169 



10 "/0. In tlie clialcolamprite the loss could be ascerlained to 

 be -SVOo. Of the endeioHte there was no more material 

 available. Therefore it could not be determined what the loss 

 in this mineral is. However, the similarity of the two minerals 

 in other respects, as well as the way in which the analysis 

 was carried on, aflord good reasons for assuming that also in 

 endeiolite the missing constituent is SiO.,. If this is assumed, 

 the result of the analysis is as follows. 











Molec 



'ular ratios 





Ab, 0, . . . 



. 59,93 



0,226 



1,24 



SiO, . 

 TiO, . 







. (11,48) 

 0,76 



0,190 ] 



} 0,199 

 0,009 j ' 



1,10 



ZrO., . 







3,78 



0,062 







Ce,0, 







4,43 



0,039 







Fe, 0, 







2,81 



0,054 







MaO . 







0,37 



0,005 



' 0,362 



2 



CuO. . 







7,89 



0,140 







K,0 . 







. 0,43 



0,004 







i\a, . 







3,58 



0,058 





*■ 



H,0 . 

 F,. . . 







. 4,14 

 0,69 



0.233 ) 



f 0,249 

 0,016 j ' 



1,36 





100,29 





-0 . . 







. 0,29 









100 



The molecular ratios of the niobic acid , the assumed 

 silicic acid, all the bases, and the water and fluorine are thus 

 approximately 1:1:2:1. The empirical formula for endeiolite 

 is consequently 



ÊNb,0,(HO),-^ÊSlO.^. 



Analogously to the structural formula of chalcolamprite 

 that of endeiolite may be written as follows: 



