is no reef, they are much more considerable, strong tides setting along them 
and sweeping out the sand and detritus. Elsewhere they communicate with the ocean 
through the channels. Were some of these latter to enlarge and become definite 
gaps, the pools would soon form a boat-channel, and wJtimately there would be a barrier- 
reef along the east side of the island. The only other point of interest about the pools 
lies in the fact that their sides are perpendicular, being held up by living Porites, 
Madrepora, Stylophora, and Heliopora—the corals which, in the dead state, form the 
boulder-zone. 
The whole shoal is doubtless growing seawards by the extension of its talus slope ; 
but the outward growth of the surface-reef appears to be considerably impeded by the 
dense Cymodocea. The latter, on the shore-flat, prevents loss and diminishes the force 
of the waves so materially that their effect on the coast behind is comparatively small. 
Indeed it is now only noticeable for a mile at the south end and for a mile northwards 
from opposite the settlement, where there are extensive formations of beach-sandstone. 
At the north end of the latter area, where there is a turtle-pond, the sandstone extends 
diagonally halfway out upon the reef, which in this position is a third of a mile across. 
This points to the washing away of the island having been responsible for much of the 
shore-flat. Further, this view is supported by the fact that the beach everywhere 
is formed of sand without coral-masses, whereas similarly-situated coral-islands are 
to seaward invariably composed largely of coral-rock or masses of corals. Otherwise it 
is necessary to suppose that Coetivy is an island of different formation to the Maldives 
and Chagos, a quite possible view. To the north the beach was interesting in having 
cliffs formed of dead Cymodocea piled up to a height of ten feet and more above 
the flat. 
