_Experiment 103._--I have already explained, how thermal _radiation_ is almost as effective in inducing contraction and retardation of growth as the more refrangible rays of the spectrum. The thermal radiation was produced by the heating of a platinum spiral, short of incandescence, by the passage of an electric current. The intensity of radiation is easily varied by adjustment of the current by means of a rheostat. The experimental specimen was a flower bud of _Crinum_. It was held by a clamp, a little below the region of growth. Stimulus was applied below the clamp so that the transmitted effect had to pass through S, the securely held tissue (Fig. 98). A feeble stimulus was applied on one side, at the indifferent point about 3 cm. below the region of growth. The positive effect of indirect stimulus reached the region of growth on the same side, bringing about an acceleration of growth with expansion and convexity, the resulting movement being _negative_ or away from the stimulus. The latent period was ten seconds, and maximum negative movement was completed in the further course of ten seconds, after which there was a recovery in the course of 75 seconds. A stronger stimulus S' gave a larger response; but when the intensity was raised still higher to S", the excitatory negative impulse overtook the positive within 15 seconds of its commencement; the convex was thus succeeded by the concave curvature (Fig. 99). Direct application of stimulus at the growing region gave rise to a positive curvature. [Illustration: FIG. 98.--Diagrammatic representation of effects of indirect and direct stimulation. Continuous arrow represents the indirect stimulation, and the curved continuous arrow above, the induced negative curvature: dotted arrow indicates the application of direct stimulus, and the dotted curve above, the induced positive curvature.] [Illustration: FIG. 99.--Tropic curvature of _Crinum_ to unilateral indirect stimulation of increasing intensities: S, S' of moderate intensity induced negative tropic effect (movement away from the stimulated side); stronger stimulus S" gave rise to negative followed by positive. Successive dots at intervals of 5 seconds Magnification 100 times.] The effect of feeble stimulus transmitted longitudinally is thus found always to induce convexity, a _negative curvature_ and movement away from stimulus. I have obtained similar responsive movement of negative sign with various plant organs, and under various forms of stimuli. Thus in the stem of _Dregea volubilis_ the longitudinally transmitted effect of light of moderate intensity was a negative curvature; direct application of light on the growing region gave, on the other hand, a positive curvature and movement towards light.