In 1865 Lord Kelvin published the results of experiments which first made evident “a very remarkable fatigue of elasticity, according to which a wire which has been kept vibrating for several hours or days through a certain range came to rest much quicker when left to itself than when set in vibration after it has been at rest for several days and then immediately left to itself.” On the strength of Lord Kelvin's statement this elastic fatigue of metals has been regarded as a definitely ascertained fact. But, quite recently (Physical Review, March 1899), Mr Joseph O. Thompson has published a paper “On the period and logarithmic decrement of a continuously vibrating wire,” in which he states that it seems probable that “for constant temperature and constant amplitude the logarithmic decrement is constant.” This conclusion is based upon his observation that the logarithmic decrement, in the case of a copper wire, when the amplitude of vibration varied from about 185° to 175°, had the same value after it had been continuously vibrated through the average arc of 180° for fifty consecutive hours as it had at the commencement of that period. Observations upon other metals seemed to support the conclusion.

Lord Kelvin also stated that fatigue caused an increase of the period of vibration. Mr Thompson finds that “no matter what metal was used, no matter whether the arc of vibration was as small as 20° or as high as 200°, no matter whether the wire was long or short, thick or thin (provided of course the breaking strength of the wire was at least twice the weight of the disc), the result was uniformly the same, namely, that when temperature and amplitude of vibration remained constant, the period of vibration was a constant quantity.”