id, and it is
still further charged with it in percolating through the surface soil in
which there is decomposing vegetable matter. As the raindrops converge
towards some one point, determined by some local accident on the
surface, and always in a line of joint, the carbonic acid attacks the
carbonate of lime with which it comes into contact, and thus a funnel is
gradually formed ending in the vertical joint below. Both funnel and
vertical joint below are being continually enlarged by this process.
This chemical action goes on until the free carbonic acid is used up.
The subterranean passages are enlarged in this manner, and what was
originally an insignificant network of fissures is developed into a
series of halls, sometimes as much as from 80 to 100 ft. high. These
results are considerably furthered by the mechanical friction of the
pebbles and sand hurried along by the current, and by falls of rock from
the roof produced by the removal of the underlying strata. In many cases
the results of this action have produced a regular subterranean river
system. The thick limestones of Kentucky, for example, are traversed by
subterranean waters which collect in large rivers, and ultimately appear
at the surface in full power. The river Axe, near Wells, the stream
flowing out of the Peak Cavern at Castleton, Derbyshire, that at
Adelsberg in Carniola, flow out of caverns in full volume. The river
Styx and the waters of Acheron disappear in a series of caverns which
were supposed to lead down to the infernal regions.

If the direction of the drainage in the rock has been altered, either by
elevations such as those with which the geologist is familiar, or by the
opening out of new passages at a lower level, these watercourses become
dry, and present us with the caves which have afforded shelter to man
and the wild animals from the remotest ages, sometimes high up on the
side of a ravine, at other times close to the level of the stream at the
bottom.

Caves, as a general rule, are as little effected by disturbances of the
rock as the ravines and valleys, which have been formed, in the main,
irrespective of the lines of fault or dislocation.

We must now examine what happens to the bicarbonate of lime which has
been formed by the action of the acid on the limestone. If a current of
air play upon the surface of the water, the carbonic acid, which floats
up the lime, so to speak, is given off and the insoluble carbonate is
deposite