81 
1911—12.] A Method of Measuring Mental Processes. 
is remarkable that in only one minute, viz. the eleventh, from the fourth 
to the fourteenth was the average working capacity attained. To take 
the middle five minutes of such a curve — as Hylan (4) suggested — would 
convey an erroneous impression of the working capacity. The constancy 
in the amount of work done from the fifth to the eleventh is no greater 
than in any other five-minute period between the fourth and the fifteenth 
minutes. Here, indeed, it is less ; for the higher of the crests occurred in 
the eleventh minute and the deepest of the “ troughs ” in the sixth minute. 
Further, not only is it not more constant, but as has just been shown, every 
value recorded in the period is below the average. 
I may not generalise upon this fact, as I investigated only seven normal 
persons, but the result at least clearly shows that Hylan’s suggestion — that 
the middle five minutes represents best the working capacity — is not accept- 
able. True, the initial spurt and the terminal spurt are both avoided, 
but the capacity for spurting is part of the capacity for performing the 
task, that is to say, is part of the mental working capacity. Indeed, I 
have just shown that the mental capacity was usually at its maximum in 
the initial spurt ; and that the next greatest output occurred at the close 
of the test time. In a daily curve, the constancy in Hylan’s middle 
period — from the fifth to the eleventh minute — was often remarkable 
(see curve of Mr A’s second test), but sometimes the rate during this time 
was not more uniform or was even less uniform than during the rest of 
the test (see curve of Mr A’s first test). The greater tendency to spurts 
and to lapses of attention which Kraepelin and others have attributed 
to the beginning and end of the task was not very apparent. A priori 
the period of greatest uniformity should fall, as they allege, in the mid 
part of the task, after the task has been properly “ got under weigh,” and 
before it has lasted long enough to induce boredom or fatigue. This 
middle period is really a method of estimating the results akin to that so 
commonly used in the investigations of the Kraepelin school — the so-called 
“ Wahrscheinliche Mittel ” estimation. 
When, however, the importance of the factors conducing to variation 
from minute to minute, in any one day, is minimised (as by the method I 
have used of taking the average of each minute over a number of successive 
days), the results are apparently not adequately represented by the average 
of the middle five-minute period. 
Another curious point to be observed in these curves — especially in 
the average curve of all the seven persons — is the absence of any evidence 
of the so-called “ Anregung ” (5), or incitation of Amberg. 
These seven normal people, so far as the mental work of a single fifteen - 
VOL. xxxii. 6 
