16 On certain Slates, Felsites, and Elvanites near Knockmahon. 



alumina were precipitated by ammonia ; ancTafter being washed 

 and ignited, their united weight was noted. 



The separation of peroxide of iron from alumina was effected 

 by dissolving the mixture in dilute nitro-hydrochloric acid, and 

 precipitated by an excess of pure caustic potassa. The peroxide 

 of iron thus obtained, after being redissolved and precipitated by 

 ammonia, was washed and weighed. The alumina was obtained 

 by acidifying the alkaline solution and precipitating by ammonia. 



A small quantity of silica was frequently obtained on attacking 

 the mixture of peroxide of iron and alumina, and its weight was 

 added to that previously found. 



In order to determine the respective amounts of protoxide 

 and peroxide of iron, a separate fusion was made on a weighed 

 quantity of the pulverized rock and sesquicarbonate of soda, care 

 being taken to cover the surface of the mixture with a layer of 

 the pure sesquicarbonate. 



The contents of the crucible were thus both fused and allowed 

 to cool in an atmosphere of carbonic acid, besides being preserved 

 from oxidation by a covering of carbonate of soda. 



The crucible and its contents were now placed in a covered 

 beaker containing dilute hydrochloric acid, and heated on the 

 sand-bath ; and as soon as carbonic acid ceased to be evolved, the 

 solution was allowed to cool, without removing the cover, and 

 the quantity of protoxide of iron present estimated by a freshly 

 standardized solution of permanganate of potassa. 



The protoxide thus found was subsequently converted, by cal- 

 culation, into sesquioxide, and deducted from the amount ob- 

 tained by direct estimation. The lime was precipitated in the 

 filtrate separated from the mixture of oxide of iron and alumina 

 by the addition of oxalate of ammonia ; and to the filtrate from 

 the lime-salt phosphate of soda was added for the precipitation of 

 magnesia. 



The percentage of the alkalies was determined by fusion with 

 caustic baryta, dissolving in hydrochloric acid, and evaporating to 

 dryness. Baryta-water and carbonate of ammonia were subse- 

 quently added, and the weight of mixed chlorides obtained. To 

 the solution of these, bichloride of platinum was added, and the 

 weight of potassa determined in the usual way. The amount of 

 soda was ascertained by deducting from the weight of mixed 

 chlorides the chloride of potassium corresponding to the potassa 

 found, and calculating the quantity of soda corresponding to 

 the remainder. 



