--------+----------------------+-----------------------
Mean 14 | Mean .3 | Mean 0

These results conclusively prove the physiological nature of the
response.

I shall in a succeeding chapter give a continuous series of
response-curves showing how, owing to progressive death from the action
of poison, the responses undergo steady diminution till they are
completely abolished.

#Effect of high temperature.#--It is well known that plants are killed
when subjected to high temperatures. I took a stalk, and, using the
block method, with torsional vibration as the stimulus, obtained strong
responses at both ends A and B. I then immersed the same stalk for a
short time in hot water at about 65 deg. C., and again stimulated it as
before. But at neither A nor B could any response now be evoked. As all
the external conditions were the same in the first and second parts of
this experiment, the only difference being that in one the stalk was
alive and in the other killed, we have here further and conclusive proof
of the physiological character of electric response in plants.

The same facts may be demonstrated in a still more striking manner by
first obtaining two similar but opposite responses in a fresh stalk, at
A and B, and then killing one half, say B, by immersing only that half
of the stalk in hot water. The stalk is replaced in the apparatus, and
it is now found that whereas the A half gives strong response, the end B
gives none.

In the experiments on negative variation, it was tacitly assumed that
the variation is due to a differential action, stimulus producing a
greater excitation at the uninjured than at the injured end. The block
method enables us to test the correctness of this assumption. The B end
of the stalk is injured or killed by a few drops of strong potash, the
other end being uninjured. There is a clamp between A and B. The end A
is stimulated and a strong response is obtained. The end B is now
stimulated, and there is little or no response. The block is now removed
and the plant stimulated throughout its length. Though the stimulus now
acts on both ends, yet, owing to the irresponsive condition of B, there
is a resultant response, which from its direction is found to be due to
the responsive action of A. This would not have been the case if the end
B had been uninjured. We have thus experimentally verified the
assumption that in the same tissue an uninjured porti