GROUND WATER IN NORTH-CENTRAL TENNESSEE

76 GROUND WATER IN NORTH-CENTRAL TENNESSEE
pp. 82-83) and thus may constitute one or more bodies of perched
ground water. Under such conditions the static level of the ground
water in wells may not define a simple surface, so that it is difficult or
impossible to recognize the regional water table. These conditions
led Martel 47 to deny the existence of a water table in limestone
regions.
Practically all the surface water that is diverted by the sink holes
is discharged into some surface stream at points of lower altitude
through tubular or fracture springs (pp. 92-95). By accretion of
water gathered by many sinks, perennial underground streams may
be formed and may flow for miles through solution passages before
reappearing at the surface. Also, a perennial surface stream may
disappear wholly or in part into a sink or solution channel, flow
underground for a distance, and reappear at the surface one or
more times. These phenomena have been summarized by Meinzer **
as follows:
Some of the large caverns contain streams that do not differ greatly from sur­
face streams. Like the surface streams, they generally flow about at the level
of the water table. As the result of a heavy rain or the rapid thawing of snow
they may receive great quantities of surface water directly through sink holes
and may, like the surface streams, become greatly enlarged. In times of flood
they may lose some of their water by percolation into the smaller crevices, but
at low stages they are fed by the body of ground water that occupies all crevices
below the water table. Barely there is a passage through the limestone that
leads from a sink hole, where the water is taken in, to a point of discharge at a
lower level, without extending down to the water table. Such a subterranean
passage is essentially a dry wash that is arched over by a natural bridge. Its
discharge is extremely irregular and ceases as soon as the flood water has disap­
peared.
Underground drainage systems of this sort are characteristic of
pure thick-bedded limestones. Davis 4fl states that the underground
channels which constitute these systems differ from surface drains in
that they are looped and do not have well-defined gradients. Mar­
tel 60 contends that the underground streams are similar to surface
rivers in that they comprise rapids, cascades, and static reaches, are
dendritic, and have floods. He contends further that the under­
ground streams differ from surface rivers in that they include very
abrupt and extreme constrictions, siphons both upright and inverted,,
abrupt falls, and pools of large volume; also in that they are looped
and in many places are partly or completely dammed by rock that
has fallen from roof or walls. Some of these differences seem not to-
be permanent characteristics, however, if the underground streama
« Martel, E. A., Nouveau traits des eaux souterraines, pp. 305-306, Paris, 1921.
« Meinzer, O. E., The occurrence of ground water in the United States: TJ. S. Geol. Survey Water-Supply-
Paper 489, p. 134, 1923.
« Davis, W. M., written communication to O. E. Meinzer, Mar. 25,1930.
M Martel, E. A., op. cit., pp. 228, 242-246.