FIXATION OF ATMOSPHERIC NITROGEN 399 



ture by electricity. The favourable temperature is said to lie 

 between 1800 and 1900 C., a higher temperature causing 

 decomposition of the preformed nitride. The presence of some 

 metallic oxides is said to be favourable to the formation of the 

 nitride, and therefore low-grade bauxite can be used, and in 

 preference to pure alumina. The gas employed is producer gas, 

 which consists roughly of one-third of its volume of carbonic 

 oxide with two- thirds of its volume of nitrogen. The simplest 

 expression for the chemical change is shown by the following 

 equation : 



A1 2 3 +3C+N 2 =2A1N+3CO, 



but there seems to be some difference of opinion as to the manner 

 in which the change is effected and the composition of the 

 nitride. The aluminium nitride can be decomposed by water or 

 solution of caustic soda : 



A1N+3H 2 0=A1(OH) 3 +NH 3 . 



When alkali is used the alumina is dissolved out leaving the 

 iron behind, and a very pure alumina may thus be separated 

 specially suitable for use in the electrolytic process by which the 

 metal is usually obtained. The aluminium nitride may also be 

 decomposed by a limited quantity of acid, sufficient to fix the 

 ammonia, while leaving the alumina insoluble. If sulphuric acid 

 is used soluble ammonium sulphate is formed as follows : 



2A1N+H 2 S0 4 +6H 2 0=2A1(OH 3 ) + (NH 4 ) 2 S0 4 . 



The formation of cyanides when carbon and nitrogen are 

 heated together to a high temperature in contact with hydrogen 

 or water vapour or with alkaline salts or silicates has long been 

 familiar. Thus it is known that hydrocyanic acid is formed 

 around the electric arc taken in ordinary moist air, and Bunsen 

 and Play fair in 1845 found cyanogen in the gases from a blast 

 furnace at Alfreton. The white incrustation which is often seen 

 at the joints of the iron furnaces consists chiefly of potassium 

 cyanide, and at one time it was supposed that this salt played an 

 important part in the reduction of iron ores. The alkali metal is 

 obviously derived from the ash of the coal and from a small 

 quantity of alkaline salt contained in the ore. Upon the recog- 

 nition of these facts attempts were made to establish a process 

 for the synthetical production of cyanides by passing nitrogen 

 over a strongly heated mixture of coal and potash. The high 



