The author presents a survey of results of an empirical research into the mental efficiency of pupils in the ninth year of the Elementary School, performed by means of a modified A-form of Dücker’s arithmetic K—L—T test* In the experiment of type one the test was given during the first teaching period and repeated (in a fortnight’s time) during the fifth period (279 subjects). In the experiment of type two the test was first given during the fifth period and repeated (in a fortnight’s time) during the first period (286 subjects). Statistically very significant differences in the quantity of the subjects’ performance between the first and the repeated performance of the test were ascertained in both types of the experiment. Very significant positive correlations were ascertained between the quantity of the subjects’ performance in the first and in the repeated performance of the test. The quantity of the subjects’ performance when the test was done for the first time was equivalent, irrespective of whether the test was done in the first period or in the fifth period; it was also equivalent when the test was repeated, irrespective of whether it was done in the fifth period or in the first period. The distribution of the quantity of the subjects’ performance in the course of the experiment differs from the theoretically normal distribution. When the test is repeated the average number in the subjects’ performance increases. When the test is repeated in the fifth period the differences ascertained are statistically very significant; when it is repeated in the first period they are insignificant. A very significant negative correlation was ascertained between the quality of the first performance and the repeated performance of the test. The distribution of the number of errors in the subjects’ performance during the experiment differs from the theoretically normal distribution. During the course of the experiment the quantity of the performance as well as the number of errors in the performance changes, but the percentage of errors remains relatively constant, i. e. the quantity of the subjects’ performance changes, but the quality of the performance remains relatively unchanged. The author is critical about Dücker’s psychological accounting for the test (content validity), which he considers excessively general and simplified, far from covering all the possible determinants of the quantity and quality of the performance. The results of the research have confirmed that the psychical efficiency of pupils and the beginning and at the end of a school day is characterized by certain differences which, however, cannot be grasped by simply finding out the overall quantity or quality of the performance, but by an analysis of the dynamic of changes in further time indices. They justify the demand for a deeper and more detailed micro-analysis of further indices of psychical efficiency, the need for discovering their causal relationships and the laws governing their changes.