1X8 Af. Berxelim's Analysis of some Minerals, [Aug. 



dines slightly to yellow. With carbonate of soda, the assay 

 decomposes without fusing, and the mass becomes greyish redf, 

 A larger quantity of the salt gives an imperfect scoria. If we 

 add borax to the assay, we obtain some traces of reduction, 

 but it is commonly manifested only by metallic striaj, when tri- 

 turated in the mortar. 



The analysis of polymignite gave 



Titanic acid 46*3 



Zirconia 14*4 



Oxide of iron 12-2 



Lime 4*2 



Oxide of manganese 2*7 



Oxide of cerium 6*0 



Yttria 11-5 



Magnesia ^ 



Potassa l J. c 



Silica r ^'■'"^''•'^ 



Oxide of tinj 



Loss 3-7 



lOO'O 



The real loss is still greater than that given above, for the 

 iron and manganese, and perhaps also the cerium, exist in the 

 mineral in the state of oxidules, whereas, in the results, they 

 are taken as oxides. The analysis of so complicated a mineraJ 

 must always be attended with very considerable loss, and it may 

 even contain some substances that have escaped detection. It 

 is evident, therefore, that the composition of a mineral like the 

 polymignite cannot be calculated. 



" I have often attempted,'' adds M. Berzelius, " to separate 

 titanic acid and zirconia from each other, but have not been 

 able to discover an infallible method for the purpose. Diluted 

 sulphuric acid separates them best, but it still dissolves a small 

 portion of titanic acid, as well as the zirconia. The carbonated 

 alkalies dissolve them alike, and nearly in the same proportions. 

 Sulphate of potash, which often does not precipitate titanium 

 from its solutions, throws it down, however, if they contain zir- 

 conia, which, in that case, carries down the titanic acid. Fluoric 

 acid acts nearly in the same manner. Infusion of galls precipi- 

 tates both titanic acid and zirconia. The analysis of polymignite 

 might be effected with sufficient accuracy, if we had a method 

 of separating these two substances ; but it also contains two 

 others which cannot be separated ; namely, yttria, oxidulous 

 manganese, the latter of which, in certain proportion, adheres 

 obstinately to the former. The best mode of separating them 

 that I know, is to dissolve them in nitric acid, evaporate the 

 solution to dryness, and keep the salts for a long tune at the 



