[Illustration: FIG. 32.--INCREASED RESPONSE WITH INCREASING VIBRATIONAL STIMULI (CAULIFLOWER-STALK) Stimuli applied at intervals of three minutes. Vertical line = .1 volt.] (_B_) 1. The vibrational stimulus was increased from 2.5 deg. to 5 deg. to 7.5 deg. to 10 deg. to 12.5 deg. in amplitude. It will be observed how the intensity of response tends to approach a limit (fig. 32). TABLE SHOWING THE INCREASED E.M. VARIATION PRODUCED BY INCREASING STIMULUS +----------------------------------------+ | Angle of Vibration | E.M.F | +----------------------------------------+ | 2.5 deg. | .044 volt | | 5 deg. | .075 volt | | 7.5 deg. | .090 volt | | 10 deg. | .100 volt | | 12.5 deg. | .106 volt | +----------------------+-----------------+ 2. The next figure shows how little variation is produced with low value of stimulus, but with increasing stimulus the response undergoes a rapid increase, after which it tends to approach a limit (fig. 33, _a_). [Illustration: FIG. 33.--RESPONSES TO INCREASING STIMULI PRODUCED BY INCREASING ANGLE OF VIBRATION (_a_) Record with a specimen of fresh radish. Stimuli applied at intervals of two minutes. The record is taken for one minute. (_b_) Record for stale radish. There is a reversed response for the feeble stimulus of 5 deg. vibration.] 3. As an extreme instance of the case just cited, I have often come across a curious phenomenon. During the gradual increase of the stimulus from a low value there would be apparently no response. But when a critical value was reached a maximum response would suddenly occur, and would not be exceeded when the stimulus was further increased. Here we have a parallel to what is known in animal physiology as the 'all or none' principle. With the cardiac muscle, for example, there is a certain minimal intensity which is effective in producing response, but further increase of stimulus produces no increase in response. 4. From an inspection of the records of responses which are given, it will be seen that the slope of a curve which shows the relation of stimulus to response will at first be slight, the curve will then ascend rapidly, and at high values of stimulus tend to become horizontal. The curve as a whole becomes, first slightly convex to the abscissa, then straight and ascending, and lastly concave. A far more pronounced convexity in the first part is shown in some cases, especially when the specimen is stale. This is due to the fact that under these circumstances response is apt to begin with an actual reversal of sign, the plant under feebler than a certain critical intensity of stimulus giving positive, instead of the normal negative, response (fig. 33, _b_).