62 CHEMICAL FERTILIZERS 



by a gas-bag arrangement or by a Cottrell electric dust 

 precipitator ; the latter alone is capable of dealing with 

 large amounts of gas. The composition of the dust obtained 

 varies from place to place, a very rich dust showing on 

 analysis 20-25 % of potassium chloride, 15 % of potassium 

 hydrogen carbonate, 5 % of potassium cyanide and small 

 traces of other substances. On the other hand some dusts 

 only contain i % of potash. Attempts to raise the amount 

 of potash in the flue dust have been made in a variety 

 of ways, but the most promising of these consists in the 

 addition of a small quantity of common salt, which produces 

 increased volatilization of potassium ; very considerable 

 improvements have been made in this way. Experiments 

 made by adding a quantity of salt sufficient to provide 

 chlorine to combine with all the potassium in the blast 

 furnace caused the amount of potash volatilized to be 

 increased three or four times. 



The cement industry provides a similar source of potash 

 compounds, but suffers from the difficulty that the cost of 

 collection must be debited exclusively to the potash industry, 

 whereas in the blast-furnace industry the gases must be 

 cleaned and their heat utilized. By the use of salt in the 

 cement kiln, potash in considerable quantities is volatilized 

 into the waste gases. The great difficulty in dealing with 

 this material is the engineering one of providing sufficient 

 plant for the removal of potash fog. The most promising 

 solution of this obstacle is the Cottrell Electrostatic method of 

 recovery (see p. 65) . To deal with these materials properly, 

 it would be necessary to erect central potash factories in 

 certain districts. Now that the subject has come under 

 investigation, it is interesting to note the fact that from iron 

 blast furnaces the potash obtained appears to be greater than 

 is accounted for in the charges. The cause of this dis- 

 crepancy may be that the ordinary methods of estimating 

 potash in these compounds are a little faulty, owing to the 

 fact that in the laboratory some potassium volatilizes during 

 the combustion of the fuel necessary to obtain an ash suited 

 to the subsequent method of analysis. The discrepancy 



