SOILS AND THEIR PROPERTIES 61 



permits, it will be advisable to separate the soil into a series 

 of layers, the first three inches, the next three inches, a 

 third three inches, and possibly a few further depths as well. 

 A very large number of samples of soil have been taken to 

 a depth of nine inches, and it is, therefore, desirable, for 

 comparative purposes, that the amount of plant food to a 

 depth of nine inches should be known, but it is often advisable 

 to have further information. The great variation of compo- 

 sition which occurs in soils from depth to depth must always 

 be borne in mind, since unless soils be sampled to a definite 

 depth, no sort of constant results can be obtained. There 

 are, however, many occasions when a soil is not nine inches 

 deep, and one is, therefore, compelled to content oneself 

 with less depth. Not infrequently within easy range of a 

 spade from the surface one may come across rock more 

 or less broken down by weathering. Many attempts have 

 been made to obtain some mechanical appliance to 'obtain 

 samples of soil with less labour than that involved in fiist 

 of all digging a hole and then obtaining a vertical slice. 

 Within the narrow limitations of particular types of soil 

 such efforts are perfectly satisfactory, but a universal 

 method for all soils has yet to be discovered, excepting the 

 more laborious method here described. All instruments of 

 the type of a boring tool become unworkable in a soil con- 

 taining many stones, whilst in humus soils they introduce the 

 serious difficulty of inaccurate measurement, owing to the 

 compression of the soil which they produce. They further 

 have the great disadvantage that the operator cannot see 

 the nature of the soil he is sampling, and an observation 

 on the spot of the actual appearance of the undisturbed 

 soil will often teach quite as much as the subsequent analysis. 

 The size of the particles of soil is a matter of great 

 practical importance. This subject has been investigated 

 very fully, and much of the literature on the subject is given 

 under such names as physical or mechanical analysis. The 

 manner in which the particles are packed together is also 

 a point of great importance. The actual size of the particles 

 is not easily altered, but the manner in which particles are 



