The last of the factors mentioned above--the effect of fatigue on the irritability of the nerve-muscle combination, or on the muscle alone--can be tested by determining variations in the least stimulus capable of causing the slightest contraction, the so-called “threshold stimulus.” As the irritability lessens, the threshold stimulus must necessarily be higher. The height of the threshold is therefore a measure of irritability. How does fatigue affect the irritability of nerve-muscle and muscle? How is the irritability of fatigued structures affected by rest? How is it influenced by adrenin or by adrenal secretion? Answers to these questions were sought in researches carried on by C. M. Gruber[85] in 1913. The Method of Determining the Threshold Stimulus The neuro-muscular arrangements used in these researches were in many respects similar to those already described in the account of experiments by Nice and myself. To avoid the influence of an anesthetic some of the animals were decerebrated under ether and then used as in the experiments in which urethane was the anesthetic. The nerve (the _peroneus communis_) supplying the _tibialis anticus_ muscle was bared and severed; and near the cut end shielded platinum electrodes were applied. These electrodes were used in fatiguing the muscle. Between these electrodes and the muscle other platinum electrodes could be quickly applied to determine the threshold stimulus and the tissue resistance. These second electrodes were removed except when in use, and when replaced were set always in the same position. Care was taken, before replacing them, to wipe off moisture on the nerve or on the platinum points. For determining the threshold stimulus of the muscle the skin and other overlying tissues were cut away from the _tibialis anticus_ in two places about 5 centimeters apart. Through these openings platinum needle electrodes could be thrust into the muscle whenever readings were to be taken. Local polarization was avoided by reinserting the needles into fresh points on the exposed areas whenever new readings were to be taken. The tendon of the _tibialis anticus_ was attached, as in the previous experiments, by a strong thread passing about pulleys to a lever which when lifted stretched a spring. During the determination of the threshold the spring was detached from the lever, so that only the pull of the lever itself (about 15 grams) was exerted on the muscle.