500 REPORT — 1882. 



by means of a solution of AgNOj, containing per gram of solution, 0-0009 grm. Ag, 

 The mean of seven experiments conducted in this way gave him the ratio : 

 Ag : RbCl :: 107-93 : 120-801 ; this when chlorine = 35 467 gives for the atomic 

 weight of rubidium, 85-344. 



The portion |3 was converted into bromide by means of pure HBr. prepared as 

 described in Mr. Huntington's paper on the atomic weight of cadmium (' Proc. 

 Amer. Acad.') it had a constant boiling point of 126° 0. Bar. 777 mm. after frac- 

 tionating six times. 



The first results were all far too high, the RbBr being contaminated with 

 platinum derived from the dish. The last two results when the weights are reduced 

 to a vacuum are not nearly so concordant as might be wished. 



EbBr 



4-2.'565 grms. gave Ag. : RbBr : : 107-93 : 165437 

 12-33595 „ „ Ag. : RbBr : : 10793 : 165-342 

 whence mean atomic weight as derived from 



RbBr when Br = 79955, 85-389. 

 These bromide determinations are given with all reserve, as they are influenced 

 to a large extent by the correction for displacement which there is reason to believe 

 is far too high. 



The extended experiments will be published shortly. The mean error of the 

 atomic weight deduced from the chloride and bromide by the method of least 

 squares is + 0-0215. At present rubidium cannot be said to conform to Proufs 

 hypothesis. 



7. On a method of obtaining Ammonia from Shoddy and Allied Substances. 



By W. Mareiott, F.G.S. 



The author has for some time been engaged in devising a method of applying 

 on a large scale tlie principle adopted in any ordinary nitrogenous estimation by 

 means of soda lime, his object being to work up such nitrogenous substances as 

 wool or hair, and convert the nitrogen into ammonia. 



Mixing shoddy (the name by which this material is known) with lime and 

 soda, and subjecting the mixture to heat in a retort, has been tried many times 

 and with different apparatuses, but the return in ammonia never paid the cost of 

 working. 



Shoddy is a bulky material, so that a retort when charged with it holds but 

 little weight. And if the size and diameter of the retort are increased to overcome 

 this, the shoddy when carbonised is such a bad conductor of heat that the matter in 

 the centre of the retort is but slowly decomposed, although a great expense is in- 

 curred for fuel. 



To obviate this, and reduce the cost of fuel, he tried the burning of the shoddy 

 in a cupola heated by combustible gases. This, after several failures, he has suc- 

 ceeded in accomplishing, and also in making the process a continuous one. 



The principle he applies for this purpose, is to generate the heat of combustion 

 in contact with the nitrogenous matters operated upon, making sure that no free 

 oxygen is left with the products of the combustion. 



The construction of the apparatus may be thus explained. 



Nos. 1 and 2 are cupolas made of boiler-plate, and lined with fire-brick. In 

 No. 1 cupola, coke and small coal is burnt, the combustion being kept up by means 

 of a fan or air-pump along with high-pressure steam. The gases generated are 

 principally carbonic oxide, hydrogen, and marsh gas, along with the nitrogen of the 

 air blown in. These mixed gases are led into the bottom of cupola No. 2, and 

 there coming into contact with a jet of air, combustion ensues, and great heat is 

 generated, -which heat has to pass up through the material in the cupola. 



The point to be strictly observed is, that no more air be admitted than is 

 required for the combustion of the gases, and no free oxygen left in the atmo- 

 sphere heating the organic matters. The ammonia generated, along with the 

 residual gases, are passed through a saturator containing sulphuric acid to absorb 



