OEIGIN OF CRYSTALLINE EOOKS. 49 



gelatinous precijîitate, which in sealed tubes, at temperatures of from 150' to 200' C, 

 was gradually changed into hexagonal plates of a potash-soda zeolite with the oxygen- 

 ratios, 1 : 3 : 6 : 2 ; having the physical characters of levynite. The residual liquid 

 was nearly free from both silica and alumina. On repeating this experiment at a higher 

 temperature, a very different result was obtained. There was an abundant separation of 

 silica in crystalline grains, with a little levynite, while an alkaline aluminate remained in 

 solution. This remarkable dissociation of the first-formed aluminous silicate into free 

 silica and soluble alumina recalls the conditions of the sexjaration of quartz already noticed 

 in § 87. The crystalline silica produced in this reaction may be either quartz or tridymite, 

 which latter form of silica, mingled with quartz, was obtained in 18*79 by Friedel and 

 Sarrasin by heating gelatinous silica with an alkaline solution to about 400' C. The 

 dissociation of alumina from silica, observed in this experiment, serves to throw light on 

 the origin of corundum and spinel. In other exiîeriments with mixtures of solutions of 

 silicate and aluminate of potash in sealed tubes at 200' C, Deville got a crystalline 

 compound with the formula of phillipsite, 1:3:8:5. Subsequently, de Schulten, in 

 similar experiments, at 180' C, with silicate and aluminate of soda, obtained crystals of 

 analcite, with the ratios, 1 : 3 : 8 : 2.*^ 



§ 99. More recent investigations in the same direction by Friedel and Sarrasin are 

 very instructive, as showing not only the generation of feldspars in the wet way, but the 

 production at will, under similar conditions, of a feldspar or a zeolite. These chemists had 

 already, by heating a mixture of silicate of alumina (precipitated from a solution of chlo- 

 ride of aluminium by silicate of potash) with an excess of a solution of silicate of potash, 

 obtained crystals of orthoclase, mingled with crystals of c[uartz or, at a more elevated tem- 

 peraiiire, of tridymite. In subsec[uent experiments, undertaken for the iDroduction of 

 albite, a similar hydrous silicate of alumina was mingled with a solution of silicate of 

 soda, (the silica and alumina in the proportions of the soda- feldspar), and heated to from 

 400' to 500' C. Instead of the anhydrous albite, however, were obtained crystals of 

 analcite, 1 : 3 : 8 : 2 ; the excess of silica, with soda and some alumina, remaining in 

 solution. When, however, an excess of silicate of soda was employed, the whole of the 

 silicate of alumina was transformed into albite.*^ Thus analcite, which is formed by the 

 action of thermal springs below *70° C, is equally produced at 180^ C, as in the experi- 

 ments of de Schulten, and at 400' C. and upwards. 



§ 100. We have thus far considered among aluminous double silicates those which 

 present the oxygen-ratio of E, : al = 1 : 3, aud have only mentioned incidentally the epidote 

 and meionite groups. The n^^merous experiments already detailed suffice to show that 

 the double silicates of alumina and alkalies, formed under very varied conditions in the 

 wet way, in the presence of an excess of alkali, always present this ratio, of 1 : 3. 

 When, however, we pass to aluminous double silicates with other protoxyd-bases, we find 

 many with the ratio, 1 : 2, as in the epidote aud meionite groups ; with 1 : 1, as in the 

 alumina-garnets and biotite ; or even 2 : 1, as in phlogopite and many hydrated alumino- 

 magnesian species of the chlorite group. The genesis of these various calcareous 



"* The results of Deville, Friedel and Sarrasin, and de Schulten in the preceding paragraphs, are cited from 

 Michel Levy and Fouqué, Synthèse des minéraux et des roches, Paris, 1882, pp. 87-13-1 and lGl-104. 

 *' Compte Rendu de l'Acad. des Sciences, le 30 Juillet, 1883. 



Section III., 1884. 7. 



