196 H. M. JOHNSON AND FRANKLIN C. PASCHAL 



laid in the rack, face down. The subject is warned by the 

 spoken words, "Ready! Now!" and the face of the card is exposed. 

 A stop-watch is started when the subject begins to print the 

 first substituted letter, and is stopped when the subject taps on 

 the desk to indicate that the last substituted letter has been 

 printed. To guard against the appearance of certain spurious 

 effects the several sets of test material are given in an irregular 

 sequence, which is different for most of the subjects. 



RESULTS OF THE TEST 



Unsuccessful tests were run on nine subjects before the method 

 was sufficiently standardized for use. Due to failure of proper 

 control of the apparatus, three other experiments were con- 

 sidered unsatisfactory. These records are not included among 

 the results presented below. 



The abscissal values of the curves of psychological performance 

 represent the middle point of the temporal period of work on 

 the given set of material. At first, attempts were made to 

 maintain a constant temporal interval, or "headway," between 

 the presentations of the several sets, but the fluctuations of 

 performance rendered this impracticable. Later, an attempt 

 was made to maintain a constant period of rest between periods 

 of work, and this attempt was more nearly though not completely 

 successful, the principal source of variability being the necessary 

 activity of the clinician. 



The scale of ordinates was arbitrarily chosen from the stand- 

 point of convenience in plotting. Each ordinate represents five 

 times the average number of letters correctly substituted per 

 minute. 



DISCUSSION OF GRAPHS 



Most of the records show an important degree of improvement 

 with practice, as compared with the initial performance. The 

 rate of improvement varies widely with the different subjects, 

 and depends in large degree on their relative composure, dis- 

 tractibility, and interest. When the mouthpiece is first inserted 



