BEHAVIOR OF VERTEBRATES 415 



the total number of mechanisms presented may vary from three 

 to nine. Only by perceiving and appropriately responding to 

 the relation which the experimenter designates as middleness can 

 the subject solve the problem." The apparatus used with the 

 pigs consisted essentially of nine stalls opening from a central 

 yard by sliding doors which were controlled by an unseen opera- 

 tor. Each stall gave access, at the other end, through a sliding 

 door controlled as before to a food trough. There was an arrange- 

 ment by which the animal after choice returned again by a side 

 way to the central court. The plan was to have the pigs learn 

 four problems: (1) to choose the first door on the right; (2) to 

 choose the second door on the left; (3) to take alternately the 

 first door on the right and the first door on the left; (4) to take 

 the middle door. In all cases it was planned to have other doors, 

 from three to nine in number open while choice was being made. 

 The two pigs which served in this experiment solved the first 

 problem in 65 and 85 trials respectively. The second problem 

 was solved more slowly because the habits formed in the pre- 

 vious experiment had to be broken up. It took 396 and 516 

 trials each. The third problem was solved also in 490 and 431 

 trials each. Problem four proved too difficult for the pigs. 

 They could succeed when the open doors were few in number 

 but when seven to nine doors were open they were confused. 

 Yerkes thinks his results indicate an approach to free ideas if 

 not the actual attainment of simple ideational behavior. Com- 

 plete numerical tables are given. Would the practice were more 

 prevalent. 



Four articles by Vincent (31), (32), (33), (34), have to do with 

 learning in mazes in which the sensory conditions were modified. 

 In the first series of experiments the true path and the cul de sacs 

 were made to differ in brightness, i. e., the true path was made 

 as white and the cul de sacs as black as possible and vice versa. 

 In the second series olfactory trails were laid in the first case in 

 the true path in the second case in the cul de sacs. In the third 

 series a maze was constructed having elevated pathways with 

 no restraining, outside walls — a condition which forced the use 

 of nose, feet and vibrissae and thus made the tactual control the 

 important one. The fourth paper gives a comparative estimate 

 of the relative effectiveness of these different senses as modes of 

 control and includes also a brief discussion of the number and 



