1822.] Weight of an Atom of Alumina. 167 



that by carbonate of soda into the second pair, and that by car- 

 bonate of ammonia into the third. The alumina on each pair 

 was edulcorated by distilled water till the water which passed 

 through ceased to produce any effect on muriate of barytes. 

 The filters were then allowed to dry in the open air. When as 

 dry as they could be made in this way, the two filters constitut- 

 ing each pair were separated from each other. The outermost 

 was put into one of the scales of the balance, and the innermost 

 still containing the whole alumina was put into the other. As 

 the two filters were exactly of the same weight, it was easy to 

 determine the exact weight of the alumina. A portion of the 

 alumina thus weighed was now detached from the filter, and 

 exposed to a strong red heat in a platinum crucible ; and from 

 the loss of weight which it sustained, it was easy to deduce the 

 loss of weight which the whole alumina would have sustained 

 had it been subjected to the same process. The following are 

 the results of these experiments : 



(1.) The alumina precipitated by the carbonate of potash 

 weighed 24*59 grains. When heated to redness, it was reduced 

 to tO-988 grains. 



(2.) The alumina precipitated by the carbonate of soda 

 T;\'eighed 24*34 grains. When heated to redness, it was reduced 

 to 10*82 grains. ' 



(3.) The alumina precipitated by the carbonate of ammonia 

 weighed 30*44 grains. When heated to redness, it was reduced 

 to 11*39 grains. It is right to mention that the slight excess of 

 ammonia had dissolved a minute portion of the alumina. This 

 was obtained by evaporation, and is included in the preceding 

 quantity. The liquid precipitated by the carbonates of potash 

 and soda afforded no traces of alumina. 



The mean of these three experiments gives 11*066 grains for 

 the alumina contained in 100 grains of alum. 



These three experiments were repeated with verv nearly the 

 same result; the only difference was, that the alumina precipi- 

 tated by carbonate of ammonia weighed 11*48 grains instead of 

 11*39 grains, which was the quantity obtained the first time. 

 This makes the mean of the three last experiments 11*096 grs. 

 ^. -The mean of all the six experiments (which must come very 

 near the truth) is 11*081 grains of alumina in 100 grains of alum. 

 Hence (i0*875 grains of alum must contain 6*74556875 grains of 

 alumina. This quantity of alumina was in combination with 15 

 grains of sulphuric acid, which is equivalent to three atoms of 

 that acid. Now sulphate of alumina, as I have ascertained by- 

 experiment, is a white salt, having an acid and astringent taste 

 similar to alum, and reddening vegetable blues as powerfully 

 as that salt. It does not owe this property to any excess of 

 sulphuric acid ; for it retains it after repeated digestions in 

 alcohol. Hence there can be no doubt that it is a compoun<l of 

 an atom of acid and an atom of alumina. t 



