| |
The
Monumental Entrance to the Palace Complex, Trench TR11
1. Conservation
At
the stone glacis which delimits the eastern extent of the Palace Complex
several fallen and slipped stones were set back in their original positions
so as to enhance the appearance of this impressive monument as well as
to slow further deterioration. At the end of the season all of the small
the sandstone fragments that had been carefully stacked in discrete piles
during the course of the excavation, once it had been realised that some
elements were carved and inscribed, were each re-examined and placed in
a total of 100 sugar sacks. Geotextile was placed on the pavement, over
the sides of the fractured, plain, granite orthostats on the north side
and pinned to the upper part of the section on the south side. Then old
car tyres were placed on top of geotextile along both the front and the
sides of the trench. These tyres were filled with soil and used as a base
on which to stack the sacks of sandstone against the walling and the south
section. A tractor was then used to backfill over the sacks. This backfilling
provided an opportunity to re-examine the excavated soil in case any fragments
might have been overlooked during excavation. Neither the re-examination
of the sandstone fragments, nor the process of backfilling, led to the
recovery of further carved or inscribed pieces: a tribute to the diligence
with which the workmen had undertaken the tedious task of searching and
examining each fragment of stone during the course of the excavation.
Although much of the newly revealed stone paving has been covered, the
tops of the granite walling remain visible. It is hoped that this temporary
measure will protect the remains from the elements and will at the same
time deter interference with the south side of the trench where further
pieces of carved and inscribed stone surely await excavation in a further
season.
2.
Aims, Methods and Progress
The
original aim of the excavation of trench TR11, following on from the successful
lifting of a portion of a beam with 197 annual growth rings in 2002, was
the recovery of burnt beams for dendrochronological study by Prof. Peter
Kuniholm and his team at Cornell University. An additional aim was the
revelation of more of the plan of the gate passage as well as the recovery
of further evidence for its embellishment in the light of the discovery,
in 2002, of the lower portions of two large ibex cut from sheet bronze.
Trench TR11 extended Trench TR01 in a westerly direction towards where
it was estimated that the inner side of the gateway might have lain. The
southern edge of the trench, which was 5.00m in width, was intended to
provide a section through the centre of the passage based on surface indications
as to the position of the southern tower (Figs 49
and 50).
As
the digging progressed it was found that the burnt and collapsed rubble
filling of the passage was extremely loose and perpetually prone to collapse.
In order to prevent injury the southern and western sides of the trench
were stepped inwards as soon as it was thought judicious to do so. Before
the end of excavation the long southern section was cut back to the original
line in such a way that it was made as straight and vertical as was deemed
compatible with safety.
An added complication to the smooth progress of excavation was the character
of the southern wall of the north tower (which formed the north limit
of the trench). The lower courses of this wall comprised neatly dressed
granite orthostats the tops of which had been carefully cut to form seating
for a thin course of, presumably, levelling stones of granite (Fig. 56).
Where the wall turned to the north, in trench TR01, there were four courses
of dressed granite stones with large completely burnt out timbers between
each. With regard to the north wall of trench TR11, the granites do not
appear to have stood higher than when they were first uncovered. Here
there was a timber, more than 25cm thick, beneath the first block above
the pavement which had totally burnt away. Loose rubble had slipped into
the resultant void. This void combined with the shattered nature of the
granite orthostats and the fact that in both walls of the tower the blocks
had slipped forward from there original positions as the timbers had disintegrated,
made excavation somewhat treacherous. Eventually the corner of the tower
collapsed, thankfully without causing injury.
In addition to the difficult logistics of working in such circumstances
as just described, the recovery of fragments of carved sandstone relief
and Old Phrygian inscription, some of which were very small indeed, meant
that every fragment of sandstone need to be closely examined. This procedure
was complicated by the very large amount of sandstone from several courses
of walling that had once stood above the granite orthostats and which
had been shattered by a combination of fire damage and impact on collapse
(Figs 52,
53
and 57).
Only once the excavation had made considerable progress was it realised
that there had also been deep and extensive disturbance by treasure seekers
sometime in the distant past. This disturbance added to the propensity
of the section to collapse and had the additional result that fragments
of architectural, sculpted and inscribed stone were distributed randomly
throughout the fill. A further difficulty was the fact that the burning
of the relief, inscription and architectural ornamentation meant that
could vary from red to black and all shades between (e.g. Figs 57
and 58).
Some of the sandstone had completely vitrified, and this observation in
a few cases holds true even of carved fragments.
As
the magnitude of the problem became apparent a system was devised of separating
all sandstone form other debris and stacking it in discrete lots, while
pieces and groups of particular interest or possible significance were
taken back to the laboratory for cleaning and further study. At the end
of the season these piles of sandstone were, as described above, carefully
re-examined, bagged in sugar sacks and stacked against the sides of the
trench to provide some protection and stability over the coming winter.
3.
The Monumental Entrance
At this preliminary stage in the excavation there appear to be a minimum
of four construction phases at this eastern end of the Palace Complex:
1, Structure A together with its stone glacis and part of the stone pavement;
2, the "Audience Hall" with yet more of the stone pavement which
is aligned with it; 3, the Monumental Entrance itself; 4, the terraces
of Structure B.
With regard to the Monumental Entance, the south-eastern corner of the
north tower has been exposed along with the surviving stretch of granite
orthostats that form the north side of the gateway. The east wall of the
tower, first revealed in trench TR01 in 2002 has four preserved courses
of neatly trimmed granite interspersed with burnt-out timber beams (Fig.
57).
The south wall comprised one course of granite orthostats beneath a levelling
course of thinner stones which were tightly fitted into carefully cut
seating (Figs 53
and 56).
In some instances the back, hidden, side of the orthostats was as well
dressed as the front. The orthostats rested on a very large timber beam
beneath which is yet another course of large cut granite which, for reasons
of safety, has only been glimpsed at. At the south-east corner of the
tower the large lower block, which is level with the top of the pavement,
is slightly offset. These granite blocks appeared to have carried another
substantial wooden beam and a minimum of five large (up to 1.50 by 0.50
by 0.50m) sandstone blocks. At the south-east corner the walling would,
therefore, have originally attained a height of some 5.00m.
The sandstone contains dovetail clamp holes for wooden clamps (Fig. 55).
Some blocks are known to posses clamps at either end, as discussed earlier
in this report. Several iron bars with nails in position at each of the
splayed ends appear to be too large for the clamp holes that have been
recorded up to now and it seems probable that these iron clamps were used
to strengthen or secure architectural timber elements (Figs 55,
63
and 64).
It is noteworthy that clamping was restricted to the sandstone, the granite
presumably being too hard and difficult for the masons to cut clamp holes.
Some of the sandstone blocks were embellished by carved three-quarter
bolsters decorated with concentric compass drawn circles on the ends (Figs
59
- 60).
Although no complete example of these large bolsters has been found, they
were presumably symmetrical and double ended. In other instances the shallow
relief carving was restricted to the ends (Fig. 58)
or with just a part of a cylindrical barrel which stood slightly proud
(Fig. 61).
All of these blocks are badly smashed and, in some cases, partially vitrified,
with the result that no complete example has yet been assembled from the
many shattered fragments. Although these blocks would appear to have formed
a row of architectural decoration at the top of the south-eastern corner
of the north tower, it is not impossible that some of these carved elements
could have come from sandstone capitals to wooden columns. It might be
speculated that the exposed ends of the horizontal wall beams at the corner
of the tower were carved with similar spool-shaped terminals. Some of
the smooth faces of the sandstone blocks retain splashes of paint which
might possibly have derived from the painting of the timber elements.
The burnt debris filling the gateway contained many fragments of burnt
mud containing reed and, sometimes, plank impressions. These had evidently
fallen from the roofs of the towers during the fire.
Approximately
one third of the way inside the gate and set against the passage wall
was a large, square, sandstone block the top of which contains a cut circular
depression almost 1.00m in diameter and some 3cm deep that formed the
seating for a large wooden pillar (Fig. 51).
It is not clear whether this pillar or column, which would have made a
dramatic visual impact, carried a capital of some sort, nor whether it
in fact supported a double-pitched roof covered with thatch. If it did
indeed carry such a roof it is more than likely that the iron bars and
bands together with the huge nails that affixed them to very substantial
wooden elements, were associated with the roof structure and the burning
of the timber frame would explain the intensity of the fire, which was
sufficient to melt both sandstone and granite, in the entrance. Further,
it might be imagined that the bronze cut-outs of an antithetical pair
of ibex that were recovered from the surface of the pavement in trench
TR01 during the 2002 season would once have been nailed to the pediment,
probably on either side of a representation of sacred tree affixed to
the king post. This pillar does not seem to have been associated with
doors for two reasons. Firstly, there is no door socket and secondly the
inclination of the pavement would seem to preclude the presence of doors
in this position.
As to the stone pavement, the surface of which had been worn to a polish,
it is quite steeply inclined until it approaches its original upper limit,
the position of which is indicated by the westernmost two stones at the
south-east corner of the exposed portion (Fig. 54).
Here, at this upper limit, the pavement appears to be more level, an observation
which is compatible with the idea that the partially excavated burnt rubble
beyond the stone paving represents a wooden threshold associated with
doors. It would be most reasonable to reconstruct doors in this position,
towards the inner end of the monumental entrance, and indeed it is possible
that the robbing in the north-west corner of the trench was a result of
looters looking for a metal door socket or other valuable embellishments.
Further excavation might resolve some of these issues if the extensive
robbing has not obliterated all of the pertinent evidence.
With regard to stratigraphy, the pavement contains several setting lines
(barely visible in Fig. 50)
which, whatever they represent, were clearly intentional. The two most
prominent of these lines, while not quite parallel with one another, follow
an alignment that is at about 30 degrees to the line of tower wall and
are very close to the alignment of the central axis of the "Audience
Hall". The evidence currently at hand does not permit clear and certain
division of the paving into phases of stratigraphic significance, although
it is more than probable that not all of the paving can have been part
of one and the same original scheme.
The stratigraphic relationship between the construction of the newly revealed
Monumental Entrance to the Palace Complex and the building of massive
stone towers of Structure A together with their supporting stone glacis
has not yet been certainly ascertained. It is however noteworthy that
Structure A has a strongly defensive character, whereas the monumental
entrance represents a display of grandeur and opulence. If materials are
any guide, the absence of sandstone in Structure A might be of chronological
significance. Whether the construction of the monumental entrance is associated
with the cutting through of the glacis or represents some intermediate
scheme is uncertain.
Following the destruction of the entrance by fire, during the course of
which, it has been postulated, the pitched timber and thatch roof together
with substantial parts of the towers to either side fell onto the pavement.
Hopes of recovering charred beams for dendrochronology were dashed because
the intensity of the burning had reduced everything to ash except by the
south-eastern corner of the north tower where some burning timber appears
to have been smothered by fallen stone blocks.
4. Inscribed
and Sculpted Monuments in the Entranceway
Somewhere
inside the entrance, probably towards the centre of the rear, was a sandstone
monument embellished with three quarter round bolsters and bolster ends
in relief. These bolsters closely resemble, on a somewhat smaller scale,
the carved elements from, perhaps, the corner of the tower. Joining fragments
make up about two thirds of the decorated architectural element of this
monument, which although the recess on the top is rectangular seems to
have been square in plan (Fig. 68).
This monument would appear to have been free-standing. Also recovered
were fragments of small-scale relief sculpture amongst which have been
identified figures in human form, a griffin and a lion - both apparently
winged, a rosette with eight petals and other fragments. There are also
vertical and horizontal elements inscribed in Old Phrygian. Prof. Claude
Brixhe reports the presence of the word for a dedication. It is very likely
that the inscribed relief sculpture and the architectural element belong
to the same monument, although this is not certain. There are, in addition,
other fragments of inscribed sandstone and a large but fragmentary piece
which resembles a stepped platform or table with a recessed top. The fact
that joining fragments, of dramatically different fire-altered colour
and sometimes only millimetres in length, were found scattered both horizontally
and vertically throughout the undisturbed burnt debris (as well as in
later disturbances) can safely be taken to indicated that the monument
was destroyed during the conflagration. Further fragments doubtless lie
beyond the excavated area, but a few of the pieces which were retrieved
have partially vitrified - with the result that some unknown portion of
the relief and inscriptions has been totally lost. These discoveries are
of considerable importance and will be the focus of much future research.
At a later, unknown, date there was very considerable disturbance to the
deposits and the structure, presumably at the hands of treasure seekers.
All of the masonry is in such poor condition that the large pits can hardly
have been dug by stone-robbers. Inscribed and relief fragments were recovered
from the fill of the robber pits as well as from the undisturbed burnt
debris, adding another element to the likely-hood of recovering further
pieces if a future season of excavation.
5. The
Function of the Monumental Gateway
This
monumental gateway led directly to the huge "Audience Hall"
which was the subject of partial excavation in previous campaigns. The
positioning of this imposing and substantial building within one discrete
area of a huge complex could be seen as supporting the proposed identification
of the building as an "Audience Hall" (rather than a temple)
located at the most public part of what appears most likely to have been
a large Palace Complex.
Work at the Ashlar building in 2003 was designed to complete the excavation
of the southern half of the larger, inner room which had been partially
investigated in 2002 (Fig. 73a).
The room was found to have been some 3.00m longer than interpretation
of the geomagnetic image had suggested (Fig. 73b),
although the position of the outer face of the rear wall of the building
could be distinctly seen on the ground.
Contrary to expectations, no internal features or installations of any
kind were discovered (Fig. 74).
It is, however, very significant that a straight and level line of burning
was seen along all of the wall faces. This line was just above the rough,
unsmoothed, lower portion of the granite blocks except in the south-west
corner where, as can be seen on the left side of Figure 75,
it was some 0.10m higher. This scorch-line surely represents a burnt wooden
surround which would thus have formed the equivalent of the sandstone
pavement in the front room that was revealed in 2002.
The existence of such a surround, with a slightly raised portion in the
south-west corner, provides explanations for the rough finish to the otherwise
smooth face of the ashlars as well as for the fact that the multi-layered
mud-plaster floor did not run up against the walls. The recognition of
this surround which, it may be imagined, was covered with cushions and
fine materials, combined with the regular re-plastering of the floor has
made a not inconsiderable difference to our conceptions of this building.
Now this larger inner room can perhaps be envisaged as having been more
comfortable and plush than the outer room with its cold stone paving.
This impression finds some confirmation from the observation that most
or all of the thick mud plaster that covered the rubble filled walling
above the faced granites was coated with white lime plaster.
There was surprisingly little burnt clay with reed impressions of the
kind generally associated with burnt roofing, suggesting perhaps that
entire the building was covered with a double pitched thatched roof and
that the clay and reeds represent an upper floor or balcony in the outer
room where (in 2002) they were found to be more concentrated.
With regard to the ashlars, no. 14 (Fig. 76),
which was not fully uncovered in 2002, was found to have a total of six
strokes while no. 19 also bore marks, perhaps in two groups. As to finds,
part of one charred wooden terminal that appears to be a miniature version
of the bolsters on the sandstone at the Monumental Entrance to the Palace
Complex perhaps represents all that has survived from sumptuous furniture.
Otherwise there were no objects other than a few iron nails and a poorly
preserved iron bracket
|
|
