The results of the investigations on four metals, viz. steel, iron, copper, and platinum, form the subject of my two first papers. In Parts I. and II. the effects on the modulus when the wire was stretched with a small load, and also with a much greater load, were examined. In this, my third paper, the investigation of these metals has been extended in several directions. The scope of the whole work has also been widened by subjecting nickel and cobalt to examination.

As the question of temperature still seemed doubtful, and as the justification given near the beginning of the second paper might not be altogether convincing, I thought it better to put the matter beyond all question. I therefore adopted the following method. Using the doublewalled tube already described, the wires were passed through it and over the wheel in the same way as in the main experiments.