come, in an electrical sense, more zinc-like.
On vibrating the copper end (inasmuch as copper would then become more
zinc-like) the difference of potential between zinc and copper ought
to be diminished, and the current flowing in the circuit would
therefore be lessened. But vibration of the zinc end ought to increase
the potential difference, and there ought to be then an increase of
current during stimulation of zinc.

[Illustration: FIG. 51.--CURRENT OF RESPONSE TOWARDS THE STIMULATED END
Hence when Cu stimulated: action current -->, normal E.M.F. diminished
(.85-.009) V.
When Zn stimulated: action current <--, normal E.M.F. increased
(.85+.013) V.]

In the particular experiment of fig. 51, the E.M.F. between the zinc
and copper ends was found to be .85 volt. This was balanced by a
potentiometer arrangement, so that the galvanometer spot came to zero.
On vibrating the zinc wire, a deflection of 33 dns. was obtained, in a
direction which showed an _increase_ of E.M.F. On stopping the
vibration, the spot of light came back to zero. On now vibrating the
copper wire, a deflection of 23 dns. was obtained in an opposite
direction, showing a _diminution_ of E.M.F. This transitory responsive
variation disappeared on the cessation of disturbance.

By disturbing the balance of the potentiometer, the galvanometer
deflection due to a known increase of E.M.F. was found from which the
absolute E.M. variation caused by disturbance of copper or zinc was
determined.

It was thus found that stimulation of zinc had increased the P.D. by
fifteen parts in 1,000, whereas stimulation of copper had decreased it
by eleven parts in 1,000. According to the old terminology, the
response due to stimulation of zinc would have been regarded as
positive variation, that of copper negative. The responses however are
not essentially opposite in character, the action current in the bar
being in both cases towards the more excited. For this reason it would
be preferable, as already said, to employ the terms positive and
negative in the sense I have suggested, i.e. positive, when the
current in the acted substance is towards the more excited, and
negative, when towards the less excited. The method of block is, as I
have already shown, the most perfect for the study of these responses.

In the experiment fig. 50, if the block is abolished and the wire is
struck in the middle, a wa