PSYCHOLOGY: R. M. YERKES 
631 
betweenness, and would be, in fact, identical with one of the sets of 
independent postulates for betweenness obtained in a forthcoming 
paper by E. V. Huntington and J. R. Kline. The transition from the 
theory of cyclic order to the theory of betweenness may thus be made 
by merely interchanging two letters in the first postulate; postulates 
II-VI are true in both theories. 
A NEW METHOD OF STUDYING IDEATIONAL AND ALLIED 
FORMS OF BEHAVIOR IN MAN AND OTHER ANIMALS^ 
By Robert M. Yerkes 
PSYCHOLOGICAL LABORATORY. HARVARD UNIVERSITY 
Received by the Academy. October 20. 1916 
Despite widespread interest in the evolution of reasoning, the com- 
parative study of ideational behavior has been neglected. Only a few 
methods of research have been devised, and these have seen scant 
service. 
Thorndike^ is responsible for the problem or puzzle-box method 
(used by him in the study of cats, dogs, and monkeys); Hamilton,^ for 
the method of quadruple choices (by which he has studied cats, dogs, 
horses, monkeys, rats, gophers, and men); Hunter,^ for the method of 
delayed reaction (applied by him to rats, raccoons, dogs, and children). 
I have perfected and applied a new method — ^that of multiple choices 
— for the detection of reactive tendencies and the study of their role 
in the attempted solution of certain types of problem. The method in- 
volves the presentation to the subject of a problem or series of problems 
whose rapid and complete solution depends upon ideational processes. 
The apparatus consists of twelve, or, in some forms, nine identical 
reaction-mechanisms, of which any number may be used for a given 
experimental observation. In the type of apparatus originally used 
for human subjects, these mechanisms are simple keys; in that which 
has been used for lower animals, they are boxes arranged side by side, 
each with an entrance door at one end and an exit door at the other, 
which may be raised or lowered at need by the experimenter through 
the use of a system of weighted cords. Under the exit door of each box 
is a receptacle in which some form of reward for correct reaction may 
be concealed until the door of the appropriate box is raised by the 
experimenter. 
It is the task of the subject to select from any group of these boxes 
whose entrance doors are raised that one in which the reward (food, for 
example) is to be presented. The experimenter in advance defines the 
