Referring to Fig. 1, A and B are two troughs rectangular in cross section attached to the supports in such positions that their axes are inclined to one another so as to form the letter V, as shown in the figure. Within these troughs slide freely the two carbon pencils, which are of circular cross section, meeting, when no current is passing, at the lower point, E. The carbon-holder, B, to the right of the figure, is rigidly attached to the framing of the lamp, but the trough, A, which carries the negative carbon, is attached to the framing by a pivot shown in the figure, and on this pivot the carbon holder can rock, its motion being controlled by the position of the armature of an electro-magnet, M, the coils of which are included in the circuit of the apparatus. By this means, the moment the current is established through the lamp, the armature is attracted, and the points of the two carbons are separated, thus forming the arc. The positive carbon, B, is held from sliding and dropping through the trough by the gentle pressure against it of the smaller carbon rod, Cą, which also slides in a trough or tube fixed in such a position that the point of contact between the two rods is sufficiently near the arc for the smaller rod to be slowly consumed as the other is burnt away; the latter in that way is permitted to slide gradually down the trough as long as the lamp is in action. The negative carbon-holder, A, is provided with a little adjustable platinum stop, E, which by pressing against the side of the conical end of the negative carbon, holds the latter in its place and prevents it sliding down the trough except under the influence of the slow combustion of the cone during the process of producing the arc. The position of the stop with respect to the conical end is determined by a small adjusting screw shown in the figure. This arrangement of stop is identical in principle with that adopted by Messrs. Siemens Brothers in their "abutment pole" lamp, and is found to work very well in practice on the negative electrodes, but is inapplicable on the positive carbons on account of the higher temperature of the latter, which is liable to destroy the metallic stop by fusion, and it is for this reason that the positive carbon in Mr. Hedges' lamp is controlled by the method we have already described. For alternating currents, however, the abutment stop may be used on both electrodes. [Illustration: Figs. 3 and 4.]