464 R. C. WALLACE 



process is one of hydrolysis ; and while it may be effected by water 

 alone, it is doubtless hastened by the carbon dioxide of the atmos- 

 phere. The gels which are formed during the process of decom- 

 position are irreversible — that is, they cannot, by the action of 

 electrolytes, pass over into sols and be in this way removed from 

 the sphere of action. They exercise, however, a selective absorp- 

 tion, alkalies being removed from the brine while the acid radicals 

 remain in solution. With potassium salts in particular this 

 property of the colloids of the soils is of importance in retaining 

 the valuable ingredients of fertilizers. This selective absorption 

 tends, therefore, to hydrolyze the chloride, and to render the solu- 

 tion more acid. The free acid reacts on the partially decomposed 

 minerals, causing further disintegration. Quartz is not affected 

 thereby, but the corrosion of quartz is in all probability due to the 

 action of alkaline carbonates. 



In short, then, the principal fact in the disintegration is the 

 intimate contact of the liquid (in a very thin him) with the atmos- 

 phere and the rock. The initial stages of the disintegration are 

 caused by the action of water in contact with air rather than by 

 that of salts in the water. Colloidal precipitation, which takes 

 place when decomposition of the silicate begins, leads to selective 

 absorption and consequent acidification of the solution, giving rise 

 in turn to further, and probably more intense, disintegration. The 

 process is continuous, gel being continuously precipitated, and 

 further selective absorption taking place. The gel, being irre- 

 versible, is not taken up as an emulsion, and can consequently be 

 removed only mechanically. 



Comparison with the action of sea-water. — On comparing the 

 action of sea-water on bowlders of similar composition to those 

 attacked by the brines, one finds an undoubtedly real difference 

 in the degree of corrosion. It is, however, a difference in degree, not 

 in process. The evidences of chemical erosion caused by the sea- 

 water are to a large extent removed by mechanical attrition caused 

 by the impact of the waves on the bowlders, and the consequent 

 rolling of the bowlders on the beach. Even in large bowlders, 

 however, where rolling is reduced to a minimum, the evidence of 

 chemical disintegration is small indeed in comparison with that of 



