294 



WALTER S. HUNTER 



Onlv after the tenth trial was the normal average error record 

 of .2 regained. It then remained constant until the end. The 

 average time record became normal after the fourteenth trial. 

 Those pigeons that became confused hesitated, upon entering the 

 maze, and invariably made an error at cul-de-sac no. i. The 

 error at no. 2 was also made as were those of turning back over 

 the true pathway. An error at no. 6 was made twice — once 

 by pigeon no. 7 on the eighth trial, again by pigeon no. 8 on the 

 seventh trial. 



TABLE IV 

 Maze Rotated 90° to the Left 



If the conditions of the experiment are kept well in mind, 

 the following explanation may not seem amiss : Assuming for 

 the time that visual sensations from without the maze are in- 

 fluential in guiding the pigeon in its reactions, the tendency 

 will be to head directly north (vide supra, fig. 2) when the maze 

 is rotated. But this is sure to result in the errors, which were 

 actually made. After the second alley is passed, in the rotated 

 maze, theoretically, the tendency will be toward the west and 

 the error at no. 6 will be avoided. The results show that this 



I 



