It is thus found that growth in plants is affected by an intensity of stimulus which is below human perception; that with increasing stimulus the latent period is diminished and the period of recovery increased; and that the induced retardation of growth increases continuously with the stimulus till at a critical intensity there is a temporary arrest of growth. I shall speak later of the effect induced by stimulus above this critical point. [Illustration: FIG. 62.--Record of a single growth-pulse of _Zephyranthes_ (Magnification 10,000 times.)] _Experiment 56._--As a further example of the capability of the Crescograph, I shall give the record of a single pulse of growth obtained with the peduncle of _Zephyranthes Sulphurea_ (Fig. 62). The magnification employed was 10,000 times, the successive dots being at intervals of one second. It will be seen that the growth pulse commences with a sudden elongation, the maximum rate being 0.4 µ per sec. The pulse exhausts itself in 15 seconds, after which there is a partial recovery in the course of 13 seconds. The period of the complete pulse is 28 seconds. The resultant growth in each pulse is therefore the difference between elongation and recovery. Had a very highly magnifying arrangement not been used, the resulting rate would have appeared continuous. In other specimens, owing probably to greater frequency of pulsation and co-operation of numerous elements in growth, the rate appears to be practically uniform. _Advantages of the Crescograph._--There is no existing method which enables us to detect and measure such infinitesimal movements and their time-relations. The only attempt made in measuring minute growth has been by observing the movement of a mark on a growing plant through a microscope. The magnification available in practice is about 250 times. The observation of the movement would itself be sufficiently fatiguing. But a simultaneous estimate of the time-relations of rapidly fluctuating changes would prove so bewildering, that accurate results from this method would be altogether impossible. A 1/12″ objective gives a linear enlargement of about 1,200 times. But the employment of this objective is impracticable in the measurement of growth elongation of an ordinary plant. With the Crescograph, on the other hand, we obtain a magnification which far surpasses the highest powers of a microscope, and it can be used for all plants. It does not merely detect growth but automatically records the rate of growth and its slightest fluctuation. The extreme shortness of time required for an experiment renders the study of the influence of a single factor at a time possible, the other conditions being kept constant. The Crescograph thus opens out a very extensive field of inquiry into the physiology of growth; and the discovery of several important phenomena mentioned in this Paper is to be ascribed to the extreme sensitiveness of the apparatus, and the accuracy of the method employed.