nd not to be
depended upon. These, too, are extremely sweet. In coarse grained gravel
and carbuncular sand the supply is surer and more lasting, and it has a
good taste. In red tufa it is copious and good, if it does not run down
through the fissures and escape. At the foot of mountains and in lava it
is more plentiful and abundant, and here it is also colder and more
wholesome. In flat countries the springs are salt, heavy-bodied, tepid,
and ill-flavoured, excepting those which run underground from mountains,
and burst forth in the middle of a plain, where, if protected by the
shade of trees, their taste is equal to that of mountain springs.

3. In the kinds of soil described above, signs will be found growing,
such as slender rushes, wild willows, alders, agnus castus trees, reeds,
ivy, and other plants of the same sort that cannot spring up of
themselves without moisture. But they are also accustomed to grow in
depressions which, being lower than the rest of the country, receive
water from the rains and the surrounding fields during the winter, and
keep it for a comparatively long time on account of their holding power.
These must not be trusted, but the search must be made in districts and
soils, yet not in depressions, where those signs are found growing not
from seed, but springing up naturally of themselves.

4. If the indications mentioned appear in such places, the following
test should be applied. Dig out a place not less than three feet square
and five feet deep, and put into it about sunset a bronze or leaden bowl
or basin, whichever is at hand. Smear the inside with oil, lay it upside
down, and cover the top of the excavation with reeds or green boughs,
throwing earth upon them. Next day uncover it, and if there are drops
and drippings in the vessel, the place will contain water.

5. Again, if a vessel made of unbaked clay be put in the hole, and
covered in the same way, it will be wet when uncovered, and already
beginning to go to pieces from dampness, if the place contains water. If
a fleece of wool is placed in the excavation, and water can be wrung out
of it on the following day, it will show that the place has a supply.
Further, if a lamp be trimmed, filled with oil, lighted, and put in that
place and covered up, and if on the next day it is not burnt out, but
still contains some remains of oil and wick, and is itself found to be
damp, it will indicate that the place contains water; for all heat
attr