The contrivance which I employ for ensuring a gradual rise of temperature, consists of a double-walled cylindrical metallic vessel; the plant is placed in the inner chamber, the walls of which are coated with electrically deposited silver and polished afterwards, and at the bottom of which there is a little water. The space between the inner and outer cylinder is filled with water, in which is immersed a coiled copper pipe. Hot water from a small boiler enters the inlet of the coiled pipe and passes through the outlet at the lower end. The water in the outer cylinder is thus gradually raised by flow of hot water in the coiled pipe. The rate of flow of hot water, on which the rate of rise of temperature depends, is regulated by a stop-cock. The air of the inner chamber in which the plant is placed, may thus be adjusted for a definite temperature. The small quantity of water in the inner chamber keeps its air in a humid condition, since dry hot air by causing dessication interferes with normal growth. METHOD OF DISCONTINUOUS OBSERVATION. [Illustration: FIG. 63.--Effect of temperature on growth, and determination of optimum temperature.] _Experiment 58._--High magnification records are taken for successive periods of ten seconds, for selected temperatures, maintained constant during the particular observation. In figure 63 is given records of rate of growth obtained with a specimen of _Kysoor_ at certain selected temperatures. It will be seen that the rate of growth increases with the rise of temperature to an optimum, beyond which the growth-rate undergoes a depression. In the present case the optimum temperature is in the neighbourhood of 35°C. METHOD OF CONTINUOUS OBSERVATION. The method of observation that I have described above is not ideally perfect, but the best that could be devised under the circumstances. A very troublesome complication of pulsations in growth, arises at high temperatures, which render further record extremely difficult. Growth is undoubtedly a pulsatory phenomenon; but under favourable circumstances, these merge practically into a continuous average rate of elongation. At a high temperature the effect of certain disturbing factors comes into prominence. This may be due to some slight fluctuation in the temperature of the chamber, or to the effect of thermal radiation from the side of the chamber. This disturbing influence is most noticed at about 45°C, rendering the record of growth above this point a matter of great uncertainty. It will presently be shown that in plants immersed in water-bath growth is often found to persist even up to 57°C.