236 THE INTELLIGENCE OF MAMMALS 
led to adopt an extreme position which has not been justi- 
fied by future experiments and which was not consistent 
with many of his own results. Though extreme, the reac- 
tion against the method of writers who based their deduc- 
tions concerning animal intelligence on stories and casual 
observations was none the less wholesome, as it succeeded 
in stimulating the study of animal intelligence by more 
careful and critical methods than those formerly employed. 
mL 

Fia. 15.—Puzzle box used in the experiments of Thorndike on cats. 
(After Thorndike.) 
In most of Thorndike’s experiments boxes were employed 
from which animals could escape by raising a lever, pulling 
a cord, or by some such simple device. In some cases vari- 
ous combinations of these devices were used. A hungry 
cat or dog was confined in the box and food was placed on 
the outside so that it could be seen. The animal in its 
efforts to get out and obtain the food would usually begin 
by biting and clawing the bars of the box. Sooner or later 
a lucky movement would raise a lever or pull a cord so that 
the door of the box would open and allow the animal to get 
the food. After this the animal would be put into the box 
