212 TRANSACTIONS OF THE 
not be made until a recess period is allowed. As the learn- 
ing continues, however, the length of the time interval re- 
quired for control becomes less and less. A stage is ulti- 
mately reached where further improvement will be a gain in 
speed, not accuracy, and this improvement is dependent upon 
series, not recess, practice. A complete learning curve is de- 
pendent upon recess period practice at first and mixed series 
and recess practice later. 
The length of the recess interval necessary for the 
attainment of the highest efficiency in accuracy varies 
enormously from subject to subject. Some subjects require 
a period five times as long as do others. For most subjects 
maximal efficiency in accuracy is secured during the early 
stages of the learning by one practice per day. 
An average group of subjects may readily be divided in- 
to those requiring long recess periods for the attainment 
of accuracy and those who can secure a high degree of ac- 
curacy with a short recess. If twenty-four hour recess 
intervals precede all practices, an average group of college 
students will differ slightly in the attainment of accuracy. 
Now if the interval preceding practice is reduced to ten 
minutes, there will be wide variations as to accuracy at- 
tained, the highest often being three to four times as high 
as the lowest. From this it is seen that individuals differ 
not so much in their ability to attain accuracy as in the 
temporal conditions underlying its attainment. 
Those subjects who show the poorest control are (1) 
those who speed up most during series practice; (2) those 
in whom this speed is little retarded by the insertion of 
short recess intervals; (3) those who are popularly called 
quick, snappy, and more accurately, nervous; (4) those most 
affected by drug stimulants such as strychnine or caffein ; 
(5) those most affected by the social stimulus or the pres- 
ence of others in the laboratory; (6) those in whom the 
efficiency from time to time is extremely variable. 
Those subjects who show the best control are (1) those 
whom the world would call somewhat slow, the phlegmatic 
type in general; (2) those who do not lose their accuracy 
in a marked way by light doses of strychnine or caffein; 
