OF THE ATOMIC WEIGHT OF ALUMINUM. 
1015 
phate prepared from the same sulphuric acid neutralised with ammonia. The quan¬ 
tities were determined by bringing the two solutions to known bulks, ascertaining 
by experiment on a sample of each how much of the respective salts was present, 
and measuring off the required volumes to be mixed. After concentrating the mixed 
solution by evaporation it was allowed to crystallise by cooling, and the crystallisation 
was repeated thrice, each time washing with a little cold water. On the last crystal¬ 
lisation pains were taken to regulate the rate of cooling so that as far as possible 
uniformly small crystalline grains were formed of about a millimetre in diameter, thus 
avoiding the liability of large crystals to contain cavities, in which mother liquor 
might be retained, and on the other hand securing the possibility of seeing with a 
lens, better than could have been done if the alum were in a still finer crystalline 
powder, that all the crystals as afterwards used were, clear and transparent, and 
showed no signs of efflorescence. 
It should be added that all the aqueous ammonia used as above for the precipitation 
of aluminum hydrate, and for its reconversion into alum, was recently and carefully 
prepared from ammonium chloride purified from alcoholic amines by boiling with nitric 
acid as recommended by Stas,'”' that the last crystallisations were effected from water 
which had been, in pursuance of the practice of the same chemist, distilled from 
potassium permanganate and hydrate, and that for these last crystallisations a large 
platinum dish was used, and care was taken that the solution was not allowed to boil, 
nor even to remain for any length of time near the boiling point, since I ascertained 
that ammonium alum, like simple ammonium sulphate, gradually gives off small 
quantities of ammonia on continued boiling of a strong solution. The very last 
crystallisation was carried out with only a sufficient quantity of the alum for a couple 
of the final experiments. 
The salt thus purified was found to be free from any ascertainable content of 
foreign substances. It gave no trace of coloration in its solution Avhen tested by a 
ferro-cyanide, tannic acid, &c., and by sulphuretted hydrogen and ammonium sulphide ; 
and spectroscopic examination showed that the fixed alkaline metals and calcium were 
absent. Silver nitrate gave no indication of chlorine. 
Ignition of ammonium alum.—Difficult ies connected with this method .—1 proposed to 
ignite a weighed quantity of this alum, whose distinct crystallisation gives it the 
advantage as to definiteness in the amount of water over the simple sulphate used by 
Berzelius, and to determine the weight of the aluminum oxide left behind, but careful 
examination of this process shoAved that two difficulties were to be feared. 
In the first place, having rapidly dried the product of the last crystallisation by 
gentle pressure between folds of smooth filtering papert free from loose fibre, portions 
were weighed off and exposed to the air at about 22° C., with the hope that before 
long a constant “ air-dried ” weight would be obtained. It had been previously ascer- 
*' Quoted in Fresenius’ ‘ Zeitschrift fur analyt. Cliemief 6 ter Jalirg., 4 tes Heft., S, 423. 
f This had been previously purified by ample washing with acid and water, and well dried. 
