by the application of a few drops of strong potash,
the area at A being left uninjured. A current is now observed to flow,
in the stalk, from the injured B to the uninjured A, as was found to be
the case in the animal tissue. The potential difference depends on the
condition of the plant, and the season in which it may have been
gathered. In the experiment here described (fig. 6, _a_) its value was
.13 volt.

[Illustration: FIG. 6.--(_a_) EXPERIMENT FOR EXHIBITING ELECTRIC
RESPONSE IN PLANTS BY METHOD OF NEGATIVE VARIATION. (_b_) RESPONSES
IN LEAF-STALK OF TURNIP TO STIMULI OF TWO SUCCESSIVE TAPS, THE SECOND
BEING STRONGER.
A and B contacts are about 2 cm. apart, B being injured. Plant is
stimulated by a tap between A and B. Stimulus acts on both A and B,
but owing to injury of B, effect at A is stronger and a negative
variation due to differential action occurs.]

A sharp tap was now given to the stalk, and a sudden diminution, or
negative variation, of current occurred, the resting potential
difference being decreased by .026 volt. A second and stronger tap
produced a second response, causing a greater diminution of P.D. by
.047 volt (fig. 6, _b_). The accompanying figure is a photographic
record of another set of response-curves (fig. 7). The first three
responses are for a given intensity of stimulus, and the next six in
response to stimulus nearly twice as strong. It will be noticed that
fatigue is exhibited in these responses. Other experiments will be
described in the next chapter which show conclusively that the response
was not due to any accidental circumstance but was a direct result of
stimulation. But I shall first discuss the experimental arrangements and
method of obtaining these graphic records.

[Illustration: FIG. 7.--RECORD OF RESPONSES IN PLANT (LEAF-STALK OF
CAULIFLOWER) BY METHOD OF NEGATIVE VARIATION
The first three records are for stimulus intensity 1; the next six are
for intensity twice as strong; the successive responses exhibit
fatigue. The vertical line to the left represents .1 volt. The record
is to be read from right to left.]

#Response recorder.#--The galvanometer used is a sensitive dead-beat
D'Arsonval. The period of complete swing of the coil under experimental
conditions is about 11 seconds. A current of 10^{-9} ampere produces a
deflection of 1 mm. at a distance of 1 metre. For a quick and accurate
method of obtaining the records, I devi