422 STELLA B. VINCENT 



always in the room, or that they had learned the order of presen- 

 tation which was not always recorded. For these reasons it was 

 determined to repeat the work under conditions which would 

 render such helps impossible. In the next experiment there 

 was an attempt to train two dogs to react to the same stim- 

 ulus tones; but in this case the forks were actuated in an ad- 

 joining room and the sound was conveyed by tubes to a point 

 just over the animal's cage. The experimenter, in another 

 room, released the cage door by means of a string. The dogs 

 were to respond to the "c" fork by mounting a chair on the 

 west side of the room and to the "e" fork by mounting one on 

 the east side. In case of a proper choice, food was dropped in 

 the chair from a chute above it. After 37 days when each dog 

 had had 505 trials, there was little evidence of discrimination 

 and the conclusion was forced that the first learning had prob- 

 ably been due to unconscious helps, to differences in the dura- 

 tion or intensity of the stimuli, or to some positive habits of 

 the animals. In the next experiment these factors were most 

 carefully watched. The details of the control are given in the 

 monograph. There was a well planned bit of apparatus used 

 which was so arranged that the animals were to go down an 

 alley, and then turn to the right or the left in response to a 

 signal which was not given until they reached the place of turn- 

 ing. It was hoped in this way to overcome the difficulties in- 

 volved in a delayed reaction. But here again after 92 days the 

 problem was not learned. The contention is faced that a pure 

 tone is too difficult of localization and is too far removed from 

 the tones to which the animal gives instinctive reaction to 

 serve as a useful stimulus for learning. In some later experi- 

 mentation the animals learned in a few days to localize sounds 

 given by buzzers but the experimentation failed to show that 

 the animals were at all sensitive to pitch differences. 



Birds. Lashley (32), from some brief work with an Ama- 

 zonian parrot, tells us that it imitated singing tones from vio- 

 lin, cello, piano and voice, and that 30 whistling tones were 

 imitated. The register for the singing tones was one octave 

 upward from c 256, for whistling tones two octaves upward 

 from f 384. The response in case of the singing tones was a 

 series of tones with many changes in pitch. The whistling tones 

 were a series of half tones beginning with f 384 and they gave 



