42 FREDERICK S. BREED 



The adoption of the discrimination method followed directly 

 from the results of Yerkes' work with the dancing mouse. I 

 have appropriated his method, my contribution being merely 

 an adaptation of the method to the study of chicks. So much 

 for the problem in a general way. 



II. Apparatus 



The apparatus used was a fan-shaped reaction box built of 

 wood and painted brown. Fig. 8 shows the ground plan, fig. 9 

 the perspective. The method of construction was as follows. 

 On a base 88 by 95.5 cm. a convenient point near one edge was 

 selected from which as a center, with radii of 38.64 cm. and 

 69 cm., respectively, arcs of circles were described. This center 

 later became the point from which the chick started at each 

 trial, and around which the entrance box was built. The nearer 

 arc marked the location of the cards first approached by the 

 chick; the more distant one outHned a boundary of the apparatus. 

 The box was made with four compartments, but only the two 

 inner ones were used in these experiments. Removable parti- 

 tions (J, fig. 8) made possible this use of only a limited por- 

 tion of the box. The height of the apparatus was 20 cm., inside 

 measurement. The covering consisted of two hfting doors, one 

 over the entrance box A and the other over the rest of the appa- 

 ratus. These were made of light wooden frames and wire netting, 

 and were so arranged that they lifted on their hinges directly 

 away from the experimenter as he stood at the entrance box. 

 The mesh of the netting was 3 cm. — as large as conditions would 

 permit, so as to interfere with the light as little as possible. 

 The exits H were closed with removable galvanized sheet iron 

 sliding doors that fitted from above in vertical grooves. The 

 end of the box about G was closed with wire netting of .86 cm. 

 mesh. Grooves at this end also provided for wooden screens 

 inside the wire of sufficient height to conceal the flock of chicks 

 in a cage just beyond and below the experiment box from the 

 chick reacting. Two sets of cards were used, one at D and 

 another at F. The card-carriers at F projected 2.55 cm. from 

 the walls, making it impossible for the exit H to be seen from 

 the chamber C. 



Following the ground plan in fig. 8 and the path of a chick 

 through the apparatus from the entrance box A, the remaining 



