strata included between them, and the directions
of the currents would agree with those observed as above.

[A] 1830. p. 399.

188. Although the electricity obtained by magneto-electric induction in a
few feet of wire is of but small intensity, and has not yet been observed
except in metals, and carbon in a particular state, still it has power to
pass through brine (23.); and, as increased length in the substance acted
upon produces increase of intensity, I hoped to obtain effects from
extensive moving masses of water, though quiescent water gave none. I made
experiments therefore (by favour) at Waterloo Bridge, extending a copper
wire nine hundred and sixty feet in length upon the parapet of the bridge,
and dropping from its extremities other wires with extensive plates of
metal attached to them to complete contact with the water. Thus the wire
and the water made one conducting circuit; and as the water ebbed or flowed
with the tide, I hoped to obtain currents analogous to those of the brass
ball (161.).

189. I constantly obtained deflections at the galvanometer, but they were
very irregular, and were, in succession, referred to other causes than that
sought for. The different condition of the water as to purity on the two
sides of the river; the difference in temperature; slight differences in
the plates, in the solder used, in the more or less perfect contact made by
twisting or otherwise; all produced effects in turn: and though I
experimented on the water passing through the middle arches only; used
platina plates instead of copper; and took every other precaution, I could
not after three days obtain any satisfactory results.

190. Theoretically, it seems a necessary consequence, that where water is
flowing, there electric currents should be formed; thus, if a line be
imagined passing from Dover to Calais through the sea, and returning
through the land beneath the water to Dover, it traces out a circuit of
conducting matter, one part of which, when the water moves up or down the
channel, is cutting the magnetic curves of the earth, whilst the other is
relatively at rest. This is a repetition of the wire experiment (171.), but
with worse conductors. Still there is every reason to believe that electric
currents do run in the general direction of the circuit described, either
one way or the other, according as the passage of the waters is up or down
the channel. Where the lateral extent of the moving water is eno