The experiment which forms the subject of this communication was originally devised as an illustration of the explanation of all kinds of microphone receivers, suggested by the beautiful experiments of Mr Blyth, on loose contacts. My idea was to replace Mr Blyth's heated point of metal by a continuous portion of the circuit which should act in the same manner.

It was obvious, for two reasons, that this part must be of small diameter; 1st, in order that the resistance, per unit of length, might be great enough to make the variation of the heating sufficient to cause sensible longitudinal extension; 2d, in order that the section might be small enough to allow sensible cooling in, say the of a second. I had reason beforehand to believe that the second of these conditions could be fulfilled in practice; because, I found in my experiments on Ohm's Law (Brit. Ass. Rep. 1876, p. 58, et seq.) that, when currents of two different strengths alternated, even with great rapidity (60 times per second) in a fine wire, the resistance was sensibly higher during the passage of the stronger current.