The following is abstracted from an article by M. E. Branly in
_La Lumière Electrique_ of May 16, 1891, and is taken from _The
Electrician_ of June 26, 1891:--
The object of this article is to describe the first results obtained in
an investigation of the variation of resistance of a large number of
conductors under various electrical influences. The substances which up
to the present have presented the greatest variations in conductivity
are the powders or filings of metals. The enormous resistance offered
by metal in a state of powder is well known; indeed, if we take a
somewhat long column of very fine metallic powder the passage of
the current is completely stopped. The increase in the electrical
conductivity by pressure of powdered conducting substances is well
known, and has had various practical applications. The variations of
conductivity, however, which occur on subjecting conducting bodies to
various electrical influences have not been previously investigated.
_The Effect of Electric Sparks._--Let us take a circuit comprising a
single cell, a galvanometer, and some powdered metal enclosed in an
ebonite tube of 1 square centimetre cross-section and a few centimetres
long. Close the extremities of the tube with two cylindrical copper
tubes pressing against the powdered metal and connected to the rest of
the circuit. If the powder is sufficiently fine, even a very sensitive
galvanometer does not show any evidence of a current passing. The
resistance is of the order of millions of ohms, although the same
metal melted or under pressure would only offer (the dimensions being
the same) a resistance equal to a fraction of an ohm. There being,
therefore, no current in the circuit, a Leyden jar is discharged at
some little distance off, and the abrupt and permanent deflection
of the galvanometer needle shows that an immediate and a permanent
reduction of the resistance has been caused. The resistance of the
metal is no longer to be measured in millions of ohms, but in hundreds.
Its conductivity increases with the number and intensity of the sparks.
Some 20 or 30 centimetres from a circuit comprising some metallic
filings contained in an ebonite cup, let us place a hollow brass
sphere, 15 to 20 centimetres in diameter, insulated by a vertical
glass support. The filings offer an enormous resistance and the
galvanometer needle remains at zero. But if we bring an electrified
stick of resin near the sphere, a little spark will pass between the
stick and the sphere, and immediately the needle of the galvanometer
is violently jerked and then remains permanently deflected. On some
fresh filings being placed in the ebonite cup, the resistance of the
circuit will again keep the needle at zero. If now the charged brass
sphere is touched with the finger, there is a minute discharge and
the galvanometer needle is again deflected. With a few accumulators
the experiment can easily be made without a galvanometer. The circuit
consists of the battery, some metallic powder, a platinum wire, and
a mercury cup. The resistance of the powder is so high that the
interruption of the circuit takes place without any sparking at the
mercury cup. If now a Leyden jar is discharged in the neighbourhood
of the circuit the powder is rendered conducting, the platinum wire
immediately becomes red hot, and a violent spark occurs on breaking the
circuit.
The influence of the spark decreases as the distance increases, but
its influence is observable several metres away from the powder,
even with a small Wimshurst machine. Repeating the spark increases
the conductivity; in fact, with certain substances successive sparks
produce successive jerks, and a gradually increasing and persistent
deflection of the galvanometer.
_Influence of a Conductor traversed by Condenser Discharges._--While
using a Wimshurst machine it was noticed that the reduction in the
resistance of the filings frequently took place before discharge. This
led me to the following experiment: Take a long brass tube, one end
of which is close to the circuit containing the metallic powder; its
other end, several metres distant from the circuit, is fairly close to
a charged Leyden jar. A spark takes place and the conductor is charged.
At the same instant, the conductivity of the metallic powder is greatly
increased.
The following arrangement, owing to its efficacy, convenience, and
regularity of action was used by me in most of my researches, and I
shall briefly call it the A arrangement (_see_ Fig. 53).
[Illustration: FIG. 53.]
[Illustration: FIG. 54.]
The source of electricity is a two-plate Holtz machine driven at from
100 to 400 revolutions. A sensitive substance is introduced into one
of the arms of a Wheatstone bridge, or into the circuit of a single
Daniell cell at a distance of some 10 metres (34ft.) from the Holtz
machine. Between the discharge knobs of the machine and the Wheatstone
bridge, and connected to the former, there are two insulated brass
tubes, A A′, running parallel to one another 40 centimetres apart.
The Leyden jars usually attached to a Holtz machine may be dispensed
with, the capacity of the long brass tubes being in some measure
equivalent to them. The knobs S were 1 mm., ·5 mm., or ·1 mm. apart.
When the plates were rotated sparks rapidly succeeded each other.
Experiments showed that these sparks had no direct effect at a distance
of 10 metres. The two tubes A A′ are not absolutely necessary, the
diminution of resistance is easily produced if only one is employed,
and in some cases, indeed, a single conductor is more efficacious.
An increase in the speed of the machine increases its action to a
marked extent. The sparks at S may be suppressed by drawing the knobs
apart, but the conductor A will still continue to exert its influence,
especially if there is a spark gap anywhere about.
_Effects of Induced Currents._--The passage of induced currents
_through_ a sensitive substance produces similar effects to those
described above. In one instance an induction coil was taken, having
two similar wires. The circuit of the secondary wire was closed through
a tube containing filings, the galvanometer being also in circuit.
Care was taken to ascertain before introducing the filings into the
circuit that the currents on make-and-break gave equal and opposite
deflections. Filings were then introduced into the circuit, the primary
being made and broken at regular intervals. The following table gives
the results obtained in the case of zinc filings:--