14 ROBERT M. YERKES 



The three methods differ so much in value, or rather in their 

 special kinds of serviceableness, that they may not be directly 

 compared. All are useful in the study of ideational and other 

 highly adaptive forms of behavior, but each has certain peculiar 

 advantages, whatever the ideational problem in question. For 

 this reason, chiefly, it has seemed to the writer important, as 

 a matter of economy and efficiency of research, to devise a form 

 of apparatus which should enable the investigator to use at will 

 any one of the three methods. 



It has not been especially difficult to plan such an apparatus, 

 for the writer has had opportunity to use, and to see used, each 

 method, and has had full advantage of the published results 

 of Hamilton and Hunter, as well as personal contact with them. 

 It may be convenient to refer to the device now to be described 

 as the convertible ideational or reactive tendency apparatus. 

 It is called an ideation apparatus, not because its usefulness is 

 limited to the study of the function of the idea, but because it 

 was originally devised as a means of discovering those types of 

 behavior which are either definitely ideational or closely akii 

 thereto. Objectivists who are offended by the term ideation 

 may substitute reactive tendency or some other equivalent term. 



The three methods for which this apparatus may be employed 

 are presented, not as the final word in the study of complex 

 behavior, but rather as the first words concerning a new ap- 

 proach to genetic problems. 



DESCRIPTION OF APPARATUS 



The apparatus consists (1) of twelve identical boxes, each 

 with an entrance door and an exit door that can be raised or 

 lowered by the experimenter from his observation stand; (2) a 

 reaction chamber in which the subject responds, as may be, to 

 a definite experimental situation, which may be described as 

 a " setting " of the various mechanisms (this setting differs for 

 the three methods, and also from trial to trial in the Yerkes' 

 method) ; (3) a release box in which the subject is confined be- 

 tween trials and from which it is admitted, at the proper mo- 

 ment, to the reaction chamber; (4) alleys for the passage of the 

 subject from the rear of the reaction mechanisms or boxes to 

 the release box; (5) twelve reward mechanisms, one for each box; 

 (6) a keyboard, or series of levers, (depending upon the size of 



