В — О— Si о 



The percentage of niobic acid is a little too high; it was 

 however not quite pure, but contained some amount of 

 tantalic acid, which has not been quantitatively determined. As 

 the atomic weight of tantalum is considerably higher than 

 that of niobium, it follows that, if the tлvo acids had been 

 separately determined, the ratios for them would have been less 

 than those now found. Also the ratio for H^ and jP is a little 

 too high. This may possibly be due to a somewhat too high 

 result of the water determination. 



Chalcolamprite thus consists of equivalent amounts of 

 the substances BNb^^O^F.^ and ESiO.^. Common pyrochlore 

 consists of the same niobium compound and , as it seems, 

 varying amounts of ÈTiO.^, where Ti may be replaced by Т/г, 

 Zr etc. On account of varying results of the analyses it has 

 not as yet been possible to establish a generally acceptable 

 formula for pyrochlore. Ramme Isb erg M has come to the 

 result that pyrochlore consists of isomorphous mixtures of 

 NaF, Èm,_0^, È.,m,,0^, and UTiO., or RTiO.,, etc. 

 Apart from the improbability of such heterogeneous combina- 

 tions replacing one another isomorphously, this requires that 

 Na and F should be present in equivalent quantities, which 

 is not always the case in pyrochlore nor ever in chalco- 

 lamprite. 



As a result of his synthetic investigations H olm q v is t '"^j 

 finds the formula NaCaNb.^Of^F for the pyrochlore-minerals. 



1) Monatsber. Berliner Akad. 1871, p. 203. 



2) Inaugural Diss. Bull, of Geol. Inst. Upsala, 1896. Vol. Ill, p. 1. 



