second, which under pronounced fatigue was reduced to 8 mm.
(3) Temperature also modifies the rate of the responsive movement. Thus in a given specimen the maximum rate of fall at the relatively low temperature of 22° C. was 10 mm. per second; it became enhanced to 105 mm. per second at 25° C., and 115 mm. per second at 31° C.
The pulvinus after the attainment of the maximum fall remains more or less persistently contracted for a short time. This is shown by the horizontal portion of the curve in fig. 12.
The Period of Recovery.—As this takes a relatively long time, its record has to be made on a slowly moving plate. The unit of time-measurement must therefore be relatively long. If the successive dots were to be made at the ordinary rate of 10 per second, they would become fused and continuous in the record. For this reason I have devised a contrivance by which the successive dots, in the recovery-portion of the curve, are placed at such intervals as to prevent overcrowding. A convenient interval is either 5 or 10 seconds.
The device for producing periodic dots at intervals, say of 10 seconds, consists of clockwork employed to interrupt the current actuating the vibrating recorder at particular intervals. A light six-rayed wheel is attached to the axis of the seconds-hand, and during the course of a single complete revolution, that is to say, in a minute, the projecting rays press the spring-key six times at intervals of 10 seconds each (fig. 13). It is only during the short interval when the key is pressed that the circuit is completed and the recorder set in vibration to make its dots. Each dot made, it should be remembered, is the result of a succession of strokes inscribed by the vibrating recorder 10 times in the second. But as the movement of the plate is slow, these successive strokes more or less superimposed make but a single large dot. Ten seconds again elapse before the next pressure of the key brings about another large dot, and in that
