In the positive bulb the stream, instead of joining the point of the electrode, left this and ran along one side of the whole length of the wire. (See effect, Plate XVII. fig. 3.) The spectroscope was applied to the neck of one of the bulbs where the stream was bright. Without the magnet a faint continuous spectrum, mainly of the blue and green, with very slight traces of the yellow and red, was seen. Upon this, five or six faint but sharp and metallic-looking lines were seen. On the magnet being excited, the continuous spectrum was not changed; but the sharp lines shone out brighter and clearer, one in the blue being especially conspicuous. These lines were measured with a micrometer; and their places being compared with Lecoq de Boisbaudran’s “Spectres lumineux,” they were easily recognized to be those of tin. On each excitation of the magnet the same brightening of the lines took place. _Effect of Magnet on Tin-Chloride Geissler Tube._ [Sidenote: Geissler tube, Sn Cl₄, examined. Glow described. Effect of the magnet.] We then examined a Geissler tube, marked Sn Cl₄. When first excited by the small coil, the spark passed freely. The glow in the bulbs was of a diffused, light purple tint; the positive electrode had a bright yellow glow around it. The capillary stream was of a sharp green-yellow, at times brightening up to a metallic-looking green. When the magnet was first employed, the tube distinctly and permanently brightened up throughout. [Sidenote: Spiral formed in positive bulb. Glow in tube extinguished.] The negative bulb was not much changed in appearance; but in the positive bulb a curious permanent and steady cloud-like spiral, of a purple colour, made its appearance, and lasted while the tube was under the magnetic influence. (See Plate XVII. fig. 12.) After a short time the tube seemed to lose a great deal of its conducting-power, and to light up in a feeble and intermittent manner, brightening only when the coil was made to work its best. While in this condition, the magnet (which had been previously disconnected) was excited, and at once what moderate glow was still shining in the tube was totally extinguished. At first it was thought some accident might have happened to the conducting-coil wires; but repeated trials satisfied us that the effect was due to the magnetic influence alone. Efforts were made, by looking to the coil and battery, to brighten up the tube as at first, but they quite failed; and it was evident some change had taken place in its conducting qualities. This tube was accidentally broken, so that we had no opportunity to renew the experiments. [Sidenote: Another tin-chloride tube tried. Glow described.]