Alumina silicate and alumina require a very strong current to effect their decomposition, and burn with a blue flame of small illuminating power.

Silica melts and volatilizes without undergoing decomposition.

Gaudoin has proposed 2 distinct methods of preparing carbon for electric rods. According to the first, he decomposes, by heat, organic matters capable of yielding pure carbon after decomposition, e.g. pitches, fats, etc. The decomposition is effected in closed retorts, or in graphite crucibles, at bright red heat. In the bottoms of the latter, are provided a tube for the liberation of volatile matters, and a second tube for feeding purposes. The gaseous products of decomposition are led into a condensing chamber, for recovery and utilization. The more or less compact carbon remaining in the retort is finely pulverized, and with it are mixed certain proportions of lampblack, and of the carbides of hydrogen previously produced by the decomposition process. These, being quite free from iron, are much superior to commercial hydrocarbons. The draw - plate or moulding apparatus employed by Gaudoin differs from that commonly used in the following important particulars:- The carbon is made to issue horizontally, at a descending angle of about 50°, guided by tubes, and supported so that the mould can be emptied without interruption and the carbon does not break under its own weight.

Gaudoin's second plan is to take dried wood, shaped in the form of the rod, and to carbonize and soak it in carbonaceous liquids. The wood is subjected to a glow distillation process, in order to drive of! the volatile matters; then washed in acids or alkalies, to remove impurities; and finally desiccated in a reducing atmosphere at very high temperature. The pores of the wood are closed by submitting it to the action of carbon chloride and various hydrocarbons under heat. This process promises to afford carbons which will burn at a slow rate, and give a steady light.

The advantage derived from closing the pores of carbons has been further attested by the success of the Sawyer and Mann rods, which are prepared in the following manner: - The carbon rod is immersed in olive - oil until it has become thoroughly saturated; while in this condition, it is included in a powerful electric current, the effect of which is to carbonize the oil in the pores and on the surface. Rods thus prepared are extremely hard, of steel grey colour on the surface, and give very constant light.

Bad carbons are undoubtedly rendered more uniform conductors by covering them with a coating of metal. A great increase of light is also secured by a slight coating of metallic bismuth, or by saturating with a solution of bismuth nitrate. It has been proposed to attain the same end by incorporating powdered copper or iron with the carbon; also by inserting a wire core in the rod, and by winding a thin strip of metal around it.

Jacquelain has pointed out that carbon for the electric light should be purer than that obtained by calcining wood; and, if not free from hydrogen, should at any rate contain no mineral impurities. He gives 3 methods for accomplishing this result: (1) By the action of a jet of dry chlorine gas directed on the carbon, raised to a light red heat; (2) by the action of potash and caustic soda in fusion; and (3) by the action of hydrofluoric acid on the finished carbons. Jacquelain has prepared carbons by all 3 methods, and has summed up in a table the photometric results of his experiments. He comes to the conclusion that the luminous power and regularity of the voltaic arc increase in direct ratio to the density, hardness, and purity of the carbons. He remarks, incidentally, that the natural graphitoid of Siberia possesses the singular and unexpected property of acquiring by purification a luminous capacity double that which it has in the natural state, and which exceeds by J that of pure artificial carbons. In passing dry chlorine gas over pulverized coal or coke heated to bright redness, all the silica, alumina, and magnesia, as well as alkalies and metallic oxides, are converted into volatile chlorides and expelled; even the hydrogen is driven off as hydrochloric acid.

The easiest method of carrying out the process on a large scale is to allow the dry chlorine gas to act upon gas carbon - from the retorts - cut into thin prisms, for 30 hours, and then raise the temperature to a bright white heat. This makes the carbon porous; in order to convert into a dense, heavy carbon, which is a good conductor and not easily combustible, the vapours of heavy tar - oils are passed slowly over the pieces of glowing carbon, when a deposition of carbon will take place within the pores of the coke. If the carbon rods are treated with fused sodic hydrate (caustic soda), the silica and alumina will be dissolved as soda silicate and aluminate, and can be removed by washing with hot water. Oxide of iron and other constituents of the ash are removed with hydrochloric acid followed by pure water. The simplest process recommended by Jacquelain, is to leave the carbons for 2 or 3 days in dilute hydrofluoric acid, at ordinary temperature, then wash well, and expose for a few hours to a slow current of tar vapours at a high temperature. (Comptes Ren - dus)

With direct currents, the positive carbon burns away at double the rate of the negative, owing to the much higher temperature which it undergoes, amounting to whiteness as compared with dull redness. With alternating currents, the carbons are consumed equally. This consumption is also completely avoided by producing the voltaic arc in vacuo.

Carbons of J in. diameter burned with a current of 75 amperes give a light equal to 400 gas - burners, each Using 500 cub. ft. per hour. The weight of carbon burned is 0 - 79 oz. per hour, requiring 2.11 ox. or 1 - 57 cub. ft. of oxygen. A 1/4 - in. crater in the positive carbon will necessitate a deduction of 15 per cent., giving 1.32 cub. ft. of oxygen per hour employed in the formation of carbonic acid.

Varley has devised an electric lamp in which be uses fine filaments in a rope - like bundle as the poles of the arc. It is said that the space between the two points is so heavily charged with incandescent carbonaceous matter that the resistance is considerably reduced, and the "light" is of much greater arc itself and not so much from the carbons, which no longer present the cup and cone formation, although possibly the filaments individually preserve the distinctive shape. An advantage is that the carbons are flexible, can be wound on a reel, and be payed out by means of clockwork. The carbons are made of pieces of rope soaked in paraffin or ozokerit, and carbonized in a crucible kept constantly filled with a hydro - carbon atmosphere.