d upward to the
surface. This view is supported by many careful observations on the
effect which certain great earthquakes have exercised on the buildings
which they have ravaged. The distinguished observer, Mr. Charles
Mallet, who visited the seat of the earthquake which, in 1854,
occurred in the province of Calabria in Italy, with great labour and
skill determined the direction in which the shock moved through some
hundreds of edifices on which it left the marks of its passage.
Platting these lines of motion, he found that they were all referred
to a vertical plane lying at the depth of some miles beneath the
surface, and extending for a great distance in a north and south
direction. This method of inquiry has been applied to other fields,
with the result that in the case of all the instances which have been
subjected to this inquiry the seat of the shock has been traced to
such a plane, which can best be accounted for by the supposition of a
fault.

The method pursued by Mr. Mallet in his studies of the origin of
earthquakes, and by those who have continued his inquiry, may be
briefly indicated as follows: Examining disrupted buildings, it is
easy to determine those which have been wrecked by a shock that
emerged from the earth in a vertical direction. In these cases, though
tall walls may remain standing, the roofs and floors are thrown into
the cellars. With a dozen such instances the plane of what is called
the seismic vertical is established (_seismos_ is the Greek for
earthquake). Then on either side of this plane, which indicates the
line but not the depth of the disturbance, other observations may be
made which give the clew to the depth. Thus a building may be found
where the northwest corner at its upper part has been thrown off. Such
a rupture was clearly caused by an upward but oblique movement, which
in the first half of the oscillation heaved the structure upwardly
into the northwest, and then in the second half, or rebound, drew the
mass of the building away from the unsupported corner, allowing that
part of the masonry to fly off and fall to the ground. Constructing a
line at right angles to the plane of the fracture, it will be found to
intersect the plane, the position of which has been in part determined
by finding the line where it intersects the earth, or the seismic
vertical before noted. Multiplying such observations on either side of
the last-mentioned line, the attitude of the underground par