trong stimulus.

From fig. 33, _b_, above, it is seen that there was an actual reversal
of response in the lower portion of the curve. It is therefore not
improbable that there may be more than one point of reversal.

In physical phenomena we are, however, acquainted with numerous
instances of reversals. For example, a common effect of magnetisation
is to produce an elongation of an iron rod. But Bidwell finds that as
the magnetising force is pushed to an extreme, at a certain point
elongation ceases and is succeeded, with further increase of magnetising
force, by an actual contraction. Again a photographic plate, when
exposed continuously to light, gives at first a negative image. Still
longer exposure produces a positive. Then again we have a negative.
There is thus produced a series of recurrent reversals. In photographic
prints of flashes of lightning, two kinds of images are observed, one,
the positive--when the lightning discharge is moderately intense--and
the other, negative, the so-called 'dark lightning'--due to the reversal
action of an intensely strong discharge.

In studying the changes of conductivity produced in metallic particles
by the stimulus of Hertzian radiation, I have often noticed that whereas
feeble radiation produces one effect, strong radiation produces the
opposite. Again, under the continuous action of electric radiation, I
have frequently found recurrent reversals.[13]

#Diminution of response under strong stimulus traced to fatigue.#--But
there are instances in plant response where the diminution effect can be
definitely traced to fatigue. The records of these cases are extremely
suggestive as to the manner in which the diminution is brought about.
The accompanying figures (fig. 34) give records of responses to
increasing stimulus. They were made with specimens of cauliflower-stalks,
one of which (_a_) showed little fatigue, while in the other (_b_)
fatigue was present. It will be seen that the curves obtained by joining
the apices of the successive single responses are very similar.

[Illustration: FIG. 34.--RESPONSES TO INCREASING STIMULUS OBTAINED WITH
TWO SPECIMENS OF STALK OF CAULIFLOWER
In (_a_) fatigue is absent, in (_b_) it is present.]

In one case there is no fatigue, the recovery from each stimulus being
complete. Every response in the series therefore starts from a position
of perfect equilibrium, and the height of the single responses increases
with increasing stimul