In place of two electrodes of similar form and dimensions, one electrode consists of a large bun-shaped disk of carbon placed with the rounded face downward. The other carbon is a fine rod of carbon of about 1/8 or 5-32 inch in diameter. The upper end of this is pointed and maintained in contact with the center of the lower surface of the disk. This rod is supported by means of a spring collar, which also forms the circuit connection. This is within about 3/4 in. of the top of the carbon, so that the 3/4 in. becomes incandescent, and the contact between the two carbons being only a point, a small electric arc is produced between the two carbons, while the electricity is at the same time passed on through the carbon disk, and the connections there attached to the next lamp. [Illustration: DIAGRAM OF CURRENT.] Referring to our diagrams, in Fig. 1 the upper carbon is shown at C, and the rod carbon at c. The former is supported by means of an adjustable jointed bracket, B, attached to the wood stand. The rod carbon is guided by the spring collar on the top of the stand, and to which the connection is made, and is supported by the fine cord running over the pulley, P. This cord is attached to the clasp, D, at the bottom of the rod, and to the balance weight, W, by which the rod is maintained in constant, practical, though not absolute contact with the disk. Round the upper part of the disk is a metal band, A, to which the circuit wire is attached, and the current thus passed on to the next lamps. At a recent trial of this lamp, the current was derived from a small Gramme electro-plating machine, requiring only 2 horse power to put it in full work. It may therefore be assumed that this was about the limit of the power at work to produce the light. At the commencement of the proceedings two lights were maintained, each stated to be equal to 320 sperm candles. At this rate the two lights would be equal to 640 candles, or 40 full power gas lights, each consuming 5 cubic feet of 16 candle gas per hour. Such gas lights, it may be observed, are not often seen, except in the argand form. The two lights burned with extreme steadiness, there being no undulation, or flickering whatever, although there was no glass globe to tone down any variations of luster. The lights were perfectly bare and unprotected, and the place where the trial was made was a workshop of moderate size. Later in the evening one light was exhibited outside the building, in an open thoroughfare, and the same perfect steadiness was observable. After the two lights had been burning for a time they were extinguished, and the current was sent through a row of ten lamps. The light per lamp was of course reduced, but there was the remarkable fact that ten lights were maintained by a comparatively weak machine, driven by an engine exerting the power of only two horses.