374 MILDRED WEST LORING 



The motor is started by hand. It is equipped with a strobo- 

 scope if a higher rate is required, but for the low rate of 5 rota- 

 tions per second a slight twist of the axle which soon gets to be a 

 knack causes the motor to get in step with the fork. 



The experimenter and subject sat on opposite sides of a small 

 table with a large black curtain stretched between so that the 

 subject could see neither the experimenter nor any of the appa- 

 ratus. The experimenter sat with the master key and the chro- 

 noscope on his right and the fork on his left, each on separate 

 stands, so that neither could affect the voice keys, which are so 

 sensitive that the passing wagons or a moving chair in an ad- 

 jacent room stops the clock. With this arrangement too, the 

 experimenter could record results and regulate the fork contact 

 without moving from his chair. The subject was given his in- 

 structions in regard to the type of response required and then the 

 list of words. The experimenter signalled the subject before 

 each word by saying " ready." Since the motor was practically 

 noiseless, the only possible distracting noise was the low hum of 

 the fork, which might have been eliminated by enclosing it in a 

 padded box, or placing it in another room. This was a constant, 

 however, throughout the experiment and apparently was not 

 noticed to any appreciable extent except by one subject who 

 worked at night when the building was absolutely quiet and all 

 outside noises were at a minimum. 



EXPERIMENT I 



The general plan of the whole problem was to secure a large 

 number of stimulus words nouns, adjectives and verbs to be 

 given to a group of subjects, in order first of all that eliminations 

 might be made of unsuitable stimulus words on the basis of the 

 reactions to them. This was called experiment I. Final con- 

 clusions were not to be based on these results because the subjects 

 were few in number and also because the lists of stimulus words 

 contained many words unfitted for the association experiment. 

 The primary object then was to detect these unsuitable words 

 by means of the reaction time and the response word, and to omit 



