the few highest courses had disappeared. This facing we followed for some
time, despite the increasing dampness of the sand, until a heavy fall of the
sides made a further descent dangerous and even impossible.
Eeturning to the rectangular pit, it will be seen that it is not of an even
depth all over, but that within the rectangle there is a deeper circular
excavation (marked 2nd level of rock) concentric with the well. This
excavation opens out to the S.E. of the well, where the sloping approach
joins it in such a way that the two viewed together give the form of a
tennis-racket. The eastern portion of the rectangle we did not excavate,
but it is clear from the surface indications that it "was similar in form to
the symmetrically opposite portion.
We can now reconstruct the appearance of the whole in ancient times.
The well was set in a deep rectangular pit, to which those who wished to
draw water gained access by the sloping approach. Whether a wall
surrounded the whole excavation it is now impossible to decide. Bound
three sides of the rectangle it is still preserved and a small length still stands
to the north-east of shaft /3. But the pieces of wall shown near to the
entrance of the slope are not raised on the rock as are those which surround
the rectangle, but built up as facing to support weak spots in the rock
itself. It is quite possible that originally the now blunted angles at the
point where the sloping passage enters the rectangle were right angles, but
there is no evidence on the point.
We have still to mention the shafts a, /S, and k. It seemed probable in
the early stages of excavation that ft at least was contemporary with and
formed part of the whole construction. But this was disproved when, after
partial clearance, we were able to perceive that its orientation was dis-
tinctly different to that of the rectangle. Confirmation of this was given
by shaft a, half of whose wall had been cut clean away by the digging out
of the sloping passage. Both these shafts belonged to the late XVIIIth
and the XlXth Dynasties. Shaft a had been badly plundered. Its
chambers were four in number opening off the S.E. end of the shaft, but
successive falls of rock had reduced them practically to a single rough cave.
Shaft yS with three chambers, also to the S.E., had suffered in a similar
manner, though it was less completely plundered. The brick shaft itself was
still covered with a solid superstructure rising in steps to the S.E. (PI. III.
fig. 8), but the thieves had entered the chambers through the roof without
troubling to descend in the shaft at all. The falls of rock had badly
broken up the wooden coffins, which were numerous. The most interesting
of the objects found were several complete sets of ushabti of the XlXth
Dynasty. Grave k contained three plundered burials, and had been