NORMAL KNEE-JERK. 67
The first curve of the chart shows the diurnal variations of the knee-jerk as determined by seven examinations. The average of all the experiments made at each examination is represented by a dot, placed at the proper height with reference to the millimetre scale, at the left side of the chart, and under the day and hour at which the examination was made. The lines connecting the dots enable the eye to readily follow the variations of the knee-jerk during each day. The larger dots represent examinations which were made directly after a meal.
On the base line of this knee-jerk chart is recorded the direction of the wind in the morning, afternoon and evening of each day.
Below are arranged, in order, the curves which show the vari- ations of the barometer, the thermometer, the electric potential of the air, and the relative humidity of the atmosphere. Each dot in each of these curves represents a separate observation, and is
f)laced at a height corresponding to the scale at the side of the after, and under the hour at which the observation was made.
The heavy cross lines show the average knee-jerk for each day, and the mean of the barometer, of the thermometer, and of the electric potential of the air for April, as determined by the obser- vations of a number of years.
Study of Plate II. — It seems to the writer that this chart is of great value from the negative evi- dence which it offers. It shows that a change in the direction of the wind, a change in the electric potential of the atmosphere, and slight changes in the relative humidity of the air, are without visible influence upon the knee-jerk, and, presumably, upon the central nervous system. It also calls the atten- tion to the fact that the variations of the barometer and the thermometer, though of the greatest import- ance in determining the height of the daily average of the knee-jerk, are secondary to hunger and fa- tigue in their effect upon its hourly variations.
. Summary of Results of Experiments of Series I. — The extent of the normal knee-jerk is continually undergoing change. So great are the variations, even when the subject is at rest, that a correct idea of the activity of the process can be gained only by averaging the results of twenty or more experiments. The average knee-jerk varies in amount at different
