§ 14. The statement commonly made is that carbonic acid enters the compound and first breaks up the union of the alkalies, potash and soda, with silicic acid, uniting with the bases and setting the silicic acid free; that it next attacks the lime and magnesia in the same way, and finally the iron, leaving the union between alumina and silica undis turbed. We would then have, in place of orthoclase, carbonates of potash, soda and iron, and hydrous silicate of alumina with free silica. This seems not to be strictly true. The real reaction seems to be that carbonic acid breaks the complex silicate into its elements and that these elements re-unite to form a large number of 'compounds better suited to the new conditions. This charge is often so complete that even the union between water and silica (which forms silicic acid) is broken, and they unite in different proportions forming an acid with different com position and properties. When drainage is defective and the ground water becomes saturated with dissolved salts, it often happens that the decomposition of a single complex silicate gives rise to a large number of simpler salts, some of which are silicates and some salts of other acids.
Van Hise in his "Treatise on Metamorphism" (pp. 372-4.) gives a list of the more common minerals and the varieties which result from their decomposition. As examples, I have selected those which follow : Orthoclase alters to allophane, biotite, cimolite, damourite, epidote, gibbsite, halloysite, kaolin, muscovite, newtonite, pyrophyllite, quartz.
Biotite alters to chlorite, diaspore, epidote, gibbsite, hematite, hydrobiotite, hypersthene, kaolin, limonite, magnetite, quartz; serpentine, sillimanite, spinel.
Hornblende alters to augite, biotite, calcite, chlorite, epidote, hematite, mag netite, quartz, serpentine, siderite.
Nephelite alters to albite (conjectural), analcite, diaspore, gibbsite, hydro muscovite (pinite), hydronephelite, kaolin, muscovite, natrolite, sodalite, thomsonite.
It must not be inferred that whenever one of these minerals decom poses all the above-named alteration products result, but simply that any group of them may be formed. It is probable that many changes not included in the above list take place.
Formation of Residual Clays.—§ 15. It has been stated above that when acid solutions enter a rock, the acids unite most readily with the alkalies, next with the alkaline earths, and then with the iron. If the water circulation is poor and the solutions come to be saturated, most of the bases will be redeposited, either as silicates or salts of other acids or both, but if the circulation is free, most of them will be carried away in solution to be deposited elsewhere. The acids of ground-water seem
to have much less attraction for aluminum than for the other bases, and so the greater portion of this base is allowed to form new com pounds with the free silicic acids, such as halloysite (Al208. 2Si08±Aq), allophane (A108. SiOz+5H20, cimolite (2Al208. 9Si0+6H20), colly rite (2Al208. Si02+9H2O), kaolin (Al208. 2Si02+2H20), schrotterite (8Al203. 3Si02+30H20), montmorillonite (Al208. 4Si02-f-H2O-{-Aq), the zeolites, etc. Some of the alumina, however, is nearly always re deposited without combination with silica in the form of gibbsite (Al208. 3H20), diaspore (Al208. H20) and other oxides or hydroxides.
If granite or a granitoid rock should completely decompose through the action of acids under conditions which. afford perfect drainage, most of the potash, soda, lime, magnesia, iron, etc., would be converted into soluble salts and carried away, while the aluminum and magnesium salts would mostly be left in the form of hydrous silicates and oxides. Such a mass would be composed of the minerals enumerated in the preceding paragraph, would be earthy in texture, and would have the properties which we assign to clay; in fact it would be what ve call a pure clay. No fixed composition could be assigned to such masses because they would contain varying proportions of these minerals and others like them. Deposits rich in allophane, collyrite, gibbsite, etc., would be high in alumina while those rich in cimolite, montmorillonite, etc., would be high in silica. As kaolin liar nearly the average composition of this group of minerals it is customary to use its name for the whole mass, and this custom is all right if we remember that deposits so named are not composed of a single mineral substance having fixed properties, but of a group of minerals whose properties vary among themselves quite widely.
A glance at the table of analyses of kaolin (§ 17) will make this point clear. Nos. 4 and 10 in the table have nearly the theoretical composition of kaolin, while 11 and 12 of this table, 7 and 8 of the table of ball clays, and 8 and 9 of that of flint clays, show too high a percentage of alumina in comparison to the silica, indicating the presence of gibbsite or some other mineral high in alumina, and Nos. 1, 2 and 6 are much too high in silica. These last may be explained by assuming the pres ence of free silica, or of minerals higher in silica than kaolin. Probably the true explanation would include both these causes.