﻿Penfield and Harper — Composition of Ralstonite. 381 



Only small quantities were used in making the above analy- 

 ses owing to the scarcity of material and the difficulty of 

 obtaining it free from thomsenolite. As far as the metals are 

 concerned the three analyses agree remarkably well with one 

 another. Nordenskiold calculated that the 50*05 per cent of 

 material which was not determined in his analysis, if assumed 

 to be fluorine would not be sufficient to satisfy all the metals, 

 and concludes that the mineral must contain some oxygen. 

 The same would be true in Penfield's analysis. Brandl deter- 

 mined the fluorine directly and found it just sufficient to satisfy 

 the metals ; he proposed the formula 3£N"a a MgCa)F 3 , 8A1F 3 , 

 6H 2 0. 



The fragments which were left after selecting material for 

 the above analysis by Penfield were carefully saved, and 

 although we could see that they contained large quantities of 

 ralstonite, still it was found impossible to separate the pure 

 mineral from thomsenolite by hand picking. After selecting 

 the purest crystal of ralstonite we could find, we took its 

 specific gravity by just floating it in the Thoulet solution and 

 found it to be 2*560; by the same method we found the 

 specific gravity of cryolite to be 2*974 and thomsenolite 2*979. 

 The great difference in specific gravity between ralstonite and 

 the other two minerals admits of a very perfect separation by 

 means of the Thoulet solution. All the material which we had 

 was crushed and made to pass an eighty-mesh sieve, the finest 

 dust was washed away by means of water and the separation 

 carried on as usual, repeating it several times so as to remove 

 the last traces of thomsenolite. Finally, the purest powder 

 was floated on a solution whose specific gravity was 2*611, a 

 few heavier particles were removed and the solution diluted to 

 the specific gravity 2*551, when all but a trace of the powder 

 sank to the bottom. This material, amounting to over twelve 

 grams, showed no impurity of double refracting thomsenolite 

 when examined under the polarizing microscope, and was used 

 in making the following analysis. 



A few remarks are necessary regarding the method of analy- 

 sis. It was found to be practically impossible to decompose 

 the mineral completely with sulphuric acid; a residue was 

 always left which could not be dissolved by further treatment 

 with sulphuric acid, nor with hydrochloric or nitric acids. For 

 the determination of the metals the mineral was fused with 

 sodium carbonate, the fusion soaked but with water, sulphuric 

 acid was added, and the solution evaporated and ignited till 

 the hydrofluoric and excess of sulphuric acids were expelled. 

 The mass was then dissolved in water and the metals determined 

 according to the usual methods. Sodium was determined once 

 by Smith's fusion method, and again by decomposing the min- 



Am. Jour. Scl— Third Series, Vol. XXXII, No. 191. — November, 1886. 

 24 



