Among experiments with processes of the second class, based upon the formation of nitrides and their subsequent decomposition, the nitrides of boron and titanium had received most attention from inventors.  The nitride of boron, which is obtained by treating boracic acid with carbon in the presence of nitrogen, when acted upon by steam, forms boracic acid again and yields the whole of its nitrogen in the form of ammonia, but the high temperature at which the first reaction takes place, and the volatility of boracic acid in a current of steam, make it impossible to utilize this reaction industrially.

There seemed to be a better chance for a process patented by M. Tessier du Mothay, who proposed to bring a mixture of nitrogen and hydrogen into contact with titanium nitride and thus to form ammonia continuously.  Titanium is the only element of which we know at present several combinations with nitrogen, and the higher of these does, on being acted upon by a current of hydrogen at an elevated temperature, produce ammonia and a lower nitride of titanium; but this lower nitride does not absorb nitrogen under any of the conditions under which we tried it, which explains the fact that if we passed a current of hydrogen and nitrogen over the higher nitride, we at first obtained a quantity of ammonia corresponding to the quantity which the nitride would give with hydrogen alone, but that the formation of ammonia then ceased completely.

Thus far we had quite failed to get the nitrogen of the air into action.

With the third class of processes, however, based upon the formation in the first instance of cyanides, we found by our very first experiments that the nitrogen of the atmosphere can be easily led into combination.  A few experiments showed that the cyanide of barium was much more readily formed than any other cyanide; so we gave our full attention from this time to the process for obtaining ammonia by means of cyanide of barium invented by MM.  Margueritte and Sourdeval.  This process consists in heating a mixture of carbonate of barium with carbon in the presence of nitrogen, and subsequently treating the cyanide of barium produced with steam, thus producing ammonia and regenerating the carbonate of barium.  A great difficulty in this process is that the carbonate of barium fuses at high temperatures, and when fused attacks fireclay goods very powerfully.

We found that this can be overcome by mixing the carbonate of barium with a sufficient quantity of carbon and a small quantity of pitch, and that in this way balls can be made which will not fuse, so that they can be treated in a continuous apparatus in which the broken briquettes can be charged from the top, and after treatment can be withdrawn from the bottom.

We found that the formation of cyanides required a temperature of at least 1,200 deg.  C., and proceeded most readily at 1,400 deg.  C., temperatures which, although difficult to attain, are still quite within the range of practical working, and we found no difficulty in obtaining a product containing 30 per cent. of barium cyanide, corresponding to a conversion into cyanide of 40 per cent. of the barium present.