XIII.] ORIGIN OF CRYSTALLINE ROCKS. 297 



source of steatite, and suggested the derivation of serpentine, 

 chlorite, and other related minerals of the crystalline schists, 

 from deposits approaching these marls in composition.* He 

 recalled, also, the occurrence of chromic oxide, a frequent ac- 

 companiment of these magnesian minerals, in the hydrated iron 

 ores of the same geological horizon with the magnesian marls 

 in France. Delesse did not, however, attempt to account for 

 the origin of these deposits of magnesian marls, in explanation 

 of which I afterwards verified Bischof 's observations on the 

 sparing solubility of silicate of magnesia, and showed that sili- 

 cate of soda, or even artificial hydrated silicate of lime, when 

 added to waters containing magnesian chloride or sulphate, gives 

 rise, by double decomposition, to a very insoluble magnesian 

 silicate. (Ante, page 122.) 



To explain the generation of silicates like the feldspars, 

 scapolite, garnet, and saussurite, I suggested that double alu- 

 minous silicates, allied to the zeolites, might have been formed, 

 and subsequently rendered anhydrous. The production of 

 zeolitic minerals observed by Daubree, at Plombieres and Lux- 

 euil, by the action of a silicated alkaline water on the masonry 

 of ancient Roman baths, was appealed to by way of illustra- 

 tion. (Ante, pages 25 and 205.) It has been shown by Daubree 

 that the elements of the zeolites were derived in part from the 

 waters, and in part from the mortar, and even the clay of the 

 bricks, which had been attacked, and had entered into com- 

 bination with the soluble matters of the water to form chaba- 

 zite. I, however, at the same time pointed out another source 

 of silicated minerals, upon which I had insisted since 1857, 

 namely, the reaction between silicious or argillaceous matters and 

 earthy carbonates in the presence of alkaline solutions. Nu- 

 merous experiments showed that when solutions of an alkaline 

 carbonate were heated with a mixture of silica and carbonate 

 of magnesia, the alkaline silicate formed acted upon the latter, 

 yielding a silicate of magnesia, and regenerating the alkaline car- 

 bonate ; which, without entering into permanent combination, 

 was the medium through which the union of the silica and the 



* Etudes sur le Metamorphisme, quarto, pp. 91. Paris, 1861. 

 13* 



