PRINCIPLES UNDERLYING METAMORPHIC PROCESSES 515 
Of the hydrolysis of the silicates little is definitely known, 
except that the ordinary solutions of alkali silicates (water glass) 
are nearly completely hydrolyzed into free alkali and colloidal 
silica. But from this it is not quite safe to conclude that the salts 
corresponding to all of the silicic acids (or aluminosilicic acids,’ if 
such salts exist) are also completely hydrolyzed under all conditions. 
For it is quite conceivable that hydrolysis of a silicate may take 
place without producing colloidal silica; this might occur either 
because the silicic acid resulting from hydrolysis of the particular 
silicate is not colloidal or because of the intervention of factors— 
temperature, or the presence of other substances in the solution— 
which enhance the real solubility of silica and thus decrease its 
tendency to appear in colloidal form. The extent of hydrolysis 
under such conditions would be that corresponding to the disso- 
ciation constant of the silicic acid, which may not be so small as is 
commonly supposed. For in the hydrolysis of silicates, as it has 
been observed, the silica becomes colloidal and is removed from 
true solution as fast as it forms, and hence its active mass in the 
solution remains exceedingly small; consequently the hydrolytic 
dissociation continues until practically all of the original silicate is 
decomposed. This therefore affords us no certain information as 
to the strength of the silicic acid, for the same phenomenon would 
be observed even if the silicic acid were of such strength (e.g., com- 
parable to acetic acid) that the extent of hydrolysis is of the order 
of o.1 per cent; furthermore, it is no absolute criterion of what 
would happen if no colloidal material were formed. 
The réle of hydrolysis is important in the kaolinization of the 
feldspars, and in many reactions occurring in solution in liquid 
water; but in all probability it is altogether subsidiary—if not 
entirely negligible—under magmatic conditions, in which case it 
is preferable to regard the silicate mixture as solvent and the water 
(or other gases) as solute. 
Most silicates react alkaline in contact with water; but accord- 
ing to F. Cornu,’ an acid reaction is given by a number of minerals, 
'To illustrate by means of an analogous case: the salts of hydrocyanic acid 
(HCN) are largely hydrolyzed, whereas there is very little hydrolysis with salts of 
hydroferrocyanic acid (H,Fe(CN)s=4HCN, Fe(CN),). 
2 Tschermak’s Miit.,.X XIV (1905), 417. 
