MAZE STUDIES WITH WHITE RAT 287 



left the true pathway, they invariably ran towards the experi- 

 menter. This positive reaction can not be considered a direc- 

 tive habit acquired in learning the maze because the conformation 

 of the maze at the old position was such as to prevent it. The 

 positive reaction can better be regarded as a feeding habit de- 

 veloped in the living cage and on the feeding table. The ex- 

 perimenter thus attracted the animal's attention because of the 

 novelty of the position and stimulated an old habit acquired 

 while the rat was being handled and fed. The arousal of this 

 habit naturally disrupted the normal functioning of the maze 

 act. In several experiments such as increasing and decreasing 

 the illumination, rotating the maze in darkened and lighted 

 environments, and rotating a heterogeneous environment, the 

 following behavior was frequently noted: Animals suffered a 

 pronounced disturbance at those points where the illumination 

 had been greatly increased. I have frequently seen animals run 

 the maze without error up to a point where an alley, customarily 

 darkened, was flooded with a beam of strong daylight. Here 

 the rat stopped suddenly, exhibited strong signs of nervousness 

 and timidity with frequent retracing in search of another path. 

 Decreasing the illumination in any part of the ma7e seemed to 

 be without effect, but a pronounced increase was effective. 

 These facts indicate that the alterations served as distractions. 

 The distractive theory is further supported by the irregular and 

 occasional character of the disturbances. This feature of the 

 results was summarized in the first paper. It refers to such 

 facts that many trials are without error, that rats are immune 

 in one experiment and susceptible in another, and that the 

 number of errors made in various trials is extremely variable. 

 If the rats are relying upon the objective data to guide their 

 conduct in the maze, it would seem that any rat should be 

 disturbed in every trial until complete adaptation is secured. 

 The fact that the disturbances occur in a perfectly haphazard 

 and accidental manner is readily interpreted on the basis of the 

 distractive theory. The disturbance is present only when the 

 alterations attract the attention of the rat, and this result 

 is largely a matter of chance. Conclusive proof of the distrac- 

 tive function is obtained from the comparative records on cover- 

 ing and uncovering the maze. Rats learned the uncovered 

 maze, — a maze with a well lighted and heterogeneous optical 



