EFFECT OF CONSTANT TEMPERATURE. The second test which I shall employ is the effect of maintenance of constant temperature, which should wipe off, as it were, traces of periodic movement. It was necessary for this investigation to maintain the plant chamber at constant temperature throughout day and night. The usual thermostat is virtually a recess in a double-walled chamber filled with water, the chamber being covered with a heat insulating material. But this contrivance is unsuitable for the plant chamber which is to contain good sized plants, and the recording apparatus. The problem of maintaining a large air-chamber at constant temperature presented many difficulties which were ultimately overcome by the device of an extremely sensitive thermal regulator. _The Thermal Regulator._--I shall in a future paper give a complete account of the large thermostatic air-chamber. The important part of the apparatus is an electro-thermic regulator which interrupts the heating electric current as soon as the temperature of the chamber is raised a hundredth part of a degree above the predetermined temperature. The automatic make and break of the current takes place in rapid succession, and the temperature of the chamber is thus maintained constant within tenth of a degree, throughout day and night. [Illustration: FIG. 203.--Abolition of diurnal movement in _Tropæolum_ under constant temperature, and its restoration under normal daily fluctuation. The upper record is of temperature and the lower of plant movement.] _Diurnal record of_ Tropæolum _under constant temperature: Experiment 212._--The normal record of geotropically curved _Tropæolum_ is already given in figure 202. In repeating the record I maintained the plant at constant temperature for 24 hours; the result of this is seen in the first part of the record (Fig. 203). The thermal record is practically horizontal, and the diurnal record of the plant shows no periodic movement. The thermal regulator was on the next day put out of operation, thus restoring the normal diurnal variation of temperature. The record of the plant is seen to exhibit once more its normal periodic movement. I have in the chapter on thermo-geotropism (p. 515) shown that the diurnal movement of a geotropically curved organ is determined in reference to the direction of force of gravity. This will be seen demonstrated in an interesting manner in the two following experiments on the effect of inversion of the plant on daily movement. DIURNAL MOVEMENT IN INVERTED POSITION.