.] METHODS OF SAMPLING - 53 



ie particles of soil, thus it serves both to bind a coarse- 

 rained sandy soil, and, by forming aggregates of the 

 nest particles, to render the texture of a clay soil more 

 pen. In determining the sizes of the constituent 

 articles of a soil, the " mechanical analysis," it is desir- 

 ble to remove the humus as far as possible, and so 

 reak up these temporary aggregates. 



Sampling of Soils. 



The first step in the analysis of any soil, mechanical 

 r chemical, consists in obtaining a sample that shall 

 dequately represent the land in question. 



In this country it is customary to take a sample down 

 a depth of 9 inches as representing the soil proper ; 

 t is, however, doubtful if this is not too deep, being 

 elow the depth to which cultivation is generally carried ; 

 >robably a 6-inch sample would more truly represent the 

 ultivated soil. In many cases it will be found that the 

 rue soil does not extend to a depth of anything like 9 

 nches, but that there is a sharp change into subsoil or 

 :ven rock before this point : e.g., on the chalk downs the 

 oil is often not more than 4 inches deep, below which 

 vhite broken chalk rock begins. In such cases the 

 ample must only be taken to the depth at which the 

 visible change begins. 



To obtain the sample two methods are generally 

 idopted. At Rothamsted a steel box, without top or 

 Dottom, 9 inches deep, and 6 inches square in section, is 

 jsed ; the sides are wedge-shaped, about f inch thick 

 it the top and tapering off to cutting edges below, 

 rhe surface, if uneven arable land, is first raked over 

 ind gently beaten level, then the box is placed in 

 position and driven down v/ith a heavy wooden rammer 

 : ill the top of the box is flush with the surrounding 

 soil. The soil enclosed by the box is then carefully dug 



