f the magnetic needle went west; the electric current was therefore
from north to south in the part of the wire passing under the needle, and
from south to north in the moving or upper part of the parallelogram. On
passing the upper part of the rectangle from east to west over the
galvanometer, the marked pole of the needle went east, and the current of
electricity was therefore the reverse of the former.

173. When the rectangle was arranged in a plane east and west, and the
magnetic needle made parallel to it, either by the torsion of its
suspension thread or the action of a magnet, still the general effects were
the same. On moving the upper part of the rectangle from north to south,
the marked pole of the needle went north; when the wire was moved in the
opposite direction, the marked pole went south. The same effect took place
when the motion of the wire was in any other azimuth of the line of dip;
the direction of the current always being conformable to the law formerly
expressed (114.), and also to the directions obtained with the rotating
ball (101.).

174. In these experiments it is not necessary to move the galvanometer or
needle from its first position. It is quite sufficient if the wire of the
rectangle is distorted where it leaves the instrument, and bent so as to
allow the moving upper part to travel in the desired direction.

175. The moveable part of the wire was then arranged _below_ the
galvanometer, but so as to be carried across the dip. It affected the
instrument as before, and in the same direction; i.e. when carried from
west to east under the instrument, the marked end of the needle went west,
as before. This should, of course, be the case; for when the wire is
cutting the magnetic dip in a certain direction, an electric current also
in a certain direction should be induced in it.

176. If in fig. 31 _dp_ be parallel to the dip, and BA be considered as the
upper part of the rectangle (171.), with an arrow _c_ attached to it, both
these being retained in a plane perpendicular to the dip,--then, however BA
with its attached arrow is moved upon _dp_ as an axis, if it afterwards
proceed in the direction of the arrow, a current of electricity will move
along it from B towards A.

177. When the moving part of the wire was carried up or down parallel to
the dip, no effect was produced on the galvanometer. When the direction of
motion was a little inclined to the dip, electricity manifested itself; an