Chemical Examination of American Minerals, 23 



minute examination of this manganese, detected in it a por- 

 tion of magnesia. I did not attempt to determine the quan- 

 tity, as it would have been very difficult to have separated so 

 small a mixture of magnesia and oxide of manganese into 

 their several parts without error. 



7. 53-25 grains of the transparent fibres of the alum were 

 dissolved in water ; the alumina was thrown down by car- 

 bonate of an)monia, the residual liquid was evaporated to 

 dryness, and the salt remaining, exposed to a red heat till it 

 ceased to give out ammoniacal fumes. The remaining salt 

 weighed 6-33 grains. Being dissolved in water, and left to 

 pontaneous evaporation, it shot into well defined crystals 

 of glauber salt. 6-23 grains of anhydrous sulphate of soda 

 contain 2*77 grains of soda. 



The alumina which had been precipitated by carbonate of 

 ammonia after edulcoration and drying on the filter, dissolv- 

 ed with efi'ervescence in muriatic acid, indicating the presence 

 of something besides alumina. The alumina vtas separated 

 by caustic ammonia, the liquid freed from alumina was evapo- 

 rated to dryness, and the residual salt exposed to a heat gra- 

 dually raised to redness to drive off the ammoniacal fumes. 

 There remained two grains of common salt, equivalent to 

 I'OG grains of soda ; thus the whole soda extracted amounted 

 to 3*83 grains. A loss amounting to 0"J7 grain of soda 

 was obviously sustained. Some of this was probably owing 

 to the mode of driving ofT the ammoniacal salts, which was 

 too rapid ; and some was still in all probability adhering to 

 the alumina. But I did not think it worth while to prosecute 

 the investigation farther, the observations in paragraph 2, 

 leaving no doubt that the whole soda, had it been collected, 

 would have saturated 5 grains of sulphuric acid, and there- 

 fore, would have amounted to 4 grains. 



Thus the constituents of 53-25 grains of the soda-alum 

 were found to be, — 



