134 The Animal Mind 



tus (the Stern tone variator) ; and to discriminate between 

 a chord containing one of these tones and a chord containing 

 the other tone. This was when the experimenter remained 

 in the room. Experiments by a much more accurate 

 apparatus, where electrically driven forks were sounded 

 from another room, while the experimenter observed the dog 

 also from an adjoining room, the tones being sounded with 

 exactly the same duration, gave precisely opposite results : 

 the dogs could not discriminate. (The test of discrimina- 

 tion was learning to turn to the right when one tone was 

 sounded and to the left when the other tone was sounded.) 

 Moreover, they could not even learn to take one turning 

 when a tone was given and the other when no tone at all 

 was given; apparently if they heard the tone at all they 

 paid no attention to it. On the other hand, the noises of 

 two electric buzzers, of different intensity, pitch, and timbre, 

 were readily discriminated and localized by the dogs. 



The same objection, that secondary clews derived from 

 the presence of the operator may account for the seeming 

 discrimination of sounds, applies to the work of Shepherd 

 on cats (677, 678) and raccoons (676). The apparent fact 

 that certain mammals are deaf to tones, while perfectly 

 able to hear noises may, as Johnson suggests, be connected 

 with the fact that even human beings cannot localize pure 

 tones with any accuracy : a sound stimulus, to have 

 practical significance, must be capable of being localized. 

 Zeliony (839) trained a cat to come from one room into 

 another when a C whistle was blown, and thought he had 

 evidence of the cat's ability to distinguish the sound of 

 this whistle from that of others differing not more than 

 a half-tone ; but the difference reacted to may have been 

 in the accompanying noises. Hahn (282) finds the bat 

 very sensitive to high-pitched sounds, but not to low ones. 



