﻿384 Penfield and Harper — Composition of Palstonite. 



Assuming, as it seems fair to do, that our results and con- 

 clusions are correct and that the formula which we have pro- 

 posed is the true one, namely, that the mineral is an isomor- 

 phous mixture of (MgNa 2 )Al' s F n , 2H 2 and (MgE"a 2 )A] 3 (OH) n , 

 2H 2 0, in which formulas fluorine and hydroxyl play the same 

 part or are isomorphous, let us see if we can in any way account 

 for the variations in the previously published analyses, espe- 

 cially between Brandl's and our own, the only two complete 

 analyses. First, we would emphasize that the greatest care was 

 used in preparing the material for our analysis ; the extremes 

 in the specific gravity of the powder which we separated were 

 2*611 and 2*551, or a variation between the lightest and heavi- 

 est of only 0*060. Second, our analysis shows that our material 

 is practically free from calcium, indicating a very complete sep- 

 aration from thomsenolite with which the ralstonite is so inti- 

 mately associated, and showing that calcium is not an essential 

 constituent of the mineral. Third, we were not limited regard- 

 ing the amount of material which we could use, as we had an 

 abundant supply of the pure mineral. From the same specimen 

 from which our material was derived, one of us by very careful 

 picking was able to obtain nearly one gram of octahedral crys- 

 tals, which he supposed were pure but which, as is shown in 

 the analysis near the beginning of this article, contained 1*67 

 per cent of calcium, showing that, a most careful and laborious 

 hand-picking had not been sufficient to free the small crystals 

 wholly from thomsenolite from which the calcium was unques- 

 tionably derived. It seems highly probable that other' investi- 

 gators have worked with material containing slight quantities 

 of thomsenolite. Grroth,* for instance, states that the material 

 which he furnished to Brandl for analysis, showed under the 

 polarizing microscope particles of a strongly double refracting 

 mineral with quadratic habit which was unquestionably thom- 

 senolite. If we assume that the mineral is free from calcium as 

 our analysis indicates and that the calcium in the other analyses 

 is all derived from thomsenolite, we should find by calculation 

 the following figures giving the per cent of thomsenolite in the 

 analyzed material and of the mineral free from thomsenolite. 



Calculated from 

 Nordenskiold. Penfield. Brandl. (MgNa 2 )Al 3 Fn.2H 2 



Thomsenolite 11*07 9*28 8-51 _ .. 



Mg 



6-20 



4-29 



3*90 



4-46 



Na 



3*95 



4-27 



5-05 



4-27 



Al 



24-27 



21-19 



23*06 



22*99 



F 









57-68 



58*25 



H 2 



15-68 



19*46 



10-17 



10*03 



99*85 100-00 

 Ratio (MgNaJ : Al = 1-17 : 3 1-06 : 3 0*97 : 3 

 * Zeitschr. Krvst., vii, 4*74. 



