OBSERVATIONS ON DETERMINATIONS OF DROUGHT-INTENSITY. 77 



V of course must be considered a variable, and therefore also /, 

 in practical cases : k depends only on the class of soil. 



11. Percolation tvhen interstices are not filled. — The supposition 

 that the interstices are filled with water can be made only when 

 ground is completely saturated, a relatively rare condition. When 

 the interstices are not filled the velocity is much less than would 

 be given by expressions like (7), viz. V = kf'I. In flow of this 

 character, what is known as the 'hydraulic radius' of the flow, is 

 diminished, and therefore the resistance to flow is increased. For 

 as the quantity diminishes indefinitely, so the resistance similarly 

 increases, and thus the dryer the soil, the slower the rate of motion 

 of the percolating liquid. The rate at which water will flow 

 downward into relatively dry soil, especially when the quantity 

 is small, is much less than its rate when the soil is saturated. 



Again, when there is any discontinuity in the liquid network, 

 the phenomena of surface tension come into play, and finally the 

 soil may contain water which is entirely prevented from moving 

 downward under the action of gravity, by the relatively greater 

 effect of molecular forces — that is those forces which come into 

 play in surface-tension. 



The effect of temperature, decidedly sensible as regards viscosity, 



is relatively smaller in this respect. Since if T denote the surface 



tension at 6 degrees Celsius, and T that at the zero of the same 



scale, then 



T=T (1-002(9) (9) 



that is to say the effect here is considerably less than -^ of the 

 viscosity effect. The influence however is the same in kind ; 

 increase of temperature promotes flow, by diminishing both 

 viscosity and surface tension. It is extremely doubtful whether 

 'partial interstitial flow' can, like 'full interstitial flow,' be made 

 satisfactorily subject to rational mathematical deduction in cases 

 such as are here considered ; it will probably have to be experi- 

 mentally determined, and expressed by purely empirical formulae. 



1-2. Exhaustion of moisture in soil by percolation. — Neglecting 

 the by no means inconsiderable effect of vegetation, soil containing 



