346 H. M. JOHNSON 



remained as controllable as it was practicable to make them under 

 the circumstances. 



The levers by which the stimuli were interchanged were 

 attached by strings running through a pulley-system to the handle 

 of the double-throw switch which controlled the charge on the 

 respective punishment grills. Thus, by merely turning the switch, 

 the respective positions of the positive and negative fields were 

 interchanged. It was thus made impossible for the animal to 

 receive a shock under the positive test-object, or to avoid it if 

 he stepped into the alley under the negative test-object. A 

 further slight change in the apparatus would have made it safe 

 in the hands of an inexperienced student. That is, installing a 

 wiring system by which the entrance-doors to the proper food- 

 compartment would open electro-magnetically when the animal 

 entered the proper alley. 



The order of presenting the standard field to the right or left 

 of the variable field was determined before each series by the 

 use of a well shuffled pack of twenty cards. The only exception to 

 the rule of strict irregularity of presentation was that the two fields 

 should not occupy the same position with respect to each other 

 through more than three successive trials during the learning 

 period, and not more than four successive trials during the ad- 

 vanced stages of discrimination. The animals practically never 

 received the stimuli in the same serial order. Of course, in twenty 

 trials the standard field would appear ten times to the right and 

 ten times to the left of the variable field. 



The following animals were used in the work herein reported: 



Dog 1 : a pure bred male English bull terrier ; six months old 

 at the beginning of the experimentation. 



Monkey 1 : an adult male Cebus, variety not established ; 

 purchased from a dealer. He died before learning the problem, 

 but certain features of his behavior are noteworthy. 



Monkey 2 : a male Cebus, nearly full grown, but not sexually 

 mature; variety not established; purchased from a dealer. 



Chicks 1 and 2 : both pure bred Indian gamecocks, six months 

 old at the beginning of the experimentation; bred by Robert 

 Stedman, Beachland, Nottingham, Cleveland, Ohio. 



All of the animals were easily tamed, and none except Chick 1 

 showed any great excitability in the experiment-box. All were 

 kept free in the same large, well-lighted room, steam-heated and 



