B. A. Keen 397 



initial rise in the coarser textured soils and the higher final value in the 

 fine textured soils, and it will be seen that the maximum value is only 

 four feet. It is practically impossible to reproduce with certainty, under 

 laboratory conditions, the soil structure as it exists in the field. Leather''' 

 has pointed out in some remarks on the permeability of soils that it is 

 not difficult to fill a series of cylinders with portions of the same soil so 

 nearly uniformly alike that the rate of flow through each one is approxi- 

 mately the same. But an agreement of duplicates thus obtained is not 

 evidence that the field conditions are reproduced by the laboratory 

 experiments, and these considerations apply equally to laboratory experi- 

 ments on the capillarity of soils. 



It is thus a matter of considerable interest and importance to obtain 

 values for the possible capillary rise in soils. Mitscherlich'', in the course 

 of an extensive investigation into various physical properties of soil, has 

 effected such a calculation. It is an indirect one, because it involves a 

 relation between the experimentally observed heat of wetting (Benet- 

 zungswiirme), and the surface of the soil particles. The calculation leads 

 to enormous values for the possible capillary rise — 2 or 3 kilometres for 

 heavy clays and loams. His maximum experimental result, however, did 

 not exceed 80 cms. over a three months period. 



In the present note is given a simple direct calculation, the results of 

 which may be taken as probable maximum values for the capillary rise. 

 The calculation is based on the following assumption: if the soil grains 

 are taken as spherical, of one size and packed in the closest possible 

 manner, then the pore space may be regarded as consisting of capillary 

 tubes having an approximately triangular cross section. The dimensions 

 of these tubes have been obtained by Slichter*^ who introduced this 

 conception and justified it in the course of an extensive mathematical 

 treatment of the flow of air and water through an "ideal" soil. The 

 mean value of the triangular cross section is •2118r^ where r is the radius 

 of the soil grain. This triangular pore changes in cross sectional area as 

 it follows the surface of the soil grains, passing alternately through 

 maximum and minimum values. The minimum value is about -HI or-, 

 which is roughly 30 per cent, less than the average mean value -211 8r-. 

 The present writer agrees with these calculations, and has used the 

 results as shown immediately below. 



Consider the height, //, which water would reach in a tube whose 



» Journ. Agric. Sci. 4 (1911-12), p. 304. 



6 Landw. Jahrb. 30 (1901), p. 361. 



« King and Slichter, 19/7t Ann. Rep. U.S. Qeol. Survey (1899), pt 2. 



