Polish Archaeology in the Mediterranean (17.2005(2007))

ESHMOUN

LEBANON

ground of the second geological terrace.
About four dozen distinctive potsherds of
the Late Roman and Early Byzantine
period were discovered scattered over an
area of approximately 3500 m2. Since the
area is cultivated, there exists the risk that
the sherds were brought there with fertile
soil trucked in from elsewhere.
Other sites identified by the survey
team (nos 50, 52, 54, 57, 58, 60, 61, 67
and 73) could not be attributed safely for
lack of distinctive material in the surface
collections. The ruins of a stone mill at site
73 dated to the 19th-20th century. Most
other sites appeared to be recent (like site
60 localized in Bisri village), but it cannot
be excluded that at least a few of them
contained also ancient vestiges.
Site 25 found in 2004 was examined as
part of the rechecking program. It lies in
steep terrain north of Bhannine Village on
the eastern side of Wadi Bhannine. It is
overgrown with grass except for the
cultivated terraces. At about 8 hectares of
area it is the biggest site detected by the
expedition so far. The pottery surface
collection represents virtually all periods
from the Hellenistic through Ottoman
times with the nearby Bhannine Village
still being inhabited today.
Continued geological reconnaissance
resulted in geological and geomorpho-
logical maps being prepared for the middle
part of valley, called the Bisri Plain,
between Bisri village and Nahr El Barouk
stream. Three superposed accumulation-
erosional terraces were found. The upper
two constitute two different erosional
surfaces of the first, lowest accumulation
terrace, which is built of lacustrine silts
and represents a lake stage of valley
development. The upper part of the second

terrace is made of sands and gravels which
covered the silts. The boundary between
the two demonstrates erosional character
with sands and gravel being deposited in
high-energy conditions representing a flood
facies (distinct near the water course and
particularly in the geological sections cut
by water in the erosional deposits). The
uppermost terrace is also built of sands and
gravels, and is of alluvial origin.
The geological development of the
valley can now be reconstructed. The first
important episode in historical times was
the appearance of a lake, which built up
behind a natural dam in the form of a huge
landslide tongue. The lake filled all of the
Bisri Plain, giving rise to the lacustrine silt
deposits which accumulated on its bottom.
It went through different stages before
drying up finally sometime in the Late
Hellenistic or early Roman period. The
lacustrine silts proved to be highly fertile
agricultural soil used throughout Roman
times. A second important change in river
regime occurred in the Late Roman period.
From a stabilized river with low flows, it
changed to an annual cycle with floods.
Deluges presumably damaged or destroyed
Roman buildings, like the Bisri Temple,
for example. It was then that the river
finally cut through the natural dam,
forming a deep eroded channel. In the end
effect, the river regime and water level
were stabilized once again.
The dating of erosional episodes and
accumulation cycles depends generally on
archaeological finds. The creation of the
lake, however, and the rate of late
lacustrine sedimentation will be determined
once a thermoluminescence analysis is
carried out of samples that the team has
taken from the sediments.

436