The muscle selected to be fatigued was usually the extensor of the right hind foot (the _tibialis anticus_), though at times the common extensor muscle of the digits of the same foot was employed. The anterior tibial nerve which supplies these muscles was bared for about two centimeters, severed toward the body, and set in shielded electrodes, around which the skin was fastened by spring clips. Thus the nerve could be protected, kept moist, and stimulated without stimulation of neighboring structures. By a small slit in the skin the tendon of the muscle was uncovered, and after a strong thread was tied tightly about it, it was separated from its insertion. A nerve-muscle preparation was thereby made which was still connected with its proper blood supply. The preparation was fixed firmly to the animal holder by thongs looped around the hock and the foot, i. e., on either side of the slit through which the tendon emerged. The thread tied to the tendon was passed over a pulley and down to a pivoted steel bar which bore a writing point. Both the pulley and this steel writing lever were supported in a rigid tripod. In the earliest experiments the contracting muscle was made to lift weights (125 to 175 grams); in all the later observations, however, the muscle pulled against a spring attached below the steel bar. The tension of the spring as the muscle began to lift the lever away from the support was, in most of the experiments, 110 grams, with an increase of 10 grams as the writing point was raised 4.5 millimeters. The magnification of the lever was 3.8. The stimuli delivered to the anterior tibial nerve were, in most experiments, single break shocks of a value barely maximal when applied to the fresh preparation. The rate of stimulation varied between 60 and 300 per minute, but was uniform in any single observation. A rate which was found generally serviceable was 180 per minute. Since the anterior tibial nerve contains fibres affecting blood vessels, as well as fibres causing contraction of skeletal muscle, the possibility had to be considered that stimuli applied to it might disturb the blood supply of the region. Constriction of the blood vessels would be likely to produce the most serious disturbance, by lessening the blood flow to the muscle. The observations of Bowditch and Warren,[76] that vasodilator rather than vasoconstrictor effects are produced by single induction shocks repeated at intervals of not more than five per second, reassured us as to the danger of diminishing the blood supply, for the rate of stimulation in our experiments never exceeded five per second and was usually two or three. Furthermore, in using these different rates we have never noted any result which could reasonably be attributed to a diminished circulation. The Splanchnic Preparation