EXCAVATIONS ON SITE OF MAUSOLEUM. 133
their parallelism with the lines of the Quadrangle, as
well as from the goodness of the masonry, seems
highly probable ; then they could hardly have been
intended to be concealed under the soil, but must
have formed a feature in the site as terrace walls.
In that case, it cannot be supposed that the south
Avail and the southern extremity of the eastern wall,
would have died away, as they do at present, into
the higher level of the rock on each side, because
such broken lines of revetement would have been
most unsightly, and it is difficult to see what pur-
pose they could have served. There is, then, but
one Avay of reconciling all these difficulties. We
must suppose that the margin of native rock
between the south and east sides of the quadrangle
and the respective revetement walls, was covered
with soil, so as to obtain a perfect level.
It is to be observed that this marginal surface
of native rock, though generally uniform in level,
presents many inequalities which it may have been
thought better to conceal by forming an artificial
terrace. The depth of soil thus laid on did not
probably exceed two feet at the edges of the terraces,
because it is evident, from the slightness of the
masonry, that the revetement walls would not have
borne much lateral pressure.
It may be remarked.that this change of level in
the peribolus would occur on the two sides on which
the platform would be approached from lower
ground, and was probably intended to set off the
elevation to greater advantage.
There would not, therefore, be the same necessity
their parallelism with the lines of the Quadrangle, as
well as from the goodness of the masonry, seems
highly probable ; then they could hardly have been
intended to be concealed under the soil, but must
have formed a feature in the site as terrace walls.
In that case, it cannot be supposed that the south
Avail and the southern extremity of the eastern wall,
would have died away, as they do at present, into
the higher level of the rock on each side, because
such broken lines of revetement would have been
most unsightly, and it is difficult to see what pur-
pose they could have served. There is, then, but
one Avay of reconciling all these difficulties. We
must suppose that the margin of native rock
between the south and east sides of the quadrangle
and the respective revetement walls, was covered
with soil, so as to obtain a perfect level.
It is to be observed that this marginal surface
of native rock, though generally uniform in level,
presents many inequalities which it may have been
thought better to conceal by forming an artificial
terrace. The depth of soil thus laid on did not
probably exceed two feet at the edges of the terraces,
because it is evident, from the slightness of the
masonry, that the revetement walls would not have
borne much lateral pressure.
It may be remarked.that this change of level in
the peribolus would occur on the two sides on which
the platform would be approached from lower
ground, and was probably intended to set off the
elevation to greater advantage.
There would not, therefore, be the same necessity