September 17, 1896] 



NA TURE 



479 



scale, I met with the most formidable difficulties, which it took 

 many years to overcome successfully. 



The very fact that ammonium chloride vapour forms so readily 

 metallic chlorides when brought in contact at an elevated 

 temperature with metals or oxides or even silicates, led to the 

 greatest difficulty, viz. that of constructing apparatus which 

 would not be readily destroyed by it. 



Amongst the metals we found that platinum and gold were 

 the only ones not attacked at all. Antimony was but little 

 attacked, and nickel resisted very well if not exposed to too 

 high a temperature, so that it could be, and is being, used for 

 such parts of the plant as are not directly exposed to heat. The 

 other parts of the apparatus coming in contact with the 

 ammonium chloride vapour I ultimately succeeded in constructing 

 of cast and wrought iron, lined with fire-bricks or Doulton tiles, 

 the joints between these being made by means of a cement 

 consisting of sulphate of baryta and waterglass. 



After means hail been devised for preventing the breaking of 

 the joints through the unequal expansion of the iron and the 

 earthenware, the plant so constructed has lasted very well. 



Oxide of nickel, which had proved the most suitable material 

 Jor the process in the laboratory, gave equally good chemical 

 results on the large scale, but occasionally a small quantity of 

 nickel chloride was volatilised through local over-heating, which, 

 however, was sufficient to gradually make up the chlorine 

 conduits. We therefore looked out for an active material free 

 from this objection. Theoretical considerations indicated mag- 

 nesia as the next best substance, but it was found that the 

 magnesium chloride formed was not anhydrous, but retained a cer- 

 tain amount of the steam formed by the reaction, which gave rise 

 to the formation of a considerable qu.antity of hydrochloric acid 

 on treatment with hot air. In conjunction with Dr. Eschellman 

 (who carried out the experiments for me), I succeeded in re- 

 ducing the quantity of this hydrochloric acid to a negligible 

 amount by adding to the magnesia a certain amount of chloride 

 of potassium, which probably has the effect of forming an 

 anhydrous double chloride. 



This mixture of magnesia and potassium chloride is, after the 

 addition of a certain quantity of china clay, made into small 

 pills in order to give a free and regular passage throughout their 

 entire mass to the hot air and other gases with which they have 

 to be treated. In order to avoid as far as possible the handling 

 and consequent breaking of these pills, I vapourise the ammonium 

 chldride in a special apparatus, and take the vapours through 

 these pills and subsequently pass hot air through, and then again 

 ammonium chloride vapour, and so on, without the pills changing 

 their place. 



The vapourisation of the ammonium chloride is carried out in 

 long cast-iron retorts lined with thin Doulton tiles, and placed 

 almost vertically in a furnace which is kept by producer gas at a 

 very steady and regular temperature. These retorts are kept 

 nearly full with ammonium chloride, so as to have as much 

 active heating surface as possible. From time to time a charge 

 of ammonium chloride is introduced through a hopper at the top 

 of these retorts, which is closed by a nickel plug. The ammonium 

 chloride used is very pure, being crystallised out from its solution 

 as produced in the ammonia soda manufacture by a process 

 patented by Mr. Gustav Jarmay, which consists in lowering the 

 temperature of these solutions considerably below o'C. by means 

 of refrigerating machinery. The retorts will, therefore, evaporate 

 a very large amount of ammonium chloride before it becomes 

 necessary to take out through a door at their bottom the non- 

 volatile impurities which accumulate in them. The ammonium 

 chloride vapour is taken from these retorts by cast-iron pipes 

 lined with tiles and placed in a brick channel, in which they are 

 kept hot, to prevent the solidification of the vapour, to large up- 

 right wrought-iron cylinders which are lined with a considerable 

 thickness of fire-bricks, and are filled with the magnesia pills, 

 which are, from the previous operations, left at a temperature of 

 about 300' C. On its passage through the pills the chlorine in 

 the vapours is completely retained by them, the ammonia and 

 water vapour formed pass on and are taken to a suitable con- 

 densing apparatus. The reaction of the ammonium chloride 

 vapour upon magnesia being exothermic, the temperature of 

 the pills rises during this operation, and no addition of heat is 

 neces.sary to complete it. The temperature, however, does not 

 rise sufficiently to satisfactorily complete the second operation, 

 viz. the liberation of the chlorine and the re-conversion of the 

 magnesium chloride into magnesium oxiile by means of air. 

 This reaction is slightly endo-therniic, and thus absorbs a small 



NO. 1403. VOL. 54] 



amount of heat, which has to be provided in one way or 

 another. I effect this by heating the pills to a somewhat higher 

 temperature than is required for the action of the air upon 

 them, viz. to 600 C. , by passing through them a current of a 

 dry inert gas free from oxygen heated by a Siemens-Cowper 

 stove to the required temperature. I use for this purpose the 

 gas leaving the carbonating plant of the ammonia soda process. 



This current of gas also carries out of the apparatus the small 

 amount of ammonia which was left in between the pills. It is 

 washed to absorb this ammonia, and after washing, this same 

 gas is passed again through the Siemens-Cowper stove, and thus 

 constantly circulated through the apparatus, taking up the heat 

 from the stove and transferring it to the pills. When these have 

 attained the required temperature, the hot inert gas is stopped 

 and a current of hot air passed through, which has also been 

 heated to 600° C. in a similar stove. The air acts rapidly upon 

 the magnesium chloride, and leaves the apparatus charged with 

 18 to 20 per cent, of chlorine and a small amount of hydro- 

 chloric acid. The chlorine comes gradually down, and when it 

 has reached about 3 per cent, the temperature of the air entering 

 the apparatus is lowered to 350° C. by the admixture of cold 

 air to the hot air from the stove ; and the weak chlorine leaving 

 the apparatus is passed through a second stove, in which its tem- 

 perature is raised again to 600° C. , and passed into another 

 cylinder full of pills which are just ready to receive the hot-air 

 current. A series of four cylinders is required to procure the 

 necessary continuity for the process. 



The chlorine gas is washed with a strong solution of chloride 

 of calcium, which completely retains all the hydrochloric acid, 

 and is then absorbed in an apparatus invented by Dr. Carl 

 Langer, by hydrate of lime, which is made to pass by a series of 

 interlocked transporting twin-screws in an opposite direction to 

 the current of gas, and produces very good and strong bleaching 

 powder, in spite of the varying strength of the chlorine gas. 

 The hydrochloric acid absorbed by the solution of calcium 

 chloride can by heating this solution be readily driven out and 

 collected. 



This process has now been in operation on a considerable 

 scale at our Works at Winnington for several years, with 

 constantly improving results, notably with regard to the loss of 

 ammonia, which has gradually been reduced to a small amount. 

 The process has fully attained my object, viz. to enable the 

 ammonia soda process to compete, not only in the production of 

 carbonate of soda, but also in the production of bleaching 

 powder, with the Le Blanc process. 



Nevertheless, I have hesitated to extend this process as 

 rapidly as I should otherwise have done, because very shortly 

 after I had overcome all its difficulties, entirely different methods 

 from those hitherto employed for the manufacture of chlorine 

 were actively pushed forward in different parts of the globe, for 

 which great advantages were claimed, but the real importance 

 and capabilities of which were and are up to this date very 

 difficult to judge. I refer to the processes for producing chlorine 

 by electrolysis. 



During the first decade of this century, Humphry Davy had 

 by innumerable experiments established all the leading facts 

 concerning the decomposing action of an electric current upon 

 chemical compounds. Amongst these he was the first to dis- 

 cover that solutions of alkaline chlorides, when submitted to 

 the action of a current, yield chlorine. His successor at the 

 Royal Institution, Michael Faraday, worked out and proved 

 the fundamental law of electrolysis, known to everybody as 

 "Faraday's Law," which has enabled us to calculate exactly 

 the amount of current required to produce by electrolysis any 

 definite quantity of chlorine. Naturally, since these two 

 eminent men had so clearly shown the way, numerous inventors 

 have endeavoured to work out processes based on these princi- 

 ples for the production of chlorine on a manufacturing scale, 

 but only during the last few years have these met with any 

 measure of success. 



It has taken all this time for the classical work of Faraday on 

 electro-magnetism to develop into the modern magneto-electric 

 machine, capable of producing electricity in sufficient quantity 

 to make it available for chemical operations on a large scale ; 

 for you must keep in mind that an electric installation sufficient 

 to light a large town wUl only produce a very moderate quantity 

 of chemicals. 



In applying electricity to the production of chlorine, various 

 ways have been followed, both as to the raw materials and as to 

 the ap]iaratus employed. While most inventors have proposed 



