The following estimate of the cost of producing 1 tou of ammonium sulphate by the steam process is given by a manufacturer of large experience :— s. d.
Gas liquor (average) .. 11 0 0 Coal (20 cwt. at 18s. 6d.) 0 18 Linte (5 cwt.) .. 0 4 0 Acid (20 cwt.) .. .. 2 5 0 Packages 0 5 0 Management, rent, &c. 0 5 0 - £15 4 2 The coal hero estimated for is largo and of good quality, such being found to pay best in the long run, except where slack is readily obtainable at a very low cost. The quantity of 20 cwt. may be considered excessive, many manufacturers being of opinion that it is possible to work with 15 cwt.
Fresenius has published the following method for preparing a solution of pure ammonia, which is very convenient for working on a small scale. Six and a half kilogrammes of ordinary commercial crystallized chloride of ammonium, and 3.5 kilogrammes of sulphate are mixed together and passed through a sieve so as to break up any large masses of the salts. The mixture is placed in the retort in layers, with intervening layers of hydrate of lime prepared by moistening 10 kilogrammes of lime with 4 of water. The whole mass is then moistened with 8 kilogrammca of water. The gas is first washed in an iron Woulfe's bottle, attached to which is a glass tube for showing the height of its contents; to one of the tubulures is fixed a glass balloon which the gas traverses before entering the refrigerator. The receiver, which may conveniently be a large carboy, contains 21 kilogrammes of distilled water. The wash bottle is supplied with a top so as to remove a portion of the liquid which passes over from the retort. The use of the chloride and sulphate possesses the advantage that the residuum is easily removed from the retort, the sulphate of lime which is formed prevents the chloride of calcium which is also formed from producing a mass, and less water is also required when operating with the mixture than with the sulphate only. After 5 or 6 hours the greater part of the ammonia is distilled, the gas which is still disengaged may be collected in another receiver with more water than is sufficient to dissolve it completely. At this period of the operation the upper part of the wash bottle is filled with
white vapours. These vapours indicate the drying of the contents of the retort and the decom position of organic matters. Ou the appearance of the vapours the receiver must be taken away and replaced by •a smaller one. This product becomes milky and is reserved for a future operation. The great point in this is, that the residue is more easily managed in its removal from the retort. If it was necessary to recover the chloride of calcium from this waste, it could be easily dissolved out with warm water. Carbonate of ammonia, chlorine, lime, copper or lead, and a little organic matter, containing carbon, are its principal impurities. A sample when made may be free from carbonic acid, but contain carbonate of ammonia from absorption of carbonic acid from the air ; the contamination with copper or lead may be avoided by using iron vessels and connecting pipes. The carbonaceous matter may arise from the salt employed, or the lutes used in making the joints. When pure, it is colourless and leaves no residuum on evaporation ; by heat in open vessels it rapidly loses its gas, consequently it should be kept in well stoppered green glass bottles in a cool place. An alcoholic solution of ammonia may be prepared by passing the gas into alcohol containing 85 per cent. to 90 per cent. instead of water.
It is a common practice to determine the strength of an ammoniacal solution by reference to its specific gravity ; it is obvious that if the solution contain any soluble matter other than the gas itself, its specific gravity cannot be taken as representing the strength or degree of saturation with the gas ; it is therefore desirable to ascertain the absence of fixed impurities by evaporating a known quantity on the water-bath. Should no residuum be ]eft, the quantity of gas present may be accurately found by reference to a table of specific gravities.
The following table given by Otto is regarded as b,.ing reliable, the determinations being made at 16°.