77 
1911-12.] A Method of Measuring Mental Processes. 
An accidental increase of rate in one was usually counterbalanced by a 
decrease in another. The average of all was probably a fair estimation of 
the normal mean rate of progress. I therefore took the average of the 
percentage increase in them on the second, third, fourth, and fifth days as 
standards by which to judge the abnormal. 
I obtained a standard of normal improvability for the five days by 
estimating the mean percentage gain between the first and the fifth days 
in the normal persons. 
As the moods from day to day might vary, and as “ habituation ” and 
exercise constantly tended to increase the working capacity, no one day 
could be taken as typical of the period. I carefully examined the third 
day * and found that it very closely resembled in many particulars the 
average of the five ; but in the plotting out of the typical work curve 
only the average of the total work of the corresponding minutes on all the 
five daysf could be taken as representative of that minute. For accidental 
spurts of one day were then neutralised or depreciated by lapses of 
attention of another, and only when in the corresponding minutes a pre- 
ponderating trend was evident in the majority of days could it stamp itself 
upon the curve of the mean work of the whole experiment. And as the 
average work curve of all the normal persons taken together further 
diminished the influence of such accidental variations, an accurate curve of 
the normal progress of mental work from minute to minute throughout 
the fifteen-minute interval was obtained. With this curve, the Normal. 
Mental Work Curve, I compared the curves of my insane cases. 
We shall first briefly consider the results yielded by the seven normal 
test persons. 
Total Output. 
The total number of units of mental work performed on the first day 
of the task varied greatly. The results were A, 610 ; B, 850 ; C, 694 ; 
D, 814 ; E, 718 ; F, 768 ; and I, 642. An important factor in this variation 
was undoubtedly previous familiarity with figures — the arithmetical training 
of the person investigated. The number of units of work done in fifteen 
minutes varied from 610 to 850. The average output for the first day was 
728. On the fifth day the corresponding results were 941, 1217, 1182, 
1124, 1154, 1086, and 1045 ; the highest was thus 1217 ; the lowest 941 ; 
the average 1107. The output on the fifth day was taken as equivalent to> 
the maximum working capacity. 
* Indicated on the charts by the black curve. 
t Indicated on the charts by the dotted line (when two curves are shown). 
