ON THE EFFECTS OF LONG-CONTINUED HEAT. 181 



lowers this specific gravity to a constant amount, and that fused silica does 

 not recover its density in cooling. Crystalline granite, as Delesse has shown, 

 passes by fusion from the density of 2G2 to that of 2*32, and Egyptian 

 porphyry from 2*76 to 248. 



Again, the felspar in granite is encrusted by the quartz, the most fusible 

 by the least fusible material, contrary to all experience of crystallization 

 either from solution or fusion. 



Lastly, all the minerals of which granite is composed have been artificially 

 produced, and their production has in every instance taken place at tempera- 

 tures far below that of the fusing-point of that rock. The first specimens 

 of artificial felspar analysed by Karsten, and measured by Mitscherlich, 

 were found in the lining of a copper furnace amongst a sublimate of zinc. 

 Mitscherlich tried to obtain the like by fusing several pounds of native felspar 

 in a porcelain furnace, and subjecting the mass to a process of slow cooling, 

 but without success*. In the Mulden smelting works, Cotta observed the 

 walls of the furnace traversed, in the joints of its masonry, and in the cracks 

 which it had undergone, by beautiful metallic veins, the sides exhibiting the 

 phenomena of impregnation and alteration as in the boundary walls of 

 natural veins, and the ores consisting of galena, blende, iron and copper 

 pyrites, purple copper, Fahl ore, native copper, &c. In like manner pyro- 



morphite (Pb 3 3? + 3- Pb CI), in well-formed six-sided prisms from the iron 

 furnace at Asbach, was found attached to the stones of the masonry. There 

 can be no doubt but that Karsten's crystals of felspar, like these, were 

 formed by gaseous sublimation ; and an analogous process would account 

 for the felspar observed by Haidinger in a basaltic cavity, under the form 

 of Laumonite, and by Bischof in a porphyritic bed, in which a Trilobite 

 also was found. 



A new view of the production of minerals has been opened by Ebelmtn, 

 who obtained the most refractory crystals of the granitic rocks, such as 

 spinel, emerald, cymophane, and corundum, by segregation in the interior 

 of a fused mass. They were formed at a heat far below that which would 

 fuse either those crystals or granite, by means of the evaporation of a fusible 

 and volatile medium. Gaudin also, on the same principle using a similar 

 alkaline solvent, and substituting sulphuric for boracic and carbonic acids 

 as the volatile ingredient, obtained the ruby. 



To the same category may be referred an experiment by Precht, who 

 having added to a transparently fused frit, weighing 1^ cwt., a considerable 

 quantity of felspar, found, after cooling, that a large portion of this mineral 

 had separated itself in foliated masses, and in several distinct crystals. 



The most important light, however, on this subject, especially in relation to 

 metamorphic phenomena, is from the experiments of Daubree on the reaction 

 of gaseous compounds upon various earthy bases. Conveying the chlorides 

 of tin and titanium over lime at heats varying from 572° to 1652° Fahr., 

 he produced crystals of tinstone and brookite ; by variations of the same 

 principle, at heats not exceeding redness, he obtained all the following mine- 

 rals : — wollastonite, staurolite, peridote, ditthene, willemite, idocrase, garnet, 

 phenakite, emerald, euclase, corundum, zircon, periclase, spinel, augite, di- 

 opside, gahnite, franklinite, haematite, felspar, and tourmaline in hexagonal 

 prisms imbedded within crystals of quartz. The process was of this descrip- 

 tion : — Chloride of aluminium, passed over lime at a red heat, produced 

 corundum ; chloride of silicium, passed in like manner over seven equiva- 

 lents of potash or soda and one of alumina, produced the different species 

 of felspar : the latter named gas, decomposed by lime at the same heat, or 



* Mr. Marshall fused a large mass of granite, and cooling it slowly obtained no crystals. 



