apparatus, are in a direct ratio to the active surfaces, it seemed to me of great importance to establish the relation between the surface of a voltaic pair and the weight capable of being supported by a bar of soft iron submitted to the magnetizing power of the current. A bar of soft iron 1½ inch in diameter by 29 inches in length, weighing 14½lbs., was bent into a horseshoe, so that the centres of the branches were seven inches apart. The bar, covered with silk, was covered by a helix of copper wire of 1¼ line thick and 35 feet long. The magnetic power was measured by means of a steelyard, and a weight supported by rollers, in order to slip easily over the arm of the lever. The surface of the soft iron armature was of a somewhat convex form, in order that the extremities of the branches, forming a flat and smooth face, should only be touched in the direction of an edge, the position of which formed a right angle with the direction of the lever. The armature was in contact with the extremities of the branches when the lever was placed horizontally. Upon the latter had been marked a scale, the divisions of which indicated the thirtieth part of the sliding weight, to which was affixed an index: it was easy to estimate the tenths of these divisions. I had taken the necessary precautions to avoid as much as possible the errors of observation arising from the disposition of the apparatus. I shall not enter here into the details of the construction of this rather complicated apparatus, which I intend to give elsewhere, as it may be useful for experiments of this kind. The electromotors which I employed consisted of copper troughs three quarters of an inch wide, and sufficiently large to enable me to immerse in them respectively the plates of zinc of 4, 16, 36, 64, 100, 144 square inches. The contact of these last with the copper was prevented by the interposition of pieces of wood. The conducting liquid, of which I had previously prepared a sufficient quantity to serve for a series of experiments, was acidulated with ten per cent. of concentrated sulphuric acid of the specific gravity of 1·840. The experiments, with the same voltaic pair, were made without interruption; but after each one precautions were taken to cleanse carefully the zinc plates, to wash the trough with water and to renew the liquid, in order to restore the same state of action. But subsequent observations convinced me that the original state is restored more certainly by exposing the plates, and especially the negative one, to a current of air, until it is perfectly dried. It will then be no longer necessary to renew the liquid so frequently, especially when the observation is confined to the primitive state. It must be acknowledged that I subsequently found the copper troughs to be ill adapted for electro-magnetic experiments; concentric cylinders, which may be plunged in the liquid, are much better. These cylinders must be fixed firmly enough to remain at the proper distance, without recurring to the interposition of wood or of any other insulating matter. Much more constant galvanic effects may be obtained if the space occupied by the liquid between the two metals be not
