36 



MEMOIR ON EMERY. 



izeci quartz; in one word, I pursued the same method as that 

 for estimating the water in emery. For the other ingredients 

 a new portion was taken, pulverized finely, and attacked either 

 by concentrated sulphuric acid or melted with carbonate of soda, 

 and afterward dissolved in hydrochloric acid with the addition 

 of a little nitric acid evaporated to dryness, and treated with 

 dilute hydrochloric acid. The liquid separated from the silica 

 is treated with an excess of caustic soda, and the filtered liquid 

 is neutralized by hydrochloric acid and the alumina precipitated 

 by carbonate of ammonia. 



The contents of the filter, which are essentially peroxide of 

 iron, are placed in a capsule, dissolved by hydrochloric acid, 

 heated and precipitated by ammonia, and thrown on a filter. 

 From the filtered solution the lime and magnesia are separated 

 in the ordinary way. The peroxide of iron remaining on the 

 filter, after being well washed and dried, is weighed and decom- 

 posed in a current of hydrogen gas. To the oxide thus reduced 

 nitric acid diluted with thirty times its weight of water is added, 

 and digested at 100° to 120° C. for about an hour, stirring fre- 

 quently, when, if the iron has been thoroughly reduced, it will 

 be taken up by the acid, and a little alumina left, which is 

 weighed and added to the first portion. Ordinarily I never 

 have found more than from one to two per cent, of alumina 

 with the oxide of iron. Care must be taken to decompose the 

 iron completely, as otherwise the iron will not be entirely taken 

 up by the acid. The mineral thus analyzed afforded as follows : 



Decomposed by 

 sulphuric acid 



Decomposed by \ 

 carb. soda J 



Decomposed by \ 

 carb. soda ( 



GO 



5" 



P 



> 

 | 



o 

 : . o X 



3 

 8- 



CD 



3 



3Q 

 B 

 CD 



9=' 



ft 



: a> 



B^ 



cd p 



: CTQ 

 : ? 



24.10 



39.8 



27.55 



6.50 



not esti- 

 mated. 



not esti- 

 mated. 



not esti- 

 mated. 



not esti- 

 mated. 



23.94 39.52 



28.05 



7.08 



0.45 



0.80 



trace. 



0.52 



23.20 



40.21 



27.25 



6.97 



0.83 



0.95 



trace. 



not esti- 

 mated. 



GOO 



0.30 



These analyses correspond to the following composition 



Silica 2 



Alumina 3 



Protoxide of iron 3 



Water 3 



The most probable formula 



AT. WEIGHT. 



PER CEN' 



113340 



23.87 



1925.88 



40.57 



1350.00 



28.44 



337.50 



7.12 



£l 3 Si+Fe 3 



Si+3fi. 



