T_) with one pole of a Daniell's battery of two cells,
and the basin, _f_, is connected with the other pole. Any vertical
movement, however slight, makes the platinum point dip into the mercury,
and thus completes the circuit. In this circuit are included two
electro-magnets, _C_ and _D_; these, during the circulation of a
current, attract their armatures, which are connected with levers. The
action of _C's_ lever is to stop the clock, _A_, which thus records, to
a half-second, the time of the occurrence of the shock, at the same
instant that the clock strikes an alarm bell, which attracts the
attention of an observer. The lever, attached to the armature of _D_, at
the first instant of the current frees the pendulum of the clock, _B_,
which was before kept from swinging, in a position out of the vertical;
the clock then acts as a time-piece, and its motion unrolls a band of
paper, _k k k_, at a rate of three metres an hour. At the same time the
armature of _D_, while attracted, presses a pencil point against the
band of paper which passes over the roller, _m_, marking on it, while
the earthquake lasts, a series of points or strokes which occupy a
length of paper corresponding to its duration, and which record the work
of the shock. After it is over the paper continues to unroll from the
drum, _i_, and passing round the clock, rolls on to the drum, _l_. If a
fresh shock occur the pencil indicates it, as before, on the paper, and
the length of blank paper between the two sets of marks is a measure of
the interval of time between the shocks. By way of additional check,
several helices, _h h h_, are hung from a stand, with small permanent
magnets suspended from their ends; below and close to these latter are
small basins, holding iron filings; into these the points of the magnets
dip, when their helices oscillate vertically, and some filings remain
sticking to the magnets as a record of the shock. One of the magnets has
a shoulder on it which moves an index hand along a graduated arc, as
shown in Fig. 2, thus again registering the amount of the vertical
movement. Such are the arrangements intended for the record of the
undulatory or horizontal elements of the wave of shock.

The following are the arrangements proposed for recording the horizontal
motions: On the stand, to the right of the clock, _A_, are set four
U-shaped glass tubes, open at their ends. One of each pair of vertical
branches must have a diameter at least dou