below 6030, will no doubt be attended with the death of
the organ. Fatigue is also found to lower the death—point.
THE THERMO-CRESCENT CURVE.
Experiment 66.——I was next desirous of devising a method by which an automatic and continuous record of the plant should enable us to obtain a curve, which would give the rate of growth at any temperature, from the arrested growth at the minimum to a temperature as high as 40°C. In order to eliminate the elements of Spontaneous variation, the entire record had to be completed within a reasonable length of time, say about 18 minutes for a rise of. as many degrees in temp Irature. This gives a rate of rise of 1°C. for one minute. Separate experiments showed that at this rate of continuous rise of. temperature there is practically no lag in the temperature assumed by thin specimens of plants. For observation during a limited range I use the slower rate of rise at 1°C per two minutes. But the result obtained by slower rise was found not to differ from that obtained with one degree rise per minute. The curve of growth is taken on a moving plate, which travels 5mm. per minute. Successive dots are made by the recording lever at intervals of a minute during which the rise of temperature is 1°C. A Thermo-crescent Curve is thus obtained, the ordinate of which represents increment of growth, and the abscissa, the time. As the temperature is made to rise one degree per minute, the abscissa also represents rise of temperature (Fig. 66). The vertical distance between two successive dots thus gives increment of growth in one minute for 1 degree rise of temperature from '1‘ to T'. If t repre- sents this length, t the interval of time (here 60 see), and m the magnifying power of the recorder, then the rate of
