We found, however, that the cyanide when exposed to the atmosphere at a temperature above 300 deg.  C. is readily destroyed under reformation of carbonate of barium, so that it is absolutely necessary to cool it down to this temperature before exposing it to the atmosphere, a fact of great importance that had hitherto been overlooked.

The operation for producing ammonia and regenerating the carbonate of barium by acting upon the cyanide with steam offers no difficulty whatever, and if the temperature is not allowed to exceed 500 deg.  C., the results are quantitative.  The regenerated carbonate of barium acts actually better than the ground witherite used in the first instance, and if care is taken that no impurities are introduced by the pitch which is used to remake the briquettes and to replace the small amount of carbon consumed at each operation, I see no reason why it should not continue to act for a very long time.

The cyanide is not acted on by carbonic oxide, but carbonic acid destroys it at high temperatures, so that it is not possible to produce it by heating the briquettes directly in a flame free from oxygen, but containing carbonic acid.  The process has, therefore, to be carried out in closed vessels, and I designed for this purpose the following apparatus: 

Clay retorts of moderate dimensions and thin walls are placed vertically in a furnace, passing through the hearth as well as through the arch of the furnace.  These are joined at the bottom to cast iron retorts of the same shape as the earthenware retort.  Through a cast iron mouthpiece on the top of the retort the material was introduced, while in the cast iron retort below the material was cooled to the necessary temperature by radiation and by the cold nitrogen gas introduced into the bottom of it.  The lower end of the cast iron retort was furnished with an arrangement for taking out from time to time small quantities of the material, while fresh material was in the same proportion fed in at the top.  As a source of nitrogen I used the gases escaping from the carbonating towers of the ammonia-soda process.  The formation of cyanide of barium from barium carbonate, carbon, and nitrogen absorbs a very large amount of heat—­no less than 97,000 calories per equivalent of the cyanide formed—­which heat has to be transmitted through the walls of the retort.  I therefore considered it necessary to use retorts with very thin walls, but I did not succeed in obtaining retorts of this description which would resist the very high temperatures which the process requires, and for this reason I abandoned these experiments.  I was at that time not acquainted with the excellent quality of clay retorts used in zinc works, with which I have since experimented for a different purpose.  I have no doubt that with such retorts the production of cyanides by this process can be carried out without great difficulty.

I believe that the process will prove remunerative for the manufacture of cyanogen products, which, if produced more cheaply, may in the future play an important role in organic synthesis, in the extraction of noble metals, and possibly other chemical and metallurgical operations.