thus sending an impulse. It is here that all previous methods have
failed. This contact should be absolute and positive, for if it is
not, the receiver will not work in unison. The contact could often be
doubled by the jarring of the instrument, thus making the receiver
jump twice. Clarke has overcome this defect by so arranging his
mechanism that the faintest contact in the primary instrument closes
two platinum points in multiple arc with it, thus making a firm and
positive contact, which is not disturbed by any jar on the primary
contact. This gives the instruments a positive start for the series of
operations, instead of the faint contact which would be given, for
example, by the light and slowly moving hand of a metallic
thermometer. The other trouble with previous methods was that the
contact points would corrode, and, in consequence of such corrosion,
the instrument would fail to send impulses. Corrosion of the contacts
is due to breaking the circuit slowly on a small surface. This is
entirely remedied by breaking the circuit elsewhere than at the
primary contact, using a quick motion, and also by giving this
breaking contact large surface and making it firm. The instrument, as
applied to a thermometer, is made as follows: From the free end of the
light spiral of a metallic thermometer fixed at the other end, an arm,
C, is attached, the end of which moves over an arc of a circle when
the temperature varies. This end carries on either side of its
extremity platinum contacts which, when the thermometer is at rest,
lie between two other platinum points, A B, carried on radial arms.
Any variation in temperature brings a point on the thermometer arm in
contact with one of these points, and thus gives the initial start to
the series of operations without opposing any friction to the free
motion of the instrument. The first result is the closing of a short
circuit round the initial point of contact, so that no current flows
through it. Then the magnets which operate one set of pawls come into