8 G. V. HAMILTON 



The first of any of the exit alleys entered by the subject is 

 allowed to give escape into the transfer box at its end. 



The monkey subject is tolled into the enclosure and confined 

 therein by the closure of the entrance alley door. He is "allowed 

 to use the first exit alley entered for escape into a court, where 

 he is at once given food. 



The human subject (only children were used in these studies) 

 is told that escape from the apparatus will be rewarded by a 

 toy credit, i.e., each escape counts one toward the number of 

 escapes necessary to secure a reward. Four toy credits secure 

 a reward on the day that is devoted to preliminary trials; after 

 that ten credits secure a reward. The human subject, like the 

 rodent and the monkey, is allowed to escape by the first exit 

 alley entered. 



Second Preliminary Trial. — The exit alley used during the 

 first preliminary trial will not afford escape, but any one of 

 the other three exit alleys may be used for this purpose. 



Third Preliminary Trial. — Either of the two exit alleys that 

 have not yet been used for escape will afford escape. 



Fourth Preliminary Trial. — -Only the one exit alley that has 

 not yet been tried will afford escape. 



These four preliminary trials, which are given to the sub- 

 ject in as rapid succession as is possible, are followed the next 

 day by the first ten of a series of 100 formal trials. Ten formal 

 trials are given daily for ten successive days. In describing the 

 method of conducting these trials and in subsequent discussions 

 of results I will use the following terms: 



Alley No. i — the exit alley at the subject's extreme left as he 

 faces the row of exit alleys from within the inclosure. 



Alley No. 2 — the second exit alley from the left. 



Alley No. 3 — the third exit alley from the left. 



Alley No. 4 — the fourth exit alley from the left. 



The Right Alley of a given trial is the only one that will afford 

 escape. 



The Impossible Alley of a given trial is the one that afforded 

 escape during the immediately preceding trial. There is 

 always one inferentially impossible alley for the subject 

 who learns that it is useless, during a given trial, to try 

 the right alley of the immediately preceding trial. 



