162 Dr, Thomson on the [March, 



we have a good many analyses of alum. But the one which 

 appears to have been nmde with the greatest care, and which 

 approaclies nearest the truth, is that of BerzeUus, first given to 

 the pubhc in the Annales de Chimie, vol, Ixxxii. p. 258. The 

 result of this analysis is as follows : 



Sulphuric acid 34*23 



Alumina 10*86 



Potash 9*81 



Water 45*00 



99*90 



Upon this analysis, it may be requisite to make a few remarks. 



(1.) His mode of determining the sulphuric acid was to dis- 

 solve 100 grains of alum in water, and precipitate the sulphuric 

 acid by means of muriate of barytes. The sulphate of barytes 

 obtained weighed exactly 99*765 grains. Now as sulphate of 

 barytes is composed of 5 sulphuric acid -f 9*75 baryte.% it is 

 obvious that 99*765 grains of the salt contain only 33*82 grains 

 of sulphuric acid instead of 34*23 grains — the quantity stated 

 by Berzelius ; so that Berzelius overrates the acid, as found by 

 his experiment, by about two-fifths of a orain. I shall show 

 hereafter that the real quantity of sulphuric acid in 100 grains 

 of alum is 32*854 grains, or nearly one grain less than the qaan- 

 tity indicated by Berzelius's experiment. 



ii (2.) The alumina was obtained by dissolving 100 grains of 

 Jtlura in water, and precipitating it by ammonia in considerable 

 excess. • The precipitate was washed and dried in a strong red 

 heat. The alumina thus obtained in one experiment weighed 

 10*86 grains ; in another 10*67 grains. This is very nearly the 

 mode which I employed. I would remark only that an excess 

 of ammonia is not necessary. If you add simply the quantity 

 required to saturate the sulphuric acid united with the alumina, 

 the whole alumina will be precipitated. The advantage of this 

 method is, that little or no alumina will be dissolved by the 

 excess of ammonia. Berzelius indeed recovered this portion by 

 evaporating the ammoniacal liquid to dryness ; but the alumina 

 in this case is apt to be carried off with the liquid as it evapo- 

 rates. This I suspect to be the reason of the small deficiency of 

 alumina in Berzelius's experiments. This deficiency in one 

 case was about one-fifth grain; and in the other, two-fifths of a 

 grain. These quantities are indeed very small ; but they have a 

 sensible effect in altering the atomic weight of alumina ; for 

 even the smallest of them amounts to nearly two per cent, of the 

 whole weight of the alumina. 



(3.) His mode of obtaining the potash was to digest 100 parts 

 of alum in a phial with carbonate of strontian and water till the 

 whole sulphate of alumina was decomposed and precipitated. 

 The filtered liquid was evaporated to dryness in a platinum. 



