532 STATE BOARD OF AGRICULTURE. 



The practical question now rises: Would identical results be obtained 

 under natural conditions? This will depend to a large extent upon tlie 

 structure and composition of the soil. If the soil is compact and con- 

 tains large amounts of colloidal material the swelling of the latter prob- 

 ably may predominate and cause a decrease in the rate of percolation. If 

 the same soil, however, is loose, and has many crevices, worm holes, etc., 

 the decreased viscosity may predominate and increase the flow of water. 

 On the other hand, it must be remembered that the temperature under 

 field conditions exists practically always in a gradient form, while in the 

 present research it was the same throughout the soil column. 



From studies on the flow of water in soils, various investigators have 

 attempted to express the relation of the soil to the movement of water 

 through it, quantitatively. To accomplish this they have assumed that 

 the porous soil is composed of bundles of capillary tubes, and that the 

 velocity of flow of liquid will consequently conform to Poiseuille's formu- 

 la, which is: i 



11 gh St r' 



^ ~8~' n ■ ~T' 



when V = volume of liquid; 

 and s = density of liquid; 

 r = radius of capillary; 

 n = viscosity of liquid; 

 1 z= length of capillary ; 

 g = gravitational constant; 

 h = head of liquid pressure; 



Such an assumption, however, is practically futile. In the first place, 

 Poiseuille's formula was deduced from the ascertained rate of flow of 

 liquid through straight tubes of approximately uniform diameter. The 

 supposed bundles of capillary tubes in the soil are irregular in area, 

 length, direction, and shape, consequently the formula is inapplicable. 

 In the second place, the foregoing research shows that the soil material 

 is dynamic and not static, and that temperature affects its physical con- 

 stitution so greatly that the velocity of flow of water through it is en- 

 tirely different from what might be expected from the known physical 

 laws. Hence, it was considered useless to attempt to apply Poiseuille's 

 formula to the foregoing experimental data. 



