ht by a nick in it, and forced downwards. In this way the circuit is
closed. Fig. 31 is a diagram of the apparatus. When the block g attached
to the pendulum catches and presses down the flipper s, the lever l l is
rocked over, so that a contact is made at k, and the current which
enters the lever l through the knife edge m, runs through the second
lever n n, down through the knife edge o, to the battery, and through
the electromagnet b which causes the armature a to be attracted. As the
block g goes on and releases s, the lever l again falls upon the rest p,
the lever n follows it a part of the way till it is stopped by the
contact q; this shortcircuits the electromagnet and prevents to a large
extent the formation of an induced current. It is claimed that sparking
is by this method almost entirely avoided. It is only when s is caught
in the notch of the block g that s is pressed down, so that the electric
attraction only takes place every few vibrations. This ingenious
arrangement makes the working of the clock nearly independent of the
strength of the battery, for if the battery is strong the impulses are
fewer and the _average arc_ remains the same. The clock is enclosed in
an airtight glass case so as to avoid barometric error. It was tested in
1905 at the Neuchatel observatory. In winter in a room of a mean
temperature of 35 deg. F. it was 1/4 sec. too slow, in summer when the
temperature was 70 deg., it was 1/2 sec. too fast. In the succeeding
winter it became .53 sec. too slow again, thus gaining a little in
summer and losing in winter. Its average variation from its daily rate
was, however, only .033 sec.

[Illustration: FIG. 30.--Hipp Electrical Clock (Peyer, Favarger et
Cie.).]

[Illustration: FIG. 31.--Contact Arrangement of Hipp Clock.]

In another system originated by G. Froment, a small weight is raised by
electricity and allowed to fall upon an arm sticking out at right angles
to the pendulum in the plane of its motion, so as to urge it onwards.
The weight is only allowed to rest on the arm during the downward swing
of the pendulum. The method is not theoretically good, as the impulse is
given at the end of the vibration of the pendulum instead of at its
middle position.

In the clock invented by C. Fery (chef des travaux pratiques at the
Ecole de Physique et Chimie, Paris), an electric impulse is given at
every vibration, not by a battery but by means of the uniform movement
of an armature which is