Konig.] OlZ [Jan, 19, 



silence. To facilitate the comparison, the results of the several analyses 

 are placed side by .side : 



Excepting Pisani's analysis, there is such close agreement between the 

 others as could only be accounted for by an identical material and the fol- 

 lowing of precisely the same methods of analysis under the same circum- 

 stances. For, if Pisani's analysis is correct, there miist be a variation in 

 the composition of the Brevig mineral. Adding the percentages of Ti Oj, 

 Zr O., Al^ O3 in Pisani's analj'sis, we obtain 16.0G, against 1(5.48 in my 

 analysis, whilst those sums in the other analyses are 11.26, 12.30, 12.23 

 respectively. But in regard to manganese, it seems evident that the min- 

 eral from Brevig and that from Colorado differ considerably, being only 

 one third in the latter of that in the former. 



It was stated above, that the astrophyllite is closely associated with 

 zircon, which occurs in all sizes from | inch to microscopic size. After 

 the astrophyllite has been picked with all possible care, and is reduced to 

 l)Owder in a mortar, an occasional harsh grit is noticed. This suggests the 

 intimate admixture of a hard, granular mineral, probably zircon, and the 

 zirconia found by analysis might only be derived from admixed zircon. 

 To test this question I decomposed the mineral with sulphuric acid of 

 medium concentration, filtered from the insoluble, and digested the residue 

 with sodium hydrate. A white, sandy residue, of great hardness, was 

 obtained, ranging from 3 to 6 per cent, in difterent samples. I analj'sed it 

 by itself, and found it to be a mixture of quartz and zircon. Since zircon, 

 even as finest powder, is not acted upon sensibly by sul]ihuric acid of 

 medium concentration, and as I had avoided the crushing of the hard 

 grains as much as possible, I feel confident to assert that whatever zirconia 

 was found in the sulphuric solution was in molecular combination as 

 astrophyllite, and not derived from admixed zircon. By the same process 

 I made sure that no titanium was present in the form of rutile or titanite, 

 or any other titanium mineral. 



I jiroceed now with the description of a method for the qualitative and 

 quantitative determination of titanium, zirconium and aluminum, which 

 I believe to be new and especially serviceable when the quantity of the 

 mixed metals is very small. The best method for separating titanium from 

 aluminum is stated to be (Rose, quant, analysis ) the boiling of 



the dilate sulphuric .solution, when titanium dioxyhydrate will precipitate. 

 This is only relatively true, depending on the relative quantities of metals 

 and of salts in the solution. If the solution be strongly acid, titanium is 

 not completely precii)itated, and in certain cases not precipitiited at all. (A 

 detailed account of the behaviour of titanium, zirconium and aluminum 



