THE SOIL AND THE PLANT 143 



the physical and chemical properties of soils and more recently 

 around the alleged excretion of toxins by plants. That the 

 physical properties of soil were of fundamental significance in 

 determining fertility was recognised by all the older agricultural 

 chemists although actual investigations were made only by 

 Schubeler [22], Wollny [30], and a few others. It was not till 

 Whitney's first papers appeared [24, 25] that the relationships 

 became clear and definite ; and when, later on, Briggs [6] and 

 King [13] published their notable researches on the movements 

 of water in soils, chemists were put in possession of useful 

 working hypotheses and learnt how to connect up the properties 

 of a soil with the properties of its component particles. It was 

 shown that the size of the particles and the extent of their 

 surfaces were highly important factors in regulating the water 

 and air relationships of soils ; further, that the known agricul- 

 tural properties of the soil could be correlated with the amounts 

 of the variously sized particles it contained, amounts readily 

 determined by a mechanical analysis. These ideas were intro- 

 duced into this country by Warington [23] and have so com- 

 mended themselves that mechanical analysis is now a recognised 

 part of the work of a soil chemist and indeed gives a better 

 basis for explaining the agricultural properties than does a 

 chemical analysis [11]. 



But we are not prepared to agree that the chemical properties 

 of soils are relatively insignificant in determining fertility. If 

 the coarser fractions of a soil are analysed they are found to be 

 almost entirely silica, but the finer portions — below 0*005 mm - 

 diameter — which play a controlling part in soil fertility are more 

 complex and contain besides the undecomposed rock material 

 a quantity of weathered silicates and decomposition products. 

 The special properties of the finer material are associated by 

 Whitney with their small dimensions, approximating in some 

 cases to molecular dimensions, and not with any particular 

 chemical compositions ; but evidence is accumulating that mere 

 smallness of size will not account for all the observed properties. 

 The view most commonly held by chemists is due to van 

 Bemmelen [1-4]. Under the conditions obtaining in the soil a 

 number of the decomposition products are deposited in a colloidal 

 form, hence the soil particles are coated with a colloidal complex 

 containing silica, aluminium and ferric hydroxides, alkaline bases, 

 phosphoric acid and humus, not in true chemical combination 



