314 Jounuil of Comparative Neurology and Psychology. 



ot" the other trials, except that minor randoin mo\'ements have 

 been eliminated, and only the principal movements chosen for 

 emphasis. The guinea pig had not succeeded in getting out by 



Tc.xt-fif^iirc y. 



going in the direction of the arrows, yet this same path was 

 chosen the second time, it having been one of the major move- 

 ments which, as a group, were previoush' successful. 



Test IV. Does the odor of the previous path furnish the stimulus? 



Age, 3 days. There is a possibilit}- that the odor of the 

 path just taken might serve as immediate indication of the path 

 to be chosen again. Therefore the reaction would be mechan- 

 ical, i. e. , in terms of immediate stimulus and response, not of 

 recollection. 



In order to test this the odor of previous trips was elimi- 

 nated by thorough scalding of the wire cages and boxes to be 

 used. The wire cages were then set upon glass instead of 

 smoked paper. The apparatus was set up as in Text-fig. 2. 



Time for guinea pig to find his way out, .33 min. No 

 random movements. The apparatus was then set up as in Text- 

 fig 4- 



Time. .83 min., the path being the one chosen on the pre- 

 vious day, i. e., including the major random movements. 



Upon a second trial of the problem at this time (the wires. 



[22] 



