The practical efforts of the factories have, indeed, not yet reached the point at which the greatest possible achievement which can be reached without over-fatigue may be secured. We called the abbreviation of the working day an experimental scheme. The question of reducing the working hours is so simple that no further special experiments are needed. But when we come to the questions of the pauses at work, the speed of work and similar factors related to fatigue, the situation is by far more complicated, and the often capricious changes in the plant have very little in common with a systematic experiment. Some well-known studies of the efficiency engineers clearly demonstrate the possibility of such systematic efforts. The best-known case is probably Taylor's study of the pig-iron handlers of the Bethlehem Steel Company. He found that the gang of 75 men was loading on the average about 12-1/2 tons per man per day. When he discussed with various managers the question of what output would be the possible maximum, they agreed that under premium work, piecework, or any of the ordinary plans for stimulating the men, an output of 18 to 25 tons would be the extreme possibility. Then he proceeded to a systematic study of the fatigue in its relation to the burden and of the best possible relation between working time and resting time. His first efforts to find formulas were unsuccessful, because he calculated only the actual mechanical energy exerted and found that some men were tired after exerting energy of 1/8 hp., while others seemed to be able to produce the energy of 1/2 hp. without greater fatigue. But soon he discovered the mistake in his figures. He had considered only the actual movements, and had neglected the period in which the laborer was not moving and was not exerting energy, but in which a weight was pulling his arms and demanding a corresponding muscular effort. As soon as this muscular achievement was taken into account, too, he found that for each particular weight a definite relation exists between the time that a man is under a heavy load and the time of rest. For the usual loads of 90 pounds, he found that a first-class laborer must not work more than 43 per cent of time working day and must be entirely without load 57 per cent. If the load becomes lighter, the relation is changed. If the workman is handling a half pig weighing 46 pounds, he can be under load 58 per cent of the day and only has to rest during 42 per cent.[43]