down into the sea, are broken into fragments and washed away.
Many instances of this nndermining may be seen around the coast;
at Bankses there is a very good example, .where a thick layer
of lava is deeply undercut. I t is, however, not so quick a process
as may at first sight appear, and such falls around the coast of St.
Helena rarely occur, though, when they do, a good slice of rock comes
away. The northern coast has in this manner been cut into by
the sea making its way inland for a distance of 1233 yards, or
nearly three-quarters of a mile, leaving a perpendicular face to the
cliff of 600 or 1000 feet. This is what we find on the leeward side,
where the sea is generally calm, excepting occasionally when disturbed
by the rollers ; but the denuding action of the Atlantic waves,
which sweep with great force before the south-east trades against the
windward side, has been much greater. There, at Holdfast Tom
near Prosperous Bay, Stone Top, and .Old Joan Point, the
Island has been worn away until perpendicular cliffs of full 2000
feet have been formed.
An effect of this kind is visible also at Ascension, which, being
a more recent formation, has been less acted on by the rollers along
its northern coast, but has been considerably denuded on that side
which is exposed to the prevailing winds. Again, at Madeira it is
seen that those parts of the coast which are exposed to the broad
sweep of the Atlantic are the most worn away.
By prolonging the direction of the strata, and making due
allowance for any displacement, the original coast line can be ascertained
with tolerable correctness, and this we find at Holdfast
Tom was once 3300 yards, or nearly two miles, further seaward than
it is at present. I f then the wearing away, in a horizontal direction,
of the coast at this exposed point, was at the rate of three
inches in a year, it would require a period of 39,600 years to elapse
since the action of the sea commenced to produce this change ; or,
taking the less exposed side of the Island, as at Ladder Hill, an
encroachment by the sea of one inch horizontally per year would
suffice to give 44,388 years. This estimate of the encroachment by
the sea on those hard iron-bound basaltic cliffs, is, to one who has
for many years carefully watched its progress, rather a minimum
than a maximum rate. Tear after year passes without any perceptible
change in the line of cliff; the undercut strata at Bankses,
already mentioned, has through a number of years remained appa