MAZE STUDIES WITH WHITE RAT 279 



corresponding values for the normals were 10 and 5.6. The 

 average variation relative to the size of the errors is thus much 

 greater for the blind group. 



Position of Cage. — Both groups contained ten rats. A smaller 

 percentage of the blinds was affected, the values being 40 and 

 70. Those blinds affected were disturbed in a greater percentage 

 of their trials (50 vs. 41), and made the greater average error 

 score (1.33 vs. .87). The blind animals require a longer dura- 

 tion of exposure to induce an effect; they were disturbed only 

 for the 24-hr. exposures, while the normals were affected by a 

 15-min. exposure. The blind rats also possessed the poorer 

 adaptive power, for the normals became so accustomed to the 

 novel situation in 24 hours that a disturbance was no longer 

 manifest. 



Rotation of Maze-. Two blinds were tested on the first type 

 of ma^e rotation, in which the three positions were tested on 

 successive days. Their records are to be compared with those 

 of ten normals. All members of both groups were disturbed. 

 The blinds made errors in a greater percentage of their trials 

 (67 vs. 65), but their average error record was much smaller 

 (3.33 vs. 6.95). With a repetition of the test the poorer adap- 

 tive ability was manifested by the animals without vision; they 

 decreased the percentage of trials in which error was present 

 from 67 to 5&, and their error record from 3.33 to 2.50. The 

 visual group on the contrary reduced their error record from 

 6.95 to 1.72 and the percentage of runs with error from 65 to 47. 



Rotation of Heterogeneous Maze Environment. The records of 

 fourteen blind animals are to be compared with those of seven 

 normal rats,. A greater percentage of blind rats was disturbed 

 (78 vs. 71). The errors of the blind group were confined to 

 a smaller percentage of the trials (38 vs. S3). The blinds gave 

 the larger error score (2.32 vs. 1;90) in spite of the fact that 

 the errors were limited to fewer trials. The discrepancy is 

 much greater when we compare the total number of errors 

 per rat (18 vs. 12). The blinds exhibited the greater range of 

 variability as to number of errors per rat; for the blinds the 

 errors ranged from 3 to 70 with a mean variation of 15. The 

 range for the normals was 9 to 15 with a mean variation of 

 2.2. The normal rats appeared to react definitely to the altered 

 conditions. With each new change of conditions the errors were 



