MINERALOGICAL AND CHEMICAL CHARACTERS. 



29 



By calculating the sesquioxides as alumina, the protoxides as 

 magnesia, and the alkalies as potash, and calculating to 100, we obtain 

 the following as an average analysis, which, omitting the small quantity 

 of magnesium fluoride, agrees nearly with the formula already given 

 for typical muscovite, namely, K 2 0. 2 H^O. 3 A1„0 3 . 6 Si0 2 :— 



SiO., . 



• 457 



AL0 3 



• 37'4 



MgO . 



. 1-4 



K 2 . 



. 10-4 



H 2 . 



. 4'9 



F 



. 0'2 



Small crystals of various micas have been artificially prepared by 

 various workers, notably by Hautefeuille and St. Gilles, von Chrust- 

 schoff and Doelter, whilst Vogt and others have recorded the for- 

 mation of some varieties of this mineral in furnace slags. Artificial 

 micas have never been, and are not likely to be, made in crystals 

 sufficiently large to be of marketable value; but Doelter's experiments 

 to test the action of sodic and magnesic fluorides on certain natural 

 silicates are particularly instructive on account of the way in which 

 they show the formation of mica at the expense of other silicate 

 minerals. In this way Doelter has shown that the variety biotite can 

 be made by the alteration of hornblende, glaucophane and garnet, a 

 phlogopite from pargasite, whilst a muscovite-like mica was formed by 

 the action of potassium fluo-silicate and aluminic fluoride on anda- 

 lusite. The last-named result is of importance, as pointed out 

 below (p. 39), on acccunt of the way it partially imitates the con- 

 ditions under which valuable muscovite appears to have formed in 

 pegmatite veins. 



C 2 



( *9 ) 



