61 LOMBARD :
the 4th, and the barometer fell only a little ; moreover, the barometer rose from the 2d to the 3d, while the knee-jerk was stationary. Again, one sees that the knee-jerk fell off from the 7th to the 8th, although the barometer was still rising. These differences can only be understood by simul- taneously comparing the three curves, and remem- bering that a rise of temperature or a fall of the barometer tends to depress the knee-jerk, while a fall of temperature or a rise of the barometer tends to elevate the knee-jerk curve.
From the 1st to the 2d the temperature rose and the barometer fell, and both of these influences acted to lessen the movement ; from the 2d to the 3d the temperature continued to rise and the barometer rose, and the counteracting influences caused the knee-jerk to remain stationary ; from the 3d to the 4th the temperature rose and the barometer fell, and the knee-jerk curve consequently fell very low ; from the 4th to the 5th the tem- perature fell markedly and the barometer rose a little, and the knee-jerk began to recover; from the 5th to the 7th the barometer rose markedly, and the slight rise of temperature which occurred, not being sufficient to counteract its influence, the knee-jerk curve rose ; from the 7th to the 8th the continually in- creasing temperature began to make itself felt, so that the process became less active, in spite of the fact that the barometer continued to rise ; from the 8th to the 11th both the temperature rose and the barometer fell, so that the knee-jerk was greatly depressed; from the 11th to the 13th the temper- ature fell and the barometer rose, and both influ- ences assisted to restore the knee-jerk ; from the
