464 Sti«li('s of a Scottish Drift Soif 



{b) Aclion of solvents. Certain substances are much more readily 

 dissolved than otiiers, e.g. in granite the felspars are more readily 

 attacked than quartz. Tliis afjaiii leads to some minerals ofcurring 

 more frequently than others in certain fractions. 



((?) Ah-^'orption and redeposilion of dissolved mailer on the surface 

 of tlie soil particles takes place through changes of temperatuic and 

 concentration, and other changes affecting the conditions existing 

 in the soil solution. 



The composition of all the fractions of a soil will lie iniluenced 

 by the extent to which these three processes have been carried during 

 the soil formation. Though these processes all operate to some extent 

 in practically every case and produce that rough general similarity in 

 chemical composition which is found to exist in particles of similar 

 grade from widely different soils, sometimes one process predominates 

 and sometimes another and hence considerable differences arise in all 

 grades of particles between soils of different history. For example, 

 where solvent action has not proceeded far and the weathering is chiefly 

 mechanical, the fractions will differ in chemical composition from those 

 of soils which have long been subjected to the action of solvents. 

 Even mechanical forces give rise to different types of chemical com- 

 position in similar mechanical separates, for the mechanical action of 

 water is more selective than that of ice, and, other things being equal, 

 there would tend to be a greater proportion of hard materials in the 

 silts and clays of glacial origin than in those due to attrition by 

 running water. 



The effect of differences in nature and origin on the chemical compo- 

 sition of the separates was drawn attention to by Dumont'. He showed 

 that two soils containing about the same proportion of potash (0-894 

 and 0-853 per cent, respectively) had very different percentages in their 

 iriechanical fractions. In one, a fine grained soil from an experimental 

 field at (Jrignon, the coarse sand contained 0-864 per cent, of potash, 

 the fine sand 0-992 jiit criil.. and the cliiy O-940 per cent. ; while in the 

 other, a coarse sandy granitic soil from ('reuse, the coarse sand contained 

 l-3;5 per cent., the fine sand 0-.")8 per cent., and the clay 0-.")l per cent. 

 When expressed as percentages of the whole soil, the dilTerence in the 

 distribution of the potash is even more striking, for the soil from (iiignon 

 contained 16-8 per cent, of clay and 17-2 per cent, of coarse sand as 

 compared with 4-5 per cent, of clay and 44 per cent, of coarse sand in 

 that from Creuse. 



» Comple.s Reniius. 19(H. 138, 21.>-217 



