The large coil was used. In all cases, with the one wire, the glow was very faint as compared with that of the closed circuit. [Sidenote: Ether vapour.] (1) _Ether Vapour._—With both wires, in company with the usual bright bands of the carbon spectrum, shading-off towards the violet, the H lines were very sharp and brilliant. With the one wire only, the carbon bands were left faintly shining, with both sides nebulous alike, and with no shading-off towards the violet. (We were not quite sure whether this was not the effect of the reduction of the light.) The H lines, though originally stronger than the carbon bands, quite disappeared from the spectrum. [Sidenote: Coal-gas.] (2) _Coal-gas._—The same effects were produced; but we thought we could detect very faint traces of the H lines. [Sidenote: Nitrogen.] (3) _Nitrogen._—The N lines, as well as those of H (also seen in the tube), were much fainter with one wire, but the H lines more so in proportion. [Sidenote: Hydrogen.] (4) _Hydrogen._—Only a marked reduction in brilliancy of the whole spectrum. [Sidenote: Oxygen, N and H.] (5) _Oxygen._—An impure tube, showing O (some of the lines hydrocarbon?), N, and H spectra simultaneously. With one wire the O lines still remained fairly bright, the N and H being only faintly seen. [Sidenote: Water-gas.] (6) _Water-gas._—Same effect. [Sidenote: Turpentine vapour.] (7) _Turpentine Vapour._—Same effect as ether, but the H lines could be faintly seen. CHAPTER XVIII. ACTION OF THE MAGNET ON THE ELECTRIC SPARK. [Sidenote: Apparatus employed.] The magnet was excited with two plates of the large battery, and the larger coil with the other two plates, the action in both cases being strong. 1. A spark from the coil was passed between two platinum wire electrodes, about three centimetres apart. [Sidenote: Spark and aura described.] It consisted centrally of a thin stream of bluish-white light, vividly bright, around which was seen a narrow, uniform, diffuse, yellow-tinted aura, which accompanied the spark in all its movements. The spark always struck across from the extreme points of the electrodes (see Plate XVII. fig. 5). [Sidenote: Effect of magnet upon the aura.] 2. On being placed between the conical poles of the excited magnet the bright thread of the spark did not change; but instead of the inconsiderable yellow-tinted aura which accompanied the unmagnetized spark, there now struck out, at right angles to the magnet-poles, a thin rosy-tinted half-disk of aura-like flame. This extended aura ran considerably along each electrode, though the spark proper still struck from the points. [Sidenote: Extended aura described.]