136 E. S. BOGARDUS AND F. G. HENKE 



of the maze thus operate disadvantageously in learning similar 

 mazes by tending to prevent the animals from picking up the 

 true path after emerging from the critical part. This disturb- 

 ance was the most pronounced and persistent in the case of 

 the fifth maze. 



After the animals once succeeded in picking up the old familiar 

 path, they almost invariably ran the rest of the maze without 

 error or hesitation. Evidently the presence of a common por- 

 tion constituting the last part of all the mazes is a highly ad- 

 vantageous feature. 



Since each maze differs from the preceding one in but one 

 respect, the question arises as to why the successive adjustments 

 should vary so enormously in difficulty as is evident in table III. 

 The determining conditions are probably very complex though 

 some of the factors were evident from the animals' behavior. 



1. Other factors being equal, that maze is the easiest in which 

 the critical position is placed nearest to the entrance. Since the 

 last common portion operates as an advantage, the longer this 

 part the easier should the maze become. In making the read- 

 justment, the animals run back and forth over the first common 

 portion. Evidently the chances of making the correct adjust- 

 ment is favored by a short runway. The shorter is this segment, 

 the more is their activity centered around the region of the 

 critical position. 



2. The adjustments which involve entrances into former cul 

 de sacs are more difficult than those which involve a short cir- 

 cuit. The truth of this proposition is evident in table III, and 

 the reasons therefor are obvious. An entrance into an alley 

 which has been effectively eliminated involves a greater viola- 

 tion of past habits than does a mere deviation from the accus- 

 tomed path. In the latter case there is also present the enticing 

 possibilities of a novel stimulus. 



3. The difficulty is increased in proportion to the distance 

 beyond the critical point at which the former path has been 

 blocked. In maze II the path was blocked at 6 while the open- 

 ing was at 5. On being stopped at 6 the rats explore around 

 in this vicinity before starting back at full speed and as a con- 

 sequence it was an easy matter to chance upon door 5. In maze 

 III the path was blocked at 3 while door 2 was open. On en- 

 countering the closed runway at 3, the animals investigate for 



