The method of measuring the stimulating value of the electric current which was used in testing the threshold was that devised by E. G. Martin[*] of the Harvard Laboratory--a method by which the strength of an induced electric shock is calculable in definite units. If the tissue resistance enters into the calculation these are called β units. When the threshold of the nerve-muscle was taken, the apparatus for the determination was connected with the nerve through the electrodes nearer the muscle. They were separated from the fatiguing electrodes by more than 3 centimeters, and arranged so that the kathode was next the muscle. When the threshold of the muscle was taken directly the apparatus was connected with the muscle through platinum needle electrodes thrust into it. The position of the secondary coil of the inductorium, in every case, was read by moving it away from the primary coil until the very smallest possible contraction of the muscle was obtained. Four of these readings were made, one with tissue resistance, the others with 10,000, 20,000, and 30,000 ohms additional resistance in the secondary circuit. Only break shocks were employed--the make shocks were short-circuited. Immediately after the determination of the position of the secondary coil, and before the electrodes were removed or disconnected, three readings of the tissue resistance were made. From these data four values for β were calculated. *[Footnote: For a full account of Dr. Martin’s method of calculating the strength of electric stimuli, see Martin: The Measurement of Induction Shocks, New York, 1912.] The strength of the primary current for determining the threshold of the nerve-muscle was usually .01 ampere, but in a few cases .05 ampere was used. For normal muscle it was .05 ampere and for denervated muscle 1.0 ampere. The inductorium, which was used throughout, had a secondary resistance of 1400 ohms. This was added to the average tissue resistance in making corrections--corrections were made also for core magnetization. The Lessening of Neuro-muscular Irritability by Fatigue The threshold for the _peroneus communis_ nerve in decerebrate animals varied from 0.319 to 2.96 units, with an average in sixteen experiments of 1.179.[*] This average is the same as that found by E. L. Porter[86] for the radial nerve in the spinal cat. For animals under urethane anesthesia a higher average was obtained. In these it varied from .644 to 7.05, or an average in ten experiments of 3.081. *[Footnote: For the detailed data of these and other quantitative experiments, the reader should consult the tables in the original papers.]