210 DE rance: underground avater-supply and river floods. 
consecutive months there was no percolation whatever ; five times 
there was no percolation for six months ; and only in one year (1860) 
was there percolation every month. The greatest percolation is 
after snow thaws, especially after frequent thaws of small falls. A 
wet winter gives abundant springs in the following antumn, but if 
that be followed by a dry winter it will obliterate the effects of the 
previous wet winter. Mr. Greaves is of opinion that a flat soil will 
never produce a flood if it be effectually nnder-drained, as it then 
resembles the percolation gauge ; such under-draining would diminish 
floods, and the water would be run off gradually and improve 
the dry-weather discharge. Mr. Greaves considers a percolation 
gauge should be 36 inches deep, and at that depth that water is 
safe from the surface loss. He suggests that in the different degrees 
of capillarity of different soils lies the cause of their variations in 
healthiness, the high capillary power of a clay soil producing a con- 
stant summer exhalation. 
Underground Water Siqjpli/ : — Pervious or permeable forma- 
tions, by gradually absorbing waters falling on the surface, which slowly 
percolate through them, act at once as filter-beds and as reservoirs, the 
capacity of which is limited by the area of absorption, and the thickness 
of the pervious bed. When rain falls upon a pervious rock overlying 
impermeable deposits, the water-line is generally near the surface, and 
forms the "plane of saturation" which is found to be slightly above that 
of the deepest valley intersecting the water-bearing rock, rising towards 
the centre of the hill, varying within certain limits, being governed by 
the amount of previous rainfall. When wells are sunk into it, and 
excessive pumping takes place, the plane of saturation is artificially 
and locally lowered, and is known as the pumping level, after a few 
hours cessation of pumping, the water rises to its original level, or 
nearly so, the point so reached is the "rest-level." The difference 
between the " rest-level " and the pumping level in some wells is as 
much as 100 feet. The area of exhaustion resembles an inverted 
cone, the apex of which rests on the point at which the pumps 
abstract the water, and the base of which is a circle at the surface 
around the well. If over pumping takes place, the "cone of 
