I have repeated Mr. Hughes’s experiments with a single Leclanché cell, and for this purpose I employed a quill, filled with five fragments of carbon, taken from the cylindrical carbons of small diameter which are made by M. Carré for the electric light. I have obtained the results which are mentioned by Mr. Hughes, but I ought to say that the experiment is a delicate one. When the pressure of the fragments of carbon against each other is too great, the current traverses them with too much force to allow the calorific effects to vary the deviation of the galvanometer, and when the pressure is too slight, the current will not pass through them. A medium degree of pressure must therefore be effected to ensure the success of the experiment, and when it is obtained, it is observed that on the approach of the hand to the tube, a deviation of 90° will, after a few seconds, diminish, so that it seems to correspond with the approach or withdrawal of the hand. But breathing produces the most marked effects, and I am disposed to believe that the greater or less deviations produced by the emission of articulate sounds when the different letters of the alphabet are pronounced separately, are due to more or less direct emissions of heated gas from the chest. It is certain that the letters which require the most strongly marked sounds, such as A, F, H, I, K, L, M, N, O, P, R, S, W, Y, Z, produce the greatest deviations of the galvanometric needle. In my paper on the conductivity of such bodies as are moderately good conductors, I had already pointed out this effect of heat upon divided substances, and I also showed that after a retrograde movement, which is always produced at once, a movement takes place in an inverse direction to the index of the galvanometer when heat has been applied for some instants, and this deviation is much greater than the one which is first indicated.