e in a
most extraordinary degree independent of those variations or modes of
action which it presents directly in its course; it consequently is of the
more value to us, as it gives us a higher relation of the power than any
that might have varied with each mode of discharge. This discharge, whether
it be by conduction through a wire with infinite velocity (1652.), or by
electrolyzation with its corresponding and exceeding slow motion (1651.),
or by spark, and probably even by convection, produces a transverse
magnetic action always the same in kind and direction.

1655. It has been shown by several experimenters, that whilst the discharge
is of the _same kind_ the amount of lateral or magnetic force is very
constant (216. 366. 367. 368. 376.). But when we wish to compare discharge
of different kinds, for the important purpose of ascertaining whether the
same amount of current will in its _different forms_ produce the same
amount of transverse action, we find the data very imperfect. Davy noticed,
that when the electric current was passing through an aqueous solution it
affected a magnetic needle[A], and Dr. Ritchie says, that the current in
the electrolyte is as magnetic as that in a metallic wire[B], and has
caused water to revolve round a magnet as a wire carrying the current would
revolve.

[A] Philosophical Transactions, 1821, p. 426.

[B] Ibid. 1832, p. 294.

1656. Disruptive discharge produces its magnetic effects: a strong spark,
passed transversely to a steel needle, will magnetise it as well as if the
electricity of the spark were conducted by a metallic wire occupying the
line of discharge; and Sir H. Davy has shown that the discharge of a
voltaic battery in vacuo is affected and has motion given to it by
approximated magnets[A].

[A] Philosophical Transactions, 1821, p. 427.

1657. Thus the three very different modes of discharge, namely, conduction,
electrolyzation, and disruptive discharge, agree in producing the important
transverse phenomenon of magnetism. Whether convection or carrying
discharge will produce the same phenomenon has not been determined, and the
few experiments I have as yet had time to make do not enable me to answer
in the affirmative.

* * * * *

1658. Having arrived at this point in the consideration of the current and
in the endeavour to apply its phenomena as tests of the truth or fallacy of
the theory of induction which I have ventured