II.] THE NATURE OF CLAY 41 



Of these fractions the first contained all the mica, 

 the others were practically pure kaolinite, yet the second 

 fraction showed none, and the third very little of the 

 characteristic properties of clay ; when dried they fell 

 or could easily be rubbed into a fine powder, only 

 the fourth and finest fraction dried into a hard co- 

 herent mass. Thus we can have material which con- 

 sists entirely of kaolinite, and yet is not clay ; such as 

 we see in natural deposits of fuller's earth, which 

 consists of kaolinite but possesses no plasticity and 

 falls on drying into a fine powder. 



In the same way a natural soil contains particles 

 of silicates of alumina of all sizes, though they only 

 begin to predominate in the fractions of finest grain. 

 In fact the physical processes of weathering, abrasion 

 in running water, cannot reduce rock materials into 

 particles below a certain size, thus the particles in the 

 fractions of finest grain are not quartz but materials 

 that have been formed by the chemical disintegration 

 of felspar and similar rock constituting minerals. 



On separating by the method to be described later 

 (p. 56) a brick earth soil into fractions according to their 

 size and analysing each fraction, the result shown on 

 accompanying table (p. 42) was obtained. 



While there is a steady decrease in the amount of 

 silica and an increase in the amount of alumina and potash 

 is the particles get smaller, there is a sudden change of 

 :omposition when the two finest fractions are reached, 

 raking the mean of a number of analyses of the clay 

 "raction of Rothamsted soils, then if all the alumina 

 nay be supposed to be combined as kaolinite A1 2 3 , 

 2Si0 2 , 2H 2 0, the clay would contain 72 to 75 per 

 :ent. kaolinite, 11 to 12 per cent, ferric oxide, with 9 to 

 to per cent, of silica, and 4 to 6 per cent, of alkalis and 

 ilkaline earths. This means that the clay contains 



