DESCRIPTION OF THE EXPEDITION. 39 
the island, separating off a part now termed Gabrielle. The actual land of Mapous de 
I'He du Coin and of Fouquet lies 150 yards behind the boulder-zone, as do Coin and Poule 
in certain positions. The north end of the latter island is being washed away by the 
tremendous current which rushes round it; its reef, too, in this position is narrow and 
certainly is not growing out to close up the passage to the north—in fact, the latter, if 
anything, appears to be widening. The islands between Coin and Poule are hollowed 
out against each other—in fact, are double-ended. From each end either a rocky point 
passes on to or behind the boulder-zone, or a sandy point 200 to 300 yards in length 
runs on to the reef parallel to the lagoon-edge. The outer points, too, are joined along 
the reef by conspicuous masses of rock, so that there is no doubt but that at one time 
Coin and Poule, at least, formed part of one elongated island (PI. 6). 
Passing to the lagoonward points, they are perhaps as extensive now as they have ever 
been. Probably the islands, as a whole, are growing towards the lagoon by piling up of 
sand * ; and if this continues some may be reunited by banks of sand on the inner parts of 
the reef. The islands themselves are generally formed of coral-rock towards the outside, 
solid in some places, loose in others, tuffe or consolidated sand-rock in the centre, and 
loose sand towards the lagoon. Where masses of rock exist in and behind the boulder- 
zone of the outer reef, they indicate that the island is being washed away, additions from 
the reef being quite negligible. The boulder-zone, too, is largely formed by the aggre- 
gation of loose fragments of coral-rock from the islands. This gradual washing away 
raises the question of the formation of the islands, but from the area investigated it is 
not one to which we can give any decided answer. The greatest height of rock above 
high-tide level we found to be 8 feet, and it was quite uncertain whether its corals were 
really in situ or not. On the whole, we incline to think that the islands were originally 
formed by the elevation of a continuous reef extending from Fouquet to Poule and 
conceivably bridging the northern passage and joining up to the reef, of which Diamant 
forms the northern island. 
The marine animals were strikingly similar to those at Salomon, but the stinging-ray 
(Trygon sp.?) is more common. We saw porpoises daily in the lagoon and whales were 
stated to enter it not infrequently. The marine plants were similar to those which live 
round most oceanic coral-islands, such as the encrusting Lithothamnia, bunches of the 
caleareous-leaved Halimeda, and a small variety of the more insignificant forms towards 
the outer edge of the reef. Between Monpatre and Gabrielle was a considerable growth 
of a grass-like cotyledonous plant (Cymodocea ciliata), a form which we never met 
with elsewhere in the Chagos and which we never saw again on the reefs until our 
arrival at Cargados. Even in Peros it has not spread from this single locality, a relatively 
still backwater between these two islands, and more probably its absence elsewhere iv 
the Chagos is due to its inability to find suitable surroundings, perhaps proper rooting- 
ground, rather than to the improbability of its seeds being floated over from the 
Seychelles region where it abounds. 
* There is such considerable change and movement of the sand in the two monsoons that we cannot make the 
statement on our own evidence alone. However, all the older inhabitants are agreed on the subject, and certainly 
the vegetation seemed to bear them out, 
