38 LOMBAKD :
Study of the Chart. — At a glance one sees that, at whatever time the examination was made, the extent of the knee-jerk varied greatly in suc- ceeding experiments. He also notices that the average of the experiments made when the subject was just out of bed, and not thoroughly roused, was low ; that in the examination made fif- teen minutes later, after the bath had been taken, it was higher ; and that an hour later, immediately after breakfast, it was still higher. From this time on, however, the knee-jerk declined, being considerably lower before lunch, and though slightly higher just after lunch, being very much lower just before dinner. After dinner it recovered somewhat, but only to fall again, if but slightly, and at bed time it was very much less than it was just after breakfast, and even less than it was just after the bath taken before breakfast.
To judge from this one day, then, there is a great difference in the extent of the knee-jerk, even in succeeding experiments, and a still greater differ- ence between experiments taken at different times in the day, the knee-jerk being greatest immediately after breakfast, and, in spite of the fact that each meal tends to increase it, being much lower at bed time.
The discussion of the reinforcements which were observed during this day will be deferred until later in the paper.
Diurnal Variation of the Knee-Jerk. — Is the diurnal variation of the knee-jerk seen on April 6th a constant phenomenon ? This question is answered by the following table of the averages, compiled from all the experiments which were taken in this series •
