HABIT FORMATION IN THE DOG 57 



The following method was employed in training: A room, 

 the floor-plan of which is shown in Fig. 4, was partitioned off 

 from a larger room, from which the observations were made 

 The problem-box F was placed in the smaller room, sixteen 

 feet from the wooden door W.D. in the diagram, and about 

 four feet directly opposite the glass door G.D. where the operator 

 stood. Food was placed in the problem-box, the door of which 

 was left open for the first week. The animal which was left to 

 run free in the larger room, was admitted through the wooden 

 door W.D. to the smaller room, where it had already become 

 accustomed to being fed. The wooden door opened by the 

 operator pulling a rope running from it to the glass door, and 

 was closed by a w^eight and pulley system after the animal 

 had passed through. After the animal had obtained the food 

 in the problem-box, it was released from the smaller room into 

 the larger through the glass door G.D. Food was again placed 

 in the problem-box and the animal readmitted through the 

 wooden door as before. Every animal soon accjuired a fixed 

 path to and from the pioblem-box, which path is indicated 

 by the dotted line and arrows in Fig. 4. By this method of 

 training previous to the setting of the problem the animal's 

 pioblem was simplified. There was no apparent disturbance 

 from the operator's movements: as soon as the animal was 

 released through the glass door G.D. it immediately ran to 

 the wooden door and awaited readmission to the animal room; 

 it was out of sight of the operator and wasted no time in trying 

 to get food in other than the prescribed way. When the animal 

 was presented to the problem-box the operator was out of the 

 loom, and in the actual experimentation, out of sight ot the 

 animal, as the larger room was darkened in the earlier stages 

 of each new problem. 



This general mode of procedure — the food-motive and the 

 removal of the operator from the animal's immediate presence — 

 has of course been used before by other experimenters and on 

 other animals. It differs from the method of Thorndike. 

 In his work the animals weie confined in the puzzle box, it 

 being assumed that the animal's desire to escape w'ould be a 

 satisfactory motive to bring about the opening of the box. 

 The animals were dropped into the box from the top. Some 

 of them being more nervous or timid than others, were more 



