268 HARVEY CARR 



maze under both conditions before the rotation tests were given. 

 One set of four rats were tested for three positions on successive 

 days when the room was well illuminated. The tests were now 

 repeated for the darkened environment and these were followed 

 by a series with an illuminated maze. The average error records 

 for the three conditions respectively were 7.15, 1.90, and 3.20. 

 The final value for the illuminated environment is thus greater 

 than that previously secured for the darkened condition in spite 

 of the fact that animals tend to adapt to these rotary shifts 

 when repeated. With a second set of six animals, complete 

 adaptation was effected for three positions while the room was 

 darkened. The room was then illuminated, and the tests were 

 repeated. A disturbance was again evident. The disturbance 

 could hardly be due to the sudden introduction of the light, 

 as the maze had been learned under these conditions, and the 

 rats had been accustomed to run the maze in its normal position 

 while the room was illuminated. The results indicate that a 

 maze rotation in reference to a well illuminated environment 

 is more disturbing than a similar one in reference to a darkened 

 environment. 



4. A sideless maze was employed in the following experiment. 

 This consists of a series of runways separated from each other 

 by open spaces four inches in width. This maze differs from the 

 standard maze usually employed in these experiments in several 

 respects: — it is less complex as to number and length of alleys, 

 the absence of sides eliminates the possibility of a contact 

 guidance in traversing the paths, and the absence of the sides 

 and the glass cover allows the animals a more intimate sensory 

 contact with the objective environment. We were interested 

 in comparing the degree of disturbance due to rotation on such 

 a maze with that exhibited by animals in the standard maze. 

 If rotation disturbs the animals because of the alteration in 

 reference to the environment, the degree of disturbance in the 

 sideless maze should be the greater. Five rats were tested. 

 The average error record, and the number of trials necessary 

 to secure adaptation were twice those for the standard maze. 

 This ratio does not adequately represent, however, the relative 

 confusion in the two cases because it neglects the greater 

 simplicity of the sideless maze. If the two mazes offered equal 

 opportunity for error, it is safe to assume that the discrepancy 



