HABIT FORMATION IN THE DOG 47 



in my results ; that the discrimination shown by the animals 

 in the preliminary experiment might not have been lost had 

 the experimenter gradually effaced himself. In other words: 

 It may be too much to expect of an animal that he form the 

 association between a sound and food in a certain box without 

 some helps in the beginning at least ; although he may be able 

 to form the association with help and yet retain it after the 

 helps have been eliminated. 



It is manifest that the justice of this criticism cannot be settled 

 a priori. Accordingly it seemed well to determine whether, and 

 if so, how readily an animal thrown on his own resources can 

 to discriminate between auditory stimuli alone. As pure 

 tones were clearly out of the question for the time being 

 at least, it was decided to use non-musical noises, which might 

 present to the dog differences in other characteristics than 

 that of pitch. 



These experiments were conducted with the same animals and 

 in the same stimulus-cage used in experiments 6 and 7. It had 

 been remo\-ed meanwhile to a building at Homewood, which 

 was more quiet than the one in the business section of the 

 city could be. The experimenter was never in the room with 

 the animals when the stimulus was given, but gave it from the 

 room adjoining, where he could not be seen. After the animal 

 had become started on the problem the experimenter did not 

 even watch her while in the act of making her choice, l)ut waited 

 until after he had heard her cross the punishment grill. Thus 

 the question of "unconscious helps" is eliminated. The order 

 of presenting the -stimulus was predetermined by the use of a 

 well shuffled pack of cards. 



The stimulus-noises were sounded on two ordinary electric 

 buzzers. In the first experiment, designated as problem 8, 

 one buzzer was placed over each of the two doors X and X', 

 but were merely hung over the edge of the cage by their flexible 

 wire, without touching the cage itself. This was to lessen the 

 probability of the animals reacting to vibration of the cage. 

 The buzzers were actuated by current from two ordinary dry 

 cells connected in series ; the current being made by the operator 

 pressing a simple contact key. Being placed over the entrance- 

 doors to the food-compartments, they could be easily and in- 



