303 
of the globe. Such formations are, on the other hand, 
utterly irreconcileable with the Wernerian theory. The 
retreat of the universal water must have taken place simul- 
taneously over the whole surface of the earth. No one 
portion of land can therefore be accounted more ancient 
or more recent than another; except in so far as the tops 
of the mountains, being the first left dry, may lay claim to 
a priority of existence. 
* The Isle of France is surrounded almost throughout 
with a coral reef, the continuity of which is nowhere broken 
but at the harbours and the mouths of the rivers. This reef 
varies greatly in diameter, forming in some parts merely a nar- 
row border, while in others it expands to the breadth of a mile 
and upwards. Its surface is in general quite level, but here 
and there intersected by narrow channels, wherein an infinite 
variety of small fish, the tenants of the reef, take refuge 
during the reflux of the tide. The extreme edge is invariably 
the shallowest part, and seems as if designedly raised as a 
bulwark against the encroachment of the sea, which breaks 
over it with a prodigious roll Beyond this barrier, the 
bottom slopes like a glacis, and soon sinks out of view. 
ing over the reef from an eleyated station, we can trace 
the windings of the channels that traverse it by the varying 
colour of the water. In the deepest parts it is of a bluish 
tint; in two or three fathoms this colour passes to green, 
which becomes paler in proportion as the depth diminishes. 
From the top of the Pouce Mountain we have a bird's-eye 
View of almost the whole island, with its pale-green border, 
veined with blue, and fringed with the foam of the everlasting 
Surge, Beyond this bright zone, the indigo colour of the 
ocean deepens on the view until lost in the aerial tints of the 
* I have already hinted my suspicion that the strong fabrics 
known by the name of Coral-recfs, are not reared up from 
the depths of the ocean, as is generally believed ; but that they 
are little more than mere incrustations growing on a base of 
Solid rock, at the depth of a few fathoms beneath the surface. 
But as this is a point that lies beyond the reach of actual 
