TRANSACTIONS OF SECTION B. 639 



over it, no chlorine is evolved at this temperature, but residual ammonia and 

 moisture may be completely driven out of the retort on stirring the mixture. If 

 the temperature be now raised to 350° C, the chlorine begins to be slowly evolved 

 on the passage of dry air, when the oxide used has been manganese ; and the per- 

 centage yield of chlorine by the peroxides of manganese is about one and a half time 

 that of any monoxide. Sesquioxides (M^Og) are unsuitable for this purpose, as 

 they yield one-third less ammonia, the other third being decomposed into nitrogen 

 and hydrogen. 



The method of volatilisation of the ammonic chloride means a high tempera- 

 ture, which means again a loss of ammonia and other disadvantages. By the 

 method of mixtures a low temperature is employed, the ammonia saved, the 

 chlorine fully evolved, and atmospheric air may be used to expel residual ammonia 

 and steam remaining in the retort ; and if the temperature be kept below 350° C. 

 air may be passing through the retort during the whole operation for the ammonia, 

 thus removing the ammonia and steam as formed, and preventing loss of ammonia 

 and the formation of hydrochloric acid when chlorine is required. 



Oxides of the heavy metals, and the alkaline earths and salts formed from these, 

 may be used for this purpose with more or less advantage. There is one oxide, 

 however, which appears superior to all others, viz., the oxides of manganese which 

 are not sesquioxide. These oxides (MnjOj and MnOo) evolve ammonia at a low 

 temperature. From 130° C. to 200° C. most of the ammonia is evolved, and at 325° C. 

 is completely evolved by these oxides. The percentage yield of chlorine by these 

 oxides is one and a half time that of the monoxide of nickel and others. For these 

 and other reasons I think they will hold their place in the new process as they have 

 in the old process of Leblanc. 



Peroxides of the composition M3O4, MOo may be used with advantage both for 

 the evolution of the ammonia, and afterwards for the chlorine on the passage of air, 

 and for the evolution of the hydrochloric acid on the passage of steam at the low 

 temperature employed. 



The workman would keep the temperature below 350° C. during the evolution 

 of the ammonia, which may be done by a mercurial thermometer placed in the 

 door of the retort. Other heavy oxides, salts, &c., require a much higher tempera- 

 ture for the evolution of the ammonia and chlorme. 



When all the chlorine or hydrochloric acid is f>volved, the residue is equivalent 

 -when reduced to the oxide with which we started, and will give the equivalent of 

 ammonia as at iirst when heated with ammonic chloride, and from this point the 

 round of operations goes on continually. 



The apparatus required would be a Stourbridge fire-clay retort, or other 

 suitable furnace or retort, heated by gas or preferably by a self-feeding coke furnace. 

 These retorts may be worked in series and connected with pipes leading to the 

 ammonia-absorbing plant, the hydrochloric condensing chambers, and the bleaching 

 powder chambers. They should also be connected with an exhauster or blower for 

 the passage through them of air or steam when chlorine or hydrochloric acid was 

 required. 



The residual oxides left after the complete evolution of the chlorine or hydro- 

 chloric acid would need washing at intervals from common salt and when man- 

 ganese has been the oxide used ; they exist chiefly as sesquioxide of manganese, 

 and before the charge of ammonic chloride is added they must be reduced. This 

 may be done by heating them with charcoal or small coal, or, better, by passing 

 the gases from the coke or other furnace used over them, when they will no longer 

 decompose the ammonia when used for the next round of operations. 



The lime now used is thus saved, the ammonia and chlorine fully recovered, 

 and reagents for this purpose recovered for use over and over again continually ; 

 and this at no other cost than the labour and fuel employed (which is small) and 

 the wear and tear of retorts. 



By this development of the ammonia-soda process the old process of Leblanc 

 has lost its last remaining prop. 



