In preceding studies it has been shown that the skeletal and cardiac musdes of patients dying of congestive heart failure were abnormally low in potassium (Harrison, Pilcher and Ewing, 1930). The loss of potassium from the skeletal musde was apparently due to edema (Pilcher, Calhoun, Cullen and Harrison, 1930). The loss of potassium from the cardiac musde was not due to edema but appeared to be related to over-work (Calhoun, Cullen, Clarke and Harrison, 1930). This condusion was based on the observation that in a given patient the ventride which had presumably been most overworked showed the greatest diminution in potassium content. In patients, however, complicating factors are often present and, before' regarding these condusions as established, it seemed wise to test them by animal experiments. The first series of observations were made on skeletal musde. This is more suitable than cardiac musde for the purpose because the musdes of one extremity can be overworked while those of the opposite limb are used as controls. METHOD Dogs were used. They received 0.3 gram of barbital per kilogram of body weight, one or two hours before the experiment was begun. The two sciatic nerves were exposed in the thigh, and one of them was stimulated rhythmically by means of electrodes connected with an induction coil and dry cell batteries. The muscles were made to Aided by a grant from the National Research Council.