making them positive. This plan greatly reduced the local action and increased the life of the cup about ten times.*

The first cost of the electrically fused magnesia filling for the cup was very high, so a cheaper material was looked for. This was found in a mixture of ground, dead-burned magnesite and Portland cement. It is mixed with water, poured between the perforated walls of the cup, and after it has set and hardened it makes a very efficient and satisfactory diaphragm.† Reasoning a priori one would expect Portland cement to be readily attacked by fused sodium hydroxide and sodium nitrate, but experience has shown that it stands the fluxing action of these two substances remarkably well.

=

The nitrogen dioxide and oxygen evolved at the positive poles are conducted by means of earthenware pipes to a number of receivers or Woulff bottles, connected together and containing water. (See Fig. 2). The NO, gas coming in contact with the water combines with it to form nitric acid, 3NO2 + H2O 2HNO3 + NO. The NO takes up a molecule of oxygen to again form NO2, and more nitric acid is formed. If it is desired to make a very strong acid for use in the manufacture of high explosives, a system of towers that automatically brings the strength of the acid up to required degree is used.

Each furnace takes a current of about 400 ampères at an average E.M.F. of 15 volts. External heat is used only when starting up and when changing the cups, which have a life of 425 to 450 hours; at other times the heat generated by the resistance to the passage of the current is sufficient to keep the electrolytes melted. Fig. 3 shows the switchboard with rheostats, meters, etc., for controlling and measuring the current, also the connections to the furnaces and the workmen engaged in dipping off sodium from the furnaces.

The general efficiency of the process is high, and although a higher voltage and, consequently, more power is required than is needed to decompose caustic soda, yet, on account

* U. S. Patent 641,438, January 16, 1901.

† U. S. Patent 641,376, January 19, 1901.

of the valuable by-product of nitric acid obtained, the cost of the sodium produced is much less than when it is made from caustic soda.

Up to ten or eleven years ago about the only use for metallic sodium, outside of the laboratory, was in the manufacture of aluminum, and when the the electrolytic method of making aluminum was discovered, it looked as if the production of sodium on a large scale would cease. It was only when electricity was also applied to the production of sodium that it could be produced cheaply enough to be used in fields that had hitherto been closed to it on account of its high cost. Chief amongst these new uses is the manufacture of alkaline cyanides so largely used in the extraction of gold from low-grade ores and tailings for electro-plating in photography, and other minor uses. Large amounts are also converted into sodium peroxide, to be used in bleaching wool, silk, feathers, etc., replacing the more expensive hydrogen peroxide in that work.

It is also used in making certain aniline colors and organic compounds, and wherever a powerful reducing agent is needed.

The conversion of sodium into sodium peroxide is easily accomplished by burning it in an excess of dry air, free from carbon dioxide, in a suitable externally-heated retort. On removal from the retort it is ground while still hot to a fine powder, and, as it deteriorates quickly on exposure to the air, it is packed and shipped in air-tight cans. It is a most powerful oxidizing agent, and must be handled with great care. The commercial article contains about 18 per cent. of available oxygen when fresh.

The old method of making potassium cyanide, first described by Leibig, was to take dehydrated ferro-cyanide of potassium and heat it with potassium carbonate—2KFe (CN), + 2 K,CO = Io KCN + 2 KCNO + Fe+CO.. The resulting cyanide contained cyanate, but this did not materially interfere with its use. An almost pure cyanide can be obtained by heating the ferro-cyanide alone until it decomposes into potassium cyanide, nitrogen, and a compound of iron and carbon K4Fe(CN)6 =4 KCN + N, + FeC,

FIG. 3.-Switchboard and connections to sodium furnaces.

=

This method entails the loss of one-third of the nitrogen : contained in the ferrocyanide, and to avoid this waste of nitrogen, Erlenmeyer proposed to add the proper amount of an alkali metal to the melted ferrocyanide-K4Fe(CN)6+ 2 Na 4 KCN + 2 NaCN + Fe. It is in this way that the most of the so-called chemically pure potassium cyanide now sold is made. The product is not a pure potassium cyanide, but a mixture of potassium and sodium cyanides. It also contains a considerable quantity of potassium carbonate, which is added to it during the course of manufacture to reduce its strength, for the combined cyanides, made according to the above formula, have a higher percentage of cyanogen than chemically pure potassium cyanide could possibly have. An inert material like potassium carbonate is therefore added in sufficient quantities to reduce the cyanogen contents to 39-40 per cent., which is equivalent to 98-100 per cent. cyanide of potassium.

Other processes have been devised for using sodium in making cyanides. One is to first convert the sodium into sodium amid by heating it in contact with ammonia gas, Na + NH, NaNH2+H, then heating the amid with carbon to form cyanide NaNH2+ C = NaCN + H2.

=

2

Another and later method by which it is claimed a better yield is obtained, is to form at a temperature of about 400° C.. a stable cyanamid from alkali amid and carbon, according to the formula-2 (NaNH2) + C = Na,N2C + H1, then reacting the produced cyanamid with a further quantity of carbon at a temperature of 800° C. to form cyanide, according to the reaction-Na,N2C + C = 2 NaCN. The objection to these methods is the large amount of expensive sodium metal needed for a given output of cyanide.

The writer has lately devised a process for using sodium in the synthetic production of sodium cyanide that gives good results, and in which the larger portion of the metallic base required is furnished in the form of caustic soda, and only a small amount of sodium is needed to finish the process. By this process a moderate sized sodium plant can produce enough metal to manufacture a large amount of the cyanide. The details of this new and interesting method of making sodium cyanide will be published later.