(3) Under very strong light the induced retardation is very great, and this persists for a long time even on the removal of light. The experiments described explains the reasons of complete recovery after moderate stimulation, and also the absence of recovery after strong stimulation. The immediate after-effect of moderate stimulation is shown to be an acceleration of rate above the normal. Returning to tropic curvature, the contraction at the proximal side induced by unilateral light is thus compensated by the accelerated rate of growth on the cessation of light. There is no such compensation in the case of strong and long continued action of light; for the after-effect of strong light shows no such acceleration as the immediate after-effect. We may perhaps go a step further in explaining this difference. Stimulus was found to induce at the same time two physico-chemical reactions of opposite signs (p. 144). One is the 'up' or A-change, associated with increase of potential energy of the system, and the other is associated with 'down' or D-change, by which there is a run-down or depletion of energy. With moderate stimulation the A-and-D effects are more or less comparable to each other. But under strong stimulation the down-change is relatively greater. Hence on cessation of moderate stimulation the increase of potential energy, associated with A-change, finds expression in enhancement of the rate of growth. The depletion of energy under strong stimulation is, however, too great to be compensated by the A-change. LATENT PERIOD OF PHOTOTROPIC REACTION. With reference to the latent period Jost thus summarizes the known results:[11] "The latent period of the heliotropic stimulus has already been determined. According to Czapek it amounts to 7 minutes in the cotyledons of _Avena_ and in _Phycomyces_; 10 minutes in hypocotyls of _Sinapis alba_ and _Beta vulgaris_, 20 minutes in the hypocotyl of _Helianthus_, and 50 minutes in the epicotyl of _Phaseolus_. If one of these organs be unilaterally illuminated for the specified time, heliotropic curvature ensues afterwards in the dark, that is to say, we meet with an after-effect in this case as in geotropism. We are quite ignorant, however, as to whether and how the latent period is dependent on the intensity of light." [11] Jost--_Ibid_, p. 473. With regard to the question of relation of the latent period to the intensity of stimulus I have shown (p. 166) that the latent period is shortened under increasing intensity of stimulus. In the case of tropic curvature induced by light, I find that the latent period is reduced under increasing intensity of light. The shortest latent period found by Czapek, as stated before, was 7 minutes. But by employing high magnification for record, I find that the latent period of phototropic action under strong light to be a question of seconds.