114

A TOPOGRAPHICAL STUDY OF ERETRIA.

the western line is a fine tower of polygonal masonry, 4 m. by 6 m.,
its outer wall still being at least 4 m. high. From immediately above
the tower, the branch-wall starts down the slope to the left, at an angle
of 11°. Just beyond this wall is the first gate of the acropolis. It is
small, only 1.6 m. Avide; but the lower courses are in excellent pre-
servation ; there is thus no doubt that this was the original width. The
branch-wall appears, so far as the ruins will admit of decision, to be
of the same nature as the main acropolis-wall a b, and was probably
built at the same time. Rather more than a third of the way down
the hillside it terminates in a tower at I. After a short break, there
comes the tower II. From this point on, a diligent search failed to
lead to the discovery of any further traces of the wall, though many
stones which have fallen from the upper line are scattered over the
ground. The first thought was that this lower wall was constructed
to include springs for the citadel fortifications ; but no traces of springs
were found in the space thus added. After a study of the northeast
entrance to the acropolis, a close examination showed that the main
purpose here was probably to form a double line of defense for the
entrance to the citadel from this direction, and at the same time to add
to the area of the acropolis. The main wall from b to d is along the
summit of a precipitous declivity, the bare rock sometimes falling 10-12
metres sheer. The branch-Avail from the gate to / is also along the
edge of a steeper portion of the hillside. Directly below the tower II
are indications that a roadway, passing close below this tower and on
between I and II, was formerly supported by a retaining-wall. This
to be sure would present, to the defenders of the tower, the "shield
side " of an enemy passing along this road ; but the lay of the land did
not allow of any other arrangement. The slope, both down the hill
without and from within up to the gateway at b, is such that a road-
way here would have been quite practicable.

The main purpose for which this wall was constructed being accom-
plished at the gate-towers I and II, it is natural to expect that from
II the line should pass as quickly as possible back to the main wall.
Though there is nothing in the space between to prove or disprove
this, at d there are slight indications that the wall may have returned
straight up the steep slope to this point. It is accordingly so shown on
the MAP. The line deg passes along the northern edge of the sum-
mit. So sharp is the fall that a substructure of smaller stones, a little
outside and below the real foundations, was deemed necessary along