22 HARRY MILES JOHNSON 



Four problems were given: 



1. Discrimination between the two tones sounded on tuning 

 forks struck by hand. 



2. Discrimination between the two tones sounded on the 

 blown variators. 



3. Discrimination between the two tones sounded on forks 

 and on large and small variators indifferently. 



4. Discrimination between chords containing one and the 

 other stimulus-tones, respectively. 



The results are summarized in the following table : 



Problem 



The tables showing the daily percentage of error are found 

 on pp. 2 4ff. The longer learning-time of Dog 2 in experiment 

 3 was due to disturbance by the falling of a piece of apparatus 

 duing an experiment. She refused to work for nine days and 

 then resumed responses of any kind only after a great deal of 

 coaxing and petting. 



Care was taken to sound the tones with varying degrees of 

 intensity. The tones of the variators can be made faint or loud 

 at will by increasing or lessening the diameter of the opening 

 of the air- valve leading to each pipe. Such a change produces 

 also a change of pitch. The latter must be corrected by re- 

 tuning to the fork. Further the same stimulus-tones can be 

 blown on any one of two or three variators. By this means 

 tones of the same fundamental pitch, but of quite different 

 timbre, were given. It was thought that this might prevent 

 the animal from reacting to constant differences of intensity 

 and timbre. The relative position of the different resonators 

 w^as also frequently interchanged, to prevent possible localiza- 

 tion from becoming a factor. 



