as soon as they are seriously studied from a psychological point of view, remains really a source for surprise.  Sometimes no more is needed than a change in the windows or in the electric lamps, by which the light can fall on the work in a psychologically satisfactory way; sometimes long series of experiments have to be made with a simple hammer or knife or table.  Often everything must be arranged against the wishes of the workingmen, who feel any deviation from the accustomed conditions as a disturbance which is to be regarded with suspicion.  In one concern I heard that the scientific manager became convinced that all the working-chairs for the women were too low and that the laborers therefore had to hold their arms in a psychophysically unfavorable position during the handling of the apparatus.  All were strongly opposed to the introduction of higher chairs.  The result was that the manager arranged for the chairs to be raised a few millimeters every evening, without the knowledge of the working-women, as soon as the factory was empty.  After a few weeks the chairs had reached the right height without those engaged in the work having noticed it at all.  The outcome was a decided increase of efficiency.

But the most rational scheme will after all be to prepare for such arrangements of tools and apparatus by systematic experiments in the psychological laboratory.  The subtlety of such investigations will lead far beyond the point which is accessible to the attempts of scientific management.  Exact experiments on attention, for instance, will have to determine how the various parts of the apparatus are to be distributed best in space if the laborer must keep watch for disturbances at various places.  Only the laboratory experiment can find the most favorable speed of the machine or can select the muscles to which the mind can send the most effective impulses.  The construction of the machine must then be adapted to such results.  In the Harvard laboratory, for instance, a practical question led us to examine which fingers would allow the quickest alternation of key movements.[29] If any two of the ten fingers perform for ten seconds the quickest possible alternation of motion, as in a trill, the experiment can demonstrate exactly the differences between the various combinations of fingers and the individual fluctuations for these differences.  With an electrical registration of the movements of the alternating fingers we studied in hundredths of a second the time for the motions of two hands and of fingers of the same hand, in order to adjust the keys of a certain machine to the most favorable impulses.