Studies in Soii. Physics, IV. 173 



is a draining under a force acting like gravity but many times 

 as stronglv. With the same centrifuge running at the same 

 speed, or witli a proj)cr ritio of speed and diameter of drum 

 whatever these may be, the cotiditions are closely reproducible 

 and it is believed that the percentages of retained water thus 

 measured stand in close and pretty rigid relation to the physical 

 nature of the soil and esi:>ccial to its behavior towards water. 

 The exact forms of these relations are not known, and, as be- 

 fore, all correlations are purely e npiri:al. Furt'iernnre, the 

 method itself is empirical and the numbers obtained on the 

 given soil types will not var\-,not only in numerical value but 

 also in relation to each other, with variations in the intensity 

 of the centrifugal force. The amount of water retained by 

 anv given soil varies with the centrifugal force applied, and as 

 the force becomes more and more intense, more and more water 

 is extracted, but, as usual, equal increments of force produce 

 progressivelv decreasing decrements of water content. The 

 amount of water retained falls (as the force is increased) toward 

 a limiting, asvnijitotic value below wliich still further increases 

 of force produce very little extraction of water. Both the 

 actual percentage of water corresponding to this asymptotic 

 condition and the intensity of the force by which it can be ob- 

 tained varv with the physical cliaracter of the soil. In sands 

 it is low and easily attained, in clays it is higher and is attained 

 onlv with greater forces. Now the forces usually employed 

 in determining the water ecjuixalent are intermediate between 

 those competent to produce the asymptotic extraction in sand ^ 

 and those necessary to produce it in clays. Consequently the 

 va'ues thus obtained for the retentivity of different soil types 

 are not comparable with each other, being made by methods not 

 physically aUke. The water equivalent, thus determined, is not 

 really a physical constant but merely an empirical determina- 

 tion made by an arbitrary method and n:jt comparable in any 

 sense with similar determinations made by slightly varied 

 methods. 



As will be seen later this condition is not insusceptible of 

 improvement, but even if it were, the practical usefulness of 

 the water equivalent would not necessarily disappear. It could 



