effected in electric clocks, that rather than go to the trouble and
expense required by such precautions, it appears far preferable to keep
an accurate time-piece in some sheltered position and use it with a
source of electricity to drive the hands of the large dial.

_Electrical Clocks._--One of the first attempts to apply electricity to
clocks was made by Alexander Bain in 1840-1850. About the same time Sir
C. Wheatstone, R. L. Jones, C. Shepherd, Paul Garnier and Louis Breguet
invented various forms of electrical time-keepers. It is not proposed
here to go into the history of these abortive attempts. Those who desire
to follow them may consult Bain, _An Account of Some Applications of the
Electric Fluid to the Useful Arts_ (1843) and _Short History of Electric
Clocks_ (1852); Sir Charles Wheatstone, _Trade Circular of the British
Telegraph Manufactory_; C. Shepherd, _On the Application of
Electro-magnetism as a Motor for Clocks_ (1851), and a list of
references in the Appendix to Tobler's _Die electrischen Uhren_
(Leipzig, 1883), and a list of books given by F. Hope Jones, _Proc.
Inst. Elec. Eng._, 1900, vol. 29. The history of electrical clocks is a
long and complicated matter, for there are some 600 or 700 patents for
these clocks in Europe and America, some containing the germs of
valuable ideas but most pure rubbish. All that can be done is to select
one or two prominent types of each class and give a brief description of
their general construction.

[Illustration: FIG. 27.--Turret Clock for Hidalgo, Mexico, driving four
8 ft. dials.]

It is in the apparently simple matter of making and keeping the
electrical contact that most of the systems of electrical time-keeping
have failed, for want of attention to the essential conditions of the
problem. In practice every metal is covered with a thin film of
non-conducting oxide over which is another film of moisture, oil, dirt
or air. Hence what is wanted is a good vigorous push of a blunted point
or edge preferably obliquely upon a more or less yielding surface so as
to get a rubbing action. Thus if the stiff spring a b (fig. 28) were
stabbed down on the oblique surface C D a good contact would invariably
result, provided that the metals employed were gold, platinum or some
not easily oxidizable metal. Or again, if a mercury surface be simply
touched with a pin, the slight sparking that is produced on making the
current will soon form a little pile of dirty oxide at