Composition of some Norwegian Minerals. 71 



the crystals sometimes are lengthened in this direction, as 

 shown by Fig. 4. The cleavage planes are very distinct, and 

 are parallel to the plane r ; and it was easy to cleave out pieces 

 of a rhombic section, the angle being about 138°, no third 

 cleavage was observed. 



The specific gravity at 60° Fahr. was found to be 5*53. In 

 analysing it I determined to follow the method employed by 

 Scheerer, and in consequence 22*78 grains were digested in 

 hydrochloric acid, by which it was readily decomposed. Much 

 water was then added, and the whole filtered from the silica ; 

 but I found that on attempting to precipitate the titanic acid 

 from this solution by boiling, as stated by Scheerer, no pre- 

 cipitation occurred ; the solution was therefore thrown down 

 by ammonia, washed, and re-dissolved in dilute sulphuric acid, 

 when a small quantity of silica remained undissolved, which 

 was filtered off, and much water then added to this solution, 

 and boiled when the titanic acid was readily thrown down and 

 weighed. On ignition it was found to be slightly tinged with 

 iron. 



In the filtrate from the ammoniacal precipitate, lime was 

 precipitated by oxalic acid ; the oxalate collected, ignited, and 

 dissolved in acetic acid, to separate some manganese which 

 was precipitated with it, and determined as sulphate. The 

 solution from this gave a precipitate of oxalate of lime. 

 And on evaporation of the filtrate a small quantity of inso- 

 luble matter remained which reacted for titanic acid before 

 blow-pipe, and was considered as such. 



The silica was treated with hydrofluoric acid, which left a 

 very small quantity of titanic acid undissolved. 



The solution from which the titanic acid had been separated 

 by boiling was now neutralized by ammonia, and precipitated 

 by oxalate of ammonia, and the yttria determined in this as 

 usual. 



The filtrate from this last precipitate was found to contain 

 iron, alumina, and glucina, which, after precipitation by hy- 

 drosulphuret of ammonia, were all determined in the usual 

 way ; the glucina being separated by carbonate of ammonia, 

 and the sesquioxide of iron freed from alumina by caustic 

 potash. The results obtained were : — 



