HABIT FORMATION IN THE DOG 51 



These results show nothing new except for the record of the 

 first day. There may be some significance in a comparison of 

 the records of Dogs i and 2 with that of Dog 4. Both the older 

 dogs required a longer time to overcome the habit formed in 

 problem 8 although the time for learning problem 8 was as 

 short for Dog 2 as for the younger animal. 



At this point the experiments had to be abandoned. I think 

 it well, however, to outline the program which it was my inten- 

 tion to carry out had conditions permitted, as I believe it a 

 safe approach to an investigation of pitch-discrimination. First, 

 it was proposed to have the animals form the simple association 

 required under the conditions of problem 8. Then, as a new 

 problem, let the two buzzers be brought nearer and nearer 

 together, until they shall be together on the table near the end 

 of the introductory alley B. If the animals can still discrimi- 

 nate, the characteristic of localization is eliminated. Then let 

 a stimulus fork be substituted for one buzzer, and if discrimina- 

 tion can be established and maintained between the two sounds, 

 substitute the second stimulus-fork for the other buzzer. If 

 the animals learn to discriminate between the two tones, varia- 

 tions of intensity and other controls can follow. Discrimina- 

 tion between noises lies well within the dog's capacity. It may 

 yet be possible to obtain reliable evidence of discrimination on 

 the basis of pitch-difterence if the animal can be brought grad- 

 ually to the point where it will be affected by the stimulus. 



SUMMARY 



The foregoing experiments have failed in showing to what 

 extent the dog is sensitive to difference of pitch. They have 

 not established that he is sensitive to pitch-difference at all. 

 While thc\^ ha\'e not proved the contrary, they should have 

 shown that we are not safe in accepting the assertion of Kal- 

 ischer, Rothmann and Swift, that the dog has an exceedingly 

 fine absolute pitch sensitivity. These experiments have shown 

 that the evidence submitted to date does not warrant such a 

 conclusion. They have also shown some of the difficulties, in 

 the way of making satisfactory tests on audition in animals, 

 and should have demonstrated what is and what is not a reli- 

 able method of investigation in this field. It should also be 



