chist in the Grecian island of Thermia (Cythnos of the
ancients), one of the Cyclades. Here also spacious halls, with rounded
and irregular walls, are connected together by narrow passages or
tunnels, and there are many lateral branches which have no outlet. A
current of water has evidently at some period flowed through the whole,
and left a muddy deposit of bluish clay upon the floor; but the erosive
action of the stream cannot be supposed to have given rise to the
excavations in the first instance. M. Virlet suggests that fissures were
first caused by earthquakes, and that these fissures became the chimneys
or vents for the disengagement of gas, generated below by volcanic heat.
Gases, he observes, such as the muriatic, sulphuric, fluoric, and
others, might, if raised to a high temperature, alter and decompose the
rocks which they traverse. There are signs of the former action of such
vapors in rents of the micaceous schist of Thermia, and thermal springs
now issue from the grottoes of that island. We may suppose that
afterwards the elements of the decomposed rocks were gradually removed
in a state of solution by mineral waters; a theory which, according to
M. Virlet, is confirmed by the effect of heated gases which escape from
rents in the isthmus of Corinth, and which have greatly altered and
corroded the hard siliceous and jaspideous rocks.[1042]

When we reflect on the quantity of carbonate of lime annually poured out
by mineral waters, we are prepared to admit that large cavities must, in
the course of ages, be formed at considerable depths below the surface
in calcareous rocks.[1043] These rocks, it will be remembered, are at
once more soluble, more permeable, and more fragile, than any others, at
least all the compact varieties are very easily broken by the movements
of earthquakes, which would produce only flexures in argillaceous
strata. Fissures once formed in limestone are not liable, as in many
other formations, to become closed up by impervious clayey matter, and
hence a stream of acidulous water might for ages obtain a free and
unobstructed passage.[1044]

_Morea._--Nothing is more common in limestone districts than the
engulfment of rivers, which after holding a subterranean course for many
miles escape again by some new outlet. As they are usually charged with
fine sediment, and often with sand and pebbles where they enter, whereas
they are usually pure and limpid where they flow out again, they must