gle spring, we may feel
assured that it is insignificant in volume when compared to that which
has been carried to the sea since the time when it began to flow. What
may have been the length of that period of time we have no data for
conjecturing. In quarrying the travertin, Roman tiles have been
sometimes found at the depth of five or six feet.

_Baths of San Filippo._--On another hill, not many miles from that last
mentioned, and also connected with Mount Amiata, the summit of which is
about three miles distant, are the celebrated baths of San Filippo. The
subjacent rocks consist of alternations of black slate, limestone, and
serpentine. There are three warm springs containing carbonate and
sulphate of lime, and sulphate of magnesia. The water which supplies the
baths falls into a pond, where it has been known to deposit a solid mass
_thirty feet thick_ in about _twenty years_.[315] A manufactory of
medallions in basso-relievo is carried on at these baths. The water is
conducted by canals into several pits, in which it deposits travertin
and crystals of sulphate of lime. After being thus freed from its
grosser parts, it is conveyed by a tube to the summit of a small
chamber, and made to fall through a space of ten or twelve feet. The
current is broken in its descent by numerous crossed sticks, by which
the spray is dispersed around upon certain moulds, which are rubbed
lightly over with a solution of soap, and a deposition of solid matter
like marble is the result, yielding a beautiful cast of the figures
formed in the mould. The geologist may derive from these experiments
considerable light, in regard to the high slope of the strata at which
some semi-crystalline precipitations can be formed; for some of the
moulds are disposed almost perpendicularly, yet the deposition is nearly
equal in all parts.

A hard stratum of stone, about a foot in thickness, is obtained from the
waters of San Filippo in four months; and, as the springs are powerful,
and almost uniform in the quantity given out, we are at no loss to
comprehend the magnitude of the mass which descends the hill, which is a
mile and a quarter in length and the third of a mile in breadth, in some
places attaining a thickness of 250 feet at least. To what length it
might have reached it is impossible to conjecture, as it is cut off,
like the travertin of San Vignone, by a small stream, where it
terminates abruptly. The remainder of the matter held in solution is