388 Peiifield and Minor — Chemical Composition and 



accepted is that of Groth* [A1(0.F,)] AlSi0 4 , corresponding 

 to an isoraorphons mixture of (AlF a ) AlSi0 4 with the andalusite 

 molecule (AlO) x\lSi0 4 in which the former predominates and 

 in which fluorine is supposed to be replaced by oxygen. 

 Rammelsbergf suggests a mixture of Al 3 SiO, and AI„SiF I0 in 

 the proportion of 5 : 1. The ratio of Si0 2 : Al a 3 : F varies 

 from 1:1: 1*50 to 1:1: 1*81 and if this could be shown to be 

 1:1:2 the composition could be expressed by either of the 

 following simple orthosilicate formulae : 



F - A1 <o\ F> A1 ^,. 



>Si or /7y> Sl 



Since it has been shown by one of us that hydroxyl so fre- 

 quently replaces fluorine, and it now seems very doubtful if 

 bivalent oxygen ever plays this role, the idea has suggested 

 itself that perhaps the variations in the percentages of fluorine 

 and the failure to yield a simple ratio are due to the partial 

 replacement of fluorine by hydroxyl. Accordingly tests were 

 made for water and it has been found to be always present. 

 This fact seems to have been generally overlooked.^: 



In testing by the ordinary closed tube method it is not 

 always evident that hydroxyl is present, since in a mineral like 

 topaz an acid vapor comes off, probably hydrofluosilicic acid, 

 instead of water. By mixing the mineral, however, with lime 

 or some other substance to hold the fluorine, water is evolved. 

 In order to determine to what extent hydroxyl is present and 

 what part it plays in the chemical composition, material from 

 a number of localities has been examined, and it will be 

 shown in the course of this article that the variations which 

 topaz shows both in chemical composition and physical proper- 

 ties result from an isomorphous replacement of fluorine by 

 hydroxyl, while a simple composition has been established 

 which can be expressed by the formula [Al(F.OH)]„Si0 4 . 



Method of analysis. — The important features of the analysis 

 were of course the accurate determination of fluorine and 

 water. For fluorine the method of Berzelius was adopted. 

 The mineral, mixed with half its weight of quartz, was fused 

 with five times the total weight of mixed sodium and potas- 



* Tabellarische Uebersicht der Mineralien, 1889, p. 106. 

 \ Mineralchemie, 1875, p. 580. 



\ As this article was going to press, we have observed in Nos. 2 and 3, vol. vi 

 of the Zeitschrift fiir anorgamsche Chemie, p. 168, March, 1894, which has just 

 come to hand, that Jannasch and Locke have also determined the presence of 

 water in topaz with the following results, but they have not discussed the relation 

 of the water to the chemical composition. 



San Luis, Mexico. Ilmen Mountains. Schneckenstein. Brazil. 



080 1-02 1-28 2-69 



