75 
1911-12.] A Method of Measuring Mental Processes. 
disease or group of diseases described as maniac depressive insanity. I 
have taken this term to include a folie circulaire,” and cases of periodic 
recurrent mania or melancholia. Most of my cases had been under my 
personal observation for several years, some had been under supervision 
in this Institution for decades. I have carefully avoided all cases in which 
I could not exclude the diagnosis of dementia prsecox, and also all cases 
of a first melancholic or maniacal attack. Many of the last were un- 
doubtedly cases of maniac depressive insanity, but I did not wish to 
introduce any element of uncertainty, so I thought it better to reject them, 
and to consider only such cases as left no possibility of doubt as regards 
the diagnosis. I had no difficulty in making the patients understand the 
method of continuous addition ; they all entered into the task with an 
astonishing degree of zest. They discussed their daily performances, and 
a keen spirit of rivalry prevailed. Their interest was easily sustained for 
several days, and some of them were considerably disappointed when their 
investigation was finished ; but as I early found that the majority of my 
cases showed a flagging interest after the fifth day, I had to limit the 
duration of the experiment to five days. As Maloney has suggested, I 
tested the patients simultaneously in groups, and took notes of their 
behaviour throughout each minute of the test, so that subsequently I 
could control and explain variations in their work. Never more than five 
patients were tested together. The experiment, therefore, consisted in 
allowing thirty cases of maniac depressive insanity to perform the 
“ Reckoning Test,” as above described, for fifteen minutes for five days. 
The experiments were conducted during the summer of 1911. Great care 
was taken that all the tests took place at exactly the same time on 
subsequent days, and the same uniform conditions of quiet and familiar 
surroundings were accurately observed during each test. As it was 
essential first to establish a normal with which to compare the results 
obtained in the insane, seven normal people were examined under the same 
conditions and for the same length of time. 
I shall first rapidly survey various factors which influence the test. In 
the fifteen-minute test of any one day an accidental lapse of attention is 
shown by a diminution in the number of units of mental work produced 
during the lapse ; similarly a spurt is shown by an increase. A lapse 
is usually followed by a spurt, and a spurt by a lapse. These variations in 
the output from minute to minute therefore tend largely to compensate the 
one for the other, so the total output is fairly characteristic of the mental 
capacity for the particular period selected. This period was kept constant. 
A person who began the test, for example, at 2.45 p.m. on one day was 
