THE COMPOSITION OF THE SOIL 95 



large number of analyses, which have given information of 

 considerable value to the student and the cultivator. The 

 method is fundamentally defective in that the grouping is quite 

 arbitrary, sharp lines being drawn where none exist in Nature, 

 and the soil is represented as a mixture of five or six different 

 substances when in point of fact the number of components is 

 indefinitely large. It is very difficult to use the resulting 

 figures for further investigations : they cannot be plotted on 

 curves or reduced to simple factors. They can, it is true, be 

 set out in columns which are very convenient for lecture 

 purposes. Some limited success has followed the attempts to 

 correlate the figures with other soil properties, as will be shown 

 later ; but in the main the actual figures obtained have not 

 proved very fruitful to investigators, though the broad general 

 results have been useful. 



Very much better results might be expected if a distribu- 

 tion curve could be obtained, showing how the particles are 

 distributed according to size. 



A certain amount of artificiality seems inevitable, and one 

 must still keep to the conventional *' effective radii " and deal 

 with the soil particles, not as they are, but as they would be 

 if they were perfect spheres. A distribution curve on these 

 lines, while not perfectly expressing the soil conditions, would 

 be a great advance on anything that we have at present. 



A serious effort to obtain such a curve has been made by 

 Sven Oden, whose elegant method of mechanical analysis 

 deserves serious attention from soil investigators. Instead of 

 ascertaining the weights of the fractions of soil falling between 

 certain limits of size, Od6n proceeds in the reverse order, and 

 ascertains the time taken for small successive equal weights to 

 fall through a column of water. A suspension of soil in water 

 is poured into a cylinder near the bottom of which is a large 

 flat plate attached to one arm of a balance, the other arm 

 being counterpoised and containing in its pan a small weight. 

 As soon as the weight of the soil particles settling on the plate 

 begins to exceed the weight in the pan the balance moves : 

 this makes an electric connection which registers the time on 



