COMPLEX NEURAL ARCS. 91 



in which the different individuals are comparable with each other and 

 relatively stable with respect to the effect of repetition. It may be 

 objected that actual articulation in reading is less common in adults 

 than silent reading. While that is undoubtedly true, it must be 

 remembered that the restraint of articulation is a refinement of develop- 

 ment. Reading was learned by actual articulation. And the passing 

 of silent reading into articulation occurs on the least provocation and 

 in the aggregate relatively often. In any event, the arousal of the 

 motor-acoustic residua is practically a universal if not a necessary 

 accompaniment to the process of understanding the printed word. The 

 nervous arcs which are involved in the articulation of familiar words 

 are relatively complex, but they are relatively constant and thoroughly 

 practiced. Of all the controlled associations, reading is probably the 

 most nearly immediate and universally practiced. Even in a mathe- 

 matical reaction the first associate which is aroused by digits, 7 times 

 8, for example, is probably not their multiple, but the auditory-motor 

 associate which is involved in reading them. 



Other things being equal, reading simple words appeared to satisfy 

 our criteria of a satisfactory experimental process better than adding 

 or any other mathematical task. Furthermore, the basal psychology of 

 the reading process has been subject to much more satisfactory analysis 

 than the mathematical processes. The adequate reaction to visual 

 verbal stimuli is about the best understood of all associations. It has 

 been experimental^ studied in connection with a considerable variety 

 of mental processes, both normal and abnormal. It has furnished 

 material for a large number of investigations in the psychology of 

 perception and attention. The conditions which determine satis- 

 factory experimentation are consequently thoroughly known and the 

 criteria of a satisfactory technique are entirely familiar to the experi- 

 mental psychologist. In all these respects, the inclusion of word-reac- 

 tion measurements in our series has been justified. In none of the 

 measurements, not even in the reflexes, have we found a lower percent- 

 age of variation within a series of observations. Notwithstanding 

 the differences between the words, the mean variation in a series of 

 24 is about 7 per cent of the reaction time. The same series of 24 four- 

 letter English words was reacted to in all our experiments by all our 

 subjects, regular and control subjects alike. 



EXPOSURE APPARATUS. 



The variety of possible instruments for giving visual stimuli under 

 experimental conditions is practically limitless. Equally limitless are 

 the experimental conditions which they may be required to satisfy. 

 There is probably no one best universal exposure apparatus. No such 

 instrument is equally good for all purposes. Any instrument is good if 

 it satisfies the specific experimental demands of the occasion. Between 



