When electrolytic contacts are made on the uninjured surfaces of the stalk at A and B, the two points, being practically similar in every way, are iso-electric, and little or no current will flow in the galvanometer. If now the whole stalk be uniformly stimulated, and if both ends A and B be equally excited at the same moment, it is clear that there will still be no responsive current, owing to balancing action at the two ends. This difficulty as regards the obtaining of response was overcome in the method of negative variation, where the excitability of one end was depressed by chemical reagents or injury, or abolished by excessive temperature. On stimulating the stalk there was produced a greater excitation at A than at B, and a current of action was then observed to flow in the stalk from the more excited A to the less excited B (fig. 6). But we can cause this differential action to become evident by another means. For example, if we produce a block, by clamping at C between A and B (fig. 14, _a_), so that the disturbance made at A by tapping or vibration is prevented from reaching B, we shall then have A thrown into a relatively greater excitatory condition than B. It will now be found that a current of action flows in the stalk from A to B, that is to say, from the excited to the less excited. When the B end is stimulated, there will be a reverse current (fig. 14, _b_). [Illustration: FIG. 14.--THE METHOD OF BLOCK (_a_) The plant is clamped at C, between A and B. (_b_) Responses obtained by alternately stimulating the two ends. Stimulation of A produces upward response; of B gives downward response.] We have in this method a great advantage over that of negative variation, for we can always verify any set of results by making corroborative reversal experiments. By the method of injury again, one end is made initially abnormal, i.e. different from the condition which it maintains when intact. Further, inevitable changes will proceed unequally at the injured and uninjured ends, and the conditions of the experiment may thus undergo unknown variations. But by the block method which has just been described, there is no injury, the plant is normal throughout, and any physiological change (which in plants will be exceedingly small during the time of the experiment) will affect it as a whole.