M. Clerc, the well known inventor of the sun lamp, has tried to overcome troubles of this sort, and has attained a simple, elegant, and at the same time cheap solution. The cut shows the arrangement. The apparatus is connected at the points, BB', with the lighting circuit. The current entering by the terminal, B', passes through the coils of a bobbin, S, before reaching the points of attachment, a and b, of the lamp, L, the normally working one. Thence the circuit runs to B. Within the coil, S, is a small hollow cylinder, T, of thin sheet iron, which is raised parallel with the axis of the bobbin during the passage of the current through the latter. At its base the cylinder is prolonged into two little rods, h and h', which plunge into two mercury cups, G and G'. The cut shows that one of the cups, G', is connected to the terminal, B', and the other, G, to the terminal, a', of the other lamp, L'. An inspection of the cut shows just what ensues when an accident happens to the first lamp while burning. The first circuit being broken at ab, the magnetizing action of the current in the bobbin ceases, the cylinder, T, descends, and the rods, h and h', dip into the mercury. It follows that the current, always starting from the terminal, B', will by means of the cups, G and G', pass through the lamp, L', to go by the original return wire to B. [Illustration] The substitution of the lamp, L, for L' is almost instantaneous. It can scarcely be perceived. It goes without saying that such an arrangement of automatic commutation is applicable to lamps with two or more filaments of which only one is to be lighted at a time. The apparatus costs little, and can be made as ornamental as desired. No exaggeration is indulged in if we pronounce it simple and ingenious. It may be used in a great variety of eases. The diameter of the wire is 55/100 (22 mm.), its length eighteen meters (60 feet), its resistance one ohm; ¾ ampere is needed to work it, and less than a watt is absorbed by it.--_Electricite_. * * * * * DEFINITIONS AND DESIGNATIONS IN ELECTROTECHNICS. We may discourse for some time to come upon the uniformity of electric language, for universal agreement is far from being established. An important step toward the unity of this language was taken in 1881 by the congress of Paris, which rendered the use of the C.G.S. system definitive and universal. This labor was completed in 1884 by the meeting of a new congress at Paris, at which a definition of the C.G.S. and practical units was distinctly decided upon. That the unit of light defined by the congress has not rapidly come into favor is due to the fact that its practical realization is not within everybody's reach.