tion of the land round the Bay of
Concepcion by several feet. The convulsion was more effectual in
lessening the size of the island of Quiriquina off the coast than the
ordinary wear and tear of the sea and weather during the course of a
whole century; but on the other hand, on the Island of St. Maria putrid
mussel-shells, still adhering to the rocks, were found ten feet above
high-water mark. Near Juan Fernandez Island a volcano uprose from under
the water close to the shore, and at the same instant two volcanoes in
the far-off Cordilleras bust forth into action.

The space from which volcanic matter was actually erupted is 720 miles
in one line and 400 miles in another line at right-angles from the
first; hence, in all probability, a subterranean lake of lava is here
stretched out of nearly double the area of the Black Sea. The frequent
quakings of the earth on this line of coast are caused, I believe, by
the rending of the strata, necessarily consequent on the tension of the
land when upraised, and their injection by fluidified rock. This rending
and injection would, if repeated often enough, form a chain of hills.

I made the passage of the Cordilleras to Mendoza, the capital of the
republic of that name, on horseback. The features in the scenery of the
Andes which struck me most were that all the main valleys have on both
sides a fringe, sometimes expanding into a narrow plain of shingle and
sand. I am convinced that these shingle terraces were accumulated during
the gradual elevation of the Cordilleras by the torrents delivering at
successive levels their detritus on the beach-heads of long, narrow arms
of the sea, first high up the valleys, then lower down and lower down as
the land slowly rose.

If this be so, and I cannot doubt it, the grand and broken chain of the
Cordilleras, instead of having been suddenly thrown up--as was till
lately the universal, and still is the common, opinion of
geologists--has been slowly upheaved in mass in the same gradual manner
as the coasts of the Atlantic and Pacific have arisen within the recent
period. The other striking features of the Cordilleras were the bright
colours, chiefly red and purple, of the utterly bare and precipitous
hills of porphyry; the grand and continuous wall-like dikes; the plainly
divided strata, which, where nearly vertical, formed the picturesque and
wild central pinnacles, but where less inclined composed the great
massive mountains on the outskirt