NORMAL KNEE-JERK. 63
Farenheit thermometer scale from 30°-70,° a scale of milli- metres, to show the extent of the knee-jerk, and the barometer scale from 29-31 inches. The curve opposite the thermometer scale shows the variations of the temperature, each of the clots connected by the lines giving the mean temperature for the corres- ponding day. The curve opposite the millimetre scale shows the variations of the knee-jerk, each dot representing the average of all the experiments taken on the corresponding day. Similarly the curve opposite the barometer scale records the variations of the barometer, each dot giving the mean of the barometer for the cor- responding day.
Study of the Diagram. — The correspondence be- tween the temperature and the knee-jerk curves is not very accurate, but one sees that on the 4th and 11th, when the temperature was high, the knee-jerk was low, while on the 1st, 7th and 14th, when the thermometer was much lower, the knee-jerk was considerably higher. In general, then, as the tem- perature rises, the knee-jerk becomes less, and as the temperature falls, the knee-jerk becomes larger.
If now one compares the knee-jerk and barometer curves, he finds the agreement to be much closer. The barometer fell, roughly speaking, from the 1st to the 4th, so did the knee-jerk ; the barometer rose from the 4th to the 8th, so did the knee-jerk ; the barometer fell from the 8th to the 11th, so did the knee-jerk ; the barometer rose from the 11th to the 13th, so did the knee-jerk ; and finally, the barom- eter fell from the 13th to the 14th, and so did the knee-jerk. In general, then, it may be said that as the barometer rises and falls the knee-jerk rises and falls.
A more careful examination, however, shows that though this general correspondence existed, the two did not agree in the extent of their variations, nor did they vary in just the same way from day to day. Thus the knee-jerk fell markedly from the 1st to
