(3) that it may be induced not merely by stimulus of light but by all forms of stimulation, (4) that the direction of the torsional response depends on the direction of the incident stimulus and the differential excitability of the organ, and (5) that there is a definite law which determines the torsional movement. EXPERIMENTAL ARRANGEMENTS. I shall first describe a typical experiment on the responsive torsion under the action of light. We have seen that in the pulvinus of _Mimosa_, light of moderate intensity and of short duration applied on the upper half induces a slow up-movement, while the stimulus of light applied below induces a more rapid down-movement. The difference is due to the fact that the lower half of the pulvinus is relatively the more excitable. Vertical light thus induces a movement in a vertical plane. But an interesting variation is induced in the response under the action of lateral light. A stimulus will be called _lateral_ when it acts on either the right or left flank of a _dorsiventral_ organ. We shall presently find that a dorsiventral organ responds to lateral stimulus by torsion. The present series of experiments were carried out with the leaf of _Mimosa_, and in order to eliminate the effect of weight and also for obtaining record of pure torsion, I employed the following device. The petiole was enclosed in a hooked support made of thin rod of glass, the petiole resting on the concavity of the smooth surface. Friction and the effect of weight is thus practically eliminated; the looped support prevented up or down movements, and yet allowed perfect freedom for torsional response. This latter is magnified by a piece of stout aluminium wire fixed at right angles to the petiole (Fig. 148). The end of the aluminium wire is attached to the short arm of a recording lever; there is thus a compound magnification of the torsional movement. The Oscillating Recorder gave successive dots at intervals which could be varied from 20 seconds to 2 minutes. Time-relations of the response may thus be obtained from the dotted record. [Illustration: FIG. 148.--Diagrammatic representation for record of torsional response. H, thin glass hook: A, aluminium wire attached to petiole for magnification of torsional movement. T, silk thread for attachment to recording lever.] With the experimental device just described, we shall be in a position to study the effect of various stimuli applied at one flank of the pulvinus--at the junction of the upper and lower halves of the organ. The observer standing in front of the leaf is supposed to look at the stem. Torsional response will then appear as a movement either with or against the hands of the clock. The torsional response, right-handed or left-handed, will presently be shown to depend on the direction of incident stimulus. In figure 149, anti-clockwise torsion is recorded as an up-curve; clockwise rotation is recorded as a down-curve.