110 rats were trained on a brightness discrimination habit. They were divided into the following 6 groups: 1) a normal-dark-shocked group, 2) an operated-dark-shocked group, 3) a normal-dark-non-shocked group, 4) an operated-dark-non-shocked group, 5) a normal-light-non-shocked group, and 6) an operated-light-non-shocked group. The nervous lesions were comparable in the 3 operated groups. The results indicate that 1) cerebral lesions cause a reversal of preference from the darker to the lighter of 2 alleys, 2) under non-shock conditions operated animals are inferior to normals but the difference is smaller when shock is used, 3) operated animals are inferior to normals in their capacity to reverse their preference. Results similar to those found for learning were also obtained in the retention tests. The author ascribes the deficiencies of the operated animals to a lowering of the "level of attention" or "vigilance" of the animal.