Vol. 6^.~] THE MOTION OF 8UB-SUEFACE WATER. 101 



the same period, expressed as inches of rainfall over the whole area ; 

 and the evaporation is that determined by the late Mr. G. J. Symons, 

 at Camden Square, London, during a period of 14 years. I have taken 

 these latter figures, because it is somewhat difficult to estimate 

 the exact evaporation over so wide an area as that of the Thames 

 Basin : the normal evaporation being largely increased by vegetable 

 growth, and by the action of capillarity in lifting towards the surface 

 t he water which would otherwise percolate to deeper zones. It is 

 interesting to note from the above that, out of an average annual 

 rainfall of 26 inches over the entire Thames basin, only 13 per cent., 

 or 3*4 inches of rain, is available for percolation. Now, although I 

 do not accept all the numerous statements of alleged permanent 

 depletion of the sub-surface water-supply, of which one hears from 

 time to time, and which may have no other foundation than an ill- 

 considered comparison of a single maximum water-level succeeding 

 a period of excessive rainfall, with a single minimum water-level 

 succeeding a period of deficient rainfall : still, it may happen that, 

 with so small an average percolation as the table shows, the 

 pumping within the area, together with the normal discharge from 

 the springs, may so closely approximate to, or even exceed, the 

 percolation, as to preclude any possibility of the recuperation of the 

 sub-surface supplies, especially during a period of long-continued 

 drought. It also follows, from a consideration of the hypothetical 

 case, that hill-masses will have a considerable influence in distorting 

 the contours ; this is well shown in the accompanying hydrological 

 map of Dorset, Wiltshire, and Hampshire (PI. V), in which both the 

 surface and underground water-contours are shown, both being 

 referred to Ordnance-datum. 



Reverting again to a consideration of hypothetical conditions, 

 we see that, if an uniformly- porous stratum with an evenly- 

 sloping surface rests conformably upon an impervious bed, which 

 crops out in a straight line and slopes uniformly with a greater or 

 less inclination in the same direction as the surface, then the 

 contour-lines of the sub-surface water will be a series 

 of lines parallel to the line of the outcrop of the im- 

 pervious bed, which will approach to or recede from 

 the line of outcrop, according as the supply of perco- 

 lating water is in excess or in defect. If, however, 

 instead of an uniform surface it be undulatory or uneven, then, 

 instead of a series of more or less parallel lines, the contours will 

 be a series of similar curves, with major axes more or less parallel 

 to the line of the outcrop, the case being a combination of the two 

 previous hypotheses ; and these assumptions are verified by observa- 

 tion in the field, as shown in figs. 2 & 3 (p. 102), not only in the 

 Chalk of Hertfordshire, but also in that of Wiltshire. 



In the first-named county, the main axes of the high-level water- 

 contours, in the area embracing the villages of Weston, Rushden, 

 Sandon, Barkway, Buckland, and Cottered, are roughly parallel to 

 the line of outcrop of the Gault-Clay, which ranges by Tilsworth, 



