B. A. Keen 399 



the values are calculated gives in fact a rectangular hyperbola for the 

 curve connecting h and r, and as r becomes very small, the value of // 

 increases exceedingly rapidly. The values for the various grain sizes 

 given in Table II refer, as already stated, to an "ideal" soil in which 

 the grains are all of one size, spherical and packed in the closest possible 

 manner. 



It is quite certain that these figures must be considerably reduced 

 for actual soils made up of a mixture of particles of all shapes and sizes 

 in which the capillary spaces are irregular in length, width and also 

 direction. The irregular shape of the grains will result in a narrow 

 capillary tube suffering a considerable enlargement of its effective cro.ss 

 section at some point, and this will greatly reduce the amount of the 

 possible capillary rise. The trapping of air in the interstices between 

 the grains will also act in the same direction. Finally in the heavier 

 types of soil containing much clay, the swelling of the colloidal portion 

 due to inhibition of water will completely close some of the openings and 

 w'ill reduce the diameter of others to such extent that the movement of 

 water through them will be extremely slow. 



The values in Table II for various soil fractions may therefore be 

 regarded as the extreme limits of capillary rise for actual soils of corre- 

 sponding average pore area. In all probability these values are con- 

 siderably in excess of those occurring in practice. On the other hand, 

 laboratory experiments give minimum values for the capillary rise, 

 because it is impossible to reproduce, as in the field, the long continued 

 oscillation of meteorological and soil water variants, which results in the 

 progressive compacting of the surface soil and subsoil into the most 

 favourable position for the production of the capillary effects. 



[Received July ^rd, 1919.) 



